15:57:31 RRSAgent has joined #css 15:57:31 logging to http://www.w3.org/2009/04/15-css-irc 15:57:36 Bert, the proposal was to copy wording directly from dev.w3.org/csswg/css3-background 15:57:42 dbaron has joined #css 15:57:48 Zakim, this will be Style 15:57:48 ok, glazou; I see Style_CSS FP()12:00PM scheduled to start in 3 minutes 15:57:51 Style_CSS FP()12:00PM has now started 15:57:53 + +1.858.216.aaaa 15:58:01 Bert, your edits for issue 108 doesn't make counter-increment: none 3 foo 4; invalid 15:58:05 zakim, +1.858.216 is me 15:58:05 +plinss; got it 15:58:12 Bert, it just says authors must not write that 15:58:19 annevk has left #css 15:58:23 Bert, since it doesn't say it's invalid, there's no requirement that UAs ignore such declarations 15:58:41 also I'm not seeing that wording in a post to www-style... 15:58:51 One at a time, I'm still looking at 102... 15:59:18 +David_Baron 15:59:20 -David_Baron 15:59:20 +David_Baron 15:59:50 Zakim, passcode? 15:59:50 the conference code is 78953 (tel:+1.617.761.6200 tel:+33.4.89.06.34.99 tel:+44.117.370.6152), anne 16:00:13 +??P26 16:00:20 Zakim, ??P26 is me 16:00:24 +anne; got it 16:01:55 +??P8 16:02:00 -??P8 16:02:43 +??P10 16:02:50 +Cesar_Acebal 16:02:51 Zakim, ??P10 is fantasai 16:02:51 +fantasai; got it 16:02:58 CesarAcebal has joined #css 16:03:43 +Bert 16:04:59 + +47.21.65.aabb 16:05:12 howcome has joined #css 16:05:19 Zakim, aabb is howcome 16:05:19 +howcome; got it 16:05:24 Zakim, who is on the phone? 16:05:24 On the phone I see plinss, David_Baron, anne, fantasai, Cesar_Acebal, Bert, howcome 16:06:18 emilyw has joined #css 16:06:45 glazou has joined #css 16:06:53 +??P31 16:07:03 zakim, P31 is me 16:07:03 sorry, emilyw, I do not recognize a party named 'P31' 16:07:10 +[Microsoft] 16:07:27 Zakim, ??P31 is emilyw 16:07:27 +emilyw; got it 16:07:29 zakim, ??P31 is me 16:07:29 I already had ??P31 as emilyw, emilyw 16:07:58 sylvaing has joined #css 16:08:07 Zakim, [Microsoft] has sylvaing 16:08:07 +sylvaing; got it 16:08:53 scribe:sylvain 16:08:56 scribenick:sylvaing 16:09:14 http://lists.w3.org/Archives/Public/www-style/2009Apr/0227.html 16:09:17 Page counters 16:09:29 hakon: this sounds complicated to me 16:09:32 elika: nooooo 16:09:47 +Melinda_Grant 16:09:53 melinda has joined #CSS 16:10:01 elika: defining interactions between counters does complicate things 16:10:30 sorry, I'm totally unable to call, my phone and SIP are both dead at this time 16:10:50 Skype! 16:10:58 melinda: if we don't define counter interactions, we can't handle certain scenarios.... 16:11:00 fantasai: I tried... 16:11:06 s/nooooo/agreed/ 16:11:46 melinda: there were issues with using section numbers in document headers 16:12:39 fantasai: use cases that were not working without counter interactions: footnote counter that resets on every page but incremented by elements in the document 16:12:39 fantasai: no, that was covered already 16:12:48 + +95089aacc 16:13:21 fantasai: also chapter-level counters that increase with every page in the chapter but reset at the next chapter (?) 16:13:37 hakon: agree with the footnote scenario, that is important. not sure whether it needs to be so complex 16:14:25 dbaron: I'm surprised it's so simple, given how complicated some of the counters stuff is :) 16:14:55 plinss: aside from general complexity, are there specific issues to be addressed ? 16:15:29 glazou_ has joined #css 16:17:57 hakon: proposal: for any named counter, only allow interaction for the oldest one 16:18:12 fantasai: counter name hiding is not the complex issue but the nesting of elements 16:18:47 dbaron: I haven't looked at the proposal yet. 16:18:50 and multiple break points 16:18:52 at the same page break 16:18:58 melinda: we should take a week to think about this further. 16:19:45 fantasai: this proposal is actually the second iteration (after melinda shot down the first one) 16:20:05 plinss: can we elaborate on how the proposal evolved to its current stage ? 16:20:55 fantasai describes the proposal's rules 16:21:10 http://lists.w3.org/Archives/Public/www-style/2009Apr/0227.html 16:22:19 It's worth being careful that the multiple break point rule doesn't kick in when an element crosses three pages. 