/irc-logs / w3c / #html-wg / 2007-09-25 / end

Options:

  1. # Session Start: Tue Sep 25 00:00:00 2007
  2. # Session Ident: #html-wg
  3. # [00:01] <hsivonen> mjs: In my experience the only thing that works is using Inkscape or a text editor and using the center point of an SVG text run as reference
  4. # [00:01] <hsivonen> mjs: Illustrator sucks
  5. # [00:01] <hsivonen> mjs: OmniGraffle Pro sucks
  6. # [00:01] <hsivonen> mjs: this stuff is *so* not ready for prime time :-(
  7. # [00:01] <Hixie> anne: how exciting
  8. # [00:01] <Hixie> anne: are they scared of publishing it or something?
  9. # [00:02] <anne> I guess, it's rather controversial :)
  10. # [00:02] <Hixie> it's weird how they seem stuck in the 90s
  11. # [00:02] <Hixie> "don't define it in detail! aah!"
  12. # [00:03] <mjs> are there any features in the XHR1 spec that aren't in any implementation?
  13. # [00:03] <Hixie> the error codes and constants, i imagine
  14. # [00:03] <anne> yeah
  15. # [00:04] <anne> both added on request from implementors...
  16. # [00:04] <Hixie> and both important
  17. # [00:04] <mjs> that doesn't seem to be their complaint though
  18. # [00:04] * hsivonen wonders which mailing list is being discussed
  19. # [00:04] <Hixie> mjs: they mentioned both, i think
  20. # [00:04] <anne> http://lists.w3.org/Archives/Member/member-webapi/2007Sep/0001.html
  21. # [00:04] <mjs> member-webapi
  22. # [00:04] <Hixie> they certainly mentioned the errors
  23. # [00:04] <hsivonen> thanks
  24. # [00:05] <Hixie> i was also amused by their request that it be versioned
  25. # [00:05] <Hixie> woah
  26. # [00:05] <Hixie> XMLHttpRequestException isn't a DOMException?
  27. # [00:06] <anne> it should just be the same, actually
  28. # [00:06] <anne> like HTMLDocument and Document
  29. # [00:06] * anne isn't really sure how to do that properly for exceptions
  30. # [00:06] <Philip> https://bugzilla.mozilla.org/show_bug.cgi?id=296904
  31. # [00:06] <Hixie> drop the interface definition
  32. # [00:06] <Hixie> and just define the constants
  33. # [00:07] <Hixie> and say somewhere that when you say to raise an exception, you mean a DOMException object
  34. # [00:07] <anne> k, cool
  35. # [00:07] <Hixie> we'll have to update dom3 core at some point to add all the new exception codes
  36. # [00:08] <Hixie> i haven't added any yet but i could do with a dozen or so new ones for html5
  37. # [00:08] <anne> maybe I should simply add SECURITY_ERR = 17 to XHR for the moment...
  38. # [00:08] <anne> if it's as simple as that
  39. # [00:08] <Hixie> mmm
  40. # [00:08] <Hixie> SECURITY_ERR
  41. # [00:08] <Hixie> that's one of the ones i need :-)
  42. # [00:08] <Hixie> isn't 17 used up?
  43. # [00:08] <Hixie> anyway
  44. # [00:08] <anne> oh, maybe 18 then
  45. # [00:08] * anne forgets
  46. # [00:08] <anne> you e-mailed a proposal for that one at some point
  47. # [00:09] <Hixie> i thought a bunch of the lower numbers were used up by one of the other specs
  48. # [00:09] <Hixie> ah
  49. # [00:09] <Hixie> how forward-thinking of me
  50. # [00:09] <anne> 18 indeed
  51. # [00:09] <anne> 0-50 is reserved for DOMCore
  52. # [00:09] <anne> there's some data on that somewhere, but I'm not sure how much it matters
  53. # [00:10] <anne> it's what the old DOM WG came up with
  54. # [00:10] <Hixie> http://lists.w3.org/Archives/Public/public-webapi/2006May/0027.html
  55. # [00:10] <Hixie> interestin
  56. # [00:10] <Hixie> well anyway
  57. # [00:10] <Hixie> we are gonna need to revamp DOM3 Core at some point
  58. # [00:10] <Hixie> but i guess we'd better start collecting the exception codes somewhere
  59. # [00:10] <Hixie> in a wiki or something
  60. # [00:11] <anne> i've been thinking about naming it something differently from "dom3 core" something like "Web DOM" as to not irritate other consumers of the DOM specifications
  61. # [00:11] <zcorpan_> DOM5 Core
  62. # [00:11] <anne> of course, as for doing actual work on the spec...
  63. # [00:12] * zcorpan_ renames his spec to ARIA5
  64. # [00:12] <Hixie> Web DOM5
  65. # [00:12] <Hixie> zcorpan_: hah
  66. # [00:14] <Hixie> http://wiki.whatwg.org/wiki/Exception_Codes
  67. # [00:14] <Hixie> there
  68. # [00:14] <Hixie> we can now track exception codes and prevent clashes
  69. # [00:14] <Hixie> if anyone knows of an exception code that isn't on that list, add it
  70. # [00:20] * Quits: Zeros (Zeros-Elip@67.154.87.254) (Quit: Leaving)
  71. # [00:21] <mjs> DOM3 Core does need some serious revision for web use
  72. # [00:21] <Hixie> Web DOM5
  73. # [00:21] <Hixie> i'll get right on that after html5, css5, svg5, http5...
  74. # [00:22] <Hixie> is anne still working on xml5?
  75. # [00:22] <anne> yeah, but some other work got in the way
  76. # [00:22] <Hixie> bbiab, afk
  77. # [00:22] <anne> anyway, going to bed
  78. # [00:27] * Quits: jgraham (jgraham@81.86.210.17) (Ping timeout)
  79. # [00:31] * Joins: hyatt (hyatt@67.125.69.250)
  80. # [00:40] * Quits: zcorpan_ (zcorpan@83.227.34.9) (Ping timeout)
  81. # [00:52] * Quits: tH (Rob@87.102.83.139) (Quit: ChatZilla 0.9.78.1-rdmsoft [XULRunner 1.8.0.9/2006120508])
  82. # [01:08] * Quits: hyatt (hyatt@67.125.69.250) (Client exited)
  83. # [01:08] * Joins: hyatt (hyatt@67.125.69.250)
  84. # [01:12] * Joins: timbl (timbl@146.115.66.146)
  85. # [01:15] * Joins: sbuluf (gc@200.49.140.210)
  86. # [01:19] * Quits: billmason (billmason@69.30.57.156) (Quit: .)
  87. # [01:35] * Quits: Sander (svl@86.87.68.167) (Quit: And back he spurred like a madman, shrieking a curse to the sky.)
