| 00:03 | <benschwarz> | foolip_: check my last two commits - https://github.com/benschwarz/developers.whatwg.org/commits/master/ |
| 00:04 | <foolip_> | that's more like it :) |
| 00:04 | <benschwarz> | also, its been deployed ;) |
| 00:04 | <foolip_> | hmm, I can't see any difference |
| 00:05 | <foolip_> | neither on the splitting or on the class=impl, whichever it was you fixed |
| 00:05 | <benschwarz> | double refresh, app cache is playing with you |
| 00:05 | <benschwarz> | (i think) |
| 00:06 | <benschwarz> | Hixie, foolip_, the spec splitter has split microdata to its own page, but for some reason its getting an extra wrapper div that the others are not. |
| 00:06 | <benschwarz> | compare the "5 microdata" header on http://developers.whatwg.org/microdata.html#microdata |
| 00:07 | <foolip_> | benschwarz, I think that's just what the source looks like |
| 00:07 | <benschwarz> | vs http://developers.whatwg.org/browsers.html#browsers |
| 00:07 | <foolip_> | to organize some stuff for W3C purposes I guess |
| 00:07 | <foolip_> | I had to fix a bug in the spec splitter related to it, I presume you got that with the pull as well |
| 00:10 | <benschwarz> | yeah… but its inconsistant. Anything we can do here @Hixie ? |
| 00:10 | <benschwarz> | http://cl.ly/330I1x3m0d2e1u0A3P3J |
| 00:10 | <foolip_> | benschwarz, does the extra nesting confuse your CSS? |
| 00:11 | <foolip_> | looks like it |
| 00:14 | <benschwarz> | yeah. sadly I had to use a > selector for those headers |
| 00:15 | <Hixie> | what's the problem? |
| 00:16 | <benschwarz> | Hixie: the microdata header has a extra wrapper div that other major sections don't have |
| 00:16 | <dglazkov> | hsivonen: are we friends? I can't decide which circle to add you to. |
| 00:17 | <Hixie> | odd |
| 00:17 | <Hixie> | oh the <div data-component> div? |
| 00:17 | <benschwarz> | yeah |
| 00:17 | <Hixie> | yeah not much i can do about that |
| 00:17 | <Hixie> | can you have your script strip it out? |
| 00:17 | <benschwarz> | if we need to |
| 00:19 | <annevk> | Ms2ger, https://bitbucket.org/ms2ger/dom-range/changeset/d3f907fe0dd7 looks kind of bogus (the text does not match the title values and e.g. previous sibling is different from preceding siblings) |
| 00:28 | <hober> | AryehGregor: email is forthcoming, but basically we much prefer option 3 to option 2, and aren't interested in option 1. |
| 00:31 | <Hixie> | benschwarz: these divs appear all over hte spec, i think it's just luck that this is the only one you've run into |
| 00:31 | <Hixie> | benschwarz: most of the others only appear in parts of the spec not included in developers.whatwg.org source material |
| 00:31 | <Hixie> | benschwarz: like websockets, webstorage, etc |
| 00:37 | <benschwarz> | Hixie: bad luck I guess :) |
| 01:12 | <Hixie> | annevk: looks like it's all up to your now. :-) |
| 01:12 | <annevk> | yeah |
| 01:13 | <annevk> | hopefully I can work with Ms2ger on it next week |
| 01:15 | <annevk> | kind of funny how it puts it under another domain I own that is completely unrelated |
| 01:15 | <Hixie> | heh |
| 01:18 | <Hixie> | anyone got a better suggestion for a name for the "Scripted-submit" flag? |
| 01:18 | <Hixie> | it's set when you use the submit() method |
| 01:18 | <TabAtkins_> | Hm, I thought Chrome had added [].foreach already. :/ |
| 01:19 | <TabAtkins_> | RAGE (╯°□°)╯︵ ┻━┻ |
| 01:19 | <Hixie> | maybe "submitted from method" |
| 01:19 | <hober> | script-triggered-submit? submitted-from-script? no, those kind of suck |
| 01:19 | <Hixie> | or "submitted from script() method" |
| 01:19 | <Hixie> | i guess it doesn't have to be short |
| 01:20 | <annevk> | "submit() flag" |
| 01:20 | <annevk> | XHR has a "send() flag" |
| 01:20 | <annevk> | though kind of different purpose |
| 01:20 | <Hixie> | i'll got with "submitted from script() method" for now |
| 01:20 | <annevk> | script() is not a method |
| 01:20 | <Hixie> | er |
| 01:20 | <Hixie> | submit() |
| 01:20 | <Hixie> | i got it right in the spec text :-P |
| 02:17 | <Hixie> | man, opera's support for <datalist> is... not per spec |
| 05:11 | <MikeSmith> | "CSE HTML Validator generates a message about missing meta description tags" |
| 05:14 | <MikeSmith> | “CSE enforces better style by letting you know when "td" and "tr" tags are not closed. Even though omitting these end tags is technically allowable in HTML 4.01, it is not recommended because it may cause rendering problems in some browsers” |
| 05:14 | <MikeSmith> | solid gold man |
| 05:38 | <erlehmann> | >Download Trial |
| 05:38 | <erlehmann> | >Buy Now |
| 05:38 | <erlehmann> | :3 |
| 05:38 | <erlehmann> | Oh, and … |
| 05:39 | <erlehmann> | >Try CSE HTML Validator Pro free for 200 validations or 30 days, whichever comes first. |
