and

[07:52:01.0000] Yeah, that's easy [07:52:02.0000] ok. How long do you expect it to take you to write that much? [07:53:00.0000] I just did in around a minute :P [07:54:00.0000] oh, ok. that's a lot quicker than it would have taken me, since I find python to be ab extremely confusing language [07:54:01.0000] s/ab/an/ [07:55:00.0000] can you send me the code for that then, so I can test it? [07:55:01.0000] http://forums.whatwg.org/viewtopic.php?p=465#465 is spam [07:56:00.0000] /me finishes tweeking it [07:56:01.0000] Lachy: (it was more than just a minute) [07:56:02.0000] Lachy: http://pastebin.ca/1272410 [07:56:03.0000] The only real problem is it runs the outliner both for that and for the TOC, and the outliner is slow [07:57:00.0000] /me wonders what Lachy is trying to achieve [07:58:00.0000] jgraham: "Replace numeric headings with the correct numeric headings for their [07:58:01.0000] depth regardless of sectioning content elements." [07:59:00.0000] I don't understand. I would understand if you said "accounting for sectioning content elements" [08:01:00.0000] like

[08:01:01.0000] becomes: [08:01:02.0000]

[08:01:03.0000] gsnedders, it doesn't seem to work [08:02:00.0000] Lachy: That's to be expected. I wrote it quickly. [08:02:01.0000] gsnedders: OK, that makes sense. I don't see how that is irrespective of sectioning content elements though :) [08:02:02.0000] jgraham: it ignores them when numbering them [08:03:00.0000] I put in in a file called replaceHeadings.py, added it to that python directory, and ran: anolis --enable=replaceHeadings test.src.html test.html [08:03:01.0000] But it uses them to deduce the numbering [08:03:02.0000] jgraham: yeah, sure [08:03:03.0000] :P [08:04:00.0000] gsnedders, test.src.html looks like this: http://pastebin.ca/1272420 [08:06:00.0000] hmm, yeah [08:08:00.0000] Lachy: oh, == and not = [08:09:00.0000] on which line? [08:10:00.0000] Lachy: http://pastebin.ca/1272422 [08:10:01.0000] Lachy: But http://pastebin.ca/1272424 is cleaner code [08:11:00.0000] (no unicode(depth + 1), just unicode(depth)) [08:11:01.0000] What's a DifferentParentException? [08:11:02.0000] oh gawd [08:12:00.0000] copy/paste error [08:12:01.0000] s/DifferentParentException/TooDeepException/ [08:13:00.0000] final version (hopefully): http://pastebin.ca/1272427 [08:13:01.0000] Also I guess section.header.tag[1].isdigit() should be section.header.tag in ("h1", "h2" [...]) but using a set for better performance [08:13:02.0000] So that it works when Hixie introduces the new h1b element [08:13:03.0000] frozenset ideally [08:13:04.0000] yes, frozenset [08:14:00.0000] /me is on;y guessing that a set will perform better for n=6 [08:15:00.0000] gsnedders, will you be including this in the releases as an optional feature, or will I need to keep my own separate copy? [08:15:01.0000] Lachy: I'll include it with 1.1 [08:15:02.0000] ok [08:16:00.0000] Lachy: I don't want to include it with 1.0, with that so close to shipping [08:16:01.0000] that'll give me time to learn more python and try and fix bugs in it [08:17:00.0000] jgraham: for Lachy's test doc it's too small to make any diff :P [08:17:01.0000] jgraham: 0.000s for both [08:19:00.0000] jgraham: no diff for HTML 5 either [08:20:00.0000] gsnedders, is there any way to avoid running the outliner twice? [08:21:00.0000] Lachy: not really [08:21:01.0000] could you somehow cache the outline and return that for subsequent calls? [08:21:02.0000] Lachy: Yeah, I'll do that for 1.1 [08:21:03.0000] ok [08:21:04.0000] Lachy: That's part of the reason why I don't want to ship this in 1.0 [08:22:00.0000] it's the third most expensive method on HTML 5 of all of Anolis [08:22:01.0000] gsnedders: The frozenset version is about 1x10^-7 seconds faster [08:23:00.0000] (per loop) [08:23:01.0000] jgraham: with what uncertainty? [08:23:02.0000] Dunno [08:24:00.0000] /me expects the uncertainty is so big that the result is more or less irrelevant [08:25:00.0000] gsnedders, for 1.1, can you make it possible for anolis