/irc-logs / w3c / #css / 2014-05-16 / end

Options:

  1. # Session Start: Fri May 16 00:00:00 2014
  2. # Session Ident: #css
  3. # [00:03] * Joins: rhauck (~Adium@public.cloak)
  4. # [00:07] * Quits: Ms2ger (~Ms2ger@public.cloak) ("nn")
  5. # [00:16] * Joins: dholbert (~dholbert@public.cloak)
  6. # [00:16] <dholbert> fantasai, ping
  7. # [00:18] <dholbert> (or anyone else): I noticed some changes to my submitted flexbox reftests, as served by the CSS Flexbox CR test suite -- it looks like they've been converted from XHTML to HTML, and in so doing, some of the references were broken. Do you know where I can find out more / report issues on that?
  8. # [00:19] <dholbert> One example: the light blue "center" div here is not centered (in the reference case) because a <!-- --> comment in the <style> block is now nerfing the centering CSS rule that follows it: http://test.csswg.org/harness/test/css-flexbox-1_dev/flexbox-align-self-vert-4/ref/flexbox-align-self-vert-4-ref/
  9. # [00:21] <dholbert> I searched on shepherd, since I believe that's where these issues are tracked, but the only version of the file there is the correct (original XHTML) one: http://test.csswg.org/shepherd/reference/flexbox-align-self-vert-4-ref/name/flexbox-align-self-vert-4/
  10. # [00:21] <dholbert> so I don't know where the file that I'm seeing in the live testsuite is coming from
  11. # [00:23] <dholbert> (It does not exist in my up-to-date clone of http://hg.csswg.org/test , either -- that only has my correct (original XHTML) version of the reference case
  12. # [00:23] <dholbert> , too.)
  13. # [00:49] * Quits: rhauck (~Adium@public.cloak) ("Leaving.")
  14. # [00:53] * Joins: rhauck (~Adium@public.cloak)
  15. # [00:55] * Quits: plh (plehegar@public.cloak) ("Leaving")
  16. # [01:23] * Joins: jdaggett (~jdaggett@public.cloak)
  17. # [01:54] * Quits: lmclister (~lmclister@public.cloak) ("")
  18. # [02:56] * Joins: sh (~sh@public.cloak)
  19. # [02:56] * Quits: sh (~sh@public.cloak) ("Page closed")
  20. # [03:16] * Joins: dauwhe_ (~dauwhe@public.cloak)
  21. # [03:16] * Quits: dauwhe (~dauwhe@public.cloak) (Ping timeout: 180 seconds)
  22. # [03:31] * Quits: rhauck (~Adium@public.cloak) ("Leaving.")
  23. # [03:32] * Joins: rhauck (~Adium@public.cloak)
  24. # [03:39] * Quits: rhauck (~Adium@public.cloak) (Ping timeout: 180 seconds)
  25. # [04:29] * Quits: darktears (~darktears@public.cloak) (Client closed connection)
  26. # [04:36] * Joins: rhauck (~Adium@public.cloak)
  27. # [05:16] * Quits: rhauck (~Adium@public.cloak) ("Leaving.")
  28. # [05:17] * Joins: lmclister (~lmclister@public.cloak)
  29. # [05:20] * Joins: rhauck (~Adium@public.cloak)
  30. # [07:03] * Joins: dwim_cloud (~uid10661@public.cloak)
  31. # [07:11] * Quits: lmclister (~lmclister@public.cloak) ("")
  32. # [07:29] * Quits: rhauck (~Adium@public.cloak) ("Leaving.")
