| 00:15 | <othermaciej> | porting Badgers should be easy |
| 00:15 | <othermaciej> | given the source assets anyway |
| 00:17 | <TabAtkins> | Yeah, if I recall Badgers is just a bunch of translations and scalings. |
| 00:20 | <Philip`> | Can you get good audio/graphics synchronisation? |
| 00:22 | <Dashiva> | Philip`: Not needed, since the original badgers desyncs |
| 00:23 | <Philip`> | Oh |
| 00:23 | Philip` | hasn't watched it enough to remember noticing that |
| 01:29 | <KaOSoFt> | <input> elements don't have "class" or "id" attribute? Just "name"? |
| 01:29 | <Philip`> | All elements have class and id attributes |
| 01:29 | <KaOSoFt> | Nevermind... just read the "Global attributes" |
| 01:29 | <KaOSoFt> | ._. |
| 01:37 | <TabAtkins> | Philip`: You have a corpus of pages to run regexps over, right? |
| 01:42 | <Philip`> | TabAtkins: Yes |
| 01:42 | <Philip`> | I have http://www.dotnetdotcom.org/ |
| 01:42 | <TabAtkins> | Oh, dotbot is you? Cool. |
| 01:43 | <TabAtkins> | So, do you index CSS files? |
| 01:43 | <Philip`> | No, it's not me |
| 01:43 | <Philip`> | I just downloaded their index sample |
| 01:43 | <TabAtkins> | Ah, kk. |
| 01:43 | <Philip`> | (which is on that page) |
| 01:44 | <TabAtkins> | kk. |
| 01:44 | <TabAtkins> | So, while I may go download that myself, for now, do you know if CSS files are indexed in that? |
| 01:44 | <Philip`> | (There's no CSS files in it) |
| 01:44 | <Philip`> | It's almost entirely text/html |
| 01:44 | <TabAtkins> | Dammit. |
| 01:44 | <TabAtkins> | kk. |
| 01:45 | <Philip`> | There's lots of inline <style>s, though |
| 01:45 | <TabAtkins> | Actually, that'd probably still be fine. |
| 01:45 | <TabAtkins> | You mind running /#[0-9a-f]{4}[^0-9a-f]|#[0-9a-f]{8}[^0-9a-f]/ over it? |
| 01:46 | <Philip`> | It's not limited to <style>s, so it finds zillions of unhelpful matches |
| 01:47 | <TabAtkins> | For that string? |
| 01:47 | <Philip`> | like URL fragments, and numeric character references |
| 01:47 | <TabAtkins> | Hrm. |
| 01:47 | <TabAtkins> | Well, I've got something else running that should hopefully cough up results over the weekend. |
| 01:48 | <Philip`> | (I assume you wanted it to be a case-insensitive regexp) |
| 01:48 | <TabAtkins> | Yus. |
| 01:49 | <Philip`> | Even the {8} one just finds <!-- webmoney attestation label#5F42F595-4775-4728-9FD2-E06986259B01 --> and <li><a href="forum/post-272136-cat-1-pg-910.html#f2290890" class="selected" title="KLIKNIJCIE">KLIKNIJCIE</a></li> etc |
| 01:49 | <TabAtkins> | Bleh, that's useless. Kk. |
| 01:51 | <Philip`> | If I put color:\s* at the front then it's more helpful |
| 01:51 | <Philip`> | www.ot-cassel.fr/index.php?IdMenu=174&IdClassement1=52 #Ligne0 { background-color:#fafete; } |
| 01:51 | <Philip`> | www.climaton.com/sticker_choice.php?city=Bankston&state=IA&page=1 <FONT STYLE="font-family:arial,verdana;font-size:10px;color:#0000;"> |
| 01:51 | <Philip`> | s4.invisionfree.com/UltimaOne/index.php?showforum=106 #submenu { border:1px solid #DDD;background-color: #FFF;font-size:10px;margin:3px 0px 3px 0px;color:#AAAA;font-weight:bold; background-image: url(http://209.85.12.236/4997/138/upload/p28044.jpg); background-repeat: repeat-x;} |
| 01:51 | <Philip`> | etc |
| 01:51 | <TabAtkins> | Hm, kk. |
| 01:52 | <TabAtkins> | Can you run it for color and background-color? |
| 01:52 | <TabAtkins> | And gimme the results? |
| 01:52 | <Philip`> | remix64.com/board/viewtopic.php?p=62229 <td class="gensmall" align="right">It is currently <b><span style='color:#10100000; font-size:10px;'>03/04/2009 - 0:50</span></b><br /></td> |
| 01:53 | <Philip`> | Can do |
| 01:53 | <TabAtkins> | Awesome. |
| 01:53 | <Philip`> | (color: will match background-color: too) |
| 01:53 | <Philip`> | Remind me to upload the results in ten minutes |
| 01:55 | <TabAtkins> | Oh, duh, of course it will. |
| 01:55 | <TabAtkins> | kk. |
| 02:01 | <Philip`> | TabAtkins: http://philip.html5.org/data/color-4-8.txt |
| 02:03 | <TabAtkins> | ^^; |
| 02:07 | <TabAtkins> | Those damned swingers, with their giant lines that I can't find the matched bits in. |
| 08:45 | <zcorpan_> | #abcdef/0.5 |
| 08:46 | zcorpan_ | thinks opacity should stay as [0,1] since people have gotten used to it by now |
| 08:47 | <zcorpan_> | #abc/0.5 |
| 08:48 | <zcorpan_> | what about color keywords combined with opacity? |
| 08:48 | <zcorpan_> | papayawhip/0.5 |
| 08:48 | <zcorpan_> | would be nice |
| 08:50 | <annevk> | rgb()/ |
