/irc-logs / mozilla / #developers / 2012-01-10 / end

Options:

# Session Start: Tue Jan 10 00:00:02 2012
# Session Ident: #developers
# [00:00] <bwinton> GRMrGecko: I'll appreciate whatever you can do. At worst, it'll make it easier for the next person to come along and add more stuff. :)
# [00:00] * Quits: rwaldron (rwaldron@moz-BDCCF091.hfc.comcastbusiness.net) (Quit: Leaving...)
# [00:00] <edmorley> has anyone created a script to parse multiple tinderbox build logs and determine which directories are already warning free on os x and linux so could have FAIL_ON_WARNINGS added?
# [00:00] <edmorley> (to save duplicating on effort)
# [00:02] <sicking> dolske: ping
# [00:02] * Joins: darktrojan (geoff@moz-57EC95C5.dsl.telstraclear.net)
# [00:03] * Joins: anant_ (anant@moz-C03D0C61.vlan426.asr1.sfo1.gblx.net)
# [00:04] * IRCMonkey42078 is now known as Tobbi
# [00:04] * Quits: anant (anant@moz-C03D0C61.vlan426.asr1.sfo1.gblx.net) (Ping timeout)
# [00:04] * anant_ is now known as anant
# [00:04] * bwinton is now known as bwinton_away
# [00:06] * Quits: jorendorff (jorendorff@moz-91590D94.hsd1.tn.comcast.net) (Quit: jorendorff)
# [00:07] * Joins: espindol_ (espindola@EFCDA10D.68E64C0A.F061A1E6.IP)
# [00:07] * Quits: espindola (espindola@EFCDA10D.68E64C0A.F061A1E6.IP) (Quit: Ex-Chat)
# [00:08] <dolske> sicking: hi!
# [00:08] * Quits: espindol_ (espindola@EFCDA10D.68E64C0A.F061A1E6.IP) (Quit: )
# [00:08] * jimm is now known as jimm-mmiab
# [00:08] * jimm-mmiab is now known as jimm-bbiab
# [00:08] <sicking> dolske: what renders the UI for the new tab-modal alerts? Do they live as part of the <browser> binding?
# [00:09] * Joins: jorendorff (jorendorff@moz-91590D94.hsd1.tn.comcast.net)
# [00:11] * Quits: adoria0000 (Mibbit@moz-B06B7389.osaka.ocn.ne.jp) (Quit: http://www.mibbit.com ajax IRC Client)
# [00:11] <dolske> sicking: there's a <stack> wrapping each tab, some glue in browser.js to handle adding/removing things from the stack, and tabprompts.xml
# [00:12] <dolske> s/browser.js/tabbrowser.xml
# [00:13] <dolske> basically the prompting code reaches up, calls getTabModalPromptBox(), and sticks a <tabprompt> into it.
# [00:13] * Quits: jet (junglecode@moz-C03D0C61.vlan426.asr1.sfo1.gblx.net) (Quit: jet)
# [00:13] <dolske> http://mxr.mozilla.org/mozilla-central/source/browser/base/content/tabbrowser.xml#346 is probably the interesting part
# [00:13] * Quits: logbot (logbot@moz-622AFC27.glob.com.au) (Client exited)
# [00:15] <sicking> dolske: so does the <stack> live in a <browser> element, or does <tabbrowser> not use <browser>s at all?
# [00:16] * Joins: sheppy (sheppy@moz-D0C71163.domainhotel-wifi.mozilla.hq)
# [00:16] <sicking> dolske: the way I had envisioned it is that <tabbrowser> contains a bunch of <tab>s which each contains a <browser> which has a <stack> with a <iframe> and some markup to do the alert
# [00:16] <gavin> the stack isn't in the browser
# [00:17] * Joins: logbot (logbot@moz-622AFC27.glob.com.au)
# [00:17] * Quits: jprmc (jprmc@F2D29657.F60B0462.67AC9B1.IP) (Ping timeout)
# [00:18] <gavin> sicking: http://mxr.mozilla.org/mozilla-central/source/browser/base/content/tabbrowser.xml#74
# [00:18] * Quits: sheppy (sheppy@moz-D0C71163.domainhotel-wifi.mozilla.hq) (Quit: sheppy)
# [00:18] <gavin> and it's just xul, not an iframe
# [00:18] * Quits: ericb2 (X@moz-9C4C3DED.fbx.proxad.net) (Quit: Success !!)
# [00:18] * Joins: jet (junglecode@moz-C03D0C61.vlan426.asr1.sfo1.gblx.net)
# [00:20] * Joins: sheppy (sheppy@moz-D0C71163.domainhotel-wifi.mozilla.hq)
# [00:20] <sicking> gavin: aren't xul <browser>s ultimately <iframe>s?
# [00:20] <sicking> gavin: inside the binding for the <browser> that is
# [00:20] <gavin> sicking: well sure, I guess - with a bunch of xbl goop
# [00:21] * Quits: sheppy (sheppy@moz-D0C71163.domainhotel-wifi.mozilla.hq) (Quit: sheppy)
# [00:21] <gavin> but the prompts are xul inserted into the xul:stack
# [00:21] <sicking> right
# [00:21] <sicking> gavin, dolske: thanks guys!
# [00:22] * Joins: njn (chatzilla@moz-6689C0E4.dyn.iinet.net.au)
# [00:22] * Quits: Mnyromyr (Mnyromyr@B2521176.7B0892CB.771966F7.IP) (Quit: ChatZilla 0.9.86 [SeaMonkey 1.1.19/2010030105])
# [00:22] * Joins: Mossop (mossop@moz-BBE3ABD.mv.mozilla.com)
# [00:22] * Quits: rs (rs@moz-C03D0C61.vlan426.asr1.sfo1.gblx.net) (Quit: ChatZilla 0.9.86.1-rdmsoft [XULRunner 1.8.0.9/2006120508])
# [00:22] * Quits: josh (josh@4E277166.F161DD3.72ECBEEF.IP) (Quit: josh)
# [00:23] * Quits: TheLink (TheLink@moz-B5AB52CB.pools.arcor-ip.net) (Quit: I has album artz!!!1)
# [00:23] * zpao|detached is now known as zpao
# [00:23] * Joins: sfleiter (stfl@moz-22F37930.hsi7.kabel-badenwuerttemberg.de)
# [00:24] * nthomas is now known as nthomas|away
# [00:25] * Quits: mdas (mdas@8CB764BC.D30B51A1.412CF160.IP) (Quit: mdas)
# [00:26] * Quits: clee (clee@moz-BBE3ABD.mv.mozilla.com) (Quit: clee)
# [00:26] * Joins: clee (clee@moz-BBE3ABD.mv.mozilla.com)
# [00:26] * Quits: jdm (jdm@moz-15BB5FE6.cable.teksavvy.com) (Client exited)
# [00:27] <Waldo> dholbert: /* ... */ should do /* / * comment body.... / */ (ugh is that unreadable) only if the comment body spans multiple lines (or if it spans one, but adding */ at the end would make it overflow line length limits)
# [00:27] <Waldo> dholbert: so what you did is right
# [00:27] <dholbert> Waldo, ok, thanks
# [00:27] * Quits: nattokirai (nattokirai@moz-6A258170.dynamic.ppp.asahi-net.or.jp) (Quit: nattokirai)
# [00:27] * Joins: ted (luser@moz-8B3C93E1.hsd1.pa.comcast.net)
# [00:31] * Quits: ajuma (ajuma@F2D29657.F60B0462.67AC9B1.IP) (Quit: ajuma)
# [00:32] * Joins: jgilbert_ (jgilbert@moz-BBE3ABD.mv.mozilla.com)
# [00:33] * Quits: jgilbert (jgilbert@moz-BBE3ABD.mv.mozilla.com) (Ping timeout)
# [00:34] * Quits: madhava (madhava@F2D29657.F60B0462.67AC9B1.IP) (Quit: madhava)
# [00:34] <ted> i assume this bug where we have extra spacing between lines got fixed and my nightly is just behind?
# [00:36] * Quits: jrmuizel (jrmuizel@F2D29657.F60B0462.67AC9B1.IP) (Input/output error)
# [00:37] <bholley> is there a better way to do what I'm trying to do here? http://pastebin.mozilla.org/1441979
# [00:38] * Joins: jrmuizel (jrmuizel@F2D29657.F60B0462.67AC9B1.IP)
# [00:38] * coop is now known as coop|afk
# [00:38] * ctalbert|mtg is now known as ctalbert
# [00:38] <bholley> I just want an nsIThreadInternal* for the current thread. This seems like way too much work
# [00:39] * Quits: imphil (philipp@moz-655EF802.customer.m-online.net) (Ping timeout)
# [00:39] <ted> hrm, guess not :-/
# [00:39] <ted> anyone else on today's linux nightly?
# [00:39] <ted> if i hit enter in a bugzilla textbox i get an extra blank line
# [00:39] <ted> like my line-height is doubled
# [00:40] <bent> bholley, nsCOMPtr<nsIThreadInternal> thread = do_QueryInterface(NS_GetCurrentThread());
# [00:40] <biesi> bholley, note that there's NS_GetCurrentThread
# [00:40] <ted> seems to be goofing up my mxr display too
# [00:40] <bholley> bent, biesi: thanks
# [00:40] <jcranmer> gosh
# [00:40] <jcranmer> how do you network people stand writing tests?
# [00:41] * Joins: jprmc (jprmc@CAC64D5C.5BCEC6DB.DA78B690.IP)
# [00:42] * mcote is now known as mcote|afk
# [00:43] * Quits: squib (squib-@moz-3F6F2A9C.ep.wisc.edu) (Quit: Leaving)
# [00:46] * Quits: Enn (enn@moz-DB6467E3.cpe.net.cable.rogers.com) (Ping timeout)
# [00:47] * Joins: gwagner (gwagner@moz-C03D0C61.vlan426.asr1.sfo1.gblx.net)
# [00:47] * Joins: sheppy (sheppy@moz-D0C71163.domainhotel-wifi.mozilla.hq)
# [00:47] <RyanVM> ted: bug 716229
# [00:50] * davehunt|mtg is now known as davehunt
# [00:51] * Quits: Archaeopteryx (itsme@moz-756328DB.cust.telecolumbus.net) (Ping timeout)
# [00:51] <RyanVM> so I discovered yesterday that CoverItLove on continuous update makes for a stuttery browser
# [00:51] <RyanVM> problem's going to be filing a useful bug since it's used for live chats during events
# [00:51] <RyanVM> CoverItLive*
# [00:52] <jwatt> do the build machines for the official Mac FF builds use MacPorts or homebrew?
# [00:53] * Joins: Archaeopteryx (itsme@moz-756328DB.cust.telecolumbus.net)
# [00:54] * Parts: BenB (ben@B911518A.DD7DF417.DDA41192.IP) (Konversation terminating Quasselstrippen)
# [00:54] * Joins: rillian_ (giles@21B7B9F2.B87E9213.6E712CE2.IP)
# [00:56] * Quits: rillian_ (giles@21B7B9F2.B87E9213.6E712CE2.IP) (Ping timeout)
# [00:56] * bc is now known as bc|afk
# [00:57] <Waldo> jlebar: I think at this rate I'm not going to have the linkedlist commentary today, unfortunately
# [00:57] <Waldo> jlebar: should be tomorrow, tho
# [00:57] <nthomas|away> jwatt: macports
# [00:57] * Quits: karl (karl@moz-1486CCF5.jetstream.xtra.co.nz) (Ping timeout)
# [00:57] * nthomas|away is now known as nthomas
# [00:57] <jwatt> nthomas: thanks
# [00:57] <nthomas> jwatt: why do you ask ?
# [00:57] <jlebar> Waldo, No worries. If I wasn't clear in the bug, I'd be comfortable with just a general signoff on "this is a good idea". But if you want to properly review, that's even better!
# [00:58] <Waldo> jlebar: "having a class for linked lists is a good idea" :-)
# [00:58] <Waldo> jlebar: but that doesn't address the half of it, really :-)
# [00:58] * Joins: mconley (mconley@moz-9C0F4EBF.cable.teksavvy.com)
# [00:58] <jlebar> :D The dependent patch won't be ready until tomorrow at the earliest, so I'm not in a rush.
# [00:59] * AaronMT is now known as AaronMT|afk
# [00:59] <jwatt> nthomas: because the page https://developer.mozilla.org/En/Developer_Guide/Build_Instructions/Mac_OS_X_Prerequisites seems to favor homebrew (by putting it before macports)
# [01:00] <nthomas> ah
# [01:00] <nthomas> the build machines have an old snapshot of macports anyway, so there's the potential for issues either way
# [01:00] <njn> bz: is it accurate to say that a |window| object is not a DOM node, but a |document| object is?
# [01:00] * njn is not a web developer
# [01:01] <jwatt> nthomas: ah, ok
# [01:01] * Joins: bbondy (bbondy@moz-28CF6D1C.home.cgocable.net)
# [01:02] * Quits: sheppy (sheppy@moz-D0C71163.domainhotel-wifi.mozilla.hq) (Quit: sheppy)
# [01:02] * Joins: willy1234x1 (willy1234x@moz-FEC54718.slkc.qwest.net)
# [01:02] <dholbert> glandium, ping?
# [01:02] <RyanVM> mccr8: ping
# [01:02] * Quits: Cameron (Cameron@moz-5911F6E0.range86-130.btcentralplus.com) (Ping timeout)
# [01:03] * Quits: bmoss (bmoss@moz-BBE3ABD.mv.mozilla.com) (Ping timeout)
# [01:03] <mccr8> RyanVM: pong
# [01:03] <RyanVM> mccr8: interested in the CoverItLive issue?
# [01:04] <mccr8> RyanVM: I don't know what that is, sorry.
# [01:04] * Joins: Enn (enn@moz-DB6467E3.cpe.net.cable.rogers.com)
# [01:04] <RyanVM> CoverItLive is a commonly-used way for doing live chats during events
# [01:04] <RyanVM> I found last night that it also makes Fx very stuttery
# [01:05] <RyanVM> but due to the nature of it being used during events, it's going to be difficult to debug without timing it well
# [01:05] <RyanVM> but ESPN should have one going tonight during the BCS championship game
# [01:05] <mccr8> okay. well, you can CC me on any bug you have filed on it.
# [01:06] <RyanVM> i'll file it in a bit and CC you
# [01:06] <mccr8> thanks.
# [01:06] <RyanVM> thank you!
# [01:06] <mccr8> maybe they have a non-live test feed or something.
# [01:06] <dholbert> glandium, un-ping
# [01:06] * Joins: bsmith (bsmith@moz-C03D0C61.vlan426.asr1.sfo1.gblx.net)
# [01:06] * shorlander-away is now known as shorlander
# [01:07] * Joins: matti (chatzilla@moz-B650AD27.dip.t-dialin.net)
# [01:07] <RyanVM> mccr8: http://www.coveritlive.com/index.php?option=com_altcaster&task=listliveevents&Itemid=223
# [01:07] <RyanVM> bingo (Sorta :))
# [01:08] * Quits: philipp64|laptop (chatzilla@moz-49802010.redfish-solutions.com) (Client exited)
# [01:08] * Joins: bmoss (bmoss@moz-BBE3ABD.mv.mozilla.com)
# [01:08] * Quits: Enn (enn@moz-DB6467E3.cpe.net.cable.rogers.com) (Ping timeout)
# [01:09] <jwatt> anyone here worked through the problems of getting m-r or m-b built using the new Xcode 4.2?
# [01:10] * Quits: jrmuizel (jrmuizel@F2D29657.F60B0462.67AC9B1.IP) (Input/output error)
# [01:10] * jimm-bbiab is now known as jimm
# [01:10] * Quits: matti (chatzilla@moz-B650AD27.dip.t-dialin.net) (services.mozilla.org (Too many invalid passwords))
# [01:11] * Joins: philipp64|laptop (chatzilla@moz-49802010.redfish-solutions.com)
# [01:12] * Quits: AutomatedTester (AutomatedT@moz-25F2601E.as13285.net) (Quit: AutomatedTester)
# [01:12] * Quits: nthomas (chatzilla@moz-75DA4CCD.dsl.telstraclear.net) (Ping timeout)
# [01:12] * Quits: vladan (vladan@F2D29657.F60B0462.67AC9B1.IP) (Ping timeout)
# [01:12] * Quits: dao (dao@moz-C1C7E228.superkabel.de) (Quit: Leaving.)
# [01:12] * Joins: Enn (enn@moz-DB6467E3.cpe.net.cable.rogers.com)
# [01:13] * wlach is now known as wlach|afk
# [01:13] <Waldo> njn: that's correct; DOM nodes are things like elements, attributenodes (although we might have removed those by not, not sure), processing instructions, documents, comment nodes, text nodes, and such
# [01:13] <Waldo> njn: window is what generally gets called "DOM 0"
# [01:13] * zpao is now known as zpao|detached
# [01:13] <Waldo> njn: and is only now-ish being standardized by whatwg and some new-ish W3C spec
# [01:13] <njn> Waldo: so the dom is everything within window.document, but not window itself?
# [01:13] <Waldo> njn: yes
# [01:14] <njn> Waldo: k, thx
# [01:14] * Joins: jrmuizel (jrmuizel@F2D29657.F60B0462.67AC9B1.IP)
# [01:14] * Joins: clokep (clokep@moz-69FB3955.c3-0.smr-ubr1.sbo-smr.ma.cable.rcn.com)
# [01:15] * Quits: Waldo (waldo@moz-2A0E3E7E.dsl.emhril.sbcglobal.net) (Quit: ChatZilla 0.9.87-4.1450hg.fc15 [XULRunner 8.0/20111108090029])
# [01:16] * Joins: nthomas (chatzilla@moz-75DA4CCD.dsl.telstraclear.net)
# [01:16] * Joins: myk1 (myk@moz-C03D0C61.vlan426.asr1.sfo1.gblx.net)
# [01:16] * Quits: myk (myk@moz-C03D0C61.vlan426.asr1.sfo1.gblx.net) (Connection reset by peer)
# [01:17] <Mook_as> attributenodes are deprecated but not removed, IIRC; and in (very far fetched) theory you can have more than one window per DOM, so they're not DOM nodes
# [01:17] * Quits: anant (anant@moz-C03D0C61.vlan426.asr1.sfo1.gblx.net) (Quit: anant)
# [01:18] * Joins: pcwalton (pcwalton@moz-5B0F92F1.hsd1.ca.comcast.net)
# [01:18] <taras> bbondy: ping
# [01:18] <taras> is the silent update stuff automagic for everybody
# [01:19] <taras> or is only a certain percentage of nightlies getting it?
# [01:20] * Quits: mike5w3c (MikeS@moz-DAFE1A45.tokyo.ocn.ne.jp) (Quit: mike5w3c)
# [01:20] * Joins: adeubank (Mibbit@moz-9B1B1DF8.allocated.csupomona.edu)
# [01:20] * Quits: northAway (northWind@moz-20E58859.cable.teksavvy.com) (Ping timeout)
# [01:21] * Quits: beaufour (beaufour@18D5CC88.C7EE4FB2.ECED8BE3.IP) (Quit: beaufour)
# [01:23] * Joins: karl (karl@538BABFE.A073F3E.97BBD552.IP)
# [01:23] * Quits: nhirata (nhirata.bu@moz-BBE3ABD.mv.mozilla.com) (Quit: nhirata)
# [01:25] <jimm> taras: everybody I believe, except 64bit builds
# [01:25] * Joins: tonymec__ (tonymec@5B2A1B79.5A39B178.277517C1.IP)
# [01:25] * tonymec__ is now known as tonymec|away
# [01:25] * Joins: nhirata (nhirata.bu@moz-BBE3ABD.mv.mozilla.com)
# [01:26] <taras> jimm: everybody who installs through installer
# [01:26] <taras> ?
# [01:26] <taras> what about people who unzip a nightly build
# [01:26] * sfleiter is now known as sfleiter|away
# [01:27] <jimm> don't think that works. it came down for me in the nightly update
# [01:27] * Quits: Archaeopteryx (itsme@moz-756328DB.cust.telecolumbus.net) (Quit: Too much information in my brain driving me insane)
# [01:28] <taras> jimm: i'm only asking because once we land bug 692255, we'll need to be able to check that updater service is indeed speeding up startup
# [01:28] * Quits: sfleiter|away (stfl@moz-22F37930.hsi7.kabel-badenwuerttemberg.de) (Ping timeout)
# [01:28] * Quits: Joeh (joe@5A3923AA.BC22908.C7CEC4ED.IP) (Quit: Joeh)
# [01:28] * Quits: jrmuizel (jrmuizel@F2D29657.F60B0462.67AC9B1.IP) (Input/output error)
# [01:28] <taras> and sounds like we need a telemetry probe to report that the updater service is running or something
# [01:29] * Joins: jrmuizel (jrmuizel@F2D29657.F60B0462.67AC9B1.IP)
# [01:31] * Joins: anant (Anant@moz-271479F2.hsd1.ca.comcast.net)
# [01:32] * Quits: jrmuizel (jrmuizel@F2D29657.F60B0462.67AC9B1.IP) (Input/output error)
# [01:32] * Quits: ashughes (ashughes@21B7B9F2.B87E9213.6E712CE2.IP) (Quit: heading home)
# [01:32] * lurking_work was under the impression that zips never updated
# [01:34] * Quits: ehsan (ehsan@F2D29657.F60B0462.67AC9B1.IP) (Input/output error)
# [01:34] * Joins: mjschranz_ (mjschranz@F3B7A1B3.694CD917.A03BB2CC.IP)
# [01:35] <nthomas> sure they do
# [01:36] <lurking_work> heh, so its hourly's that don't anymore due to a change to take them off the nightly channel - Bah!
# [01:36] <jimm> the service needs to be installed, so, if an install/update doesn't run, it's not going to be present. the old update mechnism would kick in in those cases.
# [01:36] <taras> njn: ping
# [01:37] <taras> khuey|away: ping
# [01:37] <njn> taras: pong
# [01:38] <taras> njn: am i correct in that we do not track dom stuff referenced by a compartment(for memory use)
# [01:38] <lurking_work> taras: I think Khuey should be 1/2 way across the Atlantic by now
# [01:38] <taras> lurking_work: k
# [01:38] <jimm> taras: I thoing the prefetch delete thing turned out to be a mixed bag, speeding up some installs while slowing others?
# [01:38] <jimm> *thought
# [01:39] <njn> taras: not sure I understand the question
# [01:39] <taras> jimm: nah it was fine
# [01:39] <njn> taras: we measure all DOM stuff reachable from window.document
# [01:39] * Quits: tH (Rob@cpc16-seac19-2-0-cust234.7-2.cable.virginmedia.com) (Ping timeout)
# [01:39] <taras> njn: can we saw compartment x is responsible for 2x as much memory consumption than y
# [01:39] <njn> taras: but if we have DOM stuff not reachable from there but reachable from JS, then yet, that won't be measured
# [01:39] <taras> s/saw/say/
# [01:39] * Joins: tH (Rob@cpc16-seac19-2-0-cust234.7-2.cable.virginmedia.com)
# [01:40] <RyanVM> mccr8 (or anyone else who knows): Does the javascript.mem.log pref require a restart to take effect?
# [01:40] <njn> taras: I don't understand that question either
# [01:40] <mccr8> RyanVM: no, I don't think so.
# [01:40] * Quits: AaronMT|afk (AaronMT@moz-DB17C53A.cpe.net.cable.rogers.com) (Quit: Textual IRC Client: http://www.textualapp.com/)
# [01:40] <taras> njn: i'd like to measure relative cost of tabs
# [01:40] <taras> or atleast compartments
# [01:40] <taras> in terms of cpu
# [01:40] * Quits: Mardak (Mardak@moz-BBE3ABD.mv.mozilla.com) (Quit: Mardak)
# [01:40] <mccr8> RyanVM: it is javascript.options.mem.log
# [01:40] <njn> taras: bug 687724 is open for per-tab memory reporting
# [01:40] <taras> however they can cause indirect cpu costs by using up memory -> causing CC
# [01:40] <RyanVM> yes, it works without restarting
# [01:41] * timA|lunch is now known as timA
# [01:41] <njn> taras: bug 674779 ?
# [01:41] <RyanVM> mccr8: I've got it logging, but so far they're looking normal
# [01:41] <njn> taras: sounds like a hard problem you're trying to solve
# [01:42] <taras> njn: shiny
# [01:42] * ewong|away is now known as ewong
# [01:42] <taras> njn: ok, that's basically what i want
# [01:43] <timA> I'm taking an existing class that implements an XPCOM interface and making it implement a new XPCOM interface. How do I avoid "ambiguous cast" errors when other code tries to static_cast<nsISupports*>(myClass*)?
# [01:43] <taras> so that + memory usage to allocate some of the gc cost to the tab
# [01:43] <njn> taras: which "that" are you referring to?
# [01:43] <taras> should do nicely
# [01:43] * Quits: adeubank (Mibbit@moz-9B1B1DF8.allocated.csupomona.edu) (Quit: http://www.mibbit.com ajax IRC Client)
# [01:43] <taras> that = 674779
# [01:43] <njn> taras: ok, but it's barely started
# [01:43] <taras> that's better than nothing :)
# [01:43] <njn> taras: and allocating the gc cost sounds really hard
# [01:43] <njn> ok, good
# [01:44] <taras> njn: a rough estimate should work to start with
# [01:44] <njn> taras: people will start bitching about the roughness about 5 minutes later, trust me :)
# [01:44] <taras> idea is if something is allocating a lot of stuff(ie due to twitter updates) we should be able to throttle settimeout + xmlhtprequest
# [01:44] <RyanVM> mccr8: is 107ms considered long for a GC?
# [01:44] <RyanVM> GC(T+149.3) Type:Glob, Total:107.2, Wait:0.2, Mark:59.2, Sweep:37.8, FinObj:10.2, FinStr:5.4, FinScr:0.8, FinShp:3.9, DisCod:3.1, DisAnl:7.5, XPCnct:2.3, Destry:0.1, End:11.9, +Chu:0, -Chu:0, Reason:Mallc
# [01:44] * Quits: cpearce (chatzilla@538BABFE.A073F3E.97BBD552.IP) (Ping timeout)
# [01:45] * davehunt is now known as davehunt|away
# [01:45] * shorlander is now known as shorlander-away
# [01:45] * Quits: mkaply (chatzilla@moz-92EDDD02.lightspeed.austtx.sbcglobal.net) (Ping timeout)
# [01:45] <Mook_as> timA: if you mean they now have problems due to multiple inheritance, they should switch to NS_ISUPPORTS_CAST ?
# [01:45] * Joins: cpearce (chatzilla@538BABFE.A073F3E.97BBD552.IP)
# [01:45] <mccr8> RyanVM: No, that's pretty good.
# [01:45] * zpao|detached is now known as zpao
# [01:46] <mccr8> RyanVM: The trick is to set your windows up so you can see the error console and the main window at the same time.
# [01:46] <mccr8> RyanVM: if you notice a bit of jankiness in your main window at about the same moment the erorr console spits something out, then you may have found the culprit.
# [01:47] * jcranmer is now known as jcranmer|away
# [01:48] <timA> Mook_as: the code that's having trouble (due to the now-multiple inheritance) is in nsCOMArray... maybe I should switch to using a different type of array?
# [01:50] <RyanVM> mccr8: OK, I'll keep an eye on it
# [01:50] * Quits: jet (junglecode@moz-C03D0C61.vlan426.asr1.sfo1.gblx.net) (Quit: jet)
# [01:51] * Quits: necolas (necolas@moz-C8C6D8F3.bb.sky.com) (Client exited)
# [01:51] <mccr8> RyanVM: great.
# [01:51] <mccr8> RyanVM: It is also possible that what is okay for normal browsing still makes something like streaming video unusuable.
# [01:52] * Joins: jet (junglecode@moz-C03D0C61.vlan426.asr1.sfo1.gblx.net)
# [01:52] <timA> Mook_as: to be more concrete, I'm making the nsDownload class implement a new XPCOM interface, and this (https://mxr.mozilla.org/mozilla-central/source/toolkit/components/downloads/nsDownloadManager.h#266) is the nsCOMArray that is now having troubles
# [01:52] * Quits: ewong (chatzilla@moz-E5D50C2E.static.netvigator.com) (Ping timeout)
# [01:53] <Mook_as> timA: sorry, the answers I can think of all involve replacing it with a different kind of array (possibly a nsTArray<nsRefPtr<nsDownload>>)
# [01:53] * Quits: nrc (chatzilla@538BABFE.A073F3E.97BBD552.IP) (Ping timeout)
# [01:54] * Quits: jprmc (jprmc@CAC64D5C.5BCEC6DB.DA78B690.IP) (Ping timeout)
# [01:55] * Joins: ewong (chatzilla@moz-E5D50C2E.static.netvigator.com)
# [01:55] * Quits: bsmith (bsmith@moz-C03D0C61.vlan426.asr1.sfo1.gblx.net) (Ping timeout)
# [01:56] <mbrubeck> margaret: XUL is burning on Aurora...
# [01:56] <timA> Mook_as: yeah that's the first thing I tried, but I ran into problems implementing this function (https://mxr.mozilla.org/mozilla-central/source/toolkit/components/downloads/nsIDownloadManager.idl#258) since "There are no enumerator objects that work on an nsTArray<T>" (https://developer.mozilla.org/en/XPCOM_array_guide#In-place_enumeration_4)
# [01:56] <mbrubeck> margaret: Nevermind
# [01:56] <mbrubeck> margaret: It's just bug https://bugzilla.mozilla.org/show_bug.cgi?id=714553
# [01:57] * Quits: ewong (chatzilla@moz-E5D50C2E.static.netvigator.com) (Quit: Reboot)
# [01:57] * Quits: bbondy (bbondy@moz-28CF6D1C.home.cgocable.net) (Ping timeout)
# [01:58] <timA> now I'm trying with nsIMutableArray, but before going too far down that path I wanted to see if anyone had ideas for simpler solutions
# [01:58] * Joins: jrmuizel (jrmuizel@moz-20EF8EAA.cpe.net.cable.rogers.com)
# [01:58] <margaret> mbrubeck: ok
# [01:59] <margaret> you had me worried!
# [01:59] <mbrubeck> sorry, I momentarily forgot that even if the change is Android-related, Android failures are *still* probably known flakiness. :P
# [02:02] * Joins: nrc (chatzilla@538BABFE.A073F3E.97BBD552.IP)
# [02:03] * Quits: gcp (gpascutto@moz-D0E475EA.access.telenet.be) (Quit: Make a new plan, Stan!)
# [02:03] <MattN> Is there any reason that I shouldn't use profDS to get the profile directory after startup? Does profD ever differ from profDS? This is for code that is called during and after startup.
# [02:05] * ctalbert is now known as ctalbert|afk
# [02:07] * Quits: damons (gnubeard@moz-BBE3ABD.mv.mozilla.com) (Quit: damons)
# [02:07] <Unfocused> MattN: technically, it's possible for profD to change during runtime, due to some ancient profile code that's only half supported
# [02:07] <Mossop> That isn't really possible. I'm not sure what happens in the case of the profile manager UI though
# [02:07] * Joins: ewong (chatzilla@moz-E5D50C2E.static.netvigator.com)
# [02:07] * Quits: jhammel (jhammel@moz-C03D0C61.vlan426.asr1.sfo1.gblx.net) (Quit: leaving)
# [02:08] * Joins: mijia (mijia@DC4232F0.766373FB.C3A57E70.IP)
# [02:09] * Joins: bretr (bret_recka@moz-BBE3ABD.mv.mozilla.com)
# [02:09] * Joins: mike5w3c (MikeS@moz-DAFE1A45.tokyo.ocn.ne.jp)
# [02:09] <MattN> Unfocused, Mossop: This is for migration code so it's after profile manager UI. Should I have my helper function in PlacesUtils just always use ProfDS since it will sometimes be called from a migrator or should I have it try profD and fallback to profDS?
# [02:10] <MattN> (migrators can run during startup and from the UI after startup)
# [02:11] * Quits: bholley (bholley@moz-BBE3ABD.mv.mozilla.com) (Quit: bholley)
# [02:11] * Quits: tH (Rob@cpc16-seac19-2-0-cust234.7-2.cable.virginmedia.com) (Ping timeout)
# [02:12] * timA is now known as timA|travelingToSunnyvaleForAp
# [02:12] * timA|travelingToSunnyvaleForAp is now known as timA|SunnyvaleAppsWorkWeek
# [02:12] * Quits: timA|SunnyvaleAppsWorkWeek (tabraldes@moz-535753DA.hsd1.or.comcast.net) (Quit: Instantbird 1.2a1pre)
# [02:13] * jcranmer|away is now known as jcranmer
# [02:14] * Quits: jimm (jmathies@moz-7F164CA1.pn.at.cox.net) (Quit: )
# [02:15] <Unfocused> pass? :)
# [02:15] <Unfocused> is ProfD not always available? i thought it was
# [02:17] <RyanVM> mccr8: *sigh* and of course, now I can't get it to reproduce (at least not on this chat)
# [02:18] * Joins: redfive (chatzilla@moz-B94E1A81.ipnetworksinc.net)
# [02:18] * Joins: ehsan (ehsan@moz-7D6E2170.cpe.net.cable.rogers.com)
# [02:19] <mccr8> RyanVM: maybe we fixed it. ;)
# [02:20] <MattN> Unfocused: not when the migrator runs during startup. I got an exception. Pass accepted ;)
# [02:21] * Quits: Pike (Pike@moz-269673CD.pool.mediaways.net) (Quit: ChatZilla 0.9.88 [Firefox 9.0.1/20111220165912])
# [02:21] * fabrice is now known as fabrice|afk
# [02:23] * Joins: Wevah (Wevah@moz-97AD33CE.stcd.qwest.net)
# [02:23] * Joins: raccettura (raccettura@moz-660B8F4B.hsd1.nj.comcast.net)
# [02:23] * Joins: alex (alex@86599527.E7FBFC77.AB369DBA.IP)
# [02:25] * Quits: MichaelR (MichaelR@moz-FB3B7308.hsd1.ma.comcast.net) (Ping timeout)
# [02:27] * Joins: Mossop_ (mossop@moz-BBE3ABD.mv.mozilla.com)
# [02:28] * Quits: Mossop (mossop@moz-BBE3ABD.mv.mozilla.com) (Ping timeout)
# [02:28] * Mossop_ is now known as Mossop
# [02:29] * Quits: sfink (chatzilla@moz-BBE3ABD.mv.mozilla.com) (Client exited)
# [02:29] <edmorley> dholbert: are you automating finding the candidate warning free directories, or by hand?
# [02:30] <dholbert> edmorley, I just picked a few dirs that I have a hand in, from https://bugzilla.mozilla.org/attachment.cgi?id=443193 , which indicates that they were warning-free ~18 months ago
# [02:30] * Quits: ehsan (ehsan@moz-7D6E2170.cpe.net.cable.rogers.com) (Input/output error)
# [02:30] <dholbert> the two bugs I filed (on SVG & tables) are the only ones I was intending to file/tweak myself at this point
# [02:31] * Quits: lurking_work (chatzilla@moz-107FCDBA.hfc.comcastbusiness.net) (Quit: ChatZilla 0.9.87 [Firefox 9.0.1/20111220165912])
# [02:31] <edmorley> ah thanks :-)
# [02:32] <dholbert> edmorley, np. you gonna be filing some bugs, too? (if so, good on you! :))
# [02:32] * Joins: njn_ (chatzilla@moz-67BF6169.dyn.iinet.net.au)
# [02:32] * Quits: njn (chatzilla@moz-6689C0E4.dyn.iinet.net.au) (Ping timeout)
# [02:32] * njn_ is now known as njn
# [02:33] <edmorley> yeah why not :-)
# [02:33] <dholbert> woot. :)
# [02:33] * Joins: tonymec (tonymec@5B2A1B79.5A39B178.277517C1.IP)
# [02:33] * Quits: dmb (dmb@moz-3565FEE9.da4.org) (Ping timeout)
# [02:33] * Quits: smooney (smooney@moz-BBE3ABD.mv.mozilla.com) (Quit: smooney)
# [02:33] * bwinton_away is now known as bwinton
# [02:33] * Parts: yuan (ywang@moz-C03D0C61.vlan426.asr1.sfo1.gblx.net)
# [02:34] * shorlander-away is now known as shorlander
# [02:34] * Joins: dmb (dmb@moz-3565FEE9.da4.org)
# [02:37] * Joins: nattokirai (nattokirai@moz-348F61F0.mozilla.or.jp)
# [02:40] * Quits: dmb (dmb@moz-3565FEE9.da4.org) (Ping timeout)
# [02:40] * Joins: dmb (dmb@moz-3565FEE9.da4.org)
# [02:42] * Quits: brendan (brendaneic@moz-BBE3ABD.mv.mozilla.com) (Quit: brendan)
# [02:43] * Quits: armenzg_away (armenzg@moz-8555CE12.cable.teksavvy.com) (Input/output error)
# [02:44] * Quits: bmoss (bmoss@moz-BBE3ABD.mv.mozilla.com) (Ping timeout)
# [02:45] * Joins: dalsh (dalsh@moz-40F12928.fbx.proxad.net)
# [02:45] * Quits: dalsh (dalsh@moz-40F12928.fbx.proxad.net) (Connection reset by peer)
# [02:46] * Quits: rniwa (rniwa@5CA6DC39.C60FE7DC.4065847B.IP) (Quit: rniwa)
# [02:47] * Joins: mkaply (chatzilla@moz-92EDDD02.lightspeed.austtx.sbcglobal.net)
# [02:47] * ChanServ sets mode: +o mkaply
# [02:47] * Quits: ewong (chatzilla@moz-E5D50C2E.static.netvigator.com) (Quit: reboot)
# [02:48] * Quits: JonathanS (JonathanS@moz-EF2236DC.geekbouncer.co.uk) (Ping timeout)
# [02:48] * Quits: wesj (wesj@moz-C03D0C61.vlan426.asr1.sfo1.gblx.net) (Input/output error)
# [02:48] * Joins: JonathanS (JonathanS@moz-EF2236DC.geekbouncer.co.uk)
# [02:49] * Joins: bmoss (bmoss@moz-BBE3ABD.mv.mozilla.com)
# [02:50] * KaiRo_away is now known as KaiRo
# [02:50] <jcranmer> this may seem like a weird question
# [02:50] * Quits: cpeterson (cpeterson@moz-C03D0C61.vlan426.asr1.sfo1.gblx.net) (Ping timeout)
# [02:51] * Quits: gandalf (zbraniecki@moz-66C04999.neoplus.adsl.tpnet.pl) (Quit: Computer has gone to sleep.)
# [02:51] <jcranmer> in practice, given PGO optimization
# [02:51] * Quits: bmoss (bmoss@moz-BBE3ABD.mv.mozilla.com) (Quit: bmoss)
# [02:51] <jcranmer> how feasible would it be to assume that two global variables immediately adjacent in the source file will be immediately adjacent post-optimization?
