/irc-logs / w3c / #css / 2011-11-08 / end

Options:

  1. # Session Start: Tue Nov 08 00:00:00 2011
  2. # Session Ident: #css
  3. # [00:01] * Joins: drublic (drublic@93.132.230.78)
  4. # [00:07] * Quits: jdaggett (jdaggett@180.235.9.33) (Quit: jdaggett)
  5. # [00:08] * Joins: gilles (gilles@144.189.101.1)
  6. # [00:14] * Quits: tantek (tantek@63.245.220.240) (Quit: tantek)
  7. # [00:22] * Joins: Mike5_ (Mike5@mcclure.w3.org)
  8. # [00:28] * Quits: Mike5 (Mike5@mcclure.w3.org) (Ping timeout)
  9. # [00:28] * Mike5_ is now known as Mike5
  10. # [01:02] * Quits: gilles (gilles@144.189.101.1) (Quit: gilles)
  11. # [01:04] * Joins: gilles (gilles@144.189.101.1)
  12. # [01:05] * Quits: arno (arno@192.150.10.200) (Quit: Leaving.)
  13. # [01:06] * Joins: arno (arno@192.150.10.200)
  14. # [01:09] * Quits: drublic (drublic@93.132.230.78) (Client exited)
  15. # [01:10] * Quits: arronei (arronei@131.107.0.112) (Ping timeout)
  16. # [01:11] * Quits: arno (arno@192.150.10.200) (Quit: Leaving.)
  17. # [01:12] * Quits: gilles (gilles@144.189.101.1) (Quit: gilles)
  18. # [01:20] * Joins: arno (arno@192.150.10.200)
  19. # [01:21] * Quits: arno (arno@192.150.10.200) (Quit: Leaving.)
  20. # [01:24] * Joins: arno (arno@192.150.10.200)
  21. # [01:27] * Joins: dcosta_ (dcosta@187.31.77.7)
  22. # [01:27] * Quits: dcosta (dcosta@187.31.77.7) (Ping timeout)
  23. # [01:27] * dcosta_ is now known as dcosta
  24. # [02:05] * Quits: dbaron (dbaron@159.63.23.38) (Quit: 8403864 bytes have been tenured, next gc will be global.)
  25. # [02:07] * Joins: jdaggett (jdaggett@202.221.217.73)
  26. # [02:16] * Quits: nimbupani (Adium@24.18.47.160) (Quit: Leaving.)
  27. # [02:31] * Quits: arno (arno@192.150.10.200) (Quit: Leaving.)
  28. # [02:32] * Quits: lhnz (lhnz@188.223.83.48) (Ping timeout)
  29. # [02:33] * Joins: lhnz (lhnz@188.223.83.48)
  30. # [03:26] * Joins: miketaylr (miketaylr@24.42.93.245)
  31. # [03:44] * Joins: arno (arno@192.150.10.200)
  32. # [04:04] * Quits: arno (arno@192.150.10.200) (Quit: Leaving.)
  33. # [05:02] * Quits: bradk (bradk@99.7.175.117) (Quit: Get MacIrssi - http://www.sysctl.co.uk/projects/macirssi/ )
  34. # [05:19] * Joins: dbaron (dbaron@173.228.28.129)
  35. # [05:59] * Joins: nimbupani (Adium@24.18.47.160)
  36. # Session Close: Tue Nov 08 06:04:07 2011
  37. #
  38. # Session Start: Tue Nov 08 06:04:07 2011
  39. # Session Ident: #css
  40. # [06:04] * Disconnected
  41. # [06:05] * Attempting to rejoin channel #css
  42. # [06:05] * Rejoined channel #css
  43. # [06:05] * Topic is 'CSS Working Group | logged at http://krijnhoetmer.nl/irc-logs/'
  44. # [06:05] * Set by dbaron on Wed Oct 12 01:04:03
  45. # [06:22] * Joins: Mike5_ (Mike5@mcclure.w3.org)
  46. # [06:27] * Quits: Mike5 (Mike5@mcclure.w3.org) (Client exited)
  47. # [06:27] * Mike5_ is now known as Mike5
  48. # [06:38] * Joins: arno (arno@208.87.61.130)
  49. # [07:01] * Quits: miketaylr (miketaylr@24.42.93.245) (Quit: miketaylr)
  50. # [07:09] * Quits: stearns (anonymous@50.132.63.33) (Quit: stearns)
  51. # [07:10] * Quits: hober (ted@174.143.153.77) (Ping timeout)
  52. # [07:10] * Quits: gsnedders (gsnedders@204.232.194.186) (Ping timeout)
  53. # [07:10] * Joins: gsnedders (gsnedders@204.232.194.186)
  54. # [08:07] * Quits: dbaron (dbaron@173.228.28.129) (Quit: 8403864 bytes have been tenured, next gc will be global.)