| 08:50 | <annevk> | rgba()/ ... |
| 08:51 | <zcorpan_> | i wouldn't apply it to rgb() or rgba() notations |
| 08:57 | <ment> | isn't "/" already used as operator? |
| 08:57 | <zcorpan_> | where? |
| 08:58 | <ment> | in css2.1 |
| 08:58 | <zcorpan_> | which property? |
| 08:58 | <ment> | not property, grammar |
| 08:59 | <zcorpan_> | / is not special in the grammar, i think |
| 08:59 | <ment> | operator: '/' S* | ',' S*; |
| 08:59 | <zcorpan_> | ah |
| 08:59 | <ment> | i know you have another layer of property-syntax parser atop of grammar-parser, but still |
| 08:59 | <zcorpan_> | still, it's not a problem, is it? |
| 09:01 | <ment> | if you don't mind ending with c++-like parser in ten years, then no, it isn't |
| 09:06 | <zcorpan_> | i wouldn't want to use an uglier character because the grammar splits it |
| 09:06 | <zcorpan_> | we use operators in rgba() too |
| 09:10 | <ment> | hmm that's true |
| 09:11 | <annevk> | border-radius uses / too |
| 09:11 | <annevk> | font too |
| 09:11 | <annevk> | background level 3 too |
| 09:12 | <annevk> | not very special |
| 09:13 | <zcorpan_> | maybe the tradition so far is to use / in shorthand properties to split values of two property values |
| 09:13 | <ment> | but this would have to be parsed in the lexer |
| 09:15 | <annevk> | zcorpan_, not in border-radius |
| 09:22 | <zcorpan_> | hsivonen: about.validator.nu doesn't talk about html5+aria+svg+mathml |
| 09:41 | <zcorpan_> | www.Adobe.com/Choice/ is a sponsored link at the top of the google search result for 'html5' |
| 09:45 | <annevk> | lowercase C? |
| 09:45 | <annevk> | hmm, uppercase C, but if you would type that in you get an error |
| 09:46 | <zcorpan_> | interesting |
| 13:47 | <Dashiva> | Oh neat, someone finally asked what content-language is supposed to be useful for |
| 18:12 | <zcorpan_> | it seems srt doens't have comments like websrt does |
| 18:12 | <zcorpan_> | what happens in srt impl when there are websrt comments? |
| 18:15 | <annevk> | websrt comments? |
| 18:15 | <micheil> | websrt? |
| 18:15 | <Philip`> | comments? |
| 18:16 | <zcorpan_> | the spec doesn't have comments apparently but http://damowmow.com/temp/srtspec does |
| 18:17 | <annevk> | copy-paste from text/event-source? |
| 18:18 | <annevk> | hmm doesn't seem like it |
| 18:28 | <zcorpan_> | Philip`: validator.nu seems to have code to warn when a meta is sniffed as an encoding decl but there's no http-equiv='content-type' |
| 18:29 | <zcorpan_> | Philip`: maybe that could be used to find pages that have such meta elements |
| 18:32 | <zcorpan_> | Philip`: for the purpose of researching the compat impact of http://www.w3.org/Bugs/Public/show_bug.cgi?id=9225 |
| 18:34 | Philip` | is probably too busy/lazy to do any research like that any time soon |
| 18:34 | <zcorpan_> | ok |
| 20:08 | <Hixie> | zcorpan_: i gave up trying to add comments |
| 20:09 | <zcorpan_> | Hixie: good :) |
| 20:46 | <zcorpan_> | http://www.chromium.org/sts seems useful |
| 20:47 | <othermaciej> | it really needs to be spec'd properly |
| 20:47 | <othermaciej> | I also hate the "includeSubdomains" option |
| 20:47 | <othermaciej> | trying to get STS into a standards body seems to be kind of stalled out :-( |
| 20:50 | <zcorpan_> | it doesn't help the first time you visit the site, unless the site is listed in the preloaded list |
| 20:51 | <zcorpan_> | maybe we don't need a header at all, just the preloaded list combined with smarter default (if server redirects to https, assume it wants https in the future also) |
| 20:56 | <othermaciej> | the most important feature of STS (IMO) is not https-always but rather forcing hard rejection of self-signed or otherwise invalid certs |
| 20:56 | <othermaciej> | it would be nice to do that for all certs but doing it on an opt-in basis seems like a good starting point |
| 20:56 | <othermaciej> | STS isn't meant to help you the first time you visit a site, it's meant to help you if you have a pre-existing relationship with the site and want to access it under untrusted network conditions |
| 21:01 | <annevk> | wasn't there a more formal IETF draft of STS? |
| 21:02 | <annevk> | nm |
| 21:04 | <theMadness> | I'm apalled at www-style, :selection gets a lengthy discussion while there's still no fallback mechanism other than the media query thing webkit implemented for SOME things. |
| 21:35 | <Hack5aw> | What Working Group? lol |
| 21:36 | <othermaciej> | annevk: I don't think there has been yet |