  88. # [01:36] * Joins: mjs_ (mjs@17.255.111.173)
  89. # [01:37] <hyatt> ,;,/
  90. # [01:38] * Quits: hyatt (hyatt@67.125.69.250) (Quit: hyatt)
  91. # [01:38] * Joins: hyatt (hyatt@67.125.69.250)
  92. # [01:38] * Quits: mjs (mjs@17.203.14.225) (Ping timeout)
  93. # [01:39] * Joins: mjs (mjs@17.203.14.225)
  94. # [01:41] * Quits: hyatt (hyatt@67.125.69.250) (Ping timeout)
  95. # [01:41] * Quits: mjs_ (mjs@17.255.111.173) (Ping timeout)
  96. # [02:31] * Joins: MikeSmith (MikeSmith@mcclure.w3.org)
  97. # [02:39] * Joins: aaronlev (chatzilla@209.6.168.245)
  98. # [02:49] * Quits: sbuluf (gc@200.49.140.210) (Ping timeout)
  99. # [02:52] * Joins: sbuluf (mwgkdo@200.49.140.157)
  100. # [03:01] * Joins: billyjack (MikeSmith@mcclure.w3.org)
  101. # [03:01] * Quits: billyjack (MikeSmith@mcclure.w3.org) (Client exited)
  102. # [03:06] * Quits: aaronlev (chatzilla@209.6.168.245) (Quit: ChatZilla 0.9.78.1 [Firefox 2.0.0.7/2007091417])
  103. # [03:15] * Quits: schepers (schepers@128.30.52.30) (Ping timeout)
  104. # [03:30] * Quits: kingryan (rking3@208.66.64.47) (Quit: kingryan)
  105. # [03:31] * Joins: aaronlev (chatzilla@209.6.168.245)
  106. # [03:35] * Joins: olivier (ot@128.30.52.30)
  107. # [03:40] * Quits: MikeSmith (MikeSmith@mcclure.w3.org) (Quit: Less talk, more pimp walk.)
  108. # [04:12] * Quits: aaronlev (chatzilla@209.6.168.245) (Connection reset by peer)
  109. # [04:27] * Joins: MikeSmith (MikeSmith@mcclure.w3.org)
  110. # [05:08] * Quits: MikeSmith (MikeSmith@mcclure.w3.org) (Quit: Less talk, more pimp walk.)
  111. # [05:26] * Joins: hyatt (hyatt@67.125.69.250)
  112. # [05:32] * Joins: hyatt_ (hyatt@67.125.69.250)
  113. # [05:34] * Quits: hyatt (hyatt@67.125.69.250) (Ping timeout)
  114. # [05:49] * Joins: schepers (schepers@128.30.52.30)
  115. # [05:53] * Joins: DougJ (djones4@74.76.28.112)
  116. # [06:02] * Quits: DougJ (djones4@74.76.28.112) (Quit: DougJ)
  117. # [06:36] * Quits: Lachy (chatzilla@124.170.203.157) (Quit: ChatZilla 0.9.78.1 [Firefox 2.0.0.6/2007072518])
  118. # [06:47] * Parts: hyatt_ (hyatt@67.125.69.250)
  119. # [07:05] * Joins: MikeSmith (MikeSmith@mcclure.w3.org)
  120. # [07:48] * Joins: jgraham (jgraham@81.86.218.186)
  121. # [07:55] * Quits: mjs (mjs@17.203.14.225) (Quit: mjs)
  122. # [08:45] * Joins: mjs (mjs@64.81.48.145)
  123. # [08:48] * Quits: sbuluf (mwgkdo@200.49.140.157) (Ping timeout)
  124. # [08:52] * Quits: mjs (mjs@64.81.48.145) (Quit: mjs)
  125. # [09:04] * Quits: Yudai (Yudai@210.174.47.126) (Ping timeout)
  126. # [09:06] * Quits: schepers (schepers@128.30.52.30) (Quit: Trillian (http://www.ceruleanstudios.com)
  127. # [09:21] * Joins: Yudai (Yudai@211.128.41.224)
  128. # [09:32] * Quits: olivier (ot@128.30.52.30) (Quit: Leaving)
  129. # [09:44] * Quits: heycam (cam@203.214.33.166) (Quit: bye)
  130. # [09:48] * Joins: schepers (schepers@128.30.52.30)
  131. # [09:51] * Joins: Lachy (chatzilla@203.214.146.132)
  132. # [10:21] * Joins: mjs (mjs@64.81.48.145)
  133. # [10:25] <anne> Hixie, why Array and not DOMObject[] ?
  134. # [10:26] <Hixie> what's the difference?
  135. # [10:26] <anne> the latter syntax is used elsewhere in the spec
  136. # [10:27] <Hixie> ah
  137. # [10:27] <anne> unsigned ... int[] for ImageData.data iirc
  138. # [10:27] <Hixie> it'll all change when i synchronise with Bindings for DOM
  139. # [10:27] <anne> heh
  140. # [10:27] <anne> hmm, it's just int[] data
  141. # [10:27] <anne> that seems wrong
  142. # [11:02] * Joins: ROBOd (robod@89.123.24.243)
  143. # [11:03] <hsivonen> anne: I fixed the recent validator.nu bugs that claimed annevankesteren.nl was invalid
  144. # [11:04] <anne> cool
  145. # [11:08] * Parts: anne (annevk@81.68.67.12)
  146. # [11:08] * Joins: anne (annevk@81.68.67.12)
  147. # [11:21] * Quits: drry (drry@210.191.160.64) (Ping timeout)
  148. # [11:21] * Joins: drry (drry@210.191.160.64)
  149. # [11:22] * Joins: zcorpan_ (zcorpan@88.131.66.80)
  150. # [11:25] * Joins: Sander (svl@86.87.68.167)
  151. # [11:27] * Quits: MikeSmith (MikeSmith@mcclure.w3.org) (Quit: Less talk, more pimp walk.)
  152. # [11:35] * zcorpan_ wonders if 2.1.1.2 in http://www.w3.org/TR/aria-state/ is implemented in firefox
  153. # [11:35] <zcorpan_> i hope not...
  154. # [11:36] <anne> wtf
  155. # [11:37] <anne> maybe comment on that part of the draft?
  156. # [11:46] <anne> zcorpan_, I suggest to limit the amount of valid roles
  157. # [11:47] <anne> "an opaque string as defined by [...]"
  158. # [11:47] <anne> maybe simply include all the strings in your doc...
  159. # [12:36] <zcorpan_> yeah, i plan to include everything in due course
  160. # [13:27] * Joins: matt (matt@128.30.52.30)
  161. # [13:46] * Quits: timbl (timbl@146.115.66.146) (Quit: timbl)
  162. # [13:59] * Quits: matt (matt@128.30.52.30) (Quit: matt)
  163. # [14:03] * Joins: myakura (myakura@124.87.255.103)
  164. # [14:09] * Joins: heycam (cam@203.214.33.166)
  165. # [14:27] <anne> seems that same-origin checks are surprisingly simple...
  166. # [14:27] <anne> normalize the URI, compare scheme, domain and port (ascii case-insensitive) and you're done
  167. # [14:28] <anne> actually, are Unicode domains domains case-sensitive or not? ...