  55. # [09:13] * Quits: fantasai (fantasai@69.162.163.148) (Connection reset by peer)
  56. # [09:14] * Joins: fantasai (fantasai@69.162.163.148)
  57. # [09:24] * Joins: drublic (drublic@93.132.237.88)
  58. # [09:37] * Quits: jdaggett (jdaggett@202.221.217.73) (Quit: jdaggett)
  59. # [09:47] * Joins: jamesn (jamesn@173.228.104.116)
  60. # [09:57] * Quits: jamesn (jamesn@173.228.104.116) (Quit: Colloquy for iPad - http://colloquy.mobi)
  61. # [10:48] * Quits: arronei_ (arronei@131.107.0.86) (Client exited)
  62. # [10:48] * Joins: arronei (arronei@131.107.0.85)
  63. # [11:39] * Quits: lhnz (lhnz@188.223.83.48) (Ping timeout)
  64. # [11:41] * Joins: lhnz (lhnz@188.223.83.48)
  65. # [12:14] * Quits: dholbert (dholbert@98.248.36.12) (Ping timeout)
  66. # [12:17] * Joins: dholbert (dholbert@98.248.36.12)
  67. # [12:23] * Joins: Mike5_ (Mike5@mcclure.w3.org)
  68. # [12:26] * Quits: dglazkov (u4270@88.198.6.68) (Ping timeout)
  69. # [12:28] * Quits: Mike5 (Mike5@mcclure.w3.org) (Ping timeout)
  70. # [12:28] * Mike5_ is now known as Mike5
  71. # [12:40] * Quits: dholbert (dholbert@98.248.36.12) (Ping timeout)
  72. # [12:41] * Joins: dholbert (dholbert@98.248.36.12)
  73. # [12:45] * Joins: dglazkov (u4270@88.198.6.68)
  74. # [13:43] * Quits: lhnz (lhnz@188.223.83.48) (Ping timeout)
  75. # [13:43] * Joins: lhnz (lhnz@188.223.83.48)
  76. # [14:27] * Joins: karl (karlcow@128.30.54.58)
  77. # [15:13] * Joins: miketaylr (miketaylr@24.42.93.245)
  78. # [15:32] * Joins: stearns (anonymous@50.132.63.33)
  79. # [15:39] * Quits: karl (karlcow@128.30.54.58) (Quit: Freedom - to walk free and own no superior.)
  80. # [16:20] * Joins: jamesn_ (jamesn@173.228.104.116)
  81. # [16:30] * Joins: anne (annevk@98.210.155.154)
  82. # [16:32] * Quits: anne (annevk@98.210.155.154) (Connection reset by peer)
  83. # [16:37] * Joins: hober (ted@174.143.153.77)
  84. # [16:37] * Quits: arno (arno@208.87.61.130) (Quit: Leaving.)