  33. # [08:16] * Joins: zcorpan (~zcorpan@public.cloak)
  34. # [08:48] * Joins: lmclister (~lmclister@public.cloak)
  35. # [08:58] * Joins: shans__ (~shans_@public.cloak)
  36. # [09:03] * Quits: lmclister (~lmclister@public.cloak) ("")
  37. # [09:11] * Quits: dwim_cloud (~uid10661@public.cloak) ("Connection closed for inactivity")
  38. # [09:17] * Joins: rodneyrehm (~rrehm@public.cloak)
  39. # [09:38] * Joins: Ms2ger (~Ms2ger@public.cloak)
  40. # [10:19] * Quits: jdaggett (~jdaggett@public.cloak) (Ping timeout: 180 seconds)
  41. # [10:21] * Quits: ed (~ed@public.cloak) ("ZNC - http://znc.sourceforge.net")
  42. # [10:29] * Joins: ed (~ed@public.cloak)
  43. # [10:31] * Quits: shans__ (~shans_@public.cloak) (Ping timeout: 180 seconds)
  44. # [10:40] * Joins: shans__ (~shans_@public.cloak)
  45. # [11:04] * Quits: shans__ (~shans_@public.cloak) (Ping timeout: 180 seconds)
  46. # [11:07] * Joins: shans__ (~shans_@public.cloak)
  47. # [11:13] * Quits: shans__ (~shans_@public.cloak) (Client closed connection)
  48. # [11:13] * Joins: shans__ (~shans_@public.cloak)
  49. # [11:21] * Quits: shans__ (~shans_@public.cloak) (Ping timeout: 180 seconds)
  50. # [11:24] * Joins: shans__ (~shans_@public.cloak)
  51. # [11:44] * Quits: shans__ (~shans_@public.cloak) (Ping timeout: 180 seconds)
  52. # [12:24] * Quits: rodneyrehm (~rrehm@public.cloak) ("This computer has gone to sleep")
  53. # [13:07] * Joins: darktears (~darktears@public.cloak)
  54. # [13:08] * Joins: rodneyrehm (~rrehm@public.cloak)
  55. # [13:20] * Quits: Ms2ger (~Ms2ger@public.cloak) (Ping timeout: 180 seconds)
  56. # [14:04] * Joins: Ms2ger (~Ms2ger@public.cloak)
  57. # [14:36] * Quits: rodneyrehm (~rrehm@public.cloak) ("leaving planet earth")
  58. # [15:37] * Quits: zcorpan (~zcorpan@public.cloak) (Client closed connection)
  59. # [15:37] * Joins: zcorpan (~zcorpan@public.cloak)
  60. # [15:44] * Quits: zcorpan (~zcorpan@public.cloak) (Ping timeout: 180 seconds)
  61. # [16:12] * Joins: shans__ (~shans_@public.cloak)
  62. # [16:46] * Quits: darktears (~darktears@public.cloak) ("Linkinus - http://linkinus.com")
  63. # [16:46] <SimonSapin> From the F2F wiki page: " CSS Grid to PR (last call ended, CR exit criteria met)
  64. # [16:47] <SimonSapin> Really? http://www.w3.org/TR/css-grid-1/ is a WD. Regardless, there are still unresolved issues in the ML
  65. # [17:01] * Quits: shans__ (~shans_@public.cloak) (Ping timeout: 180 seconds)
  66. # [17:16] * Quits: Ms2ger (~Ms2ger@public.cloak) ("nn")
  67. # [18:23] * Joins: lmclister (~lmclister@public.cloak)
  68. # [19:16] * Joins: rhauck (~Adium@public.cloak)
  69. # [19:38] * Joins: zcorpan (~zcorpan@public.cloak)
  70. # [19:45] * Quits: zcorpan (~zcorpan@public.cloak) (Ping timeout: 180 seconds)
  71. # [21:08] * Joins: jet (~junglecode@public.cloak)
  72. # [21:21] * Quits: rhauck (~Adium@public.cloak) ("Leaving.")