# [02:52] <derf> It's certainly not guaranteed, and I wouldn't rely on it in _my_ code.
# [02:52] <derf> Are they the same type?
# [02:52] <jcranmer> if you want context
# [02:52] * Joins: bretr_ (bret_recka@moz-BBE3ABD.mv.mozilla.com)
# [02:52] * Quits: bretr (bret_recka@moz-BBE3ABD.mv.mozilla.com) (Ping timeout)
# [02:52] * bretr_ is now known as bretr
# [02:53] * jcranmer pauses a bit
# [02:53] <jcranmer> no, that still wouldn't be enough
# [02:53] <jcranmer> never mind
# [02:53] <jcranmer> I'm trying to find out hacks that would let me set outputTimeStamp in prlog.c
# [02:53] * Quits: cpearce (chatzilla@538BABFE.A073F3E.97BBD552.IP) (Ping timeout)
# [02:54] <jcranmer> via jsctypes
# [02:54] * Quits: odin_ (Odin@moz-66D3502E.zone5.bethere.co.uk) (Ping timeout)
# [02:54] * Joins: KaIRC (robert@moz-EBD91930.adsl.highway.telekom.at)
# [02:54] * Quits: KaiRo (robert@moz-29AB93CC.adsl.highway.telekom.at) (Ping timeout)
# [02:55] <jcranmer> I'm pretty sure "reinitialize nspr" is a bad idea
# [02:56] * Quits: Mossop (mossop@moz-BBE3ABD.mv.mozilla.com) (Ping timeout)
# [02:56] * Quits: myk1 (myk@moz-C03D0C61.vlan426.asr1.sfo1.gblx.net) (Ping timeout)
# [02:57] * zpao is now known as zpao|detached
# [02:58] * Quits: akeybl_ (akeybl@moz-BBE3ABD.mv.mozilla.com) (Quit: Leaving...)
# [02:58] * Quits: @mkaply (chatzilla@moz-92EDDD02.lightspeed.austtx.sbcglobal.net) (Quit: ChatZilla 0.9.87-rdmsoft [XULRunner 1.9.0.17/2009122204])
# [02:58] * Quits: biesi (cbiesinger@moz-76ED95A4.corp.tfbnw.net) (Ping timeout)
# [02:58] * Quits: Mook_as (mook@moz-1FCC0032.activestate.com) (Quit: ChatZilla 0.9.88-rdmsoft [XULRunner 1.9.2.13/20101203074205])
# [03:02] * Quits: bent (chatzilla@moz-C03D0C61.vlan426.asr1.sfo1.gblx.net) (Quit: ChatZilla 0.9.88-rdmsoft [XULRunner 1.9.2.19/20110707195905])
# [03:02] * Quits: kumar (kmcmillan@moz-BBE3ABD.mv.mozilla.com) (Quit: kumar)
# [03:04] * Quits: Ami_Ty (Amie@moz-BBE3ABD.mv.mozilla.com) (Quit: Ami_Ty)
# [03:04] * Joins: jdm (jdm@moz-9AEDE212.cpe.net.cable.rogers.com)
# [03:06] * Joins: dRdR (dRdR@357E2FA1.983BBEF4.60C86946.IP)
# [03:06] * Quits: sicking (chatzilla@moz-C03D0C61.vlan426.asr1.sfo1.gblx.net) (Ping timeout)
# [03:07] * Quits: By-Tor (bytor@moz-46974D0B.dyn.optonline.net) (Quit: Leaving)
# [03:09] * Joins: odin_ (Odin@moz-66D3502E.zone5.bethere.co.uk)
# [03:09] * Quits: clee (clee@moz-BBE3ABD.mv.mozilla.com) (Quit: clee)
# [03:09] * Joins: Mardak (Mardak@moz-4FA48382.hsd1.ca.comcast.net)
# [03:10] * Quits: dRdR (dRdR@357E2FA1.983BBEF4.60C86946.IP) (Client exited)
# [03:11] * bwinton is now known as bwinton_away
# [03:12] * Quits: jet (junglecode@moz-C03D0C61.vlan426.asr1.sfo1.gblx.net) (Quit: jet)
# [03:12] * Quits: nhirata (nhirata.bu@moz-BBE3ABD.mv.mozilla.com) (Quit: nhirata)
# [03:13] * Quits: andreasn (andreasn@moz-D8A35C52.a199.priv.bahnhof.se) (Ping timeout)
# [03:15] * Quits: mccr8 (mccr8@moz-BBE3ABD.mv.mozilla.com) (Quit: mccr8)
# [03:15] * Parts: mevans (mevans@moz-5BBCACEF.hsd1.ca.comcast.net)
# [03:15] * Quits: redfive (chatzilla@moz-B94E1A81.ipnetworksinc.net) (Ping timeout)
# [03:17] * Joins: ewong (chatzilla@moz-E5D50C2E.static.netvigator.com)
# [03:22] * jwir3 is now known as jwir3|away
# [03:23] * jmaher|afk is now known as jmaher
# [03:23] * shorlander is now known as shorlander-away
# [03:23] * Joins: brendan (brendaneic@moz-BBE3ABD.mv.mozilla.com)
# [03:29] * Quits: DGMurdockIII (dgmurdocki@moz-E933B63.hsd1.il.comcast.net) (Quit: get satisfied! :: core-networks.de ««« (Gamers.IRC) »»» gamersirc.net ::)
# [03:29] * Quits: faramarz (faramarz@moz-BBE3ABD.mv.mozilla.com) (Quit: faramarz)
# [03:30] * Joins: By-Tor (bytor@moz-46974D0B.dyn.optonline.net)
# [03:30] * Quits: chewey (chewey@moz-FC53C82F.dip0.t-ipconnect.de) (NickServ (GHOST command used by chewey_))
# [03:30] * Joins: chewey (chewey@moz-D9216754.dip0.t-ipconnect.de)
# [03:33] * Joins: fabrice (fabrice@moz-94F028C6.hsd1.ca.comcast.net)
# [03:39] * Quits: KaIRC (robert@moz-EBD91930.adsl.highway.telekom.at) (Input/output error)
# [03:40] * Quits: juanb (jbecerra@moz-BBE3ABD.mv.mozilla.com) (Quit: juanb)
# [03:40] * Quits: gkw (gkw@moz-BBE3ABD.mv.mozilla.com) (Quit: Instantbird 1.1)
# [03:43] * Quits: ewong (chatzilla@moz-E5D50C2E.static.netvigator.com) (Quit: reboot)
# [03:44] * Quits: asac_ (asac@moz-F66A6FC4.pppoe.wtnet.de) (Ping timeout)
# [03:45] * Joins: bsmith (bsmith@moz-C03D0C61.vlan426.asr1.sfo1.gblx.net)
# [03:47] * Joins: asac (asac@moz-AF572D83.pppoe.wtnet.de)
# [03:50] * Quits: brendan (brendaneic@moz-BBE3ABD.mv.mozilla.com) (Quit: brendan)
# [03:50] * Quits: Jesse (jruderman@moz-BBE3ABD.mv.mozilla.com) (Quit: Jesse)
# [03:50] * Joins: larfdesk (Adam_Hinke@moz-F92153ED.longlines.com)
# [03:51] * Joins: ewong (chatzilla@moz-E5D50C2E.static.netvigator.com)
# [03:53] * Quits: tonymec (tonymec@5B2A1B79.5A39B178.277517C1.IP) (Input/output error)
# [03:54] * shorlander-away is now known as shorlander
# [03:54] * mjschranz_ is now known as mjschranz
# [03:56] * Joins: tonymec (tonymec@5B2A1B79.5A39B178.277517C1.IP)
# [03:56] * Quits: hub (hub@21B7B9F2.B87E9213.6E712CE2.IP) (Input/output error)
# [03:56] * Joins: rniwa (rniwa@60A74940.D6CCE4AE.77834EAA.IP)
# [03:58] * Joins: dRdR (dRdR@357E2FA1.983BBEF4.60C86946.IP)
# [03:58] * sancus is now known as sancus_
# [03:58] * sancus_ is now known as sancus
# [03:59] * Quits: fabrice (fabrice@moz-94F028C6.hsd1.ca.comcast.net) (Ping timeout)
# [04:00] * shorlander is now known as shorlander-away
# [04:02] * Quits: clokep (clokep@moz-69FB3955.c3-0.smr-ubr1.sbo-smr.ma.cable.rcn.com) (Quit: Instantbird 1.2a1pre)
# [04:02] * Quits: azakai (alon@moz-BBE3ABD.mv.mozilla.com) (Ping timeout)
# [04:05] * Joins: sworkman_ (sworkman@moz-BBE3ABD.mv.mozilla.com)
# [04:05] * Quits: sworkman_ (sworkman@moz-BBE3ABD.mv.mozilla.com) (Quit: sworkman_)
# [04:06] * Quits: RyanVM (chatzilla@moz-D04D3C77.phlapa.fios.verizon.net) (Quit: ChatZilla 0.9.88 [Firefox 12.0a1/20120107171323])
# [04:06] * Quits: sworkman (sworkman@moz-BBE3ABD.mv.mozilla.com) (Ping timeout)
# [04:06] * Quits: mjschranz (mjschranz@F3B7A1B3.694CD917.A03BB2CC.IP) (Client exited)
# [04:07] * Joins: clokep (clokep@moz-69FB3955.c3-0.smr-ubr1.sbo-smr.ma.cable.rcn.com)
# [04:09] * Quits: shorlander-away (shorlander@moz-853043D6.dhcp.insightbb.com) (Quit: Quit)
# [04:10] * Joins: cpearce (chatzilla@moz-56BC07D4.xdsl.xnet.co.nz)
# [04:12] * Quits: jwatt (roslea@jwatt.irc.users.mozilla.org) (Ping timeout)
# [04:13] * Joins: jwatt (roslea@jwatt.irc.users.mozilla.org)
# [04:14] * Quits: asadotzler (asa@moz-BBE3ABD.mv.mozilla.com) (Ping timeout)
# [04:16] * Joins: fabrice (fabrice@moz-104CC309.mv.mozilla.com)
# [04:16] * Joins: lilmatt (mwillis@moz-B7412EC8.lightspeed.snfcca.sbcglobal.net)
# [04:17] * Quits: edmorley (edmorley@moz-384CD16.range86-166.btcentralplus.com) (Quit: nn)
# [04:17] * Joins: juanb (jbecerra@moz-F1012875.hsd1.ca.comcast.net)
# [04:18] * Joins: hub (hub@moz-E2FCA694.figuiere.net)
# [04:19] * Quits: mbrubeck (mbrubeck@moz-755AD63.hsd1.wa.comcast.net) (Ping timeout)
# [04:19] * Quits: mike5w3c (MikeS@moz-DAFE1A45.tokyo.ocn.ne.jp) (Quit: mike5w3c)
# [04:19] * jhopkins is now known as jhopkins|away
# [04:22] * Quits: mconley (mconley@moz-9C0F4EBF.cable.teksavvy.com) (Input/output error)
# [04:23] * Quits: darktrojan (geoff@moz-57EC95C5.dsl.telstraclear.net) (Quit: darktrojan)
# [04:30] * Joins: mike5w3c (MikeS@moz-DAFE1A45.tokyo.ocn.ne.jp)
# [04:31] * Joins: rniwa_ (rniwa@moz-E171DA5.sfba.hfc.comcastbusiness.net)
# [04:32] <nattokirai> wondering if anyone can point me at a good example of off-main-thread i/o in gecko
# [04:32] * Quits: rniwa (rniwa@60A74940.D6CCE4AE.77834EAA.IP) (Ping timeout)
# [04:32] * rniwa_ is now known as rniwa
# [04:34] * jmaher is now known as jmaher|afk
# [04:35] * Joins: darktrojan (geoff@moz-57EC95C5.dsl.telstraclear.net)
# [04:37] * Quits: jrmuizel (jrmuizel@moz-20EF8EAA.cpe.net.cable.rogers.com) (Input/output error)
# [04:39] <jdm> how do I get a checkbox onchange function to trigger when I programmatically check the input element?
# [04:39] * Joins: flx_ (flx@590D15DC.A7CE92B2.30E47D40.IP)
# [04:40] * Joins: Jesse (jruderman@moz-537BCF9.hsd1.ca.comcast.net)
# [04:41] * cjones is now known as cjones-dinner
# [04:41] * philor|away is now known as philor
# [04:44] * Quits: nrc (chatzilla@538BABFE.A073F3E.97BBD552.IP) (Ping timeout)
# [04:48] * Quits: jlebar (jlebar@moz-3F3A6302.dyn.columbia.edu) (Input/output error)
# [04:48] * Quits: clokep (clokep@moz-69FB3955.c3-0.smr-ubr1.sbo-smr.ma.cable.rcn.com) (Quit: Instantbird 1.2a1pre)
# [04:51] * Quits: dRdR (dRdR@357E2FA1.983BBEF4.60C86946.IP) (Client exited)
# [04:55] <@bz> jdm: you don't
# [04:55] * Quits: jgriffin (jgriffin@moz-4FBFA41D.hsd1.wa.comcast.net) (Quit: jgriffin)
# [04:55] <jdm> yeah, I ended up faking a call to the function
# [04:55] <@bz> jdm: the assumption is that if script is changing stuff it can do whatever else it wants to
# [04:55] * @bz hacks up code some more
# [04:58] * Joins: mauke_ (mauke@moz-6A01E0D8.superkabel.de)
# [04:59] * Quits: mauke (mauke@moz-6A01E0D8.superkabel.de) (Ping timeout)
# [05:02] * mauke_ is now known as mauke
# [05:05] * Quits: jimb (user@moz-F4EC06CC.hsd1.or.comcast.net) (Ping timeout)
# [05:07] <@bz> can I have tbpl show me two revs only?
# [05:07] * Quits: Enn (enn@moz-DB6467E3.cpe.net.cable.rogers.com) (Ping timeout)
# [05:07] <@bz> or can I only do one rev or the whole page?
# [05:09] * Quits: dveditz (dveditz@moz-BBE3ABD.mv.mozilla.com) (Quit: dveditz)
# [05:10] <philor> seems like once upon a time tbpl took the same params as pushloghtml
# [05:11] <philor> which I always get wrong because they're undocumented, and my addressbar history is full of my wrong memory
# [05:12] * Joins: regen (Miller@moz-DFCD66B7.adsl.dynamic.seed.net.tw)
# [05:14] <@bz> philor: perfect, thanks
# [05:14] * Joins: bholley (bholley@moz-FCAF9AAB.hsd1.ca.comcast.net)
# [05:14] <@bz> philor: at least I hope it's perfect. ;)
# [05:15] <philor> yeah, could well be a two-rewrites-ago memory
# [05:15] * Quits: hub (hub@moz-E2FCA694.figuiere.net) (Input/output error)
# [05:15] <@bz> well
# [05:15] <@bz> https://tbpl.mozilla.org/?tree=Try&fromchange=560929a3b65c&tochange=05ba0d82edeb works fine
# [05:15] <@bz> but we'll know for sure once I have talos results and tick those ticks
# [05:16] <@bz> (which is why I want to load two results at once, of course)
# [05:16] <philor> mmm, compare-talos
# [05:16] <philor> wonder whether mconnor has deployed yet
# [05:16] * Joins: asadotzler (asa@F259CFB7.9EF26CBA.EB5E51FC.IP)
# [05:16] <philor> nyet
# [05:16] <philor> note to self: fork that into tbpl
# [05:17] <@bz> philor: yes, he has
# [05:17] <@bz> philor: or at least I used it earlier today and it worked
# [05:17] <@bz> philor: http://perf.snarkfest.net/compare-talos/breakdown.html?oldTestIds=9660404&newTestIds=9661405&testName=tdhtml_paint
# [05:18] <philor> huh, doesn't work for me, wonder whether it over-caches
# [05:18] <philor> indeed it does
# [05:18] * Quits: njn (chatzilla@moz-67BF6169.dyn.iinet.net.au) (Client exited)
# [05:20] * nthomas is now known as nthomas|away
# [05:20] * Quits: jdm (jdm@moz-9AEDE212.cpe.net.cable.rogers.com) (Client exited)
# [05:21] * Quits: dbradley (dbradley@moz-8FACAA93.fuse.net) (Ping timeout)
# [05:21] * Quits: jwatt (roslea@jwatt.irc.users.mozilla.org) (Ping timeout)
# [05:23] * Quits: jorendorff (jorendorff@moz-91590D94.hsd1.tn.comcast.net) (Quit: jorendorff)
# [05:24] * Quits: bnicholson (bnicholson@moz-BBE3ABD.mv.mozilla.com) (Ping timeout)
# [05:24] * Quits: darktrojan (geoff@moz-57EC95C5.dsl.telstraclear.net) (Ping timeout)
# [05:25] * Joins: dbradley (dbradley@moz-8FACAA93.fuse.net)
# [05:27] * Quits: lilmatt (mwillis@moz-B7412EC8.lightspeed.snfcca.sbcglobal.net) (Quit: lilmatt)
# [05:31] * glob|away is now known as glob
# [05:36] * philor is now known as philor|away
# [05:37] * Joins: ddahl (ddahl@moz-6971CF66.hsd1.il.comcast.net)
# [05:38] * Joins: bnicholson (bnicholson@moz-D1AB96C1.hsd1.ca.comcast.net)
# [05:39] * Quits: bretr (bret_recka@moz-BBE3ABD.mv.mozilla.com) (Quit: bretr)
# [05:42] <mconnor> philor|away: please fork that into tbpl
# [05:42] <mconnor> _please_
# [05:44] <roc> bz: what do you think about delaying firing mousemove until the refresh driver, as well?
# [05:45] * cjones-dinner is now known as cjones
# [05:45] * philor|away is now known as philor
# [05:47] * Quits: rjohnson19 (rjohnson19@moz-9148485F.hsd1.ma.comcast.net) (Quit: Leaving)
# [05:51] * Quits: karl (karl@538BABFE.A073F3E.97BBD552.IP) (Ping timeout)
# [05:51] * Joins: akeybl (akeybl@moz-BBE3ABD.mv.mozilla.com)
# [05:51] * Quits: akeybl (akeybl@moz-BBE3ABD.mv.mozilla.com) (Input/output error)
# [05:53] * Joins: ctopper (craig@moz-42075866.hsd1.or.comcast.net)
# [06:04] * Quits: terrence (terrence@moz-BBE3ABD.mv.mozilla.com) (Ping timeout)
# [06:07] * Joins: smooney (smooney@moz-3C7A0050.hsd1.ca.comcast.net)
# [06:07] <@bz> roc: I think it sounds like a great idea
# [06:08] <@bz> roc: and not something I wanted to try tonight
# [06:08] <roc> like Western civilization eh?
# [06:08] <roc> ok
# [06:08] <@bz> roc: for one thing, I'm not quite sure how we go about firing mousemove.
# [06:08] * Quits: Polynomial-C (Poly-C@moz-6226FA4E.dip.t-dialin.net) (Ping timeout)
# [06:08] * Joins: Poly-C (Poly-C@moz-84686AC3.dip.t-dialin.net)
# [06:08] <@bz> roc: would need to dig some
# [06:08] <@bz> roc: the patch for doing it for the synth event, assuming I got it right, is almost surprisingly painless
# [06:09] <roc> yeah
# [06:09] <derf> Hmm, xiphmont was complaining about that the other day.
# [06:09] <@bz> roc: https://hg.mozilla.org/try/rev/05ba0d82edeb if you want to look
# [06:09] <derf> 22:31:17 < xiphmont> Ah, nightly has introduced something new and annoying--- it is apparently processing mouse movement async from mouseup.
# [06:09] <roc> mousemove fires in response to OS events, so we'd have to do something nontrivial
# [06:09] <derf> 22:31:34 < xiphmont> And there's some delay before processing the mouseups.
# [06:09] <roc> I already looked :-)
# [06:09] <derf> 22:31:47 < xiphmont> So ~ every time I click on a tab, it accidentally tears it off into a new window
# [06:10] <roc> that sounds like a regression from bz's work :-(
# [06:10] <glob> asa, are you running cuttherope.ie on nightly? (doesn't work for me :( )
# [06:10] <@bz> mm
# [06:10] <derf> I didn't know who to blame until just now.
# [06:10] <@bz> regression from what work?
# [06:10] * Joins: m_kato (m_kato@moz-348F61F0.mozilla.or.jp)
# [06:10] <@bz> roc: yeah, the os event thing is why I didn't want to mess with it...
# [06:10] * Quits: eflores (AndChat@538BABFE.A073F3E.97BBD552.IP) (Quit: Bye)
# [06:11] <roc> your patch to flush on mousemove landed on nightly,right?
# [06:11] <@bz> derf: which nightly is xiph using?
# [06:11] <@bz> roc: good question
# [06:11] <@bz> roc: no
# [06:11] <derf> bz: He said that on the 8th.
# [06:11] <@bz> roc: I landed it morning of the 9th on inbound; merged to m-c mid-day on the 9th
# [06:12] <@bz> roc: so will be in the nightlies that start building in a few hours
# [06:12] <derf> Today is the 9th.
# [06:12] <@bz> derf: precisely
# [06:12] <derf> Well, I guess not for bz anymore.
# [06:12] <@bz> derf: also precisely. ;)
# [06:12] <nigelb> Today is 10th ;)
# [06:12] <@bz> derf: and it's not for roc anymore either
# [06:12] <@bz> derf: and hasn't been for a good long while
# [06:12] <nigelb> roc lives in the future
# [06:13] <roc> I like it here
# [06:13] <nigelb> I'm midway
# [06:13] <derf> Yes. I can't wait to join him in it next week.
# [06:13] <@bz> derf: anyway, I'd be interested in hearing whether tomorrow's nightly also has the problem xiph was talking about
# [06:13] <@bz> derf: and if so, whether he's willing to bisect
# [06:13] <derf> bz: Okay. I will harass him about it.
# [06:13] <@bz> derf: thanks!
# [06:13] <derf> I have other things I need to bother him about anyway.
# [06:14] * Quits: romeo (romeo@moz-18AD2F99.k744.webspeed.dk) (Quit: Leaving)
# [06:15] <@bz> roc: well, preliminary data suggests that this might fix the regression, maybe
# [06:15] <@bz> roc: on both layers1 and scrolling
# [06:16] <@bz> roc: http://perf.snarkfest.net/compare-talos/breakdown.html?oldTestIds=9660404&newTestIds=9661405&testName=tdhtml_paint was the regression
# [06:16] <@bz> roc: http://perf.snarkfest.net/compare-talos/breakdown.html?oldTestIds=9669767&newTestIds=9669588&testName=tdhtml_paint is my try push
# [06:16] <roc> cool
# [06:16] <roc> I would expect it to
# [06:16] <@bz> roc: going to file a bug and all that
# [06:17] * @bz even has some passing tests
# [06:17] * Quits: Honza (chatzilla@E07BF19C.5BB5597D.D0083327.IP) (Ping timeout)
# [06:18] * Joins: Mook (mook@5365CE80.16C74E88.6F478678.IP)
# [06:18] * Joins: karl (karl@moz-C7DA82DE.jetstream.xtra.co.nz)
# [06:19] * Joins: Honza (chatzilla@E07BF19C.5BB5597D.D0083327.IP)
# [06:19] * Joins: terrence (terrence@moz-BBE3ABD.mv.mozilla.com)
# [06:20] * Quits: lsumar (lsumar@538BABFE.A073F3E.97BBD552.IP) (Quit: This computer has gone to sleep)
# [06:21] * Quits: terrence (terrence@moz-BBE3ABD.mv.mozilla.com) (Ping timeout)
# [06:22] * Quits: rniwa (rniwa@moz-E171DA5.sfba.hfc.comcastbusiness.net) (Client exited)
# [06:22] * Joins: rniwa (rniwa@60A74940.D6CCE4AE.77834EAA.IP)
# [06:22] <@bz> roc: https://bugzilla.mozilla.org/attachment.cgi?id=587241&action=edit
# [06:27] * Joins: dRdR (dRdR@moz-792B0937.uwaterloo.ca)
# [06:28] * Quits: caillon (caillon@moz-359E39FC.lightspeed.sntcca.sbcglobal.net) (Quit: caillon)
# [06:29] * Quits: fzzzy (donovan@moz-2B41AF9B.lightspeed.mtvwca.sbcglobal.net) (Quit: fzzzy)
# [06:31] * Joins: gkw (gkw@moz-A7D8CA2A.hsd1.ca.comcast.net)
# [06:31] * Quits: roc (chatzilla@538BABFE.A073F3E.97BBD552.IP) (Ping timeout)
# [06:35] * Quits: mike5w3c (MikeS@moz-DAFE1A45.tokyo.ocn.ne.jp) (Quit: mike5w3c)
# [06:37] <philor> bz: did https://hg.mozilla.org/try/rev/a1c61f01b1ed really need that ancient parent?
# [06:37] * Joins: yuan (ywang@moz-6A0DF30E.hsd1.ca.comcast.net)
# [06:38] <philor> that's back in the "crash during Win M1 in a way that will then break the next talos job that slave does on some other tree" era
# [06:39] <@bz> philor: I didn't realize that tree was that old. :(
# [06:40] <@bz> philor: I just grabbed a clean tree, wrote the code, and pushed it to try
# [06:40] <philor> bz: better hope nobody broke you in the last month, then :)
# [06:40] <@bz> philor: yep
# [06:40] <@bz> philor: I will. ;)
# [06:40] <@bz> philor: but since it might take another month to get review, I'm not worrying about it for now
# [06:42] <Jesse> bz: sounds like you need a nightly cron job that ensures you have at least 4 clean trees in the morning
# [06:43] * Quits: gkw (gkw@moz-A7D8CA2A.hsd1.ca.comcast.net) (Quit: Instantbird 1.1)
# [06:43] * Joins: gkw (gkw@moz-104CC309.mv.mozilla.com)
# [06:43] <@bz> Jesse: mmm
# [06:43] <@bz> Jesse: would be a waste of time, usually
# [06:44] * @bz will likely spend the next 3 days cursing android and debugging one single tree
# [06:45] <Jesse> it wouldn't be a waste of *your* time, though ;)
# [06:47] * Joins: mbrubeck (mbrubeck@moz-755AD63.hsd1.wa.comcast.net)
# [06:48] <@bz> well
# [06:48] <@bz> that depends
# [06:51] * Quits: juanb (jbecerra@moz-F1012875.hsd1.ca.comcast.net) (Quit: juanb)
# [06:53] * Joins: redwood (chatzilla@moz-3D9AA642.dial1.atlanta1.level3.net)
# [06:53] * Joins: Ami_Ty_ (Amie@moz-7D22EB0A.dsl.static.sonic.net)
# [06:53] * Quits: pcwalton (pcwalton@moz-5B0F92F1.hsd1.ca.comcast.net) (Quit: nbvcx)
# [06:54] * Joins: josh (josh@moz-75EC6735.dlth.qwest.net)
# [06:56] * Joins: janv (varga@moz-C1261AFF.flarion.as5628.telecom.sk)
# [06:57] * Joins: magsout (magsout@moz-E559D13.fbx.proxad.net)
# [06:58] * Quits: asac (asac@moz-AF572D83.pppoe.wtnet.de) (Ping timeout)
# [07:04] * Quits: sewardj (sewardj@moz-C79D927B.cable.virginmedia.com) (Ping timeout)
# [07:05] * Quits: raccettura (raccettura@moz-660B8F4B.hsd1.nj.comcast.net) (Quit: raccettura)
# [07:07] * Joins: mike5w3c (MikeS@moz-DAFE1A45.tokyo.ocn.ne.jp)
# [07:07] * Quits: derf (derf@moz-4168F490.net) (Ping timeout)
# [07:08] * Joins: asac (asac@C761F36F.E4499D91.B3DDC20A.IP)
# [07:08] <blizzard> hmm
# [07:08] <blizzard> anyone know where I file bugs against pdf.js?
# [07:09] * Joins: Mossop (mossop@moz-66A39E88.dsl.dynamic.sonic.net)
# [07:09] <KWierso> blizzard: https://github.com/mozilla/pdf.js has an issue tracker set up
# [07:11] * @bz grumbles about tantek claiming things about "mozilla"
# [07:14] * Quits: smooney (smooney@moz-3C7A0050.hsd1.ca.comcast.net) (Quit: smooney)
# [07:15] * Quits: Mossop (mossop@moz-66A39E88.dsl.dynamic.sonic.net) (Ping timeout)
# [07:16] * Quits: alex (alex@86599527.E7FBFC77.AB369DBA.IP) (Client exited)
# [07:17] * Quits: mcote|afk (mcote@moz-7CCC926E.mc.videotron.ca) (Ping timeout)
# [07:18] * Joins: biesi (cbiesinger@moz-5EE692A7.ca.hfc.comcastbusiness.net)
# [07:18] <blizzard> bz: :(
# [07:19] <@bz> blizzard: it's like he's _trying_ to force our hand here
# [07:19] <@bz> blizzard: and I don't understand why
# [07:20] <nigelb> bz: where is this?
# [07:21] <@bz> nigelb: whatwg
# [07:21] <nigelb> AH.
# [07:21] <nigelb> *reads*
# [07:22] <nigelb> wait a minute. Doesn't he work at Mozilla? :/
# [07:22] * Quits: gkw (gkw@moz-104CC309.mv.mozilla.com) (Quit: Instantbird 1.1)
# [07:22] <@bz> yes
# [07:22] <@bz> so?
# [07:22] * Joins: jamesr (jamesr@moz-C40B3BE3.hfc.comcastbusiness.net)
# [07:22] <nigelb> for a while I thought it was someone from the outside.
# [07:23] * @bz shrugs
# [07:23] <@bz> we have lots of people working for us
# [07:24] <nigelb> Also. I saw what you mean.
# [07:25] <Jef91> Does this compiler error make sense to anyone? Building on Debian ARM - http://paste.debian.net/151634/
# [07:26] <redwood> if i submit a crash report and include my email, does my email show up in any public-facing list?
# [07:26] <nigelb> No
# [07:27] <nigelb> It goes on the web UI for the crash reporter and only admins can see it.
# [07:27] <nigelb> This is so that they can get in touch with you to reproduce the crash or something like that.
# [07:28] <redwood> so none of the crashes you see by following links from about:crashes have that email info?
# [07:28] <glob> redwood, see http://www.mozilla.org/en-US/legal/privacy/firefox.html "crash-reporting feature"
# [07:28] <redwood> kk
# [07:28] <nigelb> thanks glob :)
# [07:28] <glob> although it states that it is "not applicable to firefox for mobile" .. i thought we had crash reporting there
# [07:29] <nigelb> maybe that poliy needs updating
# [07:29] <nigelb> *policy
# [07:30] * glob files a bug
# [07:33] * Joins: Mossop (mossop@moz-66A39E88.dsl.dynamic.sonic.net)
# [07:33] <Jef91> Does this compiler error make sense to anyone? Building on Debian ARM - http://paste.debian.net/151634/
# [07:34] * Joins: pnemsak (Miranda@moz-3D67D819.rainside.sk)
# [07:36] * Quits: yuan (ywang@moz-6A0DF30E.hsd1.ca.comcast.net) (Quit: yuan)
# [07:36] * Joins: bretr (bret_recka@moz-EC82158.hsd1.ca.comcast.net)
# [07:42] * Joins: darktrojan (geoff@moz-57EC95C5.dsl.telstraclear.net)
# [07:42] <philor> bz: um, which one of the things you pushed to inbound broke http://mxr.mozilla.org/mozilla-central/source/layout/reftests/bugs/478811-2.html?force=1 ?
# [07:43] <@bz> philor: looking
# [07:43] * philor would find it a bit more explicable if the answer was "neither one"
# [07:43] * Parts: zandr (zandr@moz-891BD824.milewski.org)
# [07:44] <@bz> philor: grrr
# [07:44] <@bz> philor: the mousemove thing
# [07:44] <@bz> philor: but it's random
# [07:45] <@bz> philor: that test run on that arch was green on try!
# [07:45] * @bz looks at test
# [07:45] * Joins: ehsan (ehsan@F0B20A8D.8458880F.57F33CED.IP)
# [07:46] <redwood> thanks glob and nigelb
# [07:46] * Quits: redwood (chatzilla@moz-3D9AA642.dial1.atlanta1.level3.net) (Quit: ChatZilla 0.9.88-rdmsoft [XULRunner 1.9.2/20100222071121])
# [07:46] <@bz> erm
# [07:46] <@bz> those are antialiasing differences around the edge?
# [07:46] <@bz> wtf?
# [07:47] <philor> thus "if the answer was 'neither one'"
# [07:47] <@bz> philor: layout/reftests/bugs/478811-4.html was orange on my try push on linux debug
# [07:48] <@bz> philor: with a similar issue: antialiasing differences along the edge
# [07:49] <@bz> I'm tempted to call it "neither one" for now...
# [07:49] * Quits: dRdR (dRdR@moz-792B0937.uwaterloo.ca) (Client exited)
# [07:50] <philor> um
# [07:51] * Quits: ehsan (ehsan@F0B20A8D.8458880F.57F33CED.IP) (Input/output error)
# [07:52] * Quits: bsmith (bsmith@moz-C03D0C61.vlan426.asr1.sfo1.gblx.net) (Ping timeout)
# [07:52] * Quits: jduell (jduell@moz-F20EC42A.hfc.comcastbusiness.net) (Ping timeout)
# [07:52] <philor> that's quite a bit of coincidence
# [07:52] * Joins: Mnyromyr (MnyroWork@moz-E2E3FF3D.tal.de)
# [07:53] <blassey> so... if you google for "dislocated hip" and then click on "news" it'll just say "(function(){})();(function(){})();"
# [07:53] <blassey> something weird going on there
# [07:54] <KWierso> works for me, blassey
# [07:54] <Unfocused> wfm too
# [07:54] <blassey> nightly on OSX here
# [07:54] <blassey> happens every time for me
# [07:54] <KWierso> oh, wait, clicking "News" just took me to the news.google.com homepage
# [07:55] <blassey> oh, I meant the "news" on the left side
# [07:55] * Quits: josh (josh@moz-75EC6735.dlth.qwest.net) (Quit: josh)
# [07:55] <KWierso> nope, clicking it from the sidebar works for me as well
# [07:55] <@bz> philor: yeah, I dunnot that I buy it. :(
# [07:55] * Joins: damons (gnubeard@moz-A41E6911.hsd1.ca.comcast.net)
# [07:55] <glob> blassey, wfm, nightly/osx
# [07:55] * Unfocused blames an addon
# [07:56] <blassey> just about:telemetry and bugzilla tweaks
# [07:56] <glob> blassey, what do you use from tweaks?
# [07:56] * Joins: RemusPop (remuspop@601F3B17.33662590.A5830293.IP)
# [07:56] <blassey> tbh... I don't think bugzilla tweaks is working anymore
# [07:57] * Quits: zuzelvp (zuzelvp@2112147D.C3507A2D.9A8C35B4.IP) (Ping timeout)
# [07:57] <philor> "what did you use when it worked" would probably be a better question
# [07:57] * Joins: derf (derf@moz-4168F490.net)
# [07:58] <blassey> restarted and it works now
# [07:59] * Quits: firebot (firebot@moz-67BF9FE6.carolina.res.rr.com) (Ping timeout)
# [07:59] <@bz> hrm
# [07:59] <@bz> http://mxr.mozilla.org/mozilla-central/source/widget/tests/test_native_mouse_mac.xul might start having random fails too...
# [08:00] * Quits: pnemsak (Miranda@moz-3D67D819.rainside.sk) (Quit: pnemsak)
# [08:00] * Joins: bsmith (bsmith@moz-C03D0C61.vlan426.asr1.sfo1.gblx.net)
# [08:00] * @bz should probably back out or something
# [08:00] <@bz> of course that was _already_ an intermittent orange
# [08:00] <@bz> I might just make it happen more often
# [08:00] <@bz> because the code is fundamentally broken
# [08:00] <@bz> (the test code)
# [08:01] <philor> and overly dependent on resolution
# [08:01] <@bz> that part worries me less
# [08:01] <@bz> the broken crap is this:
# [08:01] <@bz> 149 utils.sendNativeMouseEvent(x, y, msg, flags || 0, elem);
# [08:01] <@bz> 150 gAfterLoopExecution = setTimeout(function () {
# [08:01] <@bz> 151 clearExpectedEvents();
# [08:01] <@bz> 152 callback();
# [08:01] <@bz> 153 }, 0);
# [08:02] <@bz> which assumes that the timeout will run after event listeners for the event
# [08:02] <@bz> which may not be the case
# [08:02] <@bz> (and various other similar things)
# [08:04] * Joins: josh (josh@moz-75EC6735.dlth.qwest.net)
# [08:04] <@bz> ok
# [08:04] <@bz> I really need to sleep
# [08:04] <philor> "remote changed build/mobile/robocop/FennecNativeAssert.java.in which local deleted" really isn't my favorite hg question
# [08:04] <@bz> philor: are you willing to just back me out if things are too orange?
# [08:04] <@bz> philor: or should I just back out now?
# [08:04] * Quits: ctopper (craig@moz-42075866.hsd1.or.comcast.net) (Quit: ctopper)
# [08:05] * @bz hates events
# [08:05] <philor> bz: trying, but updating a tree to back you out because they already are too orange is asking me all these questions...
# [08:05] * Joins: MarcoZ (marco.zehe@moz-97ABA480.dip.t-dialin.net)
# [08:05] <@bz> ah
# [08:05] <@bz> ok, I'll just back out
# [08:05] <@bz> one sec
# [08:07] <@bz> backed out
# [08:09] <philor> thx
# [08:11] * glob is now known as glob|away
# [08:12] * Quits: ewong (chatzilla@moz-E5D50C2E.static.netvigator.com) (Quit: Reboot)
# [08:12] * Quits: asac (asac@C761F36F.E4499D91.B3DDC20A.IP) (Ping timeout)
# [08:12] * Joins: asac (asac@moz-E7708FD5.pppoe.wtnet.de)
# [08:12] * Quits: biesi (cbiesinger@moz-5EE692A7.ca.hfc.comcastbusiness.net) (Ping timeout)
# [08:15] * philor sends that tree to the chipper
# [08:16] <darktrojan> heh
# [08:16] * Joins: anky (anky@95BED6FC.B421193.A3D1B221.IP)
# [08:16] * Quits: anky (anky@95BED6FC.B421193.A3D1B221.IP) (Quit: Leaving...)
# [08:19] * cjones is now known as cjones-brb
# [08:20] <darktrojan> Unfocused, ping
# [08:20] <Unfocused> pong
# [08:21] <darktrojan> should I just fix this detail view icon bug and let the squabble over its size continue?
# [08:21] * Joins: wolfiR (wolfiR@moz-B660BA51.dip0.t-ipconnect.de)
# [08:21] <Unfocused> yep
# [08:21] <darktrojan> k
# [08:22] <darktrojan> patch coming your way any minute now
# [08:22] <Unfocused> \o/
# [08:23] * Joins: firebot (firebot@moz-67BF9FE6.carolina.res.rr.com)
# [08:25] <Unfocused> darktrojan: and if you feel inclined to do so, feel free to take that other bug and fix it the way we originally meant to, before a certain person changed their mind
# [08:25] <darktrojan> heh
# [08:25] * Joins: Yoric (Yoric@moz-920DB13B.fbx.proxad.net)
# [08:25] <Unfocused> so i can so r+ that now
# [08:25] <darktrojan> fwiw, 48px icons in list view make no difference
# [08:25] * Quits: bnicholson (bnicholson@moz-D1AB96C1.hsd1.ca.comcast.net) (Input/output error)
# [08:25] <Unfocused> er, cos i can*
# [08:25] <darktrojan> I can't fix the other, there are no images
# [08:25] * Unfocused is having a typo day :\
# [08:25] <Unfocused> ah, yea :\
# [08:26] <Unfocused> i'm sure i could bribe shorlander into providing such icons
# [08:26] * Quits: Mossop (mossop@moz-66A39E88.dsl.dynamic.sonic.net) (Ping timeout)
# [08:27] * Quits: josh (josh@moz-75EC6735.dlth.qwest.net) (Quit: josh)
# [08:27] * Joins: ewong (chatzilla@moz-E5D50C2E.static.netvigator.com)
# [08:27] <darktrojan> is that attachment really what the os x puzzle piece looks like? 'sugly :(
# [08:28] * Joins: Mitch (chatzilla@moz-FA99DC5F.sbr802.nsw.optusnet.com.au)
# [08:30] <Unfocused> hm, yea, wonder if all the icons could do with an update - the general in-content style is a more muted now days (and will be more-so with the upcoming changes)
# [08:34] * Joins: brendan (brendaneic@moz-A286C218.hsd1.ca.comcast.net)
# [08:35] * Joins: ctopper (craig@C3495DA.BA3DBA56.AE2B2F80.IP)
# [08:39] * Joins: Enn (enn@moz-DB6467E3.cpe.net.cable.rogers.com)
# [08:41] * Quits: bsmith (bsmith@moz-C03D0C61.vlan426.asr1.sfo1.gblx.net) (Ping timeout)
# [08:42] * Quits: Mook (mook@5365CE80.16C74E88.6F478678.IP) (Quit: gone)
# [08:44] * fabrice is now known as fabrice|zZz
# [08:46] * Quits: se (DS@moz-B62ECC33.wi.res.rr.com) (Ping timeout)
# [08:46] * Joins: pnemsak (Miranda@moz-3D67D819.rainside.sk)
# [08:46] * Joins: se (DS@moz-B62ECC33.wi.res.rr.com)
# [08:49] * Quits: damons (gnubeard@moz-A41E6911.hsd1.ca.comcast.net) (Quit: damons)
# [08:49] * Quits: brendan (brendaneic@moz-A286C218.hsd1.ca.comcast.net) (Quit: brendan)
# [08:53] * Quits: anant (Anant@moz-271479F2.hsd1.ca.comcast.net) (Quit: Leaving)
# [08:54] * cjones-brb is now known as cjones
# [08:56] * Joins: jhk (jiggy@8E6C34C1.A3F9767A.1C37C358.IP)
# [08:57] * Joins: pascalc (chatzilla@moz-DDE14210.rev.sfr.net)
# [09:00] * Joins: whimboo (whimboo@moz-216F9AFA.superkabel.de)
# [09:01] * Quits: pascalc (chatzilla@moz-DDE14210.rev.sfr.net) (Ping timeout)
# [09:04] * philor is now known as philor|away
# [09:04] * darktrojan is looking at areweprettyyet
# [09:04] <darktrojan> answer: no
# [09:04] * Joins: waschtl (waschtl@moz-A4ECE553.hsi4.kabel-badenwuerttemberg.de)
# [09:05] <larfdesk> Bug 650180
# [09:06] * Joins: jhorak (jhorak@moz-107AD163.redhat.com)
# [09:10] * Quits: bretr (bret_recka@moz-EC82158.hsd1.ca.comcast.net) (Quit: bretr)
# [09:10] * Joins: micahg (micahg@moz-B376CDB0.business.broadband.hu)
# [09:10] * Quits: micahg (micahg@moz-B376CDB0.business.broadband.hu) (Client exited)
# [09:10] * Joins: micahg (micahg@moz-B376CDB0.business.broadband.hu)
# [09:11] * Joins: mcote|afk (mcote@moz-A8369C00.mc.videotron.ca)
# [09:12] <MarcoZ> Good morning! Does anybody have the bug number handy for the silent updates without UAC on Windows? This was supposed to have landed on central already, but I'm still getting UAC dialogs.
