| 00:20 | <zewt> | cool, people talking about webvtt features in terms of "we should do this since some other format does it, and this feature doesn't need to be consistent with other parts of webvtt since some other format doesn't do it" |
| 04:42 | <Hixie> | annevk: you aware of http://lists.w3.org/Archives/Public/public-whatwg-archive/2012Aug/thread.html#msg210 ? |
| 06:25 | <annevk> | Hixie: seemed like a duplicate of the Mozilla stuff based on http://lists.w3.org/Archives/Public/public-whatwg-archive/2012Aug/0218.html |
| 06:26 | <annevk> | Hixie: which is discussed on public-webapps already |
| 08:41 | <annevk> | The reason Chrome does the weird resolving btw is I think because parsing of URLs in Chrome never fails |
| 08:42 | <annevk> | so you never get a SYNTAX_ERR of any kind, but network errors down the road |
| 11:48 | <annevk> | Hixie: if you wake up soon enough: http://www.qwertee.com/product/there-s-no-place-like-home/ |
| 11:49 | <annevk> | (via rune) |
| 12:03 | <annevk> | Is https://bugzilla.mozilla.org/show_bug.cgi?id=303478 also about handling changes to <base>? |
| 12:03 | <annevk> | (And note that behavior is now defined in HTML...) |
| 12:03 | <annevk> | (Somewhat different from how bz explains it afaict.) |
| 12:44 | <Ms2ger> | Lachy, thanks for http://dvcs.w3.org/hg/resources/rev/6a3a7087e2e2 |
| 12:49 | <annevk> | can someone @opera.com tell sigbjorn Opera is the only browser to escape \ in the query string? He'll know what to do |
| 12:49 | <annevk> | sof that is |
| 12:51 | <MikeSmith> | yay XPath http://google-opensource.blogspot.co.uk/2012/09/wicked-good-xpath-faster-javascript.html |
| 12:58 | <jgraham> | Kind of hard not to read "pseudo-java" in place of "closure" |
| 12:58 | <jgraham> | annevk: Done |
| 12:59 | <AryehGregor> | jgraham, do you know (or can you find out) what makes <div contenteditable style="background:green"></div> not collapse to zero height in Opera? |
| 12:59 | <AryehGregor> | Apparently WebKit hardcodes contenteditable stuff in their layout logic somewhere, which seems kind of unpleasant. |
| 12:59 | <jgraham> | AryehGregor: I can ask at least |
| 13:00 | <AryehGregor> | Everyone but Gecko does it, though, so it should probably be specced. |
| 13:00 | <AryehGregor> | I'm hoping maybe Opera will have some good-enough 80% workaround that can be easily specced. |
| 13:00 | <AryehGregor> | Like maybe "editing hosts need to behave as though they're affected by the following CSS rule". |
| 13:04 | <jgraham> | AryehGregor: We seem to insert a <br> |
| 13:04 | <AryehGregor> | Hmm, interesting. |
| 13:05 | <AryehGregor> | So it's a parser special-case? |
| 13:05 | <AryehGregor> | That seems more appealing than a CSS special case, although still not ideal. |
| 13:05 | <jgraham> | I don't know that it's in the parser exactly |
| 13:07 | <AryehGregor> | jgraham, is https://dvcs.w3.org/hg/resources the authoritative copy that changes should be pushed to these days? |
| 13:07 | <jgraham> | Oh, it is in the parser |
| 13:07 | <jgraham> | AryehGregor: Yes |
| 13:07 | <jgraham> | http://software.hixie.ch/utilities/js/live-dom-viewer/saved/1745 |
| 13:43 | <annevk> | zcorpan: https://twitter.com/SilentImp/status/244061433627811840 |
| 13:56 | <annevk> | argh |
| 13:56 | <jgraham> | < sof> jgraham: tell annevk to try 12.50, we fixed that one |
| 13:57 | <annevk> | \0\n\t gives %00%0A%09 as you expect |
| 13:57 | <annevk> | but remove the \0 and they're all gone |
