/irc-logs / w3c / #css / 2009-02-03 / end

Options:

  1. # Session Start: Tue Feb 03 00:00:00 2009
  2. # Session Ident: #css
  3. # [02:03] * Quits: shepazu (schepers@128.30.52.30) (Quit: shepazu)
  4. # [04:24] * Quits: dbaron (dbaron@63.245.220.241) (Quit: 8403864 bytes have been tenured, next gc will be global.)
  5. # [04:50] * Joins: dbaron (dbaron@98.234.51.190)
  6. # [04:54] * Quits: dbaron (dbaron@98.234.51.190) (Ping timeout)
  7. # [04:58] * Joins: dbaron (dbaron@98.234.51.190)
  8. # [05:16] * Quits: fantasai (fantasai@71.138.141.152) (Ping timeout)
  9. # [05:32] * Quits: dbaron (dbaron@98.234.51.190) (Ping timeout)
  10. # [05:41] * Joins: dbaron (dbaron@98.234.51.190)
  11. # [07:49] * Joins: shepazu (schepers@128.30.52.30)
  12. # [08:19] * Quits: dbaron (dbaron@98.234.51.190) (Quit: 8403864 bytes have been tenured, next gc will be global.)
  13. # [11:27] * Quits: Lachy (Lachlan@85.196.122.246) (Quit: This computer has gone to sleep)
  14. # [11:40] * Joins: Lachy (Lachlan@213.236.208.22)
  15. # [12:13] * Joins: myakura (myakura@221.191.119.20)
  16. # [16:06] * Joins: ChrisL (ChrisL@128.30.52.30)
  17. # [16:10] * Quits: shepazu (schepers@128.30.52.30) (Quit: shepazu)
  18. # [16:18] * Quits: Bert (bbos@mcclure.w3.org) (Ping timeout)
  19. # [16:35] * Joins: Bert (bbos@mcclure.w3.org)
  20. # [16:38] <anne> Bert, can you review http://tc.labs.opera.com/css/namespaces/syntax-012.xml ?
  21. # [16:38] <anne> I wonder why it fails in Gecko. I thought they had a pretty good CSS parser
  22. # [16:39] <anne> as in, the one that actually matches the specification
  23. # [16:39] <Bert> Can it wait half an hour?
  24. # [16:40] <anne> sure
  25. # [16:41] <Bert> By fail you mean it doesn't apply the CSS?
  26. # [16:42] <anne> it applies some CSS, but not the CSS it needs to apply to pass
  27. # [16:43] * anne trackbot
  28. # [16:43] <Bert> Hmm, all my browsers seem to apply the CSS. That's strange, the doc is sent as appplication/xml but there is no <?xml-stylesheet?> anywhere...
  29. # [16:44] <anne> hmm, my IRC client is acting weird
  30. # [16:44] <anne> ah, it's ok again
  31. # [16:44] <anne> Bert, namespace magic, the question is why it's not green
  32. # [16:46] <Bert> Namespaces don't have semantics in generic XML, they only separate names in different spaces. It's not said anywhere that html:link is a link, except in XHTML.
  33. # [16:46] <anne> can we place leave that discussion on the side?
  34. # [16:46] <anne> (HTML5 actually defines that, I think)
  35. # [16:47] <Bert> Sure, but please don't make tests that rely on sniffing.
  36. # [16:47] <Bert> The Web would be a rather less useful place if we generalized that awful concept.
  37. # [16:47] <anne> (it's not sniffing)
  38. # [16:47] <anne> (it's namespace dispatching while parsing)
  39. # [16:48] <Bert> Whatever it is, it is counter to the XML standard.
  40. # [16:48] <anne> not really, although it's true that nobody has written down the details
  41. # [16:49] <Bert> Why don't you send the doc as XHTML? Then it has semantics.
  42. # [16:49] <anne> I guess I'll ask someone else to review the test :)
  43. # [16:53] <Bert> Another thing that is wrong is the missing attribute on the <style> attribute. As far as I know, TYPE is required.
  44. # [16:58] <anne> not in HTML5
  45. # [17:00] <Bert> Namespace b is undefined. The url() contains invalid characters.
  46. # [17:01] <Bert> The document looks even less like HTML5 than like XHTML :-)
  47. # [17:03] <Bert> (Also, we can't make tests in HTML5, because HTML5 doesn't exist yet.)
  48. # [17:03] <anne> what invalid characters?
  49. # [17:03] <Bert> The single quotes.
  50. # [17:04] <anne> url('test') is correct, no?
  51. # [17:04] <anne> it's just preceded and followed by a comment, or is that not ok?
  52. # [17:05] <Bert> No comments inside tokens :-) We just clarified that text in the spec last month.
