/irc-logs / w3c / #html-wg / 2007-06-28 / end

Options:

  1. # Session Start: Thu Jun 28 00:00:00 2007
  2. # Session Ident: #html-wg
  3. # [00:01] * Quits: kingryan (rking3@71.202.121.218) (Quit: kingryan)
  4. # [00:07] * Quits: heycam (cam@124.168.130.154) (Ping timeout)
  5. # [00:13] * Quits: gavin (gavin@74.103.208.221) (Ping timeout)
  6. # [00:18] * Joins: gavin (gavin@74.103.208.221)
  7. # [00:19] * Joins: kingryan_ (rking3@64.81.240.149)
  8. # [00:20] * Quits: kingryan_ (rking3@64.81.240.149) (Quit: kingryan_)
  9. # [00:21] * Joins: kingryan (rking3@64.81.240.149)
  10. # [00:24] * Parts: hasather (hasather@80.203.71.22)
  11. # [00:26] * Quits: myakura (myakura@58.88.37.26) (Quit: Leaving...)
  12. # [00:26] <Philip`> Hmph, someone keeps reading the public-html archives through the same caching proxy as me, so it keeps giving me old cached versions with no "next message" link :-(
  13. # [00:34] * Quits: tH (Rob@87.102.84.66) (Quit: ChatZilla 0.9.78.1-rdmsoft [XULRunner 1.8.0.9/2006120508])
  14. # [00:38] * Joins: heycam (cam@130.194.72.84)
  15. # [00:39] * Quits: DanC (connolly@128.30.52.30) (Client exited)
  16. # [00:42] * Joins: DanC (connolly@128.30.52.30)
  17. # [00:43] * Joins: mjs (mjs@17.255.104.100)
  18. # [00:43] * Joins: polin8 (polin8@209.176.7.3)
  19. # [01:19] * Parts: billmason (billmason@69.30.57.156)
  20. # [01:23] * Quits: polin8 (polin8@209.176.7.3) (Quit: polin8)
  21. # [01:29] * Joins: sbuluf (xge@200.49.140.231)
  22. # [01:52] * Quits: mjs (mjs@17.255.104.100) (Quit: mjs)
  23. # [01:53] * Joins: mjs (mjs@17.255.104.100)
  24. # [02:04] * Joins: karl (karlcow@128.30.52.30)
  25. # [02:21] * Quits: gavin (gavin@74.103.208.221) (Ping timeout)
  26. # [02:26] * Joins: gavin (gavin@74.103.208.221)
  27. # [02:43] <karl> http://developer.apple.com/documentation/AppleApplications/Reference/SafariHTMLRef/Articles/StandardHTMLTags.html
  28. # [02:43] <karl> http://developer.apple.com/documentation/AppleApplications/Reference/SafariHTMLRef/Articles/HTMLExtensions.html#//apple_ref/doc/uid/TP40002066-SW1
  29. # [02:44] <karl> hmmm
  30. # [02:44] <karl> A new search type, <input type="search">, that allows you to create a Mac OS X-style search box.
  31. # [02:45] <karl> been created for widgets it seems
  32. # [02:45] <karl> http://developer.apple.com/macosx/dashboard.html
  33. # [02:53] <mjs> it was originally made for our RSS view
  34. # [02:54] * Quits: zcorpan_ (zcorpan@84.216.41.153) (Ping timeout)
  35. # [02:54] <karl> oooh
  36. # [02:55] <karl> thanks
  37. # [02:55] <karl> mjs: is there a spec in developer doc which describes how it is working
  38. # [02:55] <karl> I searched a bit without good results
  39. # [02:55] <mjs> karl: I don't remember, I asked adele to write something up based on the implementation recently
  40. # [02:56] <karl> many thanks
  41. # [02:58] * Joins: olivier (ot@128.30.52.30)
  42. # [03:06] * Quits: mjs (mjs@17.255.104.100) (Quit: mjs)
  43. # [03:28] * Joins: mw22 (chatzilla@84.41.169.151)
  44. # [04:11] * Joins: rburns (robburns@67.167.7.187)
  45. # [04:26] * Parts: rburns (robburns@67.167.7.187)
  46. # [04:28] * Quits: gavin (gavin@74.103.208.221) (Ping timeout)
  47. # [04:34] * Joins: gavin (gavin@74.103.208.221)
  48. # [04:42] * Joins: mjs (mjs@24.4.61.129)
  49. # [04:47] * Quits: Sander (svl@71.57.109.108) (Quit: And back he spurred like a madman, shrieking a curse to the sky.)