| 13:57 | <annevk> | in Gecko |
| 13:58 | <annevk> | jgraham: yeah sorry, thought of that later :/ |
| 14:04 | <annevk> | jgraham: updated to 1497 for Opera Next, removed Opera.app to ensure I use the latest |
| 14:04 | <matt444> | http://www.w3.org/2012/05/sysapps-wg-charter.html |
| 14:04 | <matt444> | That page has a Join link which is dead |
| 14:05 | <annevk> | matt444: it's not an actual charter either |
| 14:05 | <annevk> | just a draft |
| 14:05 | <annevk> | matt444: anyway, you want to bug the W3C with that |
| 14:05 | <matt444> | ah, ok thanks |
| 14:07 | <annevk> | jgraham: still broken in the Opera Next 9.50... |
| 14:07 | <annevk> | euh, 12.50 |
| 14:08 | <jgraham> | I have build 1577 |
| 14:09 | <jgraham> | You might need to get it from the Desktop Team Blog |
| 14:16 | <annevk> | in other news, did anyone know browsers strip U+0009, U+000A, and U+000D liberally from anywhere in URL paths |
| 14:17 | <annevk> | also from the query string and fragment afaict |
| 14:17 | <annevk> | but only for hierarchical paths |
| 14:17 | <annevk> | of course... |
| 14:20 | <annevk> | installing 1583 now |
| 14:22 | <annevk> | jgraham: works, cool |
| 14:22 | <annevk> | jgraham: but not cool that the update command fails... |
| 14:23 | <jgraham> | Yeah, the Desktop people are quite reluctant to put builds onto the auto update for reasons that I don't entirely understand |
| 14:27 | Ms2ger | likes his auto-updates |
| 14:28 | <Ms2ger> | Also, since annevk apparently isn't going on vacation in the immediate future, I'll have to |
| 14:29 | <Ms2ger> | (until the 14th or so) |
| 14:29 | <annevk> | I was hoping you were going to type more |
| 14:29 | <annevk> | like "I'll have to demonstrate I'm a better editor" or some such |
| 14:29 | <annevk> | but have fun! |
| 14:29 | <Ms2ger> | Heh |
| 14:29 | <Ms2ger> | Thanks :) |
| 14:58 | Ms2ger | waves |
| 18:14 | <TabAtkins> | Anyone got a link to a document explaining the W3C policy of not allowing tech requirements in specs for things without sufficiently open licenses? |
| 20:34 | <Hixie> | seems crazy that we allow authors to add and remove text tracks from in-band text tracks while they're parsing |
| 21:17 | <jgraham> | MikeSmith: What the hell are you doing posting useful informative content to Google+? |
| 21:19 | <cabanier1> | tabatkins: you mean the w3c patent policy? |
| 21:23 | <Hixie> | jeez, zcorpan |
| 21:23 | <Hixie> | making <track> only work when there's a media element parent is a bitch |
| 21:23 | <Hixie> | i've no idea how to even do it |
| 21:35 | <TabAtkins> | cabanier1: Maybe, yeah. |
| 21:52 | <cabanier1> | http://www.w3.org/Consortium/Patent-Policy-20040205/ |
| 23:00 | <Hixie> | i wonder if there are other platforms where the people designing the APIs have to basically treat authors as the enemy... :-/ |
| 23:01 | Hixie | is trying to work out how to do the text track selection in a way that does what authors want, what users want, what vendors want, and yet is not going to just fall apart when authors start randomly removing track elements or poking at the DOM |
| 23:01 | <TabAtkins> | I, on the other hand, am figuring out the easiest way to automate this railroad-diagram creation. Existing libraries don't do what I want. |
| 23:02 | <TabAtkins> | At least, not in the exact visual style I want. |
| 23:02 | <Hixie> | railroad-diagram? that sounsd like way more fun |
| 23:03 | <TabAtkins> | Well, it's SVG. And nothing to do with trains. So, depends on your proclivities. |