# [09:13] * Joins: chrisccoulson (chr1s@moz-B376CDB0.business.broadband.hu)
# [09:13] <KWierso> MarcoZ: I can get the bug number in a second, but it only works for signed builds
# [09:13] <KWierso> if you're using x64 builds, those aren't signed
# [09:13] <darktrojan> and not for 64bit
# [09:13] <MarcoZ> KWierso: I'm using 32 bit nightlies.
# [09:14] <Unfocused> 64bit builds are people too
# [09:14] <MarcoZ> But on Win7 X64.
# [09:14] * glob|away is now known as glob
# [09:14] <KWierso> MarcoZ: https://bugzilla.mozilla.org/show_bug.cgi?id=481815
# [09:14] <KWierso> I think there's a few other conditions that have to be met
# [09:15] <KWierso> but do you see the MozillaMaintenance service in the services list in Task Manager?
# [09:18] * Joins: Ms2ger (Ms2ger@C99CCEB1.C1534583.37724B0D.IP)
# [09:18] * Quits: nattokirai (nattokirai@moz-348F61F0.mozilla.or.jp) (Quit: nattokirai)
# [09:20] * Quits: Enn (enn@moz-DB6467E3.cpe.net.cable.rogers.com) (Ping timeout)
# [09:20] * Joins: pascalc (chatzilla@moz-8F2953B2.rev.sfr.net)
# [09:21] * Quits: Stan (Stan@moz-9D59757B.dip.t-dialin.net) (Ping timeout)
# [09:22] <KWierso> MarcoZ: bug 481815 comment 483 has the conditions for when the service will be installed
# [09:24] * Quits: jhk (jiggy@8E6C34C1.A3F9767A.1C37C358.IP) (Ping timeout)
# [09:25] * Joins: Mano (chatzilla@moz-7E1DCE13.red.bezeqint.net)
# [09:25] * Parts: whimboo (whimboo@moz-216F9AFA.superkabel.de)
# [09:26] * Quits: crowder (crowder@moz-6274D06D.sd.sd.cox.net) (Ping timeout)
# [09:29] <MarcoZ> KWierso: I have the reg key mentioned, but the service is not installed. I have the output for sc query MozillaMaintenance captured.
# [09:30] <MarcoZ> It has some form of exit code.
# [09:30] * Joins: roc (chatzilla@C0ACF8B.5E1E9EEA.613E47D1.IP)
# [09:30] <MarcoZ> And the update process is always run as administrator, and my account has those privileges if I tell UAC so.
# [09:31] * Joins: jhk (jiggy@8E6C34C1.A3F9767A.1C37C358.IP)
# [09:31] * Joins: crowder (crowder@moz-6274D06D.sd.sd.cox.net)
# [09:32] * Quits: jhk (jiggy@8E6C34C1.A3F9767A.1C37C358.IP) (Connection reset by peer)
# [09:32] * Joins: imphil (philipp@moz-655EF802.customer.m-online.net)
# [09:33] * Joins: jhk (jiggy@8E6C34C1.A3F9767A.1C37C358.IP)
# [09:37] <dolske> Jef91: i'd suspect your arm arch isn't currently supported. http://mxr.mozilla.org/mozilla-central/source/xpcom/glue/arm.h
# [09:37] * Joins: protz (jonathan@moz-7F6750F6.xulforum.org)
# [09:38] <dolske> oh, hey, I'm way in backscroll. But i don't think anyone else answered?
# [09:39] * Joins: gandalf (zbraniecki@moz-66C04999.neoplus.adsl.tpnet.pl)
# [09:39] <NeilAway> bah, why is download manager using an nsCOMArray of a concrete type?
# [09:40] <Jef91> oh bleh that sucks dolske
# [09:40] <Jef91> Any way to get around that?
# [09:40] <Jef91> Do you know if there are any pre-built packages for Debian ARM?
# [09:40] <dolske> i've no idea. maybe ask the #mobile folks?
# [09:42] * NeilAway thwaps sdwilsh
# [09:42] * NeilAway thwaps biesi
# [09:42] * NeilAway thwaps mconnor
# [09:45] * Quits: ctopper (craig@C3495DA.BA3DBA56.AE2B2F80.IP) (Quit: ctopper)
# [09:45] * Joins: dao (dao@moz-C1C7E228.superkabel.de)
# [09:48] * Quits: chrisccoulson (chr1s@moz-B376CDB0.business.broadband.hu) (Ping timeout)
# [09:49] <dolske> yes, that sounds productive.
# [09:50] <dolske> maybe thwap beng just for good measure?
# [09:52] * Quits: janv (varga@moz-C1261AFF.flarion.as5628.telecom.sk) (Quit: This computer has gone to sleep)
# [09:53] * Joins: Enn (enn@moz-DB6467E3.cpe.net.cable.rogers.com)
# [09:55] * Joins: janv (varga@moz-C1261AFF.flarion.as5628.telecom.sk)
# [09:56] * Quits: Enn (enn@moz-DB6467E3.cpe.net.cable.rogers.com) (Ping timeout)
# [09:57] * Quits: m_kato (m_kato@moz-348F61F0.mozilla.or.jp) (Quit: Leaving...)
# [10:06] * Joins: mak (chatzilla@moz-C36E0FF8.retail.telecomitalia.it)
# [10:07] * Joins: past (past@moz-2DAAD734.dsl.dyn.forthnet.gr)
# [10:08] * Joins: daim (David_Mart@779E3E00.1773D26C.C0FF2207.IP)
# [10:10] * Joins: robhawkes (robhawkes@moz-33A339B7.dsl.cnl.uk.net)
# [10:16] * Joins: gabor (gabor@moz-3B57BCD1.catv.pool.telekom.hu)
# [10:16] <NeilAway> dolske: sadly this code postdates him
# [10:17] * Joins: necolas (necolas@moz-C8C6D8F3.bb.sky.com)
# [10:18] * Quits: dao (dao@moz-C1C7E228.superkabel.de) (Quit: Leaving.)
# [10:21] <Ms2ger> Hrm, I seem to have broken el.className
# [10:21] * Joins: Standard8 (Standard8@B7F1AE36.48015583.54C3481B.IP)
# [10:21] * Quits: cpearce (chatzilla@moz-56BC07D4.xdsl.xnet.co.nz) (Ping timeout)
# [10:22] * Quits: sgimeno (santiago.g@A922400D.48362F18.BCC03455.IP) (Client exited)
# [10:23] <Ms2ger> Ah, peterv's fault
# [10:23] * Quits: bholley (bholley@moz-FCAF9AAB.hsd1.ca.comcast.net) (Quit: bholley)
# [10:27] * Joins: sgimeno (santiago.g@A922400D.48362F18.BCC03455.IP)
# [10:27] * Quits: jgilbert_ (jgilbert@moz-BBE3ABD.mv.mozilla.com) (Ping timeout)
# [10:29] * Joins: smaug (chatzilla@moz-F779FD4F.elisa-mobile.fi)
# [10:29] * ChanServ sets mode: +o smaug
# [10:32] * glob is now known as glob|away
# [10:34] * Joins: msucan (msucan-@FA9E8863.56E67207.699550A1.IP)
# [10:39] <glandium> looks like i broken android again. and windows
# [10:40] <glandium> mmmm interestingly, i only broke xul android
# [10:43] * Joins: mastoris (mastoris@6C80CE82.6958D9C9.BF6BBFAF.IP)
# [10:45] * Joins: bsmith (bsmith@moz-E34A3D24.dsl.pltn13.sbcglobal.net)
# [10:45] * Quits: Bas (chatzilla@moz-B4DB3C59.ftth.concepts.nl) (Ping timeout)
# [10:46] <glandium> doh, i forgot to clobber
# [10:48] * Quits: jamesr (jamesr@moz-C40B3BE3.hfc.comcastbusiness.net) (Quit: jamesr)
# [10:49] * Quits: mastoris (mastoris@6C80CE82.6958D9C9.BF6BBFAF.IP) (Quit: Leaving)
# [10:49] * Joins: mastoris (mastoris@6C80CE82.6958D9C9.BF6BBFAF.IP)
# [10:49] * Joins: glazou (glazou@moz-204094DD.disruptive-innovations.fr)
# [10:49] <glazou> bonjour
# [10:50] * mattwoodrow is now known as mattwoodrow|away
# [10:50] * Parts: mastoris (mastoris@6C80CE82.6958D9C9.BF6BBFAF.IP) (Leaving)
# [10:50] * Joins: mastoris (mastoris@6C80CE82.6958D9C9.BF6BBFAF.IP)
# [10:51] * Quits: Mano (chatzilla@moz-7E1DCE13.red.bezeqint.net) (Input/output error)
# [10:51] * Joins: Bas (chatzilla@moz-B4DB3C59.ftth.concepts.nl)
# [10:53] * mattwoodrow|away is now known as mattwoodrow
# [10:53] <glandium> each time i use the clobberer, it's a PITA.
# [10:54] <glandium> the clobberer is also a nice example of a web app that makes firefox just stall completely
# [10:54] <nthomas|away> it doesn't like being drip fed an enormous table very much
# [10:55] <nthomas|away> if you know some php and want to make it better then we can set you up with help
# [10:55] <glandium> nthomas|away: even once the table is loaded, checking boxes is horrible
# [10:56] <glandium> seriously, can't we put someone from the webdev team on it?
# [10:56] <nthomas|away> IIRC it's just a straight form to POST
# [10:56] * Quits: karl (karl@moz-C7DA82DE.jetstream.xtra.co.nz) (Ping timeout)
# [10:56] <nthomas|away> probably can, they're looking developer tools more
# [11:02] * Joins: florian (florian@moz-87C33FDA.kimsufi.com)
# [11:03] * Joins: bjarne (bjarne@moz-46A85847.nextgentel.com)
# [11:06] * Joins: Mano (chatzilla@moz-7E1DCE13.red.bezeqint.net)
# [11:08] * Joins: Cwiiis (cwiiis@moz-F15E698.croy.cable.virginmedia.com)
# [11:09] * Joins: Archaeopteryx (itsme@moz-756328DB.cust.telecolumbus.net)
# [11:10] * Joins: edmorley (edmorley@moz-384CD16.range86-166.btcentralplus.com)
# [11:15] * Quits: stux (stux@moz-FD6B9EC6.hfc.comcastbusiness.net) (Ping timeout)
# [11:16] * Quits: ddahl (ddahl@moz-6971CF66.hsd1.il.comcast.net) (Ping timeout)
# [11:16] * Quits: Jef91 (Kristi@moz-FDB75EDD.hsd1.il.comcast.net) (Ping timeout)
# [11:16] * Joins: Jef91 (Kristi@moz-FDB75EDD.hsd1.il.comcast.net)
# [11:16] * Joins: stux (stux@moz-FD6B9EC6.hfc.comcastbusiness.net)
# [11:16] * Joins: ddahl (ddahl@moz-6971CF66.hsd1.il.comcast.net)
# [11:18] <darktrojan> well that's a different spam from the usual: "You'll wonder how you ever got along without learning UNIX"
# [11:18] * Quits: Archaeopteryx (itsme@moz-756328DB.cust.telecolumbus.net) (Quit: Too much information in my brain driving me insane)
# [11:19] * Quits: mijia (mijia@DC4232F0.766373FB.C3A57E70.IP) (Quit: mijia)
# [11:20] * Quits: pascalc (chatzilla@moz-8F2953B2.rev.sfr.net) (Ping timeout)
# [11:20] * Quits: regen (Miller@moz-DFCD66B7.adsl.dynamic.seed.net.tw) (Quit: regen)
# [11:20] * Joins: kaze (kaze@moz-A2A3C9E6.vlan402.asr1.cdg3.gblx.net)
# [11:21] * Quits: flx_ (flx@590D15DC.A7CE92B2.30E47D40.IP) (Ping timeout)
# [11:23] * Joins: gcp (gpascutto@moz-D0E475EA.access.telenet.be)
# [11:28] * ewong is now known as ewong|away
# [11:30] * Joins: dao (dao@moz-A2C27D22.dip.tu-dresden.de)
# [11:31] * Quits: imphil (philipp@moz-655EF802.customer.m-online.net) (Ping timeout)
# [11:32] * Joins: Archaeopteryx (itsme@moz-756328DB.cust.telecolumbus.net)
# [11:33] * Quits: MarcoZ (marco.zehe@moz-97ABA480.dip.t-dialin.net) (Ping timeout)
# [11:34] * Quits: jhk (jiggy@8E6C34C1.A3F9767A.1C37C358.IP) (Connection reset by peer)
# [11:36] * mattwoodrow is now known as mattwoodrow|away
# [11:40] * davehunt|away is now known as davehunt
# [11:41] * Quits: mak (chatzilla@moz-C36E0FF8.retail.telecomitalia.it) (Connection reset by peer)
# [11:41] * Joins: mak (chatzilla@moz-C36E0FF8.retail.telecomitalia.it)
# [11:44] * Joins: jacek (jacek@moz-5D707D3B.psi.wroc.pl)
# [11:47] * Joins: Goldorak (chatzilla@805CD605.E12797EB.187A1082.IP)
# [11:51] * mattwoodrow|away is now known as mattwoodrow
# [11:51] * Joins: jwatt (roslea@jwatt.irc.users.mozilla.org)
# [11:51] * Joins: rniwa_ (rniwa@moz-E171DA5.sfba.hfc.comcastbusiness.net)
# [11:51] * Quits: rniwa_ (rniwa@moz-E171DA5.sfba.hfc.comcastbusiness.net) (Quit: rniwa_)
# [11:52] * Joins: AutomatedTester (AutomatedT@moz-D4F73363.as13285.net)
# [11:52] * Quits: Mano (chatzilla@moz-7E1DCE13.red.bezeqint.net) (Ping timeout)
# [11:53] * Joins: tH (Rob@cpc16-seac19-2-0-cust234.7-2.cable.virginmedia.com)
# [11:53] * Quits: rniwa (rniwa@60A74940.D6CCE4AE.77834EAA.IP) (Ping timeout)
# [11:54] * Joins: Mano (chatzilla@moz-7E1DCE13.red.bezeqint.net)
# [11:55] <Mitch> darktrojan: Maybe if I learn UNIX I can use a mainframe!
# [11:55] <darktrojan> I don't know how I get along without it
# [11:56] <Mitch> freelance writer
# [11:59] * Joins: pranavrc (pranavrc@2282E225.2B2E4944.C28326FD.IP)
# [12:00] * Joins: karl (karl@moz-C7DA82DE.jetstream.xtra.co.nz)
# [12:01] <nigelb> Mitch++
# [12:02] * Quits: decoder (quassel@moz-216446B9.own-hero.net) (Ping timeout)
# [12:05] * Joins: decoder (quassel@moz-216446B9.own-hero.net)
# [12:10] * mattwoodrow is now known as mattwoodrow|away
# [12:14] * Joins: jhk (jiggy@8E6C34C1.A3F9767A.1C37C358.IP)
# [12:17] * Joins: pascalc (chatzilla@moz-A2A3C9E6.vlan402.asr1.cdg3.gblx.net)
# [12:17] * Joins: andreasn (andreasn@moz-D8A35C52.a199.priv.bahnhof.se)
# [12:17] * Quits: @smaug (chatzilla@moz-F779FD4F.elisa-mobile.fi) (Ping timeout)
# [12:19] * Quits: karl (karl@moz-C7DA82DE.jetstream.xtra.co.nz) (Ping timeout)
# [12:20] * Joins: clokep (clokep@moz-69FB3955.c3-0.smr-ubr1.sbo-smr.ma.cable.rcn.com)
# [12:21] <glazou> what's the correct channel to ask a question about partial MAR generation?
# [12:24] * Quits: andreasn (andreasn@moz-D8A35C52.a199.priv.bahnhof.se) (Ping timeout)
# [12:24] <darktrojan> that's probably more of a "who" than a "where" question
# [12:24] <glazou> anyone you could recommend then ?
# [12:25] <darktrojan> no idea :)
# [12:25] <glazou> thanks darktrojan :-)
# [12:25] <darktrojan> I try
# [12:26] <glazou> Chris Cooper is the author of https://wiki.mozilla.org/UpdateGeneration
# [12:26] * khuey|away is now known as khuey
# [12:27] <khuey> Ms2ger: how do I get to your house from CDG?
# [12:27] * bc|afk is now known as bc
# [12:28] <glazou> you're on this side of the atlantic khuey ??
# [12:28] <khuey> I am
# [12:28] <glazou> hey welcome :-)
# [12:28] <khuey> thanks
# [12:28] <khuey> it's pretty nice so far, they even made all the signs in english too for silly americans like me
# [12:29] <glazou> lol
# [12:29] <darktrojan> what's all this, Ms2ger takes visitors now?
# [12:29] <glazou> I think Ms2ger should even invite all local mozillians for a drink
# [12:29] <glazou> ;-)
# [12:29] * khuey supports this
# [12:31] * Quits: Yoric (Yoric@moz-920DB13B.fbx.proxad.net) (Input/output error)
# [12:31] * Quits: mak (chatzilla@moz-C36E0FF8.retail.telecomitalia.it) (Connection reset by peer)
# [12:32] <darktrojan> glazou, jhopkins|away might be of use to you by the look of this comment https://bugzilla.mozilla.org/show_bug.cgi?id=575317#c12
# [12:32] * glazou pings ccooper about MARs through twitter
# [12:32] <glazou> looking
# [12:33] * darktrojan has been waiting for that bug to be fixed for a long time :(
# [12:33] * khuey mumbles about annoying airport wifi
# [12:33] <glazou> khuey: CDG wifi gives you 15mins, then change your mac address to get extra 15mins :)
# [12:33] <darktrojan> probably better than no wifi
# [12:33] * Joins: ericb2 (X@moz-9C4C3DED.fbx.proxad.net)
# [12:34] <khuey> glazou: I found a hot spot that gives me six hours in exchange for my email address
# [12:34] <glazou> spamspamspam
# [12:34] <glazou> khuey: http://www.yopmail.com/en/ can be your friend :-)
# [12:35] <darktrojan> blackhole@...
# [12:35] <khuey> mailinator.com ;-)
# [12:35] <glazou> yep, similar
# [12:35] <glazou> but mailinator is well known and a lot of sites block it
# [12:35] <glazou> I mean they refuse such email addresses
# [12:35] <nigelb> smart
# [12:35] <khuey> yeah, I've seen that before
# [12:35] <khuey> worked here though
# [12:35] <nigelb> I've seen hotspots that want my phone number instead.
# [12:35] <nigelb> (they sms a key)
# [12:35] <glazou> wow
# [12:36] <khuey> heh
# [12:36] <glazou> I would never do that
# [12:36] <khuey> that certainly wouldn't work
# [12:36] <khuey> cdma phone ftl
# [12:36] * khuey gets on a vpn
# [12:36] * khuey is now known as khuey|away
# [12:36] <darktrojan> cdma! :-/
# [12:36] * darktrojan recalls selling cdma phones... 10 years ago
# [12:36] * khuey|away is now known as khuey
# [12:36] <khuey> boo
# [12:36] <khuey> they're blocking non-standard ports
# [12:38] * Quits: AutomatedTester (AutomatedT@moz-D4F73363.as13285.net) (Quit: AutomatedTester)
# [12:38] <gabor> mrbkap: do you have some time for a talk about wrappers I think the jetpack addons would need? not necessary now just some time in the near future...
# [12:38] <nigelb> khuey: always setup one machine with ssh on 443 for situations like these ;-)
# [12:39] <darktrojan> heh
# [12:39] <khuey> nigelb: yeah ... I need to convince IT to start running one
# [12:39] <nigelb> khuey: if you'd like, I can give you access to mine, as long as you don't do anything evil :)
# [12:40] <khuey> thanks for the offer
# [12:40] <khuey> I don't really need it though
# [12:40] * Joins: chrisccoulson (chr1s@moz-B376CDB0.business.broadband.hu)
# [12:40] <khuey> and I won't be here that long
# [12:41] <nigelb> aha
# [12:41] <nigelb> where are you headed?
# [12:42] <khuey> germany
# [12:42] <nigelb> Hrm, where are you usually located?
# [12:43] <nigelb> I assumed somewhere in europe based on your awake times
# [12:43] <khuey> east coast of the us
# [12:43] <khuey> i've just been a morning person the past few months
# [12:44] <nigelb> o_O
# [12:44] <nigelb> I never imagined you were in the US :)
# [12:45] * Quits: dao (dao@moz-A2C27D22.dip.tu-dresden.de) (Quit: Leaving.)
# [12:46] <khuey> well I'm not today :-P
# [12:55] * Joins: deLta30 (quassel@8E6C34C1.A3F9767A.1C37C358.IP)
# [12:58] * Joins: KaiRo (robert@moz-D2CAD9EE.adsl.highway.telekom.at)
# [13:00] <glandium> khuey: hey welcome on this side
# [13:00] <khuey> thanks
# [13:03] * mattwoodrow|away is now known as mattwoodrow
# [13:03] * Quits: Tobbi (Tobbi@16BAC97A.933EA279.AC7F8427.IP) (Ping timeout)
# [13:05] <glandium> apparently, osx is happy with a xmalloc request of 42GB
# [13:05] <glandium> (osx 64 bits)
# [13:06] * Quits: darktrojan (geoff@moz-57EC95C5.dsl.telstraclear.net) (Quit: darktrojan)
# [13:07] * Joins: dao (dao@moz-A2C27D22.dip.tu-dresden.de)
# [13:09] * Quits: clokep (clokep@moz-69FB3955.c3-0.smr-ubr1.sbo-smr.ma.cable.rcn.com) (Quit: Instantbird 1.2a1pre)
# [13:10] * Joins: zuzelvp (zuzelvp@2112147D.C3507A2D.9A8C35B4.IP)
# [13:11] * Quits: bsmith (bsmith@moz-E34A3D24.dsl.pltn13.sbcglobal.net) (Ping timeout)
# [13:11] * Quits: micahg (micahg@moz-B376CDB0.business.broadband.hu) (Ping timeout)
# [13:11] * Quits: chrisccoulson (chr1s@moz-B376CDB0.business.broadband.hu) (Ping timeout)
# [13:12] * khuey is now known as khuey|away
# [13:12] * Quits: deLta30 (quassel@8E6C34C1.A3F9767A.1C37C358.IP) (Ping timeout)
# [13:22] * Joins: graememcc (chatzilla@moz-3E78DB35.range86-148.btcentralplus.com)
# [13:24] * Quits: macmaN (chezburger@moz-DA23949E.dyn.estpak.ee) (Ping timeout)
# [13:26] * Quits: cjones (cjones@moz-45913895.socal.res.rr.com) (Quit: Leaving)
# [13:28] <hsivonen> :-( Firefox chrome code is observer-y, too. I thought just legacy Gecko code was.
# [13:31] * khuey|away is now known as khuey
# [13:34] <hsivonen> is dump(); supposed to work in nsBrowserGlue.js?
# [13:35] * Ziggy|AWAY is now known as Ziggy_Maes
# [13:36] * Joins: necolas_ (necolas@moz-C8C6D8F3.bb.sky.com)
# [13:36] * Quits: necolas (necolas@moz-C8C6D8F3.bb.sky.com) (Connection reset by peer)
# [13:36] * Joins: jfkthame_afk (jfkthame@43709DB9.5BC345F5.9542EC20.IP)
# [13:36] * jfkthame_afk is now known as jfkthame
# [13:36] <hsivonen> what's the deal with all the nsIObserver use instead of plain JS callbacks in chrome?
# [13:37] * mattwoodrow is now known as mattwoodrow|away
# [13:39] * Joins: IanN (iann@4837CDFE.2ECE621A.E5F36E28.IP)
# [13:40] * Joins: regen (Miller@moz-DFCD66B7.adsl.dynamic.seed.net.tw)
# [13:40] * Quits: regen (Miller@moz-DFCD66B7.adsl.dynamic.seed.net.tw) (Quit: regen)
# [13:40] * Quits: tH (Rob@cpc16-seac19-2-0-cust234.7-2.cable.virginmedia.com) (Ping timeout)
# [13:40] * Joins: macmaN (chezburger@moz-DA23949E.dyn.estpak.ee)
# [13:42] * Quits: glazou (glazou@moz-204094DD.disruptive-innovations.fr) (Quit: bbl)
# [13:43] * Joins: ewong_zzz (chatzilla@moz-FBF18AE4.static.netvigator.com)
# [13:43] * ewong_zzz is now known as ewong
# [13:45] * Joins: IRCMonkey41560 (Tobbi@16BAC97A.933EA279.AC7F8427.IP)
# [13:50] * Joins: mak (chatzilla@moz-C36E0FF8.retail.telecomitalia.it)
# [13:51] * Quits: GRMrGecko (GRMrGecko@moz-9EDAEFD9.client.mchsi.com) (Quit: Textual IRC Client: http://www.textualapp.com/)
# [13:51] * Joins: Yoric (Yoric@moz-A2A3C9E6.vlan402.asr1.cdg3.gblx.net)
# [13:55] * Quits: Wevah (Wevah@moz-97AD33CE.stcd.qwest.net) (Quit: HARR)
# [13:56] * Joins: raccettura (raccettura@moz-660B8F4B.hsd1.nj.comcast.net)
# [13:56] * Joins: tH (Rob@cpc16-seac19-2-0-cust234.7-2.cable.virginmedia.com)
# [13:59] <derf> khuey: The wonderful thing about hotspots that want your e-mail address is they can't actually mail you to confirm it's real and expect you to be able to check it before giving you access.
# [14:03] * Joins: regen (Miller@moz-A481CEE6.adsl.dynamic.seed.net.tw)
# [14:03] * khuey is now known as khuey|away
# [14:06] * Quits: jhk (jiggy@8E6C34C1.A3F9767A.1C37C358.IP) (Connection reset by peer)
# [14:07] * Joins: chrisccoulson (chr1s@moz-B376CDB0.business.broadband.hu)
# [14:07] * Quits: chrisccoulson (chr1s@moz-B376CDB0.business.broadband.hu) (Client exited)
# [14:07] * Joins: jrmuizel (jrmuizel@moz-20EF8EAA.cpe.net.cable.rogers.com)
# [14:07] * Joins: chrisccoulson (chr1s@moz-B376CDB0.business.broadband.hu)
# [14:08] * glob|away is now known as glob
# [14:09] * Joins: espindola (espindola@EFCDA10D.68E64C0A.F061A1E6.IP)
# [14:09] * Joins: micahg (micahg@moz-B376CDB0.business.broadband.hu)
# [14:09] <NeilAway> hsivonen: such as?
# [14:09] * Quits: micahg (micahg@moz-B376CDB0.business.broadband.hu) (Connection reset by peer)
# [14:10] <Mitch> khuey|away: You'll be able to see if Ms2ger was programmed from the machine world.
# [14:10] * Quits: zuzelvp (zuzelvp@2112147D.C3507A2D.9A8C35B4.IP) (Ping timeout)
# [14:12] <hsivonen> NeilAway: nsBrowserGlue.js for example observes a zillion topics instead of various APIs taking callbacks
# [14:13] <mrbkap> gabor: yeah, I do!
# [14:13] <mrbkap> gabor: alexandre was interested as well... but he is sick today.
# [14:14] <hsivonen> what mechanism is supposed to set up the bookmarks for browser/components/places/tests/browser/browser_library_infoBox.js ?
# [14:14] <gabor> mrbkap: well, I don't think we can get to the end of it at once anyway... so I thought we can start it and then later we talk about it again with alex as well
# [14:14] <hsivonen> the default import in nsBrowserGlue.js?
# [14:15] <gabor> mrbkap: I pretty much know what we need and have some idea about how we could do it... but ofc I have no clue if it's feasable or not and would be great if you could put some thoughts in it
# [14:16] <gabor> since you know a LOT more about wrappers and security than me obviously... so when do you have some time for it?
# [14:18] * Ziggy_Maes is now known as Ziggy|AWAY
# [14:18] * khuey|away is now known as khuey
# [14:21] <espindola> ted, ping on 715528
# [14:22] * bwinton_away is now known as bwinton
# [14:22] <espindola> dolske, ping on 714960
# [14:23] * bwinton is now known as bwinton_away
# [14:26] <mrbkap> gabor: pretty much any time during the day.
# [14:26] <hsivonen> browser/components/places/tests/browser/browser_library_infoBox.js confuses me
# [14:28] <hsivonen> debugging shows that it calls nsPlacesImportExportService::ImportHTMLFromURI
# [14:28] * Joins: Suresh (chatzilla@B4FB7175.505142EE.EB06F97B.IP)
# [14:28] <hsivonen> which is called once in the whole codebase in nsBrowserGlue.js
# [14:28] <ted> espindola: sorry, reviewed
# [14:28] <espindola> thanks
# [14:28] <hsivonen> but I don't see control reaching that line with my patch applied
# [14:28] * khuey is now known as khuey|away
# [14:28] <hsivonen> I wonder what I have managed to break
# [14:30] <gabor> mrbkap: how about 3pm? (37 minutes from now)
# [14:33] * Joins: Enn (enn@F2D29657.F60B0462.67AC9B1.IP)
# [14:34] <tbsaunde> so, is there a way I can make qpush do something more sain than consider the patch fully applied and the tree in a normal state if a file hhas conflicts?
# [14:34] <NeilAway> hsivonen: so, each API should have its own callback interface, rather than reusing the observer notification model?
# [14:35] * Quits: regen (Miller@moz-A481CEE6.adsl.dynamic.seed.net.tw) (Quit: regen)
# [14:36] <Ms2ger> khuey|away, North, and call me once you crossed a border
# [14:36] <NeilAway> hsivonen: plus, of course, everything needs to have its own callback functions, rather than reusing the observer service
# [14:36] <hsivonen> NeilAway: yeah. it seems sensible to give a callback to a thingy that does something asynchronous instead of having to add observers somewhere different and remembering to clean them up
# [14:39] <NeilAway> hsivonen: ok, so I looked in nsBrowserGlue.js and found no callbacks at all
# [14:40] * Joins: jhk (jiggy@8E6C34C1.A3F9767A.1C37C358.IP)
# [14:41] <glandium> espindola: I don't think you're helping the tenfourfox people with your push
# [14:41] * Quits: merinui (merinui@moz-61C7235E.osk2.eonet.ne.jp) (Quit: Leaving...)
# [14:42] <espindola> glandium, there is no 10.6 for ppc, so setting the variable in the cofigure will not make a big difference....
# [14:42] * Joins: mconley (mconley@moz-9C0F4EBF.cable.teksavvy.com)
# [14:43] * Quits: dao (dao@moz-A2C27D22.dip.tu-dresden.de) (Quit: Leaving.)
# [14:43] <espindola> policies (what os is target for example), should normally live on the mozconfig files imho. Specially now that they are in tree
# [14:43] <espindola> ted, ping 715930 too
# [14:44] <ted> getting there :)
# [14:44] <espindola> thanks
# [14:44] <Mano> hsivonen: remind me, your new import implementation - module or component?
# [14:46] * ted loves "add reviewer to CC list"
# [14:46] <hsivonen> Mano: JSM
# [14:47] <Mano> hsivonen: ah, so, please, just callbacks :)
# [14:47] <Mano> no observer overkill please
# [14:47] <Mano> semi-standard in toolkit.
# [14:47] <hsivonen> Mano: I support the legacy observer topics plus a callback for new code
# [14:48] <Mano> hsivonen: why is it needed?
# [14:48] <hsivonen> Mano: there's existing code that observes the topics
# [14:48] <Mano> the new module doesn't support the old iface somehow, right?
# [14:48] <Mano> in-tree/addons?
# [14:48] <hsivonen> Mano: it doesn't support the old entry point, no
# [14:49] * davehunt is now known as davehunt|away
# [14:49] <Mano> hsivonen: so no reason to support both
# [14:49] <Mano> just fix our callers
# [14:49] * Joins: jprmc (jprmc@moz-7F2FF3EB.cpe.net.cable.rogers.com)
# [14:49] <hsivonen> Mano: that's more code for me to rewrite :-(
# [14:49] <Mano> hsivonen: could you point me to users?
# [14:50] <Mano> if there's a valid use case for observer, we shouldn't do callbacks
# [14:50] <Ms2ger> ted, yeah, that's really nice
# [14:50] <Mano> if there's not, in-tree code should be fixed.
# [14:51] <hsivonen> Mano: http://mxr.mozilla.org/mozilla-central/search?string=bookmarks-restore-success
# [14:51] * Joins: joey (chatzilla@moz-EFCB4CBF.princetowncable.com)
# [14:51] <Ms2ger> Followup!
# [14:51] <Mano> hsivonen: not so much...
# [14:52] <hsivonen> Mano: the sync thing might be a valid use case
# [14:52] <Mano> looking at it now
# [14:53] <Mano> hsivonen: so, they try to find out if there's a "batch"
# [14:53] <Mano> right?
# [14:53] <Mano> if so, no, that's not a valid use case, because the bookmarks api will be async too sometime soon
# [14:54] <mrbkap> gabor: hey, yeah, that works.
# [14:55] <Ms2ger> 12 minutes....
# [14:55] <Mano> hsivonen: having said that, i think there's some way to listen to a a batch (mak: ping)
# [14:55] <mak> listen to a batch for what?
# [14:55] <Mano> mak http://mxr.mozilla.org/mozilla-central/source/services/sync/modules/engines/bookmarks.js#1281
# [14:55] <gabor> mrbkap: alright, shall we do vidyo?
# [14:56] <mak> Mano: yes sync uses the notifications to ignore imports
# [14:56] <Mano> mak: that's bad
# [14:56] <Mano> well, at least given the current notifications
# [14:57] <Mano> because they don't work for profile migrators
# [14:57] <mak> it wouldn't be needed if we'd export guids, but we don't
# [14:57] <Mano> or any other sort of a "batch"
# [14:57] <Mano> so
# [14:57] <Mano> any way to listen for when a batch starts
# [14:57] <Mano> and ends
# [14:57] * Mano is checking
# [14:57] <decoder> im seeing this assertion during make package on m-c: http://pastebin.mozilla.org/1443288
# [14:57] <mrbkap> gabor: yeah.
# [14:57] <decoder> known problem? havent found anything in bugzilla
# [14:58] <hsivonen> hmm. better console errors from broken chrome code cloud have saved me a lot of time today.
# [14:58] <mak> Mano: I don't understand what wouldn't work, the migrators import bookmarks from outside
# [14:58] <mak> Mano: what sync ignores is bookmarks restored to a previously tracked profile
# [14:59] * Joins: armenzg (armenzg@88D845D1.5BCEC6DB.DA78B690.IP)
# [14:59] <Mano> mak: ah, i see
# [14:59] <Mano> hsivonen: ok, so
# [15:00] <Mano> seems like you want the caller for import html to notify
# [15:00] * armenzg is now known as armenzg_walking
# [15:00] <Mano> and only in the case of restore
# [15:00] * Quits: armenzg_walking (armenzg@88D845D1.5BCEC6DB.DA78B690.IP) (Input/output error)
# [15:00] <Mano> not any-import
# [15:00] * khuey|away is now known as khuey
# [15:00] <Mano> meaning you won't have to change sync code, for now
# [15:01] <Mano> mak: and when are guids happening?
# [15:01] <mak> Mano: btw, just in case you can ping rnewman to double check... I don't remember exactly if we care about imports for perf reasons
# [15:01] <khuey> Ms2ger: can I have a phone number?
# [15:01] <mak> Mano: guids are automatically added on entries creation
# [15:01] <Ms2ger> Can I have yours? :)
# [15:01] <Mano> mak: so why aren't they used here?
# [15:02] <khuey> Ms2ger: my US cell #?
# [15:02] <khuey> sure
# [15:02] <khuey> it won't do you much good for the next few weeks though
# [15:02] <mak> Mano: because nobody had the time to add guids to html and json exports
# [15:02] <Ms2ger> Rather pointless, then
# [15:02] <Mano> mak: i see
# [15:02] * Quits: necolas_ (necolas@moz-C8C6D8F3.bb.sky.com) (Connection reset by peer)
# [15:03] <khuey> Ms2ger: cdma ftl
# [15:03] <Mano> hsivonen: ok, so, just find which import call is for restore, fix it to notify, and i think you're ok.
# [15:03] <mak> Mano: btw, I still don't get what's the problem in porting the cpp code to js related to these notifications...
# [15:03] * Joins: necolas (necolas@moz-C8C6D8F3.bb.sky.com)
# [15:03] <Mano> mak: we're adding a callback
# [15:03] <mak> ok, though sync cannot listen to that callback
# [15:03] <Ms2ger> Never heard of
# [15:03] <Mano> mak: but sync only cares for restore
# [15:03] <Mano> not any import
# [15:04] <Mano> not really the same
# [15:04] <mak> is bookmarks-restore-success sent for any kind of import?
# [15:05] <mak> the engine is wiped for bookmarks-restore-success, so it depends when we fire that
# [15:05] <mak> (I may have mirecalled, I don't look at that code from some time)
# [15:06] <Mano> mak: http://mxr.mozilla.org/mozilla-central/source/toolkit/components/places/nsPlacesImportExportService.cpp#2087
# [15:06] * rail_away is now known as rail
# [15:06] <Mano> apparently, yes.
# [15:06] <glandium> wtf https://tbpl.mozilla.org/php/getParsedLog.php?id=8444382&tree=Try
# [15:07] * Joins: imphil (philipp@moz-655EF802.customer.m-online.net)
# [15:07] <mak> Mano: ok then sync cares about any kind of import, though I think it should handle "html-initial" but not "html"
# [15:07] <mak> since the initial one replaces, the not initial one just adds
# [15:07] <khuey> gcc is on strike
# [15:07] <Mano> mak: i don't think this makes sense at all
# [15:07] <Ms2ger> Use clang
# [15:08] <Mano> mak: there's few kinds of batches (1) html initial (2) other-browser initial (3) restore (4) other-browser, not initial
# [15:08] <mak> Mano: I think the scope was to avoid duplicates and useless work, so on import/restore all remote content is wiped to be replaced by the new import
# [15:08] <mak> Mano: but if the new import does not replace, is probably overkill
# [15:08] <Mano> mak: that doesn't work for migrators.
# [15:09] <Mano> so the #1 overkill was never fixed
# [15:09] <mak> but that initial happens on an empty profile
# [15:10] * Joins: ehsan (ehsan@F0B20A8D.8458880F.57F33CED.IP)
# [15:10] * Quits: mconley (mconley@moz-9C0F4EBF.cable.teksavvy.com) (Input/output error)
# [15:10] <Mano> mak: wait, is sync already hooked up on profile creation?
# [15:10] <mak> no, not yet
# [15:10] <Mano> will it?
# [15:10] <mak> there was discussion on providing a "migrate from sync"
# [15:11] <mak> to setup a new profile
# [15:11] <Mano> mak: but that deosn't involve html-import
# [15:11] <mak> hm, nop
# [15:11] <Mano> so, html-initial isn't a case either
# [15:12] <Mano> for sync, that is.
# [15:12] <mak> right, excluding addons
# [15:13] <Mano> we're left with html-restore, json-restore, html-not-initial, and migrators
# [15:13] * Joins: davidb (davidb@F2D29657.F60B0462.67AC9B1.IP)
# [15:13] <Mano> current sync only takes care of the first
# [15:13] <Mano> and html-not-inital
# [15:14] <Mano> i suggest fixing it to work just for html-restore, for now, and figure out a better solution later, after the bookmarks api becomes async
# [15:15] <mak> I still don't understand why we can't continue doing what we do for now, and figure it out later... firing a couple topics is not hard...