| 05:39 | <erlehmann> | COMEDY GOLD |
| 05:39 | <erlehmann> | let me unpack the strategic popcorn reserves. |
| 11:16 | <jgraham> | foolip: I read machineval and "machin eval" and wonder what a machin is… |
| 11:16 | <jgraham> | I think this suggests industrial injury caused by over-exposure to js |
| 11:36 | <hsivonen> | Does this mean that TimBL doesn't approve of the idea of not having Microdata spec say how Microdata maps to RDF? http://lists.w3.org/Archives/Public/www-tag/2011Jul/0120.html |
| 11:52 | <erlehmann> | hsivonen, is that a rhetorical question? |
| 11:59 | <hsivonen> | erlehmann: no. I was semi-expecting Tim to be OK with Microdata not defining a mapping to RDF |
| 11:59 | <hsivonen> | erlehmann: since then Microdata would step on RDF's toes less than by defining a mapping that sucks |
| 12:00 | <jgraham> | Well he seems to want it to define a mapping that doesn't suck |
| 12:00 | <erlehmann> | i should probably learn about the reasons why a mapping would have to suck |
| 12:00 | <jgraham> | I wonder if a concrete proposal for such a mapping exists |
| 12:02 | <hsivonen> | I think it would be happier for everyone if a group of people who want to use such a mapping drove the definition of the mapping |
| 12:02 | <hsivonen> | instead of Hixie slapping together something when Hixie obviously isn't going to use the mapping |
| 12:07 | <erlehmann> | sounds reasonable. |
| 12:17 | <hsivonen> | TabAtkins: in a presentation of yours you said that new PDF readers don't support old PDF features |
| 12:17 | <hsivonen> | TabAtkins: any concrete examples? |
| 12:18 | <hsivonen> | TabAtkins: (I'm assuming that archivists aren't using the insecure parts of PDF that allow applicaiton invocation) |
| 12:19 | <hsivonen> | TabAtkins: IIRC, the only feature that PDF has ever dropped was LZW compression, but LZW *de*compression is still supported |
| 12:20 | <hsivonen> | hmm. Actually, I recall once seeing a batch of legacy PDFs whose fonts only worked in Adobe Reader--not in Evince |
| 13:48 | <hsivonen> | Hixie: is http://www.jenitennison.com/blog/node/161 on your feedback radar? |
| 14:07 | <erlehmann> | hsivonen, that is a perfect trolling opportunity to demand namespaces, again! :D |
| 14:08 | <Ms2ger> | Don't we have enough trolling opportunities already? |
| 15:48 | karlcow | proposed text for the itemid http://www.w3.org/Bugs/Public/show_bug.cgi?id=13452#c2 |
| 15:57 | <smaug____> | is there a coding style for webidl interfaces? |
| 16:00 | <Ms2ger> | Yes :) |
| 16:00 | <Ms2ger> | </unhelpful> |
| 16:01 | <smaug____> | bah |
| 16:01 | <Ms2ger> | Two-space indentation for everything except for read-write attributes, those are on " readonly ".length |
| 16:01 | <smaug____> | how to handle {} |
| 16:02 | <Ms2ger> | Not on a new line |
| 16:02 | <smaug____> | looks like HTML spec uses different coding style than gecko |
| 16:03 | <Ms2ger> | Much less indentation for methods, yes |
| 16:03 | <Ms2ger> | s/indentation/spaces in the middle of the line/ |
| 16:16 | hsivonen | subscribes to www-tag |
| 16:17 | <jgraham> | Poor you |
| 16:21 | <jgraham> | (I suppose the less cynical respose is "I hope this means that you are considering standing for election; the general usefulness of the tag might be increased if it wasn't almost-entirely composed of linked data theoreticians rather than people with an interst in improving the actual web) |
| 16:21 | <jgraham> | s/)/")/ |
| 16:24 | <Ms2ger> | But Linked Data is the actual web! |
| 16:24 | Ms2ger | goes back to his cave |
| 17:10 | <Hixie> | hsivonen: it was, but i didn't see anything actionable in a brief scan. Anything specifically I should be looking at? |
| 17:29 | <hsivonen> | Hixie: addressing the use case of supporting both SEO and extraction to calendar for event data |
| 17:30 | <Hixie> | you mean in a situation where you're targetting two UAs who use different vocabularies for the same thing? |
| 17:31 | <Hixie> | i don't see any good way to do that |
| 17:31 | <Hixie> | does anyone actually do the extraction to calendar using microdata? |
| 17:31 | <Hixie> | if they just use iCal then that might be the better way to go |
| 17:32 | <hsivonen> | Hixie: I'm not aware of anyone actually doing calendar extraction, but you wrote a spec for it, no? |
| 17:32 | <Hixie> | i wrote a spec for it assuming anyone would implement it |
| 17:32 | <hsivonen> | Hixie: yeah, targeting two UAs that use different vocabularies for the "same" thing on the high level but different in details |
| 17:32 | <Hixie> | if tehy don't implement it, it's kind of academic |
| 17:33 | <hsivonen> | Hixie: isn't it too early to tell? |
| 17:33 | <hsivonen> | Hixie: but it's not too early to tell that Google like their own thing more than iCalendar for the SEO use case |