to read directly from standard input rather than requiring an input file, so that I can maintain a collection of source files and run: cat section1.html section2.html ... | anolis output.html [08:26:00.0000] otherwise, I need to write to a temporary file first [08:26:01.0000] gsnedders: Dunno. It seemed to be about 5x10^-8 but I only ran it twice [08:27:00.0000] jgraham: That's still 50% uncertainty, which isn't very certain at all [08:27:01.0000] jgraham: But still quicker all the time, interestingly [08:27:02.0000] /me shrugs [08:27:03.0000] Lachy: Yeah, I've been thinking about that too [08:28:00.0000] Lachy: The question is how I tell apart the first argument [08:28:01.0000] jgraham, what are you using to measure the execution time down to a milliionth of a second? [08:28:02.0000] gsnedders, use a command line parameter that means read from standard input [08:29:00.0000] Lachy: yeah, that's probably simplest [08:29:01.0000] /me motions towards http://bugs.gsnedders.com/projects/show/anolis [08:30:00.0000] and also add one to allow it to be output directly to standard output too [08:30:01.0000] /me will flood that with feature requests later [08:51:00.0000] /me finds Google SearchWiki to be actually useful for once [08:51:01.0000] since it lets me delete the really annoying search results on faqs.org for RFCs, since it's HTML and an ugly font and I just want the IETF text versions [09:30:00.0000] Lachy: Fixed the infinite loop [12:55:00.0000] Images are similar to embedded media only if they are also embedded. [12:55:01.0000] Images from IMG tags are not logically similar to embedded media. [12:56:00.0000] Their behaviour with respect to aspect ratio can be different. [12:56:01.0000] is there any way to turn off section numbering for all sections in a document in LaTeX? [12:58:00.0000] Sure, you just have to redefine a hidden macro with 5 @�s in the name :-) [12:58:01.0000] ah, \setcounter{secnumdepth}{0} [12:58:02.0000] :P [13:00:00.0000] gsnedders: Edit menu, Find "\section", Replace "\section*", I think [13:00:01.0000] Philip`: That doesn't work for me. [13:01:00.0000] Oh [13:01:01.0000] Philip`: The section occurs in a file \input{}'d into two documents, and only one of them doesn't want numbering [13:02:00.0000] gsnedders: Write a Makefile that duplicates the file and calls sed on one copy [13:02:01.0000] Philip`: No. [13:02:02.0000] :) [13:02:03.0000] \setcounter{secnumdepth}{0} works fine [13:46:00.0000] /me gives dmitry a link to http://www.google.com/sponsoredlinks?q=programmers [13:48:00.0000] He could hire one of those teams that offered to build an application solving the halting problem [14:02:00.0000] Hixie, why would you bother responding to Dmitry at all? [14:02:01.0000] politness [14:02:02.0000] SOMEONE ON THE INTERNET IS WRONG [14:04:00.0000] That shows one result: EEPROM / Flash Programmer. [14:06:00.0000] hm, there's a lot of results herein the US [14:06:01.0000] I get lots in Norway too [14:06:02.0000] Not stretching is not a use case for IMG, right. [14:08:00.0000] That is why IMG-loaded images are stretched. [14:08:01.0000] Dashiva, all my results seem to be US sites. Does google redirect you to google.no? [14:08:02.0000] /me hates when Google does that [14:08:03.0000] No, I've set it to not redirect [14:09:00.0000] But you can go to .no and it'll give lots of .no answers :) [14:09:01.0000] and its especially annoying since my google cookie seems to expire frequently and I keep getting redirected back to google.no [14:09:02.0000] That's a feature [14:09:03.0000] I only got one result from the google.no [14:09:04.0000] There was lots of anger when google had cookies that lasted to 2038 :) [14:10:00.0000] http://www.google.no/sponsoredlinks?q=programmerere&btnG=S%C3%B8k+i+sponsede+koblinger [14:10:01.0000] oh, yeah, the expiring cookie is ok for that reason. But the fact that google redirects is a bug [14:10:02.0000] I think the keyword should be in Norse. [14:10:03.0000] We stopped speaking Norse a few centuries ago, alas [14:10:04.0000] Google should provide a way to respect Accept-Language headers. [14:11:00.0000] I wonder if it would work if I set Accept-Language: en-au,en,no;q=0 [14:11:01.0000] Lachy: google _does_ respect accept-language. For all languages except for en-* :( [14:12:00.0000] aargh! [14:12:01.0000] that's so damn annoying! [14:12:02.0000] Yeah. I hate, hate, hate it. [14:12:03.0000] I'm already sending en,nb-NO;q=0.9,nb;q=0.8,no-NO;q=0.7,no;q=0.6 [14:12:04.0000] then we need a way to tell google that i really mean I only speak english [14:12:05.0000] what's nb-NO? [14:12:06.0000] Frequently tried to contact them to at least honor en-au, en-nz, en-ca e.a. which won't be set by default anywhere. but no.... :( [14:12:07.0000] bokmål [14:15:00.0000] How's your Norwegian coming along, Lachy? [14:16:00.0000] jeg forstår ikke norsk, men jeg snakker bare litt [14:17:00.0000] translation for those of us not acquainted with Nordic languages? [14:17:01.0000] I can read a few words, and understand spoken words like polse and kvittering [14:17:02.0000] "I don't understand norwegian, but I only speak a little" (bad wording included) [14:17:03.0000] *pølse [14:18:00.0000] Dashiva, what's the correct way to phrase that in norwegian? [14:18:01.0000] I guess Nordic languages aren't big on capitalizing proper nouns/adjectives, are they? [14:18:02.0000] What exactly did you want to say? [14:18:03.0000] jcranmer: proper nouns, yes [14:18:04.0000] that I don't understand it, but I can say a few words [14:18:05.0000] "men jeg kan litt/kan noen ord" [14:19:00.0000] I'd love to learn a Nordic language someday, but I suppose I should work on other languages in my list [14:19:01.0000] Although "forstår ikke" is a fairly strong way of not understanding [14:19:02.0000] is there another way? [14:19:03.0000] "snakker ikke" might fit better [14:20:00.0000] "Jeg snakker ikke norsk, men jeg forstår en del ord og uttrykk" perhaps [14:21:00.0000] ok, that sort of works [14:21:01.0000] it's weird that google translate translates "norsk" into "English" though [14:22:00.0000] That is weird [14:23:00.0000] is "Jeg kanne har cheezburger" the correct translation for "I can has cheezburger"? [14:23:01.0000] I'll use the only foreign lang I know: "Je ne comprends pas , mais je peux parler un peu." [14:23:02.0000] "kan har" [14:24:00.0000] jcranmer, is that French? [14:24:01.0000] (kanne isn't a valid conjugation in any case) [14:25:00.0000] yep [14:26:00.0000] zcorpan, I could make that text less specific. How about: "Search for a header cell with matching id in that table." (re: http://krijnhoetmer.nl/irc-logs/whatwg/20081201#l-162) [14:26:01.0000] Although the have duality own/receive isn't present in Norwegian 'ha' [14:26:02.0000] jcranmer: I always think comprendre is spelt wrong :\ [14:27:00.0000] zcorpan, or rather "For each token, search for a header cell with matching id in that table." [14:27:01.0000] surprendre looks wrong to me until I look at the past participle... surpris [14:41:00.0000] Philip`, 4 instances of "their" changed to "thier". (re: http://krijnhoetmer.nl/irc-logs/whatwg/20081201#l-422) [14:41:01.0000] (uploading) [14:41:02.0000] BenMillard, I hope you meant "thier" changed to "their" :-) [14:44:00.0000] BenMillard, re the table header sniffing discussion earlier, were there really a lot of cases where that resulted in false positives? [14:46:00.0000] and would there be no way to distinguish them. e.g. by only detecting them in the first row or column? [14:54:00.0000] Lachy, I just got to that point in the logs. It wasn't so much the quantity of problems, it's that fact that totally reasonable use cases (such as using to mark a data cell's content as being particularly important) screw up the rest of the table. :( [14:55:00.0000] Lachy, and was used about as often for that as , IIRC. [14:56:00.0000] Lachy, I can put a note in the "Ben's Advice" section of the comparison document about that if you like? [14:58:00.0000] but in those cases, wouldn't those cells