# [15:15] * Quits: jhk (jiggy@8E6C34C1.A3F9767A.1C37C358.IP) (Ping timeout)
# [15:15] <Mano> mak: it's a new api
# [15:15] * catlee-away is now known as catlee
# [15:16] <Mano> no reason to make it use deprecated stuff on v0
# [15:16] <Mano> unless there's a good use case. apparently there isn't
# [15:16] <Mano> mak: i do agree we should call this topics, but just for restore, by the caller.
# [15:17] <mak> it may be fine, but we need feedback from Sync
# [15:17] <Mano> right
# [15:17] <mak> cause I don't know the server side enough
# [15:17] * Joins: Pike (Pike@moz-269673CD.pool.mediaways.net)
# [15:17] <Mano> mak: if sync does it for perf, they need to take care of json imports too
# [15:18] <Mano> restore*
# [15:18] <mak> json fires those notifications
# [15:18] <mak> with "json" data
# [15:18] <mak> iirc
# [15:18] <mak> you can check the buig that added those for more info
# [15:19] * Joins: AaronMT (AaronMT@moz-DB17C53A.cpe.net.cable.rogers.com)
# [15:19] <hsivonen> mak: I'd be happy to emulate the old notifications and make removing them Someone Else's Problem
# [15:19] <mak> that's what I was trying to suggest
# [15:20] * Joins: glazou (glazou@moz-204094DD.disruptive-innovations.fr)
# [15:20] <Mano> hsivonen: fine, but please file a bug and document it
# [15:20] <hsivonen> Mano: ok
# [15:23] * Quits: Suresh (chatzilla@B4FB7175.505142EE.EB06F97B.IP) (Connection reset by peer)
# [15:23] * Quits: Wes-- (chatzilla@moz-BEF0C255.page.ca) (Client exited)
# [15:23] * Quits: Mano (chatzilla@moz-7E1DCE13.red.bezeqint.net) (Ping timeout)
# [15:23] * jfkthame is now known as jfkthame_afk
# [15:23] * Joins: ajuma (ajuma@F2D29657.F60B0462.67AC9B1.IP)
# [15:24] * bwinton_away is now known as bwinton
# [15:24] * Joins: romeo (romeo@moz-18AD2F99.k744.webspeed.dk)
# [15:25] <ewong> when I do a |make -f client.mk| where does it detect the |--enable-application=browser|?
# [15:28] <Standard8> in configure
# [15:28] * Joins: beaufour (beaufour@18D5CC88.C7EE4FB2.ECED8BE3.IP)
# [15:29] <ewong> configure.in or just configure?
# [15:29] * Joins: jlebar (jlebar@moz-3F3A6302.dyn.columbia.edu)
# [15:30] * coop|afk is now known as coop
# [15:30] <ewong> Standard8 ooh.. just the person re: ccrework.. is there anything I can help with?
# [15:31] * ctalbert|afk is now known as ctalbert
# [15:31] <ewong> of course, I'll need to learn a great deal about buildconfig
# [15:31] <Standard8> ewong: I'd only recommend it if you've got good knowledge about buildconfig... its the complicated bits we need now
# [15:32] <ewong> hrm.. nope.. just starting :)
# [15:32] * Joins: Mano (Mano@BCB6104A.CF830E9.C818C838.IP)
# [15:34] * jmaher|afk is now known as jmaher
# [15:37] * khuey is now known as khuey|away
# [15:38] * Joins: armenzg (armenzg@F2D29657.F60B0462.67AC9B1.IP)
# [15:39] * Joins: bbondy (bbondy@moz-28CF6D1C.home.cgocable.net)
# [15:39] <ted> this cuttherope.ie game doesn't load in my windows nightly
# [15:40] <mak> ted: wfm
# [15:40] * Joins: caillon (caillon@moz-359E39FC.lightspeed.sntcca.sbcglobal.net)
# [15:40] <ted> it shows a bunch of bubles with a 00% in one
# [15:40] <mak> well that 00 should proceed
# [15:40] <ted> it's not :-(
# [15:40] <mak> it took a while but worked here
# [15:41] <jlebar> bz, a window calls iframe.addMyCallback(function() { ... });. When I fire the callback, I need to push a context; is it the window's context? That means I need to hold on to a weak ref to the window? What about the callback itself -- does that also need to be a weak ref?
# [15:42] <ted> odd
# [15:42] <ted> i wonder if they're using flash and flashblock is killing it?
# [15:42] * Joins: dao (dao@moz-A2C27D22.dip.tu-dresden.de)
# [15:42] <ted> doesn't work in my linux nightly either (where i don't have flash installed)
# [15:43] <ted> yup
# [15:43] <ted> lame
# [15:44] * Joins: dRdR (dRdR@moz-3D8840A3.uwaterloo.ca)
# [15:44] <ewong> Standard8 will the comm extensions (i.e venkman, etc) be moved to mc/extensions?
# [15:44] <Standard8> ewong: unknown
# [15:45] <ewong> ok..
# [15:47] * Joins: lmandel (lmandel@7F36F6DF.1D61B048.412CF160.IP)
# [15:48] * Joins: newbie (kvirc@697AC6B3.8C6159FA.9A06DD32.IP)
# [15:49] * Joins: Suresh (chatzilla@50054847.E0A9EEE1.EB06F97B.IP)
# [15:50] * davehunt|away is now known as davehunt
# [15:51] * Joins: MarcoZ (marco.zehe@moz-97ABA480.dip.t-dialin.net)
# [15:51] <glob> ted, doesn't work for me either in nightly
# [15:52] <ted> glob: you need flash installed (and not flashblock)
# [15:52] <espindola> are we having network problems on try?
# [15:52] <espindola> https://tbpl.mozilla.org/php/getParsedLog.php?id=8445133&tree=Try&full=1
# [15:52] <ted> or whitelist it in flashblock
# [15:52] * Quits: Archaeopteryx (itsme@moz-756328DB.cust.telecolumbus.net) (Quit: Too much information in my brain driving me insane)
# [15:52] <espindola> Temporary failure in name resolution
# [15:52] <espindola> :-(
# [15:52] <glob> ted, ah, for sound
# [15:52] * Joins: jimm (jmathies@moz-7F164CA1.pn.at.cox.net)
# [15:53] <glob> it's much smoother on nightly than chrome :D
# [15:53] <ted> nice :)
# [15:53] * Joins: kats (kats@moz-DE13FC16.compute-1.amazonaws.com)
# [15:53] <ted> glob: lame that it couldn't fail gracefully there without flash...
# [15:53] * Quits: lmandel (lmandel@7F36F6DF.1D61B048.412CF160.IP) (Quit: lmandel)
# [15:53] <glob> ted, did you read the comment about that in the source?
# [15:53] <ted> no
# [15:53] * Parts: kats (kats@moz-DE13FC16.compute-1.amazonaws.com)
# [15:54] <glob> // html5 audio is unreliable in many browsers so its only enabled by default on
# [15:54] <glob> // IE9 or greater. You can force html5 audio by setting the querystring:
# [15:54] * Quits: Mano (Mano@BCB6104A.CF830E9.C818C838.IP) (Client exited)
# [15:54] <beltzner> espindola: http://status.mozilla.com/ looks OK, though I'm not sure if tryserver is reflected on any of those
# [15:54] <beltzner> espindola: oh, DNS, I see - try 8.8.4.4?
# [15:55] <beltzner> glob: status shows bugzilla as performing poorly for a week - I bet you know about that, but out of curiousity, what's up with that?
# [15:55] <beltzner> also: I'm in Perth again next week!
# [15:55] <glob> beltzner, DDoS
# [15:56] <beltzner> glob: whoa, really? just on bugzilla?
# [15:56] <glob> beltzner, yeah, just bmo
# [15:56] <beltzner> we should totally pack up the Internet and take it away from the kiddies
# [15:56] <jlebar> Ah, this explains why |time curl| suddenly went from 20s to 2s.
# [15:56] * Joins: mconley (mconley@F2D29657.F60B0462.67AC9B1.IP)
# [15:56] * Joins: jorendorff (jorendorff@moz-91590D94.hsd1.tn.comcast.net)
# [15:56] <jlebar> beltzner, I bet it's because of the version numbers. :-p
# [15:57] <reuben> ~rumors~ say it was accidental via twitter, someone posted a large dependency tree
# [15:58] <jlebar> and far be it from bugzilla to employ any kind of caching.
# [15:58] <glob> reuben, that happened at the same time, but there were definitely multiple DDoS's going on
# [15:58] <reuben> ouch
# [16:00] <glob> jlebar, totally agree with the lack of caching sentiment. lots of foundations to build before that can happen :(
# [16:00] <jlebar> glob, I'm actually really happy to hear that the slowness was due to an attack rather than total incompetence on our part. :)
# [16:01] * bhearsum|afk is now known as bhearsum
# [16:01] <bsmedberg> mats and others
# [16:01] <bsmedberg> the bug you want is not bug 646275
# [16:01] <glob> jlebar, i'm not sure which is worse :P
# [16:01] * Quits: graememcc (chatzilla@moz-3E78DB35.range86-148.btcentralplus.com) (Ping timeout)
# [16:01] * Joins: mburns (mburns@moz-2BF04693.mirwin.net)
# [16:02] * Quits: tonymec (tonymec@5B2A1B79.5A39B178.277517C1.IP) (Quit: going to bed)
# [16:02] <bsmedberg> is "vporof" on IRC?
# [16:02] <bsmedberg> mats: its bug 712032
# [16:02] * Quits: Mardak (Mardak@moz-4FA48382.hsd1.ca.comcast.net) (Quit: Mardak)
# [16:02] * Quits: @bz (bzbarsky@moz-69B5879F.bstnma.fios.verizon.net) (Ping timeout)
# [16:03] <jlebar> mounir, totally off topic for the bug: Why does asking whether we're on a metered connection or not (3g vs wifi) require privilege?
# [16:04] <mats> bsmedberg: ok, sorry for the spam
# [16:04] <bsmedberg> jlebar: because it can be used to profile users by affluence maybe?
# [16:05] * bsmedberg notes that 3g vs wifi isn't a very good metered-vs-notmetered metric anyway
# [16:05] <jlebar> bsmedberg, The network API has a way to directly read metered vs not metered.
# [16:06] <jlebar> I guess there's some fingerprinting entropy specifically in "not metered and 3g", and you might be able to detect 3g by other means.
# [16:06] <jlebar> That seems like a somewhat far-fetched attack, though.
# [16:06] * Quits: dao (dao@moz-A2C27D22.dip.tu-dresden.de) (Quit: Leaving.)
# [16:06] * Quits: Mnyromyr (MnyroWork@moz-E2E3FF3D.tal.de) (Input/output error)
# [16:06] <ewong> is there a command for make such that it dumps the line # of the configure during |make -f client.mk|?
# [16:07] <bsmedberg> ewong: the line # of the configure?
# [16:07] * mcote|afk is now known as mcote
# [16:07] <ewong> bsmedberg: yes
# [16:07] * Joins: smaug (chatzilla@moz-958D39B4.elisa-mobile.fi)
# [16:07] * ChanServ sets mode: +o smaug
# [16:07] <bsmedberg> the question needs to be rephrased...? You mean a configure error?
# [16:08] <ewong> bsmedberg: no.. not configure error.. I mean, for command that configure executes, it also displays the line #
# [16:08] <ewong> sorry.. umm lemme see if I can rephrase this.
# [16:09] * Joins: kaie (kaie@moz-FC6C3955.dip.t-dialin.net)
# [16:09] <ewong> i.e. 1021 - checking for java... /c/Windows/System32/java
# [16:09] <ewong> (1021 is a fake line #.. don't really know the line # of that check)
# [16:09] <bsmedberg> not really, you'd have to change autoconf itself
# [16:09] <bsmedberg> but config.log probably has most of the details
# [16:10] * jfkthame_afk is now known as jfkthame
# [16:10] <ewong> ahh right
# [16:11] <ewong> oh.. just noticed in the config.log... ./configure isn't executed from the top to the bottom.. it jumps around
# [16:11] <ted> configure is a shell script
# [16:11] <ted> it gets executed like a shell script...
# [16:11] <ted> glob: ick, that's terrible
# [16:11] <ted> glob: i mean, i know audio has tons of bugs
# [16:11] <ted> but i think it's probably pretty good in firefox
# [16:12] * Quits: magsout (magsout@moz-E559D13.fbx.proxad.net) (No route to host)
# [16:12] <ewong> ok.. that complicates things a bit..
# [16:12] * Joins: magsout (magsout@moz-E559D13.fbx.proxad.net)
# [16:12] <ted> view source is a little sad
# [16:12] <ted> html5 = "everything is a div"
# [16:12] <ewong> oh..
# [16:12] <ewong> smaug ping
# [16:13] <espindola> catlee, all the failures were on bm-xserve23
# [16:13] <espindola> can you check if it is broken?
# [16:14] <ted> glob: they went through the trouble to encode the intro video as ogg and everything
# [16:14] * Joins: jhk (jiggy@8E6C34C1.A3F9767A.1C37C358.IP)
# [16:14] <bsmedberg> is infra flaky? I got tryserver errors "abort: error: Temporary failure in name resolution"
# [16:14] <@smaug> ewong: pong
# [16:14] <ewong> smaug : re bug #693172, can you clarify what you mean by your last comment? I am to take v2 and just add the nsDOMEventTargetWrapperCache changes to nsDOMEventTargetHelper?
# [16:15] * IRCMonkey41560 is now known as Tobbi
# [16:15] <catlee> espindola: what are you talking about?
# [16:15] <espindola> catlee, https://tbpl.mozilla.org/?tree=Try&rev=ab3d0d91c852
# [16:15] <espindola> https://tbpl.mozilla.org/php/getParsedLog.php?id=8445412&tree=Try for example
# [16:16] <espindola> error: Temporary failure in name resolution
# [16:16] <@smaug> ewong: yes, take v2 and add the changes
# [16:16] <ewong> smaug ok.. I thought I did.. but I think I got a bit confused..
# [16:16] <@smaug> I want to avoid all the extra code moves from file to some other file since hg annotate is so bad
# [16:17] <@smaug> ewong: in your latest patch you move some stuff to nsDOMEventTargetWrapperCache
# [16:17] * Joins: askalski (akuda@moz-6A36EC49.ip.abpl.pl)
# [16:17] <@smaug> and then move back
# [16:17] <@smaug> or something like that
# [16:17] * @smaug re-reads the patch
# [16:17] <askalski> hsivonen: hi, you there?
# [16:17] <@smaug> ewong: the patch is adding new code to nsDOMEventTargetWrapperCache.h
# [16:18] <@smaug> and that is a file we should remove
# [16:18] <espindola> rail, would you mind putting the nasm rpms we use on the bots somewhere?
# [16:18] * Joins: rwaldron (rwaldron@moz-BDCCF091.hfc.comcastbusiness.net)
# [16:18] <espindola> s/nasm/yasm/
# [16:18] <ewong> ahhh.. yes.. got confused there.. thanks for the review! will try to get it fixed soon..
# [16:18] <rail> espindola: sure, one min
# [16:18] <glandium> jlebar: technically, nothing prevents native code to poke at the low-level interfaces to guess if we are on wifi or 3g. The high-level APIs require a permission that is very broad and grants completely unrelated things
# [16:19] * coop is now known as coop|afk
# [16:19] <jlebar> glandium, Are we talking about the same thing, here? I was asking mounir about proposed b2g interfaces.
# [16:19] <glandium> jlebar: ah, i thought you were talking about android
# [16:20] <mounir> jlebar: no, that doesn't require privileges
# [16:20] <mounir> jlebar: asking for the specific type requires privileges
# [16:20] * Joins: graememcc (chatzilla@moz-3E78DB35.range86-148.btcentralplus.com)
# [16:20] <jlebar> mounir, okay, cool.
# [16:20] <askalski> Hi, anyone uses Eclipse here?
# [16:20] <jlebar> askalski, roc uses it, last time I checked.
# [16:20] <jlebar> glandium, Interesting to know about Android, though.
# [16:20] <askalski> roc, : hi, you there?
# [16:21] <jlebar> askalski, He's in New Zealand, so expect strange hours. :)
# [16:22] * Joins: mkaply (chatzilla@moz-92EDDD02.lightspeed.austtx.sbcglobal.net)
# [16:22] * ChanServ sets mode: +o mkaply
# [16:22] <ewong> khuey posted a link to the build config introduction.. might anyone have a link handy? it's not https://developer.mozilla.org/en/Building_with_Profile-Guided_Optimization#Introduction
# [16:22] * Joins: mdas (mdas@8CB764BC.D30B51A1.412CF160.IP)
# [16:22] * philor|away is now known as philor
# [16:22] <jlebar> smaug, ping re https://bugzilla.mozilla.org/show_bug.cgi?id=710231#c21 ?
# [16:22] <jcranmer> espindola: ping
# [16:23] * sancus is now known as sancus_
# [16:23] * sancus_ is now known as sancus
# [16:23] <askalski> jlebar, : ok. how to get his e-mail then?
# [16:23] <ewong> ahh I think this is it : https://developer.mozilla.org/en/How_Mozilla%27s_build_system_works
# [16:23] <ewong> yup it is.. thanks!
# [16:23] <Ms2ger> firebot, roc?
# [16:23] <firebot> Ms2ger: well, duh. roc is mailto:roc@ocallahan.org, the creator of the new viewmanager
# [16:24] <Ms2ger> Interesting claim to fame
# [16:24] <espindola> jcranmer, pong
# [16:24] <Ms2ger> askalski, ^
# [16:24] <jcranmer> espindola: what version of clang do tinderbox builders use, do you know?
# [16:24] <wolfiR> glandium: I get a segmentation fault in xpcshell when populate_startupcache is called and found Bug 672756 where you did something
# [16:24] <askalski> Ms2ger, : roc
# [16:24] <@smaug> jlebar: looking...
# [16:25] <jlebar> smaug, thanks.
# [16:25] <espindola> jcranmer, somewhat old 3.0. Let me check
# [16:25] <jlebar> smaug, The contexts always get me messed up. :-/
# [16:25] <wolfiR> glandium: happens for a custom application (not ff); any idea
# [16:25] * Quits: beaufour (beaufour@18D5CC88.C7EE4FB2.ECED8BE3.IP) (Ping timeout)
# [16:25] <@smaug> jlebar: you could kick mrbkap or bholley to remove the requirement to push cx
# [16:25] <espindola> I am working on building a new one (and gcc, and changing how we build them...) but is very low priority unfortunately
# [16:25] <@smaug> jlebar: last time I asked, cx pushing was still needed
# [16:26] <jlebar> smaug, Yeah, last time I asked too. But the question is *which* cx? :)
# [16:26] <glandium> wolfiR: get a backtrace
# [16:26] * Joins: sgautherie (chatzilla@moz-D7B69DC4.fbxo.proxad.net)
# [16:26] * Quits: ewong (chatzilla@moz-FBF18AE4.static.netvigator.com) (Quit: GoneToSleep)
# [16:26] <espindola> jcranmer, r145057 I think
# [16:26] <espindola> jcranmer, http://people.mozilla.org/~raliiev/clang/
# [16:26] <jcranmer> that sounds relatively newish
# [16:27] <Ms2ger> !summon ehsan
# [16:27] <ehsan> Ms2ger: hi
# [16:27] <@smaug> jlebar: in this case the context you get from iframe
# [16:27] <Ms2ger> Did you see my last comment in bug 716207?
# [16:27] * Joins: beaufour (beaufour@18D5CC88.C7EE4FB2.ECED8BE3.IP)
# [16:27] <jlebar> smaug, Okay, why?
# [16:28] <@smaug> jlebar: does BrowserFrameProgressListener have a pointer to the element?
# [16:28] <rail> espindola: http://people.mozilla.org/~raliiev/yasm/
# [16:28] <jlebar> smaug, If it doesn't, it can...
# [16:28] <espindola> rail, thanks
# [16:28] <@smaug> it doesn't have
# [16:28] <rail> np
# [16:29] <mrbkap> smaug: That depends on compartment-per-global, still.
# [16:29] <@smaug> mrbkap: ok
# [16:29] <jlebar> smaug, Wait, on bugzilla, you said the opposite of what I understood on IRC.
# [16:29] <@smaug> ?
# [16:30] <jlebar> smaug, I push the cx of the iframe, or the cx of the window which registered the callback?
# [16:30] <@smaug> you could keep a weakref to window or element
# [16:30] <Ms2ger> mrbkap, so when is cpg happening?
# [16:30] <@smaug> same cx anyway
# [16:30] <mrbkap> Ms2ger: We need to wait until the end of the beta cycle, iirc.
# [16:30] <@smaug> usually
# [16:30] <jlebar> smaug, The iframe element.
# [16:30] <mrbkap> Ms2ger: We're making sure that nobody's using JS off the main thread first.
# [16:30] <Ms2ger> Oh, right
# [16:30] <jlebar> smaug, Not the global of the script which registered the listener.
# [16:30] <ehsan> Ms2ger: commented
# [16:30] <@smaug> jlebar: using iframe would make this consistent with event handlers
# [16:30] <ehsan> need to get to the office
# [16:30] <ehsan> brb
# [16:30] <Ms2ger> ehsan, ta
# [16:30] * Quits: ehsan (ehsan@F0B20A8D.8458880F.57F33CED.IP) (Input/output error)
# [16:31] * Quits: graememcc (chatzilla@moz-3E78DB35.range86-148.btcentralplus.com) (Connection reset by peer)
# [16:31] <jlebar> smaug, Okay, I think I understand. Although I don't really understand why it runs in the iframe window's context. Like...what's the general rule here?
# [16:31] <@smaug> jlebar: expect in some non-common cases pushing context using element or window ends up pushing the same thing
# [16:32] <@smaug> jlebar: oh, no, I don't mean the cx inside iframe
# [16:32] <@smaug> I mean the context of iframe
# [16:32] <jlebar> smaug, The context of the window containing the <iframe>, right.
# [16:32] <@smaug> you can do nsCxPusher pusher; pusher.Push(iframe);
# [16:32] <jlebar> smaug, Right.
# [16:32] <ted> wolfiR: there's a bug that gaston filed where he was hitting a similar problem on OpenBSD
# [16:33] <jlebar> smaug, My question is: Why? What is the general rule? Or, what would go wrong if I pushed, say, the global of the script which called addInnerStateListener()?
# [16:33] <@smaug> jlebar: it affects to the security checks
# [16:34] * Quits: jhk (jiggy@8E6C34C1.A3F9767A.1C37C358.IP) (Connection reset by peer)
# [16:34] <@smaug> jlebar: the important part is to push some cx which doesn't have more privileges than the iframe
# [16:34] <wolfiR> glandium: in which directory is that called, (or actually from which script?
# [16:34] <jlebar> smaug, So in general, I need to push the context of the window being accessed, so we check security properly?
# [16:34] <@smaug> window being accessed?
# [16:34] * Joins: evilpie (evilpie@moz-37567C56.pools.arcor-ip.net)
# [16:34] <@smaug> I don't any any window being accessed here
# [16:34] <glandium> wolfiR: look up in the log, you'll have the xpcshell command line
# [16:34] <@smaug> s/any any/any/
# [16:34] <jlebar> smaug, Well, the <iframe> lives in a window.
# [16:35] <@smaug> it lives in a document
# [16:35] <jlebar> smaug, ...which lives in a window?
# [16:35] * Quits: glazou (glazou@moz-204094DD.disruptive-innovations.fr) (Quit: bbl)
# [16:35] <glandium> wolfiR: it's in toolkit/mozapps/installer/packager.mk, but it's painful to debug from there
# [16:35] <@smaug> elements don't always live in an window
# [16:35] * Joins: AutomatedTester (AutomatedT@moz-D4F73363.as13285.net)
# [16:36] <wolfiR> glandium: I get this error /bin/sh: Zeile 1: 15771 Speicherzugriffsfehler /local/mozilla/hg/ezperanto-build/release/dist/bin/run-mozilla.sh /local/mozilla/hg/ezperanto-build/release/dist/bin/xpcshell -g "$PWD" -a "$PWD" -f /local/mozilla/hg/mozilla-ezperanto/toolkit/mozapps/installer/precompile_cache.js -e "populate_startupcache('GreD', 'omni.jar', 'startupCache.zip');"
# [16:36] <jlebar> smaug, Hm, that's true.
# [16:36] * Joins: hub (hub@83874EA1.EB7C1AF9.6F478678.IP)
# [16:36] <wolfiR> glandium: calling that manually in objdir gives: +++ Failed to get ScriptSecurityManager service, running without principalsSpeicherzugriffsfehler
# [16:36] <@smaug> jlebar: nsCxPusher takes context from GetContextForEventHandlers
# [16:36] <wolfiR> glandium: sorry for the german message ;-)
# [16:36] <jlebar> smaug, I'm just trying to understand how I can figure this out in the future without interrogating you. :)
# [16:37] <glandium> wolfiR: are you running that in a chroot?
# [16:37] * Joins: jhk (jiggy@8E6C34C1.A3F9767A.1C37C358.IP)
# [16:37] <wolfiR> glandium: no
# [16:37] <Standard8> hrm, how do I log an error message to the error console without writing a bunch of code or throwing?
# [16:37] <glandium> edmorley, mak: does either of you plan a m-i merge soon(ish)?
# [16:37] <@smaug> jlebar: take the context in which something lives if possible
# [16:37] <mak> glandium: I made one earlier, I don't plan another one, personally
# [16:38] <glandium> wolfiR: i know i've seen such messages when /proc is not mounted
# [16:38] * Joins: jdm (jdm@moz-9AEDE212.cpe.net.cable.rogers.com)
# [16:38] <jlebar> smaug, Where "something" is "something I'm accessing, or touching." Okay.
# [16:38] <jlebar> smaug, I think I understand enough to be dangerous. :)
# [16:38] <@smaug> jlebar: if not, use the same context which was used when adding the callback (this could be more right, but is just usually complicated )
# [16:38] <wolfiR> calling it manually in gdb just reveals main (argc=5, argv=0x7fffffffdcd8, envp=0x7fffffffdd08) at /local/mozilla/hg/mozilla-ezperanto/js/src/xpconnect/shell/xpcshell.cpp:1916
# [16:38] <wolfiR> 1916 cb->findObjectPrincipals = FindObjectPrincipals;
# [16:39] * Joins: dao (dao@moz-F86049E5.superkabel.de)
# [16:39] <jlebar> smaug, Oh, that would work too? Okay.
# [16:39] <Standard8> oh reportError
# [16:39] <wolfiR> but the +++ Failed to get ScriptSecurityManager service, running without principals does not happen when called from make
# [16:39] <glandium> wolfiR: ok, so what triggers that message is that $PWD is not the right value
# [16:39] <@smaug> jlebar: then you just need to make sure you don't use wrong context if you have several callbacks
# [16:40] <wolfiR> glandium: that's why I asked where ($PWD) it should run
# [16:40] <jlebar> smaug, sure.
# [16:40] * rail is now known as rail_away
# [16:40] <jlebar> smaug, And I need to make sure that the contexts are still alive.
# [16:40] * rail_away is now known as rail
# [16:40] * Quits: gandalf (zbraniecki@moz-66C04999.neoplus.adsl.tpnet.pl) (Quit: Computer has gone to sleep.)
# [16:40] <@smaug> jlebar: oh, also, if you need to call several callbacks in a row (for loop or such), you need to call (Re)Push for each callback
# [16:40] <glandium> wolfiR: check packager.mk or the log, there should be a cd before the xpcshell invocation. i think it's either dist/bin in the objdir or the directory containing xpcshell (when building against a xul sdk)
# [16:41] <jlebar> smaug, oic.
# [16:41] * Joins: vladan (vladan@F2D29657.F60B0462.67AC9B1.IP)
# [16:41] <@smaug> jlebar: this really sucks, but apparently we need to wait for compartment changes
# [16:41] * Quits: RemusPop (remuspop@601F3B17.33662590.A5830293.IP) (Client exited)
# [16:41] <jlebar> smaug, It's better now -- at least I have some idea what I'm supposed to do!
# [16:41] <jlebar> smaug, I appreciate the help.
# [16:41] <@smaug> :)
# [16:42] <@smaug> jlebar: in general, if you have EventTarget object involved, just do nsCxPusher pusher; NS_ENSURE_STATE(pusher.Push(eventtarget));
# [16:42] * Joins: josh (josh@moz-75EC6735.dlth.qwest.net)
# [16:42] * khuey|away is now known as khuey
# [16:43] <wolfiR> glandium: found it, stacktrace now is: #0 nsTArray<nsAutoPtr<mozilla::CacheData>, nsTArrayDefaultAllocator>::AppendElements<mozilla::CacheData*> (this=0x0, array=
# [16:43] <wolfiR> 0x7fffffffd748, arrayLen=1) at ../../../dist/include/nsTArray.h:830
# [16:43] <wolfiR> #1 0x00007ffff5ea9c5b in mozilla::Preferences::AddBoolVarCache (aCache=0x7ffff7775438, aPref=
# [16:43] <wolfiR> 0x7ffff6baa52e "privacy.donottrackheader.enabled", aDefault=0)
# [16:43] <wolfiR> at /local/mozilla/hg/mozilla-ezperanto/modules/libpref/src/Preferences.cpp:1474
# [16:43] <gabor> mrbkap: I've sent you an email as I promissed, I hope it makes somewhat sense...
# [16:43] <wolfiR> brb (15 minutes)
# [16:43] <glandium> wolfiR: never seen that. I guess you'll have to debug it yourself ;)
# [16:43] * Cwiiis is now known as CwiiisAway
# [16:44] * Quits: asadotzler (asa@F259CFB7.9EF26CBA.EB5E51FC.IP) (Ping timeout)
# [16:44] <jcranmer> espindola: actually, your version of clang is 3.1
# [16:44] * Joins: asadotzler (asa@F259CFB7.9EF26CBA.EB5E51FC.IP)
# [16:44] * Quits: pnemsak (Miranda@moz-3D67D819.rainside.sk) (Connection reset by peer)
# [16:45] * Quits: AutomatedTester (AutomatedT@moz-D4F73363.as13285.net) (Quit: AutomatedTester)
# [16:45] * Ziggy|AWAY is now known as Ziggy_Maes
# [16:45] <espindola> jcranmer, I think that is correct
# [16:45] <decoder> I just reproduced the NS_ABORT_IF_FALSE(mName, "Name must be nonnull"); assertion during make package with a normal debug build
# [16:45] <espindola> I had a fix just after 3.0 branched...
# [16:46] <decoder> anyone else seeing this? (in xpcom/glue/BlockingResourceBase.h)
# [16:46] <jcranmer> espindola: include/clang/Basic/Version.inc is pretty clear about it :-)
# [16:46] <espindola> :-)
# [16:46] * Joins: madhava (madhava@F2D29657.F60B0462.67AC9B1.IP)
# [16:46] * khuey is now known as khuey|away
# [16:47] <Ms2ger> Uh, why is OSX debug broken on try?
# [16:47] <Ms2ger> abort: error: Temporary failure in name resolution :/
# [16:48] * Joins: ashughes (ashughes@83D384FE.DFA2E3D3.C3C3A9D0.IP)
# [16:48] <hsivonen> askalski: pong
# [16:49] <askalski> hsivonen, : hi
# [16:49] * Quits: beaufour (beaufour@18D5CC88.C7EE4FB2.ECED8BE3.IP) (Quit: beaufour)
# [16:49] <askalski> hsivonen, : have you managed to stop eclipse from complaining about undefined everything (like gboolean, gint...) or you just set "disable warnings" and ignore it?
# [16:50] <hsivonen> askalski: I don't recall configuring that
# [16:50] * Joins: ctopper (craig@C3495DA.BA3DBA56.AE2B2F80.IP)
# [16:51] <hsivonen> (doesn't mean I haven't configured it and forgotten)
# [16:51] <askalski> hsivonen, : ok, thanks. if I got it I will write it down
# [16:53] <Ms2ger> mats, dammit, I successfully avoided r+ing dzbarsky's patch, and now the log still claims I did! :)
# [16:54] * Joins: beaufour (beaufour@moz-723527FF.nycmny.fios.verizon.net)
# [16:54] * Joins: Mardak (Mardak@moz-7B0110AD.mv.mozilla.com)
# [16:54] <firebot> Check-in: http://hg.mozilla.org/mozilla-central/rev/0a5fdc470f9a - Rail Aliiev - bug 714542: Mac build uses pretty naming for *.checksums files - rename mac checksums file. r=ted
# [16:55] * Joins: damons (gnubeard@moz-A41E6911.hsd1.ca.comcast.net)
# [16:55] * Joins: hhillen (hhillen@moz-81563C98.ip.telfort.nl)
# [16:55] * bear-afk is now known as bear
# [16:55] * Quits: Mardak (Mardak@moz-7B0110AD.mv.mozilla.com) (Ping timeout)
# [16:55] * Joins: calatalee (Mibbit@F6A9DE18.A1146449.57460053.IP)
# [16:56] * Quits: calatalee (Mibbit@F6A9DE18.A1146449.57460053.IP) (Quit: http://www.mibbit.com ajax IRC Client)
# [16:57] <mrbkap> gabor: cool, I'll reply hopefully tomorrow morning.
# [16:58] * Quits: fabrice|zZz (fabrice@moz-104CC309.mv.mozilla.com) (Ping timeout)
# [16:58] <gabor> mrbkap: alright, sounds great
# [16:59] * coop|afk is now known as coop
# [17:02] <decoder> anyone familiar with xpcshell who could help me to track down a startup assertion?
# [17:02] <mrbkap> decoder: ?
# [17:02] * Joins: andreasn (andreasn@moz-BB6A8755.bredband.comhem.se)
# [17:02] <decoder> mrbkap: I tried building m-c today in debug mode, and make package fails. i tracked it down to the xpcshell not starting up:
# [17:03] <decoder> http://pastebin.mozilla.org/1443398
# [17:03] <decoder> (im in the installer directory because make package ran it there when it first failed)
# [17:04] <mrbkap> ew
# [17:04] <mrbkap> decoder: out of curiosity, what is the stack to that?
# [17:04] <mrbkap> decoder: but I don't know what would cause that.
# [17:05] <decoder> mrbkap: mozconfig is this: http://pastebin.mozilla.org/1443401
# [17:05] <decoder> ill try getting a stack now
# [17:05] * Joins: c0smikdebris (c0smikdebr@C4166EFD.7E460B98.BE90E62C.IP)
# [17:05] * Quits: dao (dao@moz-F86049E5.superkabel.de) (Ping timeout)
# [17:05] * Joins: dao (dao@moz-F86049E5.superkabel.de)
# [17:06] * Quits: jrmuizel (jrmuizel@moz-20EF8EAA.cpe.net.cable.rogers.com) (Input/output error)
# [17:06] <jlebar> What's the right way to call a function taking nsISupports* (say, do_GetWeakReference) on a concrete pointer which inherits from nsISupports multiple times?
