Options:
- # Session Start: Mon Jun 29 00:00:00 2009
- # Session Ident: #whatwg
- # [00:01] * Quits: jacobolus_ (n=jacobolu@c-98-248-43-68.hsd1.ca.comcast.net) (Remote closed the connection)
- # [00:01] * Joins: annevk2 (n=annevk@5ED2D27A.cable.ziggo.nl)
- # [00:10] * Joins: Lachy (n=Lachlan@85.196.122.246)
- # [00:21] * Quits: annodomini (n=lambda@wikipedia/lambda)
- # [00:34] * Quits: danbri (n=danbri@unaffiliated/danbri) ("going back to danbri.org")
- # [00:41] * Quits: annevk2 (n=annevk@5ED2D27A.cable.ziggo.nl) (Remote closed the connection)
- # [00:57] * Quits: dolske (n=dolske@c-76-103-40-203.hsd1.ca.comcast.net)
- # [01:03] * Quits: T-- (n=ttepasse@p5B016E94.dip.t-dialin.net) ("?Q")
- # [01:15] * Joins: dglazkov (n=dglazkov@c-69-181-143-54.hsd1.ca.comcast.net)
- # [01:16] * Quits: Lachy (n=Lachlan@85.196.122.246) ("This computer has gone to sleep")
- # [01:17] * Joins: Lachy (n=Lachlan@85.196.122.246)
- # [01:21] * Quits: svl (n=me@ip565744a7.direct-adsl.nl) ("And back he spurred like a madman, shrieking a curse to the sky.")
- # [01:27] * Joins: jacobolus (n=jacobolu@adsl-69-228-190-230.dsl.snfc21.pacbell.net)
- # [01:30] * Parts: Midler (n=midler@79.138.206.25.bredband.tre.se) ("Leaving.")
- # [01:32] * Joins: dolske (n=dolske@nat/mozilla/x-833459c5b074860a)
- # [01:34] * Quits: heycam (n=cam@124-168-58-6.dyn.iinet.net.au) ("bye")
- # [01:39] * Quits: dglazkov (n=dglazkov@c-69-181-143-54.hsd1.ca.comcast.net)
- # [01:50] * Joins: annodomini (n=lambda@wikipedia/lambda)
- # [02:00] * Quits: dolske (n=dolske@nat/mozilla/x-833459c5b074860a) (Read error: 110 (Connection timed out))
- # [02:02] * Joins: MikeSmith (n=MikeSmit@EM114-48-57-61.pool.e-mobile.ne.jp)
- # [02:10] * Quits: jacobolus (n=jacobolu@adsl-69-228-190-230.dsl.snfc21.pacbell.net) (Remote closed the connection)
- # [02:10] * Joins: dolske (n=dolske@c-76-103-40-203.hsd1.ca.comcast.net)
- # [02:36] * Joins: hdh (n=hdh@hdh-1-pt.tunnel.tserv20.hkg1.ipv6.he.net)
- # [02:39] * Quits: weinig (n=weinig@c-67-180-35-124.hsd1.ca.comcast.net)
- # [02:41] * Joins: heycam (n=cam@clm-laptop.infotech.monash.edu.au)
- # [02:44] * Joins: weinig (n=weinig@c-67-180-35-124.hsd1.ca.comcast.net)
- # [02:49] * Joins: jorlow (n=jorlow@c-67-180-199-19.hsd1.ca.comcast.net)
- # [02:52] * Quits: hdh (n=hdh@hdh-1-pt.tunnel.tserv20.hkg1.ipv6.he.net) (Remote closed the connection)
- # [02:52] <annodomini> Are there any features in HTML5 (other than the different parsing rules) that are currently available only in the HTML syntax and not in the XHTML syntax?
- # [02:55] * Joins: taf2_ (n=taf2@68.49.245.59)
- # [02:59] * Joins: jacobolus (n=jacobolu@64.9.242.183)
- # [03:01] * Quits: ZombieL (n=e@unaffiliated/zombieloffe)
- # [03:10] * Quits: tantek (n=tantek@adsl-63-195-114-133.dsl.snfc21.pacbell.net)
- # [03:12] <MikeSmith> annodomini: lack of support for content from arbitrary namespaces
- # [03:20] <annodomini> MikeSmith: That's a feature missing from the HTML syntax, right? I'm wondering if I'll be missing anything (other than HTML parsing, and IE compatibility) if I use XHTML.
- # [03:22] <MikeSmith> annodomini: lack of support for content from arbitrary namespaces could be considered an actual feature, not a missing feature
- # [03:22] <MikeSmith> annodomini: anyway, no, you're not going to be missing anything
- # [03:22] <annodomini> Ah, now I see.
- # [03:23] <MikeSmith> but a big caveat is that if you are dynamically aggregating content from other places, you need to be very careful to make sure that content doesn't contain any well-formedness errors
- # [03:24] <annodomini> Right, that's why I mentioned the HTML parser.
- # [03:28] <Hixie> annodomini: <noscript>
- # [03:28] <Hixie> annodomini: maybe a few others
- # [03:28] <Hixie> annodomini: <svg> and <mathml> without the namespace stuff
- # [03:30] <annodomini> Hixie: Sure, you're going to have to use namespaces if you're using XML. I'm curious why <noscript> isn't supported?
- # [03:31] <Hixie> <noscript> is implemented as a weird parsing hack
- # [03:31] <Hixie> the parser actually acts differently for script-enabled browsers vs script-disabled
- # [03:31] <heycam> document.write() is another one
- # [03:31] <Hixie> which we can't do in xml
- # [03:31] <Hixie> yep, document.write() doesn't work in xml either
- # [03:32] <annodomini> Hmm. OK, so there's a non-trivial set of functionality that's not available in the XHTML syntax. Good to know.
- # [03:33] * Joins: tantek (n=tantek@adsl-63-195-114-133.dsl.snfc21.pacbell.net)
- # [03:34] <Hixie> nothing good is missing :-)
- # [03:42] <annodomini> I'm basically wondering if I want to use XSLT to generate HTML5, if I should be trying to use XML output or HTML output.
- # [03:42] <annodomini> If it's just parser hacks like <noscript> and document.write(), I can probably work with XML.
- # [03:43] <Hixie> i'd suggest not using xslt, but xml output should be good if you do want to use it (not sure how it works with IE though)
- # [03:44] <annodomini> Why would you recommend against xslt? I'm just looking into it now, so I'm curious what the disadvantages are.
- # [03:45] <Hixie> i'm just biased against it because i get a headache when i try to use it :-)
- # [03:45] <heycam> i heartily recommend xslt!
- # [03:45] <heycam> xslt2, that is. xslt1 is slightly painful.
- # [03:46] <annodomini> (according to my googling, you can cram xhtml through IE if you lie about its mime type and leave off the XML processing instruction because it triggers quirks mode)
- # [03:47] <annodomini> Hixie: Ah, OK. I'll decide on whether it hurts my head once I use it a bit more. At a quick glance it looks fairly reasonable for using as a basic template system; I wouldn't want to use it for much more than that.
- # [03:48] <Hixie> i tried to use it to turn an xml file of numbers into an html file of slightly different numbers and that didn't go well
- # [03:48] <Hixie> that probably soured my opinion of it
- # [03:48] <Hixie> though generally speaking i'm scared of anything that's turing complete using angle brackets in the basic syntax
- # [03:48] <Hixie> e.g. also C++ templates :-)
- # [03:52] <annodomini> heycam: What are the benefits of xslt2?
- # [03:52] <annodomini> And do the browsers support it? I'm using browser xslt support for quickly testing thing out.
- # [03:52] <heycam> annodomini, two that i find useful are being able to define your own functions with <xsl:function>
- # [03:53] <heycam> and being able to tunnel template parameters
- # [03:53] <heycam> so you don't need to explicitly pass them in
- # [03:53] <heycam> i don't think any browsers support xslt2
- # [03:53] <heycam> i just use saxon for offline document processing
- # [03:53] <MikeSmith> lack of document.write() support in XHTML is a feature
- # [03:54] <MikeSmith> support for document.write() in HTML is a bug
- # [03:55] <annodomini> Yeah, I don't mind losing that.
- # [03:57] <MikeSmith> annodomini: fwiw, I use XSLT1 for generating HTML for some things. It is painful in places, but I guess I've learned what the limitations and problems are and how to deal with them.
- # [03:58] <MikeSmith> I've been too lazy to learn XSLT2
- # [03:58] <MikeSmith> when I have time, I plan to learn it by stealing from XSLT2 work that heycam did recently for generating the SVG spec
- # [03:59] <annodomini> I'm looking at it only as a basic template system, and not anything fancy, so I think I'll stick with XSLT1 for now.
- # [04:00] <heycam> annodomini, when you get to the point where you're writing recursive templates to step through strings a character at a time to do some simple string manipulation, that's when you know it's time to upgrade to xslt2
- # [04:00] <MikeSmith> yeah
- # [04:00] <annodomini> Right. Sounds good.
- # [04:00] <MikeSmith> limitations on string manipulation in XSLT1 is a joke
- # [04:01] <MikeSmith> though there are some EXSLT things that most XSLT processors
- # [04:01] <MikeSmith> extensions that make some things easier
- # [04:07] <MikeSmith> Hixie: does the spec define where escaping text spans are allowed?
- # [04:07] <MikeSmith> are they only allowed in CDATA and RCDATA elements?
