| 00:12 | <grom358> | Test case: http://pastebin.com/pRUqGS9G .. The spec parses the hello string as the end of the script tag. |
| 00:14 | <grom358> | that does not seem like the wrong behavior to me |
| 00:16 | <grom358> | because if the javascript is in external file then you don't get that behavior |
| 05:02 | <zewt> | what the holy hell |
| 05:03 | <zewt> | firefox went from a regular annoying fullscreen "hit button to exit" notification, to an obnoxious intrusive fullscreen dimming (absurd), and apparently that wasn't stupid enough, since now it opens a dialog on top of the fullscreened page that you have to click away |
| 05:03 | zewt | switches youtube back to flash; screw it |
| 05:04 | <zewt> | people oversealously contriving make-believe security issues that don't actually exist have taken over all UX design in firefox; 9/11fox |
| 05:05 | <zewt> | end disgusted rant |
| 06:55 | <zcorpan> | Hixie: +00:10.000 --> 00:04.000 looks wrong |
| 06:55 | <zcorpan> | http://html5.org/r/7442 |
| 06:56 | <zcorpan> | Hixie: the example rendering also looks wrong since it doesn't minimize delta |
| 06:59 | <zcorpan> | http://dev.w3.org/html5/webvtt/ doesn't follow the style rules for CLA -- the w3c logo isn't in the top left corner |
| 07:03 | <zcorpan> | s/style rules for CLA/CG report requirements/ |
| 07:23 | ashemedai | chuckles |
| 07:23 | <ashemedai> | The irony of webplatform.org not passing validation |
| 07:27 | <AryehGregor> | Where's the correct XHR spec these days? |
| 07:27 | <AryehGregor> | This? http://dvcs.w3.org/hg/xhr/raw-file/tip/Overview.html |
| 07:27 | <AryehGregor> | Looks right, last updated yesterday and edited by annevk. |
| 07:38 | <zcorpan> | AryehGregor: it says *previous* editor is annevk |
| 07:38 | <zcorpan> | AryehGregor: if you want the spec annevk edits, it's http://xhr.spec.whatwg.org/ |
| 07:39 | <AryehGregor> | Ah, okay. |
| 07:39 | <AryehGregor> | Thanks. |
| 08:17 | <annevk> | yeah dunno how that is going to work out... |
| 08:55 | <feross> | hey, anyone here? |
| 08:55 | <feross> | so, I just wrote this blog post about how to attack the javascript fullscreen api: http://feross.org/html5-fullscreen-api-attack/ |
| 08:56 | <feross> | and i realized that this isn't the most effective way to get the api fixed |
| 08:56 | <feross> | it's actually not really an api issue as much as a UX/UI design issue |
| 08:56 | <feross> | i'm curious what you recommend i do |
| 08:57 | <feross> | file a bug on the spec? file a bunch of browser bugs? |
| 09:06 | <AryehGregor> | feross, you could post to the spec mailing list. IIRC, the API doesn't allow the page to accept more than highly limited user input in fullscreen mode, so attackers can't actually get them to type anything in without abruptly exiting full-screen mode. Also, Flash has such a feature already and it's not widely used for phishing. |
| 09:07 | <feross> | AryehGregor: thanks for the reply. i was able to use keyboard input while in fullscreen mode in both firefox and chrome, so despite what the spec says i think that browsers are allowing this to happen. safari blocks keyboard input. |
| 09:07 | <AryehGregor> | The spec doesn't seem to give useful guidance. Definitely a spec bug. http://dvcs.w3.org/hg/fullscreen/raw-file/tip/Overview.html#security-and-privacy-considerations |
| 09:08 | <AryehGregor> | I haven't been following discussions. |
| 09:08 | <AryehGregor> | I'd post to the appropriate spec mailing list. |
| 09:09 | <feross> | okay, will do that. thanks for the help. |
| 09:10 | <jgraham> | SO I recall there being discussion about this before |
| 09:10 | <AryehGregor> | Needless to say, phishing has been discussed very extensively with this feature, so I'm guessing the people involved will have answers, but I'm not one of them, so I don't. :) |
| 09:11 | <jgraham> | The use case for keyboard input is, of course, fullscreen games |
| 09:11 | <AryehGregor> | Yes, lots and lots of discussion. IIRC, security issues delayed the feature for a lot of time. |
| 09:11 | <AryehGregor> | jgraham, the idea I saw floated was that most games could make use with a very limited set of keys, like arrow keys and so on, that would be useless to phishers. |
| 09:11 | <AryehGregor> | Arrows, Enter, space, whatever. |
| 09:11 | <jgraham> | Well that clearly isn't true |
| 09:11 | <AryehGregor> | Ones that wanted more would have to get some type of elevated permission. |
| 09:12 | <AryehGregor> | It's true for a lot of games, like simple platformers. |
| 09:12 | <AryehGregor> | You could allow arrows, Enter, space, and modifier keys with no real problems. |
| 09:12 | <jgraham> | I thought the consensus was that having a clickthrough dialog was enough that keyboard input could be allowed |
| 09:12 | <jgraham> | Sure, but people are trying to do more advanced games on the web |
| 09:12 | <AryehGregor> | I don't remember, I only followed part of the discussions. |
| 09:13 | <jgraham> | What keys does bananabread sdepend on? |
| 09:13 | <AryehGregor> | It seems like the current spec doesn't distinguish between allowing keyboard input and not. |
| 09:13 | <AryehGregor> | Some earlier versions did. |
| 09:13 | <zcorpan> | AryehGregor: games need more than the arrows. e.g. i need about half the keyboard for quake |
| 09:13 | <AryehGregor> | Sure, but a lot of games aren't Quake. |
| 09:14 | <zcorpan> | so? |
| 09:14 | <AryehGregor> | More complicated games certainly need the whole keyboard. |
| 09:14 | <jgraham> | Bananabread already uses way more keys |
| 09:14 | <AryehGregor> | So the games that don't need much of the keyboard could get away with less annoying security stuff. |
| 09:14 | <AryehGregor> | The ones that do can opt in to more annoying security stuff. |
| 09:14 | <AryehGregor> | That's what an earlier API version did. |
| 09:14 | <AryehGregor> | Seems the current one doesn't. |
| 09:16 | <zcorpan> | yeah ok. i recall that being proposed |
| 09:16 | <jgraham> | I'm not sure "make the experience simpler for a subset of use cases" is a winning proposition compared to "make the experience consistent across all use cases" |
| 09:16 | <AryehGregor> | It seems the conclusion was to agree with you. |
| 09:17 | <AryehGregor> | Doesn't that kind of stink for, e.g., full-screen video? |
| 09:17 | <AryehGregor> | Or is the idea that that gets special-cased somehow, using nonstandard magic? |
| 09:17 | <zcorpan> | if the user trusts youtube.com, he can always allow youtube.com |
| 09:17 | <jgraham> | (do video players typically use more than just arrow keys for keyboard a11y?) |
| 09:18 | <jgraham> | Plus tab, plus modifiers |
| 09:18 | <AryehGregor> | Dunno. |
| 09:20 | <jgraham> | Seems like youtube also supports 0-9 to move to various points in the video + home + end |
| 09:21 | <jgraham> | (and space) |
| 09:33 | <annevk> | AryehGregor: I kinda think keyboard input should work, if it's made clear to the user that he entered fullscreen and that he can exit it, it should all be okay... |
| 09:34 | <AryehGregor> | It's pretty easy to get a substantial fraction of users to not notice that they entered fullscreen. |
| 09:34 | <AryehGregor> | Even if 90% will spot it, that doesn't matter -- phishing is cheap, and you still get the 10%. |