  50. # [04:54] * Quits: mjs (mjs@24.4.61.129) (Quit: mjs)
  51. # [05:16] * Quits: gsnedders (gsnedders@81.132.88.104) (Ping timeout)
  52. # [06:02] * Joins: Lachy_ (chatzilla@124.168.24.114)
  53. # [06:03] * Joins: MikeSmith (MikeSmith@mcclure.w3.org)
  54. # [06:05] * Quits: Lachy (chatzilla@203.158.61.14) (Ping timeout)
  55. # [06:05] * Lachy_ is now known as Lachy
  56. # [06:07] * Quits: Lachy (chatzilla@124.168.24.114) (Quit: ChatZilla 0.9.78.1 [Firefox 2.0.0.4/2007051502])
  57. # [06:18] * Quits: MikeSmith (MikeSmith@mcclure.w3.org) (Client exited)
  58. # [06:19] * Joins: MikeSmith (MikeSmith@mcclure.w3.org)
  59. # [06:27] <heycam> is there a spec for XSLT processing from script in browsers?
  60. # [06:27] <heycam> like the XSLTProcessor in mozilla (and opera?)
  61. # [06:29] * Joins: Lionhear1 (robin@66.57.69.65)
  62. # [06:30] * Quits: Lionheart (robin@66.57.69.65) (Ping timeout)
  63. # [06:53] <karl> heycam: what kind of information you are looking for?
  64. # [06:53] <heycam> i'm just wondering if there is one, and if there isn't, whether it should be specced
  65. # [06:54] <karl> I still do not understand :)
  66. # [06:54] <karl> You mean a description of the Javascript code implementing XSLT?
  67. # [06:55] <heycam> no, a specification for using the XSLT engine built in to the browser, from script
  68. # [06:55] <heycam> like: http://developer.mozilla.org/en/docs/Using_the_Mozilla_JavaScript_interface_to_XSL_Transformations
  69. # [06:55] <heycam> yes it's in opera 9 too it seems
  70. # [07:01] <karl> heycam, you want to have a defined XSLT API?
  71. # [07:01] <heycam> yes
  72. # [07:02] <karl> heycam: http://xml.apache.org/xalan-j/trax.html
  73. # [07:03] <heycam> seems webapi might be interested in it: http://lists.w3.org/Archives/Public/public-webapi/2006Mar/0021.html
  74. # [07:04] <heycam> yes trax seems to be the standard for java
  75. # [07:08] * Quits: gavin (gavin@74.103.208.221) (Ping timeout)
  76. # [07:13] * Joins: gavin (gavin@74.103.208.221)
  77. # [07:48] * Quits: heycam (cam@130.194.72.84) (Quit: bye)
  78. # [08:15] * Joins: Lachy (chatzilla@124.168.24.114)
  79. # [08:32] * Quits: olivier (ot@128.30.52.30) (Quit: Leaving)
  80. # [08:33] * Joins: olivier (ot@128.30.52.30)
  81. # [08:42] * Quits: olivier (ot@128.30.52.30) (Quit: Leaving)
  82. # [08:43] * Quits: karl (karlcow@128.30.52.30) (Quit: Where dwelt Ymir, or wherein did he find sustenance?)