  73. # [21:31] * Quits: jet (~junglecode@public.cloak) (jet)
  74. # [21:37] <dauwhe_> liam: ping
  75. # [21:38] <liam> dauwhe_, pong
  76. # [21:38] <dauwhe_> thinking about footnotes and regions
  77. # [21:38] <dauwhe_> and their relation to @page
  78. # [21:39] <dauwhe_> since it's hard to separate footnotes from pagination
  79. # [21:39] <dauwhe_> did you see that Peter Sorotokin has built an implementation of the old page template idea in epub: http://sorotokin.com/adaptive-layout/
  80. # [21:40] <dauwhe_> It uses regions to move content around
  81. # [21:40] <liam> negative, how are you finding these things? :-)
  82. # [21:40] <liam> i tried books.js but didn't get it to work in chrome, i blame bitrot
  83. # [21:40] <dauwhe_> book.js works in webkit nightly
  84. # [21:40] <dauwhe_> didn't work in Chrome for me, either, even with all the flags turned on
  85. # [21:41] <liam> s/chrome/chromium/ to be fair
  86. # [21:41] <dauwhe_> I'm not up to speed on the history of page slots/page templates in CSS. Peter is using @ident inside @page to define the regions
  87. # [21:42] <liam> templates good, I use XSLT a lot :-)
  88. # [21:42] <dauwhe_> which is nicely similar to @footnote inside @page in GCPM, and matches up with how CSS3-page does running headers
  89. # [21:42] <liam> yeah, i am into doing things in similar ways
  90. # [21:43] <liam> I tried to push for border images to work like the page header regions too
  91. # [21:43] <dauwhe_> I'm just hoping that what we do for footnotes could be based on defining the regions inside @page...
  92. # [21:43] <liam> (and got support, including from Tab, but not yet)
  93. # [21:43] <liam> yes, I would like that
  94. # [21:43] <dauwhe_> ... rather than the book.js idea of just building lots of divs with JS to flow into.
  95. # [21:43] <liam> it's also how we did it for XSL-FO, and it does make it a lot easier to build authring tools
  96. # [21:44] <dauwhe_> excellent.
  97. # [21:44] * Joins: zcorpan (~zcorpan@public.cloak)
  98. # [21:44] <dauwhe_> Another issue is the footnote reference. book.js has to wrap footnotes in three spans. The inner span gets flow-into, the outer spans are needed to create the reference
  99. # [21:44] <liam> and to support marginalia and e.g. both per-column footnotes and page-level footnotes, each with possible different formatting
  100. # [21:45] <dauwhe_> unfortunately you can't send ::before to a different flow than the element itself :)
  101. # [21:45] <liam> i had wanted to go to the f2f to discuss this with you & alan, but didn't get travel approval so will try & be on the 'phone - maybe even a google hangout will work so i can see video
  102. # [21:45] <dauwhe_> absolute positioning of page margin boxes inside @page might address some of Daniels issues with the GCPM/CSS3 page model.
  103. # [21:46] <liam> for marginalia you need relative positioning too
  104. # [21:47] <dauwhe_> True; I was just thinking of something beyond the existing model where the margin boxes are locked in place
  105. # [21:47] <liam> this adaptive layout is nice but slow, and we need it to be accessible and search-engine-visible... which raises whole other questions about jumping directly to page 6
  106. # [21:48] <liam> https://www.w3.org/Style/XSL/Group/2008/06/footnote-examples/pages/VisibleLanguage-31-2-140/ has some marginalia
  107. # [21:48] <liam> and https://www.w3.org/Style/XSL/Group/2008/06/footnote-examples/pages/Kieckhefer-ForbiddenRites-276/ has per-column footnotes
  108. # [21:49] <liam> and you have an example (and so do I in that bunch) of footnotes that behave like CSS-floated inline blocks with auto width and wide margins, i.e. inline if there's room
  109. # [21:49] <dauwhe_> In that first link, is the top of the marginal note defined to place two lines down from the reference?
  110. # [21:50] <liam> I believe so yes
  111. # [21:50] <dauwhe_> An interesting choice :)
  112. # [21:50] <liam> yeah, it was "arty" I expect
  113. # [21:51] <dauwhe_> If it was really arty, a note on the last line of the last page would appear at the beginning of the book--the circle is unbroken :)
  114. # [21:51] <liam> the reference to the figure on the right is not explicit but it's probably anchored to the word "work" 2 lines above
  115. # [21:51] <liam> :)
  116. # [21:51] <liam> I looked at some other issues of the journal at MIT press bookstore in Cambridge MA but they are not so interesting
  117. # [21:52] <dauwhe_> That second link will give me nightmares. Parallel texts with multiple independent footnote streams
  118. # [21:52] <liam> it's not too bad, the designer left plenty o froom between the body copy and the footnote region
  119. # [21:52] <dauwhe_> easy enough to typeset by hand. not so easy for us.
