/irc-logs / freenode / #whatwg / 2009-09-08 / end

Options:

# Session Start: Tue Sep 08 00:00:00 2009
# Session Ident: #whatwg
# [00:00] * Joins: nessy (n=nessy@203-214-73-15.dyn.iinet.net.au)
# [00:01] * Quits: Creap (n=Creap@vemod.brg.sgsnet.se) (Remote closed the connection)
# [00:05] * Quits: maikmerten_ (n=maikmert@BAE3af2.bae.pppool.de) (Remote closed the connection)
# [00:07] * Joins: dglazkov_ (n=dglazkov@c-67-188-0-62.hsd1.ca.comcast.net)
# [00:11] * Quits: dglazkov_ (n=dglazkov@c-67-188-0-62.hsd1.ca.comcast.net) (Client Quit)
# [00:14] * Quits: heycam (n=cam@210-84-56-211.dyn.iinet.net.au) ("bye")
# [00:15] * Joins: Rik` (n=Rik`@pha75-2-81-57-187-57.fbx.proxad.net)
# [00:23] * Joins: cardona507 (n=cardona5@c-67-180-160-250.hsd1.ca.comcast.net)
# [00:24] * Quits: cardona507 (n=cardona5@c-67-180-160-250.hsd1.ca.comcast.net) (Client Quit)
# [00:24] * Quits: dglazkov (n=dglazkov@72.14.224.1) (Read error: 110 (Connection timed out))
# [00:24] * Joins: cardona507 (n=cardona5@c-67-180-160-250.hsd1.ca.comcast.net)
# [00:31] * Joins: dglazkov (n=dglazkov@c-67-188-0-62.hsd1.ca.comcast.net)
# [00:39] * Joins: kinetik_ (n=kinetik@121.98.132.55)
# [00:39] * Quits: kinetik (n=kinetik@121.98.132.55) (Read error: 131 (Connection reset by peer))
# [00:43] * Quits: svl (n=me@ip565744a7.direct-adsl.nl) ("And back he spurred like a madman, shrieking a curse to the sky.")
# [00:44] * Quits: cardona507 (n=cardona5@c-67-180-160-250.hsd1.ca.comcast.net)
# [00:46] * Quits: dglazkov (n=dglazkov@c-67-188-0-62.hsd1.ca.comcast.net)
# [00:48] * Quits: ZombieLoffe (n=e@unaffiliated/zombieloffe) (Client Quit)
# [00:48] * Joins: heycam (n=cam@clm-laptop.infotech.monash.edu.au)
# [00:50] * Joins: kristallpirat (n=kristall@c-base/crew/kristall)
# [00:51] * Joins: cardona507 (n=cardona5@c-67-180-160-250.hsd1.ca.comcast.net)
# [00:55] * Quits: GPHemsley (n=GPHemsle@pdpc/supporter/student/GPHemsley) ("This computer has gone to sleep")
# [00:56] * Joins: benward (n=benward@c-24-5-83-43.hsd1.ca.comcast.net)
# [00:57] * Quits: benward (n=benward@c-24-5-83-43.hsd1.ca.comcast.net) (Remote closed the connection)
# [01:03] * Quits: MikeSmith (n=MikeSmit@EM114-48-146-158.pool.e-mobile.ne.jp) (Read error: 104 (Connection reset by peer))
# [01:04] * Quits: epeus (n=KevinMar@c-67-164-14-96.hsd1.ca.comcast.net) ("The computer fell asleep")
# [01:04] * Joins: epeus (n=KevinMar@c-67-164-14-96.hsd1.ca.comcast.net)
# [01:18] * kinetik_ is now known as kinetik
# [01:22] * Quits: epeus (n=KevinMar@c-67-164-14-96.hsd1.ca.comcast.net) (Read error: 110 (Connection timed out))
# [01:25] * Quits: cardona507 (n=cardona5@c-67-180-160-250.hsd1.ca.comcast.net)
# [01:33] * Quits: Kalms (n=rasmuska@81.161.185.108)
# [01:37] * Joins: roc (n=roc@203-97-204-82.dsl.clear.net.nz)
# [01:38] * Joins: cardona507 (n=cardona5@c-67-180-160-250.hsd1.ca.comcast.net)
# [01:44] * Joins: doublec (n=doublec@203-97-204-82.dsl.clear.net.nz)
# [01:47] * Joins: steevel (n=steevel@irc.kodslav.org)
# [01:50] * Joins: epeus (n=KevinMar@c-67-164-14-96.hsd1.ca.comcast.net)
# [01:57] * Quits: annevk3 (n=annevk@83.85.115.44)
# [02:00] * Joins: benward (n=benward@98.210.154.133)
# [02:02] * Joins: yutak_home (n=kee@M006079.ppp.dion.ne.jp)
# [02:02] * Joins: shepazu (n=schepers@pool-96-252-125-220.bstnma.fios.verizon.net)
# [02:05] * Quits: Rik` (n=Rik`@pha75-2-81-57-187-57.fbx.proxad.net) (Read error: 60 (Operation timed out))
# [02:05] * Joins: tantek (n=tantek@adsl-69-107-12-74.dsl.pltn13.pacbell.net)
# [02:09] * Quits: benward (n=benward@98.210.154.133)
# [02:10] * Joins: GPHemsley (n=GPHemsle@pdpc/supporter/student/GPHemsley)
# [02:21] * Joins: Unixmonkey (n=Unixmonk@ppp-69-208-150-144.dsl.ipltin.ameritech.net)
# [02:27] * Quits: yutak_home (n=kee@M006079.ppp.dion.ne.jp) ("Ex-Chat")
# [02:28] * Joins: wakaba_ (n=wakaba_@122x221x184x68.ap122.ftth.ucom.ne.jp)
# [02:35] * Joins: Unixmonkey_ (n=Unixmonk@ppp-69-208-150-144.dsl.ipltin.ameritech.net)
# [02:36] * Quits: Unixmonkey (n=Unixmonk@ppp-69-208-150-144.dsl.ipltin.ameritech.net) (Read error: 104 (Connection reset by peer))
# [02:36] * Unixmonkey_ is now known as Unixmonkey
# [02:42] * Quits: doublec (n=doublec@203-97-204-82.dsl.clear.net.nz) ("Leaving")
# [02:42] * kristallpirat is now known as StationsPatient
# [02:44] * Quits: Unixmonkey (n=Unixmonk@ppp-69-208-150-144.dsl.ipltin.ameritech.net)
# [02:45] * Joins: doublec (n=doublec@203-97-204-82.dsl.clear.net.nz)
# [02:54] * Joins: boblet (n=boblet@p1254-ipbf304osakakita.osaka.ocn.ne.jp)
# [02:55] * Joins: MikeSmith (n=MikeSmit@EM114-48-201-185.pool.e-mobile.ne.jp)
# [02:59] * Joins: Unixmonkey (n=Unixmonk@ppp-69-208-150-144.dsl.ipltin.ameritech.net)
# [03:04] * Joins: dglazkov (n=dglazkov@c-67-188-0-62.hsd1.ca.comcast.net)
# [03:05] <AryehGregor> MikeSmith, do you have any insight into <http://bugzilla.validator.nu/show_bug.cgi?id=641>, or know who might? (FWIW: <http://validator.nu/?doc=http://www.wikipedia.org>)
# [03:05] * MikeSmith looks now
# [03:07] * Quits: MikeSmith (n=MikeSmit@EM114-48-201-185.pool.e-mobile.ne.jp) ("Tomorrow to fresh woods, and pastures new.")
# [03:07] * Quits: boblet (n=boblet@p1254-ipbf304osakakita.osaka.ocn.ne.jp)
# [03:09] * Joins: MikeSmith (n=MikeSmit@EM114-48-201-185.pool.e-mobile.ne.jp)
# [03:12] <MikeSmith> AryehGregor: hsivonen would definitely have insight into that, if he had time to look at in detail. but I think his time for validator.nu bug fixing is limited right now. so in the mean time, I can take a look at it unless he beats me to it.
# [03:12] <AryehGregor> Thanks. It's not urgent.
# [03:13] * Joins: jwalden (n=waldo@c-98-248-40-206.hsd1.ca.comcast.net)
# [03:14] <MikeSmith> OK, I've already been looking at that same code over the weekend, for the other bug you reported
# [03:14] <MikeSmith> http://bugzilla.validator.nu/show_bug.cgi?id=640
# [03:14] <MikeSmith> I already have a fix written up for that
# [03:14] <MikeSmith> partial fix
# [03:14] <AryehGregor> Great.
# [03:15] <AryehGregor> I'm sure some of the codes are actually wrong. When I can figure out which ones, I'll talk to our language people and get them replaced with more correct ones.
# [03:15] <MikeSmith> AryehGregor: actually, most of them are not outright wrong -- there's a number of them that are valid, but deprecated
# [03:16] <MikeSmith> like "zh-yue" -- that should just be "yue" instead
# [03:16] <AryehGregor> Yeah, we know about most of those, I think.
# [03:16] <AryehGregor> We've sort of accreted them over the years.
# [03:16] <MikeSmith> well, the validator should not actually report those as errors
# [03:16] <AryehGregor> Once we've set up a wiki, it's a pain to change the domain, and it doesn't really hurt anything to keep it, so we tend to leave it alone.
# [03:16] <MikeSmith> yeah, understood
# [03:16] <AryehGregor> ("we" not really including me, I'm not a sysadmin and can't change the domains)
# [03:17] <AryehGregor> It's easy to change the lang="" attribute on that one page, though.
# [03:17] <MikeSmith> OK. well please leave it for now if you can
# [03:17] <MikeSmith> until we get this bug resolved
# [03:17] <AryehGregor> No problem.
# [03:17] <MikeSmith> It's a good test case.
# [03:17] <MikeSmith> so anyway, validator.nu should just report those as warnings, not errors
# [03:18] <MikeSmith> because they are not strictly conformance errors
# [03:18] <MikeSmith> the problem is, in the current validator.nu datatype-checking code, there's no mechanism for emitting warnings about datatypes of attribute values
# [03:19] <MikeSmith> it is currently only set up to emit error messages
# [03:19] <AryehGregor> There are probably other cases where you'd want warnings for attribute values, but I can't think where offhand.
# [03:20] <MikeSmith> yeah, I'm sure we will find some others, anyway
# [03:20] * Joins: jorlow (n=jorlow@220.109.219.244)
# [03:20] <MikeSmith> AryehGregor: so I need to ask hsivonen for some guidance about what he thinks the best way to refine the code to emit warnings
# [03:20] <AryehGregor> Okay, no problem. As I say, not urgent.
# [03:20] <MikeSmith> OK
# [03:22] * Joins: murr4y (n=murray@85.84-49-67.nextgentel.com)
# [03:34] * Quits: murr4y` (n=murray@85.84-49-67.nextgentel.com) (Read error: 110 (Connection timed out))
# [03:42] * Quits: cardona507 (n=cardona5@c-67-180-160-250.hsd1.ca.comcast.net)
# [03:43] * Joins: boblet (n=boblet@p1254-ipbf304osakakita.osaka.ocn.ne.jp)
# [03:46] * Quits: Unixmonkey (n=Unixmonk@ppp-69-208-150-144.dsl.ipltin.ameritech.net)
# [03:52] * Quits: dglazkov (n=dglazkov@c-67-188-0-62.hsd1.ca.comcast.net)
# [03:56] * Quits: MikeSmith (n=MikeSmit@EM114-48-201-185.pool.e-mobile.ne.jp) ("Tomorrow to fresh woods, and pastures new.")
# [04:35] * Joins: MikeSmith (n=MikeSmit@dhcp-247-107.mag.keio.ac.jp)
# [04:36] * Joins: weinig_ (n=weinig@c-67-180-35-124.hsd1.ca.comcast.net)
# [04:41] * Joins: lazni (n=lazni@118.71.169.21)
# [04:45] * Quits: othermaciej (n=mjs@c-69-181-42-237.hsd1.ca.comcast.net)
# [04:50] * Joins: othermaciej (n=mjs@c-69-181-42-237.hsd1.ca.comcast.net)
# [04:51] * Quits: MikeSmith (n=MikeSmit@dhcp-247-107.mag.keio.ac.jp) ("Tomorrow to fresh woods, and pastures new.")
# [04:51] * Quits: weinig (n=weinig@c-67-180-35-124.hsd1.ca.comcast.net) (Read error: 110 (Connection timed out))
# [04:51] * weinig_ is now known as weinig
# [04:56] * Joins: MikeSmith (n=MikeSmit@tea12.w3.mag.keio.ac.jp)
# [04:57] * Joins: dglazkov (n=dglazkov@c-67-188-0-62.hsd1.ca.comcast.net)
# [05:09] <Hixie> http://www.bigmouthmedia.com/live/articles/html-5-draft-hyperlink-auditing-and-how-it-will-i.asp/6308/
# [05:15] <AryehGregor> Is anyone implementing ping?
# [05:16] <Hixie> mozilla has an implementation which they disabled while the spec was unstable
# [05:16] <Hixie> i expect other browsers will implement it in due course
# [05:16] * Joins: benward (n=benward@98.210.154.133)
# [05:17] * Quits: benward (n=benward@98.210.154.133) (Remote closed the connection)
# [05:18] <othermaciej> someone expressed interest in implementing it for WebKit, citing content publishers who want to use it, and I asked them to get those publishers to say so on the record
# [05:19] <AryehGregor> Could it realistically be useful in the near future? Does it support any kind of useful fallback?