be more likely to occur in the middle of the table, rather than the first row or first column, and if the entire first row/column used or , wouldn't that be a reasonable indication that their meant to be headers? [14:59:00.0000] Lachy, that's an interesting idea. [15:00:00.0000] Lachy, it becomes a question of how much heuristics Hixie would let us get away with. :) [15:00:01.0000] i don't want any heuristics, i want the algorithm to reflect the actual semantics of the language [15:01:00.0000] in particular, i really don't want anything magic like [15:01:01.0000] yeah, those things often have unexpected "gotchas!" when it comes to real tables anyway [15:01:02.0000] ok, fine [15:02:00.0000] Lachy, for example, I've seen tables where row headers use but it had an entire row is made of to highlight the data in it. There are also cases where is used to create "section headers" midway through the table. So you could add an extra layer of heuristics to disambiguate those cases...but then there's another layer you could add, and another...ad infinitum. :) [15:06:00.0000] I'll add a "No Heuristics" bit to the advice section, to make clear that my experience with studying data tables supports Hixie's design choice on this. (Basically, they open a can of worms which is difficult to close.) [15:17:00.0000] Lachy, ESPN (and a few other places) uses to apply the boldness to header cells rather than . So...how much heuristics do you want? :P [15:17:01.0000] at least the tables using bold header cells can be retrofitted to with only the text alignment making a visible difference [15:18:00.0000] or they could use with no visible difference (if HTML5 supports that) [15:19:00.0000] or use and add text-align:foo in the relevant bit of CSS [15:20:00.0000] BenMillard, don't use the slippery slope fallacy. I only suggested because it seemed like a common, easily detectable method that people use for headers [15:24:00.0000] Lachy, ok. Check the list under "Using with any attributes and a or a as the only child": http://projectcerbera.com/web/study/2007/tables/ [15:24:01.0000] Lachy, it's not that common. [15:26:00.0000] what rfc defines the ftp: scheme? [15:28:00.0000] Hixie: http://www.iana.org/assignments/uri-schemes.html seems to give the answer [15:29:00.0000] thanks [15:57:00.0000] Lachy, what part of my reasoning was a slippery slope fallacy? Was it comparing to ? [15:58:00.0000] <|tbb|> hello all, im searching for a simple tutorial how to use tag, what i want is to read the content of the srcfile into variables [15:59:00.0000] /me wonders if there's a name for the fallacy of falsely calling an argument fallacious 2008-12-02 [16:00:00.0000] BenMillard: "... So you could add an extra layer of heuristics to disambiguate those cases...but then there's another layer you could add, and another...ad infinitum" "ESPN uses ... how much heuristics do you want?" [16:02:00.0000] hmm, maybe that's the continuum fallacy, not slippery slope http://en.wikipedia.org/wiki/Slippery_slope http://en.wikipedia.org/wiki/Continuum_fallacy [16:03:00.0000] <|tbb|> anyone familar with this event-source thing? [16:03:01.0000] yes [16:03:02.0000] <|tbb|> what im trying to do is create an offline app, that loads serval stuff when it becomes online [16:04:00.0000] <|tbb|> ive found out that thing with the manifest file where you can define whats from cache and whats from network [16:05:00.0000] <|tbb|> but i dont know how to read that stuff, or better i dont know how the sourcefile should look like [16:05:01.0000] <|tbb|> for example [16:06:00.0000] <|tbb|> how can i read and output that file [16:06:01.0000] I don't understand, you seem to be confusing offline webapps with event-source, which are 2 different things [16:09:00.0000] <|tbb|> maybe point 3 descripes it better -> http://www.browseking.com/cgi-bin/nph-bk1.pl/010110A/http/www.w3.org/TR/2008/NOTE-offline-webapps-20080530/ [16:10:00.0000] what is browseking.com and why are you proxying the spec through that server? [16:11:00.0000] <|tbb|> found