| 00:23 | <ap> | Hixie: no-op? By saying "no-go", I meant that exposing load() on HTML documents may be unacceptable for compatibility |
| 00:23 | <Hixie> | ah, yeah, that's possible |
| 00:25 | <ap> | Hixie: so FWIW, Document===XMLDocument + existing XMLDocumentLoader.load still seems like a good plan to me |
| 00:25 | <ap> | Hixie: unless some vendor wants to test the waters with Document.load |
| 00:25 | <Hixie> | having XMLDocument.load not be the actual load() is not web compatible |
| 00:26 | <Hixie> | (webkit goes down a number of fallback codepaths because of that) |
| 00:26 | <ap> | Hixie: that Sarissa thing seems pretty unique, although admittedly rather common on the Web |
| 00:28 | <ap> | Hixie: introducing Document.load might the best path overall indeed. Hopefully someone remembers if there were actual known compatibility issues with that |
| 00:36 | <Hixie> | ap: it would mean "load()" would refer to document.load() and not window.load() in all event handler attributes (e.g. onload="load()" would change meaning) so you're probably right that it's a no-go. I commented on the bug again. |
| 00:36 | <Hixie> | http://www.w3.org/Bugs/Public/show_bug.cgi?id=14037 |
| 00:42 | <ap> | Hixie: that's an interesting combination of ideas that I didn't consider. It's pretty far from what WebKit implements now, so I can't confidently say that it's OK for us, but it seems OK |
| 00:42 | <Hixie> | it's basically what bz proposed before |
| 00:42 | <Hixie> | i was really hoping we could find a way to not have the onreadystate magic, which is why i had moved away from it before |
| 00:43 | <Hixie> | but i don't see a way around it |
| 00:43 | <ap> | Hixie: it's the "Regular XML documents <…> keep returning a regular Document object" part that I found new |
| 00:44 | <Hixie> | ah |
| 00:45 | <Hixie> | well it's basically what you said would be confusing in comment 27 :-) |
| 00:45 | <Hixie> | but i don't know that we have an option really |
| 00:46 | <Hixie> | i don't think it'd be that confusing though, it just means that createDocument() returns a special Document, effectively |
| 00:50 | <Hixie> | Lachy: yt? |
| 00:50 | <Hixie> | Lachy: (pondering :scope, <style scoped>, querySelector()'s refNodes) |
| 07:29 | <jgraham> | So, for reference, the right first step when doing anything python related is to set up a virtualenv |
| 07:29 | <jgraham> | http://pypi.python.org/pypi/virtualenv |
| 07:30 | <jgraham> | Hixie: Thanks for the test. I'll regenerate the W3C copy of the testsuite |
| 08:28 | <annevk> | why is sicking not on IRC when he's sending email? |
| 09:01 | <MikeSmith> | annevk: yeah! there should be a law |
| 09:02 | <MikeSmith> | hmm, didn't MS submit tests for the history API? |
| 09:03 | <MikeSmith> | oh, I guess not yet |
| 09:03 | <jgraham> | MikeSmith: We submitted tests for the popState stuff. Is that what you're thinking of? |
| 09:04 | <MikeSmith> | jgraham: dunno what I was thinking of :) |
| 09:04 | <annevk> | sicking was asking about the history API on twitter |
| 09:05 | <MikeSmith> | https://twitter.com/#!/SickingJ/status/114230538134499328 |
| 09:08 | <annevk> | So the domparsing-ref is escalated into an issue |
| 09:08 | <annevk> | I have no idea what that means |
| 09:13 | <annevk> | oh yes |
| 09:13 | <annevk> | jgraham is going to TPAC |
| 09:13 | <jgraham> | Oooh, I am? |
| 09:13 | <jgraham> | ;) |
| 09:14 | <annevk> | just don't worry if you wake up in a strange place in a month and a half |
| 09:16 | <MikeSmith> | annevk: it means we got some good times ahead |
| 09:16 | <MikeSmith> | and/or it means the sky is falling |
| 09:17 | <annevk> | it means jgraham can go to meetings and I can relax in SF :p |
| 09:21 | <MikeSmith> | heh |
| 09:23 | <MikeSmith> | you know, I've been thinking, the really large number of things that have gotten added to the platform over the last 3 years especially, all this stuff getting implemented, e.g., latest news about new features in IE10 |