- # [04:07] <MikeSmith> or are they allowed in normal elements and foreign elements?
- # [04:11] <MikeSmith> hmm, I guess the answer is that anything that might look like any escaping text span in normal elements or foreign elements is actually a comment
- # [04:12] <MikeSmith> and that anything that looks like a comment in CDATA or RCDATA contents is actually an escaping text span
- # [04:15] * Quits: jacobolus (n=jacobolu@64.9.242.183) (Remote closed the connection)
- # [04:19] * Joins: karlcow (n=karl@nerval.la-grange.net)
- # [04:24] <MikeSmith> Hixie: I think it's a mistake to use the term "CDATA elements" in the spec
- # [04:24] <MikeSmith> It would be worthwhile to try to come up with some better term
- # [04:24] <Hixie> file a bug if you come up with a better term :-)
- # [04:24] <Hixie> afk
- # [04:25] <MikeSmith> OK
- # [04:26] <MikeSmith> (problem with the word "CDATA" is that SGML jackasses managed to ensure that there would forever be ambiguity and confusion around it - http://www.flightlab.com/~joe/sgml/cdata.html)
- # [05:12] * Quits: tantek (n=tantek@adsl-63-195-114-133.dsl.snfc21.pacbell.net)
- # [05:24] * Joins: tyoshino_ (n=tyoshino@220.109.219.244)
- # [05:26] * Quits: taf2_ (n=taf2@68.49.245.59)
- # [05:27] * Quits: shikieso1 (n=shiki@220.109.219.244) (Read error: 60 (Operation timed out))
- # [05:36] <Dashiva> Is it just me, or is Shelley sounding kinda like a conspiracy theorist lately...
- # [05:36] * Joins: jacobolus (n=jacobolu@adsl-69-228-190-230.dsl.snfc21.pacbell.net)
- # [05:39] * Quits: hamaji (n=hamaji@220.109.219.244) (Read error: 113 (No route to host))
- # [05:39] * Joins: hamaji (n=hamaji@220.109.219.244)
- # [05:41] * Quits: ukai (n=ukai@220.109.219.244) (Read error: 113 (No route to host))
- # [05:41] * Quits: tyoshino (n=tyoshino@220.109.219.244) (Read error: 113 (No route to host))
- # [05:41] * Joins: shikiesos (n=shiki@220.109.219.244)
- # [05:47] * Joins: dglazkov (n=dglazkov@c-69-181-143-54.hsd1.ca.comcast.net)
- # [05:50] * Joins: ukai (n=ukai@220.109.219.244)
- # [06:00] * Quits: jorlow (n=jorlow@c-67-180-199-19.hsd1.ca.comcast.net)
- # [06:53] * Quits: dolske (n=dolske@c-76-103-40-203.hsd1.ca.comcast.net)
- # [06:59] * Joins: dolske (n=dolske@c-76-103-40-203.hsd1.ca.comcast.net)
- # [07:00] * Quits: weinig (n=weinig@c-67-180-35-124.hsd1.ca.comcast.net)
- # [07:03] * Quits: roc (n=roc@203-97-204-82.dsl.clear.net.nz)
- # [07:05] * Quits: MikeSmith (n=MikeSmit@EM114-48-57-61.pool.e-mobile.ne.jp) ("Tomorrow to fresh woods, and pastures new.")
- # [07:08] * Joins: ap (n=ap@194.154.88.34)
- # [07:10] * Quits: jacobolus (n=jacobolu@adsl-69-228-190-230.dsl.snfc21.pacbell.net) (Remote closed the connection)
- # [07:16] * Joins: zdobersek (n=zan@cpe-92-37-68-112.dynamic.amis.net)
- # [07:19] * Quits: annodomini (n=lambda@wikipedia/lambda)
- # [07:19] * Quits: ap (n=ap@194.154.88.34)
- # [07:20] * Joins: ap (n=ap@194.154.88.34)
- # [07:22] * Joins: MikeSmith (n=MikeSmit@u-211130160118.hotspot.ne.jp)
- # [07:24] * Quits: dglazkov (n=dglazkov@c-69-181-143-54.hsd1.ca.comcast.net)
- # [07:24] * Joins: weinig (n=weinig@c-67-180-35-124.hsd1.ca.comcast.net)
- # [07:35] * Joins: dave_levin (n=dave_lev@72.14.224.1)
- # [07:37] * Quits: weinig (n=weinig@c-67-180-35-124.hsd1.ca.comcast.net)
- # [07:48] * Joins: annevk2 (n=annevk@5ED2D27A.cable.ziggo.nl)
- # [07:48] <annevk2> fwiw, UTF-8+names is not an actual encoding
- # [07:48] <annevk2> why did it end up in the spec?
- # [07:49] <hsivonen> feed autodetection is just a blog post, too :-)
- # [07:50] <hsivonen> but UTF-8+names definitely shoudln't be in any spec
- # [07:50] <hsivonen> having variants of UTF-8 around is evil
- # [07:50] <hsivonen> like CESU-8
- # [07:50] * Joins: roc (n=roc@121.74.138.213)
- # [07:54] <Dashiva> If they're going to move it into unicode, doesn't it make sense to use the unicode names?
- # [07:56] <hsivonen> whose "they" and is UTF-8+names anything but a blog post that went nowhere?
- # [07:56] <Dashiva> Apparently there's an i-d
- # [08:03] * Joins: maikmerten (n=merten@ls5dhcp196.cs.uni-dortmund.de)
- # [08:09] * Joins: gsnedders (n=gsnedder@c83-252-195-71.bredband.comhem.se)
- # [08:13] * Quits: annevk2 (n=annevk@5ED2D27A.cable.ziggo.nl) (Read error: 60 (Operation timed out))
- # [08:13] * Joins: tantek (n=tantek@adsl-63-195-114-133.dsl.snfc21.pacbell.net)
- # [08:19] * Joins: svl (n=me@ip565744a7.direct-adsl.nl)
- # [08:20] * Joins: jacobolus (n=jacobolu@c-98-248-43-68.hsd1.ca.comcast.net)
- # [08:21] * Quits: svl (n=me@ip565744a7.direct-adsl.nl) (Client Quit)
- # [08:24] <Philip`> Dashiva: There's a thing written with xml2rfc and linked from a blog post, but that seems to be about it
- # [08:26] * Joins: harig (n=opera@59.90.71.35)
- # [08:26] * Parts: harig (n=opera@59.90.71.35)
- # [08:34] * Joins: annevk2 (n=annevk@535573A4.cable.casema.nl)
- # [08:35] * Quits: heycam (n=cam@clm-laptop.infotech.monash.edu.au) ("bye")
- # [08:36] * Quits: dave_levin (n=dave_lev@72.14.224.1)
- # [08:39] * Quits: maikmerten (n=merten@ls5dhcp196.cs.uni-dortmund.de) (Remote closed the connection)
- # [08:39] * Joins: heycam (n=cam@clm-laptop.infotech.monash.edu.au)
- # [08:41] * Joins: maikmerten (n=merten@ls5dhcp196.cs.uni-dortmund.de)
- # [08:47] * Quits: annevk2 (n=annevk@535573A4.cable.casema.nl)
- # [08:50] * Joins: Maurice (n=ano@a80-101-46-164.adsl.xs4all.nl)
- # [09:00] * Joins: Mrmil (n=ut_ollie@host-77-236-204-8.blue4.cz)
- # [09:12] * Quits: gsnedders (n=gsnedder@c83-252-195-71.bredband.comhem.se)
- # [09:29] * Joins: cyberpea1 (n=james@fedora/cyberpear)
- # [09:30] * Joins: zdobersek1 (n=zan@cpe-92-37-65-246.dynamic.amis.net)
- # [09:36] * Joins: zdobersek2 (n=zan@cpe-92-37-64-200.dynamic.amis.net)
- # [09:37] * Quits: zdobersek1 (n=zan@cpe-92-37-65-246.dynamic.amis.net) (Read error: 60 (Operation timed out))
- # [09:37] <takkaria> all these debates about data on htmlwg remind me of a line from Stephen Colbert
- # [09:38] * Joins: zdobersek1 (n=zan@cpe-92-37-67-249.dynamic.amis.net)
- # [09:39] <takkaria> "Statistics are just collections of facts, facts about the real world, and as we all know, reality has a liberal bias"
- # [09:40] <MikeSmith> heh
- # [09:40] <MikeSmith> brilliant
- # [09:40] <MikeSmith> takkaria: you quoted that wrong
- # [09:40] <MikeSmith> [[
- # [09:40] <takkaria> I'm sure I did, I couldn't find an accurate quote online
- # [09:40] * Quits: cyberpear (n=james@fedora/cyberpear) (Read error: 110 (Connection timed out))
- # [09:40] <MikeSmith> takkaria: no, not that
- # [09:41] <MikeSmith> [[
- # [09:41] <MikeSmith> Statistics are just collections of "facts", "facts" about the real world, and as we all know, "reality" has a liberal bias.
- # [09:41] <MikeSmith> ]]
- # [09:41] <MikeSmith> never forget your scare quotes, grasshopper
- # [09:41] * Joins: zdobersek3 (n=zan@cpe-92-37-65-76.dynamic.amis.net)
- # [09:41] <MikeSmith> have you learned nothing for your Shaolin training?
