/irc-logs / w3c / #css / 2012-04-06 / end

Options:

  1. # Session Start: Fri Apr 06 00:00:01 2012
  2. # Session Ident: #css
  3. # [00:40] * Joins: dbaron (dbaron@159.63.23.38)
  4. # [02:02] * Quits: drublic (drublic@77.2.134.18) (Client exited)
  5. # [02:10] * Joins: jdaggett (jdaggett@202.221.217.73)
  6. # [02:12] <TabAtkins> sylvaing: Pretty sure our incorrect handling of 3d rotations is a known bug.
  7. # [02:17] * Joins: divya1 (Adium@192.150.10.200)
  8. # [02:28] * Quits: divya1 (Adium@192.150.10.200) (Quit: Leaving.)
  9. # [03:01] * Joins: glenn (gadams@174.29.112.105)
  10. # [03:15] * Quits: arno (arno@192.150.10.200) (Quit: Leaving.)
  11. # [03:40] * Quits: dbaron (dbaron@159.63.23.38) (Quit: 8403864 bytes have been tenured, next gc will be global.)
  12. # [04:04] * Quits: glenn (gadams@174.29.112.105) (Client exited)
  13. # [04:11] * Quits: jet (jet@159.63.23.38) (Quit: jet)
  14. # [04:33] * Joins: glenn (gadams@174.29.112.105)
  15. # [06:58] * Quits: paul___irish (paul___iri@205.186.165.150) (Ping timeout)
  16. # [07:25] * Quits: logbot (logbot@110.173.227.145) (Ping timeout)
  17. # [07:28] * Joins: logbot (logbot@110.173.227.145)
  18. # [08:20] * Joins: tantek (tantek@70.36.139.112)
  19. # [08:25] * Joins: christianz (62a4d766@78.129.202.38)
  20. # [08:26] <christianz> can i propose CSS4 properties here?
  21. # [08:27] <fantasai> no, send them to www-style
  22. # [08:27] <fantasai> (you can, but they won't be tracked very well, so are unlikely to get much response)
  23. # [08:27] <christianz> i'm getting referred all around; Tantek told me to come here
  24. # [08:27] <christianz> is www-style another IRC channel?
  25. # [08:29] <tantek> christianz - this is a good place for quick informal discussions
  26. # [08:30] <tantek> which I'd say is a good thing to do before spending the time on proposals
  27. # [08:31] <fantasai> tantek: you should warn people that discussions here could wind up being as asynchronous as email... not a very high chance of getting an immediate reply
  28. # [08:32] <fantasai> esp not during Pacific work hours
  29. # [08:32] <tantek> fantasai - if you think such a warning is necessary, perhaps you could add it to http://wiki.csswg.org/communications ;)
  30. # [08:32] <fantasai> tantek: I don't think it's necessary
  31. # [08:33] <fantasai> tantek: It's only necessary if you are asking people to come here
  32. # [08:33] <fantasai> tantek: in order to get answers to questions
  33. # [08:33] <fantasai> tantek: or a response to a proposal
  34. # [08:33] <fantasai> tantek: or whatever
  35. # [08:33] <tantek> that page directs people to come here too
  36. # [08:33] <fantasai> tantek: yeah, but it doesn't give them any expectation of getting a response
  37. # [08:33] <fantasai> tantek: it just says where our IRC discussions happen
  38. # [08:33] <tantek> it's a reasonable IRC default to get answers to questions etc.
  39. # [08:34] <tantek> so if you think further context/warning would be helpful, go ahead and add it to the communications page
  40. # [08:34] <fantasai> I think the comm page is fine
  41. # [08:34] <christianz> uh oh, i started a fight
  42. # [08:35] * fantasai has no idea what tantek was thinking, asking someone to come here and expect a response at 11:30pm on a Thursday night, unless he was planning to respond himself...
  43. # [08:37] <tantek> hey fantasai - if you think something is necessary, then do it, rather than asking others to do so.
  44. # [08:38] <fantasai> I did. I suggested christianz send proposals to www-style.
  45. # [08:38] <fantasai> Since nothing here gets tracked formally.
  46. # [08:39] <tantek> and I suggested that it's better to have quick short discussions first rather than spend time first writing up a proposal for www-style
  47. # [08:39] <tantek> the irc channel is archived
  48. # [08:39] <tantek> per /topic
  49. # [08:39] <fantasai> great, you two have a nice discussion, I think most other people are asleep.