  85. # [16:40] * Quits: jamesn_ (jamesn@173.228.104.116) (Quit: Colloquy for iPad - http://colloquy.mobi)
  86. # [17:04] * Joins: arno (arno@192.150.10.200)
  87. # [17:45] * Quits: drublic (drublic@93.132.237.88) (Ping timeout)
  88. # [17:57] * Joins: gilles (gilles@144.189.101.1)
  89. # [18:10] * Quits: gilles (gilles@144.189.101.1) (Quit: gilles)
  90. # [18:15] * Joins: drublic (drublic@77.2.136.225)
  91. # [18:16] * Joins: dbaron (dbaron@173.228.28.129)
  92. # [18:23] * Joins: Mike5_ (Mike5@mcclure.w3.org)
  93. # [18:23] * Joins: Ms2ger (Ms2ger@91.181.45.148)
  94. # [18:28] * Quits: Mike5 (Mike5@mcclure.w3.org) (Ping timeout)
  95. # [18:28] * Mike5_ is now known as Mike5
  96. # [18:29] * Joins: gilles (gilles@144.189.101.1)
  97. # [18:32] * Quits: gilles (gilles@144.189.101.1) (Quit: gilles)
  98. # [18:38] * Joins: gilles (gilles@144.189.101.1)
  99. # [18:52] * Quits: stearns (anonymous@50.132.63.33) (Quit: stearns)
  100. # [19:05] * Quits: lhnz (lhnz@188.223.83.48) (Connection reset by peer)
  101. # [19:05] * Joins: lhnz (lhnz@188.223.83.48)
  102. # [19:32] * Quits: gilles (gilles@144.189.101.1) (Quit: gilles)
  103. # [19:32] * Quits: dbaron (dbaron@173.228.28.129) (Quit: 8403864 bytes have been tenured, next gc will be global.)
  104. # [19:35] * Quits: miketaylr (miketaylr@24.42.93.245) (Quit: miketaylr)
  105. # [19:38] * Joins: stearns (anonymous@192.150.22.5)
  106. # [19:46] * Joins: karl (karlcow@128.30.54.58)
  107. # [20:02] * Joins: tpod (tpod@159.63.23.38)
  108. # [20:02] * Joins: gilles (gilles@144.189.101.1)
  109. # [20:09] * Joins: dbaron (dbaron@63.245.220.240)
  110. # [20:26] * Quits: gilles (gilles@144.189.101.1) (Quit: gilles)
  111. # [20:57] * Joins: paul_irish (paul_irish@216.239.45.4)
  112. # [21:19] * Quits: brianman (brianman@131.107.0.76) (Ping timeout)
  113. # [21:20] * Joins: brianman (brianman@131.107.0.86)
  114. # [21:26] * Quits: tpod (tpod@159.63.23.38) (Ping timeout)
  115. # [21:36] * Joins: dcosta_ (dcosta@187.31.77.7)
  116. # [21:37] * Quits: dcosta (dcosta@187.31.77.7) (Ping timeout)
  117. # [21:37] * dcosta_ is now known as dcosta
  118. # [22:03] * Quits: Ms2ger (Ms2ger@91.181.45.148) (Quit: nn)
  119. # [22:05] * Joins: miketaylr (miketaylr@24.42.93.245)
  120. # [22:19] * Quits: dcosta (dcosta@187.31.77.7) (Ping timeout)
  121. # [22:24] * Joins: anne (annevk@72.14.229.87)
  122. # [23:05] * Joins: plinss_ (plinss@192.6.114.30)
  123. # [23:09] <brianman> Tab, you around?
  124. # [23:09] * fantasai waves
  125. # [23:09] <brianman> Hi Elika, Did my 4 issues capture the current status correctly?
  126. # [23:09] <fantasai> I think issue 1 is not an issue
  127. # [23:09] <fantasai> nobody thinks it should be allowed :)
  128. # [23:09] <fantasai> it doesn't make any sense
  129. # [23:09] <brianman> Let me relist them for archiving...
  130. # [23:10] <brianman> 1 - Brian - proposal to disallow "circle to 5px 7px"
  131. # [23:10] <brianman> 2 - Tab - proposal to disallow "circle to 5px"
  132. # [23:10] <brianman> 3 - Brian - proposal to disallow "ellipse to 5px"
  133. # [23:10] <brianman> 4 - Cover/Contain - Allow/remove for CSS3? Alias discussion has a mix. Current ED has them removed.
  134. # [23:10] <brianman> So at least three of us agree on 1.
  135. # [23:10] <fantasai> Yeah, so 1 is a non-issue. A circle with two radii is mathematically impossible per definition of a circle :)
  136. # [23:10] <brianman> Your thoughts on 2
  137. # [23:11] <fantasai> My thoughts are, we should disallow circle to <percentage> because we'd need to specify which dimension to take the percentage
  138. # [23:11] <fantasai> and that's a feature add
  139. # [23:11] <fantasai> But I'd prefer to leave circle to <length> in, since for authors if it's a circle, they'll think of it as being a circle
  140. # [23:11] <brianman> What about Tab's CSS4 idea ... circle auto 50px ?