| 17:35 | <Philip`> | Could support multiple independent microdata layers, like |
| 17:35 | <Hixie> | realistically i'd be very surprised if anyone ever did calendar extraction with microdata |
| 17:35 | <Philip`> | <li itemscope[0] itemtype[0]="http://schema.org/Event"; itemid[0]="/2011/oscon/" itemscope[1] itemtype[1]="http://microformats.org/profile/hcalendar#vevent"; itemid[1]="/2011/oscon/"> |
| 17:35 | <Hixie> | but if they do we can fix the problem then |
| 17:36 | <hsivonen> | Hixie: why don't you zap the spec text that suggests it, then? |
| 17:37 | <Hixie> | it provides a guide to people who want to write their own vocabularies |
| 17:38 | <Hixie> | plus, it does solve the problem if anyone ever does want to solve it |
| 17:38 | <hsivonen> | ok |
| 17:39 | <Hixie> | looks like sicking wants us to drop the <datagrid>-fallback-to-<select> feature |
| 17:39 | <Hixie> | anyone have any opinions on that? |
| 17:40 | <Michael> | If I knew who "us" was maybe |
| 17:40 | <Michael> | missing context :D |
| 17:40 | <hsivonen> | I thought the feature made sense |
| 17:40 | <Hixie> | us = the web |
| 17:40 | <Ms2ger> | The spec |
| 17:40 | <Ms2ger> | Or that |
| 17:40 | <hsivonen> | but I don't know if it is actually used by authors |
| 17:40 | <hsivonen> | or if it's too soon to analyze that |
| 17:41 | <Michael> | Is the fallback in a library like Modernizr or? |
| 17:41 | <Michael> | oh in the spec? |
| 17:41 | <Michael> | seems like a weird thing to have in a spec |
| 17:42 | <Ms2ger> | What, features? |
| 17:42 | <Michael> | A fallback |
| 17:42 | <Michael> | I thought fallbacks were the jobs of 3rd party developers |
| 17:42 | <hsivonen> | Michael: how do you mean? |
| 17:43 | <Michael> | I think I'm still missing context. |
| 17:43 | <Michael> | I'm going to hush |
| 17:44 | <Michael> | hsivonen, I just saw "fallback from data-grid to select" |
| 17:44 | <Michael> | and thought "Shouldn't that be the job of a 3rd party tool like modernizr?" |
| 17:44 | <Michael> | But I don't know what spec you're talking about or really even who you guys are other than being related to whatwg.org |
| 17:45 | <hsivonen> | Michael: s/grid/list/ |
| 17:45 | <Michael> | sorry |
| 17:46 | <hsivonen> | Michael: the fallback is easy to make work without any JS at all |
| 17:46 | <Michael> | Are you talking about the html5 spec? |
| 17:46 | <hsivonen> | Michael: yes |
| 17:46 | <hsivonen> | Michael: we also have scriptless fallback from <video> to Flash |
| 17:46 | <Michael> | interesting |
| 17:47 | <Philip`> | Michael: The second example in http://whatwg.org/html#the-list-attribute where it uses <select> for non-scripted fallback |
| 17:47 | <hsivonen> | though it sucks somewhat due to the lack of a baseline codec for video |
| 17:47 | <hsivonen> | "Unless this bug is substantially addressed, then I will continue to reopen it |
| 17:47 | <hsivonen> | until it is elevated to an Issue recorded against HTML5 for resolution by the |
| 17:47 | <hsivonen> | WG as a whole." |
| 17:48 | <hsivonen> | should I point out that that's not how the Process works? |
| 17:48 | <Michael> | Philip`, It makes sense. I suppose I'm just not used to seeing fallbacks in specifications |
| 17:48 | <Michael> | What is Sicking's reason for wanting it removed? |
| 17:48 | <Ms2ger> | hsivonen, no |
| 17:49 | <Ms2ger> | People who want to process-troll should first read the process |
| 17:49 | <Hixie> | Ms2ger++ |
| 17:50 | <Philip`> | Michael: HTML has always been designed to allow pages to use new features with relatively graceful degradation in old browsers that don't support them, like using <script><!-- ... --></script> to hide the script text from pre-<script> browsers |
| 17:50 | <Michael> | good point! |
| 17:50 | <Michael> | I didn't know that was in the spec. I thought that was a 'hack' someone found |
| 17:50 | Michael | sits |
| 17:50 | <Ms2ger> | Philip`, I hope you're not trying to call that one a success :) |
| 17:51 | <Hixie> | hey it did work |
| 17:51 | <Hixie> | even if it caused us undue pain over the yeras |
| 17:51 | <Philip`> | (http://dev.w3.org/html5/html-design-principles/#degrade-gracefully) |
| 17:51 | <adactio> | I woud add to that: the design of the datalist element (and the way that it can contain a select element for older UAs) is one of the best examples of designing for backwards compatibility that I've seen. |
| 17:51 | <Ms2ger> | I guess the </script> handling was the real problem |
| 17:52 | <adactio> | (I wrote a bit about the design of datalist here: http://adactio.com/journal/4272/ ) |
| 17:54 | <Philip`> | Ms2ger: It was presumably successful at getting people to use scripts without worrying about how it looked all ugly in Netscape 1, given how many people started using scripts :-) |
| 18:06 | <dglazkov> | good morning, Whatwg! |