# [05:20] * AryehGregor hasn't gotten up to that part of the spec, apparently, or perhaps has just forgotten
# [05:20] <othermaciej> it could be useful it you don't care about losing link tracking in browsers that don't support it
# [05:20] <othermaciej> it could be useful if you UA-sniff and serve different content to browsers that support it
# [05:24] <othermaciej> I think you can use the "ping" IDL attribute to detect from script if ping is supported
# [05:25] * Joins: primal1 (n=primal1@pool-98-112-164-140.lsanca.fios.verizon.net)
# [05:27] * othermaciej mulls over the SHOULD-level UI requirement
# [05:28] <othermaciej> I'm pretty sure we would not add UI for hyperlink auditing to MobileSafari
# [05:29] <othermaciej> not because I've asked but because we have so very few prefs for it in general
# [05:30] <othermaciej> on desktop Safari we'd need good wording
# [05:30] <othermaciej> I'm not sure "Enable hyperlink auditing" would be comprehensible
# [05:31] <othermaciej> and "Don't let sites track your browsing" would be a lie
# [05:32] <Hixie> my suggestion has been as follows:
# [05:33] * Joins: cardona507 (n=cardona5@c-67-180-160-250.hsd1.ca.comcast.net)
# [05:33] <Hixie> have a single trackbar control that goes from "maximum compatibility" to "maximum privacy"
# [05:33] <Hixie> with maybe three notches
# [05:33] <Hixie> the first enables third-party cookies, etc
# [05:33] <Hixie> the third disabled ping, all cookies, etc
# [05:34] <Hixie> the middle one is the default
# [05:34] <othermaciej> our only big switch toggle is Private Browsing mode
# [05:34] <othermaciej> we would definitely disable ping in that mode, though I'm not sure that satisfies the intent of the requirement
# [05:34] <Hixie> you already have a "third party cookies" pref
# [05:35] <Hixie> i'd remove that and other privacy-related controls and simplify the whole thing to the aforementioned switch
# [05:35] <Hixie> but then, i'm no ui expert
# [05:35] <othermaciej> I think the UI issue here is that normal people don't care about these settings and never touch them, but privacy nuts want to know exactly what is happening
# [05:37] <Hixie> sure, but for such small demographics you'd have extensions to expose the details
# [05:37] <Hixie> i'm just talking about the main UI for all users
# [05:37] <othermaciej> we don't currently have a way in the UI to block Referer or non-cookie forms of state other than Private Browsing
# [05:38] * Quits: tantek (n=tantek@adsl-69-107-12-74.dsl.pltn13.pacbell.net)
# [05:38] <othermaciej> Safari doesn't currently have extensions (other than unsupported hacks)
# [05:39] <othermaciej> I guess based on that, the logical conclusion is that Private Browsing should be the UI to turn off ping, but I suspect some people will want to turn it off without crippling their browsing experience quite that much
# [05:41] <othermaciej> I will ask our UI people what they think
# [05:41] * othermaciej is unclear on the other issue
# [05:42] <Hixie> other issue?
# [05:42] <othermaciej> I'm not sure why using POST for hyperlink auditing is any worse than the possibility of POSTing a form with .submit(), or using an XMLHttpRequest with POST
# [05:43] <Hixie> it's not
# [05:43] <Hixie> using POST is the right thing to do
# [05:43] <othermaciej> I'm looking at the issue tracker issues about hyperlink auditing
# [05:43] <othermaciej> I don't actually understand that one
# [05:43] <othermaciej> ping has to be a non-idempotent method to work
# [05:45] * Joins: tantek (n=tantek@adsl-69-107-12-74.dsl.pltn13.pacbell.net)
# [05:47] * Quits: tantek (n=tantek@adsl-69-107-12-74.dsl.pltn13.pacbell.net) (Client Quit)
# [06:18] * Joins: ifette (n=ifette@220.109.219.244)
# [06:19] * Joins: benward (n=benward@98.210.154.133)
# [06:20] <ifette> othermaciej: I am not sure that there is a clear mapping between privacy and ping
# [06:21] <ifette> for instance, if I'm concerned ping isn't working, i can just force a user through a redirect
# [06:21] <othermaciej> ifette: privacy as a general concept, or privacy as in Private Browsing mode?
# [06:21] <ifette> (and cause the user to incur latency)
# [06:21] <ifette> othermaciej: the latter
# [06:22] <othermaciej> presumably Private Browsing mode would not reveal to the server up front that <a ping> won't do anything
# [06:22] <othermaciej> (likewise for any other way of turning it off)
# [06:22] <ifette> presumably you can figure out that the user is in private browsing mode based on some of the other discussions around local storage and that API being disabled, and so can fall back to redirects :)
# [06:23] <ifette> I just don't get all the concern about a ping. Great, so you know the user clicked a link... okay. Any messaging to users is going to be deceptive for the majority of them. I can't just have a checkbox that says "disable <a ping>", but if I try to say "disable link tracking"
# [06:23] <ifette> that would just be dishonest given that there's so many other ways to do it
# [06:24] <ifette> Frankly, I would much rather say "look, websites can figure out if you clicked on a link. It's been that way for a long time. Live with it. <a ping> lets life move on as it has, but can make your browsing faster
# [06:24] <othermaciej> that would be an argument to have no UI at all to turn it off
# [06:24] <ifette> ... by not having to deal with redirects etc"
# [06:24] <ifette> othermaciej: excellent idea!
# [06:24] <Hixie> ifette: you can detect if the browsers supports ping
# [06:25] <jorlow> Hixie: doesn't that support ifette's argument?
# [06:25] <Hixie> jorlow: how so?
# [06:26] <ifette> What's the real use case here? What's the concern? If it's "evil google will know all the links I'm clicking", that information is already known, and furthermore if I want to be evil, I can say "If I never got an a ping response with this cookie, then for this cookie don't use a ping and use redirects instead"
# [06:26] <ifette> this is valuable information to webmasters
# [06:26] <jorlow> if aping isn't supported (as it wouldn't be in private browsing) then the page could just fall back to other methods
# [06:26] <ifette> they will get it one way or another
# [06:26] <othermaciej> ifette, jorlow: if you guys think there should be no UI requirement for ping, not even a SHOULD-level requirement, then you should make that case to Hixie
# [06:26] <ifette> we might as well make it not suck for users
# [06:26] <ifette> I will happily do so, if I can find the right thread. I will admit to being swamped via email.
# [06:27] <Hixie> the two main reasons we (google) want ping="" is (a) allowing us to honour user privacy desires without them having to use a cookie to opt-out, and (b) latency improvements from not redirecting.
# [06:27] <othermaciej> there is no current email thread
# [06:27] <Hixie> so if the browser supports ping, we wouldn't route around it being disabled
# [06:27] <othermaciej> ifette: I'm just saying that if you think there should be no browser UI to turn it off, then you shouldn't argue with me, because I didn't make or ask for the spec requirement to have it
# [06:28] <ifette> We wouldn't route around it, others might
# [06:28] <jorlow> hm....i guess it's true that aping makes it easier for sites to "do the right thing" WRT privacy
# [06:28] <jorlow> I was only looking at it from the "evil" side of things
# [06:28] <Hixie> i'm sure there are sites who would indeed route around it
# [06:28] <Hixie> but hey, that's their problem
# [06:28] <ifette> othermaciej: wasn't trying to argue with you, was simply stating that tying it to private browsing might not make sense
# [06:28] <othermaciej> it's true that UI to turn it off won't help you against a malicious site
# [06:29] <othermaciej> ifette: if there was UI for it at all, then I think tying it to private browsing would make sense, because private browsing tries to turn off everything
# [06:29] <jorlow> in that case, i suppose linking it with private browsing might make sense
# [06:29] <jorlow> at least in safari
# [06:30] <ifette> It seems to me that the following are probable: 1) if a ping isn't reliable / everyone turns it off because it's something that the media gets hyped up about like cookies, no one will use it and just use redirects instead 2) any option presented to users would be unintelligable and over promise or seem like a no-brainer to check c) none of this will matter to "evil" sites
# [06:30] <othermaciej> under private browsing, a malicious site could tell what links you follow off their page by other means, but could not tie it back to you due to lack of cookies
# [06:30] <othermaciej> though to be fair, we don't yet have a way to turn off your IP address
# [06:30] <othermaciej> well, cookies can be turned off, but in practice very few people do
# [06:30] <ifette> othermaciej: i view private browsing as "don't leave a trace of what I do on my computer", not "please use TOR and make me as anonymous as possible"
# [06:30] <Hixie> if it's "something that the media gets hyped up about like cookies", then people won't turn it off.
# [06:31] * Joins: cying (n=cying@adsl-75-18-225-53.dsl.pltn13.sbcglobal.net)
# [06:31] <ifette> othermaciej: if you say "private browsing mode turns off <a ping>", what do you say to a real user?
# [06:31] <othermaciej> most sites consider cookies reliable enough for their needs, even though there are a small minority of users who turn them off entirely
# [06:31] <othermaciej> a few sites use Flash cookies instead, though you can turn off plugins to get around those
# [06:31] <Hixie> i think if the preference is a serious one like i proposed earlier, and not just one that disabled just this feature, that it won't be turned off just for the sake of it, but only by people who actually want their privacy protected
# [06:31] <othermaciej> ifette: our private browsing explanation dialog is hardly a paragon of clarity as is
# [06:32] <ifette> othermaciej: Yes, people still use cookies, not the best example. Was trying to say that it's something real users don't well understand, and that in general when you mention cookies people think "bad" without understanding
# [06:32] <ifette> othermaciej: :)
# [06:32] <othermaciej> >95% of people will never change any privacy-related preferences
# [06:33] <ifette> Hixie: wtf. "Want their privacy protected" -- if you say that to users, what user is going to say "No, I don't want my privacy protected, let's leave this unchecked" :)
# [06:33] <othermaciej> on the other hand, lack of ping would probably break the web much less visibly than lack of cookies or lack of plugins
# [06:33] <ifette> I guess my biggest beef is that I don't want people permanently using "private mode" because tat's not a good user experience
# [06:33] <ifette> so I resist tying the option to that
# [06:33] <Hixie> ifette: hence my proposal, which isn't such a checkbox. see above.
# [06:34] <othermaciej> I don't think anyone permanently uses Private Browsing mode, or would
# [06:34] <othermaciej> but I don't think a separate setting for <a ping> is worth its weight in pixels
# [06:34] <jorlow> btw, where can i find a copy of the latest 'a ping' proposal?
# [06:34] <othermaciej> and I don't think Hixie's suggested tri-state setting would be very understandable
# [06:34] <othermaciej> jorlow: it's in the HTML5 spec
# [06:34] <Hixie> jorlow: http://whatwg.org/html5
# [06:34] <othermaciej> look for "hyperlink auditing"
# [06:34] <ifette> Correct -- but then there comes a question, if it's only tied to private browsing mode, is that really a good option?
# [06:35] <ifette> if i'm a user freaked out by a ping
# [06:35] <othermaciej> private browsing mode does lots of things that we don't offer as normal user settings
# [06:35] <jorlow> ahh...thanks http://www.whatwg.org/specs/web-apps/current-work/multipage/history.html#hyperlink-auditing
# [06:35] <ifette> sigh, sure
# [06:35] <Hixie> if you're a user freaked out by ping, then you get an extension that gives you fine-grained control
# [06:35] <othermaciej> if users really freaked out about it, we would probably add a separate setting
# [06:35] <Hixie> most users aren't going to have a clue or care
# [06:36] <othermaciej> if too many people used the setting, then it would indeed undermine the value of the feature
# [06:36] <ifette> the sad part is that tying it to private browsing mode might have an opposite effect
# [06:36] <ifette> let's say people use private browisng mode for porn
# [06:36] <othermaciej> but if that many people are that freaked out about it, then omitting the setting would just make browsers with <a ping> lose market share
# [06:36] <ifette> that would imply that porn sites would not get good results from a ping
# [06:36] <ifette> and therefore would still use redirects
# [06:36] <ifette> and so the effect of turning it off in private browsing mode would be minimal
# [06:37] <jorlow> well, users shouldn't be freaked out....because nothing is worse than the state of things today
# [06:37] <othermaciej> they shouldn't be, but the concern about having a preference is based on the theory that a lot of them will
# [06:37] <jorlow> but of course, that doesn't necessarily mean the media won't write a story to the contrary
# [06:37] <ifette> I think my preference is really to say "Look, sites can know what link you click. That's the way it is today, we're not changing anything except reducing some latency, live with it. If you want more controls over your data, let's have a more comprehensive discussion and figure soething out"
# [06:38] <othermaciej> my expectation: only a small minority of users will care enough to take action
# [06:38] <jorlow> Well, maybe the preference can be a check box with a somewhat cryptic name
# [06:38] <othermaciej> failing to give those people a means to take action without switching browsers will generate a huge amount of negative PR
# [06:38] <jorlow> i mean, most users don't even know what a cookie is...even today
# [06:38] <ifette> jorlow: try getting that past ben ;-)
# [06:38] <ifette> othermaciej: Yes, except here "take action" is kinda lame
# [06:38] <jorlow> 'disable hyperlink auditing'....if users really care, they'll know what it is
# [06:39] <jorlow> and users won't turn it off in a misguided attempt at 'privacy'
# [06:39] <othermaciej> ifette: even if it's a UI placebo, failing to have it will probably make a vocal minority very upset
# [06:39] <ifette> othermaciej: wouldn't be the first time
# [06:39] <jorlow> that's probably true........
# [06:39] <ifette> othermaciej: we took out the options to enable sslv2
# [06:39] <ifette> othermaciej: and all sorts of other crap options
# [06:39] <ifette> othermaciej: e.g. we don't let users disable javascript
# [06:40] <ifette> othermaciej: but we're supposed to let users disable <a ping> ?
# [06:40] <Hixie> just tie it to the same as the third party cookie pref, like i suggested
# [06:40] <Hixie> along with the other privacy related features
# [06:40] <ifette> Hixie: Where is your suggestion
# [06:41] <ifette> Hixie: Want to make sure I don't misinterpret it. I really don't like tying it to third party cookies if I understand you correctly, as they have nothing to do with each other
# [06:41] <Hixie> that way people who really do value privacy over compatibility get what they want, and the others don't change the pref unknowingly
# [06:41] <othermaciej> ifette: Safari has very few prefs, and I wouldn't be keen on adding one for this
# [06:41] <ifette> Hixie: I may not want some advertiser to know what I'm doing on Google, but I'm fine with Google knowing what I'm searching for on Google
# [06:41] <Hixie> http://krijnhoetmer.nl/irc-logs/whatwg/20090908#l-134
# [06:41] <Hixie> ifette: ^
# [06:41] <ifette> ty
# [06:41] <othermaciej> Hixie: in the Safari UI, we don't like to face people with an arbitrary tradeoff with unknowable consequences
# [06:41] <othermaciej> like, we'd never have a slider with "speed" on one end and "security" on the other
# [06:42] <ifette> Hixie: eep
# [06:42] <Hixie> othermaciej: you already have "Accept cookies: ( ) Always ( ) Never ( ) Only sites I visit"
# [06:42] <ifette> +1 maciej
# [06:42] <Hixie> othermaciej: which is pretty much as arbitrary
# [06:43] <othermaciej> sure, but we present that in terms of concrete result, not "more compatible" vs "more private"
# [06:43] <othermaciej> likewise we don't present "Enable Java" as "Make some sites work, but damn will they be slow and there's probably horrible security consequences"
# [06:43] <ifette> Hixie: If I don't let users turn off something as major as JavaScript, I'm sure as hell not letting them turn off some random thing like <a ping>. Down that road lies the path to hell. I am also not tying it to Incognito, as I think again it has little to do with it and won't actually solve any problem
# [06:44] <othermaciej> or "Enable plug-ins" as "Support casual games and annoying ads"
# [06:44] <ifette> Hixie: I'm happy to send an email, but I would not support anything beyond a MAY for the option
# [06:44] * Joins: SuperDot_iPod (n=superdot@adsl-75-61-83-64.dsl.pltn13.sbcglobal.net)
# [06:45] <ifette> (and frankly saying that you MAY have a checkbox option is really not saying much)
# [06:45] <Hixie> othermaciej: ok, then label the three notches as what they disabled or enable :-)
# [06:46] <ifette> Frankly, unless you're talking a per-site basis, I don't know that options about Flash or Java make sense
# [06:46] <Hixie> ifette: it would be politically inconvenient for me to change the "should" to a "may", but i don't think it matters -- "should" just means you have to have a good reason, and "We want our UA to be usable" is a good reason
# [06:46] <othermaciej> I browse with Java turned off all the time
# [06:47] <ifette> othermaciej: Java yes, I concede you that. Flash, not so much.
# [06:47] <othermaciej> well, not having Flash is livable on my iPhone, probably not so much on a desktop browser
# [06:47] <ifette> othermaciej: stop making good points :P
# [06:48] <roc> we need a stalking horse browser to ship <a ping> and take all the flak
# [06:48] <othermaciej> but we do get frequent user requests to augment the general "Enable plug-ins" setting with a separate "Enable Flash", so they can turn off just Flash
# [06:49] * Quits: cardona507 (n=cardona5@c-67-180-160-250.hsd1.ca.comcast.net)
# [06:49] <othermaciej> roc: I think ifette just volunteered Chrome - presumably with the catchy slogan, "What's the concern? Evil google will know all the links I'm clicking?"