  83. # [10:56] * Disconnected
  84. # [10:56] * Attempting to rejoin channel #html-wg
  85. # [10:56] * Rejoined channel #html-wg
  86. # [10:56] * Topic is 'HTML WG http://www.w3.org/html/wg/ logged: http://krijnhoetmer.nl/irc-logs/'
  87. # [10:56] * Set by Zeros on Mon Apr 30 23:38:28
  88. # [13:01] * Disconnected
  89. # [13:01] * Attempting to rejoin channel #html-wg
  90. # [13:01] * Rejoined channel #html-wg
  91. # [13:01] * Topic is 'HTML WG http://www.w3.org/html/wg/ logged: http://krijnhoetmer.nl/irc-logs/'
  92. # [13:01] * Set by Zeros on Mon Apr 30 23:38:28
  93. # [13:02] * Quits: polin8 (polin8@24.184.204.6) (Quit: polin8)
  94. # [13:03] * Joins: polin8 (polin8@24.184.204.6)
  95. # [15:07] * Disconnected
  96. # [15:07] * Attempting to rejoin channel #html-wg
  97. # [15:07] * Rejoined channel #html-wg
  98. # [15:07] * Topic is 'HTML WG http://www.w3.org/html/wg/ logged: http://krijnhoetmer.nl/irc-logs/'
  99. # [15:07] * Set by Zeros on Mon Apr 30 23:38:28
  100. # [16:16] <krijnh> Ping
  101. # [16:16] * Disconnected
  102. # [16:16] * Attempting to rejoin channel #html-wg
  103. # [16:16] * Rejoined channel #html-wg
  104. # [16:16] * Topic is 'HTML WG http://www.w3.org/html/wg/ logged: http://krijnhoetmer.nl/irc-logs/'
  105. # [16:16] * Set by Zeros on Mon Apr 30 23:38:28
  106. # [16:16] * Joins: gsnedders (gsnedders@81.132.88.104)
  107. # [18:24] * Disconnected
  108. # [18:24] * Attempting to rejoin channel #html-wg
  109. # [18:24] * Rejoined channel #html-wg
  110. # [18:24] * Topic is 'HTML WG http://www.w3.org/html/wg/ logged: http://krijnhoetmer.nl/irc-logs/'
  111. # [18:24] * Set by Zeros on Mon Apr 30 23:38:28
  112. # [18:24] <schnitz_> gsnedders, thats fine
  113. # [18:24] <mjs> schnitz_: other people have suggested it at various times
  114. # [18:24] <schnitz_> mjs, ok, didn't know that
  115. # [18:25] * Quits: anne (annevk@213.236.208.22) (Ping timeout)
  116. # [18:25] <schnitz_> gsnedders, trying to understand what you find elegant here, can you tell my any product that had ever a version jump in that range?
  117. # [18:26] <schnitz_> s/my/me
  118. # [18:26] <mjs> Java
  119. # [18:27] <schnitz_> Java?
  120. # [18:27] <mjs> 1.4 to 5
  121. # [18:27] <gsnedders> 1.4.2 to 5.0
  122. # [18:27] <hsivonen> Solaris
  123. # [18:27] <schnitz_> hmmm...
  124. # [18:27] <schnitz_> LOL
  125. # [18:27] <schnitz_> right
  126. # [18:27] <gsnedders> which is an even bigger jump
  127. # [18:27] <schnitz_> :-)
  128. # [18:27] <schnitz_> do we feel good about this (seriously)?
  129. # [18:28] <gsnedders> Solaris went from 2.6 to 7, BTW
  130. # [18:28] <hsivonen> schnitz_: about what? the naming debate? no.
  131. # [18:28] <Lachy> schnitz_: are you suggesting that XHTML2 will be the ultimate successor to XHTML 5/1.5?
  132. # [18:28] <schnitz_> hsivonen, no the jump
  133. # [18:28] <schnitz_> Lachy, nooooo....