| 09:24 | <MikeSmith> | those are clear signs that we are all doing something wrong |
| 09:24 | <MikeSmith> | forgetting what we're here for |
| 09:25 | <zcorpan> | MikeSmith: explain |
| 09:26 | <MikeSmith> | well, I mean, we should spend less time on that stuff -- forget about, say, writing test cases for the history API |
| 09:26 | <MikeSmith> | and instead we should use that time to discuss stuff like "consider reducing verbosity when talking about code points" |
| 09:27 | <zcorpan> | ah |
| 09:30 | <MikeSmith> | man, I like it when people drop links to the HTML version of the ES5 specs |
| 09:30 | <MikeSmith> | https://twitter.com/#!/wycats/status/114237748843524097 |
| 09:31 | <annevk> | MikeSmith, hahaha |
| 09:31 | <MikeSmith> | but now TC39 is going to ruin it by making ES6 and obsoleting it all |
| 09:32 | <annevk> | ECMAScript should move to the W3C I think |
| 09:32 | <MikeSmith> | yeah, it should |
| 09:32 | <MikeSmith> | should have been there to begin with |
| 09:32 | <annevk> | having to join a separate standards organization with its own fees and rules is somewhat annoying |
| 09:32 | <annevk> | and I think at that point the reason we're not joining |
| 09:33 | <annevk> | Bert Bos would have been annoyed, but that's okay; I'm sure he's no fan of XSL-FO either |
| 09:33 | <MikeSmith> | I don't think Bert was to blame for that decision |
| 09:34 | <MikeSmith> | I've heard other names |
| 09:34 | <MikeSmith> | the discussion is probably documented somewhere |
| 09:34 | <MikeSmith> | karlcow would probably know |
| 09:34 | <MikeSmith> | anyway, ancient history |
| 09:35 | <MikeSmith> | for now, it would be nice at least if ECMA and the TC would join us here in this century and actually make the normative spec available in HTML form |
| 09:35 | <MikeSmith> | with actual useful cross-references and such |
| 09:37 | <MikeSmith> | https://twitter.com/#!/reinmarpl/status/113995021006536704 |
| 09:37 | <MikeSmith> | Edward V Berard: ‘Walking on water and developing software from a specification are easy, if both are frozen’ |
| 09:38 | <MikeSmith> | cool, I think I have found a way to discover smart people on twitter |
| 09:38 | <MikeSmith> | just look for people who tweet about es5.github.com |
| 09:39 | <MikeSmith> | it's like an intelligence test |
| 09:50 | <woef> | Did you guys notice the "Adobe warns" article was apparently pulled from techworld? |
| 09:51 | <erlehmann> | tetris time ▙▄ |
| 09:52 | <jgraham> | zcorpan: Umm, depending on which part of the history API sicking had in mind, http://w3c-test.org/html/tests/submission/Opera/historyinterface/ |
| 09:52 | <jgraham> | (you retweeted, so you are guilty by association) |
| 09:56 | <zcorpan> | jgraham: oooh, wasn't aware we had released history api tests |
| 09:56 | <zcorpan> | nice! |
| 09:57 | <zcorpan> | (if i have reviewed them, i have a bad memory) |
| 09:58 | <jgraham> | Hehe |
| 09:58 | <jgraham> | I don't remember who reviewed them. Possibly I did |
| 09:59 | <jgraham> | I at least reemmber reading them |
| 10:02 | <jgraham> | It would be nice if someone could review them for the W3C |
| 10:02 | <jgraham> | So that they can go into approved and Microsoft can't pretend they don't exist |
| 10:03 | <annevk> | does Microsoft have such a bad implementation then? |
| 10:04 | <annevk> | would be nice if the files had somewhat more descriptive names |
| 10:04 | <jgraham> | In the tests? Yeah |
| 10:04 | <jgraham> | But they don't really group logically |
| 10:05 | <jgraham> | (also, you can't review them, in case you were planning to) |
| 10:05 | <jgraham> | (since apparently you are horribly biased by working at Opera) |
| 10:05 | <annevk> | how did http://w3c-test.org/html/tests/approved/attributes/anchor_href.htm ever get approved? |
| 10:05 | <annevk> | or http://w3c-test.org/html/tests/approved/audio/audio_001.htm for that matter |
| 10:06 | <annevk> | seems that all the tests that are not using testharness.js should go back for review |