| 18:09 | <Ms2ger> | Evening dglazkov |
| 18:09 | <Ms2ger> | Is it me or are you always getting on later? :) |
| 18:10 | <dglazkov> | early meeting, just now got to the keyboard |
| 18:11 | <dglazkov> | <insert excuse> |
| 18:18 | <webben> | Hixie: I'm confused ... I thought you'd given up on speccing datagrid for the moment anyway? |
| 18:19 | <Hixie> | datalist, not datagrid |
| 18:19 | <Hixie> | i misspoke |
| 18:20 | <Michael> | ^^ my fault |
| 18:21 | <Ms2ger> | ^^ Hixie's fault for inventing those names :) |
| 18:21 | <Michael> | lol |
| 18:21 | <Michael> | btw - thank you guys for writing specs :) |
| 18:21 | <Michael> | Disney has embraced HTML5 and we're all loving it. |
| 18:22 | <Hixie> | i've kind of given up on datagrid for now |
| 18:22 | <Hixie> | waiting for dglazkov to finish fixing xbl |
| 18:25 | dglazkov | is procrastinating writing Python code |
| 18:26 | <Ms2ger> | Hey, python is fun! |
| 18:26 | <Michael> | *cough*ruby |
| 18:26 | <Michael> | jk python is also great |
| 18:28 | <erlehmann> | Michael, has disney also embraced open standards? |
| 18:28 | <Michael> | For our software? No |
| 18:29 | <Michael> | We have envious competitors that we have to protect ourselves from unfortunately |
| 18:30 | <Michael> | We did release TEA as open source years back but closed it. Now TEA is deprecated and being replaced. |
| 18:31 | <erlehmann> | what is TEA? |
| 18:31 | <Michael> | A terrible view template language based on Java |
| 18:32 | <Michael> | The idea was to limit developers from low level functions like accessing the filesystem etc |
| 18:32 | <Michael> | http://www.pdmfc.com/tea-site/info/index.html |
| 18:32 | <erlehmann> | Michael, i meant with regards to A/V codecs in HTML5. last i checked, i lived in a country where firefox is the majority browser. |
| 18:32 | <erlehmann> | >best features from Scheme, Tcl and Java |
| 18:32 | <erlehmann> | heh :D |
| 18:32 | <Michael> | haha |
| 18:33 | <remysharp> | if I want to make use of setCustomValidity should I be setting on form.onsubmit or is there an event before the submit? |
| 18:33 | <remysharp> | (I'm getting mixed results between chrome and opera) |
| 18:33 | <Michael> | erlehmann, I'm honestly not sure. That's a good question |
| 18:33 | <annevk> | remysharp, onblur maybe? |
| 18:34 | <annevk> | kind of depends on the UI you want I guess |
| 18:34 | <remysharp> | annevk: say I just want to set the email validation fail message to "that's wrong" |
| 18:34 | <remysharp> | annevk: I'm doing that onblur? that seems...weird... |
| 18:36 | <remysharp> | annevk: in fact, it needs to be oninput instead of onblur, because user can hit enter |
| 18:36 | <remysharp> | annevk: but then I'm setting the custom validity message on every key stroke |
| 18:36 | <remysharp> | that feels really wrong. |
| 18:37 | <erlehmann> | remysharp, why not set it if it is not set? |
| 18:37 | <erlehmann> | hehe ;) |
| 18:37 | <erlehmann> | (i know, i know.) |
| 18:37 | <remysharp> | erlehmann: yeah.... you're going to have to clarify "it if it is not" bit :-\ |
| 18:37 | <annevk> | remysharp, it could be submit and blur |
| 18:38 | <Rik`> | remysharp: onformsubmit should be ok, what's the problem? |
| 18:38 | <remysharp> | annevk: so chrome doesn't get to the submit event if the fields are invalid - so I can't set at that point |
| 18:38 | <remysharp> | so what I'm trying to figure out is if chrome is wrong, or if I'm wrong |
| 18:38 | <remysharp> | (and normally - the browsers are 'righter' than me!) |
| 18:39 | <annevk> | remysharp, oh yeah |
| 18:39 | <remysharp> | working example: http://jsbin.com/icejow/3/edit |
| 18:39 | <annevk> | remysharp, Chrome is correct |
| 18:39 | <remysharp> | annevk: okay, I suspected so |
| 18:40 | <remysharp> | which means, really, I need to hook a pre-submit event |
| 18:40 | <remysharp> | otherwise I'm listening for every keystroke |
| 18:40 | <remysharp> | every keystroke example: http://jsbin.com/icejow/4/edit#html,live - obviously works - but feels utterly wrong |
| 18:40 | <Rik`> | remysharp: onformsubmit and oninvalid ? |
| 18:41 | <remysharp> | Rik`: there's an oninvalid event? |
| 18:41 | <remysharp> | that's what I'm looking for then - one mo /me tests... |
| 18:41 | <annevk> | oh yeah there is |
| 18:41 | <remysharp> | hmm |
| 18:41 | <annevk> | and the event is called invalid |
| 18:42 | <annevk> | one day I will write that terminology blog post |
| 18:42 | <remysharp> | yeah, that's not flying |
| 18:42 | <remysharp> | might be my test - but I doubt it..: http://jsbin.com/icejow/5/edit#html,live |
| 18:43 | <Ms2ger> | Which? |
| 18:43 | <remysharp> | which what? |
| 18:43 | <Ms2ger> | (@annevk) |
| 18:43 | <Rik`> | remysharp: addEventListener('invalid') |