- # [09:42] <takkaria> no :'(
- # [09:42] <takkaria> it is pretty bad when people making serious arguments start reminding me of Colbert, though
- # [09:43] * Quits: zdobersek (n=zan@cpe-92-37-68-112.dynamic.amis.net) (Read error: 110 (Connection timed out))
- # [09:44] <MikeSmith> takkaria: well, arguments of partisans from both/all sides remind me of Colbert sometimes
- # [09:44] <MikeSmith> or Beckett
- # [09:44] <MikeSmith> and everybody gets their turn to abuse the scare quotes
- # [09:44] <MikeSmith> that's the great fun of them
- # [09:45] <MikeSmith> scare quotes discriminate against no one
- # [09:45] <MikeSmith> they welcome all of us to use the freely and liberally when they suit or purposes or agenda
- # [09:46] <Philip`> As long as nobody dares use unbalanced quotes, I'm happy
- # [09:48] <Philip`> Hmph, Google retiring Page Creator will break some of my canvas tests :-(
- # [09:48] <MikeSmith> Philip`: that makes me think, I dispense with using scare quotes at all, and just say at the end of each message, (scare quotes implied)
- # [09:48] <MikeSmith> that leaves it up to reader to insert them where necessary
- # [09:48] <MikeSmith> adds some interactivity
- # [09:49] <Philip`> Adds some "interactivity", eh?
- # [09:50] <Philip`> Wow, you're right
- # [09:50] <MikeSmith> you got it, man
- # [09:50] <MikeSmith> we're starting a new trend
- # [09:50] <MikeSmith> #whatwg innovating, as usual (scare quotes implied)
- # [09:51] * takkaria grins
- # [09:51] * takkaria grins (sscare quotes implied)
- # [09:51] <takkaria> yeah, I prefer the latter
- # [09:51] <MikeSmith> heh
- # [09:51] <MikeSmith> on that note, I got to drop off
- # [09:51] <MikeSmith> back later
- # [09:51] <Dashiva> "takkaria" grins
- # [09:51] * Quits: MikeSmith (n=MikeSmit@u-211130160118.hotspot.ne.jp) ("Tomorrow to fresh woods, and pastures new.")
- # [09:53] * Quits: zdobersek2 (n=zan@cpe-92-37-64-200.dynamic.amis.net) (Read error: 110 (Connection timed out))
- # [09:54] * Joins: myakura (n=myakura@p1145-ipbf7105marunouchi.tokyo.ocn.ne.jp)
- # [09:54] * Joins: zdobersek (n=zan@92.37.78.42)
- # [09:55] * Quits: zdobersek1 (n=zan@cpe-92-37-67-249.dynamic.amis.net) (Read error: 110 (Connection timed out))
- # [09:56] * Quits: tantek (n=tantek@adsl-63-195-114-133.dsl.snfc21.pacbell.net)
- # [09:56] * Quits: Lachy (n=Lachlan@85.196.122.246) ("Leaving")
- # [09:56] * Joins: Lachy (n=Lachlan@85.196.122.246)
- # [09:59] * Joins: svl (n=chatzill@a194-109-2-36.dmn.xs4all.nl)
- # [10:01] * Quits: zdobersek3 (n=zan@cpe-92-37-65-76.dynamic.amis.net) (Read error: 60 (Operation timed out))
- # [10:02] * Joins: MikeSmith (n=MikeSmit@EM114-48-11-250.pool.e-mobile.ne.jp)
- # [10:16] * Joins: annevk2 (n=annevk@92.65.193.107)
- # [10:20] * Quits: heycam (n=cam@clm-laptop.infotech.monash.edu.au) ("bye")
- # [10:33] * Quits: shepazu (n=schepers@adsl-150-136-214.rmo.bellsouth.net) ("Core Breach")
- # [10:37] <takkaria> Dashiva: how Zen
- # [10:44] * Joins: ROBOd (n=robod@89.122.216.38)
- # [10:48] * Joins: shepazu (n=schepers@adsl-150-136-214.rmo.bellsouth.net)
- # [10:56] * Joins: tantek (n=tantek@adsl-63-195-114-133.dsl.snfc21.pacbell.net)
- # [10:58] * Joins: cyberpea2 (n=james@fedora/cyberpear)
- # [10:59] <jgraham> BTW I had a look at the mozilla layout table detection code and it is really really simple but hard to implement outside a browser because it requires access to computed style
- # [10:59] <jgraham> I supsect it is not that effective but it would be interesting to test
- # [10:59] <takkaria> heh, Shelley is brilliant
- # [11:01] <Philip`> jgraham: That also makes it hard to execute on a large sample of pages, because you need to have downloaded all the stylesheets too, so you can only really run it live and not on pre-downloaded sets of pages
- # [11:02] <jgraham> Philip`: Yes, I know
- # [11:05] <Philip`> takkaria: Because of some specific quote, or just in general?
- # [11:06] <takkaria> "However, perhaps if there were more people involved in both the effort and the decision process, you wouldn't have to resort to such trickery. Because this form of trickery will not succeed.", amongst others
- # [11:06] <Philip`> s/quote/email or blog post or blog comment or Twittering/
- # [11:06] <takkaria> and the brilliant accusations of Hixie having a bias against experts
- # [11:07] <Hixie> i couldn't work out what she meant by trickery, that came out of left field for me
- # [11:07] <takkaria> I would love to know what Shelley's mental caricature of Hixie is
- # [11:08] <Philip`> takkaria: Ah, okay - I didn't bother reading any emails in that thread that were longer than my screen
- # [11:08] <takkaria> Hixie: I can't figure it out either, maybe she's accusing you of tricking people into thinking you're unbiased or something?
- # [11:09] <takkaria> but it won't succeed, so you shouldn't bother trying
- # [11:09] * Quits: webben (n=benh@dip5-fw.corp.ukl.yahoo.com) ("leaving")
- # [11:09] * Joins: webben (n=benh@dip5-fw.corp.ukl.yahoo.com)
- # [11:10] * Quits: cyberpea1 (n=james@fedora/cyberpear) (Read error: 110 (Connection timed out))
- # [11:11] * Joins: pauld (n=pauld@host86-133-196-121.range86-133.btcentralplus.com)
- # [11:11] * Joins: ZombieLoffe (n=e@unaffiliated/zombieloffe)
- # [11:11] <Dashiva> Sometimes I get curious about what would happen if Hixie said "Okay, whatever" and stopped editing the W3C version. It's a shame we can't spawn a parallel universe to see the outcome without having to suffer it ourselves.
- # [11:12] * Joins: cyberpear (n=james@fedora/cyberpear)
- # [11:14] <takkaria> Nick Levinson is also an eternal source of amusement
- # [11:15] <takkaria> I think he must just not quite grok how the web works
- # [11:16] <Dashiva> Does he realize that downloading a web page to view it is already copying the contents? We're pirates just by using browsers.
- # [11:19] * Joins: Phae (n=phaeness@gateb.thls.bbc.co.uk)
- # [11:20] <takkaria> there's a FOX News clip I saw where there's a congressman talking about his effort to make 2010 the "Year of the Bible". there's a bit in it where he talks about the government "stealing our children's futures" by the way we're dealing with the recession, and I couldn't help thinking "it's alright, they'll find new ones on piratebay"
- # [11:23] * Quits: cyberpea2 (n=james@fedora/cyberpear) (Read error: 110 (Connection timed out))
- # [11:29] * Joins: virtuelv (n=virtuelv@pat-tdc.opera.com)
- # [11:30] <Philip`> Fortunately human population numbers are increasing, so if we steal from future generations then we'll each get a significant benefit and the costs will be much more diluted by the time future generations have to pay for them
- # [11:31] <jgraham> Philip`: Aren't populations only increasing in poor areas of the world?
- # [11:31] <jgraham> s/only/mainly/ I guess
- # [11:31] <Philip`> Also, standards of living increase over time, so we're stealing from the rich (i.e. future people) and giving to the poor (i.e. us)
- # [11:32] <takkaria> http://www.youtube.com/watch?v=F25hYD5vs_4 it's quite a brilliant little clip actually
- # [11:32] <Philip`> jgraham: That's a bonus, since those people will end up paying a larger share of the costs instead of our children
- # [11:32] <takkaria> he tries to sell making 2010 year of the bible as "it's all about freedom, really"
- # [11:33] * Joins: remysharp (n=remyshar@remysharp.plus.com)
- # [11:33] <Dashiva> Freedom to follow rules
- # [11:35] <takkaria> he even lets himself go on a mini-rant about the state taking all of our money and private property at teh end, proper republican stuff
- # [11:35] <takkaria> and presumably our children's futures, too
- # [11:36] <Philip`> Does everyone on American TV always shout, or is it just bad audio balancing in that clip?
- # [11:36] <takkaria> on Fox, they generally shout
- # [11:37] <Dashiva> Is anything Nick wrote in this last post on-topic?
- # [11:38] <remysharp> hi there, regarding the implementation status popup on the spec page - are the implementation statuses maintained, i.e. updated regularly?
- # [11:38] * Quits: roc (n=roc@121.74.138.213)
- # [11:38] <remysharp> I'm specifically looking at DnD - my tests have it working in Firefox 3.5 *only* - and not webkit as the docs suggest
- # [11:40] <Philip`> remysharp: http://developer.apple.com/documentation/appleapplications/conceptual/safarijsprogtopics/Tasks/DragAndDrop.html makes it sound like it's implemented in WebKit
- # [11:43] <remysharp> Looks aren't everything ;-)
- # [11:43] <remysharp> s/look/sound/
- # [11:43] <remysharp> darnit - anyway.
