| 03:26 | <Giacomo-> | morning |
| 03:27 | <Giacomo-> | I was looking at a website's source after the owner said that it's using html5 and noticed a lot of tags inside an <article> tag. Lik <header> div's some p's and footer. Is that a correct use of the tag? |
| 03:27 | <Giacomo-> | Website is http://ryanmerrill.net/ |
| 03:32 | <AryehGregor> | Looks legit: http://validator.nu/?doc=http://ryanmerrill.net/ |
| 03:34 | <Giacomo-> | I think the validator is a bit too forgiving :P |
| 03:35 | <Giacomo-> | I checked it there before coming here, but would you agree that the article tag isn't so appropriate there? |
| 03:35 | <Giacomo-> | Maybe a section would of been better |
| 03:36 | <Hixie> | the <article> seems fine |
| 03:36 | <Hixie> | the <hgroup>s at the top are a little more dubious |
| 03:37 | <Hixie> | should probably just be one of them or something |
| 03:37 | <Hixie> | not really clear what the right outline would be |
| 03:37 | <Giacomo-> | hm, I see |
| 04:51 | <MikeSmith> | dammit |
| 04:52 | <othermaciej> | MikeSmith: ? |
| 04:53 | <MikeSmith> | othermaciej: ah, nothing specific, just a general curse.. several things not working as expected |
| 04:53 | <MikeSmith> | e.g., I'm trying the Check for Updates option in Opera.. I select it and nothing happens |
| 04:54 | <MikeSmith> | and my XChat not working like it should |
| 04:54 | <MikeSmith> | othermaciej: btw, I tried Colloquy and really liked it, but it didn't work as expected with this "tircd" thing that I use for twitter |
| 04:55 | <othermaciej> | MikeSmith: what's tircd? |
| 04:55 | <othermaciej> | IRC-to-twitter gateway? |
| 04:56 | <MikeSmith> | yeah |
| 04:56 | <boblet> | oh hey MikeSmith |
| 04:56 | <MikeSmith> | boblet: hey man |
| 04:56 | <MikeSmith> | you been trying to ping me for a while |
| 04:56 | <boblet> | heh—wondering if I need to book a park bench for next Wednesday or not :) |
| 04:57 | <MikeSmith> | yeah, sure, definitely |
| 04:57 | <boblet> | thanks man (as always) |
| 04:58 | <boblet> | also did you get any info about the speech? just an hour on whatever, huh? |
| 04:58 | <MikeSmith> | yeah, I think so |
| 04:58 | <boblet> | hope my Japanese is in working order that day |
| 04:59 | <MikeSmith> | mine will be in its usual order, namely bad |
| 04:59 | <boblet> | hehe. I figure lots of code examples np |
| 05:00 | <MikeSmith> | I will do what I can but will probably end up being lazy and falling back to English and kataka-izing prolificly |
| 05:01 | <MikeSmith> | as far as my part, I'm just going to do a walk-through of the validator.nu sources and demo |
| 05:02 | <MikeSmith> | btw, I think Hatano-san will be making a visit to Osaka some time soon |
| 05:02 | <MikeSmith> | to do a presentation that we helped line up for him |
| 05:03 | <MikeSmith> | boblet: hey, I didn't tell you but Isshiki-san and Ashimura-san and I were in Osaka a couple weeks ago |
| 05:03 | <MikeSmith> | I didn't call you because it was a whirlwind visit |
| 05:03 | <boblet> | MikeSmith: I think Hatano-san is here for next week right? Hopefully gonna catch up in the weekend |
| 05:04 | <boblet> | heh, np. what’s with all the activity down here? *nothing* happens here :/ |
| 05:04 | <MikeSmith> | dunno when he's going, didn't realize it was that soon |
| 05:04 | <MikeSmith> | boblet: visiting friends at Panasonic |
| 05:04 | <boblet> | Kyoto if I’m not getting confused |
| 05:04 | <boblet> | aah, nice |
| 05:06 | <MikeSmith> | we arrived at lunchtime then had a couple meetings and then left pretty early because we had something else we had to go to the next day in Tokyo |
| 05:07 | <boblet> | yikes. you weren’t kidding |
| 05:09 | <MikeSmith> | btw, thanks for that smoked meat. That stuff is amazing. ate it with some australian red wine, because I was hungry and had neither a razor nor a bottle of Laphroig close at hand |
| 05:10 | <boblet> | good stuff. yeah Aussie red works well—will bring some up for next month… |
| 05:11 | boblet | wishes <address> could be used inline (want to add copyright info to same sentence) |
