| 01:47 | <MikeSmith> | :%s/\(^.*<h[1-6].\+\)\@<!\(\([AB]\.\)\?\([1-9][0-9]\?\.\)\+[1-9][0-9]\?\)\([^0-9]\)/<a href="#x\2">\2<\/a>\5/gc |
| 03:50 | <variable> | MikeSmith - trying to use regex for HTML ? |
| 10:31 | <annevk> | replied to Paul Duffin anyway |
| 10:31 | <annevk> | his arguments were extremely weak |
| 10:31 | <annevk> | also wrong, which made it easier :) |
| 10:32 | <annevk> | he suggested replacing -moz-x -webkit-x -o-x etc. with a single namespaced x |
| 10:32 | <annevk> | that is not just using namespaces, that is changing the entire approach |
| 10:32 | <annevk> | completely identical to the -wd1-x proposal from a couple of weeks ago |
| 10:32 | <annevk> | except that it uses the cumbersome namespaces syntax |
| 10:49 | Ms2ger | wishes that thread would die already |
| 10:50 | <annevk> | yeah well |
| 10:50 | <annevk> | hey btw, why are you no IE yet for the WebAPPS WG? |
| 10:50 | <annevk> | and if you become one, do you want to become co-editor again as Ms-two-ger? ;p |
| 10:51 | <Ms2ger> | I'll deal with that tomorrow |
| 10:51 | <annevk> | alright |
| 10:51 | <Ms2ger> | Doesn't matter much to me |
| 10:51 | <annevk> | btw, do you know if within a repo you can create another repo? |
| 10:51 | <annevk> | mike asked me if I could put it on dvcs.w3.org/hg/webapps/somewhere |
| 10:51 | <Ms2ger> | No |
| 10:52 | <annevk> | right, that's what I thought |
| 10:52 | <annevk> | so I should just ask for a top-level DOM Core repo on http://dvcs.w3.org/hg/ |
| 10:52 | <Ms2ger> | It would have made more sense to put that repo at hg/webapps/tests, IMO |
| 10:52 | <jgraham> | annevk: Or just make it a subdir of the other repository |
| 10:53 | <jgraham> | You might even be able to preserve history if you care |
| 10:53 | <Ms2ger> | This isn't CVS ;) |
| 10:53 | <annevk> | jgraham, you argued the other day that repos should be self-contained entities |
| 10:53 | <jgraham> | annevk: Well yeah they should |
| 10:54 | <jgraham> | Bu otoh it isn't obviously harmful to have all of WebApps in a single repository |
| 10:54 | <annevk> | indexeddb has a top-level repo |
| 10:54 | <annevk> | we can have one too |
| 10:54 | <annevk> | domcore is important enough |
| 10:55 | <jgraham> | OK, well if indexdb has set a precedent that is fine |
| 11:08 | <annevk> | emailed request for clone, cc'd www-archive |
| 11:42 | <annevk> | so I'm trying to update html5-diff |
| 11:42 | <annevk> | but really, nothing significant has happened |
| 11:42 | <annevk> | current list |
| 11:42 | <annevk> | <li>Numerous changes to the HTML parsing algorithm based on |
| 11:42 | <annevk> | implementation feedback.</li> |
| 11:42 | <annevk> | <li>The <code>hidden</code> attribute now works for table-related |
| 11:42 | <annevk> | elements.</li> |
| 11:42 | <annevk> | <li>The <code>canvas</code> <code>getContext()</code> method is now |
| 11:42 | <annevk> | defined to be able to handle multiple contexts better.</li> |
| 11:42 | <annevk> | <li>The <code>prefetch</code> link relationship can now be used on |
| 11:42 | <annevk> | <code>a</code> elements.</li> |
| 11:42 | <annevk> | <li>The <code>datetime</code> attribute of <code>ins</code> and |
| 11:42 | <annevk> | <code>del</code> no longer requires a time to be specified.</li> |
| 12:59 | <karlcow> | someone wants to create a breadcrumb in the URL bar with nav elements, I could see links used for this with title attributes for the labels of breadcrumbs. Missing a bit of hierarchy. I wonder if it would take off though http://www.uxmag.com/technology/making-the-url-bar-useful-again |