| 09:35 | <AryehGregor> | More generally, Flash already has fullscreen and it doesn't seem to have caused catastrophic security failures. If we match its features, we don't have to worry much. If we're more liberal, like allowing unrestricted keyboard input when Flash doesn't (does it?), that's a lot riskier because it's uncharted territory. |
| 09:44 | <annevk> | crippling functionality of the web platform lets proprietary platforms take over; it's not exactly a zero-sum game |
| 09:46 | <ShaneHudson> | Is the WHATWG involved in this new wiki the W3C have created? Looks nice a nice site to destroy w3schools! |
| 09:46 | <AryehGregor> | No, but I don't think "it's made clear to the user" is good enough by itself. A substantial number of users aren't going to even know what "fullscreen" means. The web is used by lots of people who aren't so clever with computers, or just aren't paying attention in the slightest. |
| 09:46 | <AryehGregor> | Anyway, I'll leave that up to the implementers. They have security teams for a reason. |
| 09:53 | <jgraham> | I like the fact that the TPAC site says "Remember: do not microblog W3C Member confidential matters". Presumably those looking to leak sekrits should write full blog entries. |
| 10:02 | <darobin> | jgraham: it's just based on experience. People don't blog confidential stuff, but they mistakenly livetweet stuff |
| 10:02 | <darobin> | not that there's all that much confidential stuff going on in the first place — finding something to leak is actually hard work :) |
| 10:04 | <jgraham> | s/microblog/publish/ sould make the whole thing less silly |
| 10:05 | <jgraham> | *would |
| 10:07 | <darobin> | in fairness, I think that not caring about such trivial details would be less silly |
| 10:09 | <jgraham> | Me? Well maybe, but I doubt I am the only one that would find it faintly ridiculous that there is a swpecific admonishment for "microblogging" but not for any other form of communication |
| 10:10 | <jgraham> | Is HTMLImageElement.complete too difficult an example to use for explaining testharness.js? Pros: has obvious list of cases, has sync and async cases. Cons: Slightly obscure, depends on networking, one case is pretty hard to test, only really uses assert_true |
| 10:12 | <jgraham> | low |
| 10:12 | <jgraham> | Ohh, focus doesn't follow brain |
| 10:36 | <zcorpan> | jgraham: .complete is pretty unique in that it can change value while script is running |
| 10:37 | <zcorpan> | (is that needed for web compat? or did we just blindly copy ie?) |
| 10:37 | <jgraham> | Yeah, that's kind of icky |
| 10:37 | <jgraham> | One reason I don't like |
| 10:37 | <jgraham> | it |
| 10:39 | <jgraham> | But I am not doing well thinking of a better alternative that naturally has sync and async parts |
| 10:40 | <jgraham> | Everything to do with <iframe>s is whacked beyond belief |
| 10:40 | <jgraham> | Maybe localStorage? |
| 10:42 | <zcorpan> | seems chrome doesn't update it while script is running. http://software.hixie.ch/utilities/js/live-dom-viewer/saved/1828 |
| 10:42 | zcorpan | files spec bug |
| 10:44 | <zcorpan> | https://www.w3.org/Bugs/Public/show_bug.cgi?id=19388 |
| 10:44 | <jgraham> | It also didn't update it in the timeout for me… |
| 10:44 | <zcorpan> | does for me |
| 10:45 | <zcorpan> | might be racy? |
| 10:45 | <jgraham> | yeah, seems to be OK if you increase the timeout to 10ms |
| 10:46 | <zcorpan> | maybe it doesn't decode the image until it paints it, i.e. until the second alert is discarded, and then there's a race between the timeout and image decode |
| 10:46 | <zcorpan> | or something, i'm just guessing here |
| 10:46 | <jgraham> | Better to write that test using sync XHR than alert, I think |
| 10:46 | <jgraham> | Or maybe even just a loop |
| 10:48 | <jgraham> | http://software.hixie.ch/utilities/js/live-dom-viewer/saved/1829 |
| 10:48 | <jgraham> | Huh? |
| 10:49 | <jgraham> | Gecko is giving false everywhere |
| 10:49 | <jgraham> | Turns out that this is a really interesting case :) |
| 10:50 | <zcorpan> | yeah, was just going to say. but it doesn't do that with my version |
| 10:50 | <jgraham> | Right, but alert is special sauce in gecko |
| 10:50 | <zcorpan> | yes |
| 10:51 | <jgraham> | Makes everything taste weird |
| 10:51 | <zcorpan> | it seems interesting to test how things interact with alert |
| 10:52 | <zcorpan> | and sync xhr.. |
| 10:53 | <jgraham> | sync XHR gives the same result as alert |
| 10:54 | <jgraham> | In gecko |
| 10:54 | <zcorpan> | jgraham: why do you need a feature with both sync and async? why not use separate features? |
| 10:55 | <jgraham> | zcorpan: Just seems easier to give the presentation an overall arc |
| 10:55 | <zcorpan> | ok |
| 10:56 | <zcorpan> | use <script>! |
| 10:56 | <jgraham> | "Here's a simple case, here's a more complex case (this is the point at which people stop listening) this is a convenience function, these are some general principles to keep in mind" |
| 10:56 | <jgraham> | Haha |
| 10:58 | <jgraham> | Homework: write a testsuite for <script> scheduling. Bonus credit for <script> + XSLT |
| 10:59 | <AryehGregor> | darobin, I'm getting an error when trying to run pegjs in webidl.js: Error: require.paths is removed. Use node_modules folders, or the NODE_PATH environment variable instead. |
| 10:59 | <AryehGregor> | Any idea what's up with that? |
| 11:00 | <darobin> | AryehGregor: I haven't touched that code in a while, so sadly no |
| 11:00 | <AryehGregor> | Oh, wait. |
| 11:00 | <AryehGregor> | I think I see. |
| 11:00 | <AryehGregor> | It was looking at an outdated local version. |
| 11:00 | <darobin> | but I know what the error is: something is using require.paths and it really shouldn't |
| 11:00 | <darobin> | ah, good |
| 11:00 | <zcorpan> | jgraham: :-) |
| 11:01 | <darobin> | AryehGregor: I'll be updating webidl.js as soon as the HTML WG gets into full testing crunch, i.e. hopefully pretty soon |
| 11:01 | <AryehGregor> | darobin, good to hear. |
| 11:01 | darobin | has to jump out, bbiab sorry |
| 11:02 | jgraham | hopes darobin's parachute opens |
| 11:03 | <AryehGregor> | darobin, I just submitted a one-line pull request for you. |
| 11:03 | AryehGregor | is reminded how vastly better git is than hg |
| 11:04 | <zcorpan> | jgraham: i'm having a hard time finding an api that isn't totally quirky and complex and has both sync and async parts |
| 11:04 | <jgraham> | Not just me then |
| 11:06 | <jgraham> | But what about localStorage? Or maybe I shoudl go for the simple parts of img.complete and then suggest that there is potential to do the more complex part and turn it into a full testsuite during the hack parts of ttwf |
| 11:09 | <zcorpan> | localStorage seems ok |
| 11:10 | <zcorpan> | can involve subframes if you want to show how that's done |
| 11:24 | <paul_irish> | hsivonen: there were policy decisions made at moz that disallowed importing MDN. |
| 11:25 | <zcorpan> | zewt: http://dvcs.w3.org/hg/xhr/raw-file/tip/Overview.html is also out of date, as it happens :-P |
| 11:25 | <darobin> | AryehGregor: PR merged, thanks! |
| 11:26 | <AryehGregor> | darobin, thanks! |
| 11:26 | <darobin> | no, thanks to you :) |