# [17:06] <decoder> mrbkap: stack: http://pastebin.mozilla.org/1443402
# [17:06] <jlebar> Do I static_cast to an arbitrary interface?
# [17:06] <Ms2ger> Yes
# [17:07] <Ms2ger> NS_ISUPPORTS_CAST, I guess
# [17:07] <Ms2ger> And note do_QueryObject
# [17:07] * khuey|away is now known as khuey
# [17:07] <jlebar> Oh, I remember NS_ISUPPORTS_CAST now. Yay.
# [17:07] <jlebar> Ms2ger, thanks!
# [17:07] <Ms2ger> Np
# [17:08] <khuey> wow, is it really only 8 am on the west coast?
# [17:08] * Quits: waschtl (waschtl@moz-A4ECE553.hsi4.kabel-badenwuerttemberg.de) (Client exited)
# [17:08] <khuey> how do people in europe ever get stuff done?
# [17:08] * davidb is now known as davidb|mtg
# [17:08] <decoder> khuey: lol
# [17:08] <Ms2ger> Stay up late
# [17:08] * Quits: damons (gnubeard@moz-A41E6911.hsd1.ca.comcast.net) (Quit: damons)
# [17:08] <decoder> khuey: im usually up till 2-3 am
# [17:09] <khuey> ah
# [17:09] <khuey> I suppose that might do it
# [17:09] <Ms2ger> smaug usually comes in when I've been up for 4 hours or so
# [17:10] * Joins: fzzzy (donovan@moz-2B41AF9B.lightspeed.mtvwca.sbcglobal.net)
# [17:10] * Joins: graememcc (chatzilla@moz-3E78DB35.range86-148.btcentralplus.com)
# [17:10] * smaug is now known as smaugMtg
# [17:10] * Quits: jhorak (jhorak@moz-107AD163.redhat.com) (Quit: Leaving)
# [17:10] <decoder> mrbkap: i have a theory
# [17:11] * Joins: Cameron (Cameron@moz-7CC6EFB4.range213-121.btcentralplus.com)
# [17:11] <decoder> i forgot I made changes to this source tree which makes NS_ABORT fatal
# [17:11] <decoder> the assert could be older
# [17:11] * Quits: bjarne (bjarne@moz-46A85847.nextgentel.com) (Quit: Leaving.)
# [17:11] <decoder> just not aborting
# [17:11] <decoder> >.<
# [17:11] <Ms2ger> NS_ABORT_IF_FALSE is always fatal
# [17:11] <NeilAway> jlebar: if you have the concrete object you may as well call GetWeakReference yourself directly
# [17:12] <decoder> hmm
# [17:12] * Quits: merike|away (quassel@moz-339D0C93.cable.starman.ee) (Ping timeout)
# [17:12] <decoder> Ms2ger: ill try reverting my patch, it also modified NS_ABORT_IF_FALSE to help the static analysis.. but if they're both fatal then i wouldnt see how that could be the reason >.<
# [17:12] * Quits: mastoris (mastoris@6C80CE82.6958D9C9.BF6BBFAF.IP) (Quit: Leaving)
# [17:13] * Joins: damons (gnubeard@moz-A41E6911.hsd1.ca.comcast.net)
# [17:13] <ted> khuey: it's bad enough being an early riser on the east coast
# [17:13] <ted> most days i finish work before lots of people in MV start
# [17:13] <mrbkap> decoder: what does your patch to NS_ABORT_IF_FALSE look like?
# [17:13] * Joins: Mardak (Mardak@moz-BBE3ABD.mv.mozilla.com)
# [17:14] <Ms2ger> Make it use JS_ASSERT?
# [17:14] <Ms2ger> Because it usually prints differently
# [17:14] <mrbkap> decoder: It looks like the assertion is asserting the opposite of what was intended.
# [17:15] <decoder> mrbkap: you are right.. -.-
# [17:15] <decoder> it's not my patch btw^^
# [17:15] <decoder> but it's asserting !_expr
# [17:15] <decoder> which is wrong
# [17:15] <decoder> it should be assert(_expr);
# [17:15] <decoder> due
# [17:15] <decoder> *DOH
# [17:15] <Ms2ger> decoder++
# [17:16] * Joins: Joeh (joe@5A3923AA.BC22908.C7CEC4ED.IP)
# [17:16] * Quits: jdm (jdm@moz-9AEDE212.cpe.net.cable.rogers.com) (Client exited)
# [17:16] <mrbkap> NS_ABORT_IF_FALSE is like the worst name ever.
# [17:16] <mrbkap> We should rname it to NS_DONT_ABORT_IF_NOT_FALSE to make it better.
# [17:17] <Mitch> NS_I_KNOW_RIGHT
# [17:17] <Ms2ger> MOZ_FAITH
# [17:17] <Ms2ger> Anyway, once we land MOZ_ASSERT-with-message, we should get rid of it
# [17:17] <decoder> Ms2ger++
# [17:18] * davidb|mtg is now known as davidb
# [17:18] <Mitch> I HOPE WE CAN USE CAPS FOR MESSAGES AS WELL AS MACROS.
# [17:18] <Ms2ger> Mitch, r-
# [17:18] <Mitch> :(
# [17:18] * Joins: dveditz (dveditz@moz-34991AF4.dhcp.cruzio.com)
# [17:18] * ChanServ sets mode: +o dveditz
# [17:18] * Joins: mjschranz (chatzilla@moz-6FE6B833.cpe.net.cable.rogers.com)
# [17:19] * khuey attempts to parse NS_DONT_ABORT_IF_NOT_FALSE
# [17:19] * Joins: ehsan (ehsan@F2D29657.F60B0462.67AC9B1.IP)
# [17:20] <evilpie> Mitch Ms2ger am i doing this right http://www.quickmeme.com/meme/35okqx/ ?
# [17:20] <decoder> khuey: deadlock detected? ;)
# [17:20] <Ms2ger> evilpie++
# [17:20] <khuey> decoder: heh
# [17:20] <Ms2ger> khuey, nerd sniped! :)
# [17:20] <Mitch> evilpie: hah
# [17:22] <jcranmer> khuey: it doesn't do anything in release builds :-)
# [17:22] * Quits: ctopper (craig@C3495DA.BA3DBA56.AE2B2F80.IP) (Quit: ctopper)
# [17:23] * jhopkins|away is now known as jhopkins
# [17:24] * Quits: beaufour (beaufour@moz-723527FF.nycmny.fios.verizon.net) (Quit: beaufour)
# [17:24] * Joins: beaufour (beaufour@moz-723527FF.nycmny.fios.verizon.net)
# [17:25] * Quits: damons (gnubeard@moz-A41E6911.hsd1.ca.comcast.net) (Quit: damons)
# [17:27] * philor is now known as philor|away
# [17:27] * Joins: jhammel (jhammel@moz-C03D0C61.vlan426.asr1.sfo1.gblx.net)
# [17:27] * Joins: kdcw (kdc@moz-F7413045.pk.shawcable.net)
# [17:28] * Joins: akeybl (akeybl@moz-52D39FF6.hsd1.ca.comcast.net)
# [17:28] * Quits: jhammel (jhammel@moz-C03D0C61.vlan426.asr1.sfo1.gblx.net) (Quit: leaving)
# [17:28] * Joins: jhammel (jhammel@moz-C03D0C61.vlan426.asr1.sfo1.gblx.net)
# [17:28] * Joins: jrmuizel (jrmuizel@F2D29657.F60B0462.67AC9B1.IP)
# [17:29] * gregglind_afk is now known as gregglind
# [17:29] * Joins: mayhemer (Miranda@B3D46202.F87A741B.F23860FD.IP)
# [17:29] * Joins: brendan (brendaneic@moz-A286C218.hsd1.ca.comcast.net)
# [17:29] * Quits: Suresh (chatzilla@50054847.E0A9EEE1.EB06F97B.IP) (Quit: Suresh)
# [17:30] * Joins: jimb (user@moz-F4EC06CC.hsd1.or.comcast.net)
# [17:31] * wlach|afk is now known as wlach
# [17:31] * Quits: Ami_Ty_ (Amie@moz-7D22EB0A.dsl.static.sonic.net) (Quit: Ami_Ty_)
# [17:31] * Joins: smooney (smooney@moz-3C7A0050.hsd1.ca.comcast.net)
# [17:32] <mbrubeck> Using Firefox (on desktop Linux) to download a file to a really slow storage device (an old USB mp3 player) seems to cause long UI hangs...
# [17:32] * Joins: mconnor|m (mconnorm@E973A01C.5BCEC6DB.DA78B690.IP)
# [17:33] * philor|away is now known as philor
# [17:34] * Joins: ejpbruel (ejpbruel@moz-5EE20326.adsl2.static.versatel.nl)
# [17:34] <ejpbruel> so, ive succesfully compiled fennec once, but now that im trying to compile it a second time, im getting errors
# [17:35] <ejpbruel> mozilla-central/toolkit/mozapps/installer/package-name.mk: No such file or directory
# [17:35] <ejpbruel> whats going on?
# [17:35] <sid0> strange, I saw a UAC prompt with today's update
# [17:36] * Quits: not_gavin (gavin@D18F6DDB.6A393516.2321E71E.IP) (Broken pipe)
# [17:36] <bhearsum> bbondy: ^^
# [17:36] * mak is now known as mak|afk
# [17:36] <bbondy> sid0 have you previously not seen UAC prompts?
# [17:36] * Quits: sgautherie (chatzilla@moz-D7B69DC4.fbxo.proxad.net) (Quit: ChatZilla 0.9.78.1 [SeaMonkey 1.5a/2007051508])
# [17:37] <sid0> bbondy: yes, for the last 3 days
# [17:37] <sid0> just today though
# [17:37] * Joins: not_gavin (gavin@D18F6DDB.6A393516.2321E71E.IP)
# [17:37] <bbondy> sid0: Could you email me your C:\ProgramData\Mozilla\logs directory zipped to netzen@gmail.com ?
# [17:38] * Joins: mwu (mwu@moz-59435430.hsd1.nj.comcast.net)
# [17:38] <sid0> bbondy: sent
# [17:39] <firebot> Check-in: http://hg.mozilla.org/mozilla-central/rev/46282354b7be - Felipe Gomes - Bug 706897. Fix intermittent test_destinationURI_annotation.xul failure. r=mak
# [17:39] <sid0> bbondy: I don't even see a log corresponding to the update
# [17:41] <bbondy> checking now
# [17:41] <bbondy> sid0 It says the maintenanceservice.log file was last modified maintenanceservice.log at 3:39am is that accurate?
# [17:42] <bbondy> I'm just using the last modified field in the zip file
# [17:42] * Joins: rillian (giles@21B7B9F2.B87E9213.6E712CE2.IP)
# [17:42] * Quits: not_gavin (gavin@D18F6DDB.6A393516.2321E71E.IP) (Ping timeout)
# [17:42] <bbondy> does that align to the update that used the uac prompt?
# [17:42] * Quits: dao (dao@moz-F86049E5.superkabel.de) (Quit: Leaving.)
# [17:42] <Ms2ger> mats, indeed you were returning an uninitialized value :)
# [17:42] <sid0> bbondy: that's 3:39 am my time. no, that would be yesterday's update
# [17:42] <bbondy> I would gues not since that update succeeded through the maintenance service, if you did an udpate after that again it seems the maintenance service wasn't attempted to be used.
# [17:43] <sid0> strange that winrar didn't use uac
# [17:43] <sid0> er, I mean it didn't use utc
# [17:43] * lsblakk|afk is now known as lsblakk
# [17:44] * Quits: mconnor|m (mconnorm@E973A01C.5BCEC6DB.DA78B690.IP) (Quit: Colloquy for iPhone - http://colloquy.mobi)
# [17:44] * Joins: TheLink (TheLink@moz-B5AB52CB.pools.arcor-ip.net)
# [17:44] * lsblakk is now known as lsblakk|buildduty
# [17:44] <bbondy> sid0: I CC'ed you on a bug I suspect is your issue, I just landed a patch that gives extra logging. Could you attach your log the next time the issue happens to you? It should be merged from mozilla-inbound to mozilla-central sometime today so it should be in your next update.
# [17:44] <sid0> bbondy: sure
# [17:45] * Joins: not_gavin (gavin@D18F6DDB.6A393516.2321E71E.IP)
# [17:45] <mats> Ms2ger: good to know you're reviewing *all* patches ;-)
# [17:45] <Ms2ger> :)
# [17:45] * Joins: mconnor|m (mconnorm@E973A01C.5BCEC6DB.DA78B690.IP)
# [17:46] <bbondy> sid0: also please answer https://bugzilla.mozilla.org/show_bug.cgi?id=715910#c20 and mention you also had a UAC prompt during update. In particular I'm only looking for an answer to the 2nd question inside Comment 20.
# [17:46] <bbondy> as a new comment.
# [17:47] <sid0> perms seem to be set correctly
# [17:47] * smaugMtg is now known as smaug
# [17:47] <sid0> but let me answer in the bug
# [17:47] <froydnj> mwu: do you have an eta on reviewing 711297?
# [17:47] <bbondy> sid0: ok please do and I'll probably have some follow up questions
# [17:47] <bbondy> possibly
# [17:48] <@smaug> Ms2ger: yes, I'm usually online from noon to 2-4am. I'm actually trying to change that right now.
# [17:49] <Ms2ger> I guess that wouldn't work too well during the work week :)
# [17:49] <@smaug> exactly
# [17:49] <bbondy> sid0: were you perhaps using an x64 build? Or a different branch like UX that might not have the change in it yet? x64 native builds do not use the service presently.
# [17:49] <khuey> well bz wanted to take the mornings for tourism and work late
# [17:49] <khuey> so that might not be so bad
# [17:49] <sid0> bbondy: no, nightly, stock x86 build
# [17:49] <bbondy> k
# [17:49] <Ms2ger> Not *that* late, I don't think :)
# [17:50] <sid0> bbondy: as I said the last few updates haven't popped up a prompt
# [17:50] * @smaug has often lunch before he really starts to work
# [17:50] <edmorley> glandium: I will be once there is enough green to warrant it (which I'll check once I've eaten)
# [17:50] <khuey> yeah I'm not staying up till 4 am ;-)
# [17:50] <bbondy> sid0: Ya just wasn't sure if you happened to open an x64 install beside the normal x86 that you use. Just had to check, thanks.
# [17:50] <@smaug> khuey: hey, 4am here is only 3am in Paris :)
# [17:51] <Ms2ger> Oh, 3AM, that doesn't sound as bad! :)
# [17:51] * Quits: mconnor|m (mconnorm@E973A01C.5BCEC6DB.DA78B690.IP) (Connection reset by peer)
# [17:51] * fabrice|afk is now known as fabrice
# [17:51] * Joins: mconnor|m_ (mconnorm@F2D29657.F60B0462.67AC9B1.IP)
# [17:52] <khuey> oh yeah, it's all fine then
# [17:52] <bbondy> I might spin off another bug after you put your comment in btw.
# [17:52] <glandium> edmorley: ok, when you do, first clobber (all but linux if you have the patience to deal with the clobberer). And ping Callek or someone else to clobber and land the c-c part of bug 701371
# [17:52] <mwu> oh, man.
# [17:53] * Quits: evilpie (evilpie@moz-37567C56.pools.arcor-ip.net) (Input/output error)
# [17:53] <mwu> froydnj: I'll stamp it
# [17:53] <mwu> but I'm not sure if you want me to review this
# [17:53] <mwu> since I don't know this code
# [17:53] <mwu> and I know telemetry even less
# [17:53] <bbondy> sid0: Can you check to make sure your preference is still on too to use the service? Options | Advanced | Update
# [17:53] * Joins: wesj (wesj@moz-C03D0C61.vlan426.asr1.sfo1.gblx.net)
# [17:53] * Joins: evilpie (evilpie@moz-37567C56.pools.arcor-ip.net)
# [17:53] <bbondy> there are some failsafe that are in place to auto turn it off btw
# [17:53] <sid0> bbondy: yes, it's ticked
# [17:54] <bbondy> k
# [17:54] * Quits: jrmuizel (jrmuizel@F2D29657.F60B0462.67AC9B1.IP) (Input/output error)
# [17:54] <froydnj> mwu: if you don't feel good about it, just bounce it to taras
# [17:55] <mwu> ok
# [17:55] * catlee is now known as catlee-lunch
# [17:55] <froydnj> mwu: thanks
# [17:55] * rail is now known as rail-lunch
# [17:55] * davidb is now known as davidb|afk
# [17:55] <Ms2ger> mounir, dammit :)
# [17:55] <mounir> Ms2ger: sorry :(
# [17:56] <Ms2ger> He's asking you for review, might as well give him review comments :)
# [17:56] * armenzg is now known as armenzg_lunch
# [17:56] <bbondy> sid0: In which way did you take the update?
# [17:56] <bbondy> for example about dialog
# [17:56] <bbondy> app was just closed and you happened to start it and it updated
# [17:56] <bbondy> etc.
# [17:56] <sid0> bbondy: an update window popped up
# [17:57] <bbondy> k
# [17:57] * khuey is now known as khuey|away
# [17:57] <bbondy> sid0: Are you set to automatically take updates in preferences? Or ask when to apply?
# [17:57] <sid0> bbondy: automatic
# [17:57] * philor is now known as philor|away
# [17:58] * Joins: faramarz (faramarz@moz-BBE3ABD.mv.mozilla.com)
# [17:58] <mounir> Ms2ger: I didn't realize that could be initialized and I didn't want to bother him with "please use NS_ENSURE_SUCCESS"
# [17:58] <bbondy> thanks
# [17:58] <Ms2ger> Should review more carefully, then :)
# [17:59] <Ms2ger> "This writes '<DL><H3>Title</H3>'"
# [17:59] <bbondy> sid0: last question, about:config do you have an entry called app.update.service.errors?
# [17:59] * Ms2ger sighs at Netscape
# [17:59] <sid0> bbondy: no
# [17:59] <bbondy> k so it didn't attempt for sure. thanks.
# [18:00] * Joins: daoberes (chatzilla@D2B84AF4.C749D475.1139E686.IP)
# [18:01] * Quits: evilpie (evilpie@moz-37567C56.pools.arcor-ip.net) (Input/output error)
# [18:02] * Quits: mjschranz (chatzilla@moz-6FE6B833.cpe.net.cable.rogers.com) (Ping timeout)
# [18:02] * Quits: smooney (smooney@moz-3C7A0050.hsd1.ca.comcast.net) (Quit: smooney)
# [18:03] * Joins: Waldo (waldo@moz-2A0E3E7E.dsl.emhril.sbcglobal.net)
# [18:03] <NeilAway> bbondy: why are you generating error 3534 ;-)
# [18:03] <bbondy> 3534?
# [18:03] * Joins: nhirata (nhirata.bu@moz-BBE3ABD.mv.mozilla.com)
# [18:04] * philor|away is now known as philor
# [18:04] * Quits: nhirata (nhirata.bu@moz-BBE3ABD.mv.mozilla.com) (Input/output error)
# [18:04] <bbondy> NeilAway: In regards to what?
# [18:05] * Joins: nhirata (nhirata.bu@moz-BBE3ABD.mv.mozilla.com)
# [18:05] * Joins: bz (bzbarsky@moz-69B5879F.bstnma.fios.verizon.net)
# [18:05] * ChanServ sets mode: +o bz
# [18:05] <@bz> so wait
# [18:05] <@bz> with the new browserid login on devmo
# [18:05] <@bz> will my old edits and my new edits look like different people, basically?
# [18:05] <Waldo> no
# [18:05] <Waldo> supposedly they carry over
# [18:05] <Waldo> I'm not entirely sure how they do this, maybe you have to use the same email?
# [18:05] * @bz wonders how
# [18:06] <@bz> ok
# [18:06] <@bz> so what does this browserid password mean?
# [18:06] <@bz> what does it protect?
# [18:06] <@bz> and WHY THE HELL DOES IT HAVE A LENGTH LIMIT?
# [18:06] <NeilAway> bbondy: ah, that's interesting, the sysmonlog service starts without error on w2k, but on w2k3 it also reports error 3534
# [18:06] * @bz cries
# [18:06] * @bz gives up on browserid for now, and perforce on editing devmo
# [18:06] <Waldo> decoder: what are you doing annotating assertions?
# [18:07] <NeilAway> bbondy: ignore me then
# [18:07] <Waldo> bz: I'd also stay away from perforce if I were you
# [18:07] <bbondy> NeilAway: what are you trying to do?
# [18:07] <@bz> waldo: har har
# [18:07] <bbondy> NeilAway: Start i manually or something?
# [18:07] * Joins: dbaron (dbaron@moz-C03D0C61.vlan426.asr1.sfo1.gblx.net)
# [18:07] * ChanServ sets mode: +o dbaron
# [18:07] <@bz> what's the right place to post about browserid stuff?
# [18:07] * Ms2ger mumbles something
# [18:07] <NeilAway> bbondy: I was wondering why starting the service manually generates error 3534, but sysmonlog does too, which I hadn't realised
# [18:07] <bbondy> NeilAway: It will stop as soon as you start, it's an on demand only and can't be always running if that's what you're trying to do.
# [18:08] <bbondy> NeilAway: ah ok cool. Ya that's fine.
# [18:08] <bbondy> NeilAway: The test I asked sid0 about above was just to determine if as an unelevated user, if he had access to the service or if he got access denied.
# [18:09] <NeilAway> bbondy: on w2k net start reports no error, and services mmc thinks it's started when it hasn't, but on w2k3 net start reports error 3534, and services mmc has this spiel about the service started and then stopped
# [18:09] * Quits: gabor (gabor@moz-3B57BCD1.catv.pool.telekom.hu) (Connection reset by peer)
# [18:09] <NeilAway> bbondy: I was just wondering why error 3534 was expected ;-)
# [18:09] * NeilAway wonders what sc start does in that case
# [18:09] <bbondy> NeilAway: I don't specifically return that, seems the SCM sets that or something, not sure.
# [18:10] * sancus is now known as sancus_
# [18:10] * sancus_ is now known as sancus
# [18:10] <jlebar> bz, #identity?
# [18:10] <Waldo> bz: probably #devmo can direct you correctly; I know there have been posts in .platform...or that
# [18:10] <bbondy> NeilAway: That spiel about the service started and stopped does mention that some services do this by design though :)
# [18:10] * bear is now known as bear-afk
# [18:11] <decoder> Waldo: no, making certain assertions failing (statically), such that the static analysis can recognize them
# [18:11] <Waldo> decoder: is this asan?
# [18:11] <decoder> Waldo: no. clang static analysis
# [18:11] <decoder> asan is a runtime thing
# [18:11] <Waldo> decoder: er, sorry, right
# [18:11] * Quits: daoberes (chatzilla@D2B84AF4.C749D475.1139E686.IP) (Ping timeout)
# [18:11] <Waldo> decoder: https://bugzilla.mozilla.org/show_bug.cgi?id=716626
# [18:12] <Waldo> sprinkle as needed
# [18:12] <jlebar> smaug, My patch queue has become a tangled mess of renames, unrenames, and re-renames. You may have to trust me that it comes out correctly. :-/
# [18:12] <@bz> jlebar: I was thinking more like email
# [18:12] <decoder> Waldo: ah yea, makes sense :)
# [18:12] <Waldo> since MOZ_NORETURN won't work given that debuggers let you return from some of our assertion methods, even without doing NeilAway-style instruction munging
# [18:12] <@bz> jlebar: unless you seriously think people are awake now
# [18:12] * Ms2ger doesn't trust jlebar
# [18:13] <Waldo> decoder: I probably should file a bug to wrap that up in...some sort of macro
# [18:13] * @bz tries #identity just in case
# [18:13] * Waldo kind of hates that MOZ_NORETURN isn't adequate for it
# [18:13] <Waldo> but, optimizations are messy
# [18:13] * Joins: gabor (admin@moz-3B57BCD1.catv.pool.telekom.hu)
# [18:13] * Joins: anky (anky@27B0D94A.B421193.A3D1B221.IP)
# [18:13] <decoder> Waldo: there is also some ifdef for static analysis, might be able to use that
# [18:14] * Quits: fzzzy (donovan@moz-2B41AF9B.lightspeed.mtvwca.sbcglobal.net) (Quit: fzzzy)
# [18:14] <jcranmer> we really need to get a static analysis builder running again
# [18:14] * Quits: anky (anky@27B0D94A.B421193.A3D1B221.IP) (Connection reset by peer)
# [18:14] <Waldo> decoder: something provided by clang?
# [18:14] <NeilAway> bbondy: indeed, if only net.exe was that clever :-)
# [18:14] <bbondy> :)
# [18:14] <jcranmer> Waldo: I think he's referring to MOZ_STATIC_ANALYSIS or similar
# [18:14] <jcranmer> which runs only if you do --enable-static-checking
# [18:14] <Waldo> hm, that exists?
# [18:15] * Waldo can't find MOZ/NS_STATIC_ANALYSIS in mxr
# [18:16] <jcranmer> NS_STATIC_CHECKING, to be precise
# [18:16] * Joins: kumar (kmcmillan@moz-D0C71163.domainhotel-wifi.mozilla.hq)
# [18:16] * Quits: imphil (philipp@moz-655EF802.customer.m-online.net) (Ping timeout)
# [18:17] <decoder> right. im defining that manually
# [18:17] * Quits: @smaug (chatzilla@moz-958D39B4.elisa-mobile.fi) (Ping timeout)
# [18:18] <decoder> also had to remove some old JS code for that to compile
# [18:18] * Joins: flx_ (flx@590D15DC.A7CE92B2.30E47D40.IP)
# [18:19] * Joins: anant (anant@moz-104CC309.mv.mozilla.com)
# [18:20] <sid0> bbondy: do you want me to try reproducing it in a VM maybe?
# [18:20] <bbondy> sid0: If you don't mind sure, it would be good to know if your update.status file contained pending-service or just pending.
# [18:20] <decoder> phone.. Waldo: ill let you know which JS part failed with that define later, maybe you can fix it=)
# [18:20] * Joins: sicking (chatzilla@moz-C03D0C61.vlan426.asr1.sfo1.gblx.net)
# [18:21] * Quits: IanN (iann@4837CDFE.2ECE621A.E5F36E28.IP) (Quit: ChatZilla 0.9.88 [SeaMonkey 2.7/20120105141124])
# [18:21] <Waldo> har har har :-P
# [18:21] <sid0> bbondy: well, I might not be able to, let me see
# [18:21] * jwir3|away is now known as jwir3
# [18:21] <bbondy> sid0: if not no worries, just happen to grab it on the subsequent updates for future days
# [18:22] * Joins: pcwalton (pcwalton@moz-7B0110AD.mv.mozilla.com)
# [18:23] * Joins: bretr (bret_recka@2557D88A.3BA624FF.D4043753.IP)
# [18:24] * Quits: hub (hub@83874EA1.EB7C1AF9.6F478678.IP) (Input/output error)
# [18:24] * Quits: kumar (kmcmillan@moz-D0C71163.domainhotel-wifi.mozilla.hq) (Ping timeout)
# [18:24] * Quits: mconnor|m_ (mconnorm@F2D29657.F60B0462.67AC9B1.IP) (Quit: Colloquy for iPhone - http://colloquy.mobi)
# [18:24] <jcranmer> espindola: ping
# [18:24] * Joins: kumar (kmcmillan@moz-D0C71163.domainhotel-wifi.mozilla.hq)
# [18:25] * Quits: kumar (kmcmillan@moz-D0C71163.domainhotel-wifi.mozilla.hq) (Quit: kumar)
# [18:25] * Joins: anky (anky@27B0D94A.B421193.A3D1B221.IP)
# [18:25] <Waldo> decoder: I filed bug 716917 on adding MOZ_DEBUGGABLE_NORETURN as an attribute that could/should be applied to JS_Assert's declaration
# [18:26] * Quits: dRdR (dRdR@moz-3D8840A3.uwaterloo.ca) (Client exited)
# [18:26] * Quits: anky (anky@27B0D94A.B421193.A3D1B221.IP) (Connection reset by peer)
# [18:27] * Joins: bretr_ (bret_recka@A73CA40C.E7870264.CFAD1666.IP)
# [18:27] <espindola> jcranmer, pong
# [18:27] * Joins: anky (anky@27B0D94A.B421193.A3D1B221.IP)
# [18:28] * Quits: bretr (bret_recka@2557D88A.3BA624FF.D4043753.IP) (Ping timeout)
# [18:28] * bretr_ is now known as bretr
# [18:28] <jcranmer> espindola: in http://hg.mozilla.org/build/puppet-manifests/file/c997e25f5170/modules/packages/manifests/devtools.pp
# [18:28] <decoder> Waldo: thx
# [18:29] <jcranmer> where could I find the rpm that corresponds to 'gcc45' ?
# [18:29] <espindola> jcranmer, you have to ask releng :-(
# [18:30] <espindola> jcranmer, http://people.mozilla.org/~raliiev/gcc/
# [18:30] <espindola> has an older version
# [18:30] <jcranmer> okay
# [18:31] * Quits: akeybl (akeybl@moz-52D39FF6.hsd1.ca.comcast.net) (Quit: Leaving...)
# [18:31] * Joins: smooney (smooney@moz-BBE3ABD.mv.mozilla.com)
# [18:33] * Joins: bmoss (bmoss@moz-BBE3ABD.mv.mozilla.com)
# [18:33] * Quits: beaufour (beaufour@moz-723527FF.nycmny.fios.verizon.net) (Quit: beaufour)
# [18:33] * Joins: smaug (chatzilla@moz-AC5A01B5.elisa-mobile.fi)
# [18:33] * ChanServ sets mode: +o smaug
# [18:33] * Joins: faramarz_ (faramarz@moz-BBE3ABD.mv.mozilla.com)
# [18:33] * Quits: faramarz_ (faramarz@moz-BBE3ABD.mv.mozilla.com) (Input/output error)
# [18:34] * Quits: faramarz (faramarz@moz-BBE3ABD.mv.mozilla.com) (Connection reset by peer)
# [18:34] * Quits: daim (David_Mart@779E3E00.1773D26C.C0FF2207.IP) (Client exited)
# [18:35] * jimm is now known as jimm-lunch
# [18:35] * Joins: johanc (chatzilla@moz-D8A1AA43.bredband.comhem.se)
# [18:36] * Joins: mjschranz (chatzilla@moz-6FE6B833.cpe.net.cable.rogers.com)
# [18:36] * Joins: jet (junglecode@moz-C03D0C61.vlan426.asr1.sfo1.gblx.net)
# [18:36] * Quits: jhk (jiggy@8E6C34C1.A3F9767A.1C37C358.IP) (Connection reset by peer)
# [18:37] * Joins: imphil (philipp@moz-655EF802.customer.m-online.net)
# [18:37] * Joins: jgriffin (jgriffin@moz-4FBFA41D.hsd1.wa.comcast.net)
# [18:39] <jorendorff> ted: i guess our official builds use make -j1 and not pymake? why's that?
# [18:39] <jorendorff> (on windows, i mean)
# [18:39] <ted> we haven't gotten around to switching
# [18:39] <ted> there's a bug!
# [18:39] * Joins: Mossop (mossop@moz-C03D0C61.vlan426.asr1.sfo1.gblx.net)
# [18:39] <ted> and we fixed most of the issues
# [18:40] <jorendorff> a bug in pymake? for shame!
# [18:40] <jorendorff> oh, you mean a bug is on file
# [18:40] <ted> right
# [18:40] <gps> there's a bug to switch it
# [18:40] * Joins: fzzzy (donovan@moz-BBE3ABD.mv.mozilla.com)
# [18:40] <gps> bug 593585
# [18:41] <ted> thanks
# [18:41] <ted> i was looking for that, couldn't find it
# [18:41] * Joins: jhk (jiggy@8E6C34C1.A3F9767A.1C37C358.IP)
# [18:41] * Joins: faramarz (faramarz@moz-BBE3ABD.mv.mozilla.com)
# [18:41] <ted> i think all the deps have been fixed
# [18:41] <ted> it's just a matter of RelEng having enough time to fix it
# [18:41] <ted> you can poke catlee-lunch and ask
# [18:42] * Quits: Yoric (Yoric@moz-A2A3C9E6.vlan402.asr1.cdg3.gblx.net) (Input/output error)
# [18:42] <jorendorff> current status: "it's in some hairy stuff" :(
# [18:42] <jorendorff> ted: i was just curious.
# [18:43] <ted> jorendorff: no, khuey fixed that
# [18:43] <ted> in one of the dep bugs
# [18:43] <ted> jorendorff: last i knew catlee had shown that it was a significant build speed win
# [18:44] * philor is now known as philor|away
# [18:44] * Joins: mccr8 (mccr8@moz-BBE3ABD.mv.mozilla.com)
# [18:44] * Quits: pranavrc (pranavrc@2282E225.2B2E4944.C28326FD.IP) (Ping timeout)
# [18:44] <jorendorff> why is it a speed win, again? just because make.py -j6 is faster than gnumake -j1?
# [18:45] <ted> that, plus pymake is faster than gnu make on windows
# [18:45] <bsmedberg> jorendorff: no actually pymake is straight-up faster
# [18:45] <ted> because it doesn't actually recurse, and it avoids shell invocations when possible
# [18:45] <bsmedberg> because we can reduce the number of processes by a couple thousand
# [18:45] <Callek> edmorley: ping?
# [18:45] <bsmedberg> and processes are expensive on windows
# [18:45] <bsmedberg> actually probably more than that now given pymake native commands
# [18:45] <jorendorff> ok, but so
# [18:45] * Joins: hub (hub@21B7B9F2.B87E9213.6E712CE2.IP)
# [18:46] <jorendorff> it's just too bad there isn't a lot of interest in pymake outside moz
# [18:46] <ted> there's not much interest in make in general :-P
# [18:46] <ted> people use it because it more or less does its job
# [18:46] <bsmedberg> well, I get patches from outside contributors
# [18:47] <ted> and the set of people using make on windows for large projects is not very large
# [18:47] <bsmedberg> but the segment of people who need make to work really well on Windows is pretty small
# [18:47] <gps> bsmedberg: can you put the official PyMake repo on GitHub?
# [18:47] <bsmedberg> aww, ted beat me to it!
# [18:47] <ted> yeah
# [18:47] <ted> :)
# [18:47] <ted> google wrote their own meta-build system to avoid this problem
# [18:47] <bsmedberg> gps: I haven't used git, that's why it's in a hg repo
# [18:47] <ted> speaking of
# [18:47] <ted> bsmedberg: i've been forcibly poking at gyp for a while here
# [18:48] <jcranmer> you know what we really need?
# [18:48] <ted> i like it in theory, but the usage leaves a lot to be desired
# [18:48] <bsmedberg> hghub?
# [18:48] <jcranmer> a way to automatically mirror some of our hg stuff onto github
# [18:48] <bsmedberg> ted: yeah...
# [18:48] <bsmedberg> ted: you want it mainly for the compilation stuff, right?
# [18:48] <hub> I am not a mercurial command
# [18:48] * bsmedberg still wants to move all the export phase into a non-make environment
# [18:48] <ted> bsmedberg: i'm not sure i want it at all
# [18:49] <ted> i like the theory of it as a descriptive language for your build
# [18:49] <ted> but in practice, all the gyp files get really hairy
# [18:49] <bsmedberg> well, I want "something that makes MSVC projects" ;-)
# [18:49] * Quits: bretr (bret_recka@A73CA40C.E7870264.CFAD1666.IP) (Quit: bretr)
# [18:49] <Waldo> hub: really? that's bizarre
# [18:49] <gps> if we switch all the build systems to PyMake, you can focus on building out the core logic in Python then slowly switch out make
# [18:49] <ted> and you wind up with autoconf-like junk stuffed inside
# [18:49] <ted> gps: it's true
# [18:49] <bsmedberg> gps: could be, but people really don't want make or pymake
# [18:49] <ted> bsmedberg: generating MSVC projects is a neat trick
# [18:49] <bsmedberg> they want MSVC and xcode
# [18:50] <ted> and on linux it generates ninja-files now
# [18:50] * wlach is now known as wlach|bbiab
# [18:50] <ted> which looks neat
# [18:50] <bsmedberg> ninja?
# [18:50] <ted> a googler's make-replacement
# [18:50] <ted> engineered to be very dumb but very fast
# [18:50] <jcranmer> there's cmake... which has problems
# [18:50] <bsmedberg> heh, we've all got one
# [18:50] <ted> http://neugierig.org/software/chromium/notes/2011/02/ninja.html
# [18:50] <gps> bsmedberg: you are talking to the person who has code to generate visual studio projects from our makefiles :)
# [18:50] <johanc> evening, I'm getting "an unexpected uncaught JS exception reported through window.onerror - is is not defined ..." when running my mochitest, any chance of some help? :)
# [18:50] * mcote is now known as mcote|bbiab
# [18:50] <ted> bsmedberg: it's built with the expectation that you are generating your build files from something like GYP
# [18:50] <gps> ted: ninja is just a builder: you still need source of truth in some other format
# [18:50] <ted> so it's not very expressive
# [18:51] <ted> gps: right, but it's a neat backend if you have GYP on the frontend
# [18:51] <gps> yup
# [18:51] <gps> this goes back to my idea - more declarative build system files that translate to other build systems (msvc, xcode, make, ninja, etc)
# [18:52] <ted> yeah, i'm just not sure i'm in love with gyp as the way to do this
# [18:52] <ted> maybe gyp with some augmentations
# [18:52] <gwagner> aurora can't run the v8 benchmarks any more? ->Content Encoding Error
# [18:52] <gps> we can get surprisingly far today by parsing Makefiles into that IL
# [18:52] * Joins: Mook_as (mook@moz-1FCC0032.activestate.com)
# [18:52] <ted> i would love a more configure-like step in front of it
# [18:52] <jcranmer> most "modern" build systems seem to go the route of "configure in them, build with msvc/xcode/make"
# [18:52] <bsmedberg> yeah, gyp support for feature stuff is pretty horrid
# [18:53] <ted> jcranmer: right, which is what gyp does
# [18:53] <ted> except for feature tests it's horrid
# [18:53] <ted> you just wind up with shit like includes="<!(pkg-config --cflags gtk+-2.0)"
# [18:53] * Quits: c0smikdebris (c0smikdebr@C4166EFD.7E460B98.BE90E62C.IP) (Ping timeout)
# [18:53] <Waldo> somebody explain to me the point of [needs review]? why is it not derivable from other bug state?
# [18:53] * Quits: hhillen (hhillen@moz-81563C98.ip.telfort.nl) (Quit: hhillen)
# [18:53] * Joins: jrmuizel (jrmuizel@F2D29657.F60B0462.67AC9B1.IP)
# [18:53] <jwir3> is bugzilla really really slow for anyone other than me?
# [18:53] <bsmedberg> "But early on in development I discovered that Scons, despite its admirable goals of correctness and ease of use, was quite slow"
# [18:54] <bsmedberg> no kidding
# [18:54] <bsmedberg> jwir3: it has been variably slow all morning, yeah
# [18:54] <ted> jwir3: apparently we've been getting DDoSed for days
# [18:54] <glob> jwir3, it's no slower than normal for me right now
# [18:54] <jwir3> hm, ok
# [18:54] * Joins: c0smikdebris (c0smikdebr@C4166EFD.7E460B98.BE90E62C.IP)
# [18:54] <bsmedberg> Waldo: whiteboard status are sometimes useful for people to scan bug lists, since bug lists don't have a simple way of including attachment state by default
# [18:55] <jwir3> DDoSed? Who would want to do that? The safari team?
# [18:55] <jwir3> (j/k)
# [18:55] <Waldo> hm, and not just a different list? oddment
# [18:55] <ted> jwir3: bored script kiddies?
# [18:55] * Quits: magsout (magsout@moz-E559D13.fbx.proxad.net) (Ping timeout)
# [18:55] <ted> who knows
# [18:55] <@bz> jwir3: twitter at least part of yesterday, apparently
# [18:55] <@bz> jwir3: or so it was rumored
# [18:55] * NeilAway doesn't want visual studio projects
# [18:56] <jcranmer> considering how much of our core makefile stuff is factored out
# [18:56] <Ziggy_Maes> NeilAway: Funny: I was just thinking the same
# [18:56] * tbsaunde wants nothing to do with $IDE
# [18:56] <johanc> http://pastebin.mozilla.org/1443485 <- le failing test
# [18:56] <jcranmer> we could probably figure out how to convert them into mozconfigure or something
# [18:57] <ted> jcranmer: gps already wrote a parser based on pymake that generates VS projects
# [18:57] <gps> https://github.com/indygreg/mozilla-central/tree/build-splendid
# [18:57] <edmorley> bz: mozilla.dev.identity for browserid aiui
# [18:57] <edmorley> callek: pong
# [18:57] <gps> that's in a state of broken right now
# [18:58] <edmorley> Callek: re coordinating 701371 and merging yeah?
# [18:58] <@bz> edmorley: ah, thanks
# [18:58] <Callek> edmorley: yea....
# [18:58] * Waldo wants nothing to do with vim, that doesn't mean other people are equally smart :-P
# [18:58] <jcranmer> ted: but then we could make it our build system and release it to the world! :-)
# [18:58] * Joins: jdm (jdm@moz-F200AB68.uwaterloo.ca)
# [18:58] <Callek> edmorley: so, land 701371 when you merge... and I can hand you the urls for sea and TB clobberer if it helps you :-)
# [18:58] <Waldo> IDEs matter
# [18:58] * jlebar has never met an IDE which understands our code.
# [18:59] <Ms2ger> Nah, gedit is fine
# [18:59] <@bz> jlebar: have you met a human who does? ;)
# [18:59] <jlebar> But maybe if you rarely use interfaces, it's OK?
# [18:59] <edmorley> Callek: I don't have a c-c tree any more, but I'll have a quick star now, pick a cset, and let you know when I'm about to merge if that wfu?