| 10:14 | <zcorpan> | annevk: send email to -testsuite |
| 10:19 | <annevk> | done |
| 12:23 | <LBP> | Hi there. is there a similar way using html5 media to pass parameters to a server if user is seeking when using the H264-server-module here: http://nginx.org/en/docs/http/ngx_http_mp4_module.html |
| 12:24 | <LBP> | many flash players implement it, but haven't found a way using in html5 media. |
| 12:32 | <zcorpan> | Hixie: http://forums.whatwg.org/bb3/viewtopic.php?f=3&t=4716&p=7434#p7434 |
| 15:38 | <zcorpan> | do we want createElement to stringify null to "" ? |
| 15:44 | <jgraham> | I don't *want* anything to |
| 15:44 | <jgraham> | web compat may or may not demand it |
| 17:14 | <timeless> | Hixie: afaict, you really do need to use Metro to test IE10 |
| 17:14 | <timeless> | (which means you need to have VTx enabled in bios typically... stupid bios) |
| 17:15 | Philip` | 's CPU doesn't even support that, regardless of BIOS :-( |
| 17:31 | <timeless> | Philip`: well, you could resize your partitions and install to metal |
| 17:32 | <timeless> | but most people try to avoid that :) |
| 17:33 | <dglazkov> | good morning, Whatwg! |
| 17:51 | <AryehGregor> | Problem: Unity has no taskbar, so how am I supposed to see an XChat window flashing when someone pings me? |
| 17:51 | AryehGregor | has started to agree with the camp that feels that Ubuntu deploys severely half-baked ideas |
| 17:52 | <zewt> | i've never used ubuntu desktop, but i've given up on ubuntu server and gone back to debian |
| 18:11 | <timeless> | AryehGregor: try Metro :) |
| 18:11 | <AryehGregor> | timeless, what's Metro? |
| 18:12 | <timeless> | well, Windows 8 |
| 18:19 | <annevk> | zcorpan, I don't get e.g. http://www.w3.org/Bugs/Public/show_bug.cgi?id=14169 |
| 18:19 | <annevk> | zcorpan, WebKit/Gecko/Presto are the ones that do null->"" rather than null->"null" for most methods |
| 18:19 | <annevk> | zcorpan, surely if the plan is to switch that default we don't need to preserve it for new methods like that? |
| 18:42 | <annevk> | Hixie, you around? Can we rename DOMSettableTokenList to DOMMutableTokenList? |
| 18:42 | <annevk> | Anyone else with opinions? |
| 18:43 | <annevk> | Hmm, never mind |
| 18:43 | <zcorpan> | annevk: they don't do null->"null" for all methods |
| 18:44 | <annevk> | no they currently do null->"" |
| 18:44 | <annevk> | but they're planning on changing that |
| 18:45 | <annevk> | I don't think that has happened yet though, unless I missed something |
| 18:45 | <zcorpan> | i mean they don't do null->"" for all methods |
| 18:47 | <annevk> | How does that affect methods were they do null->""? |
| 19:06 | <TabAtkins> | shepazu: Still need me, or do you have it under control? |
| 19:07 | <shepazu> | hey, TabAtkins |
| 19:07 | <shepazu> | I think we got it… one of a million details I overlooked in rushing to announce in time |
| 19:07 | <shepazu> | thanks for the "encouragement" :) |
| 19:07 | <TabAtkins> | Sorry for calling you a bad person... but you were. ^_^ |
| 19:08 | <shepazu> | the newest code isn't in place yet, but it's much improved already |
| 19:08 | <shepazu> | oh, do you think I designed that page? you are mad, sir! |
| 19:08 | <TabAtkins> | Ah, that's a relief. |
| 19:08 | <shepazu> | I have neither the talent to make something look that good, nor the abilitty to make code so bad |
| 19:09 | <shepazu> | * ability |
| 19:09 | <shepazu> | no, it was an outside design firm that I had limited bandwidth with |
| 19:14 | <timeless> | url? |
| 19:14 | <TabAtkins> | http://w3.org/conf |
| 19:15 | <shepazu> | TabAtkins: Tantek also sent some recommendations for the code, which we've tried to integrate |
| 19:15 | <shepazu> | there's still one pesky bit though... |
| 19:15 | <annevk> | wow |
| 19:15 | <annevk> | no date in the URL |
| 19:15 | <shepazu> | annevk: :D |
| 19:15 | <annevk> | I wonder how many people on the Team lost their teeth trying to get that through |