- # [11:44] <remysharp> no - it doesn't work for love nor money when I've tried it.
- # [11:44] * Joins: roc (n=roc@121.74.138.213)
- # [11:44] <Rik`> remysharp: webkit implements DnD pretty much like IE
- # [11:44] <remysharp> and I believe I've tried taking fairly tiny steps
- # [11:45] <Rik`> remysharp: http://www.whatwg.org/demos/2008-sept/dnd/dnd.html
- # [11:48] <remysharp> right, that's perfect - thanks. I need to compare against what and where I got my code from...
- # [11:48] <remysharp> thanks!
- # [12:03] <remysharp> sorry Rik` one last thing - I've just been testing now, and finding that I can't drag a div (which is the source of all my problems!) is that outlined in the specs, and I've just missed it?
- # [12:04] <Rik`> in webkit you'll have to use -webkit-user-drag: element
- # [12:05] <remysharp> good lord. okay, thanks - perfect - that cracked it
- # [12:06] <remysharp> native DnD is suppose to ease the burden of having pure JS solutions, but having to jump though random hoops like that (via webkit) just make it harder :-(
- # [12:06] <Rik`> this is a recent API
- # [12:07] <remysharp> yeah, I guess that's fair enough.
- # [12:07] <remysharp> one more question if you don't mind?
- # [12:08] <remysharp> I couldn't get the DnD to start in Firefox unless I specifically used dataTransfer.setData - is it a requirement that /something/ should be including in the drag
- # [12:08] <remysharp> ?
- # [12:09] <Rik`> i don't know
- # [12:09] <remysharp> I meant more from the spec pov - or same answer? :-)
- # [12:09] <Rik`> the mozilla documentation is pretty complete normally
- # [12:09] <Rik`> I just played with that on friday which is why I have some answers :)
- # [12:10] * Quits: annevk2 (n=annevk@92.65.193.107)
- # [12:14] * Quits: remysharp (n=remyshar@remysharp.plus.com) ("Gotta shoot - "peeyaow"")
- # [12:17] * Quits: Lachy (n=Lachlan@85.196.122.246) ("This computer has gone to sleep")
- # [12:27] * Joins: Lachy (n=Lachlan@pat-tdc.opera.com)
- # [12:27] * ap is now known as ap|away
- # [12:39] * Quits: Lachy (n=Lachlan@pat-tdc.opera.com) ("Leaving")
- # [12:39] * Joins: Lachy (n=Lachlan@pat-tdc.opera.com)
- # [12:39] * Quits: Hish (n=chatzill@mail2.n-e-s.de) (Read error: 104 (Connection reset by peer))
- # [12:44] * Quits: mpt (n=mpt@canonical/launchpad/mpt) (Remote closed the connection)
- # [12:45] * Quits: tantek (n=tantek@adsl-63-195-114-133.dsl.snfc21.pacbell.net)
- # [12:48] * Joins: mpt (n=mpt@canonical/launchpad/mpt)
- # [12:59] * Joins: zdobersek1 (n=zan@cpe-92-37-78-92.dynamic.amis.net)
- # [13:08] * Quits: karlcow (n=karl@nerval.la-grange.net) (Remote closed the connection)
- # [13:08] * Joins: karlcow (n=karl@nerval.la-grange.net)
- # [13:10] <Hixie> ok, content models should be easier to navigate now
- # [13:10] <Hixie> because the category definitions have the list of elements that aply
- # [13:10] <Hixie> apply
- # [13:13] * Quits: zdobersek1 (n=zan@cpe-92-37-78-92.dynamic.amis.net) (Read error: 60 (Operation timed out))
- # [13:15] <Lachy> Hixie, it would be easier to have a table of all elements that indicated which categories each is in
- # [13:15] <Hixie> we'll have that later too
- # [13:15] <Lachy> I had planned to make a table for that
- # [13:15] <Hixie> http://www.whatwg.org/specs/web-apps/current-work/#index
- # [13:15] <Lachy> ok
- # [13:18] * Quits: zdobersek (n=zan@92.37.78.42) (Read error: 110 (Connection timed out))
- # [13:20] * Quits: Amorphous (i=jan@unaffiliated/amorphous) (Read error: 110 (Connection timed out))
- # [13:22] * Joins: heycam (n=cam@124-168-58-6.dyn.iinet.net.au)
- # [13:22] * Joins: Amorphous (i=jan@unaffiliated/amorphous)
- # [13:37] <takkaria> hmm, abarth's sniffing draft doesn't define behaviour for application/octet-stream
- # [13:38] <gavin_> is that ever sniffed?
- # [13:38] <gavin_> I thought that was consistently ignored by all sniffers
- # [13:39] <takkaria> not by Opera, it seems
- # [13:39] <takkaria> and IE's docs say that it tries to determine whether it's text or binary and does stuff based on that
- # [13:41] <takkaria> it certainly seems that it's worth sniffing some application/octet-stream stuff, like video or audio, just so the UA knows what app to hand it off to
- # [13:41] <Lachy> http://html5doctor.com/
- # [13:44] * ap|away is now known as ap
- # [14:04] * Quits: maikmerten (n=merten@ls5dhcp196.cs.uni-dortmund.de) ("Verlassend")
- # [14:26] <Lachy> http://stackoverflow.com/questions/1055214/is-there-a-way-to-make-html5-video-fullscreen
- # [14:27] <Lachy> it's a shame we don't have an API to make full screen videos
- # [14:27] <Lachy> IMHO, the arguments against doing so are bogus, since Flash can already do it
- # [14:29] <Philip`> Do the arguments include the argument that you could take a screenshot of a browser then put it in <video> then make it fullscreen and spoof the browser chrome?
- # [14:29] <Lachy> although, it's entirely possible for browsers to provide their own full screen UI
- # [14:30] <Philip`> At least Flash has a popup message saying to press Escape to leave fullscreen mode, or whatever it says
- # [14:30] * Joins: tndH (n=Rob@86.154.227.105)
- # [14:30] <Philip`> (though irritatingly my keyboard's volume keys also cause it to leave fullscreen mode)
- # [14:30] <Lachy> sure, and there's no reason a browser couldn't do something similar
- # [14:31] <Lachy> a browser would likeliy have to show video controls when the mouse moves anyway, which makes it even more difficult for such an attack to be successful
- # [14:31] <roc> there are clever ways to neutralize that popup message
- # [14:31] <Lachy> roc, how?
- # [14:32] <roc> I can't find the link
- # [14:33] <Lachy> if it is possible, it seems like it would have to be a bug in Flash
- # [14:33] <roc> basically you display a lot of messages on the screen that look just like Flash's one
- # [14:33] <roc> the user is totally confused and doesn't know which one to read
- # [14:33] <roc> then it all goes away and you carry on
- # [14:34] <gavin_> it's a bug in humans ;)
- # [14:34] <roc> I do think we should support fullscreen video
- # [14:35] <roc> but I think it shouldn't be under script control
- # [14:35] <Philip`> You could show some text saying "Close your eyes for five seconds, then open them again for a surprise!", then switch to fullscreen while the user's eyes are closed
- # [14:35] <Lachy> ok, file a bug with The Human Genome Project to get that fixed.
- # [14:36] <roc> alright, sleep time
- # [14:36] * roc is now known as rocSleep
- # [14:36] <Lachy> Philip`, browsers would have to apply techniques similar to popup blockers, so that a script can't enter full screen without the user clicking
- # [14:36] <Lachy> (or some other user interaction)
- # [14:37] <Philip`> You could show some text saying "Close your eyes, click your mouse, wait five seconds, then open your eyes again for a surprise!", then switch to fullscreen while the user's eyes are closed
- # [14:37] <Lachy> browsers should of course provide their own native full screen UI, such as double clicking the video or selecting full screen from the context menu
- # [14:38] <Lachy> but then the user would probably move their mouse, see the video controls and realise that it's a video.
- # [14:39] <Lachy> Also, the attacker would have a hard time mimicking a user's individual desktop to to fool the user that easily
- # [14:39] * Quits: archtech (n=stanv@83.228.56.37)
- # [14:41] <Lachy> basically, until someone can deomonstrate a practical attack vector that isn't simply hypothesising about ludicrous scenarious, I won't believe it's a serious issue
- # [14:42] * Joins: gsnedders (n=gsnedder@pat.se.opera.com)
- # [14:42] * Joins: pmuellr (n=pmuellr@nat/ibm/x-8f97de42ae454db8)
- # [14:45] <Lachy> also, the attack has to have some benefit. Simply being able to show something on a user's screen isn't really useful, unless it could somehow be used to, e.g., steal passwords or gain unauthorised access
- # [14:46] <rocSleep> I sympathize with that point of view. However, there have been several cases of people saying that, and then going ahead and standardizing behaviour which everyone ends up depending on, and then those ludicrous scenarios become reality
- # [14:46] <rocSleep> :visited-based history sniffing for example
- # [14:47] <Philip`> I'd guess a useful fullscreen mode would need to allow pages to supply custom controls, for the same reasons that non-fullscreen <video> allows custom controls
- # [14:47] <Lachy> rocSleep, not doing something because of some unproven hypothetical attack vectors that don't accually allow an attacker to do anything is silly
- # [14:47] <Philip`> in which case you could use the custom controls to interact with the fooled user
- # [14:47] <Lachy> Philip`, no custom controls on full screen video
- # [14:48] <Philip`> Lachy: That makes it much less useful
- # [14:48] <Lachy> that would be a silly and unnecessary thing to allow
- # [14:48] <gavin_> I don't think fullscreen video would need to allow custom controls to be useful
- # [14:48] <Lachy> no it doesn't
- # [14:48] <Philip`> Why would it be more silly and unnecessary than on non-fullscreen videos?