# [18:59] * Quits: mbrubeck (mbrubeck@moz-755AD63.hsd1.wa.comcast.net) (Ping timeout)
# [18:59] <Callek> edmorley: if you'll be merging soonish that WFM
# [19:01] * Quits: dbradley (dbradley@moz-8FACAA93.fuse.net) (Ping timeout)
# [19:01] * davidb|afk is now known as davidb
# [19:01] * Joins: cpeterson (cpeterson@moz-C03D0C61.vlan426.asr1.sfo1.gblx.net)
# [19:01] * Joins: gkw (gkw@moz-A7D8CA2A.hsd1.ca.comcast.net)
# [19:02] * Joins: stevee (Miranda@moz-BEBDF855.cable.virginmedia.com)
# [19:02] * Joins: dbradley (dbradley@moz-8FACAA93.fuse.net)
# [19:02] * bear-afk is now known as bear
# [19:02] * Joins: azakai (alon@moz-BBE3ABD.mv.mozilla.com)
# [19:02] * Quits: tH (Rob@cpc16-seac19-2-0-cust234.7-2.cable.virginmedia.com) (Ping timeout)
# [19:03] <jdm> jhammel: bugs ahoy has deep linking now :)
# [19:03] <Ms2ger> <ted> a googler's make-replacement
# [19:03] <Ms2ger> <ted> engineered to be very dumb but very fast
# [19:03] <jhammel> jdm: i noticed! awesome :)
# [19:03] * Joins: tH (Rob@cpc16-seac19-2-0-cust234.7-2.cable.virginmedia.com)
# [19:03] <Ms2ger> Isn't that how they all start?
# [19:03] <ted> heh
# [19:03] <ted> Ms2ger: not necessarily
# [19:03] <jhammel> jdm: i've already fixed my permalink
# [19:03] <jdm> excellent!
# [19:04] <Ms2ger> Well, I guess you can start out dumb and slow too
# [19:04] <jhammel> ted: some start off very dumb and very slow? ;)
# [19:04] <Waldo> jdm: speaking of bugs ahoy, you might want to have it recognize [lang=c] and maybe make a C/C++ grouping
# [19:04] <jdm> Waldo: I'm pretty sure it does recognize lang=c at this point
# [19:04] <Waldo> oh really?
# [19:04] <jdm> but I'll update the actual text in the list to represent that
# [19:04] <Ms2ger> lang=c and good first bug?
# [19:04] <Ms2ger> Isn't that an oxymoron?
# [19:04] <ehsan> Ms2ger: will you watch the try results for bug 716215?
# [19:04] <Ms2ger> I can
# [19:05] * rail-lunch is now known as rail
# [19:05] <Ms2ger> ehsan, looks good, still
# [19:05] <edmorley> callek: merging as soon as i've submitted the clobberer page
# [19:06] * mdas is now known as mdas|mtg
# [19:06] <@bz> ms2ger: why?
# [19:06] <ehsan> Ms2ger: thanks :)
# [19:06] * Joins: juanb (jbecerra@moz-BBE3ABD.mv.mozilla.com)
# [19:06] <jdm> Waldo: actually, the text already says C/C++ :)
# [19:06] <Waldo> hm, another change?
# [19:07] <Ms2ger> bz, I don't like C? :)
# [19:07] <@bz> ms2ger: C is easy
# [19:07] * Joins: bsmith (bsmith@moz-E34A3D24.dsl.pltn13.sbcglobal.net)
# [19:07] * Joins: fabrice1 (fabrice@moz-BBE3ABD.mv.mozilla.com)
# [19:07] <@bz> ms2ger: at least it seemed fine coming from basic
# [19:07] * Joins: Mano (chatzilla@moz-D0BA9E6D.red.bezeqint.net)
# [19:07] <Ms2ger> Heh
# [19:07] <Waldo> C is for cookie...omnomnom
# [19:07] * Joins: Sander (chatzilla@moz-B871F4D3.direct-adsl.nl)
# [19:07] <@bz> ms2ger: once someone explained pointers to me
# [19:07] <@bz> (someone == my dad, iirc)
# [19:07] * Ms2ger came from... Pascal?
# [19:07] * Quits: fabrice1 (fabrice@moz-BBE3ABD.mv.mozilla.com) (Max SendQ exceeded)
# [19:07] <@bz> well, I guess basic and logo
# [19:08] <@bz> and two-column proofs
# [19:08] * glob is now known as glob|away
# [19:08] <@bz> which are useless as proofs, but surprisingly ok as an intro to algorithms, sorta
# [19:09] * Joins: jduell (jduell@moz-F20EC42A.hfc.comcastbusiness.net)
# [19:09] * Quits: azakai (alon@moz-BBE3ABD.mv.mozilla.com) (Ping timeout)
# [19:09] * Joins: bretr (bret_recka@A73CA40C.E7870264.CFAD1666.IP)
# [19:09] <@bz> teach you the whole methodical step by step thing
# [19:09] * Joins: fabrice1 (fabrice@moz-BBE3ABD.mv.mozilla.com)
# [19:09] * Ms2ger looks that up
# [19:09] <Callek> edmorley: ping me as soon as you push, I'll clobber and push myself then :-)
# [19:09] * catlee-lunch is now known as catlee
# [19:10] * jhammel is now known as jhammel|mtg
# [19:10] * Joins: zandr (zandr@moz-891BD824.milewski.org)
# [19:10] * Quits: joey (chatzilla@moz-EFCB4CBF.princetowncable.com) (Ping timeout)
# [19:11] * Joins: sworkman (sworkman@moz-BBE3ABD.mv.mozilla.com)
# [19:11] <edmorley> Callek: pushed :-)
# [19:11] * Quits: chrisccoulson (chr1s@moz-B376CDB0.business.broadband.hu) (Ping timeout)
# [19:11] * Joins: bholley (bholley@moz-FCAF9AAB.hsd1.ca.comcast.net)
# [19:11] * Quits: jdm (jdm@moz-F200AB68.uwaterloo.ca) (Client exited)
# [19:12] * Quits: jimb (user@moz-F4EC06CC.hsd1.or.comcast.net) (Input/output error)
# [19:12] * Joins: terrence (terrence@moz-BBE3ABD.mv.mozilla.com)
# [19:12] * Quits: andreasn (andreasn@moz-BB6A8755.bredband.comhem.se) (Ping timeout)
# [19:12] * Quits: mccr8 (mccr8@moz-BBE3ABD.mv.mozilla.com) (Quit: mccr8)
# [19:12] * Joins: joey (chatzilla@moz-EFCB4CBF.princetowncable.com)
# [19:13] <Callek> edmorley: same, so hopefully all is good :-)
# [19:13] <edmorley> hi bholley
# [19:13] <edmorley> Callek: thanks :-)
# [19:13] <bholley> hi edmorley!
# [19:13] * coop is now known as coop|lunch
# [19:14] * Quits: madhava (madhava@F2D29657.F60B0462.67AC9B1.IP) (Quit: madhava)
# [19:14] * Joins: madhava (madhava@F2D29657.F60B0462.67AC9B1.IP)
# [19:14] * Joins: mccr8 (mccr8@moz-BBE3ABD.mv.mozilla.com)
# [19:15] <Ms2ger> Hi bholley!
# [19:15] * Joins: cpearce (chatzilla@moz-56BC07D4.xdsl.xnet.co.nz)
# [19:15] <bholley> hi Ms2ger!
# [19:15] * Joins: andreasn (andreasn@moz-BB6A8755.bredband.comhem.se)
# [19:15] <Ms2ger> You may have noticed some patches in your review queue :)
# [19:16] * Joins: ashughes_ (ashughes@1B3CF998.EDC95493.A20E6FDD.IP)
# [19:16] <edmorley> ted: do we have a bug filed for this? https://tbpl.mozilla.org/php/getParsedLog.php?id=8450024&full=1&branch=mozilla-inbound#error0
# [19:16] <edmorley> I can't seem to find one
# [19:17] * Joins: lsumar (lsumar@4548E2C6.EE84D258.11F528CC.IP)
# [19:17] <Ms2ger> Don't think so
# [19:17] * Quits: ashughes (ashughes@83D384FE.DFA2E3D3.C3C3A9D0.IP) (Ping timeout)
# [19:17] * ashughes_ is now known as ashughes
# [19:17] <@bz> bholley: you saw Ms2ger's mail about DOM classes?
# [19:17] * armenzg_lunch is now known as armenzg
# [19:17] <bholley> bz: yeah
# [19:17] <bholley> bz: looks like I should start with bug 563659
# [19:18] <@bz> bholley: well, so....
# [19:18] <@bz> bholley: let's talk about this
# [19:18] <Ms2ger> (I was thinking of looking at that, but do go ahead :))
# [19:18] <@bz> bholley: how we approach this depends on the actualy binding unwrapping mechanism
# [19:18] <@bz> bholley: er, on the actual
# [19:18] <bholley> Ms2ger: if you want to look at it go for it - I'm pretty swamped
# [19:19] <@bz> bholley: the way things work right now is that we get the nsisupports and then upcast
# [19:19] * zpao|detached is now known as zpao
# [19:19] <@bz> bholley: and the casting method we use works such that if X inherits from Y then an nsISupports pointer to X will always unwrap to Y
# [19:19] <Ms2ger> bholley, that took long :)
# [19:19] <ted> edmorley: not that i've seen
# [19:19] <@bz> bholley: (I don't recall exactly how peterv set that up, but that's the effect)
# [19:19] * Quits: mjschranz (chatzilla@moz-6FE6B833.cpe.net.cable.rogers.com) (Ping timeout)
# [19:20] <@bz> bholley: agreed so far?
# [19:20] <bholley> bz: so this is in the QS unwrapping code?
# [19:20] <@bz> bholley: yes
# [19:20] <@bz> bholley: the castNative stuff, iirc
# [19:20] <bholley> bz: and we upcast, rather than QI?
# [19:21] <Ms2ger> In some cases
# [19:21] <@bz> bholley: for some unwrap targets, yes
# [19:21] * Joins: biesi (cbiesinger@moz-76ED95A4.corp.tfbnw.net)
# [19:21] <@bz> bholley: the ones explicitly whitelisted
# [19:21] <bholley> bz: in qsconf?
# [19:21] <@bz> bholley: http://mxr.mozilla.org/mozilla-central/source/dom/base/nsDOMClassInfoID.h#67
# [19:21] <@bz> bholley: in there
# [19:21] * Quits: MarcoZ (marco.zehe@moz-97ABA480.dip.t-dialin.net) (Quit: l8er)
# [19:22] <@bz> bholley: though qsconf needs to be involved too, since it needs to say that we want to unwrap to something in this list
# [19:22] <bholley> bz: oh right, those
# [19:22] <@bz> ok
# [19:22] * Joins: bent (chatzilla@moz-C03D0C61.vlan426.asr1.sfo1.gblx.net)
# [19:22] <@bz> so that's the current setu
# [19:22] <@bz> er, setup
# [19:23] * Quits: Jesse (jruderman@moz-537BCF9.hsd1.ca.comcast.net) (Quit: Jesse)
# [19:23] <@bz> for the new world, I was considering something a bit different maybe
# [19:23] <@bz> specifically, as follows:
# [19:24] * Joins: azakai (alon@moz-BBE3ABD.mv.mozilla.com)
# [19:24] <@bz> 1) There is a distinct JSClass for each (most derived prototype, impl class) pair
# [19:24] * Joins: jimb (user@moz-F4EC06CC.hsd1.or.comcast.net)
# [19:24] <@bz> (in the common case a most derived prototype only has one impl class anyway)
# [19:24] * Joins: beaufour (beaufour@moz-723527FF.nycmny.fios.verizon.net)
# [19:24] <@bz> 2) Off this jsclass we hang a set of bits describing which types impl class can cast to
# [19:25] <@bz> that is, types that sit on the inheritance chain between the impl class and its primary nsISupports; the jsobject has a pointer to this last
# [19:25] * Quits: pascalc (chatzilla@moz-A2A3C9E6.vlan402.asr1.cdg3.gblx.net) (Ping timeout)
# [19:25] * Quits: flx_ (flx@590D15DC.A7CE92B2.30E47D40.IP) (Ping timeout)
# [19:25] * davehunt is now known as davehunt|away
# [19:26] <@bz> 3) when unwrapping to type Foo, check the jsclass for the corresponding bit; if it's present just cast, else unwrapping fails
# [19:26] * Joins: rs (rs@moz-BBE3ABD.mv.mozilla.com)
# [19:26] * Quits: lsumar (lsumar@4548E2C6.EE84D258.11F528CC.IP) (Quit: This computer has gone to sleep)
# [19:26] <bholley> bz: makes sense
# [19:26] <@bz> The drawback is that we need one bit per DOM most-derived thing, basically
# [19:26] <@bz> so a few hundred bit
# [19:26] <@bz> er, bits
# [19:26] <@bz> and checking them will therefore take a bit of wrangling (e.g. we can't just do a simple shift-by-the-enum-value thing)
# [19:27] * Quits: asadotzler (asa@F259CFB7.9EF26CBA.EB5E51FC.IP) (Ping timeout)
# [19:27] <@bz> the other drawback is that each DOM class has to make sure to set up its bits correctly
# [19:27] <@bz> on the jsclass
# [19:27] <Ms2ger> Bigints? :)
# [19:27] <@bz> the benefit is that nsDocumentFragment can just not set the "element" bit and it'll all work
# [19:27] * Joins: akeybl (akeybl@moz-BBE3ABD.mv.mozilla.com)
# [19:28] <@bz> in that it'll set the "node" bit (node stuff will unwrap to nsINode)
# [19:28] <bholley> bz: hmmm
# [19:28] * Quits: jimb (user@moz-F4EC06CC.hsd1.or.comcast.net) (Ping timeout)
# [19:28] <@bz> and the "document fragment" bit (with document fragment proto methods unwrapping directly to nsDocumentFragment)
# [19:28] <Ms2ger> (Nothing there)
# [19:29] * Quits: past (past@moz-2DAAD734.dsl.dyn.forthnet.gr) (Input/output error)
# [19:29] <@bz> sure
# [19:29] <@bz> but conceptually
# [19:29] * Quits: pcwalton (pcwalton@moz-7B0110AD.mv.mozilla.com) (Quit: pcwalton)
# [19:29] <@bz> bholley: so obvious questions:
# [19:29] * Joins: jimb (user@moz-104CC309.mv.mozilla.com)
# [19:29] * Joins: kumar (kmcmillan@moz-D0C71163.domainhotel-wifi.mozilla.hq)
# [19:29] <@bz> 1) Is what I describe above sane?
# [19:30] <@bz> 2) Does it let us handle the cases Ms2ger found easily?
# [19:30] <bsmedberg> josh: ok, I have a testcase in mochitest form
# [19:30] <bsmedberg> I'll attach it to the bug?
# [19:30] * Joins: Boriss (FlyingToas@moz-C03D0C61.vlan426.asr1.sfo1.gblx.net)
# [19:30] * Quits: kaze (kaze@moz-A2A3C9E6.vlan402.asr1.cdg3.gblx.net) (Ping timeout)
# [19:31] <@bz> bholley: and maybe 3) How many bits will we end up with?
# [19:31] <bholley> bz: the bit system will leave us with a lot of goop, but i guess that's ok
# [19:31] * Joins: camd (camerondaw@moz-D00736A2.hsd1.ca.comcast.net)
# [19:31] <@bz> bholley: what sort of goop?
# [19:31] <bholley> bz: especially if we're code-gen-ing
# [19:31] <josh> bsmedberg: that would be great, thanks! patch format so i can just apply to my tree would be best
# [19:32] <@bz> bholley: there's the manual goop to set the right bits, possibly
# [19:32] <@bz> bholley: I agree that's a concern
# [19:32] <@bz> fwiw
# [19:32] * Joins: jruderman (jruderman@moz-537BCF9.hsd1.ca.comcast.net)
# [19:32] <@bz> mozilla% find . -name "nsIDOM*Element.idl" | wc -l
# [19:32] <@bz> 115
# [19:32] * Joins: hhillen (hhillen@moz-2A42E9B3.upc-d.chello.nl)
# [19:32] <@bz> mozilla% find . -name "nsIDOM*.idl" | wc -l
# [19:32] <@bz> 356
# [19:32] <bholley> bz: I'm wondering if there's a more general way to do it. But maybe anything else would be less performant
# [19:32] * Joins: pcwalton (pcwalton@moz-7B0110AD.mv.mozilla.com)
# [19:33] <@bz> So we'll definitely need at least 128 bits
# [19:33] <@bz> bholley: well, that's a good question
# [19:33] * Quits: jruderman (jruderman@moz-537BCF9.hsd1.ca.comcast.net) (Quit: jruderman)
# [19:33] * Joins: jdm (jdm@moz-F200AB68.uwaterloo.ca)
# [19:33] * Joins: Jesse (jruderman@moz-537BCF9.hsd1.ca.comcast.net)
# [19:33] * Quits: jhk (jiggy@8E6C34C1.A3F9767A.1C37C358.IP) (Ping timeout)
# [19:33] <@bz> bholley: in some ways, we're ok with having a more-performant way that works for nodes and a few other things and a slower more general way, if it makes things simpler
# [19:33] <@bz> bholley: I'm just not sure it does
# [19:33] <@bz> and probably at most 512 bits?
# [19:34] <Ms2ger> [For now]
# [19:34] <@bz> Ms2ger: sure
# [19:34] * bholley is wondering if we could somehow just hang the iids off the jsclass. But I guess that would only help for QI
# [19:34] <@bz> Ms2ger: do you expect more than 150 new DOM proto object things to be addded in the near future
# [19:34] <bholley> and it could be heavier
# [19:34] * jimm-lunch is now known as jimm
# [19:34] * mdas|mtg is now known as mdas
# [19:34] * Joins: sfink (chatzilla@moz-BBE3ABD.mv.mozilla.com)
# [19:34] <@bz> bholley: right, checking for a single bit in a set of 128 or 512 seems strictly better than a bunch of 128-bit compares
# [19:35] <Ms2ger> Before we rewrite all this again, you mean? :)
# [19:35] <@bz> bholley: we can totally hang IIDs off the jsclass
# [19:35] <bholley> bz: do we really need all the permuations though?
# [19:35] <@bz> bholley: well, no
# [19:35] <@bz> bholley: but it would simplify codegen
# [19:35] <@bz> bholley: otherwise we need a concept of which interfaces can or can't be implemented on the same object
# [19:35] <@bz> bholley: and even then, we might run into issues
# [19:36] <@bz> bholley: so thinking about it more, we do need all permutations, sorta
# [19:36] * Joins: bnicholson (bnicholson@moz-BBE3ABD.mv.mozilla.com)
# [19:36] <bholley> bz: I mean, with codegen, we could just burn things in directly somehow
# [19:36] * Quits: florian (florian@moz-87C33FDA.kimsufi.com) (Quit: Instantbird 1.2a1pre)
# [19:36] <@bz> bholley: or more precisely, we need to be able to definitively say whether any given DOM object casts to any particular class
# [19:36] <bholley> bz: with IDs, rather than bits
# [19:36] <@bz> bholley: not sure what you mean
# [19:36] * Quits: Jesse (jruderman@moz-537BCF9.hsd1.ca.comcast.net) (Quit: Jesse)
# [19:36] * Joins: Jesse (jruderman@moz-537BCF9.hsd1.ca.comcast.net)
# [19:37] <bholley> bz: well, if we have nsIBaz : public nsIBar : public nsIFoo, we could either have one bit for each of those, or just call that explicit combination "42"
# [19:37] * Quits: gkw (gkw@moz-A7D8CA2A.hsd1.ca.comcast.net) (Quit: Instantbird 1.1)
# [19:37] * Quits: jimb (user@moz-104CC309.mv.mozilla.com) (Ping timeout)
# [19:38] * Joins: cjones (cjones@moz-45913895.socal.res.rr.com)
# [19:38] * Joins: jimb (user@moz-E31CD2CB.mozilla.org)
# [19:38] <@bz> bholley: so in general, when we have that setup, we will have some methods on Baz.prototype, some on Bar.prototype, and some on Foo.prototype
# [19:38] * bholley notes that this would probably increase codesize a lot
# [19:38] <@bz> bholley: and the Bar.prototype methods will need to work on anything that inherits from nsIBar, whether it's an nsIBaz or not
# [19:39] <bholley> bz: but don't we know explicitly all the things that inherit from nsIBar?
# [19:39] <@bz> bholley: so I was thinking we would implement that by having a concrete nsBarImpl class corresponding to nsIBar
# [19:39] * Joins: clee (clee@moz-D0C71163.domainhotel-wifi.mozilla.hq)
# [19:39] <@bz> bholley: and a way for an object to indicate it's castable to nsBarImpl
# [19:39] <@bz> bholley: in general, yes
# [19:40] <bholley> bz: so we _could_ enumerate all the combinations, and it could be less than 512 bits
# [19:40] <@bz> yes
# [19:41] <@bz> then the Bar.prototype methods would need to check the number hanging off the JSClass off a list of all possible things that could inherit from BAr
# [19:41] <@bz> yes?
# [19:41] <bholley> bz: hm, yeah....
# [19:42] * @bz is thinking the case when Bar == Node
# [19:42] * Quits: bretr (bret_recka@A73CA40C.E7870264.CFAD1666.IP) (Quit: bretr)
# [19:42] <@bz> which is sorta a common case
# [19:42] <bholley> bz: well, we could just do that checking in the unwrap code, no?
# [19:42] * Joins: rniwa (rniwa@moz-E171DA5.sfba.hfc.comcastbusiness.net)
# [19:42] <@bz> we're talking about the unwrap code right now!
# [19:43] * Quits: rniwa (rniwa@moz-E171DA5.sfba.hfc.comcastbusiness.net) (Client exited)
# [19:43] <bholley> bz: oh, ok. I'm imagining a setup where we have a lookup table from the ID combination to nsBarImpl
# [19:43] * Joins: rniwa (rniwa@60A74940.D6CCE4AE.77834EAA.IP)
# [19:43] * Quits: Enn (enn@F2D29657.F60B0462.67AC9B1.IP) (Ping timeout)
# [19:43] <@bz> ok
# [19:43] <@bz> so basically each DOM proto class has a corresponding hashtable
# [19:44] <@bz> which maps ID combinations to booleans (does/doesn't do this dom proto)?
# [19:44] <mayhemer> josh: do you think it is a good idea to talk about the SFO meeting at a public phone call?
# [19:44] * Quits: fabrice1 (fabrice@moz-BBE3ABD.mv.mozilla.com) (Ping timeout)
# [19:44] * wlach|bbiab is now known as wlach
# [19:44] * bholley thinks
# [19:45] * mdas is now known as mdas|afk
# [19:46] * Quits: anky (anky@27B0D94A.B421193.A3D1B221.IP) (Client exited)
# [19:47] * Joins: Ami_Ty (Amie@moz-7B0110AD.mv.mozilla.com)
# [19:47] * Joins: northAway (northWind@moz-20E58859.cable.teksavvy.com)
# [19:48] * Joins: rshetty (quassel@3797C315.FB55C362.D30E9BEF.IP)
# [19:48] <bholley> bz: I was envisioning more of a scenario where we generate lots of stubs
# [19:48] <bholley> bz: but I don't really have a great picture of how it would work
# [19:49] * Quits: pcwalton (pcwalton@moz-7B0110AD.mv.mozilla.com) (Quit: pcwalton)
# [19:49] <bent> bz, hey, got a sec?
# [19:49] <@bz> bholley: so what stubs would you generate in the above situation?
# [19:49] <@bz> bent: sure
# [19:49] <bent> bz, i need a new NodeList-ish type thing for telephony
# [19:50] <bent> bz, are there any docs or anything on how to modify the codegen stuff?
# [19:50] * Joins: pcwalton (pcwalton@moz-7B0110AD.mv.mozilla.com)
# [19:50] <@bz> bent: other than the code itself and the bugs, no
# [19:50] <bent> or do i just add something to the
# [19:50] <bent> ok
# [19:50] * Joins: Boriss_ (FlyingToas@moz-C03D0C61.vlan426.asr1.sfo1.gblx.net)
# [19:50] * bent digs in
# [19:51] <@bz> bent: if you write down things as you go, I'm happy to read it over after. ;)
# [19:51] <bholley> bz: per-concrete-class stubs
# [19:51] <bent> heh ok
# [19:51] * Joins: gabor_ (admin@moz-3B57BCD1.catv.pool.telekom.hu)
# [19:51] <bholley> bz: but I think that's just moving the memory burden over to codesize, probably not worth it
# [19:51] <@bz> bholley: I don't think that's doable
# [19:51] <@bz> bholley: think Node.prototype.appendChild
# [19:52] <@bz> bholley: this needs to correspond to a single C++ method, right?
# [19:52] * Quits: Boriss (FlyingToas@moz-C03D0C61.vlan426.asr1.sfo1.gblx.net) (Ping timeout)
# [19:52] * Boriss_ is now known as Boriss
# [19:52] <@bz> bholley: and needs to work for any |this| that's a Node
# [19:52] <@bz> bholley: in particular, Node.prototype.appendChild.call(someNode, someChild) should work
# [19:52] * Quits: sworkman (sworkman@moz-BBE3ABD.mv.mozilla.com) (Quit: sworkman)
# [19:53] * Quits: gabor (admin@moz-3B57BCD1.catv.pool.telekom.hu) (Ping timeout)
# [19:53] * gabor_ is now known as gabor
# [19:53] * Joins: sworkman (sworkman@moz-BBE3ABD.mv.mozilla.com)
# [19:53] <bholley> bz: well, nsBarImpl could have a stub that would do the right thing. But I agree that it's not practical
# [19:53] * Quits: andreasn (andreasn@moz-BB6A8755.bredband.comhem.se) (Ping timeout)
# [19:53] * Joins: rclick (rclick@EA4B602E.692CFA54.E430BA25.IP)
# [19:53] * Quits: sworkman (sworkman@moz-BBE3ABD.mv.mozilla.com) (Input/output error)
# [19:53] * Joins: sworkman (sworkman@moz-BBE3ABD.mv.mozilla.com)
# [19:54] <@bz> bholley: that just pushes off the problem, I think
# [19:54] <@bz> bholley: if I understand the issue correctly
# [19:54] <@bz> bholley: in the end, to handle Foo.prototype.fooMethod.call(someObject) we need to be able to tell whether that object is a Foo
# [19:54] <bholley> bz: it's more or less the hashtable idea, but reversing the mapping
# [19:54] <@bz> and to do it reasonably quickly
# [19:54] <bholley> bz: given a concrete class, we have a list of the interfaces it implements
# [19:55] <bent> bz, will it ever be possible with your idea to do something like this: var MyObject = { __proto__: EventTarget.prototype }; var obj = new MyObject(); obj.addEventListener(...) ?
# [19:55] <bholley> bz: rather than the reverse
# [19:55] * Joins: flx_ (flx@moz-F4992FEC.uwaterloo.ca)
# [19:55] * Joins: Ami_Ty_ (Amie@moz-BBE3ABD.mv.mozilla.com)
# [19:55] <@bz> bent: absolutely not
# [19:55] <bent> bz, i.e. implement something in JS that derives from DOM things?
# [19:55] <bent> ok
# [19:55] * Quits: Ami_Ty (Amie@moz-7B0110AD.mv.mozilla.com) (Ping timeout)
# [19:55] * Ami_Ty_ is now known as Ami_Ty
# [19:55] <@bz> bent: in fact we allow that right now, and are trying to remove it
# [19:55] * Joins: andreasn (andreasn@moz-BB6A8755.bredband.comhem.se)
# [19:56] <@bz> bent: and webidl actually forbids it right now
# [19:56] * bholley is removing it
# [19:56] <@bz> bholley: hm
# [19:56] <bent> well... we ran into a bunch of prototyping stuff that would be great to do all from JS
# [19:56] <bent> but it doesn't work so we have to do it in C++
# [19:56] <bent> which i'm personally fine with, but others have a special love for js
# [19:56] <@bz> bent: prototyping as in "rapid prototyping"?
# [19:57] <@bz> bent: or as in "something with JS prototypes"?
# [19:57] <bent> yes, but if it had worked it could have stuck
# [19:57] <@bz> bholley: so say we have a list of interfaces for each concrete class
# [19:57] <bent> "rapid
# [19:57] <bent> "
# [19:57] <@bz> bholley: by "concrete" do you mean most-derived, or something else?
# [19:57] <bholley> bz: yrs
# [19:57] <bholley> bz: yes
# [19:57] <@bz> bholley: fine
# [19:57] <@bz> bholley: so...
# [19:57] <@bz> bholley: that might work
# [19:58] <@bz> bholley: each JSClass, instead of having bits, could just have a list of integer ids
# [19:58] <@bz> bholley: of the things that object can unwrap to
# [19:58] <@bz> bholley: and then each DOM proto has an integer that identifies which concrete class it corresponds to
# [19:58] <@bz> bholley: and instead of a bit test, you just walk the list doing == compares
# [19:59] <@bz> bholley: (if I understand what you're proposing)
# [19:59] * Joins: dRdR (dRdR@moz-3D8840A3.uwaterloo.ca)
# [20:00] <bholley> bz: I was imagining that it would just be one ID, i.e. the most-derived class
# [20:00] * Joins: fabrice1 (fabrice@moz-BBE3ABD.mv.mozilla.com)
# [20:00] <bholley> bz: oh wait
# [20:00] <@bz> bholley: so each jsclass has one id?
# [20:00] * Quits: newbie (kvirc@697AC6B3.8C6159FA.9A06DD32.IP) (Ping timeout)
# [20:00] <bholley> bz: I see what you're saying. Sure. (I was imagining this list would be in a stub, rather than the JSClass, but it doesn't matter)
# [20:00] <@bz> bholley: but then each proto object needs a list of most-derived classes that inherit from it
# [20:01] <@bz> so the "have a list" idea is not a bad one
# [20:01] <@bz> because typically the list is short
# [20:01] <@bz> for HTML elements, say, about 5 entries
# [20:01] <@bz> (HTMLFooElement, HTMLElement, Element, Node, EventTarget)
# [20:01] * Quits: cpearce (chatzilla@moz-56BC07D4.xdsl.xnet.co.nz) (Ping timeout)
# [20:01] * Quits: fabrice1 (fabrice@moz-BBE3ABD.mv.mozilla.com) (Quit: Leaving.)
# [20:01] <@bz> And if we use 16-bit IDs (should be plenty)
# [20:02] <@bz> that's 10 bytes
# [20:02] <@bz> and only 5 compares
# [20:02] <@bz> which may well be faster than bit-testing on 128+bit bitsets
# [20:02] <@bz> otherwise it's basically the same as the bitset idea
# [20:03] <bholley> bz: right
# [20:03] * Quits: andreasn (andreasn@moz-BB6A8755.bredband.comhem.se) (Ping timeout)
# [20:03] * Joins: lurking_work (chatzilla@moz-107FCDBA.hfc.comcastbusiness.net)
# [20:03] <@bz> bholley: so does that seem reasonable?
# [20:03] <@bz> bholley: if it does, then back to the "does this let us handle the cases ms2ger found?" question
# [20:04] <bholley> bz: sure
# [20:04] * Joins: andreasn (andreasn@moz-BB6A8755.bredband.comhem.se)
# [20:04] <bholley> bz: I thought the idea was that we'd fix the Ms2ger cases
# [20:04] * CwiiisAway is now known as Cwiiis
# [20:05] <camd> bc: ping
# [20:05] <bc> camd: pong
# [20:05] * Joins: asadotzler (asa@moz-BBE3ABD.mv.mozilla.com)
# [20:06] <camd> hi bc: smaug tells me you may have an addon that tracks gc and cc times and gathers them in a log?
# [20:06] <bc> camd: sort of.
# [20:06] <camd> I'm trying to help with a regression range on the GC and/or CC issues
# [20:06] * Quits: sworkman (sworkman@moz-BBE3ABD.mv.mozilla.com) (Ping timeout)
# [20:06] <camd> bc: what does it do?
# [20:06] * Quits: bsmith (bsmith@moz-E34A3D24.dsl.pltn13.sbcglobal.net) (Ping timeout)
# [20:06] <bc> http://bclary.com/projects/spider/spider/spider.xpi
# [20:06] <@bz> bholley: well
# [20:06] <@bz> bholley: the idea was to find what those cases are and see what was needed to fix them
# [20:07] <bc> camd: a general overview is at http://bclary.com/projects/spider/
# [20:07] <@bz> bholley: it's a much shorter list than I expected, actually
# [20:07] * Quits: jfkthame (jfkthame@43709DB9.5BC345F5.9542EC20.IP) (Quit: jfkthame)
# [20:07] * Joins: bsmith (bsmith@moz-E34A3D24.dsl.pltn13.sbcglobal.net)
# [20:07] <bholley> bz: do we know if it's comprehensive?
# [20:07] * Joins: sworkman (sworkman@moz-BBE3ABD.mv.mozilla.com)
# [20:07] * Quits: @dbaron (dbaron@moz-C03D0C61.vlan426.asr1.sfo1.gblx.net) (Quit: 8403864 bytes have been tenured, next gc will be global.)
# [20:07] <@bz> bholley: for dom nodes, it should be modulo ms2ger-error
# [20:07] * Quits: rshetty (quassel@3797C315.FB55C362.D30E9BEF.IP) (Connection reset by peer)
# [20:07] <@bz> bholley: (much like human error but without prejudging things)
# [20:07] <bholley> bz: heh
# [20:07] <bc> you need to set the pref to report the GC and CC times. If all you want to do is load pages then you can just kick off the spider and collect stdout/stderr to a log file. You'll probably want to enable dump as well.
# [20:08] <Ms2ger> bz, why thank you, people tend to be so insensitive to call me human without even knowing me
# [20:08] * Joins: jamesr (jamesr@D961E49B.D6CCE4AE.77834EAA.IP)
# [20:08] <@bz> ms2ger: ;)
# [20:09] * Joins: Enn (enn@F2D29657.F60B0462.67AC9B1.IP)
# [20:09] <camd> bc: ok, let me take a look. thanks!
# [20:09] * Joins: dbaron (dbaron@moz-C03D0C61.vlan426.asr1.sfo1.gblx.net)
# [20:09] * ChanServ sets mode: +o dbaron
# [20:09] * Joins: mbrubeck (mbrubeck@moz-3CE92E7A.tmodns.net)
# [20:10] <bholley> bz: I confess to not knowing much about DocumentFragment. From the spec though it looks like it should inherit nsINode though, right?
# [20:10] <bc> camd: if you want to perform some operations per web page that is loaded you can create a "userhook" set of functions which can be called prior to or after page loads
# [20:10] <Ms2ger> bholley, yep
# [20:10] <@bz> bholley: yes
# [20:10] <@bz> bholley: but really, as long as we prevent DocumentFragment from unwrapping to Element that's all we really need
# [20:10] <camd> bc: what would be an example of a userhook you might want?
# [20:10] <@bz> bholley: however we get there
# [20:10] * Quits: hhillen (hhillen@moz-2A42E9B3.upc-d.chello.nl) (Quit: hhillen)
# [20:11] <bholley> bz: so, we could just not set that bit
# [20:11] <catlee> ehsan: https://bugzilla.mozilla.org/show_bug.cgi?id=716229
# [20:11] <@bz> bholley: in my proposed setup, yes
# [20:11] * Joins: gkw (gkw@moz-BBE3ABD.mv.mozilla.com)
# [20:11] * Quits: ashughes (ashughes@1B3CF998.EDC95493.A20E6FDD.IP) (Quit: heading to the office)
# [20:11] <bc> camd: http://bclary.com/projects/spider/tests/objecthooks.js reports on embed and object tags for example. i didn't use any when i did those test runs for smaug though. just start/quit after each page.
# [20:12] <bholley> bz: and in the ID one, I suppose
# [20:12] * jhammel|mtg is now known as jhammel
# [20:13] <ehsan> catlee: nice!
# [20:15] * Quits: jdm (jdm@moz-F200AB68.uwaterloo.ca) (Client exited)
# [20:15] <@bz> bholley: yes
# [20:15] <jlebar> joe, ping?
# [20:15] <@bz> bholley: I think at this point I'm happier with the id setup than the bitfield
# [20:15] <joe> jlebar: in development mtg, can you reping?
# [20:15] <@bz> bholley: so "my proposed setup" is the id one, with a list of ids per jsclass
# [20:15] <jlebar> joe, yep.
# [20:16] <bholley> bz: and that list can be variable-length, since JSClasses can be variable-length, right?
# [20:16] <@bz> bholley: Well
# [20:16] <bholley> bz: (or rather, JSClasses can contain static lists of variable length)
# [20:16] <@bz> bholley: we can just subclass JSClass
# [20:17] <@bz> bholley: and add whatever thing we want to our subclass
# [20:17] <bholley> bz: sure
# [20:17] <@bz> bholley: and add a bit to jsclass that indicates that we really have one of our subclass instances
# [20:17] <bholley> bz: yeah, that all sounds good
# [20:18] <bholley> bz: and then we'd just have a config file that listed every concrete class and what interfaces it implements?
# [20:18] <@bz> ok
# [20:18] <@bz> bholley: that's one option
# [20:18] * philor|away is now known as philor
# [20:18] <@bz> bholley: the other option is that concrete class files would just have a table in them
# [20:18] * Quits: flx_ (flx@moz-F4992FEC.uwaterloo.ca) (Ping timeout)
# [20:19] <Ms2ger> One big config file? I'd hope it's autogenerated, then :)
# [20:19] * bholley likes doing things at compile-time, if possible
# [20:19] <@bz> bholley: the important part is that there be some sort of array of integers that's static that the static jsclass struct can point to
# [20:19] * coop|lunch is now known as coop|mtg
# [20:19] <@bz> bholley: if we can manage it, great
# [20:19] <bholley> bz: yeah
# [20:19] <@bz> bholley: much like classinfo now... ;
# [20:19] <Ms2ger> Bah, classinfo :)
# [20:20] <bholley> bz: we do need to allocate those integers somehow
# [20:21] * Joins: newbie (kvirc@697AC6B3.8C6159FA.9A06DD32.IP)
# [20:21] * Quits: mfinkle (mfinkle@moz-564248AF.hsd1.pa.comcast.net) (Quit: ZNC - http://znc.sourceforge.net)
# [20:21] * Quits: newbie (kvirc@697AC6B3.8C6159FA.9A06DD32.IP) (Quit: KVIrc 4.1.1 Equilibrium http://www.kvirc.net/)
# [20:22] <@bz> bholley: static InterfaceList[] = { .... } ; ?
# [20:22] <bholley> bz: no, I mean making sure that the IDs don't collide
# [20:22] <@bz> Ah, yes
# [20:22] <@bz> well
# [20:22] <bholley> bz: (allocation in integer-space, not memory-space)
# [20:22] <@bz> enum InterfaceList {
# [20:22] <@bz> }
# [20:22] <bholley> bz: so we'd list all the interfaces in one file?
# [20:22] <Ms2ger> #include "nsDOMClassInfo.cpp"
# [20:23] <@bz> bholley: seems like that's the obvious way to avoid collisions
# [20:23] * Joins: mfinkle (mfinkle@moz-564248AF.hsd1.pa.comcast.net)
# [20:23] <bholley> bz: yeah
# [20:23] <@bz> bholley: we can generate this file at compile time somehow if desired
# [20:23] <bholley> bz: but the bottom line is, it sounds like spec-aligning isn't necessarily a prereq for the new dom bindings
# [20:23] * Ms2ger desires
# [20:23] <@bz> bholley: right
# [20:24] <@bz> bholley: again, as long as the setup we just discussed is enough to handle the cases ms2ger found
# [20:24] <@bz> bholley: it can handle documentfragment; can it handle the others?
# [20:24] <askalski> hi, got simple questions
# [20:24] * bholley didn't look at the others too closely
# [20:24] <ejpbruel> askalski: famous last words :)
# [20:24] <@bz> could you?
# [20:24] <askalski> how to make make build faster? what I mean is re-compile
# [20:24] * mcote|bbiab is now known as mcote
# [20:25] <ejpbruel> askalski: have you tried to use ccache?
# [20:25] <askalski> recompile only what have been modified
# [20:25] * Parts: jet (junglecode@moz-C03D0C61.vlan426.asr1.sfo1.gblx.net)
# [20:25] <askalski> ccache... let me look
# [20:25] * mdas|afk is now known as mdas
# [20:25] <ejpbruel> asklaski: that does exactly what you want
# [20:25] * Joins: jet (junglecode@moz-C03D0C61.vlan426.asr1.sfo1.gblx.net)
# [20:25] <askalski> well I don't see that on my .mozconfig
# [20:25] <ejpbruel> askalski: add the following line to your mozconfig after youve installed it: ac_add_options --with-ccache=/usr/local/bin/ccache
# [20:26] <bholley> bz: I certainly can, though out of the 3 of us I know the least about this stuff
# [20:26] * Joins: redfive (chatzilla@moz-B94E1A81.ipnetworksinc.net)
# [20:26] * bholley goes to look
# [20:26] <askalski> ok, some more questions then :D
# [20:26] <dholbert> ejpbruel, I believe --with-ccache no longer takes an argument
# [20:26] <edmorley> askalski: what platform?
# [20:27] <dholbert> (or maybe it's optional? In any case, I just have --with-ccache, not --with-ccache=/foo/bar)
# [20:27] <askalski> I forgot how to set fake home directory for firefox build, so it does not open my stable-build profile
# [20:27] <askalski> linux
# [20:27] <roc> bz: I'm not sure exactly what problem you guys are trying to solve, but there's a standard set of tricks for compiling upcasts and checked downcasts efficiently
# [20:27] <ejpbruel> dholbert: it would make sense for it to be optional
# [20:27] <edmorley> askalski: have you seen https://developer.mozilla.org/en/Mozilla_Build_FAQ#Making_builds_faster ?
# [20:27] <bent> joe: gfx.content.azure.enabled already exists on my nightly, and is set to true... is there some other pref you meant to include in the wiki?