| 19:16 | <zewt> | no date = "don't know whether this is recent enough for me to care about clicking" heh |
| 19:16 | <annevk> | zewt, clearly you are inexperienced with dated URLs on w3.org |
| 19:16 | <shepazu> | I was actually hoping for w3conf.org, but there is a preference for "cool uris" |
| 19:16 | <zewt> | yeah that /conf link definitely should have a date, since it's dated information |
| 19:16 | <TabAtkins> | zewt: Given that WebIDL is still behind a 2006 date, that's not a good metric. |
| 19:16 | <zewt> | that's "not knowing how or when to use dates", which is a different matter entirely, heh |
| 19:17 | <shepazu> | having persistent tools at W3C behind dated URIs is exceedingly unintuitive |
| 19:18 | <shepazu> | annevk: actually, it wasn't so hard getting /conf, I was pleased to find |
| 19:18 | <zewt> | unimpressed with this vincent guy working on the "mouse lock" stuff |
| 19:18 | <TabAtkins> | zewt: ? |
| 19:18 | <zewt> | he's of the "ignore feedback i don't like" variety, apparently |
| 19:19 | <shepazu> | who isn't? |
| 19:19 | <zewt> | heh :P |
| 19:19 | <TabAtkins> | zewt: Since I sit next to him and help him do the standards work, I'm pretty sure that's not true. |
| 19:20 | <zewt> | well, he's entirely ignored my feedback, and that's the metric i use :) |
| 19:20 | <TabAtkins> | I'd have to look at your feedback. Maybe you just gave bad feedback. ^_^ |
| 19:21 | <shepazu> | zewt: are you trying to force feedback on joysticks? |
| 19:22 | <TabAtkins> | mouselock, not joystick. |
| 19:22 | <shepazu> | shut up, TabAtkins, your response wasn't funny |
| 19:22 | <TabAtkins> | Ah, I see. Well, same to you apparently. ^_^ |
| 19:22 | <shepazu> | :D |
| 19:22 | <zewt> | just feels like he's speccing something that will dead end nontrivial features of hardware, with the "we can do that in v2" excuse (that doesn't work if the v1 api is designed in a way that can't be logically extended in that way) |
| 19:23 | <shepazu> | come on, "force feedback on joysticks" is funny |
| 19:23 | <TabAtkins> | I maintain that it wasn't topical enough to be funny. |
| 19:23 | <zewt> | i'm going to lock your ... i can't be bothered to think up the rest of this joke |
| 19:24 | <annevk> | he didn't reply to your emails? |
| 19:25 | <annevk> | is mouse lock getting published as WD or is this some Google effort that sort of looks like a spec but is not really? |
| 19:25 | <TabAtkins> | The former. |
| 19:25 | <TabAtkins> | We're trying to get it in as an official deliverable of the Events WG. |
| 19:25 | <annevk> | last I saw it was a Google Docs document... |
| 19:25 | <TabAtkins> | Yes, that was the initial draft work. |
| 19:25 | <Philip`> | Someone needs to make an API for web pages to control the vibration motors in console game controllers attached to PCs - lots of sites could probably make good use of that to enhance videos |
| 19:26 | <annevk> | then technically they are required to reply to your comments in some manner |
| 19:26 | <TabAtkins> | Philip`: That'll be in v2 of the joystick API. It'll be limited to joysticks that are actually active, so you can't just vibrate every joystick connected. |
| 19:26 | <annevk> | though if nobody else agrees with you or feels like speaking up it might be a lost cause anyway |
| 19:26 | <zewt> | the thread pretty much ended with me giving some more feedback, and then some crickets ... just not much of an incentive to try to help out with that API |
| 19:27 | <annevk> | it's pretty easy for WGs to deal with formal objections when it comes down to it |
| 19:27 | <TabAtkins> | zewt: What's your mailing list name? |
| 19:27 | <annevk> | I wonder mostly how that API will tie in with <dialog> |
| 19:27 | <TabAtkins> | annevk: I don't think there's any connection between the two. |
| 19:28 | <Philip`> | http://wastedseconds.com/?p=1917 - there's a whole genre of games based on the vibration feature |
| 19:28 | <zewt> | tab: Glenn |
| 19:28 | <annevk> | TabAtkins, isn't there for <dialog modal> to some extent? hmm |