  50. # [08:39] * Joins: jet (jet@67.169.43.128)
  51. # [08:40] <tantek> I see no reason to discourage discussion, nor make snarky remarks about having one.
  52. # [08:41] <tantek> specifically, *public* discussion
  53. # [08:41] <christianz> i don't know that you did wrong by sending me here
  54. # [08:41] <fantasai> it's fine as long as you're ok with most people being asleep
  55. # [08:41] <christianz> i just thought i would see who might be on now
  56. # [08:41] <fantasai> the channel is pretty biased towards Pacific time
  57. # [08:41] <christianz> you never know when we are dealing globally
  58. # [08:42] <tantek> right
  59. # [08:42] <christianz> yeah, and I'm in the Pacific time zone, so i know what you mean
  60. # [08:42] <leaverou> for 20 minutes you guys are discussing where the proposal should be posted. I'm not saying anyone in particular is at fault, but this is the complete opposite of efficiency
  61. # [08:43] <tantek> and often times people are awake much past (or before) their default/expected/probabilistic time zones
  62. # [08:43] <christianz> partly my fault; longtime CSS coder stumbling around trying to figure out how to submit my suggestion
  63. # [08:43] <tantek> leaverou - indeed
  64. # [08:43] <christianz> "and often times people are awake much past (or before) their default/expected/probabilistic time zones" true that
  65. # [08:43] <fantasai> yep, I'm normally not here at this hour :)
  66. # [08:44] <christianz> glad i caught you; and the other two
  67. # [08:44] <leaverou> tantek: FWIW I do agree with fantasai that it would be best if he was sent to www-style, as more people would see it and that's where proposals should be posted.
  68. # [08:44] <fantasai> I just don't want people giving advice that results in proposals being posted here, not getting any response, and never making it to www-style
  69. # [08:44] <tantek> leaverou - is it more desirable to have more people see something first? or to spend fewer people's time first? which of those two is more efficient overall in terms of # of people's time spent?
  70. # [08:44] <leaverou> but since he came here, it's kinda inefficient to debate this for half an hour, I guess one of us could forward it to the mailing list if christianz doesn't want to post there
  71. # [08:45] <christianz> then the point is for me to know posting my suggestion here might be better posted elsewhere; i'm fine with that
  72. # [08:45] <tantek> and in my experience, most "proposals" tend to be premature and benefit from bouncing ideas off a smaller crowd before writing up anything formal for a mailing list
  73. # [08:45] <christianz> then i'll post my suggestion here
  74. # [08:45] <christianz> and now
  75. # [08:46] <leaverou> tantek: Not every subscriber in the mailing list reads everything, so I don't see how it would be wasting their time. Most people after a certain time develop their own way to filter stuff quickly
  76. # [08:46] <christianz> my idea is for border-inset and border-offset properties to mimic what they do in print where sometimes a box's border (solid, dotted, dashed or whatever) is actually a little inside or outside the box
  77. # [08:46] <leaverou> tantek: true. But IRC is expected to be more synchronous. And an idea can be refined just fine on the mailing list, if it's good
  78. # [08:47] <leaverou> christianz: have you tried outline and outline-offset?
  79. # [08:47] <christianz> that effect can now be mimicked with multiple DIVs
  80. # [08:47] <christianz> hmm, outline-offset...
  81. # [08:47] <leaverou> christianz: no, you don't need multiple divs. You can use an outline (same syntax as border) and outline-offset with a negative or positive length to do what you want
  82. # [08:48] <leaverou> by spec it doesn't have to be rectangular, but in practice, for block elements it almost always is
  83. # [08:48] * fantasai isn't entirely clear what the idea is here
  84. # [08:48] <tantek> leaverou - in experience with WHATWG and microformats.org, things are often much more quickly resolved/iterated/advanced in IRC than in the mailing list.
  85. # [08:48] <fantasai> whatwg and microformats also have an IRC channel that is alive nearly 24/7
  86. # [08:48] <christianz> now they should complement it with outline-inset
  87. # [08:49] <leaverou> christianz: outline-inset is basically outline-offset with a negative length
  88. # [08:49] <fantasai> christianz: try a negative outset :)
  89. # [08:49] <christianz> i could illustrate very quickly with an imahe
  90. # [08:49] <leaverou> if I'm understanding what you mean
  91. # [08:49] <christianz> aaaah, cool; i think you are
  92. # [08:49] <christianz> *image
  93. # [08:49] * fantasai isn't sure what's meant, is the problem having multple borders or the placement of the border wrt other things?
