| 00:01 | <tantek> | Hixie - ok, will take your word that ;) |
| 01:22 | <rniwa> | hayato_: yt? |
| 01:23 | <hayato_> | rniwa: just on keyboard. |
| 01:23 | <rniwa> | hayato_: what was the reason we have createdCallback as opposed to using constructor? |
| 01:24 | <hayato_> | rniwa: I am afraid that I am not right person to answer apis of custom elements. |
| 01:24 | <rniwa> | hayato_: dimitri? |
| 01:24 | <hayato_> | rniwa: yeah, dominic or dimitri is right person |
| 01:24 | <rniwa> | dglazkov: ping |
| 01:27 | <rniwa> | hayato_: okay, I guess I'll ask about it on public-webapps then thans |
| 01:27 | <rniwa> | thanks* |
| 01:27 | <hayato_> | rniwa: np |
| 01:27 | <rniwa> | hayato_: so... how do you feel about adding something like "template" option to document.register dictionary |
| 01:27 | <rniwa> | hayato_: where template refers to a template element |
| 01:28 | <rniwa> | hayato_: and then we'll auto-create a shadow DOM when this option is specified |
| 01:28 | <rniwa> | hayato_: we can define all of that stuff as an extension to custom elements specifications in the shadow DOM spec. |
| 01:28 | <rniwa> | hayato_: that'll allow us to at least ship the parts of shadow DOM that works for custom elements |
| 01:29 | <rniwa> | hayato_: without risking shipping underspec'ed/broken API for attaching shadow DOM to any element |
| 01:32 | <hayato_> | rniwa: I am afraid I have to read relevant topics on public-webapps carefully to answer that :) |
| 01:32 | <rniwa> | hayato_: okay |
| 01:32 | <hayato_> | rniwa: let me catch up. |
| 01:38 | <Domenic_> | has someone documented how custom elements will work with es6 class syntax? document.register doesn't seem very declarative-syntax-friendly |
| 01:44 | <rniwa> | Domenic_: document.register's second object takes an object with prototype property |
| 01:44 | <rniwa> | Domenic_: so if you did |
| 01:44 | <rniwa> | class MyButton extends HTMLElement { |
| 01:44 | <rniwa> | ... |
| 01:44 | <rniwa> | } |
| 01:44 | <rniwa> | document.register('my-button', MyButton) should work |
| 01:46 | <Domenic_> | rniwa: ah, thank you! |
| 01:47 | <rniwa> | Domenic_: but I don't know why we can't use constructor instead of createdCallback for initializing an element. |
| 01:47 | <rniwa> | Domenic_: it's so awkward in the world where we have ES6 classes |
| 01:47 | <Domenic_> | rniwa: I remember dglazkov (I think) explaining to me once something about inheritance... but then that makes no sense since you can just call super() in subclass constructors, and classes by default do that I think. |
| 01:47 | <rniwa> | because then people can naturally define constructor() inside MyButton :/ |
| 01:48 | <rniwa> | Domenic_: yeah |
| 01:48 | <rniwa> | Domenic_: we have to create the underlying DOM element first anyway |
| 01:48 | <rniwa> | Domenic_: i.e. super class |
| 01:48 | <rniwa> | Domenic_: or "super element" first |
| 01:48 | <rniwa> | Domenic_: so I don't understand what the point of having a separate createdCallback |
| 01:49 | <rniwa> | Domenic_: we were just looking at an example with ES6 classs |
| 01:49 | <rniwa> | Domenic_: and createdCallback looked just so awkward |
| 01:50 | <Domenic_> | rniwa: yeah i would love if you could get rid of it. i forget the reasoning. will monitor your progress closely. |
| 01:51 | <Domenic_> | also, vote for me for tag and i'll get rid of createdCallback! ;) |
| 01:52 | <rniwa> | Domenic_: LOL |
| 01:52 | <rniwa> | Domenic_: so... I'll just make you a TAG and then send a formal objection in your way? |
| 01:52 | <rniwa> | Domenic_: sounds legit. |