| 11:26 | <darobin> | hg? I thought only Ms2ger and W3C still used that |
| 11:29 | <AryehGregor> | Sadly, I deal with the W3C a lot. :( |
| 11:29 | <AryehGregor> | Also, Mozilla does, and I haven't switched to the git mirrors yet. |
| 11:32 | <annevk> | yeah I don't really get the reverse xhr forking |
| 11:32 | <annevk> | not really sure what their plan is either |
| 11:35 | <annevk> | hmm |
| 11:40 | <darobin> | AryehGregor: any group that makes you use hg, don't hesitate to ask that it moves to GH |
| 11:40 | <AryehGregor> | darobin, do you think it would make sense to make things like .members or .inheritance be empty arrays instead of null if they're not present? |
| 11:40 | <AryehGregor> | GH? |
| 11:40 | <darobin> | GitHub |
| 11:41 | <darobin> | AryehGregor: yeah, I think that would make sense, there's not useful distinction between null and empty here and it would help code against the API |
| 11:41 | <AryehGregor> | Sounds great. |
| 11:41 | AryehGregor | will probably write a patch soon, but is in the middle of something else |
| 11:41 | <AryehGregor> | (namely, documenting idlharness.js) |
| 11:41 | <darobin> | or, you know, if you're the primary committer to something that a WG is doing, you can just move to GH and tell them later... |
| 11:41 | <darobin> | whee, more docs! |
| 11:44 | <odinho> | 13:42 < AryehGregor> (namely, documenting idlharness.js) <3 |
| 11:49 | <hsivonen> | paul_irish: the policy decision of having used ShareAlike or something else? |
| 11:50 | <hsivonen> | paul_irish: on your blog post, you mentioned MDN, Wikipedia and Stackoverflow. All of those are SA, so deciding not to make WebPlatform.org SA is the policy decision that prevents importing from any of them. |
| 12:07 | <divya> | hsivonen: why is SA necessary. |
| 12:08 | <divya> | hsivonen: also where is license info for webplatform.org |
| 12:08 | <AryehGregor> | hsivonen, so that things can be imported from other sources that use SA, presumably. |
| 12:13 | <hsivonen> | divya: MDN, Wikipedia and Stackoverflow have SA, so SA is necessary to import content from them |
| 12:15 | <hsivonen> | divya: it seems like a tremendously bad use of contributor time to re-express SA content in order to clean it of copyright and, therefore, copyleft. See half-day tasks: http://docs.webplatform.org/wiki/WPD:Getting_Started |
| 12:16 | <divya> | hsivonen: it seems right now MDN turns a blind eye to scraped content that is not rendered in CC-BY-SA |
| 12:16 | <divya> | hsivonen: I was not aware they need to re-express SA content |
| 12:16 | <divya> | i assumed they could just scrape it. |
| 12:17 | <jgraham> | That wouldn't be SQ though |
| 12:17 | <jgraham> | *SA |
| 12:17 | <hsivonen> | divya: failure to enforce previously (I’m taking you at your word) doesn’t mean that it’s reasonable for webplatform.org to rely on being able to violate a license |
| 12:18 | <divya> | hsivonen: i didnt know about SA restrictions. |
| 12:18 | <divya> | ideally this content should be free to use by whoever where-ever |
| 12:18 | <divya> | i would love for commercial products to rely on this for their documentation. |
| 12:18 | <hsivonen> | CC-BY-SA allows the use in commercial docs |
| 12:19 | <divya> | how so? if it is 'share alike' |
| 12:19 | <divya> | what use is restricted? |
| 12:19 | <hsivonen> | divya: you have to share alike but you can use it for commercial purposes |
| 12:19 | <hsivonen> | divya: NC is the license component that’s non-Free and prohibits commercial use |
| 12:19 | divya | wikis |
| 12:19 | <hsivonen> | (Opera’s curriculum was CC-BY-SA-NC) |
| 12:20 | <divya> | right so Share Alike means you must use sam licensing |
| 12:21 | <hsivonen> | divya: right. but the license doesn’t prohibit commercial use |
| 12:21 | <divya> | +e |
| 12:21 | <divya> | yeah but i do not think that is worth enforcing, especially given companies have preferences for different licenses. |
| 12:21 | <divya> | either way, well we need to fix this licensing thing. |
| 12:23 | <divya> | i honestly think CC-BY-SA is a model that would make companies baulk at using content from that source. |
| 12:23 | <divya> | have had experiences with it at Adobe. |
| 12:24 | <hsivonen> | divya: plenty of companies seem to be OK with shipping a copyleft kernel on their phones |
| 12:24 | <hsivonen> | divya: the first copyleft-licensed Adobe thing I can think of is the Source Sans font |
| 12:24 | <hsivonen> | the first I can think of that is |
| 12:24 | <divya> | what is 'copy-left'? |
| 12:24 | <hsivonen> | not necessarily Adobe’s first |
| 12:25 | <hsivonen> | divya: the concept of share alike more generally than the CC ShareAlike in particular |
| 12:25 | <jgraham> | divya: SA |
| 12:26 | <AryehGregor> | So in JavaScript, how can I do something like a derived class calling a base class' constructor? I currently have A and B, where A.prototype and B.prototype both have C.prototype as their prototype. I want new A() and new B() to have the same effect. |
| 12:26 | <divya> | hsivonen: i thought it was apache license. no? |
| 12:26 | <AryehGregor> | I guess I could do function A() { C.call(this); }. |
| 12:26 | <AryehGregor> | Hmm, yeah, let me do that. |
| 12:27 | <divya> | bah its some other license called open font license. |
| 12:27 | <hsivonen> | divya: right |
| 12:28 | <hsivonen> | oh and Adobe seems to be distributing WebKit, parts of which are also copylefted |
| 12:28 | <divya> | hsivonen: yeah seems SA from reading the open font license. |
| 12:28 | <divya> | hsivonen: all i am saying is lawyers @ adobe do not like using licenses that are not the licenses they recommend. |
| 12:29 | <divya> | unless it goes through a long process of review and then figuring out if it is worth the risk |
| 12:29 | <divya> | i am sure companies like mozilla, adobe, apple have $$$s to go through such reviews |
| 12:29 | <divya> | not many companies do. |
| 12:29 | <divya> | so having it a 'share-alike' means some small companies either just risk it or do not bother. |
| 12:30 | <divya> | individuals dont give a damn of course. |
| 12:30 | <hsivonen> | my point is that Mozilla doesn’t own the copyright on all MDN content, StackExchange doesn’t for StackOverflow conten and Wikimedia doesn’t for Wikipedia, so you can’t just get SA waivers from a handful of parties to import content |
| 12:31 | <divya> | yeah it is unfortunate. I do not think SE content can be imported per se anyway. |
| 12:31 | <divya> | licensing suckss |
| 12:32 | <hsivonen> | to me, it would make sense to prefer compatibility with the existing body of potentially importable content than to prefer compatibility with the sensibilities of *potential* contributors |
| 12:33 | <divya> | i understand the cost of existing compatibility is significantly lesser, but we should worry about the cost of this model after a few months and finding how it is working or not. |
| 12:35 | <zcorpan> | why is a new header not better for <http://www.w3.org/mid/op.wlwusmc2a3v5gv@hr-desk>; ? |
| 12:41 | <AryehGregor> | darobin, would you be okay with me breaking up some of the long lines in grammar.peg as I make other changes? It's hard to read diffs with lines that long. |
| 12:43 | <darobin> | AryehGregor: I'm okay with you changing whatever you need for this to be easier for you man |