  141. # [23:11] <fantasai> and not a subclass of ellipse
  142. # [23:11] <fantasai> that looks confusing to me
  143. # [23:12] <fantasai> I think that
  144. # [23:12] * Quits: karl (karlcow@128.30.54.58) (Quit: Freedom - to walk free and own no superior.)
  145. # [23:12] <fantasai> circle to 50% cover
  146. # [23:12] <fantasai> or circle to 50% contain
  147. # [23:12] <fantasai> would make more sense
  148. # [23:12] <fantasai> and be more useful
  149. # [23:12] <brianman> Well...he really said "circle to auto 50%", I removed the % temporarily.
  150. # [23:12] <fantasai> yeah, I still think that's confusing
  151. # [23:12] <fantasai> it implies that circle can take two radii
  152. # [23:12] <fantasai> which is nonsense
  153. # [23:13] <fantasai> But that's a discussion we can have later
  154. # [23:13] <brianman> well, that can be refined, but th emain point was whether to exclude "circle with #s" from CSS3 so we have flexibility in CSS4
  155. # [23:13] <fantasai> I'm pretty sure that we'll allow 'circle to <length>' in CSS4 as a subset of whatever it is we pick
  156. # [23:13] <fantasai> TabAtkins: ping
  157. # [23:14] <fantasai> So my reasoning is
  158. # [23:14] <fantasai> 'circle to <length>' is almost certainly a subset of what we'd want for CSS4
  159. # [23:14] <brianman> I can accept that.
  160. # [23:15] <fantasai> And for authors 'ellipse to <lengt>' where the <length> specifies a circle, is probably a little weird and unexpected
  161. # [23:15] <fantasai> since colloquially, we refer to circles as circles, and non-circular ellipses as ellipses
  162. # [23:15] <fantasai> even though a circle is a subclass of ellipse
  163. # [23:15] <brianman> so 3...you agree with disallow?
  164. # [23:16] <brianman> (not sure if that's a word, but whatev)
  165. # [23:16] <fantasai> for 3...
  166. # [23:16] <fantasai> I like disallowing 1 length
  167. # [23:16] <brianman> for ellipse
  168. # [23:16] <fantasai> but I'm not sure about disallowing 1 percentage
  169. # [23:16] <fantasai> yeah
  170. # [23:16] <fantasai> ellipse to 30% makes sense to me
  171. # [23:16] <fantasai> it will be elliptical
  172. # [23:16] <brianman> so "ellipse to 30% 30%" == "ellipse to 30%"?
  173. # [23:17] <fantasai> yeah
  174. # [23:17] <brianman> While that might be mildly convenient, I worry it's unnecessary and potentially confusing.
  175. # [23:17] <brianman> similar to avoiding % in circle
  176. # [23:17] <fantasai> % in circle has an ambiguity problem
  177. # [23:17] <brianman> also note current (WD) syntax doesn't allow single param ellipse
  178. # [23:17] <fantasai> we don't have that with ellipses
  179. # [23:18] <brianman> I worry things like inheritance and compute value produce confusion
  180. # [23:18] <brianman> if you inherit "ellipse to 30%" it may be not obvious what happens to the inheritor
  181. # [23:18] <fantasai> it's no less obvious than "ellipse to 30% 30%"
  182. # [23:18] <brianman> (namely that you expand the 30% to 30% 30%, then compute, then inherit)
  183. # [23:19] <fantasai> uh, no
  184. # [23:19] <brianman> (not compute, inherit, expand)
  185. # [23:19] <fantasai> inheritance happens on the percentage
  186. # [23:19] <fantasai> not on the resolved percentage
  187. # [23:19] <brianman> sorry...my brain was on EM
  188. # [23:19] <brianman> you're right
  189. # [23:19] <fantasai> :)
  190. # [23:19] <brianman> % are immune to that mess
  191. # [23:19] <brianman> (so far)
  192. # [23:19] <fantasai> no, they are. It used to not be the case, and we changed it :)
  193. # [23:19] <fantasai> I think font-size is an exception... but anyway
  194. # [23:20] <brianman> Anyway, I could live with it but I think we're safer excluding the "ellipse one-radius" entirely
  195. # [23:20] <brianman> 4 - did we have any progress on that one?