  120. # [21:53] <liam> for XSL we agreeed https://www.w3.org/Style/XSL/Group/2008/06/footnote-examples/pages/Charles-II-Pseudepigrapha-Ahikar-745/ and https://www.w3.org/Style/XSL/Group/2008/06/footnote-examples/pages/Charles-II-Pseudepigrapha-Ahikar-747/ were out of scope
  121. # [21:53] <liam> oh,a nd https://www.w3.org/Style/XSL/Group/2008/06/footnote-examples/pages/Charles-II-Pseudepigrapha-IV-Ezra-4-566/ from the same book
  122. # [21:53] <liam> oops got to go
  123. # [21:53] <dauwhe_> No worries. Thanks for the discussion!
  124. # [21:54] * dauwhe_ is now known as dauwhe
  125. # [22:00] * Quits: dauwhe (~dauwhe@public.cloak) (Client closed connection)
  126. # [22:00] * Joins: dauwhe (~dauwhe@public.cloak)
  127. # [22:09] * Quits: zcorpan (~zcorpan@public.cloak) (Client closed connection)
  128. # [22:09] * Joins: zcorpan (~zcorpan@public.cloak)
  129. # [22:10] * liam back
  130. # [22:11] * Joins: rhauck (~Adium@public.cloak)
  131. # [22:15] <dauwhe> Thanks for the posts on initial caps. Very helpful.
  132. # [22:16] * Quits: zcorpan (~zcorpan@public.cloak) (Ping timeout: 180 seconds)
  133. # [22:20] * Joins: zcorpan (~zcorpan@public.cloak)
  134. # [22:24] * Quits: zcorpan (~zcorpan@public.cloak) (Client closed connection)
  135. # [22:24] <liam> :)
  136. # [22:25] * Joins: zcorpan (~zcorpan@public.cloak)
  137. # [22:25] <liam> i've been talking to people about the nearest Farsi equivalent to a drop cap (there isnt' one really, but software must not explode when it's asked for)
  138. # [22:26] <liam> and i'm glad you found my text in the XSL-FO 2 draft
  139. # [22:26] <liam> I hope to get something written up about footnotes and regions too, today or tomorrow.
  140. # [22:29] <dauwhe> I have a new draft in progress of the initial cap section in CSS Inline. Hoping to have something up during the F2F.
  141. # [22:29] <liam> awesome
  142. # [22:30] <dauwhe> I wonder if at least some images would work with similar coding to the current ED.
  143. # [22:30] <liam> I have *not* talked about accessibility
  144. # [22:30] <liam> probably, but probably not the one in my blog example, which is why I chose it
  145. # [22:31] <dauwhe> I've seen quite a few cases where you'd just need to align the top of the img with the cap height of first line, and the bottom with the baseline of the Nth line.
  146. # [22:31] <dauwhe> That wouldn't help for your recent example, though.
  147. # [22:31] <liam> (it's not an uncommon use case to have the image extend; James' example with accents is an excellent one)
  148. # [22:32] * Quits: zcorpan (~zcorpan@public.cloak) (Ping timeout: 180 seconds)
  149. # [22:32] <dauwhe> I wonder if the lengths defined there could be used to tweak things appropriately.