| 19:29 | <TabAtkins> | annevk: No, mouselock hides the cursor and starts reporting only deltas. Different entirely from modal dialogs. |
| 19:29 | <zewt> | (nb: I'm just grumping on IRC; if I wanted to put the needed energy into it I'd ping the thread myself) |
| 19:30 | <annevk> | interesting |
| 19:30 | <annevk> | guess I should actually read it |
| 19:30 | <TabAtkins> | zewt: Ah, okay. We discussed that a lot. It's... hard. |
| 19:30 | <TabAtkins> | annevk: Yeah, it's mainly useful for doing FPSes in-browser, though there are a couple more use-cases. |
| 19:30 | <zewt> | TabAtkins: if you'll pardom the annoying non-implementor perspective, I don't see what's hard about it |
| 19:31 | <TabAtkins> | zewt: The problem is reporting something useful. As far as we know, unfiltered input doesn't necessary correspond to anything sane. |
| 19:31 | <zewt> | raw inputs are in the device's DPI, which is reported by the device itself and can be queried |
| 19:31 | <TabAtkins> | In many cases you *would* want un-accelerated input, but still have it filtered in screen space. |
| 19:31 | <TabAtkins> | zewt: I'm not down with the specifics, but I believe it's more complicated than that. |
| 19:32 | <TabAtkins> | zewt: I can't recall if it's because devices lie, or there's another layer of stuff you have to worry about, or what. |
| 19:32 | <TabAtkins> | I can talk to Vincent today if you'd like. |
| 19:32 | <zewt> | as long as it converts to a sensible resolution it's perfectly usable--that is, if you mean to report 400 DPI, then reporting in 800 DPI is probably okay, but clearly 10000 DPI is not |
| 19:33 | <zewt> | that's the same as the OS's default mouse cursor sensitivity; it mostly needs to be something sane |
| 19:33 | <zewt> | (same principle, I mean, not "same units" or anything) |
| 19:34 | <timeless> | shepazu: oh... that conference is in Seattle |
| 19:34 | <timeless> | shepazu: um |
| 19:34 | <timeless> | could /conf/ please highlight Seattle? |
| 19:34 | <timeless> | there's too much text for me to find that keyword |
| 19:35 | <Hixie> | annevk: send mail (though why would we change that?) |
| 19:35 | <timeless> | ideally Seattle would be below ' |
| 19:35 | <timeless> | '15 and 16' |
| 19:35 | <Hixie> | nessy: i should be in in about 25 min |
| 19:35 | <shepazu> | timeless: yes, the next round of fixes has that change exactly |
| 19:35 | <timeless> | +1 |
| 19:35 | <shepazu> | just waiting for it to get pushed |
| 19:35 | <timeless> | can i get a ping when it's pushed? |
| 19:35 | timeless | hates polling |
| 19:36 | <shepazu> | also adds microforomats, thanks to tantek's suggestion |
| 19:36 | <shepazu> | timeless: I have no clue, out of my hands… I sent them the changes late last night |
| 19:36 | <TabAtkins> | +1 to the microformats! Yay! |
| 19:36 | <timeless> | interesting |
| 19:36 | <shepazu> | timeless: but you already have the info…. you don't need to look again…. |
| 19:36 | <timeless> | the ics file is hosted on w3conf.org |
| 19:37 | <timeless> | and it claims it's in Bellevue WA not Seattle |
| 19:37 | <shepazu> | timeless: Bellevue is right there |
| 19:37 | <timeless> | it's 10mi/15mins away! |
| 19:37 | <shepazu> | 9 miles from Seattle city center |
| 19:37 | <timeless> | across two bridges! |
| 19:38 | <TabAtkins> | Rent a car, commie. |
| 19:38 | <timeless> | http://maps.google.com/maps?q=Bellevue,+WA,+United+States+to+Seattle,+WA,+United+States&saddr=Bellevue,+WA,+United+States&daddr=Seattle,+WA,+United+States&hl=en&sll=43.641852,-79.612715&sspn=0.010916,0.022724&geocode=FQl61gIdmV23-ClBYKajz2uQVDFlB9DqglTPug%3BFcJp1gIdWVy1-ClVM-iTLBCQVDGa1URpRmUlEA&vpsrc=0&t=h&z=13 |
| 19:38 | <shepazu> | 20 minutes from SEA |
| 19:38 | <timeless> | most places won't rent to people w/o driver's licenses |
| 19:38 | <TabAtkins> | There's a way to fix that. |
| 19:38 | <shepazu> | timeless: take a bus :) |
| 19:38 | <timeless> | yeah, actually, starting to work on that today was on my todo list |