| 18:44 | <remysharp> | Rik`: ah - no form.oninvalid love. |
| 18:44 | <Rik`> | I don't know, I'm guessing :) |
| 18:44 | <remysharp> | Rik`: ah - I thought you were stating |
| 18:44 | <remysharp> | then no - that didn't fly either |
| 18:46 | <Rik`> | remysharp: works for me |
| 18:46 | <remysharp> | Rik`: in chrome? |
| 18:46 | <Rik`> | yeah |
| 18:46 | <Rik`> | but chrome dev |
| 18:46 | <remysharp> | Rik`: which build? |
| 18:46 | <remysharp> | frackin' dev builds :( |
| 18:46 | <Rik`> | 14.0.835.8 |
| 18:47 | <remysharp> | what's annoying is chrome has the oninvalid event listener - it just doesn't look like it's firing on 12. |
| 18:47 | <remysharp> | okay, if that's a bug - oninvalid sounds like the right place to be listening |
| 18:48 | <remysharp> | Rik`: you get the pirate speak error message? |
| 18:48 | <Rik`> | yes |
| 18:48 | <annevk> | Ms2ger, like tag vs element, but then event vs event handler, event type vs event object |
| 18:48 | <Rik`> | remysharp: oh no, right |
| 18:48 | <Ms2ger> | Please do |
| 18:49 | <Rik`> | remysharp: but does chrome support setCustomValidity ? |
| 18:49 | <remysharp> | Rik`: yeah, it does |
| 18:50 | <remysharp> | okay, I just think oninvalid is broken in Chrome - doesn't ever seem to fire |
| 18:50 | <Rik`> | that might be true |
| 18:50 | <Rik`> | I just hate Chrome's implementation of webforms… |
| 18:51 | <remysharp> | well, Opera has them licked, Firefox seems not to care, webkit gets what Chrome has, and Safari's release cycle is so fucking slow I stopped caring. It's a bit of a ball ache really. |
| 18:52 | <Ms2ger> | remysharp, look better at Fx |
| 18:52 | <Rik`> | yeah Firefox has the best non-broken implementation in my mind |
| 18:52 | <Rik`> | but there's not a lot of APIs |
| 18:52 | <Rik`> | (and I might be biased) |
| 18:52 | hober | points out that Safari just released... |
| 18:52 | <remysharp> | which Firefox? I'm on 5 |
| 18:52 | <Rik`> | remysharp: 5 |
| 18:53 | <Rik`> | well 4 also |
| 18:54 | <remysharp> | really? maybe I've been overlooking Fx - well - at least they have the same bug chrome have regarding this oninvalid |
| 18:55 | <Rik`> | volkmar: ping volkmar for forms discussion ;) |
| 18:56 | <remysharp> | anyway, cheers. gotta file some bugs then I guess. |
| 18:56 | <volkmar> | remysharp: what bug do we have with the invalid event? |
| 18:57 | <Rik`> | remysharp: volkmar implemented webforms in Firefox |
| 18:57 | <volkmar> | Rik`: webforms 2 |
| 18:57 | <volkmar> | i don't want to take any credits for the old ones :D |
| 18:57 | <remysharp> | volkmar: I'm right in saying that Fx doesn't support: date, datetime, time, month, etc, range, color? |
| 18:58 | <volkmar> | remysharp: yes |
| 18:58 | <remysharp> | but it does have datalist? |
| 18:58 | <volkmar> | remysharp: though, you will have a hard time finding a good implementation of those... |
| 18:58 | <volkmar> | remysharp: datalist/list, yes |
| 18:58 | <remysharp> | and no number support |
| 18:58 | <remysharp> | I guess I'm not really asking |
| 18:59 | <remysharp> | can I assume they're making their way in? |
| 18:59 | <volkmar> | remysharp: number is nearly ready |
| 18:59 | <remysharp> | cool |
| 19:00 | <remysharp> | and oninvalid doesn't work?: http://jsbin.com/icejow/5/edit |
| 19:02 | <volkmar> | remysharp: could you explain me why you are doing that: |
| 19:03 | <volkmar> | http://jsbin.com/icejow/5/edit |
| 19:03 | <volkmar> | oups |
| 19:03 | <volkmar> | email.setCustomValidity(''); |
| 19:03 | <remysharp> | to reset the error flag that gets set on setCustomValidity('foo') |
| 19:03 | <volkmar> | if (!email.validity.valid) { |
| 19:04 | <volkmar> | but setCustomValidity makes the element valid |
| 19:04 | <volkmar> | and then you check for it being invalid |
| 19:04 | <remysharp> | yep |
| 19:04 | <volkmar> | anyway |
| 19:04 | <volkmar> | actually, i'm in a herry |
| 19:04 | <volkmar> | brb |
| 19:13 | <remysharp> | volkmar: ignore that if you like - a new version that just listens for invalid - that doesn't fire: http://jsbin.com/icejow/7/edit#html,live |
| 19:15 | <volkmar> | remysharp: change 'email.form.oninvalid' to 'email.oninvalid' and it will work |
| 19:16 | <volkmar> | the form element isn't expected to receive the invalid event |
| 19:20 | <remysharp> | volkmar: but shouldn't the form fire the invalid event too? |
| 19:22 | <volkmar> | remysharp: http://www.whatwg.org/specs/web-apps/current-work/multipage/association-of-controls-and-forms.html#statically-validate-the-constraints |
| 19:22 | <volkmar> | remysharp: the invalid event is fired at the invalid fields |
| 19:22 | <volkmar> | the form never gets an invalid event according to the specs |
| 19:22 | <remysharp> | awesome - so shouldn't the event be removed from the form element then? |