# [06:49] <roc> works for me
# [06:49] <ifette> othermaciej: While I might be fine with that string, I'm pretty sure legal would shoot it down :(
# [06:49] <Hixie> on a completely different front, i'm working through some microdata feedback
# [06:49] <Hixie> and i have to say
# [06:50] <Hixie> it's really
# [06:50] <Hixie> really
# [06:50] <Hixie> really hard to do a good job when you care so little about the technology
# [06:50] <ifette> Can we add some more markup?
# [06:50] <ifette> <main> or something?
# [06:50] <ifette> ;-)
# [06:50] <ifette> and plese, let's have some more OWL stuff in there
# [06:50] <ifette> (sorry, I'm re-living some nightmares, waiting for @jorlow to slap me awake)
# [06:53] <jorlow> ifette: what are you talking about? it's 1:47pm...you're still sleeping?
# [06:53] * Quits: dglazkov (n=dglazkov@c-67-188-0-62.hsd1.ca.comcast.net)
# [06:53] <othermaciej> I think we should upgrade Web Ontology Language to Ontology Representation Language for You
# [06:54] <othermaciej> because I'm pretty sure there's no better OWL than ORLY
# [07:01] * Joins: cardona507 (n=cardona5@c-67-180-160-250.hsd1.ca.comcast.net)
# [07:02] * Quits: cardona507 (n=cardona5@c-67-180-160-250.hsd1.ca.comcast.net) (Client Quit)
# [07:10] * Quits: roc (n=roc@203-97-204-82.dsl.clear.net.nz)
# [07:17] * Quits: mpilgrim (n=mpilgrim@rrcs-96-10-240-189.midsouth.biz.rr.com) (Read error: 104 (Connection reset by peer))
# [07:18] * Joins: mpilgrim (n=mpilgrim@rrcs-96-10-240-189.midsouth.biz.rr.com)
# [07:20] * Quits: vs-hs_ (i=david@shell.etttretresju.net) (Remote closed the connection)
# [07:20] * Joins: vs-hs (i=david@shell.etttretresju.net)
# [07:20] * Quits: epeus (n=KevinMar@c-67-164-14-96.hsd1.ca.comcast.net)
# [07:23] <MikeSmith> othermaciej: you feeling any better than you were at the end of last week?
# [07:24] <othermaciej> MikeSmith: I'm down from "violently severely sick" to "really sick"
# [07:24] <MikeSmith> geez
# [07:24] <MikeSmith> any idea what it is? virus?
# [07:24] <othermaciej> probably
# [07:25] <othermaciej> some sort of flu I think, hopefully not one of the animal flavored varieties
# [07:25] * Quits: broquaint (i=5f566a0b@spc2-brig11-0-0-cust222.asfd.cable.virginmedia.com) (Remote closed the connection)
# [07:25] <othermaciej> I don't have a fever any more but I am still really congested, throat is sore, stomach somewhat upset, etc
# [07:26] <othermaciej> Saturday was really terrible though, I hadn't realized how excruciating it is to fly with a bad cold/flu - what with the air pressure changes
# [07:26] <MikeSmith> yeah
# [07:29] <jorlow> eww...yeah :-(
# [07:29] <jorlow> the only time i've ever gone to the doctor for flu like symptoms is when i had to fly
# [07:30] <jorlow> and i was sure glad i did
# [07:32] * Quits: GPHemsley (n=GPHemsle@pdpc/supporter/student/GPHemsley) ("This computer has gone to sleep")
# [07:33] <othermaciej> well, this was after getting hopped up on all the non-prescription meds I could get in the middle of the night in Cambridge
# [07:33] <othermaciej> (thank goodness for that 7-Eleven in Tech Square)
# [07:34] <MikeSmith> yeah, that place seems to be about the only thing close by that's open late
# [07:36] <othermaciej> I ended up going there around 2 AM in hopes of cold medicine and fluids
# [07:36] <othermaciej> there were a bunch of MIT kids there
# [07:36] <othermaciej> reminds me of my younger days
# [07:38] <othermaciej> I'm glad such archectypes as Purple Haired Girl, Guy with Nerdy Slogan T-Shirt, Asian Guy in Cargo Shorts and Hippie Girl in Long Dress are still to be found at the ol' Institute
# [07:38] <MikeSmith> heh
# [07:44] * Joins: roc (n=roc@121-72-191-237.dsl.telstraclear.net)
# [07:44] * Joins: KevinMarks (n=KevinMar@c-67-164-14-96.hsd1.ca.comcast.net)
# [07:46] * Quits: stevepsu (i=sjstrutt@204.15.224.234) (Read error: 145 (Connection timed out))
# [07:55] <Hixie> man, the bottom part of movie posters are a huge pain in the asss to typeset
# [07:55] <Hixie> i wonder what tools they use to do it
# [07:56] <othermaciej> you mean where they have all the tiny little mandatory credits?
# [07:56] * Joins: jacobolus (n=jacobolu@dhcp-0059871802-99-6d.client.student.harvard.edu)
# [07:57] <Hixie> yeah
# [07:57] <othermaciej> what is so hard about typesetting it?
# [07:57] <Hixie> the "EDITED BY" bit
# [07:57] <Hixie> check the bottom of http://www.comingsoon.net/imageGallery/Ninja_Assassin/large/hr_Ninja_Assassin_poster.jpg out
# [07:57] <Hixie> look for "PRODUCTION DESIGNER"
# [07:57] <Hixie> notice the following:
# [07:58] <Hixie> 1. the gap between the start of the longest of those two words to the word before it is consistent across all such cases
# [07:58] <Hixie> 2. the top of PRODUCTION lines up with the top of the line, the bottom of DESIGNER with the bottom of the line
# [07:58] <Hixie> 3. the line height of that two-line thingy is proportional to the line height of the main text
# [07:59] <othermaciej> I presume this is all down to careful selection of font sizes
# [07:59] <hsivonen> Hixie: are you trying to replicate the layout using CSS?
# [07:59] <Hixie> hsivonen: actually i was trying to do it using pages, but it would also be quite hard to get right in CSS
# [08:00] <Hixie> i actually can't find a non-hacky way to do it in Pages
# [08:00] <othermaciej> presumably in a layout tool with guide lines this is all pretty simple
# [08:00] <Hixie> othermaciej: well you can't just overlap text, since if you then edit any one part of it, you'd have to manually retypeset the rest
# [08:00] <Hixie> othermaciej: so it has to be something that supports nested two-line text
# [08:00] <Hixie> which Pages doesn't
# [08:01] <Hixie> in CSS you can do it using display:inline-block, but then getting the font sizes right is impossible
# [08:01] <Hixie> since you don't know your font metrics ahead of time
# [08:01] <othermaciej> I guess you'd want something like FrameMaker
# [08:01] <Hixie> i guess
# [08:01] <othermaciej> I'm not overly familiar with Pages
# [08:02] <othermaciej> but yeah, you're right, this is a typographical oddity that you rarely see anywhere but movie posters
# [08:11] <MikeSmith> hsivonen: if/when you have some time, I wanted to ask you if you think it would be worthwhile to add ability to the validator.nu for emitting warnings to the datatype-checking code
# [08:11] <MikeSmith> hsivonen: as opposed to just errors
# [08:12] <hsivonen> MikeSmith: it could well be worthwhile. the hard question is how to do it
# [08:12] <MikeSmith> yeah
# [08:12] <hsivonen> the datatypes are not supposed to have side effects
# [08:12] <MikeSmith> ah
# [08:13] <hsivonen> and in principle, it's ok for Jing to call into a datatype, get an error, call into another datatype, not get an error and prefer the latter and not show the first error
# [08:13] * Joins: Mrmil (n=ut_ollie@host-77-236-204-8.blue4.cz)
# [08:14] <hsivonen> also, in principle, the Java interface stuff should be such that the datatype jar can be used in a vanilla Jing
# [08:15] <MikeSmith> hsivonen: OK, I begin to see now
# [08:15] <hsivonen> the whole thing has been designed for a boolean notion of validity without any shades in between
# [08:15] <MikeSmith> that's the real problem
# [08:15] <virtuelv> I'd have thought that the bottom of those movie posters is laid out completely in Photoshop or Illustrator
# [08:15] <virtuelv> in other words, manually, with a lot of effort and pain
# [08:17] <MikeSmith> hsivonen: I have come to think that think that boolean notion of validity is a serious flaw of grammar-based schema checking
# [08:17] <hsivonen> MikeSmith: maybe filing a bug against upstream Jing could elicit some insights
# [08:17] <MikeSmith> OK, I can try that
# [08:18] <hsivonen> MikeSmith: I agree with your observation about grammars
# [08:18] <hsivonen> Hixie: OmniGraffle may work better for this use case than Pages
# [08:19] <Hixie> possibly
# [08:20] <MikeSmith> hsivonen: the specific problem that led me to ask about this is the bug about language tags that AryehGregor reported a couple of days ago
# [08:20] * Joins: Kalms (n=rasmuska@81.161.185.108)
# [08:20] <MikeSmith> http://bugzilla.validator.nu/show_bug.cgi?id=640
# [08:20] <ttepasse> I would expect handcrafted SVG in #whatwg.
# [08:20] <Hixie> you'd expect SVG in #whatwg? :-)
# [08:20] <hsivonen> ttepasse: Pages is already shiny Cocoa payware
# [08:20] <ttepasse> s/SVG/canvas 2d/ ;)
# [08:21] <hsivonen> MikeSmith: given the constraints, I suggest making stuff that should be conforming not emit any messages for now even if it's obsolete/deprecated
# [08:21] <hsivonen> which reminds me that I should redeploy V.nu
# [08:22] <ifette> Any chance you can use Ruby, with appropriate CSS to get the sizing right?
# [08:22] * Quits: doublec (n=doublec@203-97-204-82.dsl.clear.net.nz) ("Leaving")
# [08:22] <ifette> (re: "Production Designer")
# [08:22] <ifette> (yes, it's a bit of a hack, but it's the only thing I can think of that might let you get two lines on one line)
# [08:23] <heycam> display:inline-box would too
# [08:25] <MikeSmith> hsivonen: not emitting messages for those cases seems like declaring defeat to me. I would personally really rather not do that unless you think there's no practical way to work around it. even if it ends up being kind of a messy way
# [08:26] <hsivonen> MikeSmith: ok
# [08:27] <Hixie> ifette, heycam: the problem with CSS is getting the font metrics right; the layout itself is easy to achieve
# [08:28] <MikeSmith> hsivonen: I'm willing to spend some time on it this week at least. If we can't some workaround figured out, then I guess can comment out the reporting about the deprecated stuff
# [08:28] <MikeSmith> hsivonen: so here's a stupid idea: we include some magic substring in the NewDatatypeException message for warning-level cases
# [08:29] <MikeSmith> and that substring would show up in messages if anybody used the datatypes outside of the v.nu system
# [08:29] * Joins: pesla (n=retep@procurios.xs4all.nl)
# [08:29] <MikeSmith> but the error-reporting backend could check for that magic substring and then make those messages warnings instead of errors
# [08:31] <MikeSmith> I don't know how do-able that is in the current backend. or whether you think it'd even be a good idea to do it at all
# [08:32] <hsivonen> MikeSmith: It's easy to put a flag on the exception so that the RELAX NG semantics of the result are "invalid" but the V.nu front end downgrades the message type (and the validity outcome)
# [08:32] <hsivonen> MikeSmith: what's hard is making the RELAX NG semantics be that the result is valid
# [08:32] <hsivonen> MikeSmith: I guess this is a "theoretical purity" kind of point
# [08:33] <hsivonen> MikeSmith: I think you can't detect it with the current schema
# [08:33] <hsivonen> MikeSmith: but in principle, it changes the generic reuse characteristics of the datatype lib
# [08:33] <hsivonen> MikeSmith: maybe in this case, it's anyway the best solution to make the exception have a flag that tells V.nu to downgrade the message
# [08:35] * Quits: cying (n=cying@adsl-75-18-225-53.dsl.pltn13.sbcglobal.net)
# [08:36] <hsivonen> It seems like there are no exceptions to the rule that non-ASCII supersets are security holes: http://mail.apps.ietf.org/ietf/charsets/msg01846.html
# [08:37] <othermaciej> I don't think browsers can stop accepting UTF-16 though
# [08:43] * Quits: eighty4 (n=eighty4@eighty4.se) (Remote closed the connection)
# [08:44] * Joins: eighty4 (n=eighty4@eighty4.se)
# [08:45] * Quits: virtuelv (n=virtuelv@125.175.251.212.customer.cdi.no) (Read error: 110 (Connection timed out))
# [08:46] * Quits: takkaria (n=takkaria@isparp.co.uk) (Remote closed the connection)
# [08:46] * Joins: takkaria (n=takkaria@isparp.co.uk)
# [08:47] <MikeSmith> hsivonen: OK, I see what you're saying about the problem still being to have RELAX NG semantic be that it's valid
# [08:47] <MikeSmith> hsivonen: so instead, what about just having the datatype library not throw the exception at all if its used outside of validator.nu?
# [08:48] <MikeSmith> I don't know how doable that might be
# [08:48] <hsivonen> MikeSmith: that's a possibility. (you could still detect the implementation by feeding V.nu a contrived custom schema, but I guess that doesn't matter in practice)
# [08:49] <hsivonen> MikeSmith: You could have a Java system property for toggling this behavior
# [08:50] <MikeSmith> OK. seems like that route is probably the best to try
# [08:50] <MikeSmith> unless we can manage to come up with something better in the mean time, I'll work on writing up a patch for that
# [08:50] <MikeSmith> some time this week
# [08:51] <hsivonen> I've repointed the validator.nu and html5.validator.nu hostnames on the DNS level
# [08:51] <hsivonen> (so that I can do a dist upgrade on the other server)
# [08:51] <hsivonen> I hope nothing breaks...
# [08:52] <hsivonen> MikeSmith: I *think* I have deployed your recent changes
# [08:52] * Quits: yutak (n=yutak@220.109.219.244) ("Leaving")
# [08:54] <MikeSmith> hsivonen: thanks. fwiw, I've been running from my workspace (with my commits) with no problems. So I don't think those will break anything.