| 05:42 | Hixie | has no idea what to do with ian's idea of uploading directories |
| 06:13 | <murr4y> | hi, i was wondering if someone had a general idea of the status of web fonts today? |
| 06:14 | <murr4y> | i mean, håkon lie made a stir around 2007/2008, and i haven't heard much about it since |
| 06:15 | <murr4y> | what's generally supported these days? tt? eot? plain opentype? woff? |
| 06:15 | <murr4y> | sorry for not having a very specific question :p just wondered if you have any opinions |
| 06:34 | <MikeSmith> | murr4y: I think the css3.info site might have some details |
| 06:35 | <MikeSmith> | murr4y: http://www.css3.info/?s=font |
| 06:38 | <murr4y> | nice, thank you |
| 08:19 | <annevk> | ooh, larry made a Formal Objection to publishing the drafts, quelle surprise |
| 08:24 | <boblet> | murr4y: for syntax check http://paulirish.com/2009/bulletproof-font-face-implementation-syntax/ |
| 08:25 | <Lachy> | annevk, I don't see Larry's objection anywhere on public-html |
| 08:25 | <hsivonen> | annevk: where? |
| 08:25 | <annevk> | see the minutes of the telcon |
| 08:27 | <Lachy> | <masinter> do I need to repeat objections? |
| 08:27 | <Lachy> | paulc: the co-chairs are aware of the formal objection |
| 08:27 | <Lachy> | rubys: it would be helpful to repeat the objection |
| 08:27 | <Lachy> | paulc: it would be helpful to people who aren't reading w3-archive email |
| 08:27 | <Lachy> | so I guess his objection was originally sent to w3c-archive |
| 08:27 | <hsivonen> | oh. I didn't know you are allowed to make FOs in secret |
| 08:29 | <hsivonen> | The section "On Standards" in http://blogs.adobe.com/jnack/2010/01/sympathy_for_the_devil.html is insightful |
| 08:30 | <Lachy> | for those with access, this is the only mention of an objection that I can find from Larry http://lists.w3.org/Archives/Member/w3c-archive/2010Feb/0108.html |
| 08:30 | <Lachy> | maybe I missed the actual objection elsewhere, since that talks about it in past tense |
| 08:32 | <zcorpan__> | it'd be awesome if adobe selled authoring tools for html5 |
| 08:33 | <zcorpan__> | i hope they'll do it |
| 08:33 | <annevk> | hsivonen, I guess it shows that the W3C approach might not really work |
| 08:33 | <Lachy> | zcorpan__, I expect them to update Dreamweaver with HTML5 support |
| 08:34 | <zcorpan__> | Lachy: i mean the tools they currently have to produce flash content |
| 08:35 | <asmodai> | Why would anyone object to publishing drafts? :S |
| 08:35 | <Hixie> | has anyone else noticed how larry's e-mails get narrower and narrower the longer they get? |
| 08:35 | <Hixie> | i wonder what causes that |
| 08:35 | <Hixie> | it messes with my head anytime i actually read all the way through his long e-mails |
| 08:36 | <annevk> | custom line breaks? |
| 08:36 | <asmodai> | Hixie: as in a funnel shape? |
| 08:36 | <Hixie> | asmodai: yeah, though left-aligned |
| 08:36 | <zcorpan__> | clearly, if you pull an email to make it longer, it has to get narrower |
| 08:36 | <Hixie> | wedge shaped |
| 08:37 | <asmodai> | mmm, are the emails angry in tone? Anger might explain rational thoughts short circuiting causing use of smaller sentences or so. |
| 08:38 | <annevk> | Hixie, fyi, redirects are a mess |
| 08:38 | <Hixie> | asmodai: not especially |
| 08:40 | <annevk> | which begs the question, what isn't? |
| 08:40 | <asmodai> | Hixie: Mmm, interesting then. |
| 08:40 | <Hixie> | http://lists.w3.org/Archives/Public/public-html/2010Jan/0539.html is an example of the wedge |
| 08:41 | <Hixie> | though it's more a step than a wedge |
| 08:42 | <asmodai> | Wonder if he hits enter manually. |
| 08:43 | <Hixie> | no response to my comment on http://blogs.adobe.com/jd/2010/02/adobe_authoring_for_html5.html :-( |
| 08:48 | <asmodai> | I'm not sure what to think of Adobe nowadays to be honest. |
| 08:48 | <asmodai> | I dislike Flash with a passion in some areas and can understand its use in some others. |
| 09:20 | <othermaciej> | Hixie: never noticed the wedge before - that's weird |
| 09:21 | <Hixie> | it's really noticable in http://lists.w3.org/Archives/Member/w3c-archive/2010Feb/0100.html |