16:22:55 fantasai: first rule deals wit the general simple case; second rule with multiple break points 16:24:34 fantasai: I would really like to hear from Michael Day(Antenna House) and David Baron on this proposal 16:25:01 melinda: we have implementations that do what we want, the spec should match them 16:25:15 fantasai: I would postpone this discussion until we hear back from AH and Prince 16:25:22 hakon: sounds good 16:25:42 s/(/,/ 16:25:44 s/)/,/ 16:26:17 hakon will ping Antenna House and Prince on the issue 16:26:42 plinss: we can put this back on next week's agenda or the week after 16:26:52 http://lists.w3.org/Archives/Public/www-style/2009Apr/0228.html 16:26:55 Borders & Backgrounds 16:27:02 http://lists.w3.org/Archives/Public/www-style/2009Apr/0228.html 16:27:14 fantasai: i'm waiting for feedback on most of these open issues 16:28:33 fantasai: we can probably close issue 28 as no change 16:28:58 fantasai: people seem to like color fallback so we may keep the feature but change the syntax 16:29:09 fantasai: only two conclusions I have at this point 16:29:23 hakon: I'd like to find a way to clip the center image 16:29:49 fantasai: none of the authors that have given feedback so far think this is worth making border-image more complex 16:29:56 hakon: but i'm an author too :) 16:30:43 hakon: this is so easy to do now; without it you need to do Photoshop work 16:30:56 hakon to look at the proposal again... 16:31:07 szilles has joined #css 16:31:39 bert: I prefer simplicity so I would not mind dropping the feature 16:32:08 hakon: it is a slight increase in complexity that saves a frequent work item 16:32:35 hakon: in fact, why not specify when you want to keep the center image i.e. the default should be to clip it 16:33:17 s/dropping/not adding/ 16:33:30 hakon: the default today is that unless you use a manual tool you're not really getting a border anymore 16:33:58 bert: but how do you define the slicing of that image into 9 proper pieces without some editing work 16:34:07 fantasai suggests a straw poll 16:34:51 plinss: if you force the author to use transparency, you force them to choose a particular format 16:35:19 plinss: another use case, the author want the same image to be used in two places, one of which includes the center but not the other 16:36:21 hakon: this is not about graphical feature creep but the default behavior of this property 16:37:00 +SteveZ 16:37:03 plinss: i'm not hearing consensus yet 16:37:21 hakon: we'll keep it as an issue 16:38:07 Topic: 2.1 16:38:07 http://wiki.csswg.org/spec/css2.1 16:38:16 column/page break discussion postponed so alexmog can participate 16:38:22 fantasai: several action items open on Saloni, can we get those reassigned? 16:39:21 plinss: someone from msft ? 16:40:10 ACTION: sylvain and arron to work on Saloni's 2.1 issues 16:40:10 Created ACTION-140 - And arron to work on Saloni's 2.1 issues [on Sylvain Galineau - due 2009-04-22]. 16:40:21 http://wiki.csswg.org/spec/css2.1#issue-112 16:40:45 http://lists.w3.org/Archives/Public/www-style/2009Apr/0099.html 16:40:58 fantasai: I'd like to adopt dbaron's proposal; it's straightforward 16:41:34 http://www.w3.org/TR/CSS21/tables.html#column-alignment 16:42:24 bert: what is the exact issue ? 16:42:43 dbaron: some people think this section means text-align applies to columns 16:43:08 dbaron: the text does not specify which elements the property applies to and the title implies that it may apply to table columns 16:43:23 fantasai: dbaron's proposal is to clearly disambiguate this 16:44:07 bert: change seems ok. but is not absolutely necessary 16:44:27 plinss, fantasai: editorial change. no conflict with implementations. 16:44:40 plinss: objections ? 16:45:00 szilles: no objection to the change, but not sure it fixes the problem 16:45:26 plinss: we're not changing any behavior just clarifying interpretation 16:46:45 RESOLVED: dbaron's proposal accepted (Issue-112) 16:46:48 http://wiki.csswg.org/spec/css2.1#issue-113 16:47:03 fantasai: we discussed this at the F2F 16:48:14 I requested an && syntax operator to the values and units module and it was requested that we also add it to 2.1 16:48:55 hakon does not like having this in CSS3 values and units; should it go into 2.1 syntax ? 