# [08:54] <MikeSmith> I didn't sync up yet to Sam's change, though
# [08:54] * Joins: yutak (n=yutak@220.109.219.244)
# [08:56] * Quits: SuperDot_iPod (n=superdot@adsl-75-61-83-64.dsl.pltn13.sbcglobal.net) ("Get Colloquy for iPhone! http://mobile.colloquy.info")
# [08:57] * Joins: SuperDot_iPod (n=superdot@adsl-75-61-83-64.dsl.pltn13.sbcglobal.net)
# [09:06] * Joins: Maurice (n=ano@80.101.46.164)
# [09:06] * Quits: heycam (n=cam@clm-laptop.infotech.monash.edu.au) ("bye")
# [09:17] * Joins: aboodman (n=aboodman@72.14.229.81)
# [09:21] * Joins: primal1_ (n=primal1@pool-98-112-164-140.lsanca.fios.verizon.net)
# [09:21] * Quits: primal1 (n=primal1@pool-98-112-164-140.lsanca.fios.verizon.net) (Read error: 104 (Connection reset by peer))
# [09:21] * primal1_ is now known as primal1
# [09:23] <Hixie> wow, running the microwave totally killed my wifi network
# [09:23] <Hixie> i wonder if the other times i have dropouts of the network is because of neighbours running microwaves
# [09:25] * Quits: benward (n=benward@98.210.154.133) ("Sleep")
# [09:25] * Quits: lazni (n=lazni@118.71.169.21) (Read error: 110 (Connection timed out))
# [09:26] * Joins: Creap (n=Creap@vemod.brg.sgsnet.se)
# [09:31] * Joins: lazni (n=lazni@118.71.169.21)
# [09:33] * Joins: broquaint (i=4776acbc@spc2-brig11-0-0-cust222.asfd.cable.virginmedia.com)
# [09:36] * Joins: heycam (n=cam@210-84-56-211.dyn.iinet.net.au)
# [09:41] * Joins: Hish (n=chatzill@mail2.n-e-s.de)
# [09:43] * Quits: SuperDot_iPod (n=superdot@adsl-75-61-83-64.dsl.pltn13.sbcglobal.net) (Remote closed the connection)
# [09:43] * Joins: SuperDot_iPod (n=superdot@adsl-75-61-83-64.dsl.pltn13.sbcglobal.net)
# [09:46] * Joins: erlehmann (n=erlehman@tmo-104-178.customers.d1-online.com)
# [09:51] * Joins: virtuelv (n=virtuelv@pat-tdc.opera.com)
# [09:54] * Quits: Hish (n=chatzill@mail2.n-e-s.de) (Read error: 104 (Connection reset by peer))
# [09:54] * Quits: Kalms (n=rasmuska@81.161.185.108)
# [09:58] * Quits: jwalden (n=waldo@c-98-248-40-206.hsd1.ca.comcast.net) ("back in the morning")
# [10:03] * Quits: lazni (n=lazni@118.71.169.21) ("Leaving.")
# [10:07] * Joins: foolip (n=philip@pat.se.opera.com)
# [10:07] * Joins: lazni (n=lazni@118.71.169.21)
# [10:08] * Quits: SuperDot_iPod (n=superdot@adsl-75-61-83-64.dsl.pltn13.sbcglobal.net) ("Get Colloquy for iPhone! http://mobile.colloquy.info")
# [10:08] * Joins: SuperDot_iPod (n=superdot@adsl-75-61-83-64.dsl.pltn13.sbcglobal.net)
# [10:08] * Joins: maikmerten (n=maikmert@BAE3af2.bae.pppool.de)
# [10:13] * Joins: mpt (n=mpt@canonical/launchpad/mpt)
# [10:14] * Joins: benward (n=benward@98.210.154.133)
# [10:17] <Hixie> gah
# [10:17] <Hixie> so hard to write an intro section for microdata
# [10:18] <Hixie> complete this sentence: "Microdata allows you to layer lists of name-value pairs on top of Web pages, so that"
# [10:18] <Hixie> ...in a way that makes Web authors think "ooh, cool"
# [10:18] <jgraham> unicorns can sprinkle magic fairy dust everywhere, causing world peace and happiness
# [10:19] <Hixie> it also has to be accurate.
# [10:19] <jgraham> Oh? I assumed that was accurate given the intensity of the feeling that the topic has generated
# [10:20] <Hixie> i believe a lot of people believe that it is accurate, yes
# [10:20] <Hixie> but sadly it is not
# [10:20] <jgraham> AH
# [10:20] <jgraham> s/AH/Ah/
# [10:20] <Hixie> here's what i came up with:
# [10:20] <Hixie> Microdata allows you to layer lists of name-value pairs on top of Web
# [10:20] <Hixie> pages, so that JavaScript scripts, search engines, and other tools can
# [10:20] <Hixie> extract that data from the page directly instead of you having to
# [10:20] <Hixie> provide this data in a separate file.
# [10:21] <benward> "Microdata is a mechanism apply structured-data objects within web pages…" ?
# [10:21] <jgraham> your "this" doesn't match your "that"
# [10:21] <benward> ‘to apply’, even
# [10:21] <Hixie> benward: the "what" is easy, it's the "so that..." part that's hard
# [10:21] <Hixie> jgraham: did s/this data/it/
# [10:22] <othermaciej> Hixie: you may want to say something about "reliably" extracting the data from the page to contrast with more hackish sorts of "screen scraping" extraction
# [10:23] <Lachy> just be honest and say "Microdata allows you to add additional, mostly useless metadata in a web page so that you can feel good about contributing to growth of the Semantic Web"
# [10:23] <boblet> “Microdata allows you to easily add machine-readable arbitrary data as name-value pairs to web pages”…?
# [10:23] <Hixie> boblet: that's still just the "what"
# [10:24] <othermaciej> Lachy: see, this is why you're not in marketing
# [10:24] <hsivonen> Hixie: indeed, it's hard to write accurate marketing intros in specs. Doing what the XHTML2 WG's specs have is easier.
# [10:24] <Hixie> Lachy: that won't provide the users in this usability study with adequate incentive for us to get useful results
# [10:24] <Hixie> hsivonen: it's not for a spec, it's for the intro we give to the usability study participants
# [10:24] <Hixie> othermaciej: now i have:
# [10:24] <Hixie> Microdata allows you to layer lists of name-value pairs on top of Web
# [10:24] <Hixie> pages, so that JavaScript scripts, search engines, and other tools can
# [10:24] <Hixie> extract that data from the page directly, instead of you having to
# [10:24] <Hixie> provide it in a separate file, or them having to use screen scraping.
# [10:24] <hsivonen> Hixie: ah
# [10:25] <othermaciej> Hixie: that's pretty good -- adding a concrete example might help too
# [10:25] <Hixie> oh they're getting a whole bunch of those
# [10:25] <Lachy> oh, I thought it was for the spec.
# [10:25] <othermaciej> all righty then
# [10:25] <Hixie> othermaciej: http://damowmow.com/playground/microdata/001/
# [10:26] * jgraham hopes the participants in this study don't read the logs
# [10:27] <boblet> …“This data can then be easily accessed by scripts, search engines and other tools, giving authors a simple yet open-ended way to extend HTML”
# [10:27] <Hixie> jgraham: if they do, they're far too close to the project :-)
# [10:27] <Lachy> Hixie, http://damowmow.com/playground/microdata/001/blog.html seems to be lacking the microdata
# [10:27] <jorlow> is there any easy way for me to know what browsers support features that have been pulled out of the html5 spec?
# [10:28] <jorlow> for example workers
# [10:28] <Hixie> boblet: closer, but it fails at the end there where it turns into blue sky nonsense :-)
# [10:28] * boblet wonders if he really understands Microdata enough to be contributing :)
# [10:28] <boblet> har
# [10:28] <Hixie> boblet: don't worry, i don't understand it enough either
# [10:28] <Hixie> Lachy: -annotated files have it
# [10:28] <Lachy> oh
# [10:29] <othermaciej> jorlow: asking, or testing
# [10:29] <boblet> all us designer-y types love blue-sky nonsense ;-) well, depending on the shade of blue of course
# [10:29] <othermaciej> jorlow: in the case of Workers, I think the answer is recent WebKit-based browsers and recent Gecko-based browsers (most notably Safair, Chrome and Firefox, assuming Chrome has actually shipped the support)
# [10:30] * Joins: primal1_ (n=primal1@pool-98-112-164-140.lsanca.fios.verizon.net)
# [10:30] <jorlow> othermaciej: it's shipping in chrome 3
# [10:30] <jorlow> which is in beta
# [10:30] <jorlow> so no opera?
# [10:31] <othermaciej> I don't believe it's in Opera yet, but testing Opera 10 would tell more definitively
# [10:31] <jorlow> and it's too bad we don't have the html5 spec annotations for these others specs... :-/
# [10:31] <othermaciej> I wouldn't put too much faith in those annotations being up to date in any case
# [10:31] <jorlow> othermaciej: well, it's just a presentation to googlers....so I don't feel like putting in that much effort :-)
# [10:32] <boblet> Microdata is primarily a way to extend HTML’s data structures though… right? To add data that can‘t be included via a predefined method like an existing element…
# [10:32] <jorlow> thanks for the pointers tho
# [10:32] <hsivonen> boblet: no, it's a way to overlay another datastructure on top of an HTML document
# [10:32] * Quits: SuperDot_iPod (n=superdot@adsl-75-61-83-64.dsl.pltn13.sbcglobal.net) ("Get Colloquy for iPhone! http://mobile.colloquy.info")
# [10:32] <hsivonen> boblet: it doesn't extend the data structures of the page itself
# [10:32] <Hixie> what hsivonen said
# [10:33] <othermaciej> Hixie: those seem like good examples
# [10:33] <Hixie> cool
# [10:34] <hsivonen> Hixie: you should ask the RDFa TF to review your study setup before hand
# [10:34] <othermaciej> too bad the blog one has to rely so much on invisible or duplicate data
# [10:34] * Quits: virtuelv (n=virtuelv@pat-tdc.opera.com) (Read error: 110 (Connection timed out))
# [10:35] <Hixie> hsivonen: i'll be lucky if i've finished the material before the study starts
# [10:35] <Hixie> othermaciej: well if you want to encode data that's not in the page, there's not much choice
# [10:36] <Hixie> unless you mean something else
# [10:36] <boblet> hsivonen: thanks for the clarification
# [10:36] * Quits: primal1_ (n=primal1@pool-98-112-164-140.lsanca.fios.verizon.net) (Read error: 104 (Connection reset by peer))
# [10:37] * Joins: primal1_ (n=primal1@pool-98-112-164-140.lsanca.fios.verizon.net)
# [10:37] <boblet> oh btw there’s been a big uptick of interest in HTML5 in Japan in the last month
# [10:37] <othermaciej> yeah, I guess a lot of it wouldn't have reasonably been in the page
# [10:37] <othermaciej> the only obvious duplication was the title itemprop that was also in a title attribute
# [10:37] * Quits: primal1 (n=primal1@pool-98-112-164-140.lsanca.fios.verizon.net) (Read error: 110 (Connection timed out))
# [10:37] * primal1_ is now known as primal1
# [10:37] <boblet> somewhat surprisingly
# [10:38] <Hixie> othermaciej: hm, yeah. none of the proposals we're going to be testing have a way around that.
# [10:38] <Hixie> othermaciej: if you have an idea to fix that i can try to
# [10:38] <Hixie> put it in one of hte variants
# [10:38] <othermaciej> do you have a list of the variants posted anywhere?
# [10:38] * Joins: Rik` (n=Rik`@pha75-2-81-57-187-57.fbx.proxad.net)
# [10:39] <Hixie> http://www.whatwg.org/specs/web-apps/current-work/temp
# [10:39] <othermaciej> and out of curiosity, what's the study methodology? will the test subjects be asked to add the annotations using various syntax options??
# [10:40] <Hixie> various things, the details are still being worked out, but basically: given some intro material, look at this example, can you describe what the annotations are?
# [10:40] <Hixie> given this page, can you add these annotations?
# [10:40] <hsivonen> Hixie: will one person be exposed to only one syntax or multiple but in random order per person?
# [10:40] <jgraham> Is there just one round of testing?
# [10:41] <Hixie> hsivonen: i think each person sees each variant, but we cycle them around so that each one gets seen as the first one an equal number of times
# [10:41] <Hixie> jgraham: yeah
# [10:42] * Quits: erlehmann (n=erlehman@tmo-104-178.customers.d1-online.com) (Read error: 104 (Connection reset by peer))
# [10:42] <jgraham> Hixie: So after the first round can you change the variants to see if some hybrid performs better than the winner from round 1?
# [10:42] <Micheil_away> oh look, it's sir hixie
# [10:42] * Micheil_away is now known as Micheil
# [10:43] <Hixie> jgraham: i doubt we'll have the time
# [10:44] <jgraham> Hixie: OK. That's a shame
# [10:44] <Hixie> jgraham: eh, we haven't had _any_ testing for the rest of the spec
# [10:44] <Hixie> jgraham: so this is an oo% better
# [10:45] <jgraham> Right, but given we do have testing in this case, it is a shame that we can't do even better tests. If you see what I mean
# [10:46] <Philip`> Like, the effort being put into it could perhaps be used more efficiently to get better results?
# [10:46] <othermaciej> usability testing tends to have diminishing returns
# [10:47] <Hixie> jgraham: if we had the resources to do more testing, i'd rather we did it on something else, like table summaries, or whatnot
# [10:47] <Hixie> jgraham: since as othermaciej says, the benefit you get from 0 to 1 units of testing is FAR more than the benefit you get from 1 to 2 units of testing
# [10:48] <jgraham> I agree that the resources would likely be better spent on another part of the spec. But that's often not the way these things work
# [10:48] <hsivonen> did Microsoft ever have an IE8 beta that added transaction methods to localStorage?
# [10:49] <hsivonen> what were the unprefixed MS extensions that Hixie complained about?
# [10:50] <othermaciej> I recall something like that
# [10:50] <othermaciej> can't remember the name of the extension or the subject line
# [10:52] * Quits: ChrisWilson (n=cwilso@nat/microsoft/x-ysshjkxwsqzmpjut) (Read error: 60 (Operation timed out))
# [10:52] <hsivonen> http://lists.w3.org/Archives/Public/public-html/2008Jun/0249.html
# [10:53] * Joins: ROBOd (n=robod@89.122.216.38)
# [10:53] * Joins: ChrisWilson (n=cwilso@nat/microsoft/x-smupmvtpzfdurjlb)
# [10:54] <Hixie> hsivonen: those methods don't do what one might think they do, iirc
# [10:55] <hsivonen> Hixie: maybe, but it sucks that MS proposed making the thing transactional, the proposal was rejected, they shipped and now we find that Chrome and Firefox need them to be transactional, too
# [10:55] <Hixie> iirc the transactional api they were proposing wouldn't have solved the problem being discussed here
# [10:55] <hsivonen> ok
# [10:56] <hsivonen> did they ever mention that IE8 is multiprocess and the specced API would have concurrency issues?
# [10:59] <othermaciej> they never suggested that as a reason for the begin/commit methods
# [10:59] <hsivonen> ok
# [10:59] <Hixie> also their api was a design disaster
# [10:59] <hsivonen> it strange that they didn't
# [10:59] <othermaciej> they said it was based on analysis of author requirements
# [11:00] * Joins: erlehmann (n=erlehman@tmo-104-178.customers.d1-online.com)
# [11:00] <boblet> later all
# [11:00] * Joins: Phae (n=phaeness@gatea.mh.bbc.co.uk)
# [11:00] * Quits: benward (n=benward@98.210.154.133) ("Sleep")
# [11:00] * Joins: primal1_ (n=primal1@pool-98-112-164-140.lsanca.fios.verizon.net)
# [11:01] * Joins: svl (n=me@ip565744a7.direct-adsl.nl)
# [11:01] * Quits: boblet (n=boblet@p1254-ipbf304osakakita.osaka.ocn.ne.jp)
# [11:04] * Quits: Phae (n=phaeness@gatea.mh.bbc.co.uk) (Client Quit)
# [11:05] * Quits: lazni (n=lazni@118.71.169.21) ("Leaving.")
# [11:06] <Hixie> http://lists.w3.org/Archives/Public/public-html-comments/2008Mar/0015.html
# [11:06] <Hixie> is their original e-mail
# [11:08] <Hixie> their solution basically is the same as what the spec says now
# [11:08] <Hixie> except that the spec implies the call to begin() on the first set, and implies the call to commit() when the script ends
# [11:08] <Hixie> s/set/access/
# [11:09] <Hixie> and except that they don't solve the problem of two gets not changing in mid-execution
# [11:12] <Micheil> wow.. the w3c expects heaps in fees to be a member..
# [11:13] <Micheil> oh.. wait, I only found the organization fees.. >_>
# [11:13] <Hixie> there's no individual membership plan for the w3c
# [11:13] <aboodman> roc: can you provide a pointer to where hixie says that the storage mutex breaks with sync script calls across origins?