| 09:22 | <Hixie> | also http://lists.w3.org/Archives/Member/w3c-archive/2010Feb/0079.html |
| 09:24 | <Hixie> | i love how http://lists.w3.org/Archives/Member/w3c-archive/2010Feb/0106.html talks about how transparency is important |
| 09:27 | <annevk> | lolz |
| 09:27 | <annevk> | so cross-document messaging is without a W3C home for now? |
| 09:27 | annevk | didn't realize that was nuked from orbit too at the W3C |
| 09:30 | <danbri> | Hixie, sorry that's a password protected link... :? |
| 09:31 | <Hixie> | surely not, the author clearly indicates that transparency is important |
| 09:34 | <danbri> | i hope all the w3c member archives eventually get archived in public, for future historians to marvel over |
| 09:35 | danbri | reads some of the webapps rechartering thread yesterday ... not so fun reading |
| 09:35 | <Philip`> | Someone should archive it all now, and post it on Wikileaks in 20 years once nobody cares that much about keeping it private any more |
| 09:36 | <danbri> | something like that seems reasonable |
| 09:36 | <danbri> | hixie could donate his inbox too |
| 09:36 | <danbri> | s/20/50/ |
| 09:37 | <Hixie> | s/50/0/? :-) |
| 09:37 | <annevk> | Hixie, http://dev.w3.org/html5/postmsg/#idl is empty? |
| 09:38 | <annevk> | Hixie, also, could you rename the draft to Web Messaging as was suggested by othermaciej? then I can simply reference it as such from html5-diff |
| 09:39 | <hsivonen> | annevk: what got nuked from the orbit and how? did one of the split out part not get adopted by webapps? |
| 09:39 | <annevk> | pretty much |
| 09:39 | <annevk> | though maybe after june |
| 09:40 | <hsivonen> | yay for time to market |
| 09:41 | Hixie | blogs |
| 09:44 | <Hixie> | annevk: done |
| 09:46 | <othermaciej> | hsivonen: that spec is probably one of the most widely implement features of HTML5 (at least as far as actually new features go) |
| 09:46 | <hsivonen> | othermaciej: fortunately implementations don't wait until CR |
| 09:46 | <othermaciej> | so it doesn't actually matter |
| 09:47 | <othermaciej> | however WebApps is likely to complete rechartering within a few months at which point it can be published as FPWD |
| 09:49 | <annevk> | was anything else split out that I didn't notice? |
| 09:56 | <meledin> | Am I the only one to think "quantum particle theory" when reading the "Message Channels" section of the HTML5 Messaging document... |
| 09:56 | <Hixie> | heh |
| 09:56 | <Hixie> | the spec even uses the term "entangled" |
| 09:56 | <Hixie> | it's not a coincidence :-) |
| 09:57 | <meledin> | That's a comfort. Means I haven't become too geeky yet. |
| 09:58 | <Hixie> | i did physics at university |
| 09:58 | <meledin> | Are there any plans to extend message channels between different hosts (by which I mean hosting environment/browser) or is it strictly in-browser currently? |
| 09:58 | <othermaciej> | annevk: I think microdata, microdata vocabularies, canvas 2d context, and postMessage/MessageChannel are the only drafts spun off since our last publication |
| 09:58 | <Hixie> | i occasionally like to give hat tips to my old subject :-) |
| 09:59 | <annevk> | oh right, microdata vocabularies |
| 09:59 | <annevk> | is that the three separate drafts? |
| 09:59 | <annevk> | who's publishing those? |
| 09:59 | <annevk> | apart from the WHATWG, of course |
| 09:59 | <Hixie> | vocabs are gone altogether currently in w3c space |
| 10:00 | <annevk> | there's still editor drafts in cvs |
| 10:00 | <annevk> | but okay |
| 10:01 | <Hixie> | it's not being updated |
| 10:02 | <annevk> | k |
| 10:02 | <annevk> | i'm adding "Specific microdata vocabularies are gone altogether in the W3C draft |
| 10:02 | <annevk> | of HTML5 and are not published as a separate draft. The WHATWG draft of |
| 10:02 | <annevk> | HTML5 still includes them." |
| 10:18 | <othermaciej> | Hixie: can you remind me of your POST test case? |
| 10:18 | <othermaciej> | Hixie: the one that demonstrates how to do a POST without the user's permission and without using scripting? |
| 10:19 | <Hixie> | it's in this directory: http://damowmow.com/playground/demos/http/002/ |