  134. # [18:28] <hsivonen> schnitz_: from 1.0 to 5? sure
  135. # [18:29] <Lachy> oh, that's the impression I got from "XHTML 1.5 *and then* XHTML 2.0 is somewhat more elegant"
  136. # [20:29] * Disconnected
  137. # [20:29] * Attempting to rejoin channel #html-wg
  138. # [20:29] * Rejoined channel #html-wg
  139. # [20:29] * Topic is 'HTML WG http://www.w3.org/html/wg/ logged: http://krijnhoetmer.nl/irc-logs/'
  140. # [20:29] * Set by Zeros on Mon Apr 30 23:38:28
  141. # [20:29] <anne> what markup are people currently using?
  142. # [20:29] <Lionheart> Well, the markup I'm currently using looks like...
  143. # [20:31] <Lionheart> <sup><a href="#acronym_ie" rel="footnote">(1)</a></sup> ... <ol><li id="acronym_ie">...
  144. # [20:31] <Lionheart> Bleah
  145. # [20:33] <Lionheart> A List Apart has an article using JavaScript to generate DOM children of <ol> for footnotes http://www.alistapart.com/articles/improvingprint
  146. # [20:34] <Lionheart> A Google search for "ol.footnotes" gets about 57 hits.
  147. # [20:35] <Lionheart> I rather expected more
  148. # [20:38] <Lionheart> div.footnotes gets about 162
  149. # [20:39] <Lionheart> Though 111 of those are div.footnotes ul|ol
  150. # [20:42] <Lionheart> Here's a page recommending <p class="footnotes">Footnotes:<br /><a id="...'><span class="footnote">(1)</span></a> ... <br /><a id="...'><span class="footnote">(2)</span></a> ... <br />...</p> -- aiyaaa.
  151. # [20:44] <Lionheart> Any given footnote might contain multiple paragraphs, so any footnote markup should be block-level not inline.
  152. # [20:47] <Lionheart> DocBook's footnote example is <para>During the installation of the product<footnote><para>In versions 2.3 and 2.4.</para></footnote> you may see messages such as these.</para> which has the same problem as the CSS float:note example
  153. # [20:47] * Quits: polin8 (polin8@209.176.7.3) (Client exited)
  154. # [20:47] * Joins: polin8 (polin8@209.176.7.3)
  155. # [20:53] <Lionheart> Let's see, and ODF's format, simplifed, goes <p>...<note><note-citation>1</note-citation><note-body><p>...</p></note-body></note></p>
  156. # [20:54] <Lionheart> Yeah, neither of those are helpful to us.
  157. # [20:57] <Lionheart> p.footnote gets about 11300 hits, though there´s a lot of noise in the results
  158. # [21:02] <Lionheart> Mostly noise.
  159. # [21:03] <Lionheart> ¨p.footnote CSS¨ filters it down to about 1030
  160. # [21:04] <Jero> Lionheart, could you explain to me why there should be a <notes> element that could be used instead of <ol>? I *still* don't see the need for it.
  161. # [21:05] <Lionheart> Jero: Several reasons. Ex. If you distribute the individual notes to the bottoms of pages theyŕe referred from, you want the <notes> element itself to disappear.
  162. # [21:07] <Jero> oh i see, so basically it's for media=print?
  163. # [21:08] <Lionheart> media=screen could distribute the footnotes into tooltips
  164. # [21:08] <Lionheart> If it wanted to
  165. # [21:10] <Lionheart> It´s for any UA that wants to process the notes structure differently, basically
  166. # [21:10] <Jero> yeah, if you put it that way I guess having a <notes> element would be nice to have
  167. # [21:12] <Jero> ok, thanks for the info, i'm off
  168. # [21:12] <Jero> later
  169. # [21:12] * Quits: Jero (Jero@213.46.207.230) (Quit: ChatZilla 0.9.78.1 [Firefox 2.0.0.4/2007051502])
  170. # [21:16] * Quits: polin8 (polin8@209.176.7.3) (Client exited)
  171. # [21:16] * Joins: polin8 (polin8@209.176.7.3)
  172. # [21:24] * Joins: dbaron (dbaron@63.245.220.242)
  173. # [21:25] <Lionheart> Bye, Jero. Tune in next week, we´ll cover bibliographies!
