/irc-logs / w3c / #css / 2012-01-14 / end

Options:

  1. # Session Start: Sat Jan 14 00:00:00 2012
  2. # Session Ident: #css
  3. # [00:03] * Quits: sylvaing (sylvaing@131.107.0.82) (Quit: sylvaing)
  4. # [00:12] * Quits: dbaron (dbaron@159.63.23.38) (Quit: 8403864 bytes have been tenured, next gc will be global.)
  5. # [00:20] * Quits: arno (arno@192.150.10.200) (Quit: Leaving.)
  6. # [00:21] * Joins: arno (arno@192.150.10.200)
  7. # [00:22] * Quits: arno (arno@192.150.10.200) (Quit: Leaving.)
  8. # [00:26] * Quits: DaveWorth (rworth@72.83.231.22) (Quit: Leaving...)
  9. # [00:26] * Joins: rworth (rworth@72.83.231.22)
  10. # [00:26] * Quits: rworth (rworth@72.83.231.22) (Client exited)
  11. # [00:26] * Joins: ChrisL (ChrisL@128.30.52.169)
  12. # [00:31] * Joins: arno (arno@192.150.10.200)
  13. # [00:46] <Hixie> fantasai: yt?
  14. # [00:46] <Hixie> fantasai: do you see anything wrong with changing how we define <nobr> and <wbr> to:
  15. # [00:47] <Hixie> nobr { white-space: nowrap; } wbr { content: \200B; } nobr wbr { white-space: normal; }
  16. # [00:49] * Joins: dbaron (dbaron@159.63.23.38)
  17. # [00:52] * fantasai looks
  18. # [01:01] <tantek> Hixie, any chance you can make <wbr> not generate extraneous characters when a user copies a span of text with a <wbr> in the markup?
  19. # [01:01] <tantek> That one detail made it unusable for me in practice
  20. # [01:01] <tantek> (Safari inserts some crap in the copied plain text)
  21. # [01:02] <TabAtkins> ::selection wbr { content: ''; }
  22. # [01:02] <tantek> (also happens with &shy; - Safari inserts a hyphen when it shouldn't because &shy; are supposed to be invisible unless there actually *is* a break)
  23. # [01:02] <fantasai> Hixie: I think that should be okay rendering-wise, but I share Tantek's concern.
  24. # [01:02] <tantek> my guess is that it might come from the generated content
  25. # [01:02] <fantasai> TabAtkins: that wouldn't work, it would change the line-breaking on selection, forcing relayout
  26. # [01:02] <TabAtkins> True.
  27. # [01:03] <tantek> anyway as an author <wbr> is dead to me until it doesn't introduce crap into copy/paste of selections that include it
  28. # [01:03] <tantek> so it's broken in Safari for example
  29. # [01:03] <tantek> if the spec doesn't fix that, then feature will stay broken
  30. # [01:03] <TabAtkins> We probably just need a special white-space value that forces a soft break regardless of surrounding context.
  31. # [01:03] <tantek> and might as well deprecate it
  32. # [01:03] <tantek> noting the problems / buggy behavior that browsers do as the reason why it is recommended not to use it
  33. # [01:05] <fantasai> Hixie: I can't actually remember what <wbr> did outside <nobr>, if it was a line break opportunity on par with a space, or if it was a lower-priority break
  34. # [01:05] <fantasai> Hixie: Though it's probably not a web-compat concern at this point.
  35. # [01:05] * fantasai would need a copy of NS4 to test
  36. # [01:07] <fantasai> Tantek: wrt &shy; would you expect the text to contain U+00AD when copied?
  37. # [01:09] <tantek> no
  38. # [01:09] <tantek> because I don't see it
  39. # [01:09] <tantek> if there is no break
  40. # [01:09] <tantek> there is no extra character(s)
  41. # [01:09] <tantek> so there should be no crap in the copy/paste
  42. # [01:09] <fantasai> tantek, U+00AD is the soft hyphen
  43. # [01:10] <fantasai> it's an invisible character, like U+200B
  44. # [01:10] <fantasai> if it's in the source, do you expect it to be copied?
  45. # [01:10] * tantek is not a big believer in copying invisible data
  46. # [01:11] <tantek> typically results in unexpected outcomes
  47. # [01:11] <tantek> if you can't see it, you don't expect it
  48. # [01:11] <fantasai> I can tell you for sure that U+200B and several other invisble characters need to be copied
  49. # [01:11] <fantasai> in the general case
  50. # [01:12] <fantasai> some characters, like RLM, WJ, and ZWNJ are only visible in certain contexts
  51. # [01:12] <stearns> if it's in the data, I think it should be in the copy. In the case of &shy; the main problem is it gets in the data as a workaround for the lack of auto-hyphenation
  52. # [01:13] <tantek> to me, &shy; is a sort of <wbr> markup and thus should not be copied in a plain text copy/paste
  53. # [01:13] <tantek> since markup isn't copied in plain text copy/paste
  54. # [01:13] <fantasai> stearns: agree
  55. # [01:14] <tantek> basically, I've tried to use these to break-up too long URLs
  56. # [01:14] <tantek> in text that is posted
  57. # [01:14] <tantek> however when people copy that URL
  58. # [01:14] <tantek> and paste it into a browser
  59. # [01:14] <tantek> they get a bunch of extra crap characters
  60. # [01:14] <tantek> = URL fails to load