this url through google while searching for a solution, sorry [16:11:01.0000] this is a more up to date version http://www.whatwg.org/specs/web-apps/current-work/multipage/offline.html#offline [16:11:02.0000] Lachy, so it was the jump from to , then? [16:11:03.0000] yes [16:12:00.0000] ok, that makes sense [16:12:01.0000] <|tbb|> ive pasted the content -> http://pastebin.ca/1272831 [16:15:00.0000] ah, that's just an example showing what the different sections of the manifest file do [16:16:00.0000] <|tbb|> that manifest file works for me, but i dont know how to integrate an network(online) file into my cached site [16:17:00.0000] the eventsource element (previously known as "event-source") allows you to send events to the client from the server. See http://www.whatwg.org/specs/web-apps/current-work/#the-eventsource-element [16:18:00.0000] basically, the client creates a persistent connection with the server and the server can send events to the client through that connection. Then the scripts in the page can listen for those events as they're fired on the eventsource element [16:19:00.0000] <|tbb|> oh, like a webservice [16:20:00.0000] this section describes the format for the events http://www.whatwg.org/specs/web-apps/current-work/#server-sent-events [16:21:00.0000] Lachy, the "can of worms" criticism of "add extra heuristics until works" is valid before jumping to , though? [16:22:00.0000] /me is learning logic today. [16:23:00.0000] I think your argument would have been valid by pointing out that there are so many other cases that would give false positives and false negatives, that the simple solution I proposed wouldn't be very effective. [16:24:00.0000] Lachy, yeah that's what I was trying to say. :) [16:24:01.0000] ok, fair enough [16:26:00.0000] Lachy, is characterising that as "it opens a can of worms which is difficult to close" valid? [16:27:00.0000] <|tbb|> lachy, the ticker.php which is descriped in that server-sent example, how does this file look like [16:27:01.0000] (that section will include a link to these logs, I'm just figuring out a short way to describe it) [16:27:02.0000] BenMillard, I don't know. possibly. [16:28:00.0000] |tbb|, http://www.whatwg.org/specs/web-apps/current-work/#parsing-an-event-stream [16:29:00.0000] ticker.php would need to have logic in it that makes the server maintain a persistent connection, and then each event the server sends would need to be in that format [16:31:00.0000] <|tbb|> how can i set the mime type of the ticker, is that correct-> header("Content-Type: text/event-stream"); [16:31:01.0000] for PHP, yes [16:31:02.0000] I don't know how to make PHP maintain a persistent connection though [16:33:00.0000] <|tbb|> for me it where enough, to send the content of the ticker.php once [16:33:01.0000] <|tbb|> maybe im thinking completly wrong [16:33:02.0000] <|tbb|> for what i want to do [16:34:00.0000] I'm not sure what you're trying to do though. Why are you using server sent events for? [16:34:01.0000] s/Why are/What are/ [16:34:02.0000] If you just want to download the immediate output of that script, and aren't doing any kind of streaming, it sounds like you could just use XMLHttpRequest instead [16:36:00.0000] |tbb|, read http://labs.opera.com/news/2006/09/01/ [16:36:01.0000] that's based on an older version of the spec, which has since changed, but it will give you a better understanding of what it does and how it works [17:04:00.0000] <|tbb|> Lachy: ive found an php example which should sent the server-time each 3 secs, http://my.opera.com/WebApplications/blog/index.dml/tag/server-sent%20events [17:06:00.0000] <|tbb|> ive changed the sented data to -> data: YHOO data: -2 data: 10 to use it with the example in whatwg example but nothing happends, any idea? oh btw, ive changed the header also to text/event-stream [17:07:00.0000] which browser are you using? [17:07:01.0000] <|tbb|> safari (iphone) [17:07:02.0000] that doesn't implement it yet [17:08:00.0000] if you want to experiment