16:49:16 plinss: I have no issues with putting this in 2.1 16:49:38 bert: it doesn't hurt 2.1 since it's not used 16:49:51 bert: ...by 2.1 16:50:04 melinda: would we be able to qualify it ? 16:50:13 fantasai: editorial change, does not define a feature 16:50:16 hakon: agree 16:51:05 szilles: if this is an editorial change, a note clarifying that this notation is specified for completeness but unused 16:51:57 s/a note/i suggest a note 16:52:10 emilyw has joined #css 16:53:45 szilles: it would helpful to have examples to describe it 16:53:54 s/helpful/be helpful 16:54:29 RESOLVED: add && to list of value syntax operators (http://wiki.csswg.org/spec/css2.1#issue-113) 16:55:00 szilles can do without the note but would like a clear usage example 16:56:12 -David_Baron 16:56:17 discussion of 2.1 issues that need action items 16:56:49 plinss: we should ask people to process their issues; assign action items to unassigned issues next week 16:57:06 -[Microsoft] 16:57:08 -SteveZ 16:57:10 -howcome 16:57:10 - +95089aacc 16:57:11 -plinss 16:57:12 -Cesar_Acebal 16:57:12 -anne 16:57:13 -emilyw 16:57:15 -fantasai 16:57:26 -Bert 16:58:05 -Melinda_Grant 16:58:06 Style_CSS FP()12:00PM has ended 16:58:08 Attendees were +1.858.216.aaaa, plinss, David_Baron, anne, Cesar_Acebal, fantasai, Bert, +47.21.65.aabb, howcome, emilyw, sylvaing, Melinda_Grant, +95089aacc, SteveZ 16:59:11 szilles has left #css 17:05:00 Bert, let me know when I should do another check on your edits 17:05:15 Bert, the issues list is current as of one hour ago 17:05:27 Bert, that is, I closed the issues you'd finished 17:06:51 fantasai, would it help if I draft proposed text changes for Selectors? 17:07:17 anne: uh, let me read hsivonen's message first :) 17:07:33 Fantasai, I fixed 102 and 103. 17:07:45 CesarAcebal has left #css 17:10:28 fantasai, the overall idea is eliminating one magic list, defining exactly how matching works in HTML in a way that works for HTML+SVG+MathML too and is consistent with the plan for solving that in the DOM, and making HTML and XHTML more consistent; it also means that parsing a style sheet would no longer depend on whether it is associated with an HTML or XML document 17:10:51 parsing a style sheet is not supposed to depend on whether it's associated with an HTML or XML document 17:11:01 I don't see anything about the magic list in hsivonen's email 17:11:08 that's in my e-mail 17:11:18 parsing and matching is intertwined in impl 17:11:19 that's an issue for HTML to solve 17:11:26 but read matching where I said parsing if you wish 17:11:30 Selectors says it follows the case-sensitivity of the language 17:12:12 the whole DOM issue seems like a mess 17:12:29 anyway 17:12:38 I do not agree with Selectors requiring lower-casing of tag selectors 17:13:11 you keep two tag tokens around, one lowercase and one input case 17:13:17 and which you use depends on the namespace 17:13:25 I'm not talking about implementation details 17:13:29 I'm talking about the spec 17:13:34 you could phrase it as a matching requirement 17:13:58 which you use matters 17:13:59 which is intended to be general enough that it works for HTML4, HTML5, XML, and FooImaginaryLanguage 17:14:23 e.g. does {XHTML namespace, HTML} match html or not 17:14:33 no, because XML is case-sensitive 17:14:43 I'm not talking about XML 17:14:53 oh, in the HTML DOM? 17:15:11 Selectors match against a DOM 17:15:25 per HTML5 HTML elements end up in a namespace 17:15:26 that'll depend on whether the DOM considers it case-sensitive 17:15:42 is {XHTML namespace, html} the same type of element as {XHTML namespace, HTML}? 17:15:46 if yes, then they'll match 17:15:49 if no, then they won't 17:15:58 right 17:16:01 if it's not defined, it's not Selectors' place to define it 17:16:10 but {XHTML namespace, html} HTML has to match 17:16:19 and at that point it becomes a Selectors problem 17:16:33 ok, back up a sec 17:16:45 did the spec change to say that HTML elements in HTML documents create XHTML DOM nodes? 