| 23:03 | <Hixie> | oh. no trains. sad now. |
| 23:03 | <TabAtkins> | gonna add an informative section to the Syntax module illustrating the syntax of each token with railroad diagrams. |
| 23:03 | <TabAtkins> | Easier to read than regex. |
| 23:03 | <Hixie> | oh, those kinds of diagrams |
| 23:04 | <Hixie> | what i consider "pascal-style syntax diagrams", because i first saw them in pascal reference books |
| 23:04 | <Hixie> | i love those, so neat |
| 23:04 | <Hixie> | so hard to represent in text :-P |
| 23:04 | <zewt> | is that the diagrams that sqlite's docs changed to a few years ago, which overnight made the docs totally unreadable :| |
| 23:04 | <TabAtkins> | I consider them "JSON-style syntax diagrams", for the same reason. |
| 23:05 | <Hixie> | yeah, the json spec uses them too |
| 23:05 | <TabAtkins> | zewt: http://dev.w3.org/csswg/css3-syntax/railroad-diagrams.svg |
| 23:05 | <Hixie> | that's awesome |
| 23:05 | <Hixie> | TabAtkins++ |
| 23:05 | <zewt> | TabAtkins: might make sense for that (no idea), but for SQL it's terrible: http://www.sqlite.org/lang_insert.html |
| 23:06 | <Hixie> | that looks awesome too, why is it terrible? |
| 23:06 | <Hixie> | surely better than the alternative bnf, let alone regex |
| 23:06 | <zewt> | i find it much harder to read than what it used to be (which I think was closer to postgres's, eg. http://www.postgresql.org/docs/current/static/sql-insert.html) |
| 23:07 | <zewt> | also it's unsearchable |
| 23:07 | <Hixie> | the postgre alternative syntax style there is not as precise by a long shot |
| 23:07 | <Hixie> | i agree that the searchability is zero |
| 23:07 | <Hixie> | that's what i meant (in part) by "hard to represent in text" |
| 23:07 | <zewt> | (perhaps svg-based ones could fix that particular problem--looks like I can search the text in that svg, but it won't actually highlight in FF or Chrome) |
| 23:08 | <TabAtkins> | zewt: I'm missing the diagram. Is it somewhere on that page I don't see? |
| 23:08 | <TabAtkins> | Oh, I clicked on postgres, duh. |
| 23:09 | <Hixie> | TabAtkins: what does your source look like? |
| 23:09 | <TabAtkins> | Check it? I'm just writing straight SVG right now. |
| 23:09 | <zewt> | TabAtkins: i don't use SVG much--is it possible to make it so search-highlighting works, or is that just a multi-browser bug? |
| 23:09 | <TabAtkins> | Now that I've done two of them, I'm looking into generating them from a simpler source. |
| 23:09 | <Hixie> | TabAtkins: oh, you're doing it by hand? |
| 23:09 | <zewt> | (dunno if that's just coloring being overridden or something) |
| 23:09 | <Hixie> | TabAtkins: aaah |
| 23:09 | <TabAtkins> | zewt: Dunno! |
| 23:09 | <Hixie> | TabAtkins: i'm interested in subscribing to your newsletter, so that i may receive the issue that describes that syntax once you find a good one :-) |
| 23:09 | <zewt> | you're very excited about dunno! |
| 23:10 | <TabAtkins> | Hixie: It's simple. All I want to do is omit some of the busywork around determining lengths of things. Anyway, when I'm done I'll share. |
| 23:10 | <TabAtkins> | Won't finish today, but it might be my weekend project. |
| 23:10 | <Hixie> | cool |
| 23:11 | <Hixie> | maybe i can replace the html syntax section with some of these if your solution scales well, and give zewt a fit |
| 23:12 | <annevk> | Hixie: next Monday, I'm gonna need some *.spec.whatwg.org URLs |