# [11:13] <Hixie> Micheil: only organisations can join
# [11:13] <Micheil> Hixie: http://www.w3.org/Consortium/membership-faq#individual
# [11:13] <Micheil> I read yes.. plus a bit more
# [11:14] <roc> aboodman: yeah
# [11:14] <Hixie> well, if you have $60k a year, i guess :-)
# [11:14] <Hixie> Micheil: you can join the whatwg or the htmlwg to participate in html5 without cost, though
# [11:14] <Hixie> Micheil: if you are just interested in html5
# [11:14] <roc> aboodman: btw can you get the GMail people to add a feature that links to a Web archive version of a Gmail message? thanks
# [11:14] <aboodman> roc, actually i'll just code it up righ tnow
# [11:15] <aboodman> roc: if you have the subject line that also works
# [11:15] * Quits: MikeSmith (n=MikeSmit@tea12.w3.mag.keio.ac.jp) ("Tomorrow to fresh woods, and pastures new.")
# [11:15] <Micheil> Hixie: well I mean, I wouldn't mind "donating" money to the w3c in the membership style basis, eg, AWIA membership costs 20$ per year (for student), which would be good as a contribution to the w3c's projects
# [11:16] * Joins: Phae (n=phaeness@gatea.mh.bbc.co.uk)
# [11:16] * Quits: primal1 (n=primal1@pool-98-112-164-140.lsanca.fios.verizon.net) (Read error: 110 (Connection timed out))
# [11:16] * primal1_ is now known as primal1
# [11:16] <Micheil> Hixie: I'm not heavily interested in the writing and production of the specs, but I would like to help support them
# [11:16] <Hixie> Micheil: if you think the w3c would spend your money wisely, sure :-)
# [11:16] <Micheil> Hixie: I'm guessing ;P
# [11:17] <roc> aboodman: http://lists.whatwg.org/htdig.cgi/whatwg-whatwg.org/2009-March/018991.html
# [11:17] <aboodman> roc: thx
# [11:17] <Micheil> Hixie: actually, who funds a lot of the costs of running whatwg.org?
# [11:18] <Hixie> Micheil: me (out of pocket)
# [11:18] <Micheil> Hixie: hmm..
# [11:18] <Hixie> it's not much, it's just part of my hosting plan at dreamhost, where i have a whole bunch of domains and stuff hosted
# [11:19] <hsivonen> Hixie: do you have enough referrals to keep your Dreamhost plan free?
# [11:19] * Joins: lazni (n=lazni@123.16.140.21)
# [11:20] <Hixie> hsivonen: not since i started using a private virtual server
# [11:20] <Hixie> but i haven't pushed referrals
# [11:21] * Quits: Lachy (n=Lachlan@85.196.122.246) ("This computer has gone to sleep")
# [11:21] <aboodman> jorlow: ping?
# [11:25] * Quits: poe (n=poe@unaffiliated/poe) ("leaving")
# [11:25] * Quits: primal1 (n=primal1@pool-98-112-164-140.lsanca.fios.verizon.net) (Read error: 54 (Connection reset by peer))
# [11:25] * Joins: primal1 (n=primal1@pool-98-112-164-140.lsanca.fios.verizon.net)
# [11:25] * Quits: primal1 (n=primal1@pool-98-112-164-140.lsanca.fios.verizon.net) (Client Quit)
# [11:27] * Quits: Rik` (n=Rik`@pha75-2-81-57-187-57.fbx.proxad.net) (Read error: 113 (No route to host))
# [11:27] <Micheil> hmm.. I wonder why firefox like's hanging whenever I view the html5 spec
# [11:27] * Joins: Rik` (n=Rik`@81.57.187.57)
# [11:28] * Joins: webben (n=benh@nat/yahoo/x-wmqnkcadlfplfqym)
# [11:28] <Hixie> Micheil: use http://whatwg.org/html5 -- it's a multipge version that firefox can handle better
# [11:28] <hsivonen> Micheil: the script runs to completion and Gecko's JS-to-DOM boundary is slow
# [11:28] <Micheil> hmm..
# [11:28] <jorlow> aboodman: ping
# [11:29] * Quits: webben (n=benh@nat/yahoo/x-wmqnkcadlfplfqym) (Client Quit)
# [11:29] * Micheil is just cross checking his little experiment in html5 with what the docs / spec say
# [11:30] <hsivonen> Micheil: also, the spec uses abnormally complex selectors, so Firefox spends a lot of time doing selector matching
# [11:30] <jorlow> er...pong :-)
# [11:30] <Micheil> :/
# [11:30] <aboodman> jorlow: i'm curious how per-origin mutexes can work in chrome
# [11:30] * Quits: othermaciej (n=mjs@c-69-181-42-237.hsd1.ca.comcast.net)
# [11:31] <Micheil> hsivonen: I might have a look at the js and see if I can see ways to make it faster
# [11:31] <aboodman> per hixie's message at the bottom of: http://lists.whatwg.org/htdig.cgi/whatwg-whatwg.org/2009-March/018991.html
# [11:31] <aboodman> it seems like that applies to chrome
# [11:31] <hsivonen> Micheil: the spec is essentially one of Hixie's evil test cases
# [11:31] <gsnedders> Micheil: Plenty alredy have :)
# [11:31] <Micheil> hmm..
# [11:31] * Joins: Lachy (n=Lachlan@pat-tdc.opera.com)
# [11:31] <hsivonen> Micheil: you could make it walk the DOM in pieces and pump the DOM walker from timeouts
# [11:32] <Hixie> i thought i already did that
# [11:32] <jorlow> aboodman: we do only have one feature that uses locks
# [11:32] * gsnedders thinks the correct solution is to get Firefox to have a quicker JS to DOM impl
# [11:32] <aboodman> yes, but there are multiple locks, right? one per-origin.
# [11:32] <jorlow> and everythign else that could deadlock chrome is already in the spec as things that implicilty release the lock
# [11:32] <jorlow> aboodman: yeah
# [11:33] <Hixie> aboodman: you can do it, it just requires joing the locks in an ungodly number of cases. i think i described it somewhere, or roc did.
# [11:33] <aboodman> if I find something that synchronously executes javascript across iframes, i win the deadlock prize, right?
# [11:33] <jorlow> aboodman: yes
# [11:33] <Micheil> Hixie: it seems slow even when loading it with javascript disabled
# [11:33] <hsivonen> gsnedders: "fast track DOM" is being worked on
# [11:34] <aboodman> resizing an iframe element synchronously fires window.onresize in the iframe content document
# [11:34] <Hixie> Micheil: then it's just slow loading the html. http://whatwg.org/html5 is a smaller file that loads quicker.
# [11:34] <roc> aboodman: really? that seems daft
# [11:34] <roc> it should be asynchronous
# [11:34] <aboodman> i am sure there are others
# [11:35] <aboodman> nobody ever thought this would be an issue before
# [11:35] <Micheil> uhh.. I hate to mention it Hixie, but that js looks rather nasty
# [11:35] <jorlow> aboodman: hopefully not
# [11:35] <Hixie> Micheil: nasty how?
# [11:35] <jorlow> even without considering mutexes, anything that's sync is bad for amny other reasons
# [11:35] * Joins: webben (n=benh@217.12.14.241)
# [11:36] <roc> you kind of want to do cross-process IFRAMEs at some point
# [11:36] <Micheil> well, just the brutal way you're using to include extra scripts: surely you could just place them at the bottom of the file, if you were worried about blocking script loads?
# [11:36] <aboodman> cross-process iframes are hard for a lot of reasons. it probably can't be done without breaking compat.
# [11:36] <Hixie> Micheil: the way the specs are generated, i can't modify anything after the table of contents
# [11:36] <Hixie> Micheil: it's a complicated thing
# [11:36] <Micheil> uh, righteo
# [11:37] <jorlow> aboodman: just cc'ed you on the code review for adding locking :-/
# [11:37] <Hixie> it's what i get for having one source file that generates 7 specs automaticaly
# [11:37] <jorlow> you may be right tho....this might be deadlock prone :-/
# [11:37] <aboodman> jorlow: my comment is: this is lame, we should make the api asynchronous.
# [11:37] <aboodman> :)
# [11:37] <jorlow> aboodman: agreed!!!!
# [11:37] <jorlow> chrome is not goign to implement a global lock for localStorage though
# [11:37] <Hixie> i'm pretty sure a naive per-origin lock will cause deadlocks
# [11:37] <Hixie> but i forget why
# [11:37] <jorlow> so hopefully we can make the per-origin thing work or (better yet) something async
# [11:38] <jorlow> Hixie: did it have to do with onscroll?
# [11:38] <Hixie> no
# [11:38] <aboodman> oh yeah, onscroll might be another.
# [11:38] <Hixie> nothing to do with weird cases like onscroll and onresize
# [11:38] <roc> onscroll is already asynchronous in Gecko
# [11:38] <Hixie> oh, it might have had to do with document.domain
# [11:38] <jorlow> i think it's OK to say that releases the locks
# [11:39] <Hixie> if you grab the lock from two different domains, then they both set document.domain to be the same domain, then call each other
# [11:39] <roc> sure, but let's just make document.domain release the storage mutex
# [11:39] <Hixie> or something
# [11:39] <Hixie> oh we could do that, true
# [11:39] <aboodman> if a bunch of things implicitly release the storage mutex, doesn't that lead to odd behavior?
# [11:39] <aboodman> in edge cases
# [11:39] <Hixie> yes
# [11:39] <roc> I actually want to make onscroll synchronous again, but scrolling to anchors doesn't need to be synchronous anyway
# [11:40] <roc> yes, we want to minimize the number of those things
# [11:40] <Hixie> scrolling to anchors is explicitly async in html5
# [11:40] <Hixie> i'd be almost ok with just making each JS statement release the mutex, frankly
# [11:40] <Hixie> well, no
# [11:40] <roc> I'm not
# [11:40] <Hixie> it'd have to be a block
# [11:40] <lazni> I can't checkout the python build file, svn says 'Could not open the requested SVN filesystem'
# [11:40] <Hixie> at least
# [11:41] * Quits: nessy (n=nessy@203-214-73-15.dyn.iinet.net.au) ("This computer has gone to sleep")
# [11:41] <Hixie> but anyway, not gonna spec that
# [11:41] <Hixie> so that's academic
# [11:41] <roc> I don't want things that should be semantics-preserving transformations to suddenly have bizarre untestable side effects
# [11:41] <Hixie> indeed
# [11:41] <jorlow> personally, i think we have plenty of 'bizarre untestable side effects'
# [11:42] <jorlow> especially plugins calling into your page
# [11:42] <jorlow> since there's no way to know when that's goign to ahppen
# [11:42] <jorlow> or has happened
# [11:42] <roc> you loaded the plugin
# [11:42] <jorlow> yes, but it's far from deterministic
# [11:42] <roc> sure
# [11:42] <Hixie> i'm happy to make plugins as much of a pain as possible
# [11:43] <roc> but at least you opted into some badness
# [11:43] <jorlow> and i think it essentially means that you can't depend on run to completion of local storage if you use ANY plugin
# [11:43] <jorlow> since you don't control that code
# [11:43] <roc> when can plugins can call into your page?
# [11:43] <jorlow> don't know, but it's in the spec
# [11:43] <roc> only during an NPAPI call or when your own script isn't running, I hope
# [11:43] <Hixie> a plugin could also screw up your DOM
# [11:44] <Hixie> so i guess you can't depend on that either?
# [11:44] <jorlow> "Whenever a script calls into a plugin, and whenever a plugin calls into a script, the user agent must release the storage mutex."
# [11:44] <jorlow> Hixie: is it possible to spec that plugins can't do anything while a script is running?
# [11:45] <Hixie> anything is possible, since the spec doesn't say what the plugin api is
# [11:45] <jorlow> is it practical
# [11:45] <jorlow> i have no idea
# [11:45] <aboodman> jorlow: nope. it is valid (imo) for a plugin to call js re-entrantly (is that a word?)
# [11:45] <aboodman> example:
# [11:46] <aboodman> google.gears.database.synchronousTransaction(function(tx) { ... });
# [11:46] <aboodman> that should be allowed.
# [11:46] <roc> gears is an NPAPI plugin?
# [11:46] <aboodman> yes, among other things
# [11:47] <roc> I don't see why the spec has to define what Gears is allowed to do
# [11:47] <aboodman> nevermind, this is off in the weeds.
# [11:48] * Quits: Rik` (n=Rik`@81.57.187.57)
# [11:48] <jorlow> aboodman: but we're initiating it
# [11:48] <jorlow> so it is determinisitc
# [11:48] <jorlow> we == the script
# [11:48] <aboodman> i don't follow
# [11:49] <jorlow> the script is calling google.gears.database.sync.... right? so that's what causes the lock to get dropped
# [11:49] <jorlow> so the web developer knows it's going to get dropped
# [11:49] <Hixie> jorlow is worried about plugin apis that enable plugins to asynchronously call into script while script is running
# [11:49] <Hixie> i don't know if any such apis exist
# [11:49] <aboodman> oh
# [11:49] <aboodman> well Gears has gears.timer
# [11:50] <jorlow> my concern is if I say "plugin, do something" but the plugin waits to change state until an undeterministic time
# [11:50] <roc> but that can't interrupt a running script
# [11:50] <roc> that wouldn't be implementable in Gecko
# [11:50] <aboodman> also gears.location, which fires every so often as the location changes
# [11:50] <Hixie> aboodman: yeah but those are all separate events, right?
# [11:50] <jorlow> but, once again, not during script execution...right?
# [11:50] <Hixie> aboodman: so the mutex is already released by that point
# [11:51] <roc> the things that would worry me are when we call into the plugin due to layout or plugin instantiation, which can happen unexpectedly (but synchronously) during script execution
# [11:51] <Hixie> i guess you could be in trouble if a plugin has its tentacles in multiple origins and intentionally waits for one to call it before returning in another
# [11:51] <aboodman> they all post messages to the browser's event loop
# [11:51] <aboodman> if that is what you mean
# [11:51] <Hixie> but that seems extreme
# [11:51] <Hixie> (and won't be a problem since we release the mutex in those cases)
# [11:52] <jorlow> aboodman: it is what i mean
# [11:52] <jorlow> Hixie: that does still seem like a problem
# [11:52] <Hixie> what does?
# [11:52] <aboodman> jorlow: i think that is the only sensible behavior, since otherwise you'd be accessing the dom/js from multiple threads.