17:16:57 there's no such thing as XHTML DOM nodes 17:17:03 there's just DOM nodes 17:17:19 HTML DOM vs XML DOM is some fiction that was never really adopted 17:17:21 Our "internal DOM" has an IsCaseSensitive() method on nodes. 17:17:30 I think you pretty much have to. 17:17:37 hsivonen is removing a bunch of that 17:17:52 He's fixing the namespace disaster that I was against from the start. 17:17:57 But I don't think he's removing IsCaseSensitive() 17:18:15 I don't see how we could distinguish tag matching without that. 17:18:24 We're not going to make "BODY { color: green}" stop working 17:18:25 based on namespace 17:18:42 I think that's a bad idea 17:18:50 for the XHTML namespace you always match lowercase 17:19:01 So you're saying we should make "BODY { color: green} " match in an XML document? 17:19:01 it's pretty neat I think 17:19:17 that'd be a side effect 17:19:25 no 17:19:28 we are not doing that 17:19:30 I think that's a bad idea. 17:19:32 why? 17:19:43 It's 10 years too late. 17:19:55 we do it on the DOM side 17:19:57 and XML is case-sensitive 17:20:10 sure, it remains case-sensitive 17:20:19 case-insensitivity is a pain 17:20:25 it's just the selection mechanism has namespace specific knowledge 17:20:39 where do you do it on the DOM side? 17:20:45 getElementsByTagName 17:20:57 ugh 17:21:04 http://lists.w3.org/Archives/Public/public-html/2009Apr/0081.html 17:21:27 outlines the general idea 17:21:58 i think that's the only sane solution here, but i'm open to suggestions 17:22:33 tag selectors in a CSS style sheet applied to an XML document should match any elements in that document case-sensitively 17:23:05 I don't care how you explain the implementation concept, but I'm opposed to anything that changes that 17:23:05 sylvaing has joined #css 17:23:12 the proposal is to change that 17:23:18 then I'm against the proposal 17:23:52 i don't see why Selectors should be different from the DOM in this respect 17:24:23 it only affects HTML elements in XML documents 17:25:10 look, I don't care what you define for weird DOM manipulating ECMAScripted edgecases 17:25:52 that import nodes from one type of document to another and other twisted things like that 17:26:27 with HTML elements I mean XHTML elements as well, to be clear 17:26:27 but on a straight-up parse from the XML 17:26:40 selector matching should be case-sensitive 17:27:07 This testcase: http://www.w3.org/Style/CSS/Test/CSS3/Selectors/current/xhtml/tests/css3-modsel-181.xml 17:27:11 should remain valid 17:27:13 whatever you do 17:27:19 i think the model where Selectors matches the same regardless of HTML or XML is more clean 17:27:58 you are entitled to your opinions 17:28:49 but you are not entitled to changing behavior that has been clearly specced and interoperably implemented for practically a decade 17:28:58 without a better reason than aesthetics 17:29:33 it makes the platform more consistent 17:29:42 because DOM matching doesn't depend on some silly HTML/XML flag 17:30:09 and conforming XHTML elements are lowercase anyway 17:30:27 and don't match uppercase selectors anyway 17:31:57 only for textArea is that a practical issue 17:32:34 anne, we are not changing the behavior of that testcase 17:33:14 since when do you authority over this? 17:33:20 have* 17:33:32 this does not seem like a reasonable debate to me 17:33:38 i'll bow out 17:34:09 anne, if you want to bring it to the WG, feel free 17:34:17 anne, but you will not get a positive response there 17:34:46 anne, there are too many implementations that pass that test case already 17:35:38 anne, if you want, I'll even keep quiet during the whole debate that ensues 17:35:51 and several might be ok with failing that test as indicated on the list 17:35:55 anyway, we'll see 17:36:11 what dbaron proposed could work too 17:39:12 Lachy has joined #css 17:40:52 maybe some pulp fiction IE abuse is in order: http://www.elliottkember.com/ie.html 17:44:40 lol 17:45:06 i love that movie 17:46:10 totally 17:47:00 kind of love the view-source idea on that page too 17:47:48 Bert, issues 102 and 103 closed 17:47:53 Bert, let me know when you complete the rest 18:11:24 Fantasai, what else do I need to complete? 