| 12:43 | <AryehGregor> | darobin, okay, great. |
| 12:56 | <zcorpan> | zewt: isn't it up to browsers' decoders to decide at which level of blurriness they want to start showing the image? |
| 13:13 | <shepazu> | hsivonen: I made the call on WebPlatform using CC-BY, rather than CC-BY-SA, and I still believe it is the right one. Yes, it will take more time and effort to build the content up to a great state, but once we get there, the content will be much more reusable, without the confusion of SA |
| 13:19 | <zcorpan> | why was Array.isArray introduced? isn't it nonsensical to introduce sugar for *one* kind of object while not solving the problem for all other kinds? |
| 13:20 | <annevk> | zcorpan: I think the problem was only with arrays |
| 13:21 | <zcorpan> | really? then what was the problem? |
| 13:21 | <annevk> | javascript:typeof [] |
| 13:21 | <annevk> | I think |
| 13:22 | <zcorpan> | Object.prototype.toString.call([]) === '[object Array]' |
| 13:25 | <zcorpan> | http://www.adobe.com/devnet/html5/articles/categorizing-values-in-javascript.html says isArray was introduced because of the different global objects problem, but that still applies for all other objects |
| 13:26 | <annevk> | http://web.mit.edu/jwalden/www/isArray.html explains |
| 13:26 | <annevk> | ' One test in this style is Object.prototype.toString.call(o) === "[object Array]", but that relies on Object.prototype.toString and Function.prototype.call not being changed (probably a good assumption but still fragile). It's also a bit more of an obvious hack than any of the other ideas.' |
| 13:29 | <zcorpan> | ah. that's true. but doesn't explain why it was only solved for arrays |
| 13:30 | <hsivonen> | shepazu: thanks |
| 13:30 | <hsivonen> | shepazu: for saying it was you—not for the call to choose CC-BY |
| 13:31 | <annevk> | zcorpan: typeof works for the other cases? |
| 13:32 | <zcorpan> | annevk: no? typeof is not useful for differentiating different objects |
| 13:34 | <shepazu> | hsivonen: uh… you're welcome? |
| 13:34 | <shepazu> | not sure what you meant by that :) |
| 13:36 | <shepazu> | to be honest, I found it disconcerting to see complaints in this channel, taking a position against the more open option, for a more restrictive content license |
| 13:37 | <annevk> | lets not generalise shepazu |
| 13:37 | <hsivonen> | shepazu: I just meant avoiding the confusion of looking like I was thanking for the license choice |
| 13:38 | <annevk> | shepazu: e.g. http://krijnhoetmer.nl/irc-logs/whatwg/20121008#l-10 |
| 13:38 | <hsivonen> | shepazu: I just think it’s a terrible waste of time to rewrite stuff to work around a license instead of going with what’s established out there |
| 13:38 | <shepazu> | like HTML5? :D |
| 13:39 | <darobin> | ouch, burn! |
| 13:39 | <hsivonen> | shepazu: from my point of view, this is another bad licensing call by the W3C |
| 13:39 | <zcorpan> | html5 wasn't rewritten to work around a license. it was rewritten because html4 was like w3schools |
| 13:39 | <shepazu> | annevk: I didn't mean to generalize… I meant "some complaints" not "all complaints" |
| 13:39 | <darobin> | <annevk> lets not generalise shepazu # yeah, having one of those is enough |
| 13:40 | shepazu | SHEPAZU ALL THE THINGS! |
| 13:40 | <annevk> | I think only hsivonen and I complained, and argued opposite sides |
| 13:40 | <annevk> | ;) |
| 13:41 | <hsivonen> | I don’t mind webplatform.org using CC-BY or CC0 even for new content |
| 13:41 | <annevk> | and I'm just calling it out because everytime someone says something here, someone broadcasts it to the outside as "the WHATWG has spoken..." |
| 13:41 | <shepazu> | annevk: that seems quite consistent with your general philosophy on licensing :) |
| 13:41 | <hsivonen> | I just think it’s terribly wasteful not to be able to import from MDN, Wikipedia or StackOverflow |
| 13:41 | <shepazu> | annevk: fair enough, but I know better |
| 13:42 | <shepazu> | hsivonen: and with all due respect to those fine sites, which I admire, I think the mistake was on their part in choosing restrictive licenses |
| 13:43 | <shepazu> | I think we made the right call for the future use of the content |
| 13:43 | <annevk> | could we not have gone for new content in the public domain? and try to move there over time reusing the existing content under their more restrictive licensing? |
| 13:43 | <annevk> | I guess time will tell how it goes |
| 13:43 | <shepazu> | annevk: yes, but that would dramatically complicate and confuse things |
| 13:44 | <hsivonen> | shepazu: CC0 avoids the confusion |
| 13:44 | <shepazu> | CC0 provides less incentive to contribute for many people who like a little recognition |
| 13:44 | <shepazu> | but yes, we considered that |
| 13:45 | <shepazu> | code is all CC0, btw |
| 13:45 | <hsivonen> | shepazu: you really want attribution to individual authors and not just to webplatform.org? |
| 13:45 | <hsivonen> | shepazu: yay for CC0 code |
| 13:45 | <shepazu> | hsivonen: for commercial uses, yes, lists of authors, not just wpd |
| 13:45 | <shepazu> | it's only fair |
| 13:46 | <hsivonen> | also terribly impractical in the long run |
| 13:46 | <shepazu> | we've already got a ton of interest in improving this site, so I think in 6-12 months the site will not suffer for a lack of imported content |
| 13:46 | <shepazu> | hsivonen: depends how it's managed |
| 13:46 | <annevk> | well, I still acknowledge contribution to specs even though they're licensed under CC0, fwiw |
| 13:47 | <annevk> | don't really need to cater to that through licensing |
| 13:47 | <shepazu> | annevk: that's because you're trying to be a good player, not eveyone would be |
| 13:47 | <shepazu> | license are not there for good players, they are there for the exceptions |
| 13:48 | <shepazu> | anyway, thanks for the interesting discussion! I have some work to do |
| 13:51 | <AryehGregor> | darobin, it looks like WebIDL doesn't support multiple inheritance, but the parser does -- is there a reason for that? Did WebIDL used to support multiple inheritance? Can I remove parser support, or do we have to worry that that will break things? |
| 13:51 | <jgraham> | I think WebIDL did used to support it |
| 13:52 | <jgraham> | But I might be wrong |
| 13:52 | <divya> | zcorpan: !!!! you never reply https://github.com/h5bp/html5please/issues/194 |
| 13:52 | <zcorpan> | divya: indeed |
| 13:53 | <divya> | just gimme the words and i will replace zcorpan i do not know enough about it to work on it! |
| 13:54 | <zcorpan> | i'm looking at it now |
| 13:54 | <divya> | thanks zcorpan !! |
| 13:54 | <darobin> | AryehGregor: WebIDL used to support it, I think there's still some IDL out there that uses it |
| 13:54 | <AryehGregor> | darobin, okay, I'll leave support in the parser. |
| 13:55 | <darobin> | part of the problem is whether these old things should be removed, or if we still want to parse old IDLs |
| 13:56 | <AryehGregor> | Inheritance |
| 13:56 | <AryehGregor> | = herit:InheritanceInner? |
| 13:56 | <AryehGregor> | { return herit ? herit : []; } |
| 13:56 | <AryehGregor> | darobin, do you know of a better way in general to do that? |