  196. # [23:21] <fantasai> Don't think so
  197. # [23:21] <brianman> I could go either way, as long as the 4 farthest and closest flavors live.
  198. # [23:21] <fantasai> ?
  199. # [23:22] <brianman> ok - <extent-implicit> = closest-corner | closest-side | farthest-corner | farthest-side
  200. # [23:22] <brianman> ok <extent-implicit> = closest-corner | closest-side | farthest-corner | farthest-side | cover | contain
  201. # [23:22] <brianman> bad - <extent-implicit> = closest-corner | contain | cover | farthest-side
  202. # [23:23] <brianman> clearer?
  203. # [23:24] <fantasai> yes
  204. # [23:25] <fantasai> so
  205. # [23:25] <fantasai> I don't like having aliases in CSS
  206. # [23:25] <fantasai> I think we generally try very hard to avoid them
  207. # [23:26] <fantasai> the first exception we made was for page-break-before
  208. # [23:26] <fantasai> and that's fairly recent
  209. # [23:26] <brianman> is auto often an alias?
  210. # [23:26] <fantasai> no
  211. # [23:26] <brianman> an alias for "do what css2 did"
  212. # [23:26] <fantasai> heh, no
  213. # [23:26] <fantasai> auto sometimes computes to a particular value
  214. # [23:27] <brianman> and "center" vs. "50%"?
  215. # [23:27] <fantasai> but it will depend on other things
  216. # [23:27] <fantasai> we do have keywords that compute to a numeric, yes
  217. # [23:28] <brianman> so it's specifically the keywordA => keywordB not keywordA => non-keyword-hardcoded-thing that bothers you?
  218. # [23:28] <fantasai> yeah
  219. # [23:28] <brianman> Not sure I agree, but ok.
  220. # [23:28] <fantasai> aliases aren't a solution to bikeshedding :)
  221. # [23:29] <fantasai> We use them to provide better usability (e.g. center/left/right vs. 50%/0%/100%)
  222. # [23:29] <brianman> left/right have additional value, relative to center
  223. # [23:29] <brianman> anchoring
  224. # [23:29] <brianman> right = "100% - offset"
  225. # [23:30] <brianman> but yes, I hear the readability argument
  226. # [23:30] <brianman> Wish I had Brad's number so we could progress further...
  227. # [23:31] <brianman> rewinding a little... I'm a little concerned about "ellipse %" allowed but "ellipse L" not allowed. I'd rather just disallow both, but I can live with it in the name of moving forward.
  228. # [23:32] <fantasai> I see the consistency argument, and I think we should allow both.
  229. # [23:32] <fantasai> Especially since the shape keyword is optional
  230. # [23:32] <fantasai> you might want to just write
  231. # [23:32] <fantasai> radial-gradient(to 5em, red, blue)
  232. # [23:32] <brianman> I think of that differently
  233. # [23:32] <brianman> I think of that as...
  234. # [23:32] <brianman> radial-gradient([circle]? to <length>, <color-stops>)
  235. # [23:33] <brianman> vs.
  236. # [23:33] <brianman> radial-gradient([ellipse]? to <length> <length>, <color-stops>)
  237. # [23:33] <fantasai> hm
  238. # [23:33] <brianman> I like both of those. I dislike the hybrids of those.
  239. # [23:33] <fantasai> i see
  240. # [23:34] <brianman> (and length or percentage for both)
  241. # [23:34] <fantasai> well, I don't feel strongly about 'ellipse to <percentage>' vs 'ellipse to <percentage>{2}'
  242. # [23:34] <brianman> but percentage for circle is tricky
  243. # [23:34] <fantasai> I think we should have brad / tab weigh in on that
  244. # [23:34] <brianman> ok
  245. # [23:34] <fantasai> yes, percentage for circle is tricky because there are multiple interpretations
  246. # [23:34] <brianman> But I think you get my mindset better on it now.
  247. # [23:35] <TabAtkins> Pong.
  248. # [23:35] <brianman> yay 3/4
  249. # [23:35] <TabAtkins> Sorry, was travelling to work.
  250. # [23:35] <brianman> Read up and then tell us when you're current.
  251. # [23:35] <fantasai> :)
  252. # [23:35] <brianman> While he's reading...
  253. # [23:35] <brianman> E - are we agree on farthest-corner lives?