# [11:52] <jorlow> "Hixie: i guess you could be in trouble if a plugin has its tentacles in multiple origins and intentionally waits for one to call it before returning in another"
# [11:52] <aboodman> so i suspose this is how all plugins behave.
# [11:53] <Hixie> jorlow: yeah but that's not a problem if that always releases the mutex
# [11:53] <jorlow> yeah...so I'm wondering if the language should just be "Whenever a script calls into a plugin the user agent must release the storage mutex."
# [11:53] <Hixie> for npapi that may be enough, yeah
# [11:54] <jorlow> things that trigger layout still could be an issue
# [11:54] <jorlow> but at least it's determinisitc
# [11:54] <jorlow> ok...time for dinner
# [11:54] <jorlow> later guys
# [11:54] <aboodman> late
# [11:55] <roc> hmm, is he in Sydney?
# [11:55] <aboodman> tokyo
# [11:58] * Joins: boblet (n=boblet@p1254-ipbf304osakakita.osaka.ocn.ne.jp)
# [12:02] * Quits: boblet (n=boblet@p1254-ipbf304osakakita.osaka.ocn.ne.jp) (Client Quit)
# [12:11] * Quits: lazni (n=lazni@123.16.140.21) ("Leaving.")
# [12:11] * Joins: annevk3 (n=annevk@5355732C.cable.casema.nl)
# [12:14] * Joins: virtuelv (n=virtuelv@pat-tdc.opera.com)
# [12:19] * Quits: maikmerten (n=maikmert@BAE3af2.bae.pppool.de) (Remote closed the connection)
# [12:20] * Quits: mpt (n=mpt@canonical/launchpad/mpt) (Remote closed the connection)
# [12:21] * annevk3 likes Lachy's UI suggestion for <input type=date> and all
# [12:21] * annevk3 hopes someone implements it
# [12:23] <Lachy> annevk3, I'll file a bug and see if we can get it implemented one day
# [12:27] <Hixie> i wish iTunes would consider different seasons of the same show to be the same show
# [12:28] <Hixie> so that i would select a show, then an episode or a season, not have to wade through seasons of other shows looking for the show i want
# [12:35] * Joins: Hish_ (n=chatzill@mail2.n-e-s.de)
# [12:46] <annevk3> I wish downloading shows was possible in the Netherlands
# [12:46] <annevk3> well, legally that is
# [12:48] <Hixie> i wish the bbc would make their shows available too
# [12:48] <Lachy> annevk3, that's the same problem in lots of countries in the world. One day the entertainment industry will learn that the lack of legal alternative is a major cause of piracy, and that it's their own fault.
# [12:48] <Hixie> i so desperately want to watch Have I Got News For You
# [12:48] <Lachy> Hixie, I haven't heard of that show. Is it available on BBC iPlayer?
# [12:48] <Hixie> in the uk, yeah
# [12:48] <Hixie> it's a political comedy show
# [12:49] <Hixie> i hate that the BBC won't sell me their shows
# [12:49] <Hixie> even Top Gear is only available on iTunes after BBC America (a) cuts bits out and (b) airs it on cable TV
# [12:49] <Hixie> they're showing season 7 now, so that's what's being added to itunes
# [12:50] <Lachy> I don't see it in the iPlayer now. Maybe it's currently not available
# [12:50] <Hixie> nevermind that bbc is up to season 13 and bbc america already put seasons 10-12 on itunes
# [12:50] * hsivonen had thought the Have I Got News For You format was Irish
# [12:50] <annevk3> I was once advised by a guy from the BBC to use torrents to get their shows after I complained how they were not available
# [12:50] <Hixie> annevk3: !
# [12:50] <Lachy> Hixie, you could do what I do and get a subscription to a VPN service and bypass their regional restrictions
# [12:51] <Hixie> Lachy: i want to download the shows to watch at my pleasure, not watch them through a frustrating flash-based player
# [12:51] <Phae> HIGNFY is older than 7 days now, so it's not available, btw, rather than regionally restricted.
# [12:51] <Phae> at the mo'
# [12:51] <Hixie> Phae: that makes even less sense
# [12:52] <Phae> just sayin', so you Lachy knew which.
# [12:52] <Phae> -you.
# [12:52] <jgraham> (it is also regionally restricted)
# [12:52] <Phae> yes.
# [12:52] <Lachy> the BBC iPlayer does let you download them using their Adobe Air based application, but they're DRM encumbered and expire after a short period
# [12:52] <Hixie> Lachy: btw, if i have a unix box in the UK, is there some sane way to do port forwarding without buying a vpn service?
# [12:52] <Phae> but since he's going around that restriction..
# [12:52] <jgraham> Hixie: SSH
# [12:52] <Lachy> Hixie, try SSH tunnelling
# [12:52] <Hixie> that requires knowing the endpoint
# [12:52] <Hixie> which is non-trivial for the iplayer
# [12:53] <Lachy> no it doesn't.
# [12:53] <Hixie> i'd have to fake dns and all kinds of stuff
# [12:53] <Hixie> no?
# [12:53] <jgraham> ssh -D {something} should let you use the remote box as a socks proxy
# [12:53] <Hixie> oooh, socks proxy
# [12:53] <Lachy> yeah
# [12:53] <Hixie> didn't know about that
# [12:53] <Hixie> well hey hey
# [12:53] <gsnedders> You could just use a HTTP proxy and download the non-DRM iPhone copies
# [12:53] <Hixie> let's see if that works
# [12:53] <Lachy> that's what I did when I streamed the olympics stuff from the NBC last year
# [12:53] * jgraham doesn't know how to do a proper tunnel but does know that
# [12:54] <Hixie> what's a good tv show that's region restricted these days
# [12:54] * Joins: ttepass- (n=ttepas--@p5B017874.dip.t-dialin.net)
# [12:54] <gsnedders> Hixie: Most sports coverage
# [12:54] <Hixie> "good" :-P
# [12:54] <gsnedders> Hixie: Some is :P
# [12:55] <Hixie> aha, top gear
# [12:55] <Lachy> top gear isn't available any more
# [12:55] <Lachy> oh, some eps are
# [12:55] <Phae> i thought top gear got aired on US TV? although i aassume rather late.
# [12:56] * hsivonen needs some kind of VPNs for dummies for setting up a tunnel between two Linksys RVS4000 boxes
# [12:56] <Hixie> http://www.bbc.co.uk/iplayer/episode/b00lkmth/Top_Gear_Series_13_Episode_2/ is region-limited
# [12:56] <Hixie> i have that episode alreayd
# [12:56] <Hixie> i just want to test it
# [12:59] <Hixie> sweet!!!!
# [12:59] <Hixie> oh yes!
# [12:59] <Hixie> how did i not know about -D before!
# [13:00] <gsnedders> ssh -D {something} where something is what?
# [13:00] <Hixie> a port
# [13:00] <gsnedders> A UK based sshd?
# [13:00] <Hixie> 8888 e.g.
# [13:00] <Hixie> ssh -D 8888 ukhost.example.com
# [13:00] <Hixie> then you set your browser to use localhost:8888 as a socks5 proxy
# [13:00] <Hixie> and away you go
# [13:00] <gsnedders> Ah
# [13:01] <jgraham> (I believe there are also wrapper tools you can use to make a browser use a SOCKS proxy even if it doesn't have built in support for it)
# [13:06] <annevk3> hsivonen, did someone already file a bug that the selected profile is reset in the UI after validating?
# [13:07] <hsivonen> annevk3: no
# [13:08] <annevk3> hsivonen, I don't have an account so if you can ta :)
# [13:08] * Quits: erikvold (n=erikvvol@96.49.192.204)
# [13:10] <hsivonen> annevk3: ok. thanks
# [13:16] * Quits: ttepasse (n=ttepas--@p5B0134C5.dip.t-dialin.net) (Read error: 110 (Connection timed out))
# [13:16] * Joins: ttepasse (n=ttepas--@91.1.58.6)
# [13:19] * Parts: Micheil (n=Micheil@124-170-84-88.dyn.iinet.net.au) ("Any inconviences that may occur are purely coincidential, and we applaud them.")
# [13:20] * Quits: ttepass- (n=ttepas--@p5B017874.dip.t-dialin.net) (Read error: 60 (Operation timed out))
# [13:25] * Joins: myakura (n=myakura@p4203-ipbf7204marunouchi.tokyo.ocn.ne.jp)
# [13:32] * Joins: ttepass- (n=ttepas--@p5B0167A7.dip.t-dialin.net)
# [13:33] * Quits: foolip (n=philip@pat.se.opera.com) (Read error: 110 (Connection timed out))
# [13:41] * Quits: ttepasse (n=ttepas--@91.1.58.6) (Read error: 145 (Connection timed out))
# [13:43] * Joins: erikvvold (n=erikvvol@96.49.192.204)
# [13:43] * Joins: maikmerten (n=maikmert@BAE3af2.bae.pppool.de)
# [13:43] * Quits: Mrmil (n=ut_ollie@host-77-236-204-8.blue4.cz) (Read error: 104 (Connection reset by peer))
# [13:47] * Joins: ttepasse (n=ttepas--@p5B015A87.dip.t-dialin.net)
# [13:53] * Hixie sets up a proxy-to-uk.sh script that does all the changing of hte proxy settings, setting up the tunnel, etc, automatically
# [13:59] * annevk3 prefers Range to be fixed first :)
# [14:00] <Hixie> range would be more work
# [14:00] <Hixie> i'd rather not have him run away because he bit off more than he can chew
# [14:01] <annevk3> fair enough, but Traversal is something we should probably not have implemented at all...
# [14:01] <annevk3> it's a Java-centric API not at all great for the Web
# [14:01] <Hixie> what would you prefer?
# [14:03] <annevk3> not having it at all
# [14:03] <Hixie> raw DOM is worse than traversal
# [14:04] <annevk3> not with previousElementSibling and all
# [14:05] <Hixie> how do i walk just the section elements and their headings?
# [14:05] <annevk3> by writing your own iterator with a filter on top?
# [14:06] <annevk3> dom traversal is mighty bloated; John Resig wrote a good post on that
# [14:07] <hsivonen> Browsers basically implement Traversal for Acid3
# [14:07] <annevk3> it's almost like the last of the parsing quirks :p
# [14:08] * hsivonen wonders how many Web developers have the patience to learn the traversal interfaces instead of doing their own filtering on top of DOM Core
# [14:10] * Quits: ttepass- (n=ttepas--@p5B0167A7.dip.t-dialin.net) (Read error: 110 (Connection timed out))
# [14:10] <Lachy> one thing that would have made traversing the DOM significantly easier is if element.childNodes had an easy way to filter out what was returned
# [14:10] * Joins: ttepass- (n=ttepas--@91.1.109.129)
# [14:11] <Lachy> but element.querySelectorAll(":scope>*"); will largely address that
# [14:13] * Joins: pmuellr (n=pmuellr@nat/ibm/x-jimyurptnnpfignq)
# [14:14] <Hixie> hsivonen: i don't think anyone actually implemented all of traversal just for acid3, they all had code to some extent or another already there iirc
# [14:14] <hsivonen> If Shelley's assertion about the non-removability of CC RDF is true, am I barred from distributing prints of SVG clip art that comes with CC metadata?
# [14:15] <hsivonen> Hixie: there was Traversal work done in Gecko only because of Acid3
# [14:15] <Hixie> oh there was work done, for sure
# [14:15] <hsivonen> Hixie: Traversal seemed like a good thing once, but now it's more like a point on a checklist
# [14:17] <annevk3> it was a point on a checklist (DOM specs to support), now it seems a bit of a maintenance burden
# [14:17] <annevk3> imo
# [14:19] * Quits: erlehmann (n=erlehman@tmo-104-178.customers.d1-online.com) ("Ex-Chat")
# [14:21] <Hixie> http://damowmow.com/playground/microdata/001/introduction
# [14:21] <Lachy> hsivonen, I doubt the removal of the CC RDF metadata is forbidden. As long as the actual requirements of the licence are met. I don't think the technology chosen to express that licence is enforcable.
# [14:22] * Quits: ttepasse (n=ttepas--@p5B015A87.dip.t-dialin.net) (Read error: 110 (Connection timed out))
# [14:22] <Lachy> An author could just as well express the licence conditions in a
# [14:25] <annevk3> Hixie, in syntax it does not explain the name/value pairs are nested inside the <div item>
# [14:25] <Hixie> yeah i'm doing that now
# [14:25] <Hixie> and subject=""
# [14:25] <ttepass-> ?Microdata allows you to mark parts of a web page as machine-readable, named data, building a list of name-value-pairs, so that? etc.? ? ?on the top of webpages? sounds a little bit like angels on an pin.
# [14:25] <hsivonen> "You must keep intact all notices that refer to this License and to the disclaimer of warranties with every copy of the Work You Distribute or Publicly Perform."
# [14:25] <hsivonen> "keep intact all copyright notices for the Work and provide, reasonable to the medium or means You are utilizing"
# [14:26] <hsivonen> I'd argue that RDF/XML is not reasonable for the text/html medium.
# [14:26] <annevk3> Hixie, syntax/examples will be different for each of the three microdata thingies?
# [14:26] <Hixie> yes
# [14:26] <hsivonen> I'd also argue that RDF triples aren't a "notice"
# [14:26] <hsivonen> but IANAL, of course
# [14:26] <Hixie> ttepass-: a lot of microdata/rdfa/microformats/rdf/etc are angels on a pin
# [14:27] <hsivonen> it would be a huge bug in the CC licenses if use on media that doesn't support RDF/XML were prohibited
# [14:28] <ttepass-> (Depends on the viewpoint, of course) My point is more that ?to layer lists on top of a page? is rather distant from the viewpoint of web authors while ?marking things up? is more intuitive.