  168. # [14:29] <zcorpan_> they are, i think
  169. # [14:30] <zcorpan_> case insensitive that is
  170. # [14:30] <zcorpan_> :)
  171. # [14:30] <anne> unicode case insensitive?!
  172. # [14:30] <zcorpan_> smörgåsbord.se == SMÖRGÅSBORD.SE
  173. # [14:30] <anne> aaah
  174. # [14:31] <anne> oh wait
  175. # [14:31] <anne> that's probably gone after you "normalize the URI"
  176. # [14:31] * anne looks
  177. # [14:31] <zcorpan_> yeah, the domain is probably lowercased when normalized
  178. # [14:32] <anne> there's a whole chapter in RFC 3987 on normalization and comparison
  179. # [14:32] <anne> I'me saved!
  180. # [14:34] <anne> ouch, even Unicode normalization norm C sticks it head around the window there
  181. # [14:37] <anne> perform schem-based normalization on both URIs and then compare the "scheme" "domain" and "port" parts
  182. # [14:37] <anne> if there's no host info... crash
  183. # [14:47] <zcorpan_> "All MathML elements should be in the MathML namespace http://www.w3.org/1998/Math/MathML" -- http://www.w3.org/TR/2007/WD-mathml-for-css-20070924/
  184. # [14:47] <zcorpan_> should?
  185. # [14:50] <anne> bug?
  186. # [14:52] <zcorpan_> gah, what's up with all this DTD stuff everywhere
  187. # [14:52] <zcorpan_> when will spec writers stop messing with DTDs
  188. # [14:53] <Philip> If they didn't use DTDs, they'd be incompatible with all the tools that make use of DTDs, like, uh, ...
  189. # [14:53] <Philip> (Are there any except the W3C Validator?)
  190. # [14:55] <anne> ok, dealt with same-origin except for data:, javascript: etc.
  191. # [14:59] <beowulf> Philip: opensp?
  192. # [15:01] * Joins: olivier (ot@128.30.52.30)
  193. # [15:01] * Joins: matt (matt@128.30.52.30)
  194. # [15:05] * Joins: icaaq (icaaaq@85.228.55.82)
  195. # [15:22] * Joins: MikeSmith (MikeSmith@mcclure.w3.org)
  196. # [15:25] <Lachy> anne, yt?
  197. # [15:25] <anne> yeah
  198. # [15:26] <Lachy> got an email from Carl at opera. He wrote "you should send a quick text, when you are ante portas..." What does "ante portas" mean?
  199. # [15:27] <Lachy> I assume it's Norwegian or something
  200. # [15:28] <anne> sounds Latin to me
  201. # [15:28] <anne> for "at the gates"
  202. # [15:28] <Lachy> btw, my plane arrives in Oslo on Oct 3 at 11:00
  203. # [15:28] <anne> cool, but I'm in NL :)
  204. # [15:28] <Lachy> ah, ok.
  205. # [15:28] * Joins: aaronlev (chatzilla@66.31.86.217)
  206. # [15:28] <Philip> beowulf: Do you know what people use OpenSP for? (I can't find any references to applications in the first hundred search results, except about HTML and Docbook validation)
  207. # [15:29] <anne> Jacques Distler used it at one point to validate MathML entries on his site
  208. # [15:29] <anne> he might in fact still do that
  209. # [15:30] <beowulf> Philip: i probably should have appended a :) to my remark. i use opensp to validate html and spell-check
  210. # [15:30] <beowulf> well, i don't use it to spell check but the results get passed to sgml-spell-check
  211. # [15:31] <beowulf> http://developer.postgresql.org/~petere/sgml-spell-checker/
  212. # [15:34] <Philip> Okay, that sounds actually useful :-)
  213. # [15:37] * Joins: timbl (timbl@128.30.7.41)
  214. # [15:45] <zcorpan_> aaronlev: do you have any feedback about http://simon.html5.org/specs/aria-proposal ? does it match firefox (ignoring that you do namespace lookup)?
  215. # [15:46] * Joins: alexf (alejandro@85.152.42.1)
  216. # [15:47] * Parts: alexf (alejandro@85.152.42.1)
  217. # [15:48] <MikeSmith> Philip - the W3C markup validator uses the sp/opensp parser also; but other than that and the spell-checker and legacy DocBook SGML tools, I don't think it was ever used much in the free-software world at least
  218. # [15:53] <aaronlev> zcorpan_: i'll look at that soon, how long will you be online?
  219. # [15:54] <zcorpan_> aaronlev: i'll be online
  220. # [15:54] * Joins: hasather (hasather@90.227.221.48)
  221. # [16:06] * Quits: olivier (ot@128.30.52.30) (Quit: Leaving)
  222. # [16:08] <hsivonen> anne: you aren't quite saved. RFC 3987 weasels around the NFC vs. NFKC issue without quite pinning down which one it wants
  223. # [16:09] <hsivonen> or that's how I read it
  224. # [16:13] <hsivonen> I wonder if SMÖRGÅS written with the Ångström sign could fool a same-domain check
  225. # [16:13] <hsivonen> I wouldn't be too surprised if you could register a .com domain with the Ångström sign
  226. # [16:25] * Quits: aaronlev (chatzilla@66.31.86.217) (Client exited)
  227. # [16:28] <anne> hsivonen, hmm, we need URL5, but I don't want to write it
  228. # [16:35] * Joins: billmason (billmason@69.30.57.156)
  229. # [16:36] * Joins: aaronlev (chatzilla@66.31.86.217)
  230. # [16:39] * Joins: myakura_ (myakura@124.87.255.103)
  231. # [16:39] * Quits: myakura (myakura@124.87.255.103) (Ping timeout)
  232. # [16:45] <hsivonen> anne: are you still following the www-archive/html4all alt thread?
  233. # [16:45] <anne> with no intention to reply, but yes
  234. # [16:45] <hsivonen> anne: is it intuitively clear to you that absence of data is better that bogus data?
  235. # [16:45] <hsivonen> than
  236. # [16:46] <anne> in the general case, sure
  237. # [16:47] <anne> (i'm not a 100% sure here as besides "bogus" I believe they also reveal there's an image there)
  238. # [16:47] * Joins: tH (Rob@87.102.83.139)
  239. # [16:48] <anne> seems quite selective replying to your points though
  240. # [16:54] <Lachy> I'm not convinced that omitted data is better than bogus data, or vice versa
  241. # [16:56] <Lachy> Based on current AT behaviour alone, reading out redundant text that can at least be comprehended is better than reading out a randomly generated file name
  242. # [16:57] <anne> that's not omitted versus bogus though
  243. # [16:57] <Lachy> the argument that ATs can and should improve in the case of omitted alt, instead of reading out a incomprehensible file name, needs more evidence and research into exactly what UAs could do in practice
  244. # [17:08] <hsivonen> Lachy: is there a reason to believe the current state of AT is a permanent anomaly?
