| 00:14 | <Hixie> | i think #whatwg hit peak sarcasm today |
| 00:14 | <Hixie> | :-P |
| 00:20 | <xtrm0> | ahah Hixie |
| 00:39 | <cwilso> | Hixie: don't sell the channel short. They can do better. :P |
| 00:53 | <JonathanNeal> | jorendorff: you around? |
| 00:53 | <jorendorff> | JonathanNeal: Not really! I'm supposed to have gone home by now. What's up? |
| 00:54 | <JonathanNeal> | I’m just lingering on that toLength bug in Firefox, and I’m wondering if the spec is right to suggest that Infinity should be capped to the max instead of, say, throwing an error. |
| 00:55 | <JonathanNeal> | I can’t think of a scenario when you would intent Infinity to mean something that isn’t Infinity. Do you know why the spec wouldn’t suggest something similar? |
| 01:31 | <JonathanNeal> | jorendorff: ^ |
| 09:44 | <caitp-> | Today HTML5 is turned into a standard so we can safely say the O.O() will be widely adopted, if it isn't already everywhere. |
| 10:26 | <annevk> | caitp-: is that a quote? |
| 10:26 | <caitp-> | yes |
| 10:41 | <foolip> | caitp-: what's O.O()? |
| 10:41 | <annevk> | Object.observe() presumably |
| 10:42 | <foolip> | um, am I right to assume that's part of ES6, not HTML? |
| 10:42 | <annevk> | foolip: yup |
| 10:42 | <annevk> | foolip: ES7 actually |
| 10:42 | <foolip> | well, great quote :) |
| 10:42 | <annevk> | if all goes well |
| 10:44 | <caitp-> | I thought you might find the quote entertaining for a number of reasons |
| 11:40 | <annevk> | jgraham: http://antirez.com/news/82 |
| 12:58 | <JonathanNeal> | Is # the official or unofficial symbol for .prototype. ? As in Array#every or Element#matches ? |
| 13:00 | <ondras> | it is used in this sense in jsdoc |
| 14:02 | <jgraham> | annevk: Yes, exactly that. |
| 16:26 | <jorendorff> | JonathanNeal: no idea |
| 16:37 | <gsnedders> | what actually changed from the original WebKit AAA? |
| 17:16 | <Domenic> | annevk: https://dom.spec.whatwg.org/#interface-nodefilter is a callback interface but it has static properties (constants) which is not allowed according to https://heycam.github.io/webidl/#dfn-callback-interface |
| 17:17 | <Domenic> | annevk: it's unclear to me what the global window.NodeFilter is supposed to be/do |
| 17:18 | <Domenic> | annevk: I think maybe NodeFilter is an interface but there is also a callback interface containing acceptNode only? |
| 17:20 | <Domenic> | annevk: I will file a bug |
| 17:42 | <SteveF__> | Domenic: have forwarded your email to list - thanks, I also talked to Rich about getting the funky list restrictions lifted |
| 17:43 | <Domenic> | SteveF__: awesome :) |
| 18:01 | <Domenic> | Ms2ger: I am trying to update my web platform tests usage but add_result_callback doesn't exist anymore? |
| 18:01 | <Ms2ger> | Domenic, looking |
| 18:03 | <Domenic> | hmm testharness.js hasn't been updated in a long time so seems likely the problem is on my end |
| 18:03 | <Ms2ger> | Or I will when my editor wakes up |
| 18:03 | <Domenic> | oh nevermind that's jgraham's fork that's old |
| 18:04 | <jgraham> | Yeah, don't look at that :) |
| 18:04 | <Ms2ger> | add_result_callback does seem to still be around |
| 18:05 | <Domenic> | yeah, maybe loading testharness.js is failing... probably my problem |
| 18:05 | <Ms2ger> | That could well be it |
| 18:06 | <Domenic> | oh yep i forgot to recursively update my submodules gooood times. |
| 18:06 | <jgraham> | Yeah, not sure what to do about that |
| 18:07 | <jgraham> | Maybe a script to bootstrap the checkout |
| 18:08 | <Ms2ger> | serve.py checking if you're on the wrong commit? |
| 18:08 | <Ms2ger> | That'd probably be painful, though |
| 18:11 | <Domenic> | it's mostly my fault because jsdom doesn't give good error messages for failing to load a script |
| 18:11 | <Domenic> | testharness.js just plain wasn't present on the filesystem |
| 18:50 | <annevk> | Domenic: are you more motivated to fix IDL now? |
| 18:50 | <Domenic> | annevk: lol, a bit |
| 18:51 | <annevk> | I forgot about that constants hack, that is quite ugly |
| 19:04 | <annevk> | http://www.w3.org/2014/10/28-webapps-minutes.html#item07 (discussion N about URLs, seemed rubys tried to keep it real) |
| 19:04 | <annevk> | One interesting tidbit there Domenic was that the TAG endorses "openstand principles" |
| 19:05 | <Domenic> | annevk: yeah... well, timbl does |
| 19:05 | <annevk> | Domenic: I see, except when it comes to longdesc or some such :p |
| 19:06 | <annevk> | You see, if the W3C actually followed those principles, the call the chairs of the HTML WG made with W3C management with regards to the 25 formal objections would not have been secret |
| 19:08 | <Domenic> | annevk: I don't understand the setter part of https://dom.spec.whatwg.org/#dom-attr-value |
| 19:09 | <annevk> | Domenic: what's unclear? |
| 19:09 | <Domenic> | annevk: what does step 2 accomplish? |
| 19:10 | <annevk> | Domenic: mutation observers |
| 19:10 | <annevk> | Domenic: if you follow the "change" link... |
| 19:10 | <Domenic> | oh i didn't realize "change" was a special algorithm, now it all makes sense |