| 13:56 | <darobin> | AryehGregor: mmmm, not that I can think of off the top of my head |
| 13:56 | <AryehGregor> | It seems like if I try this, I get an error because "herit" is undefined: Inheritance = (":" w herit:ScopedNameList)? { return herit ? herit : []; } |
| 13:56 | <AryehGregor> | Okay. |
| 13:57 | <darobin> | but then again I'm not a PEG expert, I just picked it up because I needed a Q&D WebIDL parser in a hurry for a project |
| 13:57 | <darobin> | hmpf, I'd expect PEG to pass herit = null there... |
| 13:57 | <darobin> | oh, no, I remember now |
| 13:57 | <darobin> | when you have a Foo?, if it doesn't match then the variable is undefined |
| 13:57 | <darobin> | which really sucks |
| 13:58 | <darobin> | you need to test that herit is defined |
| 13:58 | <darobin> | I think that when I get around to this I'll rewrite without PEG — it'll make for something faster, too |
| 14:00 | <AryehGregor> | typeof herit == "undefined" seems to break things in weird ways. |
| 14:05 | <zcorpan> | divya: replied |
| 14:06 | <divya> | thanks zcorpan !! |
| 14:07 | <darobin> | AryehGregor: I'm doing something else at this second, but the way I've tracked these things down before has been to look at the generated code to see why it's weird |
| 14:07 | <AryehGregor> | I looked, and didn't see any mention of the label. |
| 14:07 | <AryehGregor> | In the code. |
| 14:07 | <AryehGregor> | But this way works, so good enough for me. |
| 14:08 | <darobin> | if it works, then great |
| 14:19 | <annevk> | Is it time to admit defeat on initCustomEvent? |
| 14:19 | <annevk> | https://www.w3.org/Bugs/Public/show_bug.cgi?id=15259 |
| 14:27 | <annevk> | so I want to add new Document() |
| 14:27 | <annevk> | but the age-old what argument is it going to take thing pops up |
| 14:28 | <annevk> | I guess I'll just add it without argument and then let people debate that bit :) |
| 14:31 | <zcorpan> | divya: i forgot to mention that web developers of course can opt to not rely on flash or java for any functionality |
| 14:42 | <annevk> | jQuery also calls it timeStamp |
| 14:56 | <jgraham> | I hate mercurial |
| 14:56 | jgraham | again trys to work out how do squash a few changesets together |
| 15:09 | <jgraham> | Oh Sweet Jesus |
| 15:09 | <jgraham> | So I just tried to use hg histedit to do this |
| 15:09 | <jgraham> | and it deleted all the files |
| 15:46 | <marcosc_> | Question: with the img@srcset attribute, it's not clear to me how commas are handled in a candidate image string's path, fragment, and query. Like, for example: <img srcset="generate.php?x=1,w,x23&title=hello, generate.php?y=3,32,23&title=hello x2"> |
| 15:51 | <annevk> | Ms2ger: http://dom.spec.whatwg.org/#dom-document |
| 15:55 | <zewt> | marcosc_: at least one descriptor always exists; the space between the url and the descriptor ends the url |
| 15:56 | <zewt> | see http://www.whatwg.org/specs/web-apps/current-work/#processing-the-image-candidates for particulars |
| 15:57 | <tantek> | jgraham does hg have a "-undo" option? |
| 15:57 | <marcosc_> | zewt: thanks |
| 15:57 | <jgraham> | tantek: It turned out that the files were still in history |
| 15:57 | <jgraham> | But it had made a commit that deleted them |
| 15:58 | <zewt> | heh, when I'm using git I make wholesale cp -a backups a lot |
| 15:58 | <jgraham> | and then refused to do anything else, possibly because I was in a path that no longer existed |
| 15:58 | <jgraham> | In the end I managed to invoke some mq magic to do what I needed |
| 15:59 | <jgraham> | But it realy does seem like everything slightly advanced is a minefield in mercurial |
| 16:22 | <marcosc_> | annevk, are you still maintaining: http://dvcs.w3.org/hg/domcore/raw-file/tip/Overview.html ? |
| 16:22 | <Ms2ger> | marcosc_, no, we edit dom.spec.whatwg.org now |
| 16:22 | <marcosc_> | Ms2ger: ok, cool. Maybe add a note to the W3C one :) |
| 16:26 | <annevk> | marcosc_: https://twitter.com/annevk/status/250503261138849792 |
| 16:27 | <annevk> | (took me longer to find that tweet to not repeat myself than it would be to just answer...) |
| 16:27 | <annevk> | marcosc_: yeah, Ms2ger, if you can add that note; I guess you're still allowed |
| 16:28 | <annevk> | marcosc_: fwiw, https://dvcs.w3.org/hg/url/raw-file/tip/Overview.html |
| 16:28 | <annevk> | marcosc_: has happened before |
| 16:28 | <marcosc_> | exactly like that would be awesome |
| 16:28 | <annevk> | it's not gonna be me |
| 16:28 | <Ms2ger> | <title class="now3c">Moved</title> |
| 16:56 | <AryehGregor> | annevk, how do you test what the name of an encoding should be? I get "UTF-8" for document.characterSet in Gecko/WebKit, and seemingly "utf-8" in Opera (except in a different test it returns the empty string), and "unicode" in IE. |
| 16:56 | <AryehGregor> | Spec seems to suggest "utf-8"; "UTF-8" seems like it makes more sense. |
| 16:57 | <AryehGregor> | (if that's the data source you're using) |
| 16:57 | <annevk> | AryehGregor: the spec does indeed suggest all lowercase |
| 16:57 | <AryehGregor> | Any reason for that, given existing browsers' behavior? |
| 16:57 | <annevk> | AryehGregor: hope we could make that work? |
| 16:58 | <AryehGregor> | Why is it worth it? Just go with uppercase. |
| 16:58 | <annevk> | AryehGregor: string identifiers are always lowercase |
| 16:58 | <AryehGregor> | Well, it's an acronym. |
| 16:58 | <AryehGregor> | Also: tagName. |
| 16:58 | <annevk> | AryehGregor: well not all is uppercase I think; e.g. windows-1252 would be Windows-1252 or some such |
| 16:58 | <AryehGregor> | Probably. |
| 16:58 | <annevk> | AryehGregor: tagName is obsoleted with localName |
| 16:59 | <annevk> | anyway, I don't feel that strongly and am taking patches or suggested renaming if you think that'll be better |
| 16:59 | <annevk> | could you please file a bug? |
| 16:59 | <AryehGregor> | I just don't want to write tests for a spec that needlessly doesn't match browsers. I'm not testing the whole spec, but I want some test or other for .characterSet in some DOM tests I'm writing right now. |
| 16:59 | <AryehGregor> | Sure. |
| 16:59 | <AryehGregor> | Where? I don't see an Encoding component in the W3C Bugzilla. |
| 17:00 | <annevk> | is it not under WHATWG? |
| 17:00 | <AryehGregor> | Oh, I see. |
| 17:00 | <annevk> | it's linked from the spec too I think |
| 17:00 | <annevk> | you're reading http://encoding.spec.whatwg.org/ right? |
| 17:00 | <annevk> | gotta go |
| 17:00 | <AryehGregor> | Yeah. |
| 17:10 | <annevk> | AryehGregor: ah, so you're not gonna test all of them? |
| 17:12 | <annevk> | TabAtkins: your twitter handle makes me laugh every time |
| 17:22 | <arv> | annevk: I'm implementing MutationObservers based on the spec and generally everything works and matches Gecko and WebKit. However, there is one issue. When appending a record to the queue both Gecko and WebKit replaces the last record if it represents the same mutation and the new record has an oldValue. |
| 17:23 | <annevk> | how does that happen? |
| 17:23 | <annevk> | also, a bug + changes requested would be a big help, not gonna be able to do much more today |
| 17:23 | <arv> | annevk: OK. I'll file a bug with a test case |
| 17:24 | <annevk> | thanks arv! |