  94. # [08:49] * Quits: jet (jet@67.169.43.128) (Quit: jet)
  95. # [08:49] <tantek> fantasai - I think the latter
  96. # [08:49] <leaverou> christianz: wait, I'll make you a demo
  97. # [08:49] <fantasai> leaverou: note that outline has some interesting z-index characteristic
  98. # [08:49] <fantasai> s
  99. # [08:50] <tantek> sounds like altering the painting of the border to be offset from the border area
  100. # [08:50] <leaverou> christianz: http://dabblet.com/gist/2317664
  101. # [08:50] <christianz> fantasai, imagine a box on your page but instead of its border hugging it it is actually the border line is spaced away somewhat from the box
  102. # [08:50] <fantasai> christianz: but what does that mean, visually?
  103. # [08:50] <fantasai> christianz: how do you define the box?
  104. # [08:51] <christianz> fantasai, look at leaverou's example
  105. # [08:51] <christianz> it is just what i was getting at
  106. # [08:51] <tantek> leaverou - good example
  107. # [08:51] <christianz> you can even change the negative offset value to a positive one and see what happens
  108. # [08:51] * Joins: jet (jet@67.169.43.128)
  109. # [08:52] <fantasai> so do you want ot drawn over the content?
  110. # [08:52] <fantasai> or do you put padding there so it's not over the content?
  111. # [08:52] <christianz> have you looked at his example yet?
  112. # [08:52] <fantasai> if it's outset from the background area, does it still take up layout space, or does it draw over things before/after/to-the-side?
  113. # [08:52] <leaverou> fantasai: multiple borders would easily solve the latter
  114. # [08:52] <leaverou> but...
  115. # [08:52] <leaverou> :/
  116. # [08:52] <fantasai> christianz: hers, yes
  117. # [08:53] <tantek> christianz - do you have a specific use-case? e.g. some print or other layout example
  118. # [08:53] <christianz> good question, it might draw over things, depends on how you lay things out
  119. # [08:53] <christianz> will take a photo from a magazine
  120. # [08:53] <leaverou> christianz: that would be useful
  121. # [08:53] <fantasai> christianz: note you can do this with border-image pretty easily
  122. # [08:54] <fantasai> christianz: it has a way to inset/outset the border-drawing area from the image
  123. # [08:54] <leaverou> fantasai: images are cumbersome
  124. # [08:54] <tantek> leaverou - I agree
  125. # [08:54] <leaverou> that's why I disagree with not making border a list
  126. # [08:54] <fantasai> leaverou: that's fine. I still don't understand how this is supposed to impact layout though
  127. # [08:54] <leaverou> IIRC the reason was that border-image covers it, but most people don't want to use images, so they use hacky workarounds instead
  128. # [08:55] <christianz> coming right up...
  129. # [08:55] <christianz> http://www.flickr.com/photos/cziebarth/7049944653/in/photostream
  130. # [08:56] <christianz> keep in mind i just hastily took that pic
  131. # [08:56] <leaverou> christianz: you can't do that currently, not without images at least
  132. # [08:56] <leaverou> or multiple divs
  133. # [08:56] <christianz> can't mimic the photo?
  134. # [08:56] <leaverou> an outline-radius property was discussed recently, but rejected
  135. # [08:56] <christianz> can't do the effect shown in the photo?
  136. # [08:57] <leaverou> yes, since the dashed outline has a radius
  137. # [08:57] <leaverou> currently, outline does not follow border-radius
  138. # [08:57] <leaverou> tantek & fantasai: correct me if I'm saying anything wrong
  139. # [08:57] <christianz> but border does, even though sometimes not very well
  140. # [08:57] <fantasai> leaverou: yeah, though that's up to the implementation. They could follow the radius
  141. # [08:57] <leaverou> yes, border does
  142. # [08:58] <leaverou> outline usually doesn't
  143. # [08:58] <fantasai> christianz: heh, yes, sometimes not very well...
  144. # [08:58] <leaverou> firefox has a nonstandard outline-radius property but doesn't apply by default (and did I mention it's nonstandard?)