| 13:05 | <annevk> | nav elements? |
| 13:05 | <annevk> | hmm, that's what rel="up up" etc. is for |
| 13:59 | <karlcow> | annevk: I guess what he calls nav element not necessary the one of html5 :) he might not know about the details of the spec |
| 14:13 | <annevk> | whoa, I haven't been following "html5" on twitter for a while |
| 14:13 | <annevk> | it seems now it's mentioned about every minute |
| 14:14 | <karlcow> | annevk: http://paper.li/tag/html5 |
| 14:15 | <annevk> | interesting, "html5" does not include "#html5" |
| 14:32 | <MikeSmith> | karlcow: are you aware of any standards for marking up dictionaries that are actually used across any production applications? |
| 14:33 | <MikeSmith> | I mean stuff that different dictionaries vendors might actually be using |
| 14:35 | <MikeSmith> | hmm, I guess in the free-software world, there is whatever format(s) that dict(1) uses |
| 14:37 | <MikeSmith> | but not sure there's even much of a format at all to that |
| 14:38 | <MikeSmith> | I think it just pulls pre-formatted text records for various DBs without doing any manipulation of them |
| 14:38 | <Philip`> | Wasn't SGML designed for marking up a dictionary? |
| 14:39 | <MikeSmith> | dunno |
| 14:39 | <MikeSmith> | I thought SGML was designed to replace GML |
| 14:39 | <MikeSmith> | SuperGML |
| 14:40 | <MikeSmith> | hey we should relay claim to the S* prefix |
| 14:40 | <MikeSmith> | for "Super" |
| 14:40 | <MikeSmith> | as an alternative to "eXtensible" |
| 14:41 | <Ms2ger> | We're using Web for that |
| 14:41 | <MikeSmith> | so XBL could be SBL, the Super Binding Language |
| 14:41 | <MikeSmith> | Ms2ger: "Web" has 2 more letters than "S" |
| 14:41 | Ms2ger | is stunned |
| 14:42 | <MikeSmith> | if we go with 3 letters, I vote for "SSS" |
| 14:42 | <MikeSmith> | SSSBL |
| 14:42 | <MikeSmith> | "extra Ses for extra Super!" |
| 14:43 | <Dashiva> | S also looks like 5 |
| 14:43 | <MikeSmith> | yeah |
| 14:43 | <MikeSmith> | so it'd be sorta backward-l33t |
| 14:43 | <Dashiva> | S > X on t-shirts |
| 14:44 | <Philip`> | I always seem to notice Microsoft employees using "super" |
| 14:44 | <Philip`> | and pretty much nobody else |
| 14:44 | <MikeSmith> | that's because they "get it" |
| 14:44 | <MikeSmith> | they understand the word "super" and super value |
| 14:45 | <MikeSmith> | *has super value |
| 14:45 | <Philip`> | http://www.theregister.co.uk/2005/08/04/microsoft_nicely/ |
| 14:45 | <Philip`> | I don't know if it's a Microsoft-specific thing or a region-specific thing |
| 14:45 | <Philip`> | but it seems quite noticeable |
| 14:46 | <MikeSmith> | ok |
| 14:46 | <MikeSmith> | point taken |
| 14:47 | <MikeSmith> | they seem to have laid claim to it already |
| 14:47 | <MikeSmith> | damn |
| 14:47 | <MikeSmith> | OK, let's revive "groovy" |
| 14:47 | <MikeSmith> | I bet they don't use that |
| 14:47 | <MikeSmith> | no wait |
| 14:47 | <MikeSmith> | Bitchin' |
| 14:48 | <MikeSmith> | the Bitchin' Binding Language |
| 14:53 | <annevk> | if we can have three leading characters lets go for XXXBL |
| 14:56 | Ms2ger | pushes "XXX DOM Core" |
| 14:56 | <annevk> | objections? |
| 14:57 | <annevk> | I'm starting to think we should just go with "DOM Core" |
| 14:57 | <annevk> | but maybe that is too close to the original |
| 14:57 | <Ms2ger> | We can try |
| 15:12 | <wirepair> | anyone here from opera? |
| 15:13 | <annevk> | somewhat |
| 15:13 | <wirepair> | is there a better way of submitting security bugs than http://www.opera.com/support/bugs/ |