| 01:52 | <Domenic_> | with my extensive knowledge of process i can assure you that it works exactly that way |
| 01:53 | <Domenic_> | </joke> (just in case) |
| 01:54 | <rniwa> | Domenic_: ah, you went to CalTech about the same time I would have gone to LOL |
| 01:54 | <rniwa> | Domenic_: I was going to attend there from 2008 through 2010 as a transfer student but CalTech wouldn't finance an international transfer student so I didn't go there. |
| 01:55 | <Domenic_> | rniwa: oh cool/sad! |
| 01:55 | <rniwa> | Domenic_: yeah... i would have loved to work on some quantum gravity craziness. |
| 01:56 | <rniwa> | although i was gonna major in CS. |
| 01:56 | rniwa | stalks |
| 01:58 | <Domenic_> | i still get sad about once a month about abandoning physics and doing computers instead... |
| 01:58 | <rniwa> | Domenic_: yeah, physics is a lot of fun. |
| 01:59 | <rniwa> | Domenic_: unfortunately, i'm not smart enough to work on quantum gravity stuff though :( |
| 01:59 | <rniwa> | seems to require some IQ > 150 |
| 02:00 | <Domenic_> | rniwa: I don't think anyone is ... I mean nobody's getting very far. Everyone just plugs away at their own stuff that is interesting but ultimately probably not going to be built on by many other people. |
| 02:00 | <Domenic_> | so it's more an experience than anything else, IMO :) |
| 02:01 | <rniwa> | Domenic_: I guess so although the recent discovery of Higg's Boson or an equivalent quantum particle would rule out a lot of alternative models out there |
| 02:01 | <rniwa> | Domenic_: I'm hoping to see the grand unification in my life time but maybe i'm too optimistic |
| 02:02 | <rniwa> | would be so cool to be working on that. |
| 02:02 | <Domenic_> | agreed on both of the above... |
| 02:03 | <rniwa> | Domenic_: on the other hand, I'm glad I don't have to have a constant fear of someone else figuring out what I've been studying for 10 years LOL. |
| 06:27 | <MikeSmith> | rniwa: spec bug around parsing that may be relevant for you https://www.w3.org/Bugs/Public/show_bug.cgi?id=24017 |
| 06:27 | <MikeSmith> | since you touched that code recently I think |
| 06:29 | <MikeSmith> | dunno what the spec change would be.. check for a document fragment as an anscenstor instead of checking for a template anscenstor? |
| 06:30 | <MikeSmith> | or as teh root I guess, not just anscestor |
| 06:30 | <rniwa> | MikeSmith: that one is for Hixie |
| 06:30 | <rniwa> | Hixie: but I think he might be having the same confusion I had earlier yesterday |
| 06:43 | <rniwa> | MikeSmith: thanks for the info anyway :) |
| 06:44 | rniwa | needs to go home now |
| 08:50 | <MikeSmith> | Ms2ger: I don't know who manages evangelism bugs overall or what the conventions are but Abdul Rauf [:haseeb] seems to have resolved a number of bugs as invalid that really should be worksforme |
| 08:50 | <MikeSmith> | and resolved them all without any comment |
| 08:50 | <MikeSmith> | https://bugzilla.mozilla.org/page.cgi?id=user_activity.html&action=run&from=-14d&who=Abdulraufhaseeb%40gmail.com |
| 08:51 | <Ms2ger> | Dunno, I guess either works |
| 08:55 | <dekiss> | how can I make email newsletter like normal html5 web page or there is something different |
| 08:55 | <dekiss> | I need it to look good on outlook 2007 too |
| 09:00 | <MikeSmith> | hsivonen: http://bugzilla.validator.nu/show_bug.cgi?id=952 |
| 09:00 | <MikeSmith> | "Can't connect to the database" |
| 10:29 | <zcorpan_> | hsivonen: i tried searching in google for an error message firefox logs in the console, to see what people are having trouble with. this one was a bit interesting http://stackoverflow.com/questions/11633162/character-encoding-not-declared-in-html-document |