  145. # [08:58] * fantasai just took screenshots from IE, since none of the other browsers could render a particular edge case sanely
  146. # [08:58] <leaverou> however, lets look at the big picture
  147. # [08:59] <leaverou> even if it was possible to use outline for this, it would still be a hack
  148. # [08:59] <fantasai> right
  149. # [08:59] <christianz> so . . . for now outline-offset can be used but only effectively on non-rounded off DIVs
  150. # [08:59] <leaverou> a hack to imitate multiple borders, basically
  151. # [08:59] <fantasai> which brings me back to the question of, how do you expect this to affect layout
  152. # [08:59] <christianz> my example wasn't so much to show border-radius but an offset outline or border
  153. # [08:59] <leaverou> christianz: either on non rounded divs OR when your outline goes that much far inside that even if it had a radius, it would've been straight by that point
  154. # [09:00] <christianz> this is just the example i was able to grab quickly from my scrapbook of design inspiration
  155. # [09:00] <christianz> other examples i have don't have rounded corners
  156. # [09:00] <leaverou> christianz: for example, in the demo I sent you
  157. # [09:00] <leaverou> you can go up to border-radius: 12px without it looking bad
  158. # [09:00] <christianz> yes, i've been tinkering with your example
  159. # [09:00] <christianz> kind of like this Dabblet thing
  160. # [09:01] <fantasai> *sigh*
  161. # [09:01] <fantasai> Alright, so can we post to www-style yet?
  162. # [09:01] <leaverou> because even if the outline had a radius, the more inset it is, the less radius it has.
  163. # [09:01] <fantasai> flag it [css4-background]
  164. # [09:01] <christianz> yes
  165. # [09:01] <fantasai> and then maybe when i ask questions about layout, someone will answer them
  166. # [09:01] <leaverou> fantasai: I'm not sure I'd want to post again about multiple borders. I think I've requested it in the past. Not sure though
  167. # [09:01] <leaverou> but it certainly has been discussed and rejected
  168. # [09:01] <christianz> it's not exactly a multiple border issue
  169. # [09:02] <christianz> i guess it sort of is
  170. # [09:02] <christianz> and sort of isn't
  171. # [09:02] <fantasai> it could be interpreted either that way or another way
  172. # [09:02] <leaverou> I'd guess that it's much more likely to be done with border becoming a list than with adding border-offset
  173. # [09:02] <leaverou> border-offset would seriously eff up the box model
  174. # [09:02] <fantasai> depends on what it means
  175. # [09:02] <leaverou> like, you'd have two padding boxes
  176. # [09:02] <christianz> mmm, maybe it would
  177. # [09:02] <leaverou> or two content boxes
  178. # [09:02] <leaverou> or both
  179. # [09:03] <fantasai> I'd say you have two border areas: the area you paint, and the area that's used for layout
  180. # [09:03] <christianz> i think the box model could stay the same but with its border placed farther in or out; but i'd have to think this through more before i could say for sure
  181. # [09:03] <fantasai> the offeset distance would add to the border layout area, but not the painting area
  182. # [09:04] <leaverou> fantasai: wouldn't that add something like extra padding? or am I misunderstanding you?
  183. # [09:04] <fantasai> leaverou: think about background-clip
  184. # [09:04] <leaverou> brb sorry
  185. # [09:05] <fantasai> christianz: just make a nice colleciton of your examples and post to www-style and say 'I want to do this' :)
  186. # [09:05] <fantasai> christianz: I can guarantee we'll look into it. I can't guarantee when we'll actually address it though.
  187. # [09:06] <christianz> URL for www-style?