| 15:17 | <annevk> | is there something with wrong with what is explained there? |
| 15:17 | <annevk> | it seems to include guidelines for submitting security-sensitive issues |
| 15:21 | <wirepair> | i'm more concerned with how long things are taking :) |
| 15:21 | <wirepair> | and was curious if there's a person i could just mail directly instead |
| 15:25 | <annevk> | oh, that sounds bad |
| 15:26 | <annevk> | if you have a bug ID I can probably do something |
| 15:27 | <wirepair> | yeah one sec |
| 16:56 | <AryehGregor> | annevk, I don't know if http://www.w3.org/Bugs/Public/show_bug.cgi?id=10267 is actually an issue in WebIDL -- I haven't read it, so it might clearly specify the result. |
| 16:57 | <AryehGregor> | It's poor form to file bugs against a spec you haven't read. :) |
| 17:09 | <annevk> | http://dev.w3.org/2006/webapi/WebIDL/#delete |
| 17:10 | <AryehGregor> | Yes, I looked at the spec and saw that part, but didn't really understand it. I think I was told that the correct answer was to silently ignore the delete, and IIRC I filed a bug against Chrome for it. |
| 17:11 | <AryehGregor> | https://bugs.webkit.org/show_bug.cgi?id=43224 |
| 17:11 | <annevk> | "If property with name P on object O has the DontDelete attribute, return false." prolly means silently ignoring |
| 17:16 | AryehGregor | notes that of the 15 WebKit bugs he's filed, 7 have gotten any response at all that was relevant to the bug, and 4 have gotten a response longer than one comment with no follow-ups |
| 17:18 | <AryehGregor> | Whereas of the 16 Mozilla bugs I've filed, only one has received no response. |
| 17:19 | <AryehGregor> | And only three have received just one response, of which one was INVALID with a legitimate reason, and one was DUPLICATE. |
| 17:21 | <AryehGregor> | And of my six Chromium bugs, four are resolved, and one I just filed a few days ago and it's in the process of being resolved. |
| 17:21 | <AryehGregor> | Maybe I should re-report all my WebKit bugs as Chromium bugs. |
| 17:24 | <annevk> | I guess there's no real project QA |
| 17:24 | <annevk> | at least of the kind that goes through incoming bug reports |
| 17:24 | <AryehGregor> | I've filed two or three Opera bugs too, but I have no idea what happened to them. Of the two IE bugs I filed, one got a "we've confirmed the problem" and the other got no response. |
| 17:25 | <AryehGregor> | Of course, this is somewhat biased by the fact that my median Mozilla bug is surely older than my median WebKit bug, but still. |
| 17:25 | <AryehGregor> | . . . also, I don't think more of the Mozilla bugs are actually *fixed*. |
| 17:27 | <annevk> | if you want follow-up to Opera bugs, feel free to give me bug numbers |
| 17:27 | <annevk> | s/numbers/IDs/ |
| 17:28 | <variable> | AryehGregor, re mozilla bugs: I've seen some take over a year to get fixed |
| 17:28 | <AryehGregor> | variable, um, there are Mozilla bugs that have been open for ten years. |
| 17:28 | <variable> | re chrome bugs, of which I've filed 1, it got fixed within 3 days |
| 17:28 | <AryehGregor> | But it's nice to at least receive an indication that someone has looked at the bug and confirmed it's an issue. |
| 17:28 | <variable> | AryehGregor, yeah I know - longest mozilla bug I've seen open/unconfirmed is 5 years |
| 17:29 | <variable> | I've seen some open longer - but with activity ;-) |
| 17:29 | <AryehGregor> | Yeah, that's my experience too. Chromium bugs get dealt with extremely quickly, Mozilla gets a response but is less likely to be fixed immediately, WebKit is usually ignored, Opera I can't tell, IE I'm not sure in general because I've only filed two bugs. |