  61. # [01:14] <fantasai> Hixie: On a completely different topic, which is in fact related, we've run across a problem with people wanting to control which elements are selectable
  62. # [01:14] <tantek> that's my concrete use case
  63. # [01:15] <fantasai> Hixie: There's apparently CSS extensions to do that in webkit and moz; the webkit one at least seems to show significant usage
  64. # [01:15] <tantek> fantasai - URL?
  65. # [01:15] <fantasai> Hixie: But this seems to be a content issue, not a style issue, and probably should be handled by HTML
  66. # [01:15] <fantasai> Hixie: Connecting these two issues, if that "unselectable" mechanism was turned on by default for <WBR>, we've solved Tantek's problem :)
  67. # [01:16] <fantasai> tantek: http://www.google.com/search?q=user-select
  68. # [01:16] <fantasai> hm, looks like MS also has a version
  69. # [01:16] <stearns> tantek: auto-hyphenation should be able to break URLs at appropriate intervals
  70. # [01:16] <stearns> (but I'm not volunteering to write that spec)
  71. # [01:17] <tantek> ah ghost of user-select past
  72. # [01:17] <tantek> that was in a very early draft of CSS3-UI
  73. # [01:17] <tantek> might have even been in my CSS-UI proposal
  74. # [01:17] * tantek would have to check
  75. # [01:17] <tantek> stearns I want a way to put in manual optional breaks in text without having it cause any copy/paste crap
  76. # [01:18] <tantek> if you have a solution, please suggest it
  77. # [01:18] <tantek> since we can't depend on auto-hyphenation
  78. # [01:18] <tantek> I've tried <wbr> and &shy; and failed
  79. # [01:18] <tantek> alternatives welcome
  80. # [01:18] <tantek> I'm not bound to any one solution
  81. # [01:18] * fantasai suggests defining <wbr> as hixie suggests with select=off :)
  82. # [01:18] <tantek> I just want the problem solved
  83. # [01:18] <tantek> fantasai - some theoretical future technology spec is not a soluition
  84. # [01:19] <tantek> I need something to use on my site today
  85. # [01:19] <fantasai> tantek: theoretical future bugfix is also not available for your site today
  86. # [01:19] <tantek> not in 5 years when I can depend on browsers having to interoperably implement it
  87. # [01:19] <tantek> fantasai I'm asking in case folks know of other techniques that work today
  88. # [01:19] <tantek> if you don't, just say you don't, no need to pretend that future solution is the only way.
  89. # [01:20] <fantasai> sorry, you didn't specify a timeline
  90. # [01:20] <tantek> and bugfixes tend to happen faster than new features
  91. # [01:20] <tantek> timeline = sooner the better
  92. # [01:21] <stearns> tantek: there's code in hyphenator.js that strips &shy;s on copy (since it inserts a bazillion of them)
  93. # [01:21] <tantek> hm interesting
  94. # [01:21] <tantek> does it also insert them on load?
  95. # [01:21] <tantek> since I would want copy/paste to work in the noJS case
  96. # [01:21] <tantek> but thanks, that's a good start
  97. # [01:22] * Joins: leaverou_ (leaverou@79.166.96.120)
  98. # [01:25] * Quits: leaverou (leaverou@89.210.10.11) (Ping timeout)
  99. # [01:25] * leaverou_ is now known as leaverou
  100. # [01:32] * Quits: nimbu1 (Adium@157.22.251.133) (Quit: Leaving.)
  101. # [02:34] * Quits: dbaron (dbaron@159.63.23.38) (Quit: 8403864 bytes have been tenured, next gc will be global.)
  102. # [03:06] * Joins: nimbu (Adium@207.239.114.206)
  103. # [03:11] * Quits: nimbu (Adium@207.239.114.206) (Quit: Leaving.)
  104. # [03:19] * Quits: arno (arno@192.150.10.200) (Quit: Leaving.)
  105. # [03:20] * Joins: nimbu (Adium@207.239.114.206)
  106. # [03:37] * Quits: nimbu (Adium@207.239.114.206) (Quit: Leaving.)
  107. # [05:14] * Joins: arronei (arronei@131.107.0.89)
  108. # [05:14] * Quits: arronei_ (arronei@131.107.0.91) (Ping timeout)
  109. # [06:41] * Quits: cyril (chatzilla@203.41.202.66) (Quit: ChatZilla 0.9.87 [Firefox 3.6/20100115144158])
  110. # [06:46] * Quits: ChrisL (ChrisL@128.30.52.169) (Ping timeout)
  111. # [09:09] * Joins: Ms2ger (Ms2ger@91.181.210.181)
  112. # [09:58] * Joins: jarek (jarek@83.27.235.18)
  113. # [10:45] * Joins: tantek_ (tantek@70.36.139.219)
  114. # [10:45] * Quits: tantek (tantek@70.36.139.219) (Connection reset by peer)
  115. # [10:45] * tantek_ is now known as tantek
  116. # [11:04] * Quits: jarek (jarek@83.27.235.18) (Quit: jarek)
  117. # [12:43] * Quits: Ms2ger (Ms2ger@91.181.210.181) (Connection reset by peer)
  118. # [12:53] * Joins: Ms2ger (Ms2ger@91.181.210.181)
  119. # [13:53] * Quits: danielfilho (danielfilh@187.31.77.7) (Ping timeout)
  120. # [13:53] * Joins: danielfilho (danielfilh@187.31.77.7)
  121. # [16:05] * Quits: tantek (tantek@70.36.139.219) (Quit: tantek)
  122. # [17:52] * Joins: nimbu (Adium@24.18.47.160)
  123. # [21:33] * Quits: Ms2ger (Ms2ger@91.181.210.181) (Quit: nn)
  124. # Session Close: Sun Jan 15 00:00:00 2012

The end :)