- # [14:48] <Lachy> when a user goes full screen, the browser's native controls should be sufficient
- # [14:48] <gavin_> and I don't think a custom-control-less fullscreen would be "must less useful"
- # [14:49] <gavin_> just like I don't think custom-control-less normal <video> isn't "must less useful"
- # [14:49] <Lachy> because custom controls generally need to integrate with the page design. But in full screen mode, the rest of the page isn't visible, so that
- # [14:49] <Lachy> isn't as necessary
- # [14:50] <Lachy> it would also be much more a much more difficult thing to do, since you'd have to somehow allow the user to overlay HTML content on top of the full screen video, which is outside of the normal browser viewport
- # [14:51] <Lachy> s/the user/the author/
- # [14:52] <gavin_> bah, how did I manage to type "must" intead of "much" twice?
- # [14:53] <Philip`> So if e.g. the BBC iPlayer wants to have a button to enable display of subtitles, and wants to switch from Flash to <video>, it will no longer be able to do that?
- # [14:53] <Philip`> (...in fullscreen mode)
- # [14:53] <Lachy> hmm, I didn't notice. I read "much" in both cases anyway :-)
- # [14:54] <gavin_> perhaps not
- # [14:54] <gavin_> but that doesn't mean that the feature is "much less useful"
- # [14:54] <Lachy> the browser would have to provide UI for enabling subtitles
- # [14:54] <gavin_> only that it doesn't address that specific use case
- # [14:54] <Philip`> Well, "less useful" then :-)
- # [14:54] <Lachy> Philip`, can you send me a link to a iplayer video with subtitles so that I can see how it works?
- # [14:54] <Philip`> Lachy: Not unless you have a UK IP address
- # [14:55] * Lachy connects to a UK based VPN to bypass the regional restrictions
- # [14:55] <Philip`> Ah
- # [14:55] <Philip`> Lachy: Try e.g. http://www.bbc.co.uk/iplayer/episode/b00lkmth/Top_Gear_Series_13_Episode_2_(new_series)/ and then play and wait a few seconds until the actual programme starts
- # [14:56] <Philip`> and then look for the "S" button
- # [15:00] * Joins: zdobersek (n=zan@cpe-92-37-66-63.dynamic.amis.net)
- # [15:02] * Joins: danbri (n=danbri@dhcp149.dagstuhl.de)
- # [15:02] * Quits: Rik|work (n=Rik|work@fw01d.skyrock.net)
- # [15:04] * Joins: Lachy_ (n=Lachlan@london.perfect-privacy.com)
- # [15:09] * Joins: Rik|work (n=Rik|work@fw01d.skyrock.net)
- # [15:13] * Quits: Lachy (n=Lachlan@pat-tdc.opera.com) (Read error: 110 (Connection timed out))
- # [15:21] * Joins: annevk2 (n=annevk@83.85.115.164)
- # [15:22] * Quits: ap (n=ap@194.154.88.34)
- # [15:22] * Quits: danbri (n=danbri@unaffiliated/danbri)
- # [15:27] * Joins: Lachy__ (n=Lachlan@pat-tdc.opera.com)
- # [15:32] * Joins: annodomini (n=lambda@wikipedia/lambda)
- # [15:37] * Quits: annodomini (n=lambda@wikipedia/lambda)
- # [15:38] * Lachy__ is now known as Lachy
- # [15:38] * Quits: Lachy_ (n=Lachlan@london.perfect-privacy.com) (Read error: 110 (Connection timed out))
- # [15:39] * Joins: cyberpea1 (n=james@fedora/cyberpear)
- # [15:40] * Joins: taf2_ (n=taf2@38.99.201.242)
- # [15:43] * Quits: Lachy (n=Lachlan@pat-tdc.opera.com) ("Leaving")
- # [15:43] * Joins: Lachy (n=Lachlan@pat-tdc.opera.com)
- # [15:48] <jgraham> I wonder if it is OK to make new years resolutions in June?
- # [15:49] <jmb> sure. just define a year as the period of time running from one June to the next
- # [15:49] * Joins: aroben (n=aroben@unaffiliated/aroben)
- # [15:49] <Philip`> The start of the year is entirely arbitrary, so it's just as good to pick any other time
- # [15:50] <Philip`> The length of the year isn't so arbitrary, given the planet we live on - you might be considered weird if you intend your resolution to last for a Martian year instead
- # [15:53] <jgraham> The length of the year seems irrelevant since new year's resolutions are defined to start at a time and theoretically last forever but in practice generaly last a few days
- # [15:56] <Philip`> Oh, okay
- # [15:56] <Philip`> But Wikipedia says "While the traditional time to make New Year's resolutions is New Year's Day, it is perfectly acceptable to wait until late January, or even early February to make these goals." so presumably June is unacceptable
- # [15:56] * Quits: cyberpear (n=james@fedora/cyberpear) (Read error: 110 (Connection timed out))
- # [15:58] <jgraham> Oh. well that's no use
- # [15:58] <Phae> you could lump it in with general "spring cleaning" type jobs.
- # [16:00] * gsnedders hopes he has a working computer now
- # [16:01] <Philip`> gsnedders: Your presence here suggests that you do
- # [16:04] <Philip`> "Thank you for your past interest in VMware. As part of our routine scheduled maintenance, we will be removing email addresses and associated subscription information from our marketing database for contacts who have not updated their profile and/or subscription preferences within the last 6 months."
- # [16:04] <Philip`> That's probably the first time I've ever seen a company intentionally reduce the reach of their email marketing
- # [16:12] * Joins: danbri (n=danbri@dhcp149.dagstuhl.de)
- # [16:13] * Quits: nessy (n=nessy@124-168-251-103.dyn.iinet.net.au) ("This computer has gone to sleep")
- # [16:14] * Joins: nessy (n=nessy@124-168-251-103.dyn.iinet.net.au)
- # [16:14] <Lachy> jgraham, the financial year ends at the end of June in some countries, so you could do a New Financial Year's Resolution
- # [16:14] * Quits: nessy (n=nessy@124-168-251-103.dyn.iinet.net.au) (Remote closed the connection)
- # [16:18] * Joins: jmb^ (n=jmb@login.ecs.soton.ac.uk)
- # [16:19] * Joins: maikmerten (n=merten@vp-c-67.cs.uni-dortmund.de)
- # [16:20] * Quits: jmb (n=jmb@login.ecs.soton.ac.uk) (Read error: 110 (Connection timed out))
- # [16:20] * Quits: MikeSmith (n=MikeSmit@EM114-48-11-250.pool.e-mobile.ne.jp) (Read error: 110 (Connection timed out))
- # [16:21] * Joins: weinig (n=weinig@c-67-180-35-124.hsd1.ca.comcast.net)
- # [16:21] * jmb^ is now known as jmb
- # [16:21] * Quits: weinig (n=weinig@c-67-180-35-124.hsd1.ca.comcast.net) (Read error: 104 (Connection reset by peer))
- # [16:21] * Joins: weinig (n=weinig@c-67-180-35-124.hsd1.ca.comcast.net)
- # [16:26] <gsnedders> What's a sane PDF reader for Ubuntu?
- # [16:27] <hsivonen> the bundled one
- # [16:28] <Philip`> cat and the PDF file format specification
- # [16:31] <Lachy> LOL
- # [16:33] <jgraham> It seems like the PDF specification is only avaliable as a PDF, which presents a bootstrapping problem
- # [16:33] <Philip`> That's not a problem - you can use Acrobat Reader to print the PDF specification
- # [16:34] * Joins: dglazkov (n=dglazkov@c-69-181-143-54.hsd1.ca.comcast.net)
- # [16:34] <Philip`> then use the printed spec and cat as a more sane PDF reader
- # [16:35] <gsnedders> jgraham: That's no different to HTML, but I think PDF claims to be logical.
- # [16:36] <Lachy> gsnedders, at least with HTML specifications, you can read HTML 2.0, implement that, which then allows you to read HTML 3.2 and 4.01, and after implementing those, you can implement HTML5, throw out your previous work on HTML 2.0, 3.2 and 4.01, and you're all done
- # [16:37] <gsnedders> Lachy: Or you could just download the PDF version of HTML 5 to start with
- # [16:38] <Lachy> but you first need to find the link to the PDF version, which is linked from the HTML version
- # [16:38] <gsnedders> Lachy: HTML can be read well enough as a text format though
- # [16:38] <Philip`> Someone should print out a copy of the HTML5 spec, then drop it in a trench on a building site, so that a future generation's equivalent of Time Team will dig it up even if there's been an intervening collapse of civilisation and loss of digital information
- # [16:39] <gsnedders> Under the rubble, preferably.