  245. # [17:09] <hsivonen> Lachy: fwiw, I'm not considering redundant data bogus here
  246. # [17:09] <hsivonen> IMG3442.jpg is bogus
  247. # [17:09] <hsivonen> "Grand Canyon" not
  248. # [17:09] <Lachy> I don't think it is a permanent state, I'd just like to know in what ways they really can be improved.
  249. # [17:10] <hsivonen> Lachy: even saying "image" is better than reading out IMG3423.jpg
  250. # [17:10] <hsivonen> certainly better than 3414313243_af83ef.jpg
  251. # [17:12] <Lachy> in some cases, reading out the filename isn't bad. e.g. src="grand-canyon.jpg", in others it's a disaster. So we need to know if it's reasonable to expect UAs to be able to distinguish between those cases.
  252. # [17:14] <anne> you can certainly do text heuristics and dictionary lookups
  253. # [17:14] <anne> that's toally reasonable
  254. # [17:17] <hsivonen> Lachy: dictionary lookups are reasonable
  255. # [17:18] <hsivonen> Lachy: also, it would be reasonable for synthetizers to allow the app to preflight as string to measure its spoken length prior to deciding whether to speak
  256. # [17:21] <Lachy> yeah, dictionaries may work to a certain extent. Though not everything is in a dictionary. In some cases, it may need some heuristics based on the phonetics or something. e.g. src="lachlanhunt.jpg" probably wouldn't show up in a dictionary, but could probably still be read.
  257. # [17:28] <aaronlev> i have a patch for firefox that i've never gotten in yet
  258. # [17:28] <aaronlev> that does some cool stuff to generate a repaired alt
  259. # [17:28] <aaronlev> but i'm not sure it needs to be specified does it?
  260. # [17:31] <anne> it should
  261. # [17:31] <anne> at least documentation
  262. # [17:31] <hsivonen> aaronlev: it doesn't the point just is that we shouldn't expect JAWS and WindowsEyes to be the end state of development
  263. # [17:31] <hsivonen> s/ the point/. The point/
  264. # [17:32] <aaronlev> hsivonen: you mean screen readers will still invent new ideas over time
  265. # [17:32] <aaronlev> or the browser will get better at helping them get repaired text
  266. # [17:33] <hsivonen> aaronlev: either
  267. # [17:33] * Parts: anne (annevk@81.68.67.12)
  268. # [17:33] <aaronlev> i would agree
  269. # [17:33] * Joins: anne (annevk@81.68.67.12)
  270. # [17:33] * Quits: anne (annevk@81.68.67.12) (Client exited)
  271. # [17:33] <hsivonen> gotta go
  272. # [17:33] * Joins: anne (annevk@81.68.67.12)
  273. # [17:33] <aaronlev> see you
  274. # [17:37] <aaronlev> zcorpan_: hi
  275. # [17:47] * Quits: schepers (schepers@128.30.52.30) (Quit: Trillian (http://www.ceruleanstudios.com)
  276. # [17:51] <zcorpan_> aaronlev: hi
  277. # [17:52] <aaronlev> zcorpan_: looking at spec now
  278. # [17:52] <zcorpan_> aaronlev: ok, i just added some examples
  279. # [17:52] <aaronlev> i'm trying to see if we can get rid of people using xhtml2:role and just move to |role|
  280. # [17:53] <aaronlev> in fact that already worked in ff2 so it's more doable than other streamlining improvements
  281. # [17:53] <zcorpan_> ok
  282. # [17:54] <zcorpan_> though xhtml2:role is still needed for svg
  283. # [17:59] * Quits: gavin (gavin@63.245.208.169) (Ping timeout)
  284. # [18:04] <aaronlev> zcorpan_: for svg you can just use xhtml:role no?
  285. # [18:04] <aaronlev> but yeah
  286. # [18:04] * Joins: gavin (gavin@63.245.208.169)
  287. # [18:04] <aaronlev> we should still support xhtml2:role for svg
  288. # [18:04] <aaronlev> i'm going to go work to phase out xhtml2:role on elements in the html/1999 namespace
  289. # [18:05] <zcorpan_> aaronlev: the spec doesn't define xhtml:role atm
  290. # [18:06] <aaronlev> what about role attribute module
  291. # [18:06] <zcorpan_> aaronlev: the role attribute module is confusing but is really defining the role attribute in no namespace on elements in the http://www.w3.org/1999/xhtml namespace
  292. # [18:06] <zcorpan_> aaronlev: not the role attribute in the http://www.w3.org/1999/xhtml namespace
  293. # [18:07] <aaronlev> zcorpan_: yeah
  294. # [18:08] <zcorpan_> i would be happy with using the xhtml namespace instead of the xhtml2 namespace, although dojo uses the xhtml2 namespace and i thought we wanted to support that :)
  295. # [18:08] <zcorpan_> i'm not so happy with supporting both
  296. # [18:08] * Joins: dbaron (dbaron@71.204.145.103)
  297. # [18:09] <aaronlev> zcorpan_: agreed
  298. # [18:09] <aaronlev> lets fix it
  299. # [18:10] <aaronlev> can you suggest wording and send to the editor Shane McCarron? You can cc me
  300. # [18:10] <aaronlev> you can suggest clarification as part of last call comments
  301. # [18:10] <aaronlev> we should not have to support xhtml:role
  302. # [18:12] <zcorpan_> aaronlev: ok
  303. # [18:14] <aaronlev> zcorpan_: I just sent you a draft note that I would send to current ARIA authors
  304. # [18:17] <zcorpan_> aaronlev: ok, it might take a while before it hits my inbox
  305. # [18:17] <aaronlev> k
  306. # [18:21] <zcorpan_> aaronlev: got it; sounds good
  307. # [18:21] <aaronlev> zcorpan_: thanks for helping with this
  308. # [18:21] <aaronlev> with all of it
  309. # [18:21] <zcorpan_> my pleasure :)
  310. # [18:21] <zcorpan_> we need to get it right in opera
  311. # [18:23] * Quits: dbaron (dbaron@71.204.145.103) (Quit: 8403864 bytes have been tenured, next gc will be global.)
  312. # [18:29] <zcorpan_> aaronlev: cc www-html-editor?
  313. # [18:29] <aaronlev> if you think so
  314. # [18:30] <aaronlev> can you cc me and rich?
  315. # [18:30] <aaronlev> schwer
  316. # [18:30] <aaronlev> at
  317. # [18:30] <aaronlev> us.ibm.com
  318. # [18:30] <zcorpan_> well, dunno, i want it archived somewhere
  319. # [18:30] <zcorpan_> sure
  320. # [18:30] <aaronlev> thanks
  321. # [18:30] <aaronlev> yeah, makes sense
  322. # [18:30] <zcorpan_> can i say that mozilla will remove support for xhtml:role?