# [20:27] <joe> bent: are you sure it isnot actually gfx.**canvas**.azure.enabled?
# [20:27] <Ms2ger> roc, "making Firefox fast"
# [20:27] <@bz> roc: we're trying to solve the problem of quickly going from a JSObject to a concrent C++ class
# [20:28] <bent> joe, ah poop, it is
# [20:28] <@bz> roc: er, concrete
# [20:28] * bent learns abc's
# [20:28] <@bz> roc: or to null if the native for that JSObject doesn't inherit from that concrete class
# [20:28] <peterv> bz: where's that ms2ger-list?
# [20:28] <@bz> peterv: I'll forward to you
# [20:28] <peterv> ah, hmm
# [20:28] <Ms2ger> peterv, sorry, you're not with the lucky few :)
# [20:29] <@bz> peterv: should have asked him to cc you to start with
# [20:29] <@bz> peterv: mail sent
# [20:29] <peterv> thx :-)
# [20:29] * peterv feels lucky now
# [20:29] <roc> we're going to have one JSClass per concrete class, right?
# [20:29] <askalski> edmorley, : do you possibly know the answer to second question?
# [20:29] <@bz> roc: yes
# [20:29] <dholbert> ejpbruel, ah, cool, looks like it does have an optional arg (http://mxr.mozilla.org/mozilla-central/source/configure.in?mark=7360-7361#7356)
# [20:29] <@bz> roc: one JSClass per most-derived concrete class
# [20:30] <jlebar> gah, I'm orange on inbound.
# [20:30] <Ms2ger> Better than on central!
# [20:30] <edmorley> askalski: use -P profile.name
# [20:30] <@bz> roc: btw, every time I touch events I get really depressed. :(
# [20:30] <bholley> Ms2ger: what are these Shared*Element classes, anyway?
# [20:31] <Ms2ger> Code size savings, IIRC
# [20:31] <askalski> no, there was some other solution, selecting $HOMEDIR or something like this in eclipse, you know that?
# [20:31] <@bz> roc: trying to decide how much time to really devote to fixing the broken-ass tests that can't handle slightly more async synthetic mousemove. :(
# [20:31] <askalski> edmorley, :no, there was some other solution, selecting $HOMEDIR or something like this in eclipse, you know that?
# [20:31] <Ms2ger> Implementing multiple elements in one C++ class
# [20:31] * Quits: johanc (chatzilla@moz-D8A1AA43.bredband.comhem.se) (Ping timeout)
# [20:31] <bholley> Ms2ger: oh, I see. The simple ones
# [20:32] <@bz> yes
# [20:32] <edmorley> askalski: sorry I don't use eclipse and not really sure what you're wanting to achieve
# [20:32] <@bz> or in the case of SharedObjectElement not necessarily simple, but "similar enough"
# [20:32] <bholley> bz: well, since they all share the same superclass hierarchy, it should be fine for us
# [20:32] <bholley> bz: anything that one can be casted to all the others can be
# [20:32] <@bz> bholley: for Shared*?
# [20:32] <roc> bz: so given the JSObject, you know what the most-derived C++ class is (from the JSClass), and you want to see if that inherits from some particular parent concrete C++ class?
# [20:33] <@bz> bholley: yeah, just have to make sure their jsclass flags list the right things
# [20:33] <Ms2ger> Shared can't necessarily be casted to one of its interfaces, though
# [20:33] * Joins: karl (karl@moz-9D2227AB.jetstream.xtra.co.nz)
# [20:33] <Ms2ger> Which is one of the conditions bz asked for :)
# [20:33] * Quits: dRdR (dRdR@moz-3D8840A3.uwaterloo.ca) (Client exited)
# [20:33] * Quits: Unfocused (u2622@moz-160C58C6.com) (Max SendQ exceeded)
# [20:33] <@bz> roc: to a first approximation
# [20:33] * Joins: Unfocused (u2622@moz-160C58C6.com)
# [20:34] <@bz> roc: what form does the "knowing" you mention take?
# [20:34] <@bz> roc: knowing of most-derived C++ class, that is
# [20:34] <mounir> glandium, khuey|away: I'm going to land a lot of C++ soon, is it still relevant to let you guys know?
# [20:35] <Ms2ger> r-
# [20:35] <Ms2ger> :)
# [20:35] <roc> a pointer to some metadata about the most-derived concrete C++ class
# [20:35] <mounir> Ms2ger: remember me to never ask you any review :)
# [20:35] <Ms2ger> mounir, I'm happy to remind you of that, if you forgot :)
# [20:36] <edmorley> jlebar: sorry think your push is causing M1 orange on inbound
# [20:36] <@bz> roc: right
# [20:36] <Ms2ger> <jlebar> gah, I'm orange on inbound.
# [20:36] <Ms2ger> edmorley, ^
# [20:36] <bholley> bz, Ms2ger: So, for nsHTMLSharedElement, we'll just need the ability for different instances of that class to have different bits
# [20:36] <@bz> roc: so right now the proposal is that this metadata just be a list of the things it inherits from
# [20:36] <edmorley> Ms2ger: ah thanks, my speed skim read failed there
# [20:36] <@bz> bholley: on the jsclass. Should be totally doable
# [20:36] * jhammel is now known as jhammel|foodz
# [20:36] <bholley> bz: yeah
# [20:37] * bholley looks at the last one
# [20:38] <peterv> so we'd have multiple JSClasses for nsHTMLSharedElement?
# [20:38] * Joins: lsumar (lsumar@E2F7352.56080C53.97BBD552.IP)
# [20:38] <bholley> peterv: yeah
# [20:38] * Quits: jet (junglecode@moz-C03D0C61.vlan426.asr1.sfo1.gblx.net) (Quit: jet)
# [20:38] * Quits: sworkman (sworkman@moz-BBE3ABD.mv.mozilla.com) (Quit: sworkman)
# [20:39] <roc> for casts to classes on the "primary inheritance chain", you can give each most-derived concrete class an array of its parent classes, where element 0 is the root class, element 1 is the subclass of the root, etc. Then the class you want to cast to will always be at the same index in that array, if it occurs at all.
# [20:39] <@bz> peterv: that's the thinking
# [20:39] <@bz> roc: the actual cast is just a no-op
# [20:39] * Quits: bmoss (bmoss@moz-BBE3ABD.mv.mozilla.com) (Ping timeout)
# [20:39] <@bz> roc: all the things we'll want to cast to are on the primary inheritance chain
# [20:39] * Joins: sworkman (sworkman@moz-BBE3ABD.mv.mozilla.com)
# [20:39] <@bz> roc: so in terms of what needs to happen with the bits of the pointer, the answer is "nothing"
# [20:40] * Joins: jet (junglecode@moz-C03D0C61.vlan426.asr1.sfo1.gblx.net)
# [20:40] <@bz> roc: we just need to check whether we in fact inherit from the whatever, and if we do return the same pointer reinterpret_cast to the other thing
# [20:40] * jwir3 is now known as jwir3|lunch
# [20:40] <@bz> roc: if not, return null
# [20:40] * @bz hopes he and roc are not talking past each other
# [20:40] <@bz> but maybe we are
# [20:41] <@bz> so ok
# [20:41] <@bz> continue?
# [20:41] <roc> so if you just want to check whether the class you want to cast (say T) to is a superclass of the concrete class (C), compute D, how deep T is in the class hierarchy (can be done statically), then get C's ancestor class table A[], and check whether A[D] == T
# [20:41] * gregglind is now known as gregglind_afk
# [20:41] <@bz> ah, cute
# [20:42] * Joins: evilpie (evilpie@moz-37567C56.pools.arcor-ip.net)
# [20:42] <@bz> have to make sure that the ancestor class tables are all the size of the biggest one
# [20:42] <@bz> but that should be doable statically
# [20:42] <roc> you don't even need an array bounds check if you make all ancestor class tables a minimum length and D < that minimum length
# [20:42] <roc> right
# [20:42] * Joins: bjacob (bjacob@F2D29657.F60B0462.67AC9B1.IP)
# [20:42] <bholley> roc: hey, that's a great idea
# [20:42] * @bz is sold
# [20:42] <bholley> bz: that would save us from walking the list
# [20:42] <@bz> yep
# [20:42] <@bz> exactly
# [20:42] * Joins: bmoss (bmoss@moz-BBE3ABD.mv.mozilla.com)
# [20:42] <bholley> roc: <3
# [20:43] <@bz> and ordering that way is pretty straightforward
# [20:43] <bholley> bz: the DOM doesn't have multiple interface inheritance, right?
# [20:43] <Ms2ger> No
# [20:43] <Ms2ger> Well, SVG does
# [20:43] <Ms2ger> But they're insane
# [20:43] <@bz> bholley: for our purposes, no
# [20:44] <bholley> Ms2ger: will we need to do something different for SVG?
# [20:44] <@bz> bholley: since we're interested in inheritance on _proto_ chains, not on interfaces
# [20:44] * Quits: bmoss (bmoss@moz-BBE3ABD.mv.mozilla.com) (Max SendQ exceeded)
# [20:44] * Joins: kaze (kaze@moz-74A6AC89.w82-123.abo.wanadoo.fr)
# [20:44] * Joins: flx_ (flx@moz-F4992FEC.uwaterloo.ca)
# [20:44] <bholley> bz: ah, I see
# [20:44] <askalski> roc
# [20:44] <Ms2ger> No, someone just needs to fix the spec
# [20:44] <askalski> roc: you got a moment?
# [20:44] * Joins: bmoss (bmoss@moz-BBE3ABD.mv.mozilla.com)
# [20:44] <@bz> bholley: unlike quickstubs, which have a method per (interface, name) pair, we will have a method per (proto object, name) pair
# [20:44] <roc> askalski: sure
# [20:44] <@bz> bholley: so any sort of weird gets flattened into a proto chain
# [20:45] <@bz> bholley: and then our stubs are hung off the proto chain
# [20:45] <jlebar> joe, ping?
# [20:45] <bholley> bz: right, ok
# [20:45] * Quits: mbrubeck (mbrubeck@moz-3CE92E7A.tmodns.net) (Ping timeout)
# [20:45] <@bz> bholley: so the stub knows how deep down the chain it is
# [20:45] <joe> jlebar: hi
# [20:45] <@bz> bholley: this is going to rock. ;)
# [20:45] <edmorley> jlebar: are you backing out, or shall I?
# [20:45] <bholley> bz: \o/
# [20:45] <jlebar> joe, So it looks like RasterImage::SourceDataComplete needs to report errors synchronously. Is that right?
# [20:45] <bholley> bz: now, there's just that SVG case
# [20:45] <joe> let me see
# [20:45] <jlebar> edmorley, How mad would you be if I pushed a test fix? That test is just busted, I think, and I'm testing now...
# [20:46] <Ms2ger> \oo/
# [20:46] <bholley> Ms2ger is a two-headed monster!
# [20:46] <Ms2ger> Uh, shouldn't have told you that
# [20:47] * bholley realizes that may have been offensive to two-headed people(s?)
# [20:47] <Ms2ger> bholley, SVG needs to be rewritten with lots of 'implements'
# [20:47] <Ms2ger> Nah, just assuming everybody who has two heads is a monster
# [20:48] <bholley> Ms2ger: we can special-case the non-monsters
# [20:48] <edmorley> jlebar: wfm if you don't mind checking/starring each M1 failure until it goes green to make sure nothing else creeps in from someone else? :-)
# [20:48] <joe> jlebar: yes, I believe so
# [20:48] <joe> jlebar: it's already called inside an asynchronous call - OnStopRequest
# [20:48] <jlebar> edmorley, okay; thanks. I just need another few minutes.
# [20:49] <bjacob> bholley: i have the association of victims of tchernobyl on the phone, are you available?
# [20:49] * Joins: sfleiter|away (stfl@moz-22F37930.hsi7.kabel-badenwuerttemberg.de)
# [20:49] <joe> what the actual hell is going on in here
# [20:49] <jlebar> joe, Okay. As I understand the current code, SourceDataComplete() runs the worker. And the worker does...however much it wants. It might not finish decoding and throw an error.
# [20:49] <joe> right
# [20:50] <jlebar> So SourceDataComplete might *not* synchronously report the error.
# [20:50] <jlebar> But it needs to...
# [20:50] <joe> is true
# [20:50] <jlebar> So I guess I'm going to make SourceDataComplete decode everything.
# [20:50] * Quits: @dbaron (dbaron@moz-C03D0C61.vlan426.asr1.sfo1.gblx.net) (Quit: 8403864 bytes have been tenured, next gc will be global.)
# [20:51] <joe> I don't know that it's necessary
# [20:51] <joe> we currently have this limitation after all
# [20:51] <joe> no need to change it
# [20:51] * Joins: yuan (ywang@moz-7B0110AD.mv.mozilla.com)
# [20:51] * Joins: dbaron (dbaron@moz-C03D0C61.vlan426.asr1.sfo1.gblx.net)
# [20:51] * ChanServ sets mode: +o dbaron
# [20:51] * Quits: sicking (chatzilla@moz-C03D0C61.vlan426.asr1.sfo1.gblx.net) (Ping timeout)
# [20:51] * Quits: rwaldron (rwaldron@moz-BDCCF091.hfc.comcastbusiness.net) (Ping timeout)
# [20:52] <jlebar> joe, Well, fixing it properly might be just as hard as fixing it to the level it was fixed before.
# [20:52] <jlebar> joe, atm, I'm failing tests I think because I never report an error synchronously in SourceDataComplete.
# [20:52] <joe> I don't think we should make SourceDataComplete() decode everything
# [20:53] <jlebar> Okay. What's the correct fix?
# [20:54] * Joins: dRdR (dRdR@moz-792B0937.uwaterloo.ca)
# [20:54] * Joins: mib_le4ky9 (Mibbit@F6A9DE18.A1146449.57460053.IP)
# [20:54] <joe> depends. What's the failure?
# [20:54] <jlebar> smaug, can you rs this test fix for the <iframe mozbrowser> patch? http://pastebin.mozilla.org/1443592
# [20:55] <jlebar> joe, https://tbpl.mozilla.org/?tree=Try&rev=eb29adb5609e Lots of them. At least some explicitly have to do with not noticing erroneous images.
# [20:55] <@smaug> jlebar: looking
# [20:55] <jlebar> smaug, (This is quite possibly the dumbest test...)
# [20:55] <joe> jlebar: we probably need to block onload until we're done decoding
# [20:56] <@smaug> jlebar: rs=me
# [20:56] <jlebar> smaug, thanks.
# [20:56] <jlebar> joe, mmm...
# [20:56] <joe> I think that OnRequestFinished or somesuch is what blocks onload (implicitly)
# [20:56] <joe> OnStopRequest
# [20:56] <joe> gah
# [20:56] <joe> how did I forget that
# [20:56] * Quits: askalski (akuda@moz-6A36EC49.ip.abpl.pl) (Ping timeout)
# [20:57] * Joins: askalski (akuda@moz-6A36EC49.ip.abpl.pl)
# [20:57] <@smaug> ehsan: I still don't understand how you get so much higher CC times
# [20:58] <@smaug> ehsan: I have plenty of tabs open and all ...
# [20:58] <ehsan> smaug: no idea :/
# [20:58] <ehsan> smaug: but the times in the recent build are much better now
# [20:58] <@smaug> I mean, I'm in 16-40ms range
# [20:58] * Quits: mib_le4ky9 (Mibbit@F6A9DE18.A1146449.57460053.IP) (Quit: http://www.mibbit.com ajax IRC Client)
# [20:58] <ehsan> I've seen multi-second pauses regularly
# [20:58] <cpeterson> nightly build question: What time do the nightly builds snapshot mozilla-central? How can I determine if a particular mozilla-central changeset made it into the nightly build? I see hg tags for AURORA/etc, but not nightly.
# [20:58] <ehsan> smaug: have you tried running with ABP?
# [20:59] <@smaug> no
# [20:59] <@smaug> ehsan: have you try to run without it?
# [20:59] <jlebar> joe, So we should finish decoding the image in OnStopRequest?
# [20:59] * Quits: janv (varga@moz-C1261AFF.flarion.as5628.telecom.sk) (Quit: This computer has gone to sleep)
# [20:59] * Quits: Boriss (FlyingToas@moz-C03D0C61.vlan426.asr1.sfo1.gblx.net) (Client exited)
# [21:00] * Joins: Boriss (FlyingToas@moz-C03D0C61.vlan426.asr1.sfo1.gblx.net)
# [21:00] <ehsan> smaug: I can do it. do you want me to?
# [21:00] <@smaug> just looking at the CC times... 18ms, 29ms ...
# [21:00] <@smaug> ehsan: that could be useful
# [21:00] * jgriffin is now known as jgriffin-lunch
# [21:00] <ehsan> smaug: ok, will do that right now
# [21:00] <ehsan> smaug: I think you should also try running with a few of the top add-ons if you want to decrease the CC times that our users experience ;)
# [21:00] <edmorley> cpeterson: ~3am pacific iirc
# [21:01] * Quits: rclick (rclick@EA4B602E.692CFA54.E430BA25.IP) (Ping timeout)
# [21:01] <cpeterson> thanks edmorley
# [21:01] <edmorley> cpeterson: you can look at about:buildconfig, the text file in the nightly ftp directory or else if you have the zip, look in application.ini
# [21:02] <joe> jlebar: no; we shouldn't fire onstoprequest until the image is done decoding
# [21:02] <@smaug> ehsan: ahaa, the size of the latest cclog is like 7% of the previous one
# [21:02] <edmorley> cpeterson: or press the down arrow on tbpl until you see the last nightly ("N" builds) on the repo you are interested in
# [21:02] <joe> jlebar: so the status tracker will have a bit for "necko is done" but should have an addition bit for "done decoding" too
# [21:02] <ehsan> smaug: really?!
# [21:02] <jlebar> joe, oic
# [21:02] <cpeterson> thanks. I didn't know about about:buildconfig
# [21:03] <ehsan> hmm
# [21:03] <@smaug> ehsan: 105068535 vs 7402363
# [21:03] <@smaug> I assume both were created the same way
# [21:03] <ehsan> smaug: I think I disabled DOM Inspector in this run... could that be relevant?
# [21:03] <@smaug> ehsan: you didn't even bz2 this log
# [21:04] <@smaug> ehsan: anything can be relevant
# [21:04] <ehsan> smaug: oh, heh yeah but that's cause I forgot ;)
# [21:04] <ehsan> and attaching it worked
# [21:04] <edmorley> cpeterson: nightly tester tools (used to and preusme still has) a "show pushlog of all changesets since the last update was run" (so works for leaving it a few days between updating nightlies)
# [21:05] <ehsan> smaug: ok, I disabled ABP as well
# [21:05] <ehsan> smaug: anything interesting in my log?
# [21:06] <@smaug> ehsan: still looking
# [21:06] <ehsan> ok
# [21:07] <askalski> hey, is it normal, that "make -f client.mk clean" consumes like 3-4gb'o'ram before my system starts swapping and I kill it?
# [21:07] * Quits: Mano (chatzilla@moz-D0BA9E6D.red.bezeqint.net) (Ping timeout)
# [21:07] * Quits: beaufour (beaufour@moz-723527FF.nycmny.fios.verizon.net) (Quit: beaufour)
# [21:07] <askalski> and is there a way to "make clean" within 6gb ram system?
# [21:07] * Joins: nrc (chatzilla@538BABFE.A073F3E.97BBD552.IP)
# [21:08] <jlebar> askalski, $rm -rf objdir
# [21:08] * Joins: AutomatedTester (AutomatedT@moz-D4F73363.as13285.net)
# [21:08] <@smaug> ehsan: well, at least I don't see any mysterious gmail documents being in the graph
# [21:08] <@smaug> ehsan: in the previous logs there was always some zombie documents
# [21:08] * Quits: jacek (jacek@moz-5D707D3B.psi.wroc.pl) (Quit: Konversation terminated!)
# [21:08] <@smaug> I don't see any such documents here
# [21:08] <askalski> jlebar, : ok, it stores all intermediate results there, right?
# [21:08] <jlebar> askalski, If you set up your mozconfig correctly, yes. :)
# [21:08] <bholley> bz: is there a way to run dromaeo without leaving my computer idle for 25 minutes?
# [21:08] <@smaug> oh, one chrome://browser/content/browser.xul
# [21:08] * Quits: dRdR (dRdR@moz-792B0937.uwaterloo.ca) (Client exited)
# [21:09] <jlebar> askalski, |$ hg status| would tell you if there's anything new in your src dir.
# [21:09] * Joins: dRdR (dRdR@moz-792B0937.uwaterloo.ca)
# [21:09] <ehsan> smaug: hrm, could dom inspector hold on to documents through js objects?
# [21:09] <askalski> jlebar, : mk_add_options MOZ_OBJDIR=@TOPSRCDIR@/../obj-ff-dbg ?
# [21:09] <jlebar> askalski, looks good!
# [21:09] <@smaug> ehsan: that is possible
# [21:10] <@smaug> I don't remember how DOMi works
# [21:10] <ehsan> smaug: note that I barely ever _used_ dom inspector
# [21:10] <@smaug> ehsan: this log was created using my latest tryserver build, right?
# [21:10] * Quits: kdcw (kdc@moz-F7413045.pk.shawcable.net) (Quit: HydraIRC -> http://www.hydrairc.com <- *I* use it, so it must be good!)
# [21:10] <Ms2ger> bholley, sure, get a second computer
# [21:10] <ehsan> smaug: yes
# [21:11] <askalski> jlebar, : OK, is there a difference between normal make and pymake?
# [21:11] <@smaug> since that build does fix 3 different classes of problems
# [21:11] <Ms2ger> askalski, pymake doesn't deadlock as much on windows
# [21:11] <jlebar> askalski, normal make will choke if you do -jN, but pymake will maybe work.
# [21:11] * Joins: beaufour (beaufour@18D5CC88.C7EE4FB2.ECED8BE3.IP)
# [21:11] <askalski> ok, but it's windows specific, right?
# [21:11] <jlebar> askalski, yes
# [21:12] <ehsan> smaug: from about:buildconfig: http://hg.mozilla.org/try/rev/c90043e27dc1
# [21:12] <askalski> ok, any more clues about speeding up compile process?
# [21:12] <askalski> I already have ac_add_options --with-ccache
# [21:12] <askalski> though it ignores the argument to my findings
# [21:12] <edmorley> askalski: did you look at the FAQ I linked?
# [21:13] * Quits: flx_ (flx@moz-F4992FEC.uwaterloo.ca) (Ping timeout)
# [21:13] <@smaug> ehsan: yep, that is the latest one
# [21:13] <askalski> edmorley, can you resend it
# [21:13] <jlebar> it shouldn't be ignoring ccache....
# [21:13] <edmorley> (2012-01-10 19:20:47) edmorley: askalski: have you seen https://developer.mozilla.org/en/Mozilla_Build_FAQ#Making_builds_faster ?
# [21:13] <ehsan> smaug: good. so it's also possible that your latest fixes have OMG SUPER AWESOME :)
# [21:14] <@smaug> they are super awesome :) I'm trying to find more things to fix.
# [21:14] <@smaug> and get all this stuff landed asap
# [21:14] * Joins: mjschranz (mjschranz@moz-57B23647.senecac.on.ca)
# [21:14] <ehsan> I'm so excited for fast CCs :)
# [21:14] <askalski> edmorley, : thanks, I must have it lost last time
# [21:14] <edmorley> np :-)
# [21:15] * jhammel|foodz is now known as jhammel
# [21:15] * coop|mtg is now known as coop
# [21:15] <askalski> ok
# [21:15] * Joins: flx_ (flx@moz-F4992FEC.uwaterloo.ca)
# [21:15] <askalski> do you set up any linux debuggers?
# [21:15] <askalski> because everyone I ask uses Visual Studio for some reason
# [21:15] <@smaug> ehsan: I *think* I still see plenty of uncollectable stuff caused by ABP in the graph. I need to think what to do with those objects.
# [21:16] <jlebar> askalski, you should be able to do |gdb dist/bin/firefox|
# [21:16] <@smaug> mccr8 did suggest something like not adding XPCVariant as roots...
# [21:16] <@smaug> that would cut tens of thousands objects from the graph in this log
# [21:16] <mccr8> smaug: really ? I don't remember that. ;)
# [21:16] <askalski> jlebar, : yes, but I was thinking about something more graphical. I spend several hours setting up eclipse, but I failed to get it working with ccache
# [21:17] <@smaug> mccr8: you mentioned something about that
# [21:17] * Joins: sicking (chatzilla@moz-C03D0C61.vlan426.asr1.sfo1.gblx.net)
# [21:17] <@smaug> mccr8: or, hmm, you were talking about WrappedJS
# [21:17] <jlebar> askalski, I've never met a graphical debugger on Linux I liked. Let me know if you have some success.
# [21:17] <@smaug> oops, my bad
# [21:17] * Joins: nattokirai (nattokirai@moz-6A258170.dynamic.ppp.asahi-net.or.jp)
# [21:17] <jhammel> s/on Linux//
# [21:17] <mccr8> smaug: heh np
# [21:17] <askalski> jlebar, : in my project (habanero3d.sf.net) I use netbeans
# [21:17] * mccr8 is now known as mccr8|lunch
# [21:17] <askalski> and it works for most of the time, very similar to visual studio
# [21:18] <roc> does anybody have a problem with http://www.cuttherope.ie/?html5audio=true ? Microsoft said they disabled HTML5 audio for Firefox due to bugs for some users, but didn't mention what those bugs are, and it works fine for me
# [21:18] <askalski> enough to not to kill myself tracing uninitialized variable, so thats's something :)
# [21:18] <jlebar> roc, how do I re-enable HTML5 audio there?
# [21:18] * Joins: Yoric (Yoric@moz-920DB13B.fbx.proxad.net)
# [21:18] <gcp> jlebar: use his URL
# [21:18] <armenzg> anyone know about browserID?
# [21:18] <jlebar> oh.
# [21:18] <jlebar> roc, sounds fine to me on Linux.
# [21:19] <armenzg> it seems I have a profile that prevents me from using it
# [21:19] <@smaug> audio seems to work fine here
# [21:19] <@smaug> very annoying audio
# [21:19] <jlebar> roc, and on Mac.
# [21:19] * Joins: rclick (rclick@EA4B602E.692CFA54.E430BA25.IP)
# [21:19] * lsblakk|buildduty is now known as lsblakk|buildduty|lunch
# [21:19] * sfleiter|away is now known as sfleiter
# [21:20] * Joins: Julian (chatzilla@moz-43E73578.greenbytes.de)
# [21:20] * Joins: bretr (bret_recka@moz-C03D0C61.vlan426.asr1.sfo1.gblx.net)
# [21:21] * Joins: Mnyromyr (Mnyromyr@B2521176.7B0892CB.771966F7.IP)
# [21:21] * Joins: ashughes (ashughes@21B7B9F2.B87E9213.6E712CE2.IP)
# [21:22] <@smaug> ehsan: ah, I think I know how to optimize your case... new builds coming soon
# [21:23] <ehsan> smaug: you mean with ABP?
# [21:23] <@smaug> yeah
# [21:23] * Quits: c0smikdebris (c0smikdebr@C4166EFD.7E460B98.BE90E62C.IP) (Quit: Computer has gone to sleep.)
# [21:23] <ehsan> cool, let me know when they're ready to test :)
# [21:24] <@smaug> ehsan: it may increase purple buffer clean up times, so need to be careful... but we'll see
# [21:24] <edmorley> cuttherope.ie not working in latest win32 nightly for me, anyone else?
# [21:24] <mwu> not working in osx nightly either
# [21:24] <mwu> doesn't think nightly is a recent browser
# [21:25] <@bz> huh
# [21:25] * @bz had no problems with nightly when he tried
# [21:25] * mdas is now known as mdas|afk
# [21:26] <lurking_work> doesn't work in 9.0.1 at my XP SP3 at work - then again this box won't do HWA either, so not so sure about html5
# [21:26] <jlebar> mwu, edmorley wfm in nightly on mac...
# [21:27] <mwu> maybe my nightly isn't up to date enough
# [21:27] * Quits: davidillsley (chatzilla@moz-CF96CCA3.range109-157.btcentralplus.com) (Client exited)
# [21:27] * Joins: squib (squib@moz-F5CA0CFB.dhcp.mdsn.wi.charter.com)
# [21:27] * Joins: davidillsley (chatzilla@moz-CF96CCA3.range109-157.btcentralplus.com)
# [21:27] <espindola> armenzg, any news on the os x debug builds?
# [21:28] <armenzg> espindola: I'm probably tackling it tomorrow morning
# [21:28] <bholley> bz: yt?
# [21:28] <espindola> armenzg, awesome. With that we should be able to stop using the 10.5 sdk tomorrow too.
# [21:28] <espindola> on m-c
# [21:28] <@bz> yes
# [21:28] <jlebar> bholley, so...want to fix bug 505385? It looks like this, or the creation of further hacks, blocks https://bugzilla.mozilla.org/show_bug.cgi?id=715308
# [21:28] <@bz> bholley: ^
# [21:28] <armenzg> espindola: for both 10.5 and 10.6 machines?
# [21:29] <bholley> bz: let's move to #content where it's less crowded
# [21:29] <armenzg> this is independent from my work, right?
# [21:29] <@bz> bholley: ok
# [21:29] <espindola> armenzg, with the debug build moved to 10.6
# [21:29] <espindola> we will not be building anything on 10.5 anymore
# [21:29] <espindola> (for m-c)
# [21:29] <armenzg> sounds good
# [21:29] <armenzg> :)
# [21:29] <espindola> armenzg, to be exact
# [21:29] <espindola> https://tbpl.mozilla.org/?tree=Try&rev=c7fec001ba99
# [21:29] <ehsan> smaug: out of curiosity, do we have docs somewhere which explains how the cycle collector works?
# [21:29] <espindola> the only thing in that that should fail
# [21:30] <bholley> jlebar: I would like it to be fixed, but I don't have the bandwidth at this particular moment
# [21:30] <espindola> is the build that will try to use a 10.6 sdk from a machine without one
# [21:30] <ehsan> smaug: you keep talking about purples and blacks and stuff, and I keep pretending that I know what you mean by them ;)
# [21:30] <espindola> so the same try tomorrow should be all green
# [21:30] <@smaug> ehsan: nsCycleCollector.cpp has documentation
# [21:31] * nthomas|away is now known as nthomas
# [21:31] * Quits: KWierso (chatzilla@moz-6E3FB3B9.desm.qwest.net) (services.mozilla.org (Too many invalid passwords))
# [21:31] <@smaug> ehsan: and name of the article
# [21:31] <ehsan> smaug: thanks, I'll take a look
# [21:31] * Joins: KWierso|outoftown (chatzilla@moz-6E3FB3B9.desm.qwest.net)
# [21:31] * KWierso|outoftown is now known as KWierso
# [21:32] * Joins: jfkthame (jfkthame@43709DB9.5BC345F5.9542EC20.IP)
# [21:32] <@smaug> ehsan: http://www.research.ibm.com/people/d/dfb/papers/Bacon01Concurrent.pdf
# [21:32] <@smaug> I think it is that
# [21:32] <Ms2ger> Bacon? Hmm
# [21:32] <bholley> jlebar: I would gladly help design the interface though
# [21:32] <jhammel> better even than that...Concurrent bacon!
# [21:32] <bholley> jlebar: and could possibly do it a bit later
# [21:33] <@smaug> Ms2ger: no no, be vegetarian
# [21:33] * jlebar wonders if I can get my patch in without fixing that bug.
# [21:33] <Ms2ger> Electricity is enough for me, but I need to like bacon to fit in :)
# [21:34] <@smaug> oh, XPCVariant has spare bits. I'm always happy when I find left-over bits I can use
# [21:34] * jgriffin-lunch is now known as jgriffin
# [21:34] <jhammel> left-over bacon bits?
# [21:35] <@smaug> I wouldn't use bacon for anything
# [21:35] * Quits: clee (clee@moz-D0C71163.domainhotel-wifi.mozilla.hq) (Quit: clee)
# [21:36] * Quits: Goldorak (chatzilla@805CD605.E12797EB.187A1082.IP) (Ping timeout)
# [21:37] <Wes_> 8O
# [21:37] <Ms2ger> 90
# [21:37] <jhammel> sold!
# [21:38] <Ms2ger> What'd I get?
# [21:38] * Joins: Goldorak (chatzilla@8C15BD3E.A9EA3755.187A1082.IP)
# [21:39] * Quits: kumar (kmcmillan@moz-D0C71163.domainhotel-wifi.mozilla.hq) (Quit: kumar)
# [21:39] <Wes_> Ms2ger: A storage locker full of some hoarder's junk
# [21:39] * jwir3|lunch is now known as jwir3
# [21:39] <Ms2ger> Dammit
# [21:40] * Ms2ger passes it to jmaher
# [21:40] * Joins: kumar (kmcmillan@moz-D0C71163.domainhotel-wifi.mozilla.hq)
# [21:40] <mwu> jlebar: cut the rope still doesn't work after updating to the latest nightly on osx
# [21:40] * ashughes is now known as ashughes|meeting
# [21:40] <jlebar> mwu, oO
# [21:40] <jduell> smaug: there's a couple bugs where people are creating new websockets (or changing document.location) during an onstop that's caused by page navigation
# [21:40] <jduell> smaug: is there something in XHR that prevents this, and if so, can we copy to websockets?
# [21:41] <jlebar> mwu, I'm on 10.7. Are you on 10.6?
# [21:41] <mwu> yeah
# [21:41] <jlebar> :shrug
# [21:41] <jduell> smaug: see bug 602286 and bug 696085
# [21:41] <mwu> yeah I dunno.
# [21:41] <@smaug> bz: ^
# [21:41] * Joins: pnemsak (Miranda@moz-7409BD53.orange.sk)
# [21:41] <jmaher> Ms2ger: you are my friend
# [21:41] <Ms2ger> Indeed, Sir.
# [21:42] * mdas|afk is now known as mdas
# [21:42] * Quits: kumar (kmcmillan@moz-D0C71163.domainhotel-wifi.mozilla.hq) (Quit: kumar)
# [21:42] * Quits: victorporof (victorporo@BCAA178E.F4308271.4873E54D.IP) (Quit: Leaving...)
# [21:42] <@smaug> jduell: you mean during unload ?
# [21:43] <@smaug> jduell: bz or sicking might remember if we have anything special. I would assume XHR to just work. (well, work and then the connection is perhaps cut)
# [21:43] <@smaug> jduell: but I don't remember the details
# [21:44] <jduell> smaug: I assume so--the onclose is being called because the page is going away
# [21:44] <@smaug> onclose?
# [21:44] <@smaug> you mean unload event listener
# [21:44] <@smaug> ah, websocket onclose
# [21:44] * Quits: jrmuizel (jrmuizel@F2D29657.F60B0462.67AC9B1.IP) (Input/output error)
# [21:44] <roc> mwu: what doesn't work?
# [21:45] <Callek> smaug: so any special news re my CC aha moment?
# [21:45] <jduell> smaug: right
# [21:45] * Joins: priya (priya@moz-BBE3ABD.mv.mozilla.com)
# [21:45] <Callek> or any (reduced) testcases to try, etc.?
# [21:45] <@smaug> Callek: sorry, nothing yet
# [21:45] <@bz> I'm not sure what the question is
# [21:45] * LegNeato is now known as LegNeato_away
# [21:45] <@smaug> jduell: what is the stack when onclose is called?
# [21:45] <@bz> fwiw, cuttherope worked fine for me on mac
# [21:45] <@bz> In a Jan 4 nightly, sorta
# [21:46] <mwu> roc: https://people.mozilla.com/~mwu/Screen%20shot%202012-01-10%20at%203.38.29%20PM.png
# [21:46] <jduell> bz: see bug 602286 and bug 69608. Users are opening new websockets (or changing document.location) during an onstop that's caused by page navigation
# [21:46] * Quits: mayhemer (Miranda@B3D46202.F87A741B.F23860FD.IP) (Quit: Miranda IM! Smaller, Faster, Easier. http://miranda-im.org)
# [21:46] <@bz> jduell: ok
# [21:46] <jduell> and apparently this is bad, and somehow blocked in XHR
# [21:46] <@bz> jduell: people do that with image loads and xhr too
# [21:46] <jduell> I'm looking to do the same for WS
# [21:46] <roc> mwu: oh, you mean their sniffing code is broken
# [21:46] <@bz> jduell: we don't block it in xhr that I know of
# [21:46] <mwu> yeah
# [21:47] <jduell> bz: mmmK. How do I stop it from happening in WS then?
# [21:47] <@bz> mwu: what build are you using there?
# [21:47] * lsblakk|buildduty|lunch is now known as lsblakk|buildduty
# [21:47] <mwu> 12.0a1 (2012-01-10)
# [21:47] <mwu> Mozilla/5.0 (Macintosh; Intel Mac OS X 10.6; rv:12.0a1) Gecko/20120110 Firefox/12.0a1
# [21:47] <@bz> jduell: need to figure out what behavior we want first
# [21:47] <jduell> bz: or do you mean, there's no bug here
# [21:47] <@bz> jduell: there are two relevant events: when the next page starts to load and when the previous page starts to unload
# [21:48] <@bz> mwu: one sec
# [21:48] * @bz is grabbing updated nightlies
# [21:48] <mwu> ok
# [21:48] <@bz> jduell: we probably cancel websockets on that first one
# [21:48] <jduell> bz: I assume at a minimum we want any WS created during onclose to go away when the page unloads. Right now we're creating a ghost WS that lives for a long time but isn't connected to a page
# [21:48] <@bz> jduell: I think the difference is that usually XHR is short-lived
# [21:49] <@bz> jduell: so even if people do it, it just finishes and then it's done
# [21:49] <jduell> bz: somehow the cancel isn't happening, at least in the bug 696085 report
# [21:49] <rclick> jlebar: ping
# [21:49] <jlebar> rclick, hey
# [21:49] <@bz> jduell: we should probably close all WS connections associated with the old page before embedding the nw one
# [21:49] <@bz> jduell: why not?
# [21:49] <@bz> jduell: we cancel, but their onclose just opens a new one, no?
# [21:50] <ehsan> has anybody got their builds failing on windows with missing refs to sqlite3_mutex_held/sqlite3_mutex_notheld?
# [21:50] <jduell> bz: right. Maybe the loop that cancels open WS's isn't seeing the new one as it iterates?
# [21:50] <rclick> jlebar: to my untrained eye, it looks like some of the failures in 715308 are due to size decodes happening async instead of sync...
# [21:50] <@bz> mwu: worksforme, Jan 10 nightly, clean profile on mac
# [21:50] <@bz> jduell: it's not a loop
# [21:50] <mwu> bz: osx 10.7?
# [21:50] <@bz> jduell: or more precisely...
# [21:50] <jlebar> rclick, could be. I just (like, after you pinged me) fixed another issue, so I can test.
# [21:50] <@bz> mwu: 10.6
# [21:51] <mwu> huh.
# [21:51] <jlebar> rclick, which ones?
# [21:51] <@bz> jduell: we just call cancel on the loadgroup
# [21:51] <rclick> jlebar: I'm thinking that if, in the size decode case, if all the data was written to the decoder in AddSourceData(), and the decoder was shutdown in SourceDataComplete, some of the failures might go away.
# [21:51] <@bz> jduell: which I assume cancels websockets
# [21:51] <@bz> jduell: but yes, wouldn't cancel new stuff that starts after that point
# [21:51] <jduell> bz: I guess I'm wondering where the right frob is to fix this. Is there somethign WS.open() can check in the page that indicates "this page is going away: fail", or do I find the code in docShell/LoadGroup and make sure it cancels the new one
# [21:51] <jlebar> rclick, Well, the hack I just added is to always process all the remaining source data in SourceDataComplete.
# [21:52] <jduell> bz: sounds like I need to make sure the loadgroup add/cancel is happening correctly.
# [21:52] <jlebar> rclick, That's not really the right thing, and I dunno what joe will have to say about it. But the alternative looks complicated, because the notifications I'd need to send are totally messed up.