| 17:29 | <AryehGregor> | Probably it varies by component somewhat. |
| 17:30 | <annevk> | might also depend on the issue |
| 17:31 | <annevk> | e.g. Chromium also has 16964 unresolved issues :) |
| 17:31 | <annevk> | which is more or less the same for every browser |
| 17:31 | <variable> | I'd love to figure out the statistics for avergage time to response and average time to completion ;-) |
| 17:31 | <AryehGregor> | But that's biased by the nature of the issues filed. |
| 17:32 | <variable> | also - chrome has far fewer features and is newer so it has less code bloat. So most issues can be dealt with fairly fast |
| 17:32 | <variable> | Mozilla has some really ugly and old code |
| 17:32 | <AryehGregor> | Probably Opera and IE get the fewest (and highest-quality?) issues filed, because of poor visibility and poor usability. WebKit and Mozilla probably more. Chromium probably the most -- it's very visible and easy to use and you don't even need to sign up, if you have a Google account. |
| 17:33 | <variable> | AryehGregor, IE had some stats about the quality of bug reports |
| 17:33 | <AryehGregor> | Yes, I saw those. |
| 17:33 | <AryehGregor> | Basically, "the general public is a lot worse at filing bug reports than QA". |
| 17:33 | <AryehGregor> | I hope so, or else you've got some pretty terrible QA. |
| 17:34 | <AryehGregor> | But QA will find different types of issues compared to the general public. |
| 17:38 | <variable> | AryehGregor, I'd like see a simpler "report a bug" function for the general public that is basically screenshot + bounding box + small comment |
| 17:38 | <variable> | and have QA/very technical audience do the rest |
| 17:38 | <dbaron> | annevk, you really think :not(a.foo) should be like :not(a):not(.foo) ? |
| 17:40 | Ms2ger | found that weird too |
| 17:40 | <AryehGregor> | That doesn't make sense. |
| 17:40 | <AryehGregor> | :not(a.foo) should match <a> and <span class=foo>, just not <a class=foo>. |
| 17:41 | <annevk> | AryehGregor, isn't that the same? |
| 17:41 | <annevk> | dbaron, oh wait, hmm |
| 17:42 | <annevk> | dbaron, not anymore |
| 17:42 | annevk | should probably stay out of :not() debates |
| 17:42 | <dbaron> | AryehGregor, also, re: mozilla bugs, a few thoughts: (1) Gecko bugs tend to get ignored when they're filed on Firefox instead of Core. (2) sometimes we have the bad habit of just ignoring bug reports we don't understand rather than saying "I don't understand this" |
| 17:43 | <dbaron> | annevk, you should probably clarify, since so far there's been unanimous disagreement with what bzbarsky suggested |
| 17:43 | <AryehGregor> | dbaron, (1) I always file against Core. (2) All but one received responses, and I'm pretty sure the remaining one is comprehensible: https://bugzilla.mozilla.org/show_bug.cgi?id=586763 |
| 17:43 | AryehGregor | is pretty sure he files good bug reports, as a rule |
| 17:44 | <Ms2ger> | Three people added themselves to the cc list, that's not ignored :) |
| 17:44 | <annevk> | dbaron, replied to Paul Duffin who came to the same conclusion as bzbarsky |
| 17:45 | <AryehGregor> | I said it got no response, not that it was ignored. :) |
| 17:45 | <AryehGregor> | Not that I'm saying it really needs a response. |
| 17:45 | <annevk> | so Chromium people fix those kind of bugs instantly? |
| 17:46 | <annevk> | seems weird |
| 17:46 | <AryehGregor> | Not necessarily, it's just a general impression I get. |