- # [16:39] <Lachy> you're assuming the person attempting to read the specification is familiar with some form of HTML, and won't be frightened away by all that markup
- # [16:42] <Dashiva> Also, if they don't have a browser, they probably don't have internet access to the PDF version either
- # [16:44] <Lachy> Dashiva, wget
- # [16:44] <Philip`> Dashiva: gopher
- # [16:44] <Lachy> Philip`, is the spec even available through the gopher protocol?
- # [16:45] <Dashiva> Lachy: If they can wget the spec, they can wget a browser
- # [16:45] <Philip`> Lachy: Don't know; if not, someone should set up a server for it
- # [16:46] <hsivonen> wonders if Workers address the threading use cases of Google Native Client
- # [16:46] <Lachy> Philip`, it depends if someone can create some market demand for such a service
- # [16:46] * hsivonen wonders...
- # [16:47] <Philip`> Lachy: It doesn't need any market demand, it just needs someone sufficiently bored and strange to set it up
- # [16:48] <Lachy> Philip`, someone like you, perhaps?
- # [16:48] <Philip`> Dashiva: Maybe they're using an IP-over-SMS protocol and will be bankrupt before they manage to download an entire browser, but would just about be able to download the HTML5 spec
- # [16:48] <Philip`> Lachy: I'm not sufficiently bored at the moment
- # [16:50] <Lachy> I wonder if anyone has implemented an IP-over-SMS service yet?
- # [16:52] <Lachy> I wonder if anyone would be bored enough to manually type in the bytes on their phone to construct an IP packet that is itself divided into 160 byte sequences?
- # [16:53] <Philip`> You could build a device out of Lego to press the buttons for you automatically
- # [16:53] <Dashiva> Surely you'd want to use base64 instead :)
- # [16:53] * Quits: virtuelv (n=virtuelv@pat-tdc.opera.com) (Read error: 104 (Connection reset by peer))
- # [16:55] <Philip`> Depends on whether you want to optimise bandwidth usage or data entry time
- # [16:56] <Lachy> Philip`, given the limitations of character entry on common mobile devices, encoding the entire packet as base64 might be the only viable option
- # [16:56] <Philip`> e.g. it'd probably be much quicker to turn off predictive typing and use base10, but you'd have to send more messages
- # [16:57] <Lachy> well, not the only viable option, since you could conceivably create your own encoding scheme
- # [16:57] <Philip`> I don't think my phone's dictionary contains SGVsbG8gd29ybGQ= so it would complain a lot when I tried to type it in
- # [16:57] <Lachy> but that would be a silly waste of time (unlike the rest of this project)
- # [16:58] <Dashiva> How about just sending one SMS per byte, having the message contain the byte value as a number 0-255?
- # [16:58] <Dashiva> Then you could have jumbo SMSes where you type more than one number in each SMS, separated by non-digits
- # [16:59] <Philip`> SMSs are a bit unreliable so you'd probably have to add sequence numbers and ACKs
- # [16:59] <Philip`> else the higher layers aren't going to cope very well with all the packet corruption
- # [17:01] <Lachy> Dashiva, once you have IP, you can then use TCP for that
- # [17:01] * Quits: dglazkov (n=dglazkov@c-69-181-143-54.hsd1.ca.comcast.net)
- # [17:01] <Philip`> You need each SMS to contain the octet, the sequence number, and a SHA512 HMAC to ensure integrity, and then it'll be fine
- # [17:02] <Dashiva> And the ACK?
- # [17:02] * Joins: archtech (n=stanv@83.228.56.37)
- # [17:03] <Lachy> you could get away with just the sequence number. Adding all the rest would take up too much of each message's 160 byte sequence
- # [17:04] <Philip`> Lachy: SMS probably loses more than one message out of every ~1500, and TCP won't recover from 100% packet loss, so you need some layer 2 error-correction mechanism
- # [17:05] <Lachy> if you get a 100% packet loss, then the connection would simply timeout and you could assume transmission failure
- # [17:05] <Philip`> Also, TCP won't help at all with UDP applications like real-time games and Skype
- # [17:05] <Lachy> SMS-over-IP isn't designed for real time applications
- # [17:06] * Parts: Mrmil (n=ut_ollie@host-77-236-204-8.blue4.cz)
- # [17:06] <Philip`> Lachy: If transmission failure is pretty much guaranteed, it's a bit of a rubbish protocol
- # [17:06] <Dashiva> You could do a regular call and use voice software to read out the bytes
- # [17:06] <Lachy> oops, IP over SMS
- # [17:06] <Philip`> so we ought to design it to work around those problems
- # [17:06] * Quits: Maurice (n=ano@a80-101-46-164.adsl.xs4all.nl) ("Disconnected...")
- # [17:07] <Philip`> Dashiva: Good thinking - the voice channel should be much more reliable than SMS
- # [17:07] <Lachy> I'd love it if someone wrote an April Fools RFC for this protocol :-)
- # [17:07] * Joins: billmason (n=billmaso@ip154.unival.com)
- # [17:08] <takkaria> mm, Skype-over-IP-over-SMS would be beautiful
- # [17:09] <Lachy> hey, you could use Twitter for multicasting
- # [17:09] <gavin_> I imagine most SMS traffic ends up being sent over an IP network at some point
- # [17:09] <Lachy> but then you'd be restricted to 140 character sequences
- # [17:09] <gavin_> so really it's IP-over-SMS-over-IP
- # [17:11] * Quits: myakura (n=myakura@p1145-ipbf7105marunouchi.tokyo.ocn.ne.jp) ("Leaving...")
- # [17:11] <Philip`> Mmm, IP-over-Twitter
- # [17:12] <gsnedders> takkaria: You're upstairs, right?
- # [17:12] <takkaria> gsnedders: yeah
- # [17:13] <hsivonen> We should move TPAC to Sweden
- # [17:13] <Lachy> IP-over-twitter would be the same as IP-over-SMS, except that the IP-over-twitter protocol would need to specify a way to detect and strip the username from the beginning of each message
- # [17:14] <Philip`> Presumably IP-over-SMS and IP-over-Twitter would be able to interoperate, using Twitter's SMS features
- # [17:14] * hsivonen notes that twitter applies the length limit by UTF-16 (or characters?) while SMS applies it by bytes
- # [17:15] <hsivonen> so twitter gives you 140 characters even if you use stuff that's not in the original GSM encoding and turns SMS into an 80-code unit UTF-16 message
- # [17:15] * hsivonen wonders how twitter's gateway deals with that
- # [17:16] <Dashiva> Splits into two messages? :)
- # [17:16] * Quits: weinig (n=weinig@c-67-180-35-124.hsd1.ca.comcast.net)
- # [17:16] * Quits: annevk2 (n=annevk@83.85.115.164)
- # [17:18] <Philip`> Truncates? Crashes?
- # [17:19] <Lachy> someone should test it. I don't have Twitter sending me SMSs, so i can't
- # [17:20] <Philip`> Apparently SMS does "160 7-bit characters, 140 8-bit characters, or 70 16-bit characters"
- # [17:21] <Philip`> (since it's 140 octets)
- # [17:21] <Philip`> (according to http://en.wikipedia.org/wiki/SMS#Message_size)
- # [17:22] <gsnedders> hsivonen: Twitter applies NFC, then limits by number of codepoints used.
- # [17:22] * Joins: annevk2 (n=annevk@535573A4.cable.casema.nl)
- # [17:22] <gsnedders> (or at least did a few months ago when I got bored and tried to work it out)
- # [17:24] <Lachy> well, if we're encoding the message as Base64, then normalising to NFC won't actually affect the message anyway
- # [17:24] * Joins: arun_ (n=arun@cpe-68-173-35-78.nyc.res.rr.com)
- # [17:25] <Philip`> The SMS spec says it's "UCS-2" and references ISO/IEC 10646
- # [17:25] * Joins: tantek (n=tantek@adsl-63-195-114-133.dsl.snfc21.pacbell.net)
- # [17:26] <hsivonen> Philip`: what's the 8-bit encoding?
- # [17:26] <takkaria> can someone with IE8 go to http://takkaria.org/tests/octet-stream/ and see if hitting 'Start' does anything?
- # [17:26] <Philip`> and doesn't give any details on e.g. how surrogates should be handled
- # [17:26] <hsivonen> the 160 char thing works with accented Western European chars
- # [17:26] <Philip`> hsivonen: "8 bit data is user defined"
- # [17:27] <hsivonen> ah.