  323. # [18:31] <zcorpan_> or that mozilla would like to remove support for it, or something
  324. # [18:43] <aaronlev> we'll remove it
  325. # [18:44] <aaronlev> zcorpan_: that's fine
  326. # [18:45] <zcorpan_> aaronlev: ok
  327. # [18:47] * Joins: Roger (roger@213.64.74.230)
  328. # [18:49] * Joins: dbaron (dbaron@63.245.220.241)
  329. # [18:55] <aaronlev> zcorpan_: ok now i'm going to go remove it from mozilla
  330. # [18:56] <zcorpan_> aaronlev: ok. i've written the email, just about to send it
  331. # [18:58] <aaronlev> zcorpan_: your spec doesn't say that if there's a prefix other than "wairole:", to look up the prefix and see if it's from the GUIRoleTaxononmy
  332. # [18:59] <zcorpan_> aaronlev: indeed
  333. # [18:59] <zcorpan_> aaronlev: i think we won't do namespace lookup in opera
  334. # [18:59] <aaronlev> k
  335. # [18:59] <aaronlev> but, are you sure it shouldn't be in the spec?
  336. # [18:59] <aaronlev> it's only a couple of lines of code
  337. # [19:00] <zcorpan_> maybe in a future version of the spec, when we actually need the extensibility
  338. # [19:00] <zcorpan_> the permitted values are compatible with such processing
  339. # [19:02] * Quits: matt (matt@128.30.52.30) (Quit: matt)
  340. # [19:02] <aaronlev> k, but will we remember to change this spec then?
  341. # [19:03] <zcorpan_> sure
  342. # [19:03] <aaronlev> So, even if that world comes to be
  343. # [19:03] <aaronlev> "wairole:" will be reserved?
  344. # [19:03] <zcorpan_> yeah
  345. # [19:03] <aaronlev> ok
  346. # [19:04] <aaronlev> but only when used on an html element?
  347. # [19:04] <aaronlev> what if you use it on svg?
  348. # [19:04] <aaronlev> can we say anything about that?
  349. # [19:04] <zcorpan_> any element
  350. # [19:05] <aaronlev> but the spec says it only applies to elemments in the html/1999 namespace
  351. # [19:05] <zcorpan_> no: "Authors may specify a role attribute in the http://www.w3.org/TR/xhtml2 namespace on *any element*, unless ..."
  352. # [19:06] <aaronlev> where's the link again?
  353. # [19:06] <zcorpan_> http://simon.html5.org/specs/aria-proposal
  354. # [19:06] <aaronlev> ok
  355. # [19:07] <aaronlev> the namespace folks might not like that
  356. # [19:07] <zcorpan_> why not? the xml: prefix is also fixed :)
  357. # [19:07] <aaronlev> it is?
  358. # [19:07] <zcorpan_> yes
  359. # [19:07] <aaronlev> ok, well then
  360. # [19:07] <aaronlev> i like it :)
  361. # [19:07] <zcorpan_> :)
  362. # [19:08] <zcorpan_> as is the no namespace namespace :)
  363. # [19:08] <aaronlev> Authors may specify a role attribute in the http://www.w3.org/TR/xhtml2 namespace on any element, unless that is an element in the http://www.w3.org/1999/xhtml namespace and it has a role attribute in no namespace specified.but will you be changing
  364. # [19:08] <aaronlev> won't we change that to
  365. # [19:08] <aaronlev> Authors may specify a role attribute in the http://www.w3.org/TR/xhtml2 on any element not in the http://www.w3.org/1999/xhtml namespace.
  366. # [19:09] <zcorpan_> yeah, that could work
  367. # [19:09] <aaronlev> becuase in that case they need to just use a no-namespace role attribute
  368. # [19:09] <zcorpan_> but the aaa: attributes should still be allowed on xhtml elements?
  369. # [19:10] <zcorpan_> for compat with firefox 2?
  370. # [19:10] <aaronlev> right, we need that sorry
  371. # [19:10] <aaronlev> ff2 supports role without a namespace
  372. # [19:11] <zcorpan_> yep
  373. # [19:11] <aaronlev> dojo is shipping before ff3 ships
  374. # [19:11] <aaronlev> so, xhtml2:role="checkbox" would work on svg?
  375. # [19:11] <aaronlev> so no namespace or wairole: would mean the same thing, for any element?
  376. # [19:13] <zcorpan_> yes
  377. # [19:13] <zcorpan_> updated the spec
  378. # [19:17] <aaronlev> zcorpan_: i think some people may feel their toes are stepped on, but it is as you say, a proposal
  379. # [19:17] <aaronlev> because it says how xhtml2 role should be used outside of html
  380. # [19:17] <aaronlev> but, i don't see why we shouldn't propose it
  381. # [19:17] <aaronlev> let them try and argue against it
  382. # [19:18] <zcorpan_> yeah. or well technically it isn't even the actual xhtml2 namespace
  383. # [19:19] <zcorpan_> the xhtml2 spec does not define that namespace
  384. # [19:19] <zcorpan_> and since namespaces are opaque strings, it could just as well be "urn:blurb"
  385. # [19:20] <zcorpan_> which the xhtml2 spec also doesn't define
  386. # [19:20] <zcorpan_> the xhtml2 spec defines the "http://www.w3.org/2002/06/xhtml2/" namespace
  387. # [19:20] <aaronlev> i'm glad you understand it
  388. # [19:20] <aaronlev> i don't
  389. # [19:20] <zcorpan_> :)
  390. # [19:20] <aaronlev> not sure i want to either
  391. # [19:21] <zcorpan_> no, namespaces are messy :)
  392. # [19:22] * Quits: myakura_ (myakura@124.87.255.103) (Quit: Leaving...)
  393. # [19:23] <zcorpan_> http://tinyurl.com/2qdvmx -- comments before i send it?
  394. # [19:24] <zcorpan_> perhaps i shouldn't mention the .../xhtml2 namespace thing
  395. # [19:26] <aaronlev> zcorpan_: i agree
  396. # [19:26] <aaronlev> The role attribute in no namespace on an element in the
  397. # [19:26] <aaronlev> http://www.w3.org/1999/xhtml namespace is a required feature in
  398. # [19:26] <aaronlev> this specification.
  399. # [19:27] <aaronlev> The grammar isn't quite right
  400. # [19:27] <aaronlev> I'm trying to figure out how to reword that
  401. # [19:28] <zcorpan_> i guess i could drop that point also, the spec doesn't really say what to do with role attributes anyway
  402. # [19:29] <aaronlev> can you send me a new version when you have it?
  403. # [19:30] <zcorpan_> http://tinyurl.com/368vgu
  404. # [19:34] <aaronlev> zcorpan_: looks good, put my last name "Leventhal" too
  405. # [19:34] <zcorpan_> aaronlev: ok. sending
  406. # [19:37] * Quits: Roger (roger@213.64.74.230) (Quit: Roger)
  407. # [19:37] * Joins: tH_ (Rob@87.102.114.133)
  408. # [19:37] <zcorpan_> http://lists.w3.org/Archives/Public/www-html-editor/2007JulSep/0032.html
  409. # [19:38] * Quits: tH (Rob@87.102.83.139) (Ping timeout)
  410. # [19:38] * tH_ is now known as tH
  411. # [19:39] <aaronlev> zcorpan_: now, about the role proposal
  412. # [19:40] <aaronlev> maybe point out that xml: is hardcoded
  413. # [19:40] <aaronlev> wairole: would be reserved in a similar way to xml:
  414. # [19:40] <zcorpan_> aaronlev: in the spec, you mean?