  174. # [21:36] <Lionheart> Bibliographic citations have some similarities to endnotes. An example from a W3C document we refer to often around here: ¨Documents with an XML MIME type are always handled in standards mode. [<cite><a href="#ref-doctype">DOCTYPE</a></cite>]¨
  175. # [21:39] <Lionheart> An MLA styled cite would go <cite>(Sivonen, 2007)</cite>
  176. # [21:41] <Lionheart> We´ve talked about <cite for=¨#q1¨> to tie cites to <blockquote> or <q>, but not about a cite referencing a bibliographic entry at the end of a document, which is more of a <cite from=¨...¨>.
  177. # [21:50] <Lionheart> It seems <cite> is used in two rather different ways: <p>As <cite for=¨#half¨>Abraham Flathead</cite> said in his call for unity, <q id=¨half¨>A home that´s cut in half usually falls over.</q> <cite href=¨#Lincoln_The_Carpentry_Years_Vol_15¨>(Sandbridge, 842)</cite></p>
  178. # [21:51] * Joins: kingryan_ (rking3@208.66.64.47)
  179. # [21:52] * Quits: kingryan (rking3@208.66.64.47) (Ping timeout)
  180. # [21:58] <Lionheart> Hm, seems like I´m monologuing.
  181. # [21:58] * Joins: mjs (mjs@17.255.100.184)
  182. # [22:03] * Quits: gavin (gavin@74.103.208.221) (Ping timeout)
  183. # [22:08] * Joins: gavin (gavin@74.103.208.221)
  184. # [22:12] <hsivonen> offtopic but why is the DVS stuff Gregory talks about distributes as VHS instead of, say, audio CD? just curious
  185. # [22:18] * Quits: ROBOd (robod@86.34.246.154) (Quit: http://www.robodesign.ro )
  186. # [22:18] * Quits: kingryan_ (rking3@208.66.64.47) (Client exited)
  187. # [22:19] * Joins: kingryan (rking3@208.66.64.47)
  188. # [22:23] * Joins: zcorpan_ (zcorpan@84.216.42.249)
  189. # [22:41] * Quits: polin8 (polin8@209.176.7.3) (Quit: polin8)
  190. # [22:43] * Joins: polin8 (polin8@209.176.7.3)
  191. # [22:50] <hsivonen> whoa! the Apple docs say they implement <layer>!
  192. # [22:50] <mjs> the docs are incorrect if they say that
  193. # [22:51] <hsivonen> mjs: "not well supported" implies some support: http://developer.apple.com/documentation/AppleApplications/Reference/SafariHTMLRef/Articles/HTMLExtensions.html#//apple_ref/doc/uid/TP40002066-SW1
  194. # [22:52] <Hixie> well, they're supported in that they end up in the dom
  195. # [22:52] <Hixie> that's _some_ support ;-)
  196. # [22:52] <mjs> hsivonen: hmm, we recognize the layer tag but I think we just make a generic element
  197. # [22:52] <mjs> hsivonen: we handle it as a block-level tag for parsing
  198. # [22:53] <mjs> hsivonen: that's it
  199. # [22:53] <hsivonen> mjs: OK. reading the doc made me expect more
  200. # [22:53] <mjs> hsivonen: khtml has some sort of half-assed support, we might have had it once but it's long-since removed
  201. # [22:55] <Hixie> safari does indeed terminate <p>s with <layer>
  202. # [22:55] * hsivonen reminisces the Mozilla <layer> killing evangelism effort
  203. # [22:56] <mjs> die, <layer>, die
  204. # [22:56] <mjs> anyway I'm not sure if our parsing hacks for it are actually useful for anything or just a relic
  205. # [22:57] <zcorpan_> ie doesn't terminate <p> with <layer>
  206. # [22:58] <zcorpan_> nor firefox or opera
  207. # [22:59] <Hixie> nor the spec :-)
  208. # [22:59] <hsivonen> Hixie: suppose input is labeled text/html; charset=UTF-16LE and the first three code units are: BOM, BOM, letter A. What's the first token?