| 19:39 | <timeless> | > Sorry, we don't have transit schedule data for a trip from Bellevue, WA to Seattle, WA at the time and date you specified. |
| 19:39 | <shepazu> | timeless: most people don't know where Bellevue is, so we say "Seattle" to avoid confusion |
| 19:39 | <timeless> | transit is 34 mins fwiw |
| 19:40 | <timeless> | 35mins during morning rush :) |
| 19:40 | <shepazu> | timeless: are you seriously bickering about these details? :) |
| 19:40 | <timeless> | oops, sorry, it's 31mins going the right direction :) |
| 19:41 | <timeless> | no, i'm actually not really allowed to leave the province much until mid december |
| 19:41 | <karlcow> | hmmm "vibrate every joystick connected". I can see business opportunities |
| 19:41 | <timeless> | otherwise i lose health coverage |
| 19:41 | <TabAtkins> | karlcow: Way ahead of you. |
| 19:41 | <timeless> | TabAtkins: are usb missile launchers covered under joysticks? |
| 19:41 | <karlcow> | yes there… was even one which was connected to ipods I think |
| 19:42 | <timeless> | http://www.thinkgeek.com/geektoys/warfare/8a0f/ |
| 19:42 | <TabAtkins> | timeless: Do they expose the HID joystick interface? |
| 19:42 | <timeless> | $25 in stock |
| 19:42 | <karlcow> | but I guess I'm not talking about the same type of joysticks |
| 19:42 | <TabAtkins> | timeless: Various people plan to add generic HID interaction to javascript. The joystick API is a specialized subset of that. |
| 19:42 | timeless | worries |
| 19:42 | <karlcow> | ah yes… OhMiBod |
| 19:42 | <timeless> | i don't know that i want random web pages to talk to medical instruments |
| 19:43 | <timeless> | and will it distinguish between USB and BT? |
| 19:43 | <TabAtkins> | timeless: Presumably there's permissions involved somewhere. |
| 19:43 | <TabAtkins> | timeless: No, I think the HID interface is agnostic as to the wire you use. |
| 19:43 | <timeless> | i generally presume that part will be done wrong :) |
| 19:43 | <TabAtkins> | timeless: That's part of why joysticks are being separated out from the general case, because they're easier to solve permissions for. ^_^ |
| 19:44 | <TabAtkins> | (Hitting a button on the joystick grants access to the current page.) |
| 19:44 | <TabAtkins> | (And you can revoke access by turning off the joystick.) |
| 19:45 | <timeless> | "turning off the joystick"?? |
| 19:45 | <timeless> | that seems like a bad requirement |
| 19:45 | <TabAtkins> | Or disconnecting, whatever. Many joysticks can be turned off, though. |
| 19:45 | <timeless> | many can't |
| 19:45 | <TabAtkins> | That's how you make your Xbox stop trying to listen to a controller. |
| 19:45 | <timeless> | and many places don't like people walking out with their joysticks or similar usb devices |
| 19:45 | <timeless> | and tend to not expose the usb plug connection |
| 19:45 | <TabAtkins> | ...what? |
| 19:46 | <timeless> | never seen a phone store that objected to people walking away with their display phones? |
| 19:46 | <zewt> | well, joystick security is fundamentally easier--it's very hard to make a convincing case for any serious security problems related to joysticks |
| 19:46 | <TabAtkins> | Presumably an arcade that exposes all its games through a web interface won't mind the active page always having access. |
| 19:46 | <zewt> | (letting scripts vibrate a joystick or read inputs that they shouldn't is annoying and should be prevented, but challenging to call a real security problem) |
| 19:47 | <timeless> | TabAtkins: you didn't say that they lose access when the page flips |
| 19:47 | <timeless> | so i presumed they don't |
| 19:47 | <TabAtkins> | zewt: Right, so false positives on an access grant aren't very worrying. |
| 19:47 | <TabAtkins> | timeless: Do you mean when the page navigates, or when the tab switches? |
| 19:47 | <timeless> | the latter |
| 19:48 | <TabAtkins> | I dunno what the spec currently says about that. |