# [14:29] <annevk3> http://code.google.com/p/pywebsocket/ cool
# [14:29] * Joins: Unixmonkey (n=Unixmonk@mail.reprographix.com)
# [14:29] <Hixie> ttepass-: not sure what "marking things up" means
# [14:30] <Hixie> in this context
# [14:31] * Joins: GPHemsley (n=GPHemsle@pdpc/supporter/student/GPHemsley)
# [14:31] <ttepass-> I'm not sure, what ?layering on top? in this context means.
# [14:31] <Hixie> ah :-)
# [14:31] <Hixie> it means that you have your HTML file
# [14:32] <Hixie> and if you imagine it as a printed sheet of paper
# [14:32] <Hixie> you then layer a transparency with the additional data on top of the paper
# [14:32] <Hixie> so that it doesn't affect either the paper, not the transparency
# [14:32] <ttepass-> Ok, that's a metaphor.
# [14:33] <ttepass-> But ... isn't Microdata with its DOM API a part of the DOM tree?
# [14:35] <ttepass-> Hm. Arguing over language an metaphors is not really productive. Feel free to ignore me. ;)
# [14:36] <Hixie> well it's part of the document, sure
# [14:36] <Hixie> i mean it uses attributes and stuff in the HTML file
# [14:36] <Hixie> but it doesn't affect the HTML document's semantics
# [14:36] <Hixie> other than adding these separate name-value pairs
# [14:36] <Hixie> which are otherwise unrelated to the content
# [14:38] <ttepass-> There I'd ask what "otherwise unrelated" means since a lot of values will be the textContent of Elements.
# [14:39] <Hixie> well like if you say <div item><button itemprop="x" ...></button></div>
# [14:39] <Hixie> ...it would be wrong to say that the button has itemprop "x"
# [14:39] <Hixie> the following would mean the same:
# [14:39] <Hixie> <div item><button ...></button><meta itemprop="x" content=""></div>
# [14:39] <Hixie> ...as far as the microdata goes
# [14:40] <Hixie> as in, the transparency isn't printed on top of the paper
# [14:40] <Hixie> it's just layered on top
# [14:40] <Hixie> anyway
# [14:40] <Hixie> 5:35am
# [14:40] <Hixie> far past bed time
# [14:40] <ttepass-> Good night.
# [14:41] * Quits: wakaba_ (n=wakaba_@122x221x184x68.ap122.ftth.ucom.ne.jp) ("Leaving...")
# [14:52] * Joins: boblet (n=boblet@p1254-ipbf304osakakita.osaka.ocn.ne.jp)
# [15:01] * Joins: dbaron (n=dbaron@c-69-140-1-234.hsd1.va.comcast.net)
# [15:04] * Joins: da3d (n=opera@h11n1fls34o986.telia.com)
# [15:04] * Quits: shepazu (n=schepers@pool-96-252-125-220.bstnma.fios.verizon.net)
# [15:05] * Joins: TabAtkins (n=chatzill@99-35-179-251.lightspeed.hstntx.sbcglobal.net)
# [15:05] <JohnResig> Lachy: ftp://ftp.mozilla.org/pub/firefox/releases/
# [15:06] <Lachy> JohnResig, what about that?
# [15:06] <JohnResig> Lachy: you twittered earlier looking for an official 2.0 release
# [15:06] <Lachy> no I didn't
# [15:06] <JohnResig> huh
# [15:06] <JohnResig> lol, I'm going crazy, nevermind
# [15:06] * Joins: BlurstOfTimes (n=blurstof@168.203.117.59)
# [15:06] <Lachy> ok
# [15:07] <JohnResig> ah, apparently it was someone with a very similar username - I follow too many people
# [15:07] <Lachy> no worries
# [15:09] <Lachy> out of curiosity, who has a nickname similar to mine?
# [15:09] <JohnResig> it was someone named 'Lachlan Hardy'
# [15:09] <Lachy> oh
# [15:09] <Lachy> I know him.
# [15:09] <JohnResig> my brain saw that and made a leap :)
# [15:16] * Joins: atha (n=atha@unaffiliated/athayde)
# [15:17] <hsivonen> Philip`: do you have test cases showing that Google's "RDFa implementation" isn't?
# [15:39] * Quits: atha (n=atha@unaffiliated/athayde) ("leaving")
# [15:43] * Joins: mpt (n=mpt@canonical/launchpad/mpt)
# [15:47] * Quits: Amorphous (i=jan@unaffiliated/amorphous) (Read error: 145 (Connection timed out))
# [15:51] * Quits: jacobolus (n=jacobolu@dhcp-0059871802-99-6d.client.student.harvard.edu) (Remote closed the connection)
# [15:57] * Joins: miketaylr (n=mtaylor@38.117.156.163)
# [15:58] * Joins: Amorphous (i=jan@unaffiliated/amorphous)
# [16:04] * Joins: MikeSmith (n=MikeSmit@EM114-48-200-157.pool.e-mobile.ne.jp)
# [16:05] * Joins: shepazu (n=schepers@31-33-83.wireless.csail.mit.edu)
# [16:05] * Joins: jacobolus (n=jacobolu@dhcp-0059871802-99-6d.client.student.harvard.edu)
# [16:06] * Joins: lazni (n=lazni@118.71.42.132)
# [16:07] * Quits: jacobolus (n=jacobolu@dhcp-0059871802-99-6d.client.student.harvard.edu) (Remote closed the connection)
# [16:09] * Joins: jacobolus (n=jacobolu@dhcp-0059871802-99-6d.client.student.harvard.edu)
# [16:09] * Joins: dglazkov (n=dglazkov@c-67-188-0-62.hsd1.ca.comcast.net)
# [16:10] * Quits: jacobolus (n=jacobolu@dhcp-0059871802-99-6d.client.student.harvard.edu) (Remote closed the connection)
# [16:12] * Quits: dglazkov (n=dglazkov@c-67-188-0-62.hsd1.ca.comcast.net) (Client Quit)
# [16:15] * Quits: jorlow (n=jorlow@220.109.219.244) (Read error: 110 (Connection timed out))
# [16:15] * Joins: Kalms (n=rasmuska@81.161.185.108)
# [16:21] * Joins: hobertoAtWork (n=hobertoa@gw1.mcgraw-hill.com)
# [16:25] * Joins: ttepasse (n=ttepas--@p5B014A41.dip.t-dialin.net)
# [16:27] * Quits: ttepass- (n=ttepas--@91.1.109.129) (Read error: 145 (Connection timed out))
# [16:31] * Joins: jacobolus (n=jacobolu@140.247.40.237)
# [16:33] * Joins: yutak_home (n=kee@M006079.ppp.dion.ne.jp)
# [16:34] * Joins: cardona507 (n=cardona5@c-67-180-160-250.hsd1.ca.comcast.net)
# [16:36] <hsivonen> anyone got the URL for the source of this gem? http://junkyard.damowmow.com/339
# [16:41] * Quits: jacobolus (n=jacobolu@140.247.40.237) (Remote closed the connection)
# [16:44] * Quits: TabAtkins (n=chatzill@99-35-179-251.lightspeed.hstntx.sbcglobal.net) (Read error: 54 (Connection reset by peer))
# [16:47] <Lachy> "The SVG is, itself, XML. Doesn't matter how you want to think of it in HTML, it is XML. " -- Shelley Powers
# [16:47] <Lachy> *facepalm*
# [16:47] * Joins: TabAtkins (n=chatzill@99-35-179-251.lightspeed.hstntx.sbcglobal.net)
# [16:47] <Lachy> that's like arguing that HTML with XHTML-like talismans is XHTML
# [16:47] <TabAtkins> hsivonen: I think it ws puysl.com
# [16:47] <TabAtkins> It doesn't seem to be attested any more, but it *does* lock Firefox.
# [16:48] <TabAtkins> (thus why I dropped from the room)
# [16:48] <miketaylr> and it does start HORRIBLE music
# [16:49] * TabAtkins wonders why they even bother with CSS when they're just going to use <font> everywhere anyway.
# [16:50] <annevk3> Lachy, it's pretty common to confuse language and syntax
# [16:50] <Lachy> I wish Shelley would understand that the text/html serialisation of SVG is a distinct, non-XML format, just like HTML is a distinct non-XML format from XHTML
# [16:50] * Joins: erlehmann (n=erlehman@tmo-104-178.customers.d1-online.com)
# [16:51] * Quits: lazni (n=lazni@118.71.42.132) ("Leaving.")
# [16:57] * Quits: mpt (n=mpt@canonical/launchpad/mpt) (Remote closed the connection)
# [16:59] * Parts: Kalms (n=rasmuska@81.161.185.108)
# [16:59] * Quits: weinig (n=weinig@c-67-180-35-124.hsd1.ca.comcast.net)
# [17:01] * Quits: erlehmann (n=erlehman@tmo-104-178.customers.d1-online.com) (Read error: 54 (Connection reset by peer))
# [17:05] * Joins: erlehmann (n=erlehman@tmo-104-178.customers.d1-online.com)
# [17:07] * Quits: Maurice (n=ano@80.101.46.164) ("Disconnected...")
# [17:10] * Joins: Steve^ (n=steve@92.40.102.75.sub.mbb.three.co.uk)
# [17:14] * Joins: dglazkov (n=dglazkov@nat/google/x-qzclyyjcrvtebfmf)
# [17:15] * Quits: webben (n=benh@217.12.14.241) (Read error: 145 (Connection timed out))
# [17:15] <erlehmann> hsivonen, i learnt DOM traversal. then i encountered the shiny new selector stuff :>
# [17:16] <TabAtkins> Heh, I never learned proper DOM traversal. I just avoided javascript until jQuery got popular. ^_^
# [17:18] * Joins: webben (n=benh@nat/yahoo/x-ynuvbpmgudlysbhx)
# [17:22] <erlehmann> TabAtkins, i have some unicorn pics to spare. wanna start the DOM Super Friends ?
# [17:22] <TabAtkins> I'd rather start the DOM Legion of Doom.
# [17:22] <erlehmann> that calls for a shoop. brb.
# [17:23] <TabAtkins> Also: WebIDL means that native javascript is always going to be saddled with a Java-like interface in an attempt to be language-agnostic, so we'll never have a DOM that's actually easy to use and leverages the unique strengths and design structure of js.
# [17:25] <annevk3> actually, Web IDL is an attempt to get away from that
# [17:25] <annevk3> and is doing pretty well
# [17:26] <TabAtkins> That'd be good. I'm honestly just name-dropping, and don't understand a lot of the crazy interface descriptions you people use. I just know that describing apis in terms of some interface language has been cited as the reason why we can't do smart things in javascript, because some languages that would implement the api don't have the relevant abilities.
# [17:26] <TabAtkins> It is good to know that people are fixing that.
# [17:28] <annevk3> what you're describing is roughly the problem with the original IDL language (OMGIDL) and the DOM specs that used it
# [17:28] <TabAtkins> Cool.
# [17:28] <TabAtkins> I am now informed, and a better consumer.
# [17:30] * Joins: paul_irish (n=paul_iri@12.33.239.250)
# [17:32] * Quits: paul_irish (n=paul_iri@12.33.239.250) (Client Quit)
# [17:32] * Joins: paul_irish (n=paul_iri@12.33.239.250)
# [17:43] * Joins: jacobolus (n=jacobolu@dhcp-0059871802-99-6d.client.student.harvard.edu)
# [17:45] * Joins: mcdave (n=Adium@cm-83-97-166-34.telecable.es)
# [17:45] * Quits: jacobolus (n=jacobolu@dhcp-0059871802-99-6d.client.student.harvard.edu) (Remote closed the connection)
# [17:48] * Joins: mcdave_ (n=mcdave@cm-83-97-166-34.telecable.es)
# [17:49] * Quits: mcdave_ (n=mcdave@cm-83-97-166-34.telecable.es) (Client Quit)
# [17:49] * Quits: virtuelv (n=virtuelv@pat-tdc.opera.com) (Read error: 110 (Connection timed out))
# [17:50] * Joins: mlpug (n=mlpug@a88-115-164-40.elisa-laajakaista.fi)
# [17:50] * Joins: tantek (n=tantek@adsl-69-107-12-74.dsl.pltn13.pacbell.net)
# [17:51] * Quits: Lachy (n=Lachlan@pat-tdc.opera.com) ("This computer has gone to sleep")
# [17:55] * Joins: dave_levin (n=dave_lev@74.125.59.65)
# [17:58] * Quits: paul_irish (n=paul_iri@12.33.239.250)
# [18:01] * Quits: pesla (n=retep@procurios.xs4all.nl) ("( www.nnscript.com :: NoNameScript 4.21 :: www.esnation.com )")
# [18:03] * Quits: erlehmann (n=erlehman@tmo-104-178.customers.d1-online.com) (Read error: 104 (Connection reset by peer))
# [18:03] * Quits: boblet (n=boblet@p1254-ipbf304osakakita.osaka.ocn.ne.jp)
# [18:03] * Joins: Lachy (n=Lachlan@85.196.122.246)
# [18:06] * Quits: StationsPatient (n=kristall@c-base/crew/kristall) ("Wünsche weiterhin guten Flug")
# [18:07] * Joins: ap (n=ap@nat/apple/x-ixrewoylqgwnxzet)
# [18:12] * Joins: paul_irish (n=paul_iri@12.33.239.250)
# [18:18] * Quits: annevk3 (n=annevk@5355732C.cable.casema.nl)
# [18:20] * Joins: sbublava (n=stephan@77.118.137.143.wireless.dyn.drei.com)
# [18:21] * Joins: Maurice (i=copyman@5ED548D4.cable.ziggo.nl)
# [18:25] * Joins: erlehmann (n=erlehman@tmo-104-178.customers.d1-online.com)
# [18:28] * Joins: jlebar (n=jlebar@nat/mozilla/x-pkgzjplybwqdnakr)
# [18:30] * Quits: myakura (n=myakura@p4203-ipbf7204marunouchi.tokyo.ocn.ne.jp) ("Leaving...")
# [18:30] * Quits: mcdave (n=Adium@cm-83-97-166-34.telecable.es) ("Leaving.")
# [18:30] * Quits: Phae (n=phaeness@gatea.mh.bbc.co.uk)
# [18:35] * Joins: mcdave (n=Adium@cm-83-97-166-34.telecable.es)
# [18:36] * Quits: tantek (n=tantek@adsl-69-107-12-74.dsl.pltn13.pacbell.net)
# [18:39] * Joins: bgalbraith (n=bgalbrai@nat/mozilla/x-ejnobuccbzjpxpsy)
# [18:50] * Quits: erlehmann (n=erlehman@tmo-104-178.customers.d1-online.com) ("Ex-Chat")
# [18:59] * Joins: mpt (n=mpt@canonical/launchpad/mpt)
# [19:10] * Joins: benward (n=benward@nat/yahoo/x-svkbinepcqpczyov)
# [19:14] * Joins: aja_ (n=aja@70.230.190.133)
# [19:16] * Quits: aja_ (n=aja@70.230.190.133) (Connection reset by peer)
# [19:16] * Joins: aja_ (n=aja@70.230.190.133)
# [19:18] * Joins: jwalden (n=waldo@nat/mozilla/x-duunneqwuajfjoca)
# [19:21] <aja_> hsivonen (or whoever): noticed that since pedagogical and polyglot have been added, should not longer always have selected for permissive option....selection isn't being "remembered"
# [19:22] * Joins: sbublava_ (n=stephan@77.118.93.102.wireless.dyn.drei.com)
# [19:23] * Joins: zdobersek (n=zan@cpe-92-37-68-167.dynamic.amis.net)
# [19:26] * Quits: mpt (n=mpt@canonical/launchpad/mpt) (Read error: 113 (No route to host))
# [19:28] * Joins: mpt (n=mpt@canonical/launchpad/mpt)
# [19:34] * Quits: sbublava (n=stephan@77.118.137.143.wireless.dyn.drei.com) (Read error: 104 (Connection reset by peer))
# [19:38] <Lachy> hsivonen, Is the polyglot checking mode still under development? I just tried it, and it failed to warn about either unquoted attributes or omitted traling slashes on void elements
# [19:39] <Lachy> hsivonen, also, will it be a fully XML-compatible polyglot check, or just a few checks for commonly requested talismans?