| 19:23 | <remysharp> | but in fact my whole test gets borked with oninvalid (though it's a useful event) |
| 19:23 | <Ms2ger> | The event *handler* is on all elements for simplicity |
| 19:24 | <remysharp> | that's a pretty poor reason to have it there. |
| 19:25 | <volkmar> | remysharp: that's the case for nearly all events handler |
| 19:26 | <volkmar> | i think the expceptions are some events for body and a couple of other elements |
| 19:27 | <AryehGregor> | Hmm. So changing your e-mail on Bugzilla is impossible, I guess, because it uses it as an ID? |
| 19:27 | <AryehGregor> | Annoying. |
| 19:28 | <remysharp> | I'm struggling to understand the use of setCustomValidity then |
| 19:28 | <remysharp> | Here's my example - now using email.oninvalid - http://jsbin.com/icejow/9/edit |
| 19:28 | <remysharp> | which works - but you have to hit submit *twice* after a corrected validation error |
| 19:29 | <remysharp> | before I was binding to the onsubmit - but that, as annevk, et al said - is wrong, onsubmit shouldn't submit before the form has validated |
| 19:29 | <remysharp> | setting it on the input event (as per the spec example) is the wrong way |
| 19:30 | <remysharp> | so does anyone have another example of how it should be used? (I'm sure you do!) |
| 19:32 | <annevk> | why is the spec example wrong? |
| 19:33 | <remysharp> | using oninput to set or remove the custom message? |
| 19:34 | <remysharp> | every keystroke it's updating the validation message - it shouldn't be tested on every keystroke |
| 19:34 | <remysharp> | that's one use case, but not a common use case |
| 19:34 | <remysharp> | if you had a input[type=email] and as you typed it told you the field was invalid, it would annoy you (or it would me) |
| 19:35 | <remysharp> | so we wait until they hit submit - i.e. the validation feedback should be set at the "trying to submit" phase |
| 19:35 | <AryehGregor> | annevk, I actually once tried to write down what cases cause HIERARCHY_REQUEST_ERR for a DOM Range algorithm, and it was ridiculously hard. |
| 19:36 | <AryehGregor> | (so I gave up and did it differently) |
| 19:36 | <AryehGregor> | The DOM is way too complicated. :( |
| 19:36 | <Ms2ger> | Heh |
| 19:36 | <Ms2ger> | Let's kill it and start over |
| 19:36 | <annevk> | remysharp, it actually should for :valid and :invalid |
| 19:36 | <annevk> | remysharp, arguably, anyway |
| 19:37 | <annevk> | remysharp, ah yeah, that's why Mozilla has :ui-invalid or some such |
| 19:37 | <AryehGregor> | Ms2ger, I wish. :( |
| 19:37 | <annevk> | hmm |
| 19:37 | <annevk> | AryehGregor, how is it done now? |
| 19:37 | <volkmar> | remysharp: at Mozilla, we are trying to push somithng that might fit you needs: https://developer.mozilla.org/en/HTML/Element/input#attr-x-moz-errormessage |
| 19:37 | <Ms2ger> | Hey, we edit the DOM specs, we can do this, right? |
| 19:37 | <Ms2ger> | Right? |
| 19:38 | <AryehGregor> | annevk, I don't remember, but nothing that would work for you. |
| 19:38 | <annevk> | Ms2ger, kill it? |
| 19:38 | <AryehGregor> | I think I just said "call insertBefore()" or something. |
| 19:38 | <remysharp> | volkmar: yeah, that's something that's specifically talked about in 2nd edition of Introducing HTML5 - which is really good - but I'm trying to get the JS side of things running |
| 19:38 | <annevk> | sure man |
| 19:38 | <annevk> | just when everyone starts to rely on DOM Core we nuke that spec from orbit |
| 19:38 | <remysharp> | volkmar: unless you can man handle all the others browsers to add this property for me :) |
| 19:39 | <annevk> | or replace it with "HAHAHAHA FOOLED YOU KTHXBAI" |
| 19:39 | <Ms2ger> | "Only available at the WHATWG" |
| 19:43 | <annevk> | Ms2ger, "gor" in DOM Range ;) |
| 19:43 | <remysharp> | right, my flight is landing - I'm sure you've had enough of me anyway. Thanks for the help - sorry for the hassling. |
| 19:44 | <AryehGregor> | Oh, I just realized why we have such a ridiculous flood of bugs from TFs now: we're almost at the deadline for LC1. |
| 19:44 | <AryehGregor> | Sigh, bureaucracy. |
| 19:44 | <annevk> | seems to get people moving somewhat |
| 19:45 | <annevk> | I wonder though when they would comment if the W3C did not have snapshots |
| 19:45 | <annevk> | Ms2ger++ for CDATA patches |
| 19:45 | <Ms2ger> | And if the answer was "not", whether that would be a good or a bad thing |
| 19:46 | <Ms2ger> | I should get back to those... |
| 19:46 | <annevk> | Yeah, for a lot of these comments it is somewhat hard to see the long term value |
| 19:46 | <annevk> | Wording always changes over time, concepts are what you really want to comment on |
| 19:47 | <Ms2ger> | So don't comment on my "gor" :) |
| 19:50 | <annevk> | I was gonna add that grammar and such is of course important, but it seems that a TF would have a somewhat different focus |