| 10:25 | <othermaciej> | Hixie: thanks |
| 10:25 | <othermaciej> | one of my favorite test cases ever :-) |
| 10:25 | <Hixie> | yeah i kinda like that one too |
| 10:39 | <virtuelv> | that's sort of hard to protect against, no? |
| 10:39 | <Hixie> | nah, not really |
| 10:39 | <Hixie> | same-origin, just avoid doing it |
| 10:40 | <Hixie> | cross-origin, use Origin or other CSRF protections |
| 10:40 | <virtuelv> | yes, in terms of protecting the server |
| 10:41 | <virtuelv> | also delivering some hash to the user in a hidden form field |
| 10:42 | <virtuelv> | I was more concerned with the "let's hide from the user that we're doing a post" |
| 10:42 | <virtuelv> | and that's a harder problem to protect against |
| 10:42 | <Hixie> | not clear that it's a problem |
| 10:43 | <virtuelv> | Hixie: it's a problem only to the extent that web sites are broken |
| 10:44 | <Hixie> | how do you mean? |
| 10:44 | <virtuelv> | (which is to say, they are, but I don't think we can fix problems that lie between a developer's chair and keyboard) |
| 10:45 | <Hixie> | same-origin, the server can do whatever it wants anyway, without the client even being involved |
| 10:45 | <Hixie> | so it's not clear to me why it matters whether you're doing a GET or a POST |
| 10:46 | <Hixie> | there are advantages to do with UI that give good reasons to do the "right" one |
| 10:46 | <Hixie> | but i don't see any particular threat from being able to "trick" the user into doing a same-origin post with the user's knowledge |
| 10:46 | <Hixie> | if there was, the whole web would have broken down long ago :-) |
| 10:46 | <Hixie> | given that you can script a .post(), etc |
| 10:47 | <Hixie> | er, a .submit(), not a .post() |
| 10:52 | <Philip`> | What makes you think the web hasn't broken down long ago? :-) |
| 10:52 | <Hixie> | if this is broken, i'd love to see it working! |
| 10:53 | <Philip`> | If this is working, I'd hate to see it broken :-p |
| 10:53 | <Hixie> | "broken" would be like the web in the 80s |
| 11:44 | <workmad3> | Hixie: the web didn't exist in the 80's (well... 89, but that's barely the 80's) |
| 11:44 | <workmad3> | although I guess not-existing is a form of broken :) |
| 11:44 | <Hixie> | that's what i meant, yeah :-) |
| 11:44 | <workmad3> | and wow, I need to start checking time-stamps and not reviving hours old arguments :) |
| 11:44 | <gsnedders> | s/80's/80s/g |
| 11:45 | gsnedders | hides |
| 11:45 | <workmad3> | gsnedders: fair point :) |
| 11:45 | <Hixie> | '80s? |
| 11:45 | <meledin> | The English language has been broken for decades... |
| 11:45 | gsnedders | is blatantly too much of a QA to allow such bugs |
| 11:48 | <gsnedders> | Hixie: That would be better still |
| 11:49 | gsnedders | remembers at joint HTML-WG/TAG meeting in '08 timbl making some comment about work on HTML going back 18.5 years, and just commenting in IRC: "/me feels like a n00b, again", causing several people in the meeting to laugh |
| 11:54 | <Hixie> | i wonder if maybe we should just punt on PUT and DELETE methods |
| 11:54 | <Hixie> | for form submission |
| 11:54 | <hsivonen> | I've learned that parser performance isn't parsing but painting |
| 11:55 | <Hixie> | it's not clear that they're really useful given XHR |
| 11:57 | <othermaciej> | hsivonen: did you cause too many incremental updates? |
| 11:58 | <hsivonen> | othermaciej: yes. |
| 11:59 | <hsivonen> | also, it's annoying that Mac and Vista have so different perf characteristics than XP&Linux |
| 12:00 | <meledin> | Hixie: Wouldn't the same line of reasoning apply to GET and POST then? |
| 12:00 | <Hixie> | i understand the use cases for GET and POST from a form's perspective |
| 12:00 | <Hixie> | PUT and DELETE are a bit less obvious |
| 12:01 | <Hixie> | when was the last time you deleted something that had a URL? |
| 12:01 | <meledin> | It stems from the church of REST |
| 12:01 | <asmodai> | Isn't that whole GET/POST/PUT/DELETE at least directly mappable to CRUD? |
| 12:01 | <Hixie> | yeah, i guess it depends how much you buy into the whole REST thing |