  209. # [22:59] <hsivonen> doh. that's an obvious case
  210. # [22:59] <Hixie> hsivonen: character 'BOM'
  211. # [22:59] <hsivonen> I meant
  212. # [22:59] <Hixie> hsivonen: character 'A' to your next question
  213. # [22:59] <Hixie> (BOM, letter A)
  214. # [23:00] <hsivonen> Hixie: suppose input is labeled text/html; charset=UTF-16 and the first six bytes are: BOM, BOM, letter A. What's the first token?
  215. # [23:00] <Hixie> character 'BOM'
  216. # [23:00] <hsivonen> Hixie: thanks. can I read this in the spec?
  217. # [23:00] <Hixie> doesn't make any difference what the encoding is, HTML5 requires the encoding layer to not eat the first BOM
  218. # [23:01] <Hixie> hm
  219. # [23:02] <Hixie> 8.2.2.3 is the section that is supposed to cover this
  220. # [23:02] <Hixie> i assume the argument is that the first paragraph eats a BOM
  221. # [23:02] <Hixie> I could make that explicit
  222. # [23:02] <hsivonen> I'd appreciate that
  223. # [23:03] * hsivonen adds void dontSwallowBom()
  224. # [23:04] <Hixie> can you point me to the relevant part of UTF-16's definition that swallows the BOM? I want to make sure my prose is compatible.
  225. # [23:05] * Joins: myakura (myakura@58.88.37.26)
  226. # [23:09] <Hixie> hsivonen: ok, committed the update
  227. # [23:14] <hsivonen> Hixie: I can't find a de jure definition. My thinking is based on common-sense implementations
  228. # [23:17] <hsivonen> I wonder if point 4. at http://unicode.org/faq/utf_bom.html#28 is in a spec somewhere
  229. # [23:17] <hsivonen> "In particular, whenever a data stream is declared to be UTF-16BE, UTF-16LE, UTF-32BE or UTF-32LE a BOM must not be used."
  230. # [23:18] <Hixie> i wounder what the rationale was
  231. # [23:19] <hsivonen> the byte order is declared in the encoding label so the first code unit is part of the logical data
  232. # [23:20] <Hixie> that seems to be the same thinking that suggests that if you have higher-level encoding declarations, you shouldn't have <meta charset>, which everyone seems to think is a bad requirement
  233. # [23:22] <hsivonen> I think the existence of the *LE and *BE labels is a bad idea
  234. # [23:24] * Hixie implements the "permitted slash" nonsense
  235. # [23:25] * Quits: polin8 (polin8@209.176.7.3) (Quit: polin8)
  236. # [23:25] <Hixie> one day i'm gonna make my own version of HTML that doesn't have to make anyone happy but me
  237. # [23:25] <Hixie> and that version of HTML will be so great it'll take over the world
  238. # [23:27] * zcorpan_ wonders what Hixie is on about
  239. # [23:27] <Hixie> just whining that <img/> is allowed :-)
  240. # [23:28] <zcorpan_> oh
  241. # [23:28] <zcorpan_> well, from a practical POV i think it's good :)
  242. # [23:29] <Lionheart> I would really like to see <img>replacement text</img> but alas I can´t have my druthers
  243. # [23:29] <Hixie> the "permitted slash" nonsense is the poster child of how you don't have to make me agree with you for me to be convinced by the strength of your arguments
  244. # [23:29] <Hixie> i wonder if i should raise a formal objection over it :-P
  245. # [23:30] <zcorpan_> :P
  246. # [23:32] <Lionheart> Time to go to class. Ciao.