| 17:48 | <AryehGregor> | I'm mostly annoyed at the lack of response to the WebKit bugs. Not even confirmed. |
| 17:49 | <dbaron> | AryehGregor, also, it's clear from the activity that bzbarsky saw it and cc:ed people on it; he just didn't say anything |
| 17:49 | <annevk> | AryehGregor, ask on webkit-dev about it? |
| 17:50 | <AryehGregor> | dbaron, my point was to compliment Mozilla's bug handling, relative to WebKit. |
| 17:50 | <AryehGregor> | Not to denigrate it. |
| 17:50 | <dbaron> | ok, well, I'm glad you think it's good. |
| 17:50 | <dbaron> | It could be better sometimes |
| 17:50 | <dbaron> | but anyway |
| 17:51 | <AryehGregor> | I can't really complain about the Mozilla bugs I've filed except to say they don't always get fixed, which is kind of inevitable. |
| 17:52 | <Ms2ger> | Patches welcome :) |
| 17:52 | <AryehGregor> | I know. |
| 17:52 | <variable> | dbaron, if a bug gets reported to the wrong component someone should really tell the reporter and and re-triage it. I don't always know which component to report issues with |
| 17:52 | <AryehGregor> | Which is why I didn't complain. :) |
| 17:52 | <AryehGregor> | Some of these are pretty trivial, maybe I'll submit a patch to fix another one. |
| 17:53 | <Ms2ger> | Good, then I don't have to do it |
| 17:53 | <dbaron> | variable, that usually happens if the component it starts with is something other than Firefox::General |
| 18:09 | <annevk> | jgraham, http://bitbucket.org/ms2ger/web-dom-core/changeset/562deb8e647e |
| 18:11 | <weinig> | Ms2ger: ping |
| 18:11 | <Ms2ger> | Pong |
| 18:11 | <variable> | AryehGregor, https://bugzilla.mozilla.org/show_bug.cgi?id=554 --> oldest open mozilla bug I could find ;-) |
| 18:11 | <weinig> | Ms2ger: hi, I am curious what the comment on http://www.whatwg.org/specs/web-apps/current-work/multipage/elements.html#embedding-custom-non-visible-data means |
| 18:12 | <variable> | MikeSmith, <MikeSmith> :%s/\(^.*<h[1-6].\+\)\@<!\(\([AB]\.\)\?\([1-9][0-9]\?\.\)\+[1-9][0-9]\?\)\([^0-9]\)/<a href="#x\2">\2<\/a>\5/gc ---> are you trying to use regex to parse HTML? |
| 18:12 | <weinig> | the implementation status says that WebKit does not pass all the relevent tests |
| 18:12 | <weinig> | and I am curious what tests (other than the two linked there) it is referring to |
| 18:12 | <Ms2ger> | Don't read too much into that |
| 18:12 | <weinig> | so I can fix them :) |
| 18:12 | <weinig> | Ms2ger: oh |
| 18:12 | <weinig> | Ms2ger: how much should I read into it |
| 18:13 | <Ms2ger> | A rough indication at the very best |
| 18:14 | <weinig> | Ms2ger: hm, ok, that is a bit confusing |
| 18:14 | <weinig> | Ms2ger: but ok |
| 18:14 | <Ms2ger> | It's got too many states to keep up to date |
| 18:15 | <weinig> | Ms2ger: is there a way we should notifying people when they are wrong? |
| 18:15 | <Ms2ger> | Change it yourself :) |
| 18:16 | <weinig> | Ms2ger: oh, fair enough |
| 18:16 | <Ms2ger> | Or file a bug |
| 18:16 | <weinig> | both very fair options, thanks! |
| 18:16 | <Ms2ger> | Np |
| 19:33 | <jgraham> | annevk: When you get Web DOM Core onto W3C, can you start using the HTML5 copy of testharnes.js |
| 19:33 | <jgraham> | At leat one of those bugfixes is a duplicate of something I already fixed |
| 19:33 | <jgraham> | Having multiple copies of things sucks |
| 19:35 | <jgraham> | Ms2ger: I don't understand what you are trying to achieve with the change on line 350 |
| 20:01 | <annevk> | yeah, we should figure out something |
| 20:01 | <Ms2ger> | To make sure that this.message is a template |