| 19:48 | <TabAtkins> | Either way, not a huge deal. Reloading the page would wipe access too. |
| 19:48 | <timeless> | i've seen various custom browser ui's |
| 19:48 | <timeless> | they tend to mess up the nav controls :) |
| 22:09 | <AryehGregor> | It's annoying that hg-git can't somehow magically make the commit id's match up. Now I have nothing unambiguous to call the commits. |
| 22:09 | <AryehGregor> | I'll just have to give the first line of the summary, I guess. |
| 22:17 | <jgraham> | That would hve to be very magic indeed |
| 22:17 | <Hixie> | TabAtkins: if people are ok with the @rule for style scoped="", any chance you can propose it to the csswg and get it okayed? |
| 22:17 | <jgraham> | You could call them someSHA1(git):someOtherSHA1(hg) |
| 22:17 | <Hixie> | TabAtkins: (the csswg gets antsy when i spec css things) |
| 22:18 | <gsnedders> | AryehGregor: It would have to change how one of them computes the sha1, which neither allows, for data integrity checking purposes |
| 22:18 | <AryehGregor> | I know. |
| 22:21 | <TabAtkins> | Hixie: Yeah, I can. |
| 22:22 | <Hixie> | cool, thanks |
| 22:22 | <TabAtkins> | Hixie: You know, you could suggest it *before* putting it in your spec... |
| 22:22 | <Hixie> | TabAtkins: that's what you're doing :-D |
| 22:24 | <Philip`> | AryehGregor: Doesn't it do something like tag the commits with the Hg revision ID? |
| 22:24 | <AryehGregor> | Philip`, you mean the git revision id? This is an hg extension, not a git extension. |
| 22:24 | Philip` | uses an SVN-to-Hg thing which does that, so he can refer to "svn.r1234" in the Hg repository |
| 22:24 | <Philip`> | AryehGregor: Oh, whichever way around it is |
| 22:25 | AryehGregor | looks |
| 22:25 | <hober> | Hixie: so is the idea to have :scope and @global but not :root? |
| 22:25 | <jgraham> | So, uh, the W3C is announcing a conference 2 months in the future? |
| 22:25 | <jgraham> | I am not an expert |
| 22:25 | <TabAtkins> | hober: Yeah. |
| 22:25 | <TabAtkins> | hober: Or, wait, no. |
| 22:25 | <AryehGregor> | Philip`, I don't see any tags or such: https://dvcs.w3.org/hg/editing |
| 22:25 | <jgraham> | But doesn't one generally ive more notice? |
| 22:25 | <Hixie> | hober: define "have"? |
| 22:25 | <jgraham> | +g |
| 22:25 | <TabAtkins> | hober: :root still exists. It refers to the root of the document, like normal. |
| 22:26 | <TabAtkins> | hober: The proposal is just that normal selectors are scoped to the scoping element, and won't match anything outside of that (so :root never matches, unless <style scoped> is a child of the root). |
| 22:26 | <TabAtkins> | hober: And then a selector in an @global block isn't scoped, but still sets :scope to the scoping element. |
| 22:26 | <jgraham> | And by "announcing" I mean "has a website with almost no information, but an annoying parallax effect thing for" |
| 22:27 | <hober> | TabAtkins: OK. That's not terribly crazy. :) |
| 22:28 | <Hixie> | woot :-) |
| 22:29 | <hober> | so "foo :scope bar" in a <style scoped> never matches unless it gets @global'ed |
| 22:29 | <TabAtkins> | Yes. |
| 22:31 | <hober> | word. |
| 23:01 | <AryehGregor> | Blech, compiz is chewing up my CPU. |
| 23:10 | <AryehGregor> | I guess I have to start using my DVCSes like real DVCSes: save patches, wait for tests to finish running, amend with new expected values before committing. |
| 23:10 | <AryehGregor> | Otherwise I can't work on multiple things at once. |
| 23:11 | <gsnedders> | AryehGregor: branches? |
| 23:11 | <AryehGregor> | gsnedders, guess so. |
| 23:11 | <gsnedders> | AryehGregor: commit/squash? |
| 23:11 | <AryehGregor> | "squash"? |
| 23:11 | <AryehGregor> | What does that do? |
| 23:13 | <divya> | combines many commits into one |
| 23:13 | <AryehGregor> | Interesting. Is that hg or git? |
| 23:15 | <AryehGregor> | Ah, mq. |
| 23:15 | <AryehGregor> | Should've guessed. |
| 23:16 | <AryehGregor> | I really wish I had a faster CPU with more cores. |
| 23:16 | AryehGregor | kicks himself |