- # [17:27] <Philip`> hsivonen: "Character table: User Specific"
- # [17:27] <hsivonen> Russian goes the 16 bit path
- # [17:27] <takkaria> yeah, it kinda relies on different phones using the same charsets
- # [17:27] <hsivonen> which is why it pays off to use ad hoc Romanizations of Russian
- # [17:27] <Lachy> I would assume the defacto standard would probably be Windows-1252
- # [17:27] <Lachy> at least in Western markets
- # [17:27] <hsivonen> Lachy: I wouldn't assume so
- # [17:28] <Philip`> The table in http://en.wikipedia.org/wiki/SMS#GSM shows the 7-bit characters
- # [17:28] <hsivonen> GSM had a crazy legacy of its own
- # [17:28] <Philip`> 8-bit sounds like it's for data (e.g. sending photos), not text
- # [17:28] * Joins: cyberpear (n=james@fedora/cyberpear)
- # [17:29] <Philip`> They've got a whole two bits for identifying the character encoding, so they could expand it with an extra choice in the future
- # [17:29] <Lachy> wow, I always thought SMS just used ASCII
- # [17:30] <hsivonen> Lachy: it was invented by people who had accented use cases
- # [17:31] <Philip`> The 7-bit thing also has escape codes to shift into a different character set, with three currently defined (Turkish, Spanish, Portugese)
- # [17:33] * Joins: annodomini (n=lambda@wikipedia/lambda)
- # [17:35] * Quits: cyberpear (n=james@fedora/cyberpear) (Operation timed out)
- # [17:35] <Philip`> If you use the single shift mechanism, then sending 0x1B 0x?? will look up the second byte in the character table for the message's language (defined in a header field)
- # [17:36] <Philip`> or there's another mode which uses a language-specific table by default, and I think uses 0x1B to escape to the standard 7-bit table
- # [17:37] <Philip`> This system does sound somewhat evolved, i.e. an accumulation of hacks
- # [17:39] * Quits: Dashiva (i=Dashiva@m223j.studby.ntnu.no)
- # [17:39] * Joins: cyberpear (n=james@fedora/cyberpear)
- # [17:39] * Quits: Lachy (n=Lachlan@pat-tdc.opera.com) ("This computer has gone to sleep")
- # [17:41] * Quits: cyberpea1 (n=james@fedora/cyberpear) (Connection timed out)
- # [17:42] * Philip` was looking at http://www.3gpp.org/ftp/Specs/archive/23_series/23.038/23038-820.zip via ftp://ftp.3gpp.org/Specs/html-info/23038.htm via http://en.wikipedia.org/wiki/SMS#cite_note-3GPP_23.038-25
- # [17:46] <hsivonen> hmm. I guess if Google Native Client takes off, it'll further cement the position of x86 instructions as the universal byte code that future CPUs need to implement
- # [17:47] * hsivonen wonders how GCC & Native Client compares to javac & HotSpot
- # [17:48] * Joins: dglazkov (n=dglazkov@nat/google/x-6eeaf0058876cab4)
- # [17:48] * Joins: bgalbraith (n=bgalbrai@nat/mozilla/x-cb094580bf8f571f)
- # [17:49] <Philip`> It's particularly bad if people use NaCl to do computation-intensive work (which is its main purpose) using SSE/SSE2/etc
- # [17:49] <Philip`> because they're probably not much fun to emulate efficiently on a different CPU architecture
- # [17:51] <hsivonen> are there autovectorizers for any non-vector languages and do they work?
- # [17:51] * Quits: maikmerten (n=merten@vp-c-67.cs.uni-dortmund.de) (Remote closed the connection)
- # [17:52] <Philip`> The Intel C++ Compiler can autovectorise loops
- # [17:52] <Philip`> and sometimes it works
- # [17:52] * Quits: svl (n=chatzill@a194-109-2-36.dmn.xs4all.nl) ("And back he spurred like a madman, shrieking a curse to the sky.")
- # [17:53] * Quits: annevk2 (n=annevk@535573A4.cable.casema.nl)
- # [17:53] <Philip`> http://software.intel.com/en-us/articles/vectorization-with-the-intel-compilers-part-i/
- # [17:54] * Joins: taf2__ (n=taf2@38.99.201.242)
- # [17:55] * Joins: Midler (n=midler@94.191.173.33.bredband.tre.se)
- # [17:56] <Philip`> (As far as I'm aware, anyone who actually cares about vectorisation in C++ will use explicit SIMD intrinsics (or write assembly code) because compilers are rarely magic enough)
- # [17:58] <hsivonen> ok
- # [18:00] <Philip`> Hmm, apparently GCC has -ftree-vectorize
- # [18:01] <Philip`> which apparently works for C and C++ and Fortran and Java etc
- # [18:03] <Philip`> (to a limited, buggy extent)
- # [18:04] * Joins: Dashiva (i=Dashiva@m223j.studby.ntnu.no)
- # [18:05] * Quits: taf2_ (n=taf2@38.99.201.242) (Read error: 110 (Connection timed out))
- # [18:06] <takkaria> gsnedders: I'm no longer upstairs
- # [18:09] * Quits: danbri (n=danbri@unaffiliated/danbri)
- # [18:11] * Quits: gsnedders (n=gsnedder@pat.se.opera.com) ("Adios Intarwebs!")
- # [18:14] * Quits: jgraham (n=jgraham@web22.webfaction.com) (Read error: 104 (Connection reset by peer))
- # [18:15] * Quits: webben (n=benh@dip5-fw.corp.ukl.yahoo.com) ("leaving")
- # [18:16] * Joins: webben (n=benh@dip5-fw.corp.ukl.yahoo.com)
- # [18:19] * Joins: mstange (n=markus@aixd3.rhrk.uni-kl.de)
- # [18:25] * Quits: Phae (n=phaeness@gateb.thls.bbc.co.uk)
- # [18:26] * Joins: weinig (n=weinig@17.246.16.228)
- # [18:26] * Joins: Maurice (i=copyman@5ED548D4.cable.ziggo.nl)
- # [18:27] * Quits: cyberpear (n=james@fedora/cyberpear) (Read error: 60 (Operation timed out))
- # [18:30] * Joins: weinig_ (n=weinig@17.203.15.205)
- # [18:35] * Quits: tantek (n=tantek@adsl-63-195-114-133.dsl.snfc21.pacbell.net) (Read error: 104 (Connection reset by peer))
- # [18:35] * Joins: tantek (n=tantek@adsl-63-195-114-133.dsl.snfc21.pacbell.net)
- # [18:36] * Quits: tndH (n=Rob@86.154.227.105) (Read error: 110 (Connection timed out))
- # [18:45] * Quits: pauld (n=pauld@host86-133-196-121.range86-133.btcentralplus.com) ("Gone for a burton")
- # [18:46] * Joins: gsnedders (n=gsnedder@c83-252-197-166.bredband.comhem.se)
- # [18:47] * Joins: cyberpea2 (n=james@fedora/cyberpear)
- # [18:48] * Quits: weinig (n=weinig@17.246.16.228) (Read error: 110 (Connection timed out))
- # [18:52] * Quits: cyberpea2 (n=james@fedora/cyberpear) (Remote closed the connection)
- # [18:53] * Quits: arun_ (n=arun@cpe-68-173-35-78.nyc.res.rr.com) ("Ruh-roh!")
- # [18:55] * Joins: sbublava (n=stephan@77.118.172.75.wireless.dyn.drei.com)
- # [18:57] * Quits: annodomini (n=lambda@wikipedia/lambda) (Read error: 104 (Connection reset by peer))
- # [18:57] * Joins: annodomini_ (n=lambda@wikipedia/lambda)
- # [18:57] * Joins: cyberpea2 (n=james@fedora/cyberpear)
- # [19:02] * Joins: jennb (n=jennb@72.14.227.1)
- # [19:03] * Quits: cyberpea2 (n=james@fedora/cyberpear) (Read error: 60 (Operation timed out))
- # [19:04] * Joins: dave_levin (n=dave_lev@72.14.227.1)
- # [19:10] * Joins: svl (n=me@ip565744a7.direct-adsl.nl)
- # [19:18] * Joins: cyberpea2 (n=james@fedora/cyberpear)
- # [19:18] * Joins: Lachy (n=Lachlan@85.196.122.246)
- # [19:29] * Joins: danbri (n=danbri@dhcp149.dagstuhl.de)
- # [19:30] * Joins: remysharp (n=remyshar@remysharp.plus.com)
- # [19:31] * Quits: dolske (n=dolske@c-76-103-40-203.hsd1.ca.comcast.net)
- # [19:36] * Quits: cyberpea2 (n=james@fedora/cyberpear) (Connection timed out)
- # [19:40] * Quits: danbri (n=danbri@unaffiliated/danbri)
- # [19:44] * Quits: dave_levin (n=dave_lev@72.14.227.1)
- # [19:44] * Quits: weinig_ (n=weinig@17.203.15.205) (Read error: 104 (Connection reset by peer))
- # [19:47] * Joins: dave_levin (n=dave_lev@72.14.227.1)
- # [19:53] * Joins: maikmerten (n=maikmert@Z9e79.z.pppool.de)
- # [19:58] * Joins: dolske (n=dolske@nat/mozilla/x-b36ca457bd62d0f4)
- # [19:59] * Joins: cyberpear (n=james@fedora/cyberpear)
- # [20:01] * Joins: bgalbraith_ (n=bgalbrai@nat/mozilla/x-e946ceb8de98c5ca)
- # [20:01] * Quits: bgalbraith (n=bgalbrai@nat/mozilla/x-cb094580bf8f571f) (Read error: 104 (Connection reset by peer))
- # [20:03] * Joins: ojan (n=ojan@nat/google/x-9686375566ceb2fc)
- # [20:07] * Joins: Phae (n=phaeness@cpc2-acto6-0-0-cust747.brnt.cable.ntl.com)
- # [20:10] * maikmerten is now known as maik|eat
- # [20:14] * Joins: jgraham (n=jgraham@web22.webfaction.com)
- # [20:23] * Joins: Jswalden86 (n=waldo@nat/mozilla/x-74abf7f1a09441c1)
- # [20:27] * Quits: gsnedders (n=gsnedder@c83-252-197-166.bredband.comhem.se)
- # [20:27] * Joins: cying (n=cying@70.90.171.153)
- # [20:30] * Joins: weinig (n=weinig@nat/apple/x-01d8833472f092b1)
- # [20:42] <gavin_> Lachy: http://www.bunnyhero.org/2008/05/10/scaring-people-with-fullscreen/ fwiw
- # [20:48] * maik|eat is now known as maikmerten
- # [20:49] * Parts: ojan (n=ojan@nat/google/x-9686375566ceb2fc)
- # [20:55] <Lachy> gavin_, as I expected, a totally ineffective and useless attack
- # [20:56] <Lachy> since the "Press ESC..." message appears right in the middle of the screen, it's the most visible
- # [20:56] <gavin_> I couldn't find it the first two times I tried
- # [20:56] <gavin_> (was looking at the center)
- # [20:56] <Lachy> granted, the demo isn't the best since it uses jibberish for the other text, making them easy to ignore
- # [20:57] <Lachy> I don't know how it's possible to miss, it's so obvious
- # [20:57] * Quits: weinig (n=weinig@nat/apple/x-01d8833472f092b1)
- # [20:58] <Lachy> also, the way I suggested implementing full screen mode in browsers would make that attack virtually ineffective, since pressing any keys would reveal the browser's native video controls
- # [21:01] * Joins: ap (n=ap@194.154.88.34)
- # [21:05] * Joins: cying_ (n=cying@70.90.171.153)
- # [21:16] * Joins: yshin (n=yshin@72.14.227.1)
- # [21:23] * Quits: cying (n=cying@70.90.171.153) (Read error: 110 (Connection timed out))
- # [21:23] * cying_ is now known as cying
- # [21:23] <Midler> gavin_: i couldnt find it either
- # [21:23] <Midler> however
- # [21:24] <Midler> making a random white & black dots (lika ants war on tv)
- # [21:24] <Midler> might make it harder to find
- # [21:27] <Midler> does html5 support webforms 2.0?