  247. # [23:33] * Quits: Lionheart (robin@66.57.69.65) (Quit: Leaving.)
  248. # [23:36] * beowulf finishes reading the backlog
  249. # [23:37] <beowulf> i'm not ever getting those 15 minutes back
  250. # [23:38] * Sander hopes someone was paying you for them. :)
  251. # [23:38] <beowulf> 'fraid not
  252. # [23:38] <Dashiva> If you can't get paid for doing something, do it while being paid to do something else :)
  253. # [23:39] <beowulf> i really want a 5 > 2 tee now
  254. # [23:40] * Sander goes trademark "web 5.0" and "web 1.5"
  255. # [23:41] <hsivonen> beowulf: we could start proudly wearing 1.5 < 2 tees
  256. # [23:41] <zcorpan_> 1.5 > 2
  257. # [23:41] <zcorpan_> :)
  258. # [23:41] <hsivonen> :-)
  259. # [23:42] <beowulf> i was reading and wondering what other name you'd suggest for xhtml5?
  260. # [23:42] <beowulf> "bert"
  261. # [23:42] <zcorpan_> i proposed "Bob"
  262. # [23:42] <beowulf> it seemed to be hinted it sholdn't contain x,h,t,m & l and all...
  263. # [23:42] <beowulf> Bob is good
  264. # [23:43] <beowulf> shorter
  265. # [23:43] <beowulf> <!DOCTYPE bob>
  266. # [23:43] <hsivonen> I don't like courting it with 5 > 2, but if we are really required to use 1.5, I might get a 1.5 > 2 shirt
  267. # [23:43] <zcorpan_> nope, that would trigger quirks mode :(
  268. # [23:43] <beowulf> boo!
  269. # [23:44] <beowulf> in the real world, no-one will use the version number anyway
  270. # [23:44] <beowulf> except the people who might possibly be using XHTML2 I suppose
  271. # [23:45] <beowulf> that's my guess
  272. # [23:45] * Joins: briansuda (briansuda@85.220.95.76)
  273. # [23:45] <zcorpan_> perhaps we should say that html uses levels, just like css
  274. # [23:45] <zcorpan_> HTML Level 5
  275. # [23:45] <zcorpan_> Version 1, Level 5
  276. # [23:45] <beowulf> HTML Level 5 with fireball attack!
  277. # [23:45] <beowulf> quirks mode again?
  278. # [23:46] <zcorpan_> that's Version 1, Level 5 with Quirks
  279. # [23:46] <hsivonen> ouch. the JDK UTF-8 decoder is not "safe"
  280. # [23:46] <hsivonen> that is, it accepts non-shortest-form
  281. # [23:49] <beowulf> actually we can't have 1.5
  282. # [23:49] <beowulf> i can't make a window 1.5 in irssi...
  283. # [23:52] <hsivonen> ouch times two. the JDK UTF-8 decoder silently accepts CESU-8
  284. # [23:54] <Hixie> nice
  285. # [23:55] <Hixie> i wonder how many sites are vulnerable as a result
  286. # [23:55] * Parts: hasather (hasather@80.203.71.22)
  287. # [23:55] <hsivonen> let's see if IBM has got it right
  288. # [23:56] <Hixie> ok i think i fixed my parser to be up to date
  289. # [23:56] <Hixie> it doesn't support entities and noscript, but that's ok
  290. # [23:57] <mjs> what is CESU-8?
  291. # [23:58] <Hixie> an abomination the likes of which rarely grace this earth
  292. # [23:58] <Hixie> it's UTF-8 except any characters not in the basic multilingual plane get encoded as their surrogate pairs, encoded as UTF-8
  293. # Session Close: Fri Jun 29 00:00:00 2007

The end :)