  53. # [17:06] <Bert> (It said comments may occur *between" tokens, it now says *outside* tokens.
  54. # [17:06] <Bert> )
  55. # [17:06] <anne> and url() is a token?
  56. # [17:06] <anne> k
  57. # [17:06] <Bert> Yes, unlike attr(), e.g.
  58. # [17:07] <anne> is url() inside background-image a token as well?
  59. # [17:08] <anne> s/inside/as value of/
  60. # [17:08] <Bert> Yes, same token. Tokenization is independent of property syntax.
  61. # [17:09] <anne> weird
  62. # [17:09] <anne> (and also understandable)
  63. # [17:11] <anne> whoa, changing that makes the whole thing apply
  64. # [17:11] * anne wonders if that's the correct error handling
  65. # [17:13] <anne> thanks Bert, I suppose I should now figure out what url(/**/''/**/) should do even though it is non-conforming
  66. # [17:13] <Bert> Haven't looked closely, but I'd say that the error recovery would just skip to the next ";" and continue parsing normally at @namespace c...
  67. # [17:15] <anne> that does not appear to be what is happening in e.g. Opera
  68. # [17:16] <anne> and what would the url() value be in that scenario?
  69. # [17:18] <Bert> There would be none. The line reads as ATKEYWORD(@namespace) COMMENT IDENT(b) COMMENT FUNCTION ("url(") COMMENT STRING('test-b') COMMENT ')' ';'
  70. # [17:18] <Bert> Which is well-formed, but doesn't mean anything.
  71. # [17:21] <anne> if it can be tokenized as such, why are comments disallowed there? (I suppose I should've followed the discussion...)
  72. # [17:22] <Bert> Maybe Opera ignores the last line because it doesn't kow what "b|" means?
  73. # [17:23] <Bert> It often can't be tokenized. URLs may contain characters that are otherwise not allowed in CSS, such as url(abc<<<def)
  74. # [17:24] <Bert> Or better example: url(abc[[[def) which has non-balanced brackets.
  75. # [17:25] <Bert> Or, which is actually likely to occur: url(path?q1=a;q2=b) with a semicolon.
  76. # [17:26] <anne> (removing b|test didn't change anything)
  77. # [17:26] <anne> (one of the first things I tried)
  78. # [17:27] <Bert> Adding a valid declaration of b does cause Opera to apply the green.
  79. # [17:27] <anne> sure
  80. # [17:27] <anne> url(abc[[[def) is acceptable?
  81. # [17:27] <anne> wow
  82. # [17:28] <Bert> And so does removing ", b|test"
  83. # [17:28] <anne> CSS syntax is even more silly than I thought
  84. # [17:28] <anne> really? if you keep the invalid rule around? that does not work for me
  85. # [17:28] <Bert> In CSS1, we didn't need quotes around URLs, the parens of url() were enough.
  86. # [17:29] <Bert> No, without the invalid rule. Haven't tried putting that back yet.
  87. # [17:29] <anne> right
  88. # [17:39] * Quits: myakura (myakura@221.191.119.20) (Quit: Leaving...)
  89. # [17:43] * Quits: Lachy (Lachlan@213.236.208.22) (Quit: This computer has gone to sleep)
  90. # [18:07] * Quits: ChrisL (ChrisL@128.30.52.30) (Client exited)
  91. # [19:37] * Joins: fantasai (fantasai@63.245.220.225)
  92. # [20:06] * Joins: dbaron (dbaron@63.245.220.225)
  93. # [20:23] * Joins: glazou (glazou@82.247.96.19)
  94. # [20:23] <glazou> hi
  95. # [20:27] <glazou> plinss, plinss_ : yt ?
  96. # [20:37] * Quits: dbaron (dbaron@63.245.220.225) (Quit: 8403864 bytes have been tenured, next gc will be global.)
  97. # [20:37] * Joins: dbaron (dbaron@63.245.220.241)
  98. # [20:44] * Quits: dbaron (dbaron@63.245.220.241) (Quit: 8403864 bytes have been tenured, next gc will be global.)
  99. # [20:44] * Joins: dbaron (dbaron@63.245.220.225)
  100. # [21:40] * Quits: glazou (glazou@82.247.96.19) (Quit: glazou)
  101. # [21:49] * Joins: Lachy (Lachlan@85.196.122.246)
  102. # [22:52] * Quits: dbaron (dbaron@63.245.220.225) (Quit: 8403864 bytes have been tenured, next gc will be global.)
  103. # [23:59] * Joins: dbaron (dbaron@63.245.220.241)
  104. # Session Close: Wed Feb 04 00:00:00 2009

The end :)