18:11:57 Issue 85 - fixing typos 18:12:06 Issue 94 - placement of example 18:12:11 maybe you did these? 18:12:18 I didn't check them because you didn't say you did 18:12:27 I believe so, let me check if I uploaded them... 18:12:27 Issue 100 - copying wording from css3-background 18:12:39 btw, I didn't check 104 18:12:43 if you could double-check that for me 18:12:47 I don't really know what to look for 18:12:54 issue 105 18:13:00 Issue 108 18:14:19 I think 100 is done. I copied what was needed according to the issue, I don't think we should change something just to make it better. 18:15:27 94 indeed already uploaded. 18:16:10 85 also 18:17:08 Double-checking 104... 18:18:47 (I noticed while editing 104 that Yves and I already proposed a different, but equivalent change some time ago. A little shorter, but not much.) 18:20:40 Bert, the 85 fix didn't make it to Changes or Errata 18:22:05 We should really stop changing CSS 2.1, if only because it's such a pain to edit a spec and a changes section and an errata list :-( 18:22:23 hehehe 18:22:32 Bert, Issue 94 didn't make it to the Changes list 18:22:52 or Errata 18:25:08 Double-checked 104. 18:27:56 I see 85 in the errata and in the changes... 18:28:20 althjough it would look better to merge C.5.8 with C.5.7. 18:32:04 Zakim has left #css 18:42:47 dbaron has joined #css 18:50:24 85 and 94 updated in errata and changes. 105 seems OK. 108 uses the exact text I sent to www-style; nobody asked for it to be changed. 18:53:35 you could also do 112 while I check :) 18:55:42 Bert, well, ok but I don't think your text actually fixes the issue for 108 18:58:12 oh, there's another change 18:58:18 The issue was whether 'none' could be a counter name. It now says, in two places, that it cannot. What else is needed? 19:00:11 I'll let it slide 19:01:58 Bert: should I close Issue 104 then? 19:08:12 Yes, 104 can be closed. 19:08:28 112 is being generated, should be uploaded in a minute or two. 19:38:39 Bert, wrt feature requirements for border-image 19:38:56 the bullet about drawing the outside edge of the border image 19:39:05 should say "at or beyond" rather than "at the border edge" 19:40:59 Did anybody request "beyond"? 19:41:24 yes 19:41:33 that was one of the major points in the discussion over Brad Kemper's propsoal 19:41:45 s/propsoal/proposal/ 19:41:51 and one of the key features in it 19:43:23 Checked in an example for && 19:43:32 to css3-background 19:43:37 you can use it to edit Issue 113 19:48:21 Borders outside the borders: seems rather over-engineered. 19:48:32 borde-rimage was nice because i wa ssimple. 19:48:46 Now it's no longer simple. I'd rather drop the whole property :-( 19:49:53 you probably should take a good look at the examples in http://www.bradclicks.com/cssplay/border-image/Thinking_Outside_The_Box.html 19:50:03 they show why it's necessary 19:50:07 anyway 19:50:24 No, they show that Brad has a lot of fantasy. 19:50:37 112 still hasn't made it to w3.org 19:50:48 did you check it in yet? 19:51:27 I can see it. 19:52:14 oh 19:52:15 yeah 19:52:21 I can see the text change too 19:52:26 but the title hasn't been changed 19:52:30 http://lists.w3.org/Archives/Public/www-style/2009Apr/0099.html 19:54:31 sylvaing has joined #css 19:58:37 The title wasn't part of David's proposal, just a "perhaps." The change itself already wasn't necessary, the change of the title even less. It would have been good in 1997, but now it just causes confusion. 19:59:51 alright, fine 20:00:31 next time, I'm going to insist on us being excruciatingly precise about which bits of a proposal we're accepting when we accept a proposal 20:00:46 I was under the impression that both changes were accepted 20:09:45 Alright, those issues closed 20:09:49 113 and you're done :) 20:11:12 Bert, let me know when you're done with 113 20:31:39 (fwiw, the scribe thought both changes were accepted as well) 20:57:38 anne has joined #css 21:06:58 sylvaing has joined #css 21:31:48 sylvaing has joined #css 21:58:14 sylvaing has joined #css 23:54:21 arronei has joined #CSS