with it, use Opera, but you have to use the old MIME type because that's what has been implemented [17:11:00.0000] <|tbb|> lachy, sorry because ive asked again, are you sure that the latest safari iphone os fw.2.2 doesnt support it? [17:12:00.0000] yes, I'm sure [17:12:01.0000] Lachy: what are we talking about? [17:12:02.0000] server sent events [17:12:03.0000] ah [17:12:04.0000] /me isn't sure if those are supported in webkit trunk [17:14:00.0000] http://www.google.com/search?q=Server-sent+events+webkit returns some discussions about reviewing patches to implement it [17:14:01.0000] first result: https://lists.webkit.org/pipermail/webkit-reviews/2007-September/019931.html [17:14:02.0000] so it looks like there was an implementation at least started [17:15:00.0000] https://bugs.webkit.org/show_bug.cgi?id=14997 still marked NEW [17:20:00.0000] gsnedders: *poke* [17:20:01.0000] http://software.hixie.ch/utilities/js/live-dom-viewer/?%3C!DOCTYPE%20html%3E...%3Cscript%3C%2Fscript%3E%3Cx%3E%3C%2Fx%3E%3Cselect%3E%3Cscript%3C%2Fscript%3E%3Cx%3E%3C%2Fx%3E%3Coptgroup%3E%3Cscript%3C%2Fscript%3E%3Cx%3E%3C%2Fx%3E%3Coption%3E%3Cscript%3C%2Fscript%3E%3Cx%3E%3C%2Fx%3E [17:20:02.0000] sigh [17:20:03.0000] /me mumbles something about ... [17:30:01.0000] yes [17:31:00.0000] why is there no second /X element in http://software.hixie.ch/utilities/js/live-dom-viewer/?%3C!DOCTYPE%20html%3E...%3Cx%3E%3Cb%3Ea%3C%2Fb%3E%3C%2Fx%3E%3Cselect%3E%3Cx%3E%3Cb%3Ea%3C%2Fb%3E%3C%2Fx%3E%3C%2Fselect%3E in IE8?? [17:32:00.0000] aargh! WTF! "Internet Explorer has modified this page to prevent a potential cross-site scripting attack." [17:32:01.0000] how the hell do I disable that annoying "feature" [17:32:02.0000] /me uploads to the clipboard from a real browser first [17:33:00.0000] You tell the author of the site you're using to add a special header to disable the XSS protection :-) [17:33:01.0000] i don't ever get that message [17:33:02.0000] did i do something wrong? [17:33:03.0000] I do get it [17:34:00.0000] how do i get it to come up? [17:34:01.0000] Hixie, copy and paste that URL above into IE8 [17:34:02.0000] it seems to detect the is an empty element [17:24:01.0000] was that the reason, or part of the reason, or was it for a different reason(s) altogether? [17:44:00.0000] there are a lot of MS fanboys. The MS MVP program is institutionalized fanboys [17:49:00.0000] it's going to be a real nightmare figuring out why (or not) a site is being displayed with the IE8 layout engine [17:56:00.0000] "As for CSS compliance I have a question: the CSS specification says that illegal property-values should be ignored. Why does IE throw an exception when an illegal property-value is assigned through script (e.g. setting element.style.width='-10px' or something)?" [17:56:01.0000] urp [17:57:00.0000] roc: what was the answer to that? [17:58:00.0000] there was no answer [18:00:00.0000] MikeSmith: that was the only reason that i recall [18:01:00.0000] Hixie: OK, thanks [18:01:01.0000] roc: that was from a recent "chat with the MS team" thing? [18:01:02.0000] comment in the IEblog [18:03:00.0000] I se [18:03:01.0000] see [18:29:00.0000] roc: btw, I made a number of changes to the script I use for generating the "Typical default display properties" sections, to deal with problems you mentioned [18:29:01.0000] cool [18:29:02.0000] see http://www.w3.org/html/wg/markup-spec/#audio-display [18:29:03.0000] or better: [18:29:04.0000] http://www.w3.org/html/wg/markup-spec/#input-display [18:37:00.0000] MikeSmith: I don't think 'box-align' is a standard property (yet) [18:38:00.0000] roc: yeah, I know. but is that not also true about box-orient and box-margin? [18:38:01.0000] yes [18:38:02.0000] I didn't notice those [18:39:00.0000] appearance and box-sizing are too [18:39:01.0000] where are all of the circle-based properties? [18:39:02.0000] the circle doesn't get any respect :( [18:40:00.0000] roc: yeah, and border-radius, and margin-start, and padding-start, and user-select [18:40:01.0000] well, those