  254. # [23:35] <brianman> (regardless of cover/contain conclusion)
  255. # [23:36] * Quits: dbaron (dbaron@63.245.220.240) (Ping timeout)
  256. # [23:36] <fantasai> I'm neutral on that point atm...
  257. # [23:36] <fantasai> but I can see that it would be good for CSS4
  258. # [23:36] <brianman> k
  259. # [23:36] <fantasai> I'd like to add an 'opposite' keyword in CSS4 to handle Brad's concerns wrt sizing
  260. # [23:36] <fantasai> So we could have
  261. # [23:36] <brianman> If we were agreed on keeping those two (f-c and c-s) then I'd say you three can decide on cover/contain among yourselves.
  262. # [23:36] <fantasai> radial-gradient(to opposite nearest-side, red, blue)
  263. # [23:37] <TabAtkins> current
  264. # [23:37] <brianman> "opposite nearest-side" == "farthest-side". tell me where i'm wrong?
  265. # [23:37] <brianman> assuming nearest==closest
  266. # [23:37] <fantasai> farthest-side might be in the other dimension
  267. # [23:37] <TabAtkins> I'd prefer "nearest-x-side", etc
  268. # [23:38] <brianman> closest*
  269. # [23:38] <TabAtkins> yeah
  270. # [23:38] <brianman> yah, that's what i recommended in email and then brad went dark
  271. # [23:38] * TabAtkins is trying to contain a curious kitten at his desk.
  272. # [23:38] <brianman> heh
  273. # [23:39] <brianman> here was the example from email...
  274. # [23:39] <brianman> background-image: -ms-radial-gradient(100px 175px, closest-side-y circle, black, black 2%, transparent 2%, transparent 98%, black 98%, black 100%, transparent);
  275. # [23:39] <brianman> back on 11/3
  276. # [23:39] <brianman> the other option is allowing two keywords
  277. # [23:39] <brianman> to mean x/y closest/farthest-ness
  278. # [23:40] <brianman> but only for the "side side" pairing
  279. # [23:40] <fantasai> brings up a question
  280. # [23:40] <TabAtkins> Yeah, I'd prefer just exlicitly naming at that point.
  281. # [23:40] <fantasai> should it be farthest-side or farthest side :)
  282. # [23:40] <TabAtkins> former
  283. # [23:40] <TabAtkins> We have precedent on that.
  284. # [23:41] <brianman> so Tab...on the 1,2,3 discussion above. your thoughts?
  285. # [23:42] <TabAtkins> disallow all 3
  286. # [23:42] <TabAtkins> for now, at least
  287. # [23:42] <brianman> Repeating in case it's useful...
  288. # [23:42] <brianman> radial-gradient([circle]? to <length>, <color-stops>)
  289. # [23:42] <brianman> radial-gradient([ellipse]? to <length> <length>, <color-stops>)
  290. # [23:42] <TabAtkins> disallow, since it's confusing that you can't use percentages yet
  291. # [23:44] <brianman> well you can use percentages in my second above
  292. # [23:44] <brianman> just not the first
  293. # [23:44] <TabAtkins> Yes.
  294. # [23:44] <TabAtkins> Sorry, was typing the "disallow" when I saw your first ine.
  295. # [23:44] <brianman> Right right.
  296. # [23:44] <TabAtkins> the second is obviously fine. ^_^
  297. # [23:45] <brianman> She went quiet.
  298. # [23:45] <brianman> While she gets her voice back...
  299. # [23:45] <brianman> So you think we should nuke cover contain or keep em?
  300. # [23:46] <TabAtkins> No real opinion, but they're easier to understand for many people, so slightly inclined to keep.
  301. # [23:46] <TabAtkins> And definitely keeping the 4 cloesest/farthest, for consistency.
  302. # [23:46] <brianman> ok, so sounds like we're on the same page.
  303. # [23:46] <brianman> A - c-c c-s f-c f-s
  304. # [23:46] <fantasai> I think we should push this issue to the WG... it's unrelated to the current syntax issue anyway
  305. # [23:47] <brianman> I'm fine with that
  306. # [23:47] <fantasai> we have the problem whether or not we're using comma syntax :)
  307. # [23:47] <brianman> As long as the 2x2 c/f-c/s stay I think we're good.