- # [21:28] <Lachy> wow, Midler, I really don't get how it's even remotely difficult to see. It might be harder if the text boxes contained readable information
- # [21:28] <Midler> not me asking
- # [21:28] <Midler> I dont even know what it is
- # [21:28] <Lachy> Midler, webforms 2 stuff was integrated into HTML5 already
- # [21:28] <Midler> ok
- # [21:29] <Lachy> basically, it's all the form controls and attributes, like <input>, <select>, <textarea>, and their associated APIs
- # [21:29] <Midler> aha
- # [21:29] <Midler> well, iam a kind of a newbie =)
- # [21:30] <Midler> therefore i hang around in here, hope to learn something
- # [21:30] <Midler> =)
- # [21:42] * Quits: pmuellr (n=pmuellr@nat/ibm/x-8f97de42ae454db8)
- # [21:42] * Joins: drostie (n=hopkins@5ED17066.cable.ziggo.nl)
- # [21:44] <Midler> Lachy: was it webforms that was shown in the google tech talk video?
- # [21:50] * Joins: gsnedders (n=gsnedder@c83-252-197-166.bredband.comhem.se)
- # [21:50] * Quits: taf2__ (n=taf2@38.99.201.242)
- # [21:51] <Lachy> Midler, if it was about new form controls in HTML5, yes
- # [21:51] <Midler> ok
- # [21:53] * Quits: bgalbraith_ (n=bgalbrai@nat/mozilla/x-e946ceb8de98c5ca)
- # [21:56] * Joins: bgalbraith (n=bgalbrai@nat/mozilla/x-4a900237cc265b46)
- # [22:02] * Quits: maikmerten (n=maikmert@Z9e79.z.pppool.de) (Remote closed the connection)
- # [22:09] * Quits: yshin (n=yshin@72.14.227.1)
- # [22:16] * Quits: Lachy (n=Lachlan@85.196.122.246) ("This computer has gone to sleep")
- # [22:17] * Joins: weinig (n=weinig@nat/apple/x-6dc9a4e0c95b0889)
- # [22:17] * Joins: dolske_ (n=dolske@nat/mozilla/x-48ed3244887683c7)
- # [22:18] * Quits: cying (n=cying@70.90.171.153)
- # [22:18] * Quits: dolske (n=dolske@nat/mozilla/x-b36ca457bd62d0f4) (Read error: 104 (Connection reset by peer))
- # [22:18] * Joins: ojan (n=ojan@nat/google/x-b6066e9f6ffacd2d)
- # [22:18] * Quits: weinig (n=weinig@nat/apple/x-6dc9a4e0c95b0889) (Client Quit)
- # [22:22] * Joins: cying (n=cying@70.90.171.153)
- # [22:26] * Joins: weinig (n=weinig@nat/apple/x-1852b97c1eeeebc2)
- # [22:27] * Quits: rocSleep (n=roc@121.74.138.213)
- # [22:27] * Quits: dolske_ (n=dolske@nat/mozilla/x-48ed3244887683c7) (Read error: 104 (Connection reset by peer))
- # [22:28] * Joins: dolske (n=dolske@nat/mozilla/x-3c8772d10e5383e7)
- # [22:37] * Joins: ojan_ (n=ojan@nat/google/x-6306fd8f8bafc927)
- # [22:38] * Quits: remysharp (n=remyshar@remysharp.plus.com) ("Gotta shoot - "peeyaow"")
- # [22:39] * Joins: kangax (n=kangax@157.130.31.226)
- # [22:40] * Quits: ojan (n=ojan@nat/google/x-b6066e9f6ffacd2d) (Read error: 60 (Operation timed out))
- # [22:45] * Quits: sbublava (n=stephan@77.118.172.75.wireless.dyn.drei.com)
- # [22:48] * Joins: Lachy (n=Lachlan@85.196.122.246)
- # [22:48] * Quits: Lachy (n=Lachlan@85.196.122.246) (Client Quit)
- # [22:49] * Joins: Lachy (n=Lachlan@85.196.122.246)
- # [22:49] * Quits: Lachy (n=Lachlan@85.196.122.246) (Client Quit)
- # [22:49] * Joins: Lachy (n=Lachlan@85.196.122.246)
- # [22:51] * Quits: bgalbraith (n=bgalbrai@nat/mozilla/x-4a900237cc265b46)
- # [22:51] * Quits: ROBOd (n=robod@89.122.216.38) ("http://www.robodesign.ro")
- # [22:54] * Joins: bgalbraith (n=bgalbrai@nat/mozilla/x-c47a1b6983622e91)
- # [22:55] * Quits: dolske (n=dolske@nat/mozilla/x-3c8772d10e5383e7)
- # [22:56] <takkaria> so, Firefox doesn't sniff application/octet-stream, Safari 4 apparently only sniffs application/octet-stream files which look like HTML, and Opera sniffs lots of thigs sent as application/octet-stream depending on file extension
- # [22:57] * Joins: dolske (n=dolske@nat/mozilla/x-031275505d3a1a9b)
- # [23:06] <takkaria> I'm pretty sure sniffing application/octet-stream as text/html is a potential security hole, actually
- # [23:06] * Joins: ttepasse (n=ttepasse@p5B0133D1.dip.t-dialin.net)
- # [23:06] * Joins: roc (n=roc@203-97-204-82.dsl.clear.net.nz)
- # [23:07] <Lachy> what does IE do?
- # [23:07] <takkaria> no idea, I've not got any way to run it ATM
- # [23:07] <Lachy> ok
- # [23:07] * Quits: svl (n=me@ip565744a7.direct-adsl.nl) ("And back he spurred like a madman, shrieking a curse to the sky.")
- # [23:07] <takkaria> you could run http://takkaria.org/tests/octet-stream/ if you wanted
- # [23:08] <takkaria> I assume it will work on IE, I don't think I use anything it can't handle
- # [23:11] * Quits: zdobersek (n=zan@cpe-92-37-66-63.dynamic.amis.net) ("Leaving.")
- # [23:12] * Quits: kangax (n=kangax@157.130.31.226)
- # [23:18] * Quits: mstange (n=markus@aixd3.rhrk.uni-kl.de) ("ChatZilla 0.9.85 [Firefox 3.6a1pre/20090629042158]")
- # [23:27] * Quits: Maurice (i=copyman@5ED548D4.cable.ziggo.nl) ("Disconnected...")
- # [23:39] * Quits: Jswalden86 (n=waldo@nat/mozilla/x-74abf7f1a09441c1) ("ChatZilla 0.9.85 [Firefox 3.6a1pre/20090629044912]")
- # [23:41] * Joins: weinig_ (n=weinig@17.246.18.189)
- # [23:42] * Quits: weinig (n=weinig@nat/apple/x-1852b97c1eeeebc2) (Read error: 60 (Operation timed out))
- # [23:43] * Parts: ojan_ (n=ojan@nat/google/x-6306fd8f8bafc927)
- # [23:43] * weinig_ is now known as weinig|meeting
- # [23:43] * Quits: ap (n=ap@194.154.88.34)
- # [23:45] * Quits: weinig|meeting (n=weinig@17.246.18.189) (Client Quit)
- # [23:51] * Quits: tantek (n=tantek@adsl-63-195-114-133.dsl.snfc21.pacbell.net)
- # [23:52] * Joins: tndH (n=Rob@james-baillie-pc083-229.student-halls.leeds.ac.uk)
- # [23:58] * Quits: Phae (n=phaeness@cpc2-acto6-0-0-cust747.brnt.cable.ntl.com)
- # Session Close: Tue Jun 30 00:00:00 2009
The end :)