# [21:52] * Joins: jrmuizel (jrmuizel@F2D29657.F60B0462.67AC9B1.IP)
# [21:53] * Quits: bholley (bholley@moz-FCAF9AAB.hsd1.ca.comcast.net) (Quit: bholley)
# [21:53] <mwu> bz: fixed it by switching to accept cookies instead of asking
# [21:53] <mwu> since we break local storage or something silently if we ask for cookies
# [21:53] * Quits: lurking_work (chatzilla@moz-107FCDBA.hfc.comcastbusiness.net) (Quit: ChatZilla 0.9.87 [Firefox 9.0.1/20111220165912])
# [21:53] * Quits: Enn (enn@F2D29657.F60B0462.67AC9B1.IP) (Ping timeout)
# [21:54] <rclick> jlebar: one moment, I don't have the test logs in front of me anymore.
# [21:55] <mwu> edmorley: ^
# [21:56] <@bz> mwu: ask for cookies is not supported, last I checked...
# [21:56] <@bz> mwu: if you mean the setting where it puts up a dialog before every cookie
# [21:56] <@bz> mwu: (or more precisely, not supported by core Gecko; Firefox exposes a ui pref for it)
# [21:56] <mwu> bz: yeah, it certainly isn't, but we have it in our prefs dialog anyway
# [21:56] <@bz> mwu: yeah, I know
# [21:57] <@bz> mwu: it's insane
# [21:57] <mwu> also racy
# [21:57] <@bz> mwu: since afaik it's kinda crashy
# [21:57] <edmorley> mwu: I don't use ask, but I have unticked third party (though a refresh with that reticked didn't seem to help earlier)
# [21:57] <mwu> every once in while, a website will lock up firefox
# [21:57] * Quits: @smaug (chatzilla@moz-AC5A01B5.elisa-mobile.fi) (Ping timeout)
# [21:57] * Quits: Ms2ger (Ms2ger@C99CCEB1.C1534583.37724B0D.IP) (Quit: nn)
# [21:57] <@bz> jduell: yeah, I'd start there
# [21:58] <jduell> bz: thanks
# [21:59] <jduell> bz: re: the other issue (changing document.location in onclose): jdm summarized it here, if you have any ideas, maybe comment: https://bugzilla.mozilla.org/show_bug.cgi?id=602286#c11
# [21:59] <mwu> edmorley: maybe there's something in your cookies exceptions list?
# [22:00] * Joins: cpearce (chatzilla@538BABFE.A073F3E.97BBD552.IP)
# [22:00] <@bz> jduell: nothing to really add
# [22:00] <jduell> bz: ok
# [22:00] <@bz> jduell: pretty sure you could use a long-poll XHR the same way
# [22:01] <jduell> bz: would you mark it WONTFIX? Or just low priority?
# [22:01] <@bz> the latter for now
# [22:02] <@bz> but also check the spec
# [22:02] <@bz> it may well require this behavior
# [22:02] * khuey|away is now known as khuey
# [22:02] <firebot> Check-in: http://hg.mozilla.org/mozilla-central/rev/fee66b45f486 - Axel Hecht - bug 716842, make sure that l10n repacks have mobile/locales/Makefile, r=aki
# [22:03] <rclick> jlebar: I was looking at https://tbpl.mozilla.org/php/getParsedLog.php?id=8433958&tree=Try . I'm guessing the failures in test_fileapi_slice.html are due to the size decode not happening soon enough.
# [22:03] <edmorley> mwu: my exceptions list is empty; however running in a clean profile (still doesn't work) but shows the flash needed bar (presume I've turned the notification off on my current profile), so guessing it's just choking on not having flash (for an html5 demo, nice)
# [22:03] <jlebar> edmorley, is there a trick to getting the tbpl summaries to load before I turn 50?
# [22:03] <mwu> oh?
# [22:03] <mwu> I loaded the game through the url that forces html5 audio
# [22:03] * mattwoodrow|away is now known as mattwoodrow
# [22:03] <jlebar> rclick, I'll try that one with the patch.
# [22:03] <edmorley> mwu: ditto
# [22:03] <mwu> weird
# [22:04] <edmorley> jlebar: not really, I just run lots of tabs in parallel
# [22:04] <edmorley> jlebar: depends on log size; sometimes you'll get an orange that pollutes the log so much you won't get it to load through tbpl, so have to use the brief view link
# [22:04] <jlebar> edmorley, I seem to recall that in the past, if I clicked on a bunch of orange numbers, it would do a bunch of requests in parallel, and then I could go back and see all of the results. But not anymore...
# [22:04] <edmorley> jlebar: ctrl+click
# [22:05] * Quits: Yoric (Yoric@moz-920DB13B.fbx.proxad.net) (Input/output error)
# [22:05] <jlebar> edmorley, I mean, a bunch of log-parsing requests.
# [22:05] * Joins: lurking_work (chatzilla@moz-107FCDBA.hfc.comcastbusiness.net)
# [22:05] * Quits: roc (chatzilla@C0ACF8B.5E1E9EEA.613E47D1.IP) (Ping timeout)
# [22:05] <edmorley> oh, in which case not sure ('in the past' probably means before my time)
# [22:05] <edmorley> philor? ^
# [22:06] * Quits: flx_ (flx@moz-F4992FEC.uwaterloo.ca) (Ping timeout)
# [22:06] <jlebar> rclick, the test goes green with the change I made. And the change was exactly what you suggested...I think. :)
# [22:07] * Joins: anky (anky@27B0D94A.B421193.A3D1B221.IP)
# [22:07] <jlebar> edmorley, I'll file a bug. This makes it basically impossible to star things.
# [22:07] * Quits: Goldorak (chatzilla@8C15BD3E.A9EA3755.187A1082.IP) (Connection reset by peer)
# [22:08] * Quits: brendan (brendaneic@moz-A286C218.hsd1.ca.comcast.net) (Quit: brendan)
# [22:09] <edmorley> thanks :-)
# [22:09] * Quits: Jef91 (Kristi@moz-FDB75EDD.hsd1.il.comcast.net) (Client exited)
# [22:09] <philor> ah, but who are you going to file it against?
# [22:09] <jlebar> philor, well, I was considering assigning it to you. I assume that in your subsequent rage, you'd figure out the right assignee.
# [22:10] * Joins: flx_ (flx@moz-F4992FEC.uwaterloo.ca)
# [22:10] * Quits: jorendorff (jorendorff@moz-91590D94.hsd1.tn.comcast.net) (Quit: Eaten by grue.)
# [22:10] <edmorley> heh
# [22:10] <joe> is there a reason to not unconditionally use <cmath> in nsMathUtils.h?
# [22:10] <philor> I meant product/component, really
# [22:10] <jlebar> philor, It's not a TPBL bug?
# [22:10] <joe> if you happen to include it before including nsMathUtils.h, isfinite is put into the std namespace
# [22:10] <philor> jlebar: dunno, what is it?
# [22:10] * Joins: eflores (AndChat@538BABFE.A073F3E.97BBD552.IP)
# [22:10] <joe> instead of the global namespace
# [22:10] <jlebar> philor, I'll cc you on the bug in a sec.
# [22:10] * KaiRo is now known as KaiRo_away
# [22:11] <philor> you click on an orange letter, and it doesn't load the summary quickly enough? that's tbpl -> bzapi -> bugzilla
# [22:11] <philor> with IT inserted between and around all of those
# [22:11] * Joins: zuzelvp (zuzelvp@2112147D.C3507A2D.9A8C35B4.IP)
# [22:12] <philor> some of them never load? filed, nobody who has looked at it has any idea how to even start debugging
# [22:12] * Quits: bmoss (bmoss@moz-BBE3ABD.mv.mozilla.com) (Ping timeout)
# [22:12] <philor> you have so many requests going to tbpl at once that you hit the per-server limit? filed
# [22:12] <jlebar> philor, https://bugzilla.mozilla.org/show_bug.cgi?id=717005
# [22:13] <rclick> jlebar: More or less. I'm just curious if we can get away with only doing that hack for size decodes (which are cheap) instead of all decodes.
# [22:13] <edmorley> philor: couldn't tbpl cache the bzapi response to avoid all this?
# [22:14] <philor> edmorley: when are you going to invalidate the cache?
# [22:14] <edmorley> every 5 mins or something
# [22:14] <jlebar> rclick, I'd approached the problem by looking at images with errors. In that case, we need to decode the whole thing in SourceDataComplete() in order to determine whether there's an error.
# [22:14] <jlebar> rclick, So it's more than just size decodes.
# [22:15] * jlebar would go for a button which says "retrieve all the orange information onscreen".
# [22:16] * Joins: myk (myk@moz-D0C71163.domainhotel-wifi.mozilla.hq)
# [22:16] * khuey is now known as khuey|away
# [22:17] * Quits: northAway (northWind@moz-20E58859.cable.teksavvy.com) (Ping timeout)
# [22:19] * Joins: dao (dao@moz-E37F9636.superkabel.de)
# [22:19] * rclick is now known as rclick|afk
# [22:20] * Joins: tmyoung (tmyoung@moz-A52EE408.alyoung.com)
# [22:21] * Quits: rillian (giles@21B7B9F2.B87E9213.6E712CE2.IP) (Quit: Ex-Chat)
# [22:21] * Joins: abwillis (abwillis@9877934.9DD4DBBF.6A7A197.IP)
# [22:23] * jwir3 is now known as jwir3|away
# [22:23] * Joins: sworkman_ (sworkman@moz-BBE3ABD.mv.mozilla.com)
# [22:24] * Quits: cpeterson (cpeterson@moz-C03D0C61.vlan426.asr1.sfo1.gblx.net) (Input/output error)
# [22:25] * Joins: alex (alex@moz-BD8D0A09.lightspeed.sntcca.sbcglobal.net)
# [22:25] * Quits: sworkman (sworkman@moz-BBE3ABD.mv.mozilla.com) (Ping timeout)
# [22:25] * sworkman_ is now known as sworkman
# [22:25] * Joins: johanc (chatzilla@moz-D8A1AA43.bredband.comhem.se)
# [22:26] * Joins: damons (gnubeard@moz-C03D0C61.vlan426.asr1.sfo1.gblx.net)
# [22:26] * mccr8|lunch is now known as mccr8
# [22:27] * Joins: mbrubeck (mbrubeck@843BAE1A.5382978A.91651C13.IP)
# [22:27] * Joins: chrisccoulson (chr1s@moz-DF69ED6E.business.broadband.hu)
# [22:28] * juanb is now known as juanb|lunch
# [22:28] * Joins: bmoss (bmoss@moz-BBE3ABD.mv.mozilla.com)
# [22:30] * Quits: chrisccoulson (chr1s@moz-DF69ED6E.business.broadband.hu) (Quit: Ex-Chat)
# [22:30] * Joins: chrisccoulson (chr1s@moz-DF69ED6E.business.broadband.hu)
# [22:31] * coop is now known as coop|mtg
# [22:32] * Joins: DGMurdockIII (dgmurdocki@moz-E933B63.hsd1.il.comcast.net)
# [22:32] * Joins: chrisccoulson_ (chr1s@moz-B376CDB0.business.broadband.hu)
# [22:32] * Joins: florian (florian@moz-87C33FDA.kimsufi.com)
# [22:33] * Quits: chrisccoulson (chr1s@moz-DF69ED6E.business.broadband.hu) (Ping timeout)
# [22:35] * Joins: cpeterson (cpeterson@moz-C03D0C61.vlan426.asr1.sfo1.gblx.net)
# [22:36] * Joins: Yoric (Yoric@moz-920DB13B.fbx.proxad.net)
# [22:37] * Quits: anant (anant@moz-104CC309.mv.mozilla.com) (Ping timeout)
# [22:37] * Joins: bholley (bholley@moz-FCAF9AAB.hsd1.ca.comcast.net)
# [22:38] * Quits: bjacob (bjacob@F2D29657.F60B0462.67AC9B1.IP) (Input/output error)
# [22:39] * Quits: sworkman (sworkman@moz-BBE3ABD.mv.mozilla.com) (Quit: sworkman)
# [22:40] * Quits: davidb (davidb@F2D29657.F60B0462.67AC9B1.IP) (Quit: davidb)
# [22:40] * Quits: Mitch (chatzilla@moz-FA99DC5F.sbr802.nsw.optusnet.com.au) (Quit: ChatZilla 0.9.88 [SeaMonkey 2.6.1/20111221001913])
# [22:40] * Quits: bmoss (bmoss@moz-BBE3ABD.mv.mozilla.com) (Ping timeout)
# [22:41] * Joins: faramarz_ (faramarz@moz-BBE3ABD.mv.mozilla.com)
# [22:41] * Quits: zuzelvp (zuzelvp@2112147D.C3507A2D.9A8C35B4.IP) (Ping timeout)
# [22:41] * Quits: myk (myk@moz-D0C71163.domainhotel-wifi.mozilla.hq) (Max SendQ exceeded)
# [22:41] * Joins: myk (myk@moz-D0C71163.domainhotel-wifi.mozilla.hq)
# [22:42] * Joins: sworkman (sworkman@moz-BBE3ABD.mv.mozilla.com)
# [22:42] * Quits: faramarz (faramarz@moz-BBE3ABD.mv.mozilla.com) (Ping timeout)
# [22:42] * faramarz_ is now known as faramarz
# [22:42] * Quits: Julian (chatzilla@moz-43E73578.greenbytes.de) (Client exited)
# [22:43] * Joins: bmoss (bmoss@moz-BBE3ABD.mv.mozilla.com)
# [22:43] <jbuck> question about https://bugzilla.mozilla.org/show_bug.cgi?id=605472 and closing tickets. In that ticket he wants TIME_UPDATE lowered to 100ms so that animation works better. In Popcorn.js we added a flag to do exactly that, using requestAnimationFrame. Is it appropriate for me to close that ticket with a link to the documentation?
# [22:43] * gregglind_afk is now known as gregglind
# [22:44] * Joins: bjacob (bjacob@F2D29657.F60B0462.67AC9B1.IP)
# [22:44] * Joins: roc (chatzilla@538BABFE.A073F3E.97BBD552.IP)
# [22:44] * Quits: andreasn (andreasn@moz-BB6A8755.bredband.comhem.se) (Ping timeout)
# [22:44] <jbuck> (He specifically mentions doing animation w/ Popcorn.js in that ticket, which is why I mention it)
# [22:45] * Joins: clee (clee@moz-D0C71163.domainhotel-wifi.mozilla.hq)
# [22:46] * Joins: mib_v0vumx (Mibbit@F6A9DE18.A1146449.57460053.IP)
# [22:46] * LegNeato_away is now known as LegNeato
# [22:47] <Unfocused> jbuck: the bug was originally for changing the behaviour of timeupdate though, so i'd recommend commenting that his use-case (popcorn.js) now uses an alternate solution
# [22:47] * Quits: florian (florian@moz-87C33FDA.kimsufi.com) (Ping timeout)
# [22:48] * Joins: anant (anant@moz-D0C71163.domainhotel-wifi.mozilla.hq)
# [22:48] * Joins: florian (florian@moz-87C33FDA.kimsufi.com)
# [22:48] * Joins: anant_ (anant@moz-104CC309.mv.mozilla.com)
# [22:48] * Quits: kaze (kaze@moz-74A6AC89.w82-123.abo.wanadoo.fr) (Ping timeout)
# [22:48] <jesup> Unfocused++
# [22:49] <Unfocused> then cdouble, who owns some of that code, can decide whether to resolve it WONTFIX or not
# [22:49] * Quits: nthomas (chatzilla@moz-75DA4CCD.dsl.telstraclear.net) (Ping timeout)
# [22:49] <Unfocused> :)
# [22:49] * Joins: kumar (kmcmillan@moz-D0C71163.domainhotel-wifi.mozilla.hq)
# [22:50] * Quits: anant (anant@moz-D0C71163.domainhotel-wifi.mozilla.hq) (Ping timeout)
# [22:50] * anant_ is now known as anant
# [22:50] <jbuck> alright, thanks
# [22:51] <roc> RTL MathML is freaky
# [22:51] * Joins: nthomas (chatzilla@moz-75DA4CCD.dsl.telstraclear.net)
# [22:51] <@bz> too many 'L's
# [22:52] * jmaher is now known as jmaher|afk
# [22:52] * Quits: protz (jonathan@moz-7F6750F6.xulforum.org) (Quit: Leaving)
# [22:52] <bholley> why do we want to avoid AssignWithConversion et al?
# [22:52] * Quits: graememcc (chatzilla@moz-3E78DB35.range86-148.btcentralplus.com) (Ping timeout)
# [22:53] <biesi> bholley, it is often the wrong thing (breaking i18n)
# [22:53] <biesi> bholley, but it is so appealing
# [22:53] * Quits: chrisccoulson_ (chr1s@moz-B376CDB0.business.broadband.hu) (Ping timeout)
# [22:53] * Joins: graememcc (chatzilla@moz-FA3F4331.range109-158.btcentralplus.com)
# [22:54] <bholley> biesi: why is NS_ConvertUTF8toUTF16 better?
# [22:54] * Quits: mjschranz (mjschranz@moz-57B23647.senecac.on.ca) (Client exited)
# [22:54] * Quits: robhawkes (robhawkes@moz-33A339B7.dsl.cnl.uk.net) (Connection reset by peer)
# [22:54] * Joins: robhawkes (robhawkes@moz-33A339B7.dsl.cnl.uk.net)
# [22:54] * davehunt|away is now known as davehunt
# [22:55] <jbuck> bz: question regarding https://bugzilla.mozilla.org/show_bug.cgi?id=682299 , when you said "The imagelib changes here don't look necessary. Please undo those." just remove all of the changes I made in /image ?
# [22:55] * Joins: chrisccoulson (chr1s@moz-B376CDB0.business.broadband.hu)
# [22:55] * coop|mtg is now known as coop
# [22:55] <Mook_as> bholley: because AssignWithConversion just lops off the high bytes according to http://mxr.mozilla.org/mozilla-central/source/xpcom/string/src/nsStringObsolete.cpp#1053
# [22:56] <biesi> bholley, because it converts from UTF-8, not from ASCII ;)
# [22:56] <bholley> biesi: ah, I see
# [22:57] <biesi> bholley, and because it makes it explicit what it converts from and to
# [22:57] * Quits: johanc (chatzilla@moz-D8A1AA43.bredband.comhem.se) (Ping timeout)
# [22:57] <biesi> oh, we moved libpr0n to /image? nice
# [22:58] * bholley did that
# [22:58] <sstangl> no more libpr0n?
# [22:58] <sstangl> aw :(
# [22:58] * Quits: robhawkes (robhawkes@moz-33A339B7.dsl.cnl.uk.net) (Quit: Linkinus - http://linkinus.com)
# [22:58] <bholley> biesi: so there's no danger in converting AssignWithConversion to NS_ConvertUTF8toUTF16?
# [22:58] <bholley> sstangl: this came up at my mozcamp talk ;-)
# [22:58] <biesi> bholley, there certainly is danger
# [22:58] <sstangl> bholley: aw :(
# [22:58] <bholley> sstangl: whoops
# [22:58] <@bz> jbuck: I think so, yes
# [22:59] <biesi> bholley, if the string is in fact in ISO-8859-1 (latin1), then that change would do the wrong thing
# [22:59] * Quits: mak|afk (chatzilla@moz-C36E0FF8.retail.telecomitalia.it) (Quit: ChatZilla 0.9.86.1-rdmsoft [XULRunner 2.0/20110318052756])
# [22:59] <bholley> sstangl: thought you were mstange, who was there ;-)
# [22:59] <roc> mstange has been sighted?
# [22:59] <@bz> bholley: where are you running into AssignWithConversion ?
# [23:00] * Quits: graememcc (chatzilla@moz-FA3F4331.range109-158.btcentralplus.com) (Quit: ChatZilla 0.9.88 [Firefox 9.0.1/20111221233052])
# [23:00] <bholley> bz, biesi: bug 716333
# [23:00] * Cwiiis is now known as CwiiisAway
# [23:00] <bholley> roc: yeah, we hung out in berlin
# [23:00] * Joins: jorendorff (jorendorff@moz-FAB36621.hsd1.tn.comcast.net)
# [23:01] <biesi> bholley, do we compile with that JSSTRINGS_ARE_UTF8 flag now?
# [23:01] <bholley> biesi: I don't know
# [23:01] <biesi> bholley, well the general answer is that you have to figure out what encoding the string actually is in
# [23:01] * Quits: bsmith (bsmith@moz-E34A3D24.dsl.pltn13.sbcglobal.net) (Ping timeout)
# [23:01] <biesi> and then use the appropriate conversion function
# [23:01] * Quits: myk (myk@moz-D0C71163.domainhotel-wifi.mozilla.hq) (Max SendQ exceeded)
# [23:01] * Joins: myk (myk@moz-D0C71163.domainhotel-wifi.mozilla.hq)
# [23:01] <@bz> bholley: ugh
# [23:02] <bholley> biesi: ok. So AssignWithConversion for ascii, NS_ConvertUTF8toUTF16 for UTF8?
# [23:02] * Quits: Yoric (Yoric@moz-920DB13B.fbx.proxad.net) (Input/output error)
# [23:02] * Quits: evilpie (evilpie@moz-37567C56.pools.arcor-ip.net) (Quit: ChatZilla 0.9.88 [Firefox 12.0a1/20120110031111])
# [23:02] <biesi> bholley, no - for ascii, you use AssignASCII (or in the other direction, NS_LossyConvertUTF16toASCII)
# [23:02] <biesi> (can also use NS_ConvertASCIItoUTF16)
# [23:03] * Quits: Joeh (joe@5A3923AA.BC22908.C7CEC4ED.IP) (Connection reset by peer)
# [23:03] <@bz> fwiw
# [23:03] <@bz> I bet the stuff you get out of jseng is a mix of utf-8 and iso-8859-1
# [23:03] <@bz> sometimes in the same string
# [23:03] <@bz> as in "broken beyond believe for non-ASCII)
# [23:03] <@bz> er, "
# [23:04] * Quits: Ami_Ty (Amie@moz-BBE3ABD.mv.mozilla.com) (Input/output error)
# [23:05] * Quits: necolas (necolas@moz-C8C6D8F3.bb.sky.com) (Client exited)
# [23:05] <bholley> bz: so what do you think we should do here?
# [23:05] * Quits: msucan (msucan-@FA9E8863.56E67207.699550A1.IP) (Quit: .)
# [23:05] * Joins: njn (chatzilla@moz-67BF6169.dyn.iinet.net.au)
# [23:05] * Quits: azakai (alon@moz-BBE3ABD.mv.mozilla.com) (Ping timeout)
# [23:06] * philor is now known as philor|away
# [23:07] * Joins: rjohnson19 (rjohnson19@moz-9148485F.hsd1.ma.comcast.net)
# [23:09] * Quits: flx_ (flx@moz-F4992FEC.uwaterloo.ca) (Ping timeout)
# [23:09] <hub> do you guess regularly check the "checking-needed" bugs?
# [23:10] <hub> s/guess/guys/
# [23:10] * Joins: Ami_Ty (Amie@moz-BBE3ABD.mv.mozilla.com)
# [23:10] <bholley> hub: "checkin-needed", and yes
# [23:10] <hub> yeah I just had the spelling wrong on IRC
# [23:10] * Joins: gandalf (zbraniecki@moz-9DD54591.neoplus.adsl.tpnet.pl)
# [23:10] <bholley> hub: put that in your keywords and edmorley will probably push it for you sometime
# [23:11] <hub> yeah I have done that already
# [23:11] <hub> just making sure
# [23:11] <bholley> hub: :-)
# [23:11] <hub> thanks
# [23:11] <@bz> bholley: other than crying?
# [23:12] * hub can't wait for L3 access
# [23:12] <@bz> bholley: I don't think there's a sane way to do this right
# [23:12] <bholley> bz: should I r+ the patch?
# [23:12] <@bz> bholley: given the APIs you have to work with
# [23:12] <@bz> bholley: I wouldn't, personally
# [23:12] <@bz> bholley: since those aren't utf-8 strings
# [23:12] <@bz> bholley: unfortunately they're not anything in particular strings....
# [23:12] * heycam|away is now known as heycam
# [23:12] * Joins: darktrojan (geoff@moz-57EC95C5.dsl.telstraclear.net)
# [23:12] <bholley> bz: doesn't the JS engine have some sort of standard?
# [23:13] * Quits: jrmuizel (jrmuizel@F2D29657.F60B0462.67AC9B1.IP) (Input/output error)
# [23:13] * Quits: bbondy (bbondy@moz-28CF6D1C.home.cgocable.net) (Ping timeout)
# [23:15] <bholley> luke: apropos the above: Do JSAPI strings like JSErrorReport::filename have any assumed encoding?
# [23:15] * Quits: jorendorff (jorendorff@moz-FAB36621.hsd1.tn.comcast.net) (Quit: Eaten by grue.)
# [23:15] * Quits: KLB (Kenneth_Ba@moz-CB90BBF2.maine.res.rr.com) (Ping timeout)
# [23:15] <bholley> luke: I'm assuming that jschar strings are utf16, but bz thinks there's no right answer for char*s
# [23:15] <luke> bholley: there is a comment in error handling code referring to weeping l10n angles,
# [23:15] <luke> bholley: so i fear the worst
# [23:16] <luke> *angels
# [23:16] * Joins: KLB (Kenneth_Ba@moz-CB90BBF2.maine.res.rr.com)
# [23:16] <luke> bholley: actually, for JSErrorReport::filename, i think it is just a simple copy of whatever JSScript::filename had,
# [23:16] <luke> bholley: and that is passed to JSAPI
# [23:16] <luke> i think it is more-or-less a blackbox to the JS engine
# [23:18] <@bz> bholley: nope
# [23:18] <@bz> bholley: js engine likes to pretend char* is an opaque byte array, not a string
# [23:18] * Quits: askalski (akuda@moz-6A36EC49.ip.abpl.pl) (Quit: Wychodzi)
# [23:18] <@bz> bholley: the filename is going to be UTF-8
# [23:18] * Quits: bmoss (bmoss@moz-BBE3ABD.mv.mozilla.com) (Ping timeout)
# [23:18] <bholley> bz: but jschar* is utf16, right?
# [23:18] <@bz> bholley: other stuff from inside the js eng itself, who knows
# [23:18] <@bz> bholley: yes
# [23:18] <@bz> bholley: well
# [23:18] <@bz> bholley: "no"
# [23:19] <@bz> bholley: jschar* is just arrays of 2-byte integers
# [23:19] <@bz> bholley: it doesn't have to be valid UTF-16
# [23:19] <@bz> bholley: and it's trivial to create JSStrings that aren't
# [23:19] * Joins: bmoss (bmoss@moz-BBE3ABD.mv.mozilla.com)
# [23:19] <@bz> bholley: though if you don't do manual char-by-char string building it will tend to be UTF-16
# [23:19] <@bz> bholley: (of course the same is true of nsString)
# [23:20] <firebot> Check-in: http://hg.mozilla.org/releases/mozilla-1.9.2/rev/a46281bc813e - Michal Novotny - Bug 504014 - Enforce RFC 3986 syntax for IPv6 literals a=dveditz
# [23:20] * Quits: chrisccoulson (chr1s@moz-B376CDB0.business.broadband.hu) (Ping timeout)
# [23:21] * Quits: bmoss (bmoss@moz-BBE3ABD.mv.mozilla.com) (Ping timeout)
# [23:21] * Joins: jrmuizel (jrmuizel@F2D29657.F60B0462.67AC9B1.IP)
# [23:22] * bwinton is now known as bwinton_away
# [23:22] * Joins: adeubank (Mibbit@moz-9B1B1DF8.allocated.csupomona.edu)
# [23:22] * Quits: AutomatedTester (AutomatedT@moz-D4F73363.as13285.net) (Quit: AutomatedTester)
# [23:22] * Joins: chrisccoulson (chr1s@moz-B376CDB0.business.broadband.hu)
# [23:23] * Joins: azakai (alon@moz-BBE3ABD.mv.mozilla.com)
# [23:23] * Joins: mjschranz (mjschranz@moz-57B23647.senecac.on.ca)
# [23:23] <bholley> biesi, bz: Ah, it looks like this is all part of bug 113234
# [23:24] * jwir3|away is now known as jwir3
# [23:24] * Quits: Tobbi (Tobbi@16BAC97A.933EA279.AC7F8427.IP) (Ping timeout)
# [23:25] * Joins: AutomatedTester (AutomatedT@moz-D4F73363.as13285.net)
# [23:25] * Quits: armenzg (armenzg@F2D29657.F60B0462.67AC9B1.IP) (Input/output error)
# [23:25] * Quits: stevee (Miranda@moz-BEBDF855.cable.virginmedia.com) (Connection reset by peer)
# [23:25] * Quits: Mnyromyr (Mnyromyr@B2521176.7B0892CB.771966F7.IP) (Ping timeout)
# [23:25] * bholley goes to run more dromaeo
# [23:25] * Joins: johanc (chatzilla@moz-D8A1AA43.bredband.comhem.se)
# [23:25] * Quits: bholley (bholley@moz-FCAF9AAB.hsd1.ca.comcast.net) (Quit: bholley)
# [23:25] <froydnj> when was the summer summit? july? or june?
# [23:25] * Quits: MMx (mmx@moz-1AFE3502.afthd.hg.tu-darmstadt.de) (Ping timeout)
# [23:26] * Quits: adeubank (Mibbit@moz-9B1B1DF8.allocated.csupomona.edu) (Quit: http://www.mibbit.com ajax IRC Client)
# [23:26] * Quits: Dagger (Dagger@moz-C46A7175.cable.virginmedia.com) (Ping timeout)
# [23:26] * Quits: imphil (philipp@moz-655EF802.customer.m-online.net) (Ping timeout)
# [23:26] * Joins: MMx (mmx@moz-1AFE3502.afthd.hg.tu-darmstadt.de)
# [23:27] * Quits: Standard8 (Standard8@B7F1AE36.48015583.54C3481B.IP) (Quit: ZNC - http://znc.in)
# [23:27] <mbrubeck> july 2010
# [23:27] <darktrojan> froydnj, the last one was the first week of july iirc
# [23:27] <froydnj> er, sorry, that should have been future tense
# [23:27] <mbrubeck> And the next one is July 2012
# [23:27] * Joins: imphil (philipp@moz-655EF802.customer.m-online.net)
# [23:27] * juanb|lunch is now known as juanb
# [23:27] <mbrubeck> We were told to expect it in the second, third, or fourth week of July.
# [23:28] * Quits: faramarz (faramarz@moz-BBE3ABD.mv.mozilla.com) (Input/output error)
# [23:28] * Joins: faramarz (faramarz@moz-BBE3ABD.mv.mozilla.com)
# [23:28] * Quits: johanc (chatzilla@moz-D8A1AA43.bredband.comhem.se) (Ping timeout)
# [23:28] <biesi> mbrubeck, do you know more details? like where it will be?
# [23:28] * Joins: tonymec (tonymec@5B2A1B79.5A39B178.277517C1.IP)
# [23:28] <mbrubeck> no, all they've told is is the rough date
# [23:28] <Waldo> fwiw, anyone who would occasionally take advantage of an /optional/ message argument to be included in MOZ_ASSERT should feel free to comment in bug 716112; this would permit both MOZ_ASSERT(arg != NULL) and MOZ_ASSERT(cond, "this is what the condition really means"), or MOZ_ASSERT(non-obvious-thing, "we can assert this because of cosmic rays in that file ten nesting levels away over there")
# [23:28] * Joins: Mnyromyr (Mnyromyr@B2521176.7B0892CB.771966F7.IP)
# [23:29] <biesi> mbrubeck, aw ok. thx
# [23:29] <mbrubeck> I assume the planners are still trying to lock down a venue.
# [23:29] * philor|away is now known as philor
# [23:29] * Joins: flx_ (flx@moz-F4992FEC.uwaterloo.ca)
# [23:30] * darktrojan checks his calendar
# [23:30] <darktrojan> woo, might not be double-booked this time
# [23:30] <darktrojan> er, triple-booked
# [23:31] * Quits: mconley (mconley@F2D29657.F60B0462.67AC9B1.IP) (Connection reset by peer)
# [23:32] * Joins: mconley (mconley@F2D29657.F60B0462.67AC9B1.IP)
# [23:32] * Joins: Dagger (Dagger@moz-C46A7175.cable.virginmedia.com)
# [23:32] * Quits: chrisccoulson (chr1s@moz-B376CDB0.business.broadband.hu) (Ping timeout)
# [23:33] * Joins: birtles (chatzilla@33CB57DB.B50A5FC0.DE671710.IP)
# [23:33] * Quits: nattokirai (nattokirai@moz-6A258170.dynamic.ppp.asahi-net.or.jp) (Quit: nattokirai)
# [23:33] <gps> Unfocused: so, bug 708134. how can I do your bidding?
# [23:34] * Joins: brendan (brendaneic@moz-BBE3ABD.mv.mozilla.com)
# [23:34] * Quits: mib_v0vumx (Mibbit@F6A9DE18.A1146449.57460053.IP) (Quit: http://www.mibbit.com ajax IRC Client)
# [23:34] <Waldo> gps: you forgot "my master"
# [23:35] * Quits: @bz (bzbarsky@moz-69B5879F.bstnma.fios.verizon.net) (Ping timeout)
# [23:37] * Joins: chrisccoulson (chr1s@moz-B376CDB0.business.broadband.hu)
# [23:38] * Joins: bmoss (bmoss@moz-BBE3ABD.mv.mozilla.com)
# [23:38] * Quits: bmoss (bmoss@moz-BBE3ABD.mv.mozilla.com) (Quit: bmoss)
# [23:39] * Joins: bmoss (bmoss@moz-BBE3ABD.mv.mozilla.com)
# [23:39] <Unfocused> gps: besides my brain dump in comment 4?
# [23:40] <gps> Unfocused: I'd like to know which brain matter has legs
# [23:40] * Quits: chrisccoulson (chr1s@moz-B376CDB0.business.broadband.hu) (Ping timeout)
# [23:40] * Quits: TheLink (TheLink@moz-B5AB52CB.pools.arcor-ip.net) (Client exited)
# [23:40] * Joins: jorendorff (jorendorff@moz-FAB36621.hsd1.tn.comcast.net)
# [23:41] <darktrojan> jaws, I'm confused by the choice of mockups in your blog post
# [23:41] <gps> there was another bug filed yesterday and I wasn't sure if/how that would impact this one
# [23:42] <Unfocused> gps: so, use RepositoryAddon.sourceURI to get the xpi location - do a dumb string replace. i'm not entirely sure we should bother with preserving non-AMO urls, since the src param is relatively common practice
# [23:42] * Quits: bmoss (bmoss@moz-BBE3ABD.mv.mozilla.com) (Max SendQ exceeded)
# [23:42] * Joins: bmoss (bmoss@moz-BBE3ABD.mv.mozilla.com)
# [23:42] * Quits: rclick|afk (rclick@EA4B602E.692CFA54.E430BA25.IP) (Ping timeout)
# [23:43] <gps> I think I can handle that
# [23:43] <jaws> darktrojan: i used the only mockups i could find
# [23:43] <jaws> darktrojan: do you know of any others?
# [23:43] <Unfocused> and for the metadata ping, add a param to getInstallFromURL, hat takes a constant similar to the AddonManager.UPDATE_WHEN_* constants
# [23:44] <jaws> darktrojan: admittedly, they are a bit outdated
# [23:44] <darktrojan> I think there might be others, I'll have a look
# [23:44] <Unfocused> i *think* that covers stuff well enough
# [23:44] <darktrojan> jaws, the first and last one are the same, to start with :)
# [23:45] <jaws> darktrojan: hmmmph, i must have done a failed copy/paste when editing the html. thanks for letting me know
# [23:45] <darktrojan> still, keep up the good work, I've been wanting that stuff to happen for ages
# [23:45] <jaws> :)
# [23:47] * Joins: clokep (clokep@moz-69FB3955.c3-0.smr-ubr1.sbo-smr.ma.cable.rcn.com)
# [23:47] * Quits: clokep (clokep@moz-69FB3955.c3-0.smr-ubr1.sbo-smr.ma.cable.rcn.com) (Connection reset by peer)
# [23:48] * Joins: clokep (clokep@moz-69FB3955.c3-0.smr-ubr1.sbo-smr.ma.cable.rcn.com)
# [23:48] * Quits: wolfiR (wolfiR@moz-B660BA51.dip0.t-ipconnect.de) (Quit: Leaving)
# [23:48] <@dbaron> what's the channel where the JS folks hang out called?
# [23:48] <gavin> #jsapi
# [23:48] <Unfocused> #jsapi ?
# [23:48] <@dbaron> yeah
# [23:48] <@dbaron> thanks
# [23:49] * Quits: clee (clee@moz-D0C71163.domainhotel-wifi.mozilla.hq) (Quit: clee)
# [23:49] * Quits: tonymec (tonymec@5B2A1B79.5A39B178.277517C1.IP) (Ping timeout)
# [23:49] * Quits: tonymec|away (tonymec@5B2A1B79.5A39B178.277517C1.IP) (Ping timeout)
# [23:50] <darktrojan> :( I wish we had a way to quickly go back to the previous page, not just the previous #target
# [23:50] <firebot> Check-in: http://hg.mozilla.org/mozilla-central/rev/83afca21c9e6 - Joe Walker - Bug 689605 - GCLI should be ready to be shipped, preffed on; r=msucan
# [23:51] <firebot> http://hg.mozilla.org/mozilla-central/rev/62cca592625d - Joe Walker - Bug 710152 - GCLI javascript completion should stop providing completions with complex JS input; r=dcamp
# [23:51] <firebot> http://hg.mozilla.org/mozilla-central/rev/1962ca549264 - Tim Taubert - merge m-c to fx-team
# [23:51] <firebot> http://hg.mozilla.org/mozilla-central/rev/8409adc46f14 - Joe Walker - Bug 704182 - webconsole forces wacky lists on GCLI; r=dao,msucan
# [23:51] <firebot> http://hg.mozilla.org/mozilla-central/rev/a64fb47161e7 - Joe Walker - Bug 704184 - Find a better way to scroll web console output; r=dcamp
# [23:51] <firebot> http://hg.mozilla.org/mozilla-central/rev/9ea2996a1f3c - Joe Walker - Bug 709748 - GCLI 'help' command causes firefox to crash; r=paul
# [23:51] <firebot> http://hg.mozilla.org/mozilla-central/rev/fdab6d891be0 - Berker Peksag - Bug 683510 - GCLI needs a 'console' command; r=msucan
# [23:51] <jwir3> since we're landing new things on inbound now, maybe it would be better if firebot reported check ins for inbound, rather than central?
# [23:51] <jaws> darktrojan: i've fixed the images in the post. thanks again
# [23:52] <jwir3> because we're basically getting reports for merges from inbound to m-c
# [23:52] * Quits: Callek (chatzilla@moz-DD17331C.dhcp.oxfr.ma.charter.com) (Ping timeout)
# [23:52] <darktrojan> jaws, that looks better
# [23:53] <darktrojan> jaws, is the grid view for the AOM actually on the todo list?
# [23:53] <darktrojan> (please say yes, it's hot)
# [23:55] * Quits: azakai (alon@moz-BBE3ABD.mv.mozilla.com) (Ping timeout)
# [23:55] * Joins: rclick|afk (rclick@moz-F5321782.dsl.covlil.sbcglobal.net)
# [23:55] * rclick|afk is now known as rclick
# [23:55] <Unfocused> doubtful
# [23:55] <Unfocused> for normal extensions, at least
# [23:55] <darktrojan> I know :(
# [23:55] <Unfocused> but i'm currently fixing up a patch to do basically that for themes/backgrounds
# [23:56] * Quits: birtles (chatzilla@33CB57DB.B50A5FC0.DE671710.IP) (Client exited)
# [23:57] * Joins: birtles (chatzilla@33CB57DB.B50A5FC0.DE671710.IP)
# [23:57] * philor is now known as philor|away
# [23:57] <Unfocused> i'd forgotten how nice that looked... might borrow some stuff from that grig view mockup
# [23:57] <Unfocused> er, grid
# [23:57] <decoder> is "make package" stripping symbols out of the binaries? can this be prevented somehow?
# [23:57] <decoder> addr2line doesnt seem to work anymore after I packaged the build
# [23:57] <decoder> but it does work before packaging
# [23:58] * Quits: Unfocused (u2622@moz-160C58C6.com) (Max SendQ exceeded)
# [23:58] * Joins: Unfocused (u2622@moz-160C58C6.com)
# [23:59] * Quits: rs (rs@moz-BBE3ABD.mv.mozilla.com) (Quit: ChatZilla 0.9.86.1-rdmsoft [XULRunner 1.8.0.9/2006120508])
# [23:59] * bhearsum is now known as bhearsum|afk
# Session Close: Wed Jan 11 00:00:01 2012

The end :)