  415. # [19:40] <aaronlev> yes, i think it can be worded appropriately into the spec
  416. # [19:40] <zcorpan_> ok
  417. # [19:41] <aaronlev> that legitimizes it, and I think a lot of people reading it won't know about that
  418. # [19:41] <aaronlev> I have a patch for Firefox, but I don't want to seek review until we get feedback on this
  419. # [19:44] <zcorpan_> added a note
  420. # [19:46] <zcorpan_> extended the note
  421. # [19:47] <aaronlev> Here's what my patch implements:
  422. # [19:47] <aaronlev> 1. The role attribute must be in no namespace for elements in http://www.w3.org/1999/xhtml namespace
  423. # [19:47] <aaronlev> 2. The role attribute must be in xhtml2 namespace for anything else
  424. # [19:47] <aaronlev> 3. If the role _value_ has no namespace prefix or the reserved prefix "wairole:", they are considered to be from the WAI roles, mapped from "http://www.w3.org/2005/01/wai-rdf/GUIRoleTaxonomy#". Either shortcut is allowed whenever role is used, and does not depend on the namespace of the element with the role.
  425. # [19:48] <zcorpan_> so you won't support xhtml2:role on html elements?
  426. # [19:49] <aaronlev> zcorpan_: what do you think the right thing to do is?
  427. # [19:49] <aaronlev> i just told everyone to use |role| and not xhtml2:role for html elements
  428. # [19:50] <aaronlev> i sent it to everyone i know using aria and to public mailing lists
  429. # [19:51] <aaronlev> i think we're ok if we remove it, as long as people don't think that's inconistent with how things should work
  430. # [19:51] <zcorpan_> ok, yeah i'm happy to remove it
  431. # [19:51] <zcorpan_> i thought it needed to be supported for dojo
  432. # [19:51] <aaronlev> zcorpan_: dojo is willing to change it
  433. # [19:52] <aaronlev> it's just one place where they set the role
  434. # [19:52] <aaronlev> i asked
  435. # [19:52] <zcorpan_> ok, great
  436. # [19:52] * zcorpan_ updates the spec
  437. # [19:52] <aaronlev> zcorpan_: i'm more willing to do this today because i got the aria-foo part checked in. now i'm ready for polish :)
  438. # [19:52] <aaronlev> the other part was bigger and harder
  439. # [19:54] <zcorpan_> ok :)
  440. # [19:54] <aaronlev> zcorpan_: so let me run this by my colleague here before we submit it anywhere, is that ok with you?
  441. # [19:54] <aaronlev> i want to make sure he's okay with it
  442. # [19:54] <zcorpan_> aaronlev: sure
  443. # [19:55] <aaronlev> is it updated now?
  444. # [19:55] <zcorpan_> now it is
  445. # [19:56] <aaronlev> thanks for writing it
  446. # [19:56] <aaronlev> you can list my name as a secondary contributor if you want
  447. # [19:56] <zcorpan_> oh of course
  448. # [19:57] <aaronlev> Should we remove "Should the last one be disallowed? Or the last two ones even?"
  449. # [19:57] <aaronlev> aren't we set now with the proposal as is?
  450. # [19:57] <zcorpan_> yeah
  451. # [19:57] <aaronlev> why is the first box under "Abstract" empty
  452. # [19:57] <zcorpan_> it will contain the abstract :)
  453. # [19:58] <aaronlev> k
  454. # [19:58] <aaronlev> I find this part difficult to undetrsand: "No other attributes are to be processed in this way — in particular not foo attributes in no namespace on elements in the http://www.w3.org/1999/xhtml namespace. "
  455. # [19:59] <aaronlev> maybe i should put my feedback in email now
  456. # [19:59] <zcorpan_> see 2.1.1.2 XHTML without prefixes in http://www.w3.org/TR/2007/WD-aria-state-20070601/#module_element
  457. # [20:00] <zcorpan_> it suggests that <div required="true"> should be equivalent to <div aaa:required="true">
  458. # [20:02] <aaronlev> which sentence says that?
  459. # [20:03] * Quits: MikeSmith (MikeSmith@mcclure.w3.org) (Ping timeout)
  460. # [20:03] <aaronlev> Wow, I didn't know you could just use "aaa:' without defining it, does it say that?
  461. # [20:03] <zcorpan_> "Prefixes can sometimes be dropped when using XHTML 1.1 For Accessible Adaptable Applications."
  462. # [20:04] <zcorpan_> aaronlev: you can if you have a validating xml processor, because the decl is defined in the dtd... but browsers don't and that will result in a YSoD in moz
  463. # [20:04] <aaronlev> that's a bad idea
  464. # [20:04] <zcorpan_> yes
  465. # [20:04] <aaronlev> i mean
  466. # [20:04] <aaronlev> none of this works in firefox
  467. # [20:04] <aaronlev> we should get rid of that
  468. # [20:04] <zcorpan_> yes
  469. # [20:05] <zcorpan_> try to get rid of the dtd stuff altogether, i'd suggest :)
  470. # [20:05] <zcorpan_> they are just a waste of time
  471. # [20:12] <aaronlev> zcorpan_: want to join the working group?:)
  472. # [20:12] <aaronlev> just long enough to help us fix the specs?
  473. # [20:12] <aaronlev> i've been heads down coding and don't even notice some of the problems like this
  474. # [20:13] <zcorpan_> yeah. sure. :) i'll hear with my manager
  475. # [20:14] <aaronlev> chaals will support you
  476. # [20:14] <zcorpan_> hmm. if we don't need to support xhtml2:role on html elements, then perhaps we can use html:role on non-html elements after all
  477. # [20:14] <zcorpan_> because people don't use role on non-html elements at all yet
  478. # [20:15] <zcorpan_> it would still be backwards compatible with firefox 2
  479. # [20:15] <aaronlev> yes we do in firefox itself
  480. # [20:15] <aaronlev> in the XUL UI
  481. # [20:16] <aaronlev> but of course we can change that as firefox changes
  482. # [20:17] <zcorpan_> i mean that we use http://www.w3.org/1999/xhtml instead of http://www.w3.org/TR/xhtml2 as the namespace for the attribute
  483. # [20:17] <aaronlev> btw they're changing the spec to say "aria:" instead of "aaa:"
  484. # [20:17] <aaronlev> but it's not public yet
  485. # [20:18] <aaronlev> zcorpan_: i have no problem wit hthat
  486. # [20:18] <zcorpan_> ok
  487. # [20:18] * Joins: Roger (roger@213.64.74.230)
  488. # [20:19] <aaronlev> zcorpan_: i check for the namespaced properties before the hyphenated ones
  489. # [20:19] <aaronlev> zcorpan_: because they can only be set dynamically
  490. # [20:19] <aaronlev> in text/html
  491. # [20:19] <zcorpan_> aaronlev: ok
  492. # [20:19] <aaronlev> so can the spec say to process those first?