| 19:10 | <Domenic> | I thought it was just trying to reassign "the context object" to a string |
| 19:10 | <annevk> | Hah, there's no cheating, follow all the links :p |
| 19:11 | <Domenic> | meh, see, ES specs are nice and algorithmic; it would be ChangeAttributeValue(this, newValue, oldValue) :P |
| 19:12 | <annevk> | I would be happy to adopt that style some day, once IDL is updated and defines it |
| 19:12 | <Domenic> | yesss jsdom's Attr might be spec compliant |
| 19:12 | <annevk> | I think for now it's too much work to introduce a new style and keep everything up-to-date |
| 19:14 | <Domenic> | yeah |
| 19:14 | <Domenic> | best for greenfield |
| 19:21 | <Domenic> | Augh converting a DOM implementation that assumes Attr : Node to one that has standalone Attr objects is hard :( |
| 19:31 | <Ms2ger> | Domenic, quite |
| 19:33 | <annevk> | Domenic: note that it's still unclear whether that will stick :( |
| 19:59 | <annevk> | Hmm, why did reddit.com add target=_blank |
| 19:59 | <Domenic> | Wait, yeah, no browsers have tried this yet |
| 19:59 | <Domenic> | Why am I making jsdom the guinea pig |
| 20:01 | annevk | writes @WHATWG tweets instead of figuring out what is wrong with the script... |
| 20:01 | <Domenic> | annevk: yeah I am scared about pages trying to insert text nodes into Attrs |
| 20:01 | <annevk> | Domenic: that part Mozilla disabled successfully |
| 20:01 | <annevk> | Domenic: Attr's never have children in Firefox |
| 20:02 | <Domenic> | nice |
| 20:02 | <Domenic> | our main problem in jsdom is we're carrying around a large currently-passing test suite from the original DOM1 + DOM2 tests |
| 20:02 | <Domenic> | so every time we make a change like this it means going back and modifying or removing a bunch of tests |
| 20:02 | <Domenic> | the alternative is to greenfield everything so we can pass the new web-platform-tests |
| 20:02 | <Domenic> | my gut says that will be harder |
| 20:03 | <Domenic> | but it's very tempting at times |
| 20:03 | <Ms2ger> | Domenic, fwiw, I've been working on submitting equivalent tests for those to wpt |
| 20:04 | <smola> | annevk: is there any chance that the normalize-to-percent-encoding-to-uppercase thing is included in the standard? |
| 20:04 | <Domenic> | Ms2ger: that would be lovely. Spec-compliant features only? |
| 20:05 | <Ms2ger> | Domenic, well, all the stuff that's been thrown out is in dom/historical.html in wpt :) |
| 20:50 | <klysium> | TabAtkins: congrats on HTML5 :D |
| 20:51 | <TabAtkins> | klysium: Thanks, it was a great achievement for me personally. |
| 20:53 | <caitp-> | is there like a nice, up-to-date fork of blink somewhere that renders nice blame views and history views? that would be nice |
| 20:54 | <caitp-> | s/fork/mirror |
| 20:57 | <ondras> | http://img.pr0gramm.com/2014/10/29/974763bdab985db9.jpg |
| 20:57 | <ondras> | oops, bad window |
| 20:57 | <ondras> | sorry. |
| 21:08 | <ebureau> | lol. nice |
| 21:14 | <annevk> | smola: perhaps, if we can get other implementations to agree |
| 21:14 | <annevk> | smola: Chrome does some normalization around percent-encoding |
| 21:14 | <annevk> | smola: also other things such as "normalizing" %61 and such |
| 21:29 | <annevk> | smola: best way to go about it is to file bug reports |
| 21:30 | <annevk> | smola: it may be however that changing casing is not an option and would break some legacy content... |
| 21:30 | <Domenic> | rubys: if you happen to watch logs, is your url parser available as an npm package? if so I may want to use it in jsdom |
| 21:31 | <Domenic> | OK time to write a Streams blog post |
| 21:55 | <annevk> | foolip: suddenly a surge in <img composite> usage |
| 21:56 | <JonathanNeal> | Domenic: thoughts on toLength turning infinity into a number? Should it throw instead? |
| 22:02 | <Domenic> | JonathanNeal: it should not throw; JS doesn't generally throw for that type of thing |
| 22:03 | <JonathanNeal> | How about a resolve to zero? In the off chance that someone passes in an object with a length of Infinity, it takes a real beating on one's system. |
| 22:04 | <JonathanNeal> | And Infinity becomes a rather hacky way of writing 9007199254740991 |
| 22:05 | <JonathanNeal> | I mean 9007199254740992 |
| 22:08 | <Domenic> | Rounding to the closest allowable length seems fine |
| 22:08 | <Domenic> | annevk: OK I wrote up a blog post, how do I post it? |
| 22:09 | <Domenic> | or Hixie |
| 22:10 | <Domenic> | TabAtkins: looks like no ability to omit editors in Bikeshed still? |
| 22:12 | <TabAtkins> | Not yet. File an issue so I won't forget? |
| 22:12 | <TabAtkins> | I can do it tonight. |
| 22:12 | <Domenic> | I wonder if actually I just want to hide the editor with CSS? |
| 22:13 | <Domenic> | the metadata seems OK |
| 22:41 | <JonathanNeal> | My patience for this flash video player constantly stalling w3’s 20th anniversary presentation has really got me blue. |
| 22:42 | <JonathanNeal> | Somewhat because it’s Flash, and mostly because I can watch it for about 10 minutes before I have to miss 2 minutes waiting for the thing to work again. |
| 22:55 | <smola> | annevk: ok, thank you |