| 20:01 | <Ms2ger> | There's way too many document.all-like things in testharness.js |
| 20:30 | <annevk> | whoa http://www.imdb.com/title/tt1521225/ |
| 20:30 | <annevk> | did not know |
| 20:31 | <AryehGregor> | Compiling Firefox seems to produce a crazy number of warnings, at least by default. I'm not used to these "compiler" things, but wouldn't it make sense to either fix them, or silence them? |
| 20:31 | AryehGregor | has heard that WebKit uses -Werror |
| 20:32 | <AryehGregor> | (plus -Wall, -Wextra, and lots of other -W's) |
| 20:33 | <Ms2ger> | We're working on them |
| 20:34 | <Ms2ger> | But yes, there are a lot |
| 20:34 | <AryehGregor> | Can't you institute a policy like "new commits must not introduce new warnings", at least? Or do you already have such a policy? |
| 20:35 | <Ms2ger> | It's hard to see the new ones among the old ones |
| 20:36 | <AryehGregor> | Well, we have computers for that, right? :) |
| 20:38 | <Ms2ger> | We tried to use -Werror on selected directories, but that would mean you have to test on all platforms before pushing |
| 20:38 | <Ms2ger> | And makes it harder to use newer compilers |
| 20:39 | <AryehGregor> | How does WebKit deal with that, I wonder? |
| 20:39 | <AryehGregor> | othermaciej was the one who mentioned WebKit uses -Werror, so I'd ask him, but he's not here. |
| 20:41 | <Ms2ger> | They have a bot that tries compiling all submitted patches, or something like that |
| 20:41 | <AryehGregor> | And Mozilla commits it and then tries compiling it? Trying to compile and run all tests before the commit succeeds makes more sense, no? |
| 20:41 | <AryehGregor> | (so people don't check out broken code by mistake) |
| 20:42 | <hsivonen> | AryehGregor: the try server doesn't flag new warnings |
| 20:42 | <AryehGregor> | (you probably aren't the best person to ask these questions of :) ) |
| 20:42 | <hsivonen> | but if you end up pushing new warnings to mozilla-central, you will get bugmail about it |
| 20:43 | <AryehGregor> | So is the total number of warnings actually decreasing? Are trivial patches to eliminate warnings welcome? I saw some quite trivial ones, like suggesting you use parentheses with || nested in &&. |
| 20:43 | AryehGregor | doubts he'd be interested in spending his time on that, but is wondering anyway |
| 20:43 | <Ms2ger> | Very welcome |
| 20:44 | Ms2ger | should actually get around to creating bugs for the warnings he's fixed locally |
| 20:44 | <hsivonen> | I don't really know what the trend is |
| 20:46 | <Ms2ger> | How widely supported is Array.isArray? |
| 20:57 | <jgraham> | Ms2ger: I think making this.message a template might be a mistake |
| 20:57 | <jgraham> | (my mistake, obviously) |
| 20:57 | <Ms2ger> | I don't disagree |
| 20:57 | <jgraham> | Ms2ger: What do you mean "document.all like things" |
| 20:58 | <jgraham> | ? |
| 20:59 | <Ms2ger> | Returning a node instead of an array with a single node in it |
| 20:59 | <Ms2ger> | So you have to check everywhere what you got |
| 20:59 | <jgraham> | In the templaing system? |
| 21:00 | <jgraham> | *templating |
| 21:00 | <Ms2ger> | template.render in particular, yes |
| 21:01 | <jgraham> | (I am not unconvinced that the whole templaing system isn't a giant clusterfuck. But I'm not sure what is good) |
| 21:01 | <Ms2ger> | Too many negatives :) |
| 21:02 | <jgraham> | I think it might be a bad idea |
| 21:03 | <Ms2ger> | What's the use case for it anyway :) |
| 21:14 | <Ms2ger> | nn |
| 21:21 | <jgraham> | Ah, Ms2ger left |
| 21:22 | <jgraham> | I guess I havr to save my defence for some other time |