  493. # [20:19] * zcorpan_ revamps the spec
  494. # [20:19] <zcorpan_> sure
  495. # [20:20] * Joins: kingryan (rking3@208.66.64.47)
  496. # [20:22] <aaronlev> sent you mail
  497. # [20:23] * Quits: mjs (mjs@64.81.48.145) (Quit: mjs)
  498. # [20:26] <zcorpan_> updated the spec
  499. # [20:26] <zcorpan_> now the email i just sent to www-html-editor no longer applies... :)
  500. # [20:26] <aaronlev> so now what
  501. # [20:27] <zcorpan_> we support html:role on non-html elements instead of xhtml2:role on non-html elements
  502. # [20:27] <zcorpan_> aaa:foo is checked before aria-foo
  503. # [20:29] <zcorpan_> i added an example to the note about unnamespaced "foo" attributes
  504. # [20:29] <zcorpan_> " (as in e.g. <div hidden="true"/>)"
  505. # [20:30] <zcorpan_> added you as a contributor
  506. # [20:30] <zcorpan_> do you have a suggestion for what to use instead of "foo"?
  507. # [20:31] <aaronlev> like [propertyname]
  508. # [20:31] <aaronlev> or reword it so the sentence doesn't need to refer to it that way
  509. # [20:32] <zcorpan_> changed it to "propertyname" -- does it read ok?
  510. # [20:33] <aaronlev> let me check in a few minutes
  511. # [20:33] <aaronlev> what should we do about the note to www-html-editor?
  512. # [20:34] <zcorpan_> i can reply to myself saying ignore this comment
  513. # [20:35] <aaronlev> before you do that
  514. # [20:35] <aaronlev> let me run the proposal by rich
  515. # [20:35] <zcorpan_> ok
  516. # [20:35] <aaronlev> and see if he's ok with what we're saying
  517. # [20:36] <zcorpan_> yup
  518. # [20:37] <aaronlev> should the spec say, in user agents that process xhtml2, they should process the xhtml2:role?
  519. # [20:38] <aaronlev> i mean, if you're actually using xhtml2, then xhtml2:role should work
  520. # [20:38] <aaronlev> otherwise the xhtml2 folks will complain about the proposal
  521. # [20:39] <zcorpan_> xhtml2 defines a role attribute in no namespace on elements in the http://www.w3.org/2002/06/xhtml2 namespace
  522. # [20:40] <zcorpan_> http://www.w3.org/2002/06/xhtml2/ even
  523. # [20:40] <zcorpan_> i could clarify that in the spec
  524. # [20:41] <aaronlev> so in html, xhtml 1.x, and xhml2, no namespace on the role attribute
  525. # [20:41] <zcorpan_> yeah
  526. # [20:41] <aaronlev> in anything else, use the xhtml/1999 namespace
  527. # [20:41] <zcorpan_> yes
  528. # [20:42] <aaronlev> let's clarify that part about xhtml2 so we don't get people there mad
  529. # [20:43] <aaronlev> ok, my colleague is fine with that
  530. # [20:43] <aaronlev> so i'm fine with you cancelling our note to html editors
  531. # [20:43] <zcorpan_> added a note about xhtml2
  532. # [20:43] <zcorpan_> good
  533. # [20:46] <zcorpan_> ok, sent
  534. # [20:47] <zcorpan_> aaronlev: the wairole prefix needs to be declared (authoring conformance criteria, not UA conformance criteria)
  535. # [20:48] <aaronlev> ok
  536. # [20:48] <aaronlev> so what's the point in hardcoding it then?
  537. # [20:48] <aaronlev> for use in xhtml?
  538. # [20:48] <aaronlev> i think i udnerstand -- just for consistency in case the author makes an error
  539. # [20:48] <zcorpan_> for compat with firefox 2
  540. # [20:48] <aaronlev> well for ff2 we only allowed that in text/html
  541. # [20:49] <aaronlev> but ok
  542. # [20:49] <zcorpan_> yeah, but that doesn't make the proposal incompatible with firefox 2
  543. # [20:50] <zcorpan_> by requiring a namespace decl for it, we forbid authors to use the wairole prefix declaratively in html (because html doesn't support namespace decls)
  544. # [20:53] * Quits: kingryan (rking3@208.66.64.47) (Ping timeout)
  545. # [20:55] * Quits: Roger (roger@213.64.74.230) (Quit: Roger)
  546. # [20:56] * Quits: gavin (gavin@63.245.208.169) (Ping timeout)
  547. # [20:59] * Joins: Roger (roger@213.64.74.230)
  548. # [21:00] <zcorpan_> ok, i'm heading home now to grab some food; will be back later tonight
  549. # [21:00] <zcorpan_> thanks for the help!
  550. # [21:00] * Joins: gavin (gavin@63.245.208.169)
  551. # [21:00] <aaronlev> zcorpan_: see you
  552. # [21:00] <aaronlev> when do you think i should run it by my colleague
  553. # [21:00] <aaronlev> ?
  554. # [21:01] <aaronlev> he's probably not ready to look at it yet, so i think i'll wait until you work on it more
  555. # [21:01] <zcorpan_> any time you like
  556. # [21:01] <aaronlev> ok
  557. # [21:01] <aaronlev> white timezone are you in?
  558. # [21:01] <zcorpan_> gmt+1
  559. # [21:01] <zcorpan_> the remaining two issues will require some research
  560. # [21:02] <zcorpan_> i might look into it tomorrow -- it's probably pretty much what [MAPPING] says
  561. # [21:02] <zcorpan_> if that's what you have implemented :)
  562. # [21:03] * zcorpan_ waves
  563. # [21:03] <Hixie> anne: no, it's not that simple (same origin); consider document.written() documents or data: documents
  564. # [21:06] * Joins: Philip_ (philip@80.177.163.133)
  565. # [21:06] * Quits: zcorpan_ (zcorpan@88.131.66.80) (Ping timeout)
  566. # [21:08] * Quits: Philip (philip@80.177.163.133) (Ping timeout)
  567. # [21:17] * Joins: mjs (mjs@17.203.14.225)
  568. # [21:22] * Joins: aaron (chatzilla@66.31.86.217)
  569. # [21:23] * Quits: aaronlev (chatzilla@66.31.86.217) (Ping timeout)
  570. # [21:23] * aaron is now known as aaronlev
  571. # [21:35] * Quits: Roger (roger@213.64.74.230) (Quit: Roger)
  572. # [21:42] * Joins: kingryan (rking3@208.66.64.47)
  573. # [21:43] * Quits: aaronlev (chatzilla@66.31.86.217) (Quit: ChatZilla 0.9.78.1 [Firefox 3.0a9pre/2007092504])
  574. # [21:43] * Joins: aaronlev (chatzilla@66.31.86.217)
  575. # [21:47] * Joins: zcorpan_ (zcorpan@83.227.34.9)
  576. # [21:49] <zcorpan_> aaronlev: i came to think of one way it might conflict with xhtml2, but only relevant for UAs that implement xhtml2
  577. # [21:49] <zcorpan_> if you specify both role and html:role on an element in the xhtml2 namespace
  578. # [21:49] <aaronlev> that should be unsupported
  579. # [21:50] <zcorpan_> the html:role should be ignored on xhtml2 elements if the ua supports xhtml2
  580. # [21:51] <aaronlev> right
  581. # [21:51] <aaronlev> agree
  582. # [21:53] <zcorpan_> fixed
  583. # [21:57] * Quits: ROBOd (robod@89.123.24.243) (Client exited)
  584. # [21:58] <zcorpan_> (that doesn't affect us though because we don't support xhtml2)
  585. # [21:59] <mjs> it seems problematic to predicate recognition of a namespaced "super-global" attribute on whether a language is supported or not
  586. # [21:59] <zcorpan_> mjs: hmm.. why?