| 12:01 | <hsivonen> | Hixie: looking for the wrath of HTTP fans again? |
| 12:01 | <Hixie> | hsivonen: no, i'm trying to work out how to deal with actual issues raised |
| 12:02 | <othermaciej> | if you make it same-origin only (and not allowed to follow cross-origin redirects) then it should be safe |
| 12:03 | <meledin> | To be fair I never used PUT/DELETE in forms at all since, err, they don't work :-) |
| 12:03 | <othermaciej> | I don't know whether it would actually be useful |
| 12:03 | <meledin> | I do POST to /delete |
| 12:04 | <meledin> | Well, the theory being that if you would PUT to, e.g. a post reply on a forum, there would be no risk of double posts since you'd overwrite the last one |
| 12:05 | <asmodai> | I still find the names of PUT and POST confusing. My mind is also nagging me about missing implementations of DELETE in some webservers or so since it is hardly used. Good chance I'm confusing it with something else though. |
| 12:06 | <othermaciej> | now I am curious how control values affect the request for a PUT or DELETE form |
| 12:06 | <meledin> | It would be useful if people doing forms subscribe to REST. Otherwise, no, not really. So to me that's the question to ask. |
| 12:09 | <othermaciej> | it looks like per HTML5, method=DELETE ignores the contents of form controls and just makes a DELETE method request to the action URI |
| 12:11 | <othermaciej> | that would make DELETE as a form action kind of useless |
| 12:11 | <annevk> | hmm, IETF fail |
| 12:12 | <othermaciej> | for PUT, the form contents are encoded into the body, but it seems unlikely that you'd want to upload a resource using any of the three formats... |
| 12:12 | <annevk> | http://tools.ietf.org/html/rfc5646 says it is bcp47 but http://tools.ietf.org/html/bcp47 says it is http://tools.ietf.org/html/rfc4647 which is not obsoleted by the first URL... |
| 12:12 | <hsivonen> | yay for BCP stability |
| 12:13 | <meledin> | Well, for DELETE I would assume if there were form contents (such as a list of items to delete) you could encode it in the URL like for GET. |
| 12:14 | <othermaciej> | yeah, but HTML5 does not say to do that, and it's not clear that would be a good behavior anyway |
| 12:14 | <othermaciej> | because then you could only delete resources that have a query in their URI |
| 12:15 | <Hixie> | yeah it'd make no sense to do that |
| 12:15 | <meledin> | I don't see why? Isn't the query ignored if the remote side isn't interested in it? |
| 12:15 | <Hixie> | DELETE would only make sense in the context of a <button> with a formaction="" attribute |
| 12:16 | <Hixie> | PUT doesn't really make any sense at all |
| 12:16 | <Hixie> | both were mostly added for parity wiht xforms back in the day |
| 12:16 | <annevk> | lots of people seemed to want them |
| 12:16 | <othermaciej> | PUT would make sense if you could upload the contents of a file (from <input type=file>) without any other junk prepended or appended |
| 12:17 | <othermaciej> | or upload the contents of a text field or contentEditable area |
| 12:17 | <meledin> | PUT doesn't literally mean what the form PUTS to the resource is the end result, though? |
| 12:17 | <othermaciej> | uploading urlencoded or multipart form data does not seem very useful |
| 12:17 | <othermaciej> | meledin: that's what http PUT is supposed to mean |
| 12:17 | <othermaciej> | (roughly) |
| 12:17 | <meledin> | My understanding is it just means it overwrites the previous data and it's deterministic in its behaviour |
| 12:18 | <annevk> | how the data is encoded has no relevance to PUT afaict |
| 12:18 | <othermaciej> | for PUT wouldn't you normally want to submit something in a form that could make sense to serve as a resource? |
| 12:19 | <othermaciej> | that's certainly how atompub uses it |
| 12:19 | <meledin> | Creating a thread on a forum, for example? PUT-ing the same form several times (e.g. reload/back/forward) doesn't create multiple threads |
| 12:20 | <Philip`> | meledin: Why doesn't it? |
| 12:21 | <Philip`> | How does the HTTP method make any difference to that, in particular? |