| 19:50 | <annevk> | well, a TF on accessibility |
| 19:53 | <Ms2ger> | We need a Grammar TF |
| 19:53 | <Ms2ger> | timeless could chair it |
| 19:53 | <annevk> | :) |
| 19:54 | <annevk> | I guess where DOM Range uses insertBefore it can be changed to use http://dvcs.w3.org/hg/domcore/raw-file/tip/Overview.html#concept-node-insert instead |
| 19:55 | <AryehGregor> | annevk, yes, I've been waiting for hooks like that to use. |
| 19:55 | <AryehGregor> | It will make the mutation stuff nicer too if everyone uses them. |
| 19:55 | <AryehGregor> | You should have the replace algorithm invoke the insert algorithm, surely? |
| 19:55 | <AryehGregor> | And define a remove algorithm too. |
| 19:56 | <AryehGregor> | Even if it's only one step right now, it would be a useful hook. |
| 19:57 | <annevk> | I'm not sure it should invoke the insert algorithm. |
| 19:57 | <annevk> | That kind of depends on how we handle mutation events |
| 19:58 | <AryehGregor> | How so? Replace is remove followed by insert, no? |
| 19:59 | <annevk> | would you get two mutation events? |
| 19:59 | <AryehGregor> | How would you not get two mutation events? You have to get both a remove event and an insert event, no? |
| 20:00 | <AryehGregor> | I assume there's no "replace" event. |
| 20:00 | <annevk> | i guess, I have not really thought about it much |
| 20:00 | <annevk> | but once mutation events are there I suspect this will be reorganized somewhat |
| 20:01 | <AryehGregor> | The only catch is you have to do all the exception checks prior to actually removing the existing node, but you can't do the insert followed by the remove because then you won't be able to replace children of Documents that you can only have one of. |
| 20:01 | <AryehGregor> | But you can definitely have a lot less copy-paste than you have now. |
| 20:01 | <AryehGregor> | Copy-paste of a large chunk of text with a few subtle differences is the worst. |
| 20:02 | <annevk> | not sure how to properly merge them |
| 20:02 | <AryehGregor> | Eh, it'll do for now. |
| 20:29 | <annevk> | Ms2ger, any ideas on how to make the boilerplate part of DOM Core easily changeable? |
| 20:29 | <Ms2ger> | Hmm? |
| 20:35 | <annevk> | Ms2ger, so we can publish one copy under a more permissive copyright |
| 20:36 | <Ms2ger> | --filter, I guess |
| 20:36 | <Ms2ger> | I don't want to end up with a Hixiespec-style trainwreck :) |
| 20:38 | <annevk> | agreed |
| 20:38 | <annevk> | I guess it is mostly things that need to be omitted and a few small things added |
| 20:38 | <annevk> | and a change of style sheet |
| 20:39 | Ms2ger | has a look |
| 20:40 | <Ms2ger> | Copyright, This Version, logo, SotD |
| 20:40 | <Ms2ger> | Anything else? |
| 20:42 | <Ms2ger> | And CSS |
| 20:43 | <Ms2ger> | Doctype/meta charset? I wouldn't bother |
| 20:45 | <AryehGregor> | Is there some short way to grab the current function as an object from within the function itself? |
| 20:47 | <annevk> | Ms2ger, add dfn.js? |
| 20:47 | <Ms2ger> | Yes |
| 20:47 | <Philip`> | AryehGregor: arguments.callee ? |
| 20:47 | <annevk> | Ms2ger, could maybe do that for the W3C Editor's draft as well |
| 20:47 | <AryehGregor> | Philip`, which throws an exception in strict mode? |
| 20:48 | <AryehGregor> | (why does it do that?) |
| 20:49 | <Ms2ger> | function foo() { ...foo... } |
| 20:49 | <annevk> | AryehGregor, messes up JIT |
| 20:50 | <AryehGregor> | Hmm, why? |
| 20:50 | <annevk> | because you can modify the function or some such? forgot the details |
| 20:50 | <AryehGregor> | Ms2ger, yeah, but that's a little annoying if the function is called areLooselyEquivalentValues or something. |
| 20:50 | <AryehGregor> | Wondering if there was a shorter way. |
| 20:50 | <AryehGregor> | annevk, can't you do the same by modifying the function's name itself? |
| 20:50 | <Ms2ger> | var callee = areLooselyEquivalentValues; |
| 20:50 | <AryehGregor> | . . . |
| 20:50 | <annevk> | http://stackoverflow.com/questions/103598/why-was-the-arguments-callee-caller-property-deprecated-in-javascript |
| 20:50 | <AryehGregor> | Clever. |
| 20:51 | <AryehGregor> | And that keeps a reference, right, so if I modify callee it will modify the actual function? |
| 20:51 | AryehGregor | is always confused by how assignment works in languages more sophisticated than C |
| 20:51 | <Philip`> | (In C terms, everything in JS is a double or a pointer) |
| 20:51 | <annevk> | heh, answered by olliej |
| 20:52 | <annevk> | can't get any better than that |
| 20:52 | <annevk> | Ms2ger, hahaha typo fix is great |
| 20:52 | <Ms2ger> | :) |
| 20:53 | <AryehGregor> | Philip`, that helps. |
| 20:53 | <Ms2ger> | I live to serve |
| 21:01 | <annevk> | How do you iterate over nodes in a Range? |