  587. # [22:00] <mjs> zcorpan_: needless lack of interop between xhtml2 UAs and non-xhtml2 UAs
  588. # [22:00] <mjs> presuamblly the rules for the xhtml namespace are the same whether or not the UA "supports" xhtml1
  589. # [22:01] <zcorpan_> yeah. true
  590. # [22:01] <mjs> I think it would make more sense to base it on the element's namespace alone, not whether that namespace is supported
  591. # [22:01] <zcorpan_> yes, i'll change it
  592. # [22:02] <zcorpan_> done
  593. # [22:02] <zcorpan_> thanks
  594. # [22:05] * Quits: timbl (timbl@128.30.7.41) (Quit: timbl)
  595. # [22:06] * Joins: timbl_ (timbl@128.30.7.41)
  596. # [22:13] * Joins: ROBOd (robod@89.123.24.243)
  597. # [22:17] <aaronlev> zcorpan_: did you see rich's comment?
  598. # [22:23] <zcorpan_> no?
  599. # [22:25] * Quits: jgraham (jgraham@81.86.218.186) (Ping timeout)
  600. # [22:26] <aaronlev> i cc'd you, it was buried in there
  601. # [22:26] <aaronlev> now i've replied to him
  602. # [22:28] <zcorpan_> perhaps it hasn't hit my inbox yet -- sometimes it takes hours before emails arrive to my @opera.com address for some reason
  603. # [22:30] * Joins: jgraham (jgraham@81.86.223.139)
  604. # [22:37] <aaronlev> zcorpan_: rich is going to explain that we will need to propose prefixless role values in svg/xul/etc. to xhtml2 wg, because they own the role module spec
  605. # [22:37] <aaronlev> and what that means, if we have to go through them
  606. # [22:39] <zcorpan_> ok
  607. # [22:41] <zcorpan_> yeah, the aria stuff effectively overrides the unprefixed roles in the module spec
  608. # [22:43] <zcorpan_> if we still want them we can say that they are also aria roles
  609. # [22:44] <zcorpan_> but they don't seem very useful -- there are html5 elements that can be used for most of them :)
  610. # [22:45] * Quits: ROBOd (robod@89.123.24.243) (Quit: http://www.robodesign.ro )
  611. # [22:48] <zcorpan_> banner -> nothing, authors won't use it because it would be too easy for ad blockers to pick up on it
  612. # [22:48] <zcorpan_> contentinfo -> <footer>
  613. # [22:48] <zcorpan_> definition -> implied with <dfn> and context
  614. # [22:48] <zcorpan_> main -> <article>
  615. # [22:49] <zcorpan_> navigation -> <nav>
  616. # [22:49] <zcorpan_> note -> <aside>
  617. # [22:49] <zcorpan_> search -> none yet
  618. # [22:49] * Quits: Sander (svl@86.87.68.167) (Quit: And back he spurred like a madman, shrieking a curse to the sky.)
  619. # [22:49] <zcorpan_> secondary -> <aside>
  620. # [22:49] <zcorpan_> seealso -> <aside>?
  621. # [22:51] * Quits: gsnedders (gsnedders@86.137.237.196) (Quit: Don't touch /dev/null…)
  622. # [22:53] * Joins: gsnedders (gsnedders@86.137.237.196)
  623. # [22:55] <aaronlev> zcorpan_: agree, and i think the wai roles should be considered a superset of those
  624. # [22:55] <aaronlev> but rich does have a point
  625. # [22:55] <aaronlev> maybe we should take what we can get without stirring things up too much
  626. # [22:56] <zcorpan_> which means?
  627. # [22:57] <aaronlev> require a prefix for wai roles, when role is used on elements not in http://www.w3.org/1999/xhtml
  628. # [22:57] <aaronlev> becuase we can just argue that http://www.w3.org/1999/xhtml is in the html-wg domain
  629. # [22:58] <aaronlev> i think you probably didn't get the next few emails back and forth
  630. # [22:58] <aaronlev> yet
  631. # [22:59] <zcorpan_> ok. it still conflicts with the role module though, because the role attribute module defines some prefixless values
  632. # [22:59] <aaronlev> but none of the prefixless values are the same as wai roles
  633. # [23:00] <aaronlev> so just say if there is no prefix, it checks boths
  634. # [23:00] <aaronlev> both
  635. # [23:00] <aaronlev> the pf won't let there be a conflict between wai role names and role attribute names
  636. # [23:01] <Lachy> zcorpan_, role=banner -> <header>
  637. # [23:02] <zcorpan_> hmm. the spec as defined now will just ignore the wairole: prefix, so e.g. role="wairole:banner" is the same as role="banner"
  638. # [23:02] <Lachy> the description is poor, but it says "he banner content typically contains the site or company logo and ..."
  639. # [23:02] <zcorpan_> Lachy: ah, ok
  640. # [23:03] <aaronlev> zcorpan_: don't you think that's okay to propose that we treat the wairoles as a superset?
  641. # [23:03] <zcorpan_> aaronlev: yeah
  642. # [23:04] <zcorpan_> and also that the predefined roles in the module spec work with a wairole: prefix
  643. # [23:04] * Joins: mjs_ (mjs@17.255.111.173)
  644. # [23:04] * Quits: mjs (mjs@17.203.14.225) (Ping timeout)
  645. # [23:05] <aaronlev> zcorpan_: i'll catch you later
  646. # [23:05] <aaronlev> gotta go
  647. # [23:06] <zcorpan_> aaronlev: ok, cya
  648. # [23:06] * Quits: aaronlev (chatzilla@66.31.86.217) (Quit: ChatZilla 0.9.78.1 [Firefox 3.0a9pre/2007092504])
  649. # [23:15] * Quits: mjs_ (mjs@17.255.111.173) (Quit: mjs_)
  650. # [23:44] * Quits: zcorpan_ (zcorpan@83.227.34.9) (Ping timeout)
  651. # Session Close: Wed Sep 26 00:00:00 2007

The end :)