| 12:21 | <annevk> | othermaciej, but the server could take the data out of the multipart and use it as resource, no? |
| 12:21 | <Philip`> | Seems like you could do it with POST if you just add some unique form identifier each time you generate the editing form |
| 12:22 | <Philip`> | which is the same as you'd have to do with PUT except there you'd put the unique ID in the URL to make it point at a new resource, I guess |
| 12:23 | <othermaciej> | annevk: could in theory - but that doesn't seem to be the intent of PUT as I understand it |
| 12:24 | <othermaciej> | annevk: for example it seems like you might want to be able to use PUT to talk to an atompub service, but PUT from a form HTML5-style would make that impossible |
| 12:24 | <othermaciej> | I do think PUT could make sense if you could identify a single control and PUT its value |
| 12:24 | <meledin> | "Unless otherwise specified for a particular entity-header, the entity-headers in the PUT request SHOULD be applied to the resource created or modified by the PUT." -- othermaciej seems right. |
| 12:25 | <meledin> | Don't see any mention of "servers can/should ignore content-type on PUT". |
| 12:28 | <annevk> | entity-header != entity body |
| 12:28 | <annevk> | but I don't really care either way :) |
| 12:29 | <annevk> | maybe not allowing them is better because they have this awkward same-origin restriction |
| 12:30 | <meledin> | Philip`: Basically, you're right. The end result is the same. In the PUT case, though, the end result is forced as part of the programming model, while in the POST case the programmer has to think. Being a dumb monkey, I like not thinking. |
| 12:32 | <Philip`> | Seems like the actual implementation requires the same amount of thinking in either case |
| 12:32 | <Philip`> | The conceptual model of resources and identifiers might make the solution easier to discover, though |
| 12:33 | <meledin> | For you and me? Yeah. For someone who hasn't encountered this particular issue before? ... what you said in fancier words. |
| 12:33 | <Philip`> | but you can use that conceptual model regardless of the implementation details |
| 12:34 | <Philip`> | and the HTML spec/language is not an effective way to teach people that model |
| 12:35 | <meledin> | I've no strong opinions on the last point. I just jumped in as devil's advocate on the reasoning "it's useless because it can be done with XHR" |
| 12:38 | Philip` | doesn't have strong opinions on any points, because he doesn't actually care about any of it |
| 12:38 | <meledin> | Apparently enough to correct someone who's wrong on the Internet ;-) |
| 12:39 | <Hixie> | oh it's not useless because it can be done with XHR |
| 12:39 | <Hixie> | it's useless in general, and, because it can be done with XHR, the use cases that remain are already possible anyway. |
| 12:39 | <Hixie> | that would be a better way to put it. |
| 12:40 | <othermaciej> | changes to the design could potentially make it useful, but it's not clear if they are worthwhile, given that you can do it with XHR |
| 12:40 | <othermaciej> | at least, now that XHR can upload a file |
| 12:42 | <annevk> | I now and then think we should value scriptless solutions a bit more, but maybe that's old school |
| 12:44 | <othermaciej> | I think scriptless solutions for common use cases are valuable |
| 12:44 | <meledin> | Scriptless is awesome, especially for stuff end-users never need to see |
| 12:44 | <meledin> | Because it is highly improbable it can go wrong |
| 12:45 | <othermaciej> | I'm not sure what distinction you are making |
| 12:46 | <meledin> | Then take it that it's worthless I guess :-) I just meant it's useful (to me) to make admin views, and debug views and such in plain HTML without JS or even CSS because it's trivial (to me). Add JS to the equation and now I have to debug |
| 13:25 | hsivonen | doubts the wisdom of making isindex a parser macro |
| 13:28 | <annevk> | isn't it like that in other browsers? |
| 13:28 | <annevk> | e.g. Opera |
| 13:28 | <annevk> | and IE too I believe |
| 13:29 | <zcorpan__> | there's so little isindex compat that we could do the sane thing |