| 23:12 | <Hixie> | annevk: good to know |
| 23:12 | <Hixie> | annevk: this is the ones that we have to dance into your account? |
| 23:12 | <zewt> | also, it's weird that <svg hidden> doesn't work |
| 23:12 | <annevk> | either that or make me some weird account |
| 23:13 | <Hixie> | annevk: or shall we just give you an account on my user |
| 23:13 | <Hixie> | that's probably easier |
| 23:13 | <annevk> | yeah, it's only going to be a simple script that pulls from github anyway |
| 23:13 | <Hixie> | if we don't have to involve dreamhost it's trivial, just say the word |
| 23:13 | <Hixie> | k |
| 23:13 | <Hixie> | guess at some point i should set up the specs/ page to point to these specs, too |
| 23:13 | <annevk> | xhr, encoding, fullscreen, cors maybe |
| 23:14 | <annevk> | and url at some point, but I've some research and writing to do for that one |
| 23:14 | <Hixie> | a mass exodus, no less |
| 23:14 | <Hixie> | i take it the w3c just shot itself in the foot? or did you time out waiting for them to |
| 23:14 | <annevk> | well, the W3C Invited Expert agreement I don't like |
| 23:14 | <annevk> | it basically prohibits forking these days |
| 23:15 | <Hixie> | yeah |
| 23:15 | <annevk> | and since I'm gonna pay for all development myself, that seems unfair |
| 23:15 | <Hixie> | (only the w3c can fork, don't you know) |
| 23:15 | <annevk> | I rather give it away as long as I still have money |
| 23:17 | <Hixie> | i wonder how many of your specs microsoft will end up forking into webapps :-) |
| 23:18 | <annevk> | if they turn them into respec like with the work of Ms2ger it's all for the worse :/ |
| 23:19 | <TabAtkins> | Hixie: Check the source of http://dev.w3.org/csswg/css3-syntax/railroad-diagrams.svg now. Does the comment at the top look good to you? |
| 23:19 | <Hixie> | hm, not sure xml would be my first choice, but i guess it works |
| 23:21 | <Hixie> | at least for these examples |
| 23:21 | <Hixie> | there are some more complex things i don't think you'd be able to represent as well |
| 23:21 | <Hixie> | do you plan to use this for property value syntax too? |
| 23:21 | <Hixie> | or just core syntax? |
| 23:24 | <Hixie> | TabAtkins: btw, for the hyphen case / <optional>, i think it's convention to put the "normal" case in the straight line, which in this case would mean the hyphen would be on the side branch |
| 23:25 | <Hixie> | TabAtkins: also, you need an element for "sequence of stuff" so that a <choice> can choose between sequences |
| 23:26 | <Hixie> | consider e.g. http://www.freepascal.org/docs-html/ref/ref79x.png |
| 23:27 | <Hixie> | TabAtkins: this diagram has an example i don't think you could do with xml easily: http://www.cosc.canterbury.ac.nz/greg.ewing/python/paramlist.png |
| 23:29 | <GPHemsley> | so... :active defaults to the whole page if an element isn't specified? |
| 23:30 | <Hixie> | no |
| 23:30 | <Hixie> | :active is equilavent to *:active |
| 23:30 | <Hixie> | *|*:active, if there's no default namespace |
| 23:30 | <Hixie> | like .active would be |
| 23:32 | <GPHemsley> | ah |
| 23:32 | <GPHemsley> | though isn't that effectively the same thing? |
| 23:34 | <Hixie> | depends on what exactly you mean by "defaults to the whole page" |
| 23:35 | <GPHemsley> | Hixie: Probably the fact that :active is *:active ;) |
| 23:44 | <Hixie> | the script that i report bugs with just told me off for not being descriptive enough for me to understand what i meant |
| 23:44 | <Hixie> | -_- |