| 10:30 | <hsivonen> | zcorpan_: thansk |
| 10:31 | <hsivonen> | MikeSmith: no idea why mariadb had died. Updated and restarted. |
| 10:31 | <zcorpan_> | hsivonen: it seems like at least some people are responding to the message by inserting encoding declarations, which is encouraging :-) |
| 10:33 | <annevk-cloud> | hsivonen: a blog entry on encoding UI would be good to document state of the art and trade offs made in 2013 |
| 10:34 | <annevk-cloud> | hsivonen: e.g. the removal of iso-8859-8-i |
| 10:36 | <hsivonen> | annevk-cloud: I'm inclined to keep quiet for a while to see what complaints arise by people noticing problems on their own. Blogging might stir up concern trolling. |
| 10:36 | <hsivonen> | annevk-cloud: I'm OK with blogging some time later, though |
| 10:37 | <annevk-cloud> | coolio |
| 10:40 | <annevk-cloud> | There is a Chrome ICU bug btw |
| 11:11 | <zcorpan_> | hmm. i think i've hit a bug in both presto and blink where <video> is garbage collected although it has a <track src> child with an onload handler |
| 11:20 | <MikeSmith> | hsivonen: thanks |
| 11:23 | <zcorpan_> | seemingly the same bug in gecko |
| 11:23 | <Ms2ger> | Sounds like a feature, then ;) |
| 11:24 | <Ms2ger> | Is this Gecko with vtt support? |
| 11:24 | <zcorpan_> | yeah, flipped the pref |
| 11:25 | <zcorpan_> | gecko doesn't seem to fire 'load' on <track> when it does load, but i see the difference in the network tab |
| 11:29 | <zcorpan_> | https://critic.hoppipolla.co.uk/d32ea7bc?review=437 |
| 11:40 | <zcorpan_> | i guess i should merge cssom and cssom view so that the wg doesn't end up publishing only one of them |
| 11:47 | <MikeSmith> | anybody up to date on what the state of support is for autocomplete= in UA? |
| 11:47 | <MikeSmith> | do any UAs support anything other than on/off for it? |
| 12:00 | <MikeSmith> | it makes me sad that when certain people file bugs my first reaction is to assume their bug reports aren't useful |
| 12:00 | <MikeSmith> | since nearly all the previous bug reports they've raised have been useless |
| 12:01 | <MikeSmith> | problem is there's the one time out of two dozen or so bug reports where they're actually maybe reporting something useful |
| 12:02 | <MikeSmith> | the bug reporter who cried wolf |
| 12:09 | <Ms2ger> | The bug reporter who cried "implement longdesc"? |
| 12:14 | <darobin> | lol |
| 12:14 | <Ms2ger> | Might have been Hixie |
| 12:16 | <Ms2ger> | (It was) |
| 12:16 | <Ms2ger> | At least for SeaMonkey |
| 12:40 | <MikeSmith> | seriously we need some way to mod individual bugzilla comments up and down |
| 12:41 | <MikeSmith> | so that the usefull comments bubble up to the top of the bug and the ass comments get buried |
| 12:45 | <jgraham> | That would work so well |
| 12:45 | <jgraham> | Particularly when bugs get picked up by reddit |
| 12:46 | <jgraham> | (actually in other cases it could maybe work OK, but dbaron's idea of having a continually updated summary might be better_ |
| 12:46 | <jgraham> | s/_/)/ |
| 12:53 | <Ms2ger> | jgraham, depends on who can vote |
| 13:20 | <zcorpan_> | <http://www.w3.org/mid/52A1CBD6.10503⊙nc>; |
| 13:37 | <MikeSmith> | I wonder how much more absurd the republishing circus can get |
| 13:37 | <MikeSmith> | the webapps wg has a choice about what it publishes and doesn't |
| 13:38 | <MikeSmith> | the W3C staff is not forcing the group to publish anything |