| 17:27 | <AryehGregor> | annevk, I'd be happy to test all of them. I can write a little harness that uses a simple data structure for all the encoding maps that you can keep updated along with the spec. |
| 17:27 | <AryehGregor> | It's just not what I was testing just now. |
| 17:28 | <AryehGregor> | (I was doing simple tests on properties of Node and its various derived interfaces.) |
| 17:29 | <annevk> | that sounds sweet |
| 17:30 | <annevk> | I saw hsivonen might do some work on encodings too for Gecko |
| 17:31 | <annevk> | I'm happy to help out too btw, getting browsers aligned on byte -> code point mapping would be a great improvement |
| 17:38 | Ms2ger | tries to remember what he was trying to fix about NoInterfaceObject |
| 17:55 | <Ms2ger> | Huh |
| 17:55 | <Ms2ger> | AryehGregor, yt? |
| 17:55 | <AryehGregor> | Ms2ger, yep. |
| 17:55 | <AryehGregor> | (probably not for too much longer, though) |
| 17:56 | <Ms2ger> | I'm looking at IdlInterface.prototype.test_self |
| 17:56 | <AryehGregor> | Okay. |
| 17:57 | <Ms2ger> | Oh |
| 17:57 | <Ms2ger> | I thought you were testing Object.getOwnPropertyDescriptor(window, this.name) twice |
| 17:58 | <Ms2ger> | But the other one was for Object.getOwnPropertyDescriptor(window[this.name], "prototype") |
| 17:59 | <Ms2ger> | Might be nice if we could make those asserts a little less verbose |
| 17:59 | <AryehGregor> | Like a wrapper of some type, yeah. |
| 18:02 | <AryehGregor> | It could also do more rigorous tests, like actually trying to write and seeing if it succeeds. |
| 18:06 | <Ms2ger> | Yeah, but that'll need more code to (try to) revert the changes |
| 21:31 | <annevk> | karlcow: I would have expected you to describe http://lists.w3.org/Archives/Public/www-tag/2012Oct/0063 as beautiful, not ironic, since you like crazy stuff |
| 21:33 | <karlcow> | what is amazing in that email is the misguided labeling and rudeness. Like the person being ballistic on me on really the wrong person. |
| 21:33 | <karlcow> | I don't get it |
| 21:33 | <annevk> | going ballistic on people is generally best avoided |
| 21:35 | <annevk> | karlcow: I also love how he transforms <i> to <em> and apparently considers himself an expert :-) |
| 21:35 | <karlcow> | double irony when I have managed my own entire Web site for a few years with XHTML 1.0 + XSLT |
| 21:36 | <Hixie> | wow, why did i follow that link |
| 21:36 | <annevk> | Hixie: dude www-tag should give you a hint |
| 21:36 | <Hixie> | i could have gone my day happily ignorant that www-tag is still discussing stuff that we resolved ten years ago |
| 21:38 | <karlcow> | "working at w3c I was an XML evil lover, now working at Opera I'm an XML hater". Some humans make me sad. Boxes, labels, and stuff. Ah yes I should add my RDF label for some others too. |
| 21:40 | <annevk> | @opera.com -> must be a WHATWG XML-hating fanboy is indeed kinda nonsense, but you do hang out here ;-) |
| 21:40 | <Hixie> | yeah i love the assumption that whatwg = xml-hating |
| 21:40 | <Hixie> | seems to rather miss the point |
| 21:40 | <annevk> | the other thing with that kind of stuff is that often the people here know much more about XML and how it works for web sites / browsers than any of those complaining |
| 21:40 | <karlcow> | annevk: I hang out on #swig too, aka Semantic Web Interest Group |
| 21:41 | <Hixie> | i take it this latest spat started because of the url spec work? |
| 21:41 | <karlcow> | I'm the evil evil evil, all sins in one cow ;) |
| 21:41 | <annevk> | Hixie: maybe, I saw a few tweets, but www-tag thinks URL spec work is inside the HTML spec... |
| 21:41 | <Hixie> | o_O |
| 21:41 | <Hixie> | if they wanted to do the url work, they had plenty of time |
| 21:42 | <Hixie> | it's not like either you or me actually wants to do it |
| 21:42 | <annevk> | in fact they can still do it if they want to :-) |
| 21:42 | <annevk> | I don't really care, I just wanna fix problems |
| 21:42 | <Hixie> | hear hear |
| 21:43 | <zewt> | i hate xml, but i do have my reasons :) |
| 21:43 | <zewt> | does this mean i can get a job at opera? |
| 21:43 | <karlcow> | ahaha |
| 21:43 | <zewt> | i guess the association is one-way :( |
| 21:43 | <annevk> | zewt: if you're serious I think you'd have a good chance of getting a job, but they might ask you to move |
| 21:44 | <karlcow> | zewt: I do not think it is a requirement, my cover letter/resume was xhtml 1.1 served as application/xhtml+xml ;) |
| 21:44 | <zewt> | heh, i moved for my current job, i'll give it a chance first :P |
| 21:44 | <karlcow> | zewt: http://www.opera.com/company/jobs/ |
| 21:44 | <annevk> | zewt: anyone with above average standards knowledge is a serious candidate for sure |
| 21:45 | <zewt> | "above average" sure sounds like "any whatsoever" |
| 21:45 | <annevk> | after I hit enter I realised that was kinda vague |
| 21:45 | <annevk> | but I'm pretty sure you fit the criteria |
| 21:45 | <zewt> | which after years of most specs being IETF-style walls of descriptive text, is a bit understandable |
| 21:47 | <zewt> | man, google docs spreadsheets is nearly unusable in FF |
| 21:48 | <annevk> | What I meant to say is people knowledgable enough about complex algorithms in standards and can comment on them to the extent they need to be changed are sure employee candidates for browser vendors |
| 21:50 | <zewt> | personally I see "complex algorithms in standards" and "complex algorithms in code" as the same thing |
| 21:50 | <TabAtkins> | they are |
| 21:50 | <annevk> | Sure, but you need to have an interest in standards to comment on the former, which is kinda important |
| 21:52 | <zewt> | also in "complex algorithms" (but in my opinion, people who can't do that picked the wrong field) |
| 21:54 | <annevk> | heh yeah |
| 21:54 | <annevk> | it's just that complex algorithms is much broader; getting complex algorithms to work the same in n implementations is somewhat different problem |
| 21:54 | <annevk> | and not for everyone |
| 21:56 | <zewt> | there's also sort of a "web mindset" that you have to be able to see, i suppose in the same way that some people understand the security mindset and some don't |
| 21:56 | <annevk> | heh, I love and TabAtkins and Hixie speculate and ask respectively about upcoming Google products |
| 21:56 | <annevk> | s/and/how/ |
| 21:57 | <TabAtkins> | Hahaha, yes. |
| 21:57 | <annevk> | reading http://www.w3.org/TR/html-design-principles/ without saying WTWTF all the time also helps |
| 21:57 | <Hixie> | I was asking about _current_ displays, not speculation :-) |
| 22:01 | <tantek> | Hixie, current displays, you mean something like http://thereisnofold.com/ ? |
| 22:02 | <Hixie> | I mean whatever Mark was referring to in http://lists.w3.org/Archives/Public/public-whatwg-archive/2012Oct/0073.html |
| 22:02 | <tantek> | btw: re: XSLT - I tend to avoid using/creating/modifying it, but it (combined with a Tidy pre-processing step) has been quite good at keeping the production-dependent h2vx.com service going (based on X2V) that people use to convert hCards to vCards and hCalendar to iCalendar. |
| 22:03 | <othermaciej> | what about current displays? |
| 22:03 | <tantek> | physical pixel density, sure |
| 22:03 | <tantek> | 3x density displays, sure is faster than printing things to test stuff ;) |