# [19:41] <aja_> Lachy: read sam's blog post?
# [19:46] <Lachy> aja_, I have now
# [19:46] * Joins: GarethAdams (n=GarethAd@pdpc/supporter/active/GarethAdams)
# [19:48] * Quits: mpt (n=mpt@canonical/launchpad/mpt) ("Ex-Chat")
# [19:48] <aja_> note that "permissive" keeps getting selected, i.e. not remembered. plolglot seemed to work okay on google.com anyway
# [19:49] <aja_> s/plolglot/polyglot/
# [19:52] * Quits: cardona507 (n=cardona5@c-67-180-160-250.hsd1.ca.comcast.net)
# [19:53] * Joins: cying (n=cying@70.90.171.153)
# [19:53] * Quits: mcdave (n=Adium@cm-83-97-166-34.telecable.es) ("Leaving.")
# [19:55] * Joins: rubys1 (n=rubys@cpe-065-190-139-141.nc.res.rr.com)
# [19:56] <rubys1> the permissive getting re-selected problem has been fixed, but the fix hasn't yet been deployed
# [19:57] <rubys1> Lachy: http://validator.nu/?doc=http%3A%2F%2Fgoogle.com%2F&profile=polyglot
# [19:59] * Quits: Steve^ (n=steve@92.40.102.75.sub.mbb.three.co.uk) (Read error: 110 (Connection timed out))
# [19:59] * Joins: weinig (n=weinig@nat/apple/x-piqadripvejanctq)
# [19:59] <Lachy> rubys1, I already looked at that
# [20:00] <Lachy> also, it doesn't answer my question about what exactly it will eventually check when complete
# [20:00] * Joins: jacobolus (n=jacobolu@140.247.158.233)
# [20:01] <rubys1> the only answer I can give is: whatever testcases Tantek and/or Lars provides. Until or unless such tests are provided, I don't intend to develop this any further.
# [20:01] <rubys1> And if such tests aren't forthcoming, I plan to remove this function.
# [20:09] * Quits: jacobolus (n=jacobolu@140.247.158.233) (Remote closed the connection)
# [20:09] * Joins: jlebar_ (n=jlebar@nat/mozilla/x-behjrokzondvrhmk)
# [20:11] * Joins: weinig_ (n=weinig@17.246.16.129)
# [20:22] * Quits: jlebar (n=jlebar@nat/mozilla/x-pkgzjplybwqdnakr) (Read error: 110 (Connection timed out))
# [20:23] * Quits: miketaylr (n=mtaylor@38.117.156.163) ("adios.")
# [20:26] * Joins: erlehmann (n=erlehman@tmo-104-178.customers.d1-online.com)
# [20:26] <Philip`> hsivonen: There's http://www.google.com/webmasters/tools/richsnippets?url=http://philip.html5.org/demos/rdfa/richsnippet.html
# [20:27] * Quits: weinig (n=weinig@nat/apple/x-piqadripvejanctq) (Read error: 110 (Connection timed out))
# [20:27] * weinig_ is now known as weinig
# [20:27] <Philip`> I presume it's easy to write valid RDFa that Google won't understand or will misinterpret, too
# [20:27] <Philip`> and I was kind of planning to look at that but I haven't got around to it
# [20:29] * Joins: zdobersek1 (n=zan@cpe-92-37-67-192.dynamic.amis.net)
# [20:31] <Lachy> rubys1, in that case, what significant differences do you perceive between the pedagogical and polyglot modes of validation?
# [20:32] <Lachy> If you only intend the polyglot checker to check the specific features that get submitted as test cases, rather than a fully XHTML-compatible polyglot document checker, why do you need both modes?
# [20:33] <Lachy> I'd rather not have a validator claim to have a polyglot checking mode, if it isn't at least fully XHTML-compatible
# [20:34] <Lachy> (I can accept it opting to not check for things like the inclusion of tbody or LF after <pre> start tags, etc.)
# [20:38] <rubys1> My sense is that Lars and Tantek have different goals... I'm simply prepared to track both.
# [20:39] <rubys1> specifically, I sense that Tantek will want some checks that Lars isn't happy with
# [20:39] * Joins: virtuelv (n=virtuelv@125.175.251.212.customer.cdi.no)
# [20:40] <Lachy> I don't know who Lars is
# [20:41] <rubys1> Lars Gunther
# [20:41] * Quits: weinig (n=weinig@17.246.16.129)
# [20:41] <Lachy> ah, ok.
# [20:42] <rubys1> My feeling is that a number of people want "XHTML", but when you ask specific questions about weird edge cases, they say "I don't want that". So, I'm trying to get people to say what they really want.
# [20:44] <rubys1> overall, I think that informational messages about implicitly closed tags would be a good thing, checking for attribute quotes is a MEH for many people, and flagging   is something that only a few crazies will want.
# [20:45] * Quits: zdobersek (n=zan@cpe-92-37-68-167.dynamic.amis.net) (Read error: 110 (Connection timed out))
# [20:46] * Quits: maikmerten (n=maikmert@BAE3af2.bae.pppool.de) (Connection timed out)
# [20:46] * Joins: zdobersek (n=zan@cpe-92-37-72-206.dynamic.amis.net)
# [20:47] * Joins: maikmerten (n=maikmert@89.61.178.80)
# [20:51] * Joins: weinig (n=weinig@nat/apple/x-tartuvlxzexbytck)
# [20:52] * Joins: mcdave (n=Adium@cm-83-97-164-135.telecable.es)
# [20:52] <MikeSmith> rubys1: qa-dev now has your latest change
# [20:54] <rubys1> MikeSmith: not as far as I can tell
# [20:55] <rubys1> http://validator.nu/ has profile, http://qa-dev.w3.org:8888/ does not
# [20:56] <MikeSmith> rubys1: can you give me a test case?
# [20:59] * Joins: Super-Dot (n=Super-Do@66-240-27-50.isp.comcastbusiness.net)
# [20:59] <rubys1> ok, test case: go to "http://qa-dev.w3.org:8888/", look for string "profile". End of test case.
# [21:01] <MikeSmith> rubys1: yeah, clearly not synced up. I'll go check
# [21:02] * Quits: mcdave (n=Adium@cm-83-97-164-135.telecable.es) ("Leaving.")
# [21:03] * Quits: zdobersek1 (n=zan@cpe-92-37-67-192.dynamic.amis.net) (Read error: 110 (Connection timed out))
# [21:03] * Joins: aroben (n=aroben@unaffiliated/aroben)
# [21:04] <MikeSmith> rubys1: It seems the problem is that I've not done the "svn switch" thing it my workspace on that machine
# [21:04] <rubys1> that would do it. :-)
# [21:05] <MikeSmith> PITA.. thanks a lot cvsdude people
# [21:05] <rubys1> I simply opted to do a clean checkout
# [21:07] * Quits: aja_ (n=aja@70.230.190.133) (Read error: 104 (Connection reset by peer))
# [21:09] <erlehmann> rubys1, what weird edge cases are there in XHTML ?
# [21:11] <rubys1> erlehmann: http://wiki.whatwg.org/wiki/HTML_vs._XHTML
# [21:14] <erlehmann> rubys1, thats not weird on the XHTML side, at least -- its consistent.
# [21:15] * erlehmann notes the XHTML side of the comparison frequently has shorter text and no „for some elements / attributes“ clauses.
# [21:17] <rubys1> erlehmann: my weblog is consistently well formed XML, and (except for one issue intentionally made to prove a point) valid both as XHTML5 and HTML5.
# [21:20] * gsnedders finds Opera has web compat bugs with the UCAS Apply website
# [21:21] * Joins: dpranke (n=Adium@nat/google/x-touqirnmubesiukw)
# [21:33] * Quits: jlebar_ (n=jlebar@nat/mozilla/x-behjrokzondvrhmk) (Read error: 60 (Operation timed out))
# [21:33] <MikeSmith> rubys1: OK, now qa-dev has the latest
# [21:36] * Quits: sbublava_ (n=stephan@77.118.93.102.wireless.dyn.drei.com)
# [21:41] * Quits: vvv (n=vvv@mediawiki/VasilievVV) (Read error: 104 (Connection reset by peer))
# [21:42] * Quits: erlehmann (n=erlehman@tmo-104-178.customers.d1-online.com) (Read error: 113 (No route to host))
# [21:43] * Quits: GarethAdams (n=GarethAd@pdpc/supporter/active/GarethAdams)
# [21:47] * Joins: vvv (n=vvv@mediawiki/VasilievVV)
# [21:51] * Quits: roc (n=roc@121-72-191-237.dsl.telstraclear.net)
# [21:59] * Joins: Rik` (n=Rik`@pha75-2-81-57-187-57.fbx.proxad.net)
# [22:01] * Quits: MikeSmith (n=MikeSmit@EM114-48-200-157.pool.e-mobile.ne.jp) ("Tomorrow to fresh woods, and pastures new.")
# [22:02] * Joins: pererik (n=pe@unaffiliated/pererik)
# [22:02] * Joins: erlehmann (n=erlehman@tmo-104-178.customers.d1-online.com)
# [22:02] * Joins: ap_ (n=ap@17.246.19.174)
# [22:09] * Joins: miketaylr (n=mtaylor@38.117.156.163)
# [22:10] * Quits: ap_ (n=ap@17.246.19.174) (Remote closed the connection)
# [22:11] * Quits: ap (n=ap@nat/apple/x-ixrewoylqgwnxzet) (Read error: 104 (Connection reset by peer))
# [22:11] * Joins: ap (n=ap@17.203.14.219)
# [22:19] * Quits: mlpug (n=mlpug@a88-115-164-40.elisa-laajakaista.fi) (Remote closed the connection)
# [22:21] <Lachy> hsivonen, bug in validator.nu:
# [22:21] <Lachy> "Error: Bad value “http://www.w3.org/1999/xlink” for the attribute “xmlns:link” (only “http://www.w3org/1999/xlink” permitted here).
# [22:21] <Lachy> From line 5, column 1; to line 7, column 48
# [22:21] <Lachy> "
# [22:21] <Lachy> note where it incorrectly says "only “http://www.w3org/1999/xlink” permitted here"
# [22:21] <Lachy> s/w3org/w3.org/
# [22:21] * Quits: maikmerten (n=maikmert@89.61.178.80) (Remote closed the connection)
# [22:22] * Joins: jacobolus (n=jacobolu@dhcp-0059871802-99-6d.client.student.harvard.edu)
# [22:26] <Lachy> hsivonen, I'll file a bug
# [22:27] * Joins: ttepass- (n=ttepas--@p5B0142FA.dip.t-dialin.net)
# [22:29] * Quits: Hish_ (n=chatzill@mail2.n-e-s.de) (Read error: 110 (Connection timed out))
# [22:30] * Quits: pmuellr (n=pmuellr@nat/ibm/x-jimyurptnnpfignq)
# [22:32] * Quits: ROBOd (n=robod@89.122.216.38) ("http://www.robodesign.ro")
# [22:33] * Joins: jlebar (n=jlebar@nat/mozilla/x-xzwlbzpufhjtnohd)
# [22:34] * Joins: mcdave (n=Adium@83.97.164.135)
# [22:35] * Quits: Creap (n=Creap@vemod.brg.sgsnet.se) (Remote closed the connection)
# [22:35] * Quits: mcdave (n=Adium@83.97.164.135) (Client Quit)
# [22:35] * Joins: mcdave (n=mcdave@cm-83-97-164-135.telecable.es)
# [22:37] * Quits: erlehmann (n=erlehman@tmo-104-178.customers.d1-online.com) (Read error: 113 (No route to host))
# [22:37] * Joins: erlehmann (n=erlehman@tmo-104-178.customers.d1-online.com)
# [22:41] * Joins: rubys (n=rubys@cpe-065-190-139-141.nc.res.rr.com)
# [22:43] * Quits: erlehmann (n=erlehman@tmo-104-178.customers.d1-online.com) (Read error: 54 (Connection reset by peer))
# [22:44] * Joins: erlehmann (n=erlehman@tmo-104-178.customers.d1-online.com)
# [22:49] * Quits: ttepasse (n=ttepas--@p5B014A41.dip.t-dialin.net) (Read error: 110 (Connection timed out))
# [22:55] * Quits: svl (n=me@ip565744a7.direct-adsl.nl) (Nick collision from services.)
# [22:55] * Joins: svl (n=me@ip565744a7.direct-adsl.nl)
# [22:59] * Joins: Creap (n=Creap@vemod.brg.sgsnet.se)
# [23:03] * Joins: and (n=and@apo29.girton.cam.ac.uk)
# [23:07] * Joins: cardona507 (n=cardona5@c-67-180-160-250.hsd1.ca.comcast.net)
# [23:09] * Quits: da3d (n=opera@h11n1fls34o986.telia.com)
# [23:13] * Joins: othermaciej (n=mjs@c-69-181-42-237.hsd1.ca.comcast.net)
# [23:15] * Quits: virtuelv (n=virtuelv@125.175.251.212.customer.cdi.no) ("Ex-Chat")
# [23:23] * Parts: rubys (n=rubys@cpe-065-190-139-141.nc.res.rr.com)
# [23:25] * ap is now known as ap|away
# [23:30] * Quits: rubys1 (n=rubys@cpe-065-190-139-141.nc.res.rr.com) ("Leaving.")
# [23:34] * Quits: BlurstOfTimes (n=blurstof@168.203.117.59) ("Leaving...")
# [23:37] * Quits: Unixmonkey (n=Unixmonk@mail.reprographix.com)
# [23:37] * Quits: shepazu (n=schepers@31-33-83.wireless.csail.mit.edu)
# [23:40] * Quits: Maurice (i=copyman@5ED548D4.cable.ziggo.nl)
# [23:42] * Joins: rkennedy (n=rkennedy@pixout.appriss.com)
# [23:50] * Quits: zdobersek (n=zan@cpe-92-37-72-206.dynamic.amis.net) ("Leaving.")
# [23:53] * Quits: webben (n=benh@nat/yahoo/x-ynuvbpmgudlysbhx) (Read error: 110 (Connection timed out))
# [23:58] * Joins: webben (n=benh@genkt-048-074.t-mobile.co.uk)
# [23:58] * Quits: webben (n=benh@genkt-048-074.t-mobile.co.uk) (Read error: 104 (Connection reset by peer))
# Session Close: Wed Sep 09 00:00:00 2009

The end :)