| 13:39 | <MikeSmith> | a bare-minimal principled stand that the group could take is that it doesn't republish a spec without making it crystal clear where the content came from |
| 13:41 | <MikeSmith> | ArtB: my recommendation is that the WebApps WG stop republishing the DOM Parsing and Serialization immediately |
| 13:41 | <MikeSmith> | and that it not start publishing it again until its able to make crystal clear who actually wrote the spec |
| 13:42 | <MikeSmith> | oop |
| 13:42 | <MikeSmith> | s |
| 13:42 | <MikeSmith> | mean that for #wam |
| 13:43 | <MikeSmith> | or #testing rather |
| 13:46 | <jgraham> | MikeSmith: I think you misspelt #w3c-secret-treehouse |
| 13:47 | <MikeSmith> | heh |
| 14:06 | <karlcow> | #wam as in "before you go go" |
| 14:09 | <jgraham> | "Last Christmas / I wrote you a spec / this year, to cause you some tears / I'll publish again with changes" |
| 14:58 | <dekiss> | after I migrated the wordpress website in the html code relative links that begin with /wp-content/.. are not working but if I change them to wp-content/.. they are working, I am confused with this, is this the normal behaviour, relative links in html code should begin with /wp-content/.. or wp-content/.. ? |
| 14:59 | <Ms2ger> | Everywhere on foo.com, /x/.. refers to foo.com/x/.. |
| 15:00 | <Ms2ger> | On foo.com/a/b.html, x/.. refers to foo.com/a/x/.. |
| 16:44 | <dekiss> | Ms2ger thank you.. sorry just saw now |
| 16:54 | <JonathanNeal> | SteveF__: you have a doppelganger. |
| 16:54 | <JonathanNeal> | Who doesn't like discussing HTML much. |
| 16:57 | <SteveF__> | JonathanNeal: probably |
| 17:59 | <zewt> | http://lists.w3.org/Archives/Public/public-webapps/2013OctDec/0768.html http://lists.w3.org/Archives/Public/public-webapps/2013OctDec/0770.html as little as I care about anything related to "process", this reply is excellently bewildering |
| 19:06 | <jamesr__> | i guess he's upset about something but doesn't want to throw accusations around? |
| 19:06 | <jamesr__> | but now i don't know what he's upset about |
| 19:34 | <MikeSmith> | zewt: jgraham pointed out that that pair of messages is only bewildering if you consider Bjorn to be a single person as a opposed to a collective |
| 19:35 | <MikeSmith> | so the first message was sent by Objector Bjorn |
| 19:35 | <MikeSmith> | and then Art's message should have got routed to the Problem Solver Bjorn part |
| 19:36 | <MikeSmith> | but that part of the collective doesn't seem to be operating at all right now and hasn't been for a long time |
| 19:37 | <MikeSmith> | so there was a kind of "does not compute" malfunction in the collective |
| 19:37 | <MikeSmith> | hence the second message |
| 19:38 | <MikeSmith> | I think for greater efficiency the collective should just throw in these kind of cases |
| 19:38 | <MikeSmith> | ProblemSolverNotFoundException |
| 19:39 | <Hixie> | "but that part of the collective doesn't seem to be operating at all right now and hasn't been for a long time" has got to be the most subtle yet cutting put-down i've heard in weeks |
| 19:40 | Hixie | doffs his hat to MikeSmith |
| 19:40 | <astearns> | 503 Service Unavailable |
| 19:44 | <MikeSmith> | no put-down intended. I'm all about Love and positive vibrations. ask anybody :) |
| 19:48 | <MikeSmith> | anyway I guess I've trolololled out my full quota for the day and should get some sleep |
| 19:48 | <MikeSmith> | nn |
| 19:52 | <gsnedders> | Anyone got suggestions on how to find the encoding of unlabelled content, given an aim to be accurate and not fast? Language-identification + dictionary matching seems like a possibly useful approach, but would take a fair bit of tweaking to be accurate I fear. |