| 22:04 | <othermaciej> | there are definitely devices with more than 3x96 dpi |
| 22:04 | <othermaciej> | the iPhone 5 is one |
| 22:04 | <othermaciej> | but I don't think there are any that out of the box run with 3x scaling of the UI |
| 22:04 | <tantek> | Hixie, best part of that email, the footer "This electronic mail message may contain confidential and privileged information from HI Corporation. If you are not the intended recipient, any disclosure, photocopying, distribution or use of the contents of the received information is prohibited." - yeah, on an archive, nice. |
| 22:04 | <Hixie> | that's what matters |
| 22:05 | <othermaciej> | (and css pixel scaling is generally based on ui scaling, not literally dividing the physical DPI by 96, which is why use of CSS DPI is generally more confusing than informative) |
| 22:06 | <Hixie> | yeah i try to use "device pixels per CSS pixels" |
| 22:06 | <annevk> | isn't CSS dpi these days dependent on how you define CSS px? |
| 22:07 | <othermaciej> | CSS dpi is 96 by definition |
| 22:07 | <Hixie> | i don't know what "CSS dpi" is |
| 22:07 | <othermaciej> | wait, it's 96 times the scale factor |
| 22:07 | <othermaciej> | I always confuse myself |
| 22:07 | <othermaciej> | CSS has a "dpi" unit |
| 22:07 | <Hixie> | oh, that |
| 22:07 | <Hixie> | who uses that |
| 22:07 | <othermaciej> | it's 96 times dppx which is "device pixels per css pixel" |
| 22:08 | <Hixie> | "device pixels per CSS pixel" is what matters |
| 22:08 | <othermaciej> | other than the css wg? probably no one |
| 22:08 | <Hixie> | and so far, excluding zoom, it seems to be either 1 or 2. |
| 22:08 | <annevk> | what's the opposite of <3? !<3? that's kinda how I feel about dppx |
| 22:09 | <Hixie> | </3 |
| 22:09 | <othermaciej> | it's certainly a hideous unit name |
| 22:09 | <TabAtkins> | That's why Images 4 has the 'x' unit which means the same thing. ^_^ |
| 22:10 | <fantasai> | othermaciej: It went through LC twice. Why u not complain earlier? :) |
| 22:10 | <annevk> | you and your aliases |
| 22:10 | <othermaciej> | fantasai: I have complained many times, and I don't need to have another conversation with the hand |
| 22:11 | Hixie | wonders what going through LC has to do with anything :-P |
| 22:11 | <fantasai> | we asked for comments |
| 22:11 | <fantasai> | nobody sent any on the name of the unit |
| 22:11 | <Hixie> | do you not ask for comments at other times? :-) |
| 22:11 | <annevk> | Hixie: it's a thing the W3C has where they think stuff is stable but it really isn't |
| 22:11 | <mkanat> | "Mr Dent, you realize the plans have been available...." |
| 22:11 | <Hixie> | annevk: how quaint |
| 22:12 | <fantasai> | Hixie: sure, we asked for comments before that too |
| 22:12 | <fantasai> | Hixie: but nobody sent any |
| 22:12 | <Hixie> | but not after? :-) |
| 22:12 | <othermaciej> | here is one example of someone complaining about the 'dppx' unit name: http://lists.w3.org/Archives/Public/www-style/2012Apr/0685.html |
| 22:12 | <othermaciej> | (w/ more on that same thread IIRC) |
| 22:12 | <annevk> | I like your use of "someone" |
| 22:13 | <Hixie> | here's another: http://lists.w3.org/Archives/Public/www-style/2012May/0915.html |
| 22:13 | <othermaciej> | I don't know if that was before or after Last Call but I did not get the impression that the CSS WG wanted me to make the same comment over and over |
| 22:13 | <Hixie> | starts with "There's been a decent bit of feedback recently about the 'dppx'" and is written by a WG member |
| 22:13 | <fantasai> | othermaciej: that was after |
| 22:14 | <annevk> | one of my problems with the CSS WG is that they adhere to da Process so much they get stuck up in it |
| 22:14 | <annevk> | e.g. one of those emails cites Process concerns |
| 22:14 | <annevk> | as if they fricking matter |
| 22:14 | <Hixie> | http://lists.w3.org/Archives/Public/www-style/2011May/0560.html "Plus, 'dppx' is a hard unit to pronounce", May 2011 |
| 22:15 | <Hixie> | presumably may 2011 is before the point at which you stopped asking for comments |
| 22:15 | <othermaciej> | both those specific comments were followed up by a bunch of discussion |
| 22:15 | <annevk> | what matters is a) has it shipped and b) does content depend on it |
| 22:15 | <TabAtkins> | Ignoring all of this, I've addressed it in the newest draft. |
| 22:15 | <Hixie> | annevk++ |
| 22:15 | <othermaciej> | so I think if the CSS WG chose not to take action on them, that is due to the CSS WG, not because "nobody sent any" |
| 22:17 | <Hixie> | (btw, this kind of thing is one reason why i guarantee that i'll respond to all substantial feedback on the whatwg list: so i can't miss any) |
| 22:17 | <fantasai> | Hixie: It is, and I missed that as being something to act on |
| 22:17 | <othermaciej> | TabAtkins: don't see anything in <http://dev.w3.org/csswg/css3-values/#resolution>; (though I do see an issue marker for a context-specific alias of 'x' in <http://dev.w3.org/csswg/css4-images/>;) |
| 22:18 | <TabAtkins> | othermaciej: That's what I meant by "latest draft". |
| 22:18 | <fantasai> | Hixie: I took that comment as "should this unit exist", and the answer was "yes, it's useful to have that functionality" |
| 22:18 | <TabAtkins> | I haven't worked on Images 4 enough to push that out of an issue and into normative text. |
| 22:18 | <Hixie> | fantasai: anyway, has it shipped and been used by anyone yet? |
| 22:19 | <othermaciej> | TabAtkins: that does save you from having to use 'dppx' in one context, though not in the 'resolution:' media query which was my original complaint as linked above |
| 22:19 | <Hixie> | fantasai: if not, then it's not too late to fix |
| 22:19 | <fantasai> | Hixie: Given the commenter there is an editor of the relevant drafts, though, and is able to raise issues and resolve issues on his own, I guess it wasn't seen as important enough to do something about at the time |
| 22:19 | <othermaciej> | technically a WG is required to give a substantive response to all comments at every point from LC onward |
| 22:19 | <fantasai> | Hixie: It's been shipped, I believe |
| 22:19 | <fantasai> | Hixie: Not sure about whether it's in release builds yet |
| 22:20 | <TabAtkins> | othermaciej: Once we get 'x' as a unit, it'll be usable in the 'resolution' MQ too. |
| 22:20 | <othermaciej> | not clear to me if the CSS WG has given a substantive reply to my comment on 'dppx' |
| 22:20 | <Hixie> | fantasai: so no content is relying on it? |
| 22:21 | <othermaciej> | TabAtkins: is the plan to add it to units & values rather than to images? |
| 22:21 | <othermaciej> | TabAtkins: if so, the issue marker is in the wrong place, no? |
| 22:21 | <tantek> | http://memegenerator.net/instance/28084024 |
| 22:21 | <fantasai> | Hixie: Probably not much |
| 22:21 | <TabAtkins> | othermaciej: It's there because the relevant thing *introducing* the unit is image-set(), and specifically WebKit's implementation fo it. |
| 22:22 | <Hixie> | fantasai: so then just fix it already :-P |
| 22:22 | <othermaciej> | TabAtkins: is it actually the plan to add it to Units & Values though? (I care more about the actual result than the process niceties here) |