  308. # [23:47] <TabAtkins> yeah
  309. # [23:47] <brianman> So E - convince tab why "circle 5px" is worth keeping
  310. # [23:48] <fantasai> My reasoning is that
  311. # [23:48] <fantasai> a) it's a subset of the syntax we will accept in L4, however we choose to solve the percentage circle problem
  312. # [23:48] <fantasai> b) authors are unlikely to think of circles as a subset of ellipse -- even though they are, colloquially we only use 'ellipse' when referring to a non-circular ellipse
  313. # [23:49] <brianman> [c) no ambiguity in desired rendering]
  314. # [23:49] <TabAtkins> one sec
  315. # [23:49] <brianman> Tab - Devil's advocate arguments?
  316. # [23:49] <TabAtkins> k
  317. # [23:50] * Joins: dbaron (dbaron@63.245.220.240)
  318. # [23:50] * fantasai waves to dbaron
  319. # [23:50] <TabAtkins> Regardless of whether authors think of circles as subsets of ellipses, they will expect some parallel structure.
  320. # [23:50] <TabAtkins> Specifically, they'll expect to be able to use percentages, because they're allowed in ellipse.
  321. # [23:51] <TabAtkins> While we all certainly agree what "circle to <length>" means, it doesn't hurt anything to leave it until level 4 where we can solve it *properly*
  322. # [23:51] <fantasai> I think if we're allowing it, we should make it valid now
  323. # [23:51] <fantasai> and solve the percentage problem in L4
  324. # [23:51] <fantasai> which will then make the parallel structure parallel
  325. # [23:51] <TabAtkins> Why is it important to expose right now?
  326. # [23:51] <fantasai> because if it's only valid in L4, nobody will benefit from it
  327. # [23:52] <brianman> concurring somewhat with tab's assertion - I really don't look forward to resolving "By Design" the bugs that will get logged on "circle 3%" not working
  328. # [23:52] <fantasai> they'll all have to learn 'ellipse to 4em 4em' as the magic incantation for a circle
  329. # [23:52] <fantasai> and they will use that indefinitely
  330. # [23:52] <brianman> ew @ em
  331. # [23:52] <fantasai> partly due to that being what they learned, and partly due to that being more backwards-compat
  332. # [23:52] <TabAtkins> Sigh. That's somewhat compelling.
  333. # [23:53] <brianman> I would posit the following...
  334. # [23:53] <brianman> If we're going to allow this "partial" flavor for circle, it should be in the grammar right in everybody's face.
  335. # [23:53] <brianman> Not prose-covered.
  336. # [23:53] <TabAtkins> Yeah.
  337. # [23:53] <TabAtkins> Alternately, we bite the bullet and allow % circles.
  338. # [23:53] <brianman> Put another way "if it's worth it, it's worth having a bigger grammar block."
  339. # [23:54] <TabAtkins> Lock ourselves into a definition right now of % being resolved relative to width.
  340. # [23:54] <brianman> expanded functionality resolution...
  341. # [23:54] <TabAtkins> Believe me, I know. ^_^
  342. # [23:55] <fantasai> My suggestion for resolving circles in L4 would be to combine <percentage> and cover/contain
  343. # [23:55] <fantasai> I think that's more useful than picking one particular dimension
  344. # [23:55] <brianman> I find the suggestion confusing. sample usage?
  345. # [23:56] <fantasai> to 100% cover would calculate a cover circle
  346. # [23:56] <fantasai> to 100% contain would calculate a contained circle
  347. # [23:57] <brianman> hm...
  348. # [23:57] <brianman> that's another way of thinking about it that leads to killing cover and contain in css3 to leave css4 flexible
  349. # [23:58] <brianman> ok, so reigning it back a bit....
  350. # [23:58] <brianman> css3
  351. # [23:59] <brianman> recommend we keep "circle <percentage>" off the table for CSS3
  352. # [23:59] <TabAtkins> You're running the conversation well. ^_^
  353. # [23:59] <TabAtkins> kk
  354. # [23:59] <brianman> heh trying ;)
  355. # [23:59] <brianman> I can accept "circle <length>" if it's part of the grammar instead of the prose.
  356. # Session Close: Wed Nov 09 00:00:00 2011

The end :)