| 19:53 | <gsnedders> | (Well, being fast isn't a problem, it just isn't the aim. Accuracy is more important.) |
| 20:36 | <SimonSapin> | gsnedders: I assume you know about this? https://pypi.python.org/pypi/charade |
| 20:39 | <gsnedders> | SimonSapin: Yeah, indeed. Need something more accurate for more documents. |
| 21:12 | <dekiss> | footer element works on older browsers? |
| 21:16 | <Hixie> | define "work" |
| 21:16 | <Hixie> | :-) |
| 21:17 | <dekiss> | the website will show you know the website will render in the browser? |
| 21:17 | <dekiss> | I am reading some articles now suggesting making those elements display block |
| 21:20 | <Hixie> | yes, footer won't break in legacy browsers |
| 21:21 | <Hixie> | it's only in edge case things that you might have trouble |
| 21:21 | <Hixie> | e.g. in legacy browsers <footer> is treated more like <span> than <div> |
| 21:21 | <Hixie> | so it doesn't close <p> elements implicitly |
| 21:21 | <Hixie> | brb |
| 21:37 | <dekiss> | Hixie thanks |
| 22:02 | <Hixie> | ok, message ports and catastrophic failures. |
| 22:02 | <Hixie> | i guess we need an event name for "the other side didn't get GC'ed, but it went away prematurely, and you are never gonna hear back from them" |
| 22:03 | <Hixie> | ontheothersidedidnotgetgcedbutdidgoawayprematurelyandyouarenevergoingtohearbackfromthem="" is a bit long |
| 22:03 | <Hixie> | oncatastrophy="" |
| 22:03 | <Hixie> | onerror="" maybe? |
| 22:31 | <smaug____> | onerror is too common |
| 22:33 | <smaug____> | though, perhaps that is a good thing |
| 22:51 | <jgraham> | ondeath |
| 22:55 | <jsbell> | On a tangent... at various points there has been the suggestion to bubble those sorts of errors up all the way to window.onerror, but today window.onerror is only set up for uncaught exceptions (with file/line number). Is that crazy talk? |
| 23:00 | <jgraham> | It's not obviously web-compatible at least |
| 23:06 | <Hixie> | jsbell: what sort of errors? like, img 'error' events? |
| 23:07 | <Hixie> | jsbell: not really clear how you'd handle them, i mean, they're not the same kinds of errors |
| 23:07 | <Hixie> | smaug____: yeah, i had the same thought (too common, but maybe that's good). not sure what the answer is. |
| 23:08 | <Hixie> | onerror is the best i've come up with so far |
| 23:08 | <Hixie> | ondeath is a bit too... strong |
| 23:09 | <jsbell> | agreed, it's not a concrete request yet. The idea is "let me just attach error logging to one place" for stuff that'll get reported back to the server. The specific case that's on my radar is Indexed DB (of course); the events already bubble up to a database object so it's not so bad. |
| 23:09 | <Hixie> | for events in the dom that you want to see you can use addEventListener() on Window to set a capturing handler |
| 23:10 | <Domenic_> | onagonizingdemise |
| 23:10 | <jsbell> | http://lists.w3.org/Archives/Public/public-webapps/2011JanMar/0342.html was the discussion, FWIW |
| 23:11 | <jsbell> | I plan to keep ignoring it |
| 23:13 | <jgraham> | Hixie: You are passing up the chance to make the web just a little bit more metal ;) |
| 23:14 | <jsbell> | on�rror |
| 23:14 | <Hixie> | jgraham: i shall have to find a way to live with myself nonetheless :-) |
| 23:14 | <Hixie> | jsbell: thankful-- i mean, unfortunately, we are limiting ourselves to ASCII characters. |
| 23:16 | <jgraham> | we could go with cheap mysticism instead: oncrossingtheveil |
| 23:21 | <juliet> | onbroken |
| 23:28 | <jgraham> | Bah, if we are making serious suggestions then onlost seems better |