  150. # [22:32] * Joins: zcorpan (~zcorpan@public.cloak)
  151. # [22:38] * Quits: zcorpan (~zcorpan@public.cloak) (Client closed connection)
  152. # [22:39] * Joins: zcorpan (~zcorpan@public.cloak)
  153. # [22:40] <liam> for Latin scripts, you use the cap height of the font for the drop cap, and the cap height of the font for running text, and align those, and align the baselines appropriately, and then leave room above and below as needed; that method works for accents too
  154. # [22:41] <liam> for images, you'd need to identify the alignment points somehow
  155. # [22:41] <liam> I do also have examples where the text comes as close as possible to the initial on each line, but I think those are better treated as a graphic runaround, which I think Sylvain has been working on
  156. # [22:43] <dauwhe> agreed on runarounds. sounds like a job for shapes/exclusions.
  157. # [22:46] * Quits: zcorpan (~zcorpan@public.cloak) (Ping timeout: 180 seconds)
  158. # [22:46] <dauwhe> Changing the subject, I'm still thinking about how to get a footnote reference with regions.
  159. # [22:48] <dauwhe> Could we extend the ::marker pseudo-element? if you use ::marker on an element, it stays behind even if the element flows into a region?
  160. # [22:49] <dauwhe> Nah, that wouldn't work for lists.
  161. # [22:55] <liam> given markup like <aside class="footnote" id="fn1"><span class="marker">*</span><span class="text">This is a footnote</span></aside>
  162. # [22:55] <liam> (I know this isn't possible because you don't know the marker, but bear with me)
  163. # [22:55] <liam> we can put the text of the footnote into the footnote region, and use ::before on that to get a marker
  164. # [22:56] <dauwhe> We actually do something like that with Prince; we just replace the content of .marker with the content property
  165. # [22:56] <liam> so then the question is how to generate the marker in the text (instead of supplying it) right?
  166. # [22:56] <liam> yes
  167. # [22:56] <dauwhe> Prince allows content everywhere, not just on pseudos.
  168. # [22:57] <dauwhe> Yep, the marker on the footnote itself is easy.
  169. # [22:57] <liam> i thought browsers did too. I really want to see a way to style content boxes, so content: "happy" "boy"; can have the "boy" in red
  170. # [22:57] <dauwhe> I forget what they do. Some epub reading systems don't like it, of course.
  171. # [22:59] <liam> so maybe we need to extend the flow-into-region concept to be able to leave a remnant behind
  172. # [22:59] <liam> ::remnant
  173. # [22:59] <dauwhe> ^_^
  174. # [22:59] <dauwhe> it's what's left over after demolishing the bikeshed :)
  175. # [23:01] <liam> :)
  176. # [23:03] <dauwhe> That reminds me of a larger issue, having a pointer to an object rather than the object itself. Complex footnotes are a markup problem in HTML because we can't put complex block structures inside p.
  177. # [23:04] <dauwhe> We actually code footnote refs separately from footnotes. Ref is a span at the callout point, FN itself is a div after the para.
  178. # [23:04] <liam> xsl:retrieve-marker :-)
  179. # [23:04] <dauwhe> With some trickery, Prince places the footnote in the correct place.
  180. # [23:05] <liam> the same often has to be done with figures - e.g. in journals where figures must be at the top of the column in which they are first referenced (and can span both cols if needed)
  181. # [23:05] <liam> formatters often cannot handle "backward floats"
  182. # [23:05] <dauwhe> Exactly. He who shall not be named has proposed solutions like target-pull()
  183. # [23:06] <liam> Våldon :) :)
  184. # [23:07] <dauwhe> A generalization of such an idea seems like it could be broadly useful in HTML, not just for bookish folks.
  185. # [23:07] <liam> retrieve marker can work with local or remote formatting, too - or a combination, as needed for a back-of-the-book index
  186. # [23:07] <liam> it's very useful
  187. # [23:08] <dauwhe> Gotta run. Thanks for the IRCing.
  188. # [23:08] <liam> ok, thank you too
  189. # [23:09] * liam hopes to be on the 'phone for at least part of the f2f
  190. # [23:33] * Quits: lmclister (~lmclister@public.cloak) (Client closed connection)
  191. # [23:33] * Joins: lmclister (~lmclister@public.cloak)
  192. # Session Close: Sat May 17 00:00:00 2014

The end :)