| 21:02 | <AryehGregor> | annevk, by writing a custom function. :) |
| 21:02 | <AryehGregor> | Which is fairly nontrivial, in fact. |
| 21:02 | <annevk> | So I heard |
| 21:02 | <AryehGregor> | I wrote a getAllContainedNodes() function in my implementation.js that seems to work. |
| 21:02 | <annevk> | gotta go, talk to you later, might extend NodeIterator at some point |
| 21:22 | <timeless> | Ms2ger: sure.. |
| 21:22 | <timeless> | but someone else has to take minutes :) |
| 21:23 | <Ms2ger> | Minutes? We're all pretty fond of confcall-less groups, here ;) |
| 21:23 | jgraham | would like to say that AFAIK the Testing TF has not provided any significant LC feedback but is doing work that will actually improve both the spec and the web |
| 21:23 | <timeless> | ok |
| 21:24 | timeless | needs to look up `TF` in a table |
| 21:25 | <Ms2ger> | jgraham, I should file my next bug as a response from the Testing TF :) |
| 21:29 | <AryehGregor> | I really need a reliable way in Chrome to get a backtrace when something goes into an infinite loop. |
| 21:29 | <AryehGregor> | I've tried leaving the debugger open and hitting pause, but it generally just crashes the tab. |
| 21:29 | <AryehGregor> | Also, the "would you like to kill this tab" dialog really needs to be non-modal. |
| 21:30 | <jamesr> | there's some level of v8 gdb support, iirc |
| 21:30 | <jamesr> | or you could just look at the native stack |
| 21:30 | <AryehGregor> | I'd kind of prefer using Web Inspector instead of gdb here. :) |
| 21:31 | <AryehGregor> | Now the infinite loops disappeared somehow, yay. |
| 21:35 | <AryehGregor> | Okay, so why don't we define how many pixels xx-small/x-small/.../xx-large are? |
| 21:35 | <AryehGregor> | I guess because if the user changes the default font size, it doesn't change explicitly-specified font sizes? |
| 21:36 | <AryehGregor> | But that doesn't make a lot of sense. Why not just give the user generic options to set a minimum font size and/or scale all text up by a certain factor? |
| 21:36 | <AryehGregor> | Or better yet, scale up everything on the page. |
| 22:22 | <AryehGregor> | Argh, fontSize is a total and absolute mess. |
| 22:22 | AryehGregor | stab stab stab |
| 22:25 | <AryehGregor> | The computed value is only related to the specified value in an impossible-to-handle way. |
| 22:25 | <AryehGregor> | The conversion is lossy and inherently varies between UAs. |
| 22:54 | <AryehGregor> | w |
| 22:54 | <AryehGregor> | a |
| 22:54 | <AryehGregor> | Ack. |
| 22:54 | <AryehGregor> | Awesome, Last Call feedback in the form of a Word document attached to a bug. http://www.w3.org/Bugs/Public/show_bug.cgi?id=13461 |
| 23:22 | <jgraham> | AryehGregor: I assume when the members of the Association of American Publishers publish dictionaries they will use this as an example of irony |
| 23:25 | <jgraham> | (unrelated fact: I am told that my friend's sister worked in Malaysia with Alanis Morrisette's English teacher. Rumours that she had fled Canada after being unable to face the shme are, however, almost certainly false) |
| 23:25 | <annevk> | Wi-Fi on the plane |
| 23:27 | <jgraham> | Or as Samuel L. Jackson would put it, motherfuckin' Wi-Fi on the motherfuckin' plane |
| 23:28 | <annevk> | SEO of specs |
| 23:29 | <annevk> | lol |
| 23:29 | <Hixie> | solid technical grounds, those |
| 23:30 | <Hixie> | which spec matches implementations? |
| 23:31 | <annevk> | I think theirs might match implementations, but if we want to get rid of EventException DOM Core is the way to go |
| 23:31 | <Hixie> | do we want to get rid of EventException? |
| 23:32 | <Hixie> | i thought we'd decided we were gonna have lots of new exception interfaces and they'd each have different names or something |
| 23:32 | <Hixie> | or was the decision something else |
| 23:32 | <annevk> | I'm not sure what the decision there was |
| 23:32 | <Hixie> | i forget what that discussion concluded |
| 23:32 | <annevk> | I don't think it has yet |
| 23:33 | <annevk> | anyway, having a new exception for one method where INVALID_STATE_ERR already matches the semantics seems dumb |
| 23:33 | <Hixie> | can't disagree there |
| 23:33 | <Hixie> | but if it's already implemented, that ship may have sailed |
| 23:33 | <Hixie> | especially if it's interoperable |
| 23:36 | <annevk> | i'm not sure it's interoperable |
| 23:36 | <annevk> | I think WebKit didn't fire it |
| 23:36 | <Hixie> | oh i thought you said the other spec matched implementations |
| 23:36 | <annevk> | well some of them |
| 23:51 | <hober> | AryehGregor: ok, replied on list |
| 23:51 | <hober> | AryehGregor: (finally) |
| 23:54 | <hober> | AryehGregor: haahahaha, I replied on the wrong thread, sorry. stop writing so many emails that I flag in my mailer as needing attention! :) |