  188. # [09:06] <fantasai> http://lists.w3.org/Archives/Public/www-style/
  189. # [09:06] <christianz> thanks
  190. # [09:06] <christianz> at least i know about outline-offset now
  191. # [09:07] <christianz> and Dabblet
  192. # [09:16] * Quits: christianz (62a4d766@78.129.202.38) (Quit: http://www.mibbit.com ajax IRC Client)
  193. # [09:44] <leaverou> fantasai: back
  194. # [09:44] <leaverou> too late I guess :/
  195. # [09:54] * Quits: jet (jet@67.169.43.128) (Quit: jet)
  196. # [10:00] * Quits: glenn (gadams@174.29.112.105) (Client exited)
  197. # [10:38] * Joins: SimonSapin (simon@82.232.219.95)
  198. # [10:40] * Joins: Ms2ger (Ms2ger@91.181.161.196)
  199. # [10:50] * Quits: jdaggett (jdaggett@202.221.217.73) (Quit: jdaggett)
  200. # [11:01] * Quits: logbot (logbot@110.173.227.145) (Ping timeout)
  201. # [11:04] * Joins: logbot (logbot@110.173.227.145)
  202. # [11:36] * Joins: drublic (drublic@95.115.32.152)
  203. # [12:04] * Quits: tantek (tantek@70.36.139.112) (Quit: tantek)
  204. # [14:00] * RRSAgent excuses himself; his presence no longer seems to be needed
  205. # [14:00] * Parts: RRSAgent (rrs-loggee@128.30.52.169)
  206. # [15:47] * Quits: drublic (drublic@95.115.32.152) (Client exited)
  207. # [17:03] * Joins: jet (jet@67.169.43.128)
  208. # [17:25] * Joins: arno (arno@192.150.10.201)
  209. # [18:02] * Joins: glazou (glazou@82.247.96.19)
  210. # [18:04] * Parts: glazou (glazou@82.247.96.19)
  211. # [18:06] * Quits: jet (jet@67.169.43.128) (Quit: jet)
  212. # [18:29] * Joins: drublic (drublic@95.115.32.152)
  213. # [18:53] * Joins: paul___irish (paul___iri@205.186.165.150)
  214. # [18:54] * Joins: jet (jet@159.63.23.38)
  215. # [18:59] * Quits: leaverou (leaverou@67.180.84.179) (Quit: leaverou)
  216. # [19:05] * Quits: drublic (drublic@95.115.32.152) (Client exited)
  217. # [20:13] * Joins: dbaron (dbaron@159.63.23.38)
  218. # [20:31] * Quits: SimonSapin (simon@82.232.219.95) (Ping timeout)
  219. # [20:31] * Joins: SimonSapin (simon@82.232.219.95)
  220. # [20:35] * Joins: glenn (gadams@71.218.123.241)
  221. # [20:43] <TabAtkins> fantasai: Haha, that was my response too, but I decided not to actually say anything. ^_^
  222. # [21:36] * Quits: SimonSapin (simon@82.232.219.95) (Ping timeout)
  223. # [21:37] * Joins: drublic (drublic@95.115.32.152)
  224. # [22:26] * Quits: Ms2ger (Ms2ger@91.181.161.196) (Quit: nn)
  225. # [22:38] * Quits: arno (arno@192.150.10.201) (Quit: Leaving.)
  226. # [22:51] <fantasai> TabAtkins: yeah, I'm tactless like that :)
  227. # [22:52] * Quits: kennyluck (kennyluck@114.43.124.199) (Quit: kennyluck)
  228. # [22:52] * Joins: kennyluck (kennyluck@114.43.124.199)
  229. # [22:53] * Quits: kennyluck (kennyluck@114.43.124.199) (Client exited)
  230. # [22:54] * Joins: kennyluck (kennyluck@114.43.124.199)
  231. # [22:55] <fantasai> TabAtkins: And I have to respond somehow: he tagged it with [css3-text].
  232. # [22:58] <kennyluck> Writing is troublesome. I hate writing too. That discussion on IRC was fine I would say.
  233. # [22:58] <kennyluck> I missed the 'myth' party.
  234. # [22:58] <kennyluck> myth: WHATWG is an organization
  235. # [22:59] <kennyluck> It's just a community of insane people who like writing mails.
  236. # [22:59] <TabAtkins> Or who are compelled to.
  237. # [22:59] <TabAtkins> It's a sickness.
  238. # [22:59] <TabAtkins> Don't make fun.
  239. # [23:00] <kennyluck> myth: standardization is slow
  240. # [23:00] <kennyluck> It's pretty obvious to me that it's implementation that's slow.
  241. # [23:01] <kennyluck> And the reason is very simple: creating Web sites/apps makes big money. Creating browsers doesn't.
  242. # [23:01] <kennyluck> myth: Web Developers don't know all these.
  243. # [23:02] <kennyluck> I think they do, but how do Web Standards make money? -webkit- and HTML5 do.
  244. # [23:26] * Quits: glenn (gadams@71.218.123.241) (Ping timeout)
  245. # [23:33] * Joins: glenn (gadams@174.29.105.49)
  246. # [23:36] <kennyluck> "It’s true the vendor prefixes have become ubiquitous, but that’s because the W3C has been slow to act, and slow to implement, and where they have proposed the spec, the browser vendors are still working out how to implement in some cases."
  247. # [23:37] <kennyluck> myth: "W3C people" and "browser vendors" are disjoint sets
  248. # Session Close: Sat Apr 07 00:00:01 2012

The end :)