ones are in CSS3 [18:40:02.0000] although Webkit's syntax might not match the CSS3 syntax for border-radius [18:41:00.0000] OK [18:41:01.0000] user-select might not be in any current CSS3 draft but it has been in the past [18:41:02.0000] so has box-sizing I suppose [18:41:03.0000] OK, well, I guess I can switch those back to e.g., _vendor_-box-orient [18:41:04.0000] box-align/box-orient/box-margin are the most glaring annoyances because they've never been in any CSS spec [18:42:00.0000] OK [18:42:01.0000] I don't think it makes sense to include non-standard properties at all, to be honest [18:42:02.0000] hmm [18:42:03.0000] most readers aren't going to know that those box properties are XUL related [18:42:04.0000] webkit uses them also [18:43:00.0000] yeah, because it implements parts of XUL layout [18:43:01.0000] ah [18:43:02.0000] these rules are really UA-specific ways of achieving a certain rendering for the replaced element [18:43:03.0000] so I'll just have the script drop the box- properties [18:43:04.0000] roc: OK, I see [18:44:00.0000] IMHO we don't need or really want to display "typical UA properties" for replaced elements [18:44:01.0000] yeah [18:44:02.0000] they're implementation details [18:45:00.0000] OK, but what about keeping the -start properties ? [18:46:00.0000] I could just have it change those to -left, but that would kind of misleading also [18:46:01.0000] MikeSmith, is your HTML markup language draft somewhere in CVS? [18:47:00.0000] I don't see it with the rest of them on dev.w3.org [18:47:01.0000] Lachy: it's in www.w3.org CVS [18:47:02.0000] ok [18:48:00.0000] just out of interest, why didn't it get put with the others? [18:49:00.0000] because systeam is asking me repeatedly not to put any more "documents" in dev.w3.org. claim that dev.w3.org is intended for hosting code and that we are essentially abusing it [18:50:00.0000] problem is that dev.w3.org is one puny machine, not load-balanced or mirrored like www.w3.org is [18:50:01.0000] IIRC the "start" and "end" variants of properties are something that is being standardized [18:50:02.0000] they have had some serious load problems with spiders crawling the HTML5 draft hosted on dev.w3.org [18:50:03.0000] MikeSmith, shouldn't the solution to that problem be to allow CVS access to certain areas of www.w3.org by WG members? [18:50:04.0000] I'd just leave them in [18:50:05.0000] heycam: yeah [18:51:00.0000] roc: OK, thanks. What do you think about keeping "appearance"? [18:51:01.0000] values of appearance are mostly obvious [18:52:00.0000] all your uses of 'appearance' are for replaced elements [18:52:01.0000] like I said, I wouldn't include any "default display properties" for replaced elements [18:52:02.0000] ok, so does the sys team have any intention of shifting specs to the other CVS server, or moving dev.w3.org onto the load balanced servers? [18:53:00.0000] roc: I guess I'm not clear on on what a replaced element is. seems like 'input, input[type="password"], input[type="search"]' is not selecting a replaced element.. or is it? [18:54:00.0000] replaced elements are elements like form controls where (at least in principle) CSS does not define the rendering of the element [18:54:01.0000] OK [18:54:02.0000] CSS treats replaced elements as black boxes basically [18:55:00.0000] it just so happens that in Gecko and Webkit, at least, we implement replaced elements using CSS internally as well [18:55:01.0000] or mostly so [18:55:02.0000] but that is an implementation detail [18:56:00.0000] I see. Is that true for "user-select" also? [18:56:01.0000] Lachy: we can shift specs ourselves to www.w3.org and use ACLs to control access. but it's maintenance PITA, and no good reason to do it except for the dev.w3.org load issues [18:56:02.0000] user-select is a CSS property, not an element [18:57:00.0000] you're listing it on 'input' and 'textarea', which are both replaced elements [18:57:01.0000] roc: yeah, I know -- I mean the actual property [18:57:02.0000] OK [18:57:03.0000]