| 22:22 | <TabAtkins> | Yeah, once we agree to *have* it, it'll move to V&U. |
| 22:23 | <Hixie> | tantek: http://imgflip.com/i/4e3k |
| 22:25 | <Hixie> | tantek: http://imgflip.com/i/4e3o |
| 22:25 | <hober> | some more comments in the general area, http://lists.w3.org/Archives/Public/www-style/2012May/0409.html and http://lists.w3.org/Archives/Public/www-style/2012Apr/0674.html |
| 22:26 | <fantasai> | Hixie: I can't make substantive changes to a spec that's in CR without CSSWG approval, especially for something that's implemented and will require a change to those implementations. We can raise it as an issue, but it doesn't mean the WG will agree to change. Seems unlikely given it's "bikeshedding". |
| 22:26 | <Hixie> | hahaha |
| 22:26 | <fantasai> | Hixie: But Tab's super excited about the 'x' unit, so maybe he can convince them :) |
| 22:26 | <Hixie> | fantasai: http://imgflip.com/i/4e3o |
| 22:27 | <Hixie> | i was going to add a second line to this one, but really it stands alone even with just the top line: http://imgflip.com/i/4e3w |
| 22:27 | <TabAtkins> | Really, we can only do that if we're willing to publish the draft outside of CSSWG space. |
| 22:27 | <zewt> | gah, why is the bug tracker sending notification emails in a fixed-width font; that's just wrong |
| 22:27 | <TabAtkins> | Which, sure, possible, but annoying. |
| 22:27 | <hober> | tantek Hixie: those are all good w3cmemes candidates... |
| 22:28 | <Hixie> | TabAtkins: if the only way you can fix a spec is by not publishing it through the csswg, one wonders why one would ever consider publishing it through the csswg in the first place |
| 22:29 | <TabAtkins> | Because it's rarely a problem, and useful in many other ways? |
| 22:31 | <Hixie> | hober: done :-P |
| 22:31 | <Hixie> | though i was enjoying the previous one |
| 22:31 | <Hixie> | (http://w3cmemes.tumblr.com/image/32354094056) |
| 22:33 | <Hixie> | hober: (we have a dupe! cleanup on aisle 5!) |
| 22:34 | <tantek> | Hixie, the multi-attachement posts break the flow of tumblr |
| 22:34 | <tantek> | attachment event |
| 22:34 | <tantek> | even* |
| 22:34 | <Hixie> | actually i thought it worked kinda well, since they were related |
| 22:35 | <Hixie> | though the cropping on insanity wolf was unfortunate |
| 22:35 | <tantek> | the linear series works well |
| 22:35 | <tantek> | where you can see the back/forth in the flow :) |
| 22:35 | <tantek> | presumably someone here has http://whatwgmemes.tumblr.com/ ? |
| 22:35 | <Hixie> | there's very little "back and forth" on w3cmemes :-P |
| 22:35 | <tantek> | Hixie, the The Office one belongs there |
| 22:36 | <Hixie> | it's mostly just f"orth" :-P |
| 22:36 | <tantek> | need more back forth on w3cmemes ;) |
| 22:36 | <Hixie> | er |
| 22:36 | <Hixie> | "forth" |
| 22:36 | <tantek> | mostly it's back |
| 22:36 | <tantek> | if it was forth, the tumblr wouldn't exist |
| 22:36 | <Hixie> | whichever is "bitching" :-P |
| 22:37 | <Hixie> | seriously though, i am finding it unbelieveable that even now the csswg can't fix a mistake that hasn't shipped yet, for purely process reasons |
| 22:37 | <Hixie> | that's such BS |
| 22:37 | <Hixie> | this is the kind of crap that the AB never believes me about |
| 22:37 | <tantek> | I thought it had shipped (implementations) ? |
| 22:37 | <tantek> | or is this a different kind of "shipped"? |
| 22:38 | <tantek> | even if it has "shipped", we should still be able to fix things that are mere aliasing like that. that's what deprecation is for ;) |
| 22:38 | <Hixie> | unless there's content depending on dppx, it can just be dropped or renamed |
| 22:45 | <smaug____> | hmm, what has w3 done to their bugzilla |
| 22:45 | <smaug____> | bugmail looks odd |
| 22:46 | <smaug____> | odd == unreadable |
| 22:46 | <smaug____> | shepazu: do you happen to know |
| 22:46 | <Hixie> | they just did an upgrade |
| 22:46 | <Hixie> | but i don't see any differences with my textmode client... |
| 22:47 | <smaug____> | I sure don't get that kind of bugmail from Moz bugzilla |
| 22:47 | <smaug____> | and I think Mozilla is using the latest bugzilla release |
| 22:47 | <Hixie> | oh it looks like they turned on html mail? |
| 22:47 | <smaug____> | I guess so |
| 22:47 | <smaug____> | really annoying |
| 22:54 | <Hixie> | damnit, bugzilla changed in a way that broke my scripts |
| 22:55 | <TabAtkins> | Hixie: It has shipped, though content dependent on it is likely verly low. |
| 22:55 | <tantek> | scrapers gonna scrape |
| 22:57 | <Hixie> | TabAtkins: that isn't what you and fantasai said before, before you just said the process wouldn't let you fix it in the wg. |
| 22:57 | <Hixie> | ew, the summaries are links now |
| 22:57 | <Hixie> | in search results |
| 22:57 | <Hixie> | way to make things unreadable |
| 22:58 | <Hixie> | ROFL |
| 22:58 | <Hixie> | merge conflict on the search page |
| 22:58 | <shepazu> | smaug____: ugh |
| 22:58 | <shepazu> | I'll ask them to change that... |
| 22:59 | <Hixie> | shepazu: while you're at it, the search page (advanced) has a visible merge conflict |
| 22:59 | <shepazu> | Hixie: thanks, I'll relay that |
| 22:59 | <Hixie> | thanks |
| 23:00 | <shepazu> | hope it'll be fixed soon |
| 23:00 | <shepazu> | sorry for the hassle |
| 23:00 | <Hixie> | no hassle, search works fine |
| 23:00 | <Hixie> | just looks ugly :-) |
| 23:02 | <hober> | Hixie tantek: deduped |
| 23:13 | <hober> | shepazu: where is the bugzilla preference to switch from html email to plain text? |
| 23:16 | <shepazu> | hober: excellent question |
| 23:16 | <shepazu> | ask smaug____ |
| 23:18 | <smaug____> | hober: ah |
| 23:18 | <smaug____> | hmm |
| 23:18 | <smaug____> | but w3 did change something very recently |
| 23:18 | <smaug____> | and they have some default no other bugzilla has |
| 23:19 | <hober> | shepazu: while i have your attention, did you see my request in #webplatform? |
| 23:19 | <smaug____> | at least not bmo nor bwo |
| 23:19 | <shepazu> | hober: I missed it |
| 23:19 | <shepazu> | what is it? |
| 23:22 | <hober> | shepazu: could you remove the <a> element around Apple in the list on webplatform.org/stewards/? Alternately, it could link to apple.com. |
| 23:24 | <shepazu> | hober: I can add a link... |
| 23:24 | <shepazu> | I need the <a> for the CSS |
| 23:26 | <shepazu> | smaug____, hober, Hixie, sent an email to sysreq |
| 23:32 | <smaug____> | shepazu: thanks |
| 23:34 | <othermaciej> | reading w3cmemes makes me alternately lol and cringe in shame |
| 23:35 | <cabanier> | smfr: I have a patch for CG blending: https://bugs.webkit.org/show_bug.cgi?id=98450 |
| 23:35 | <tantek> | "I need the <a> for the CSS" - LOL. And this is why stubbornella advocates styling classes, not elements. |
| 23:36 | <cabanier> | smfr: my problem yesterday was a silly typo |
| 23:36 | <tantek> | othermaciej, the joker one is gold |
| 23:43 | <othermaciej> | this one is all "defensive, much?" http://w3cmemes.tumblr.com/post/29103302932 |
| 23:43 | <othermaciej> | (in response to http://w3cmemes.tumblr.com/post/29085196102/disaster-girl-has-an-interesting-take-on-the ) |
| 23:43 | <hober> | heh |
| 23:44 | <hober> | yes, even the shady cabal behind w3cmemes have differences of opinion from time to time :) |