| 13:30 | <jgraham> | Would anyone notice if we coated isindex in concrete and dumped it in the harbour in the middle of the night? |
| 13:30 | <annevk> | sane being dropping it altogether? |
| 13:32 | <hsivonen> | in Opera, the macro is rooted in an isindex element |
| 13:33 | <hsivonen> | in WebKit, the text field in the macro is an element with the name isindex |
| 13:33 | <hsivonen> | well, I'll remove the isindex tests from the DOM Level 2 HTML test suite and see if I can get r+ |
| 13:34 | <hsivonen> | isindex being a macro has caused me disproportionate grief compared to its usefulness |
| 13:34 | <annevk> | don't you need to implement the special processing semantics for isindex form submission? |
| 13:34 | <hsivonen> | annevk: Yes! there's more grief ahead of me before the parser can be shipped! |
| 13:35 | hsivonen | gestures angrily in the general direction of isindex |
| 13:35 | <annevk> | <keygen> might be tricky too |
| 13:35 | <annevk> | given how that's implemented in Gecko |
| 13:35 | <zcorpan__> | sane being parsing it like a proper void element |
| 13:35 | <annevk> | somewhat unfortunate how a disproportionate amount of time goes to bogus features of the past |
| 13:36 | <jcranmer> | if people dropped support for isindex, who would break? |
| 13:44 | <hsivonen> | sent email. I wasn't particularly eloquent, because I'm pretty annoyed with isindex right now |
| 13:46 | <jgraham> | Seriously do we have any examples of actual sites that break due to <isindex> being broken? |
| 13:46 | <jgraham> | s/broken/missing/ |
| 13:47 | <hsivonen> | I'm even more angered when I consider the Concorde effect and how much effort I've already put into isindex |
| 13:52 | <Philip`> | http://likelylotusblue.com/cgi-bin/webglimpse/ |
| 13:52 | <Philip`> | You'd break that kind of thing if <isindex> wasn't supported at all |
| 13:53 | <Philip`> | Oops, that one's broken anyway |
| 13:53 | <Philip`> | http://likelylotusblue.com/cgi-bin/webglimpse/home/llbadmin/likelylotusblue-www |
| 13:53 | <Philip`> | That's slightly less broken but still broken |
| 13:54 | <annevk> | not very convincing examples :) |
| 13:56 | <Philip`> | http://costard.lbl.gov/cgi-bin/man/man2html - there's one |
| 13:56 | <Philip`> | It works if you enter something like "cat" |
| 13:57 | <annevk> | sweet |
| 14:07 | Philip` | fails to find any other examples that actually work |
| 15:05 | <annevk> | ooh, am I being tentatively accused of FUD on public-html? |
| 15:05 | <annevk> | I wonder how depressing it would be to find the last useful email on that list |
| 15:07 | <annevk> | oh, there was exactly 1 non-process related email today |
| 15:07 | <gsnedders> | How did that get through? |
| 15:08 | <Dashiva> | Maybe the process needs to be improved to prevent that from happening |
| 15:08 | <annevk> | and yesterday around 3 or 4 |
| 15:09 | <annevk> | it's sad, though for sure the Process moguls can declare victory for now |
| 15:13 | <TabAtkins> | I have GMail use nearly identical colors for whatwg, htmlwg, and webapps list emails, so I can't tell which is which without specifically looking for it. |
| 15:17 | <annevk> | also, http://twitter.com/shelleypowers/status/8296745940 didn't last very long |
| 15:18 | <jgraham> | And you are surprised why? |
| 15:19 | <annevk> | was that in reply to TabAtkins? |
| 15:19 | <Lachy> | someone should raise a objection about the violation of process that allowed that non-process related e-mail to get through. |
| 15:20 | <annevk> | you might upset the Process moguls |
| 15:24 | <asmodai> | Hixie: Well, your post made it to the front of reddit ;) |
| 15:24 | <asmodai> | oh wait, second page |
| 15:25 | <annevk> | it's really no wonder people prefer the WHATWG list if you study the numbers a bit |
| 15:29 | <annevk> | link to the reddit entry? |
| 15:32 | <TabAtkins> | Heh, just challenged Lawson to a drink-off. |
| 15:32 | <TabAtkins> | Now just have to find an excuse to be in the same place as him. |
| 15:37 | <Lachy> | TabAtkins, it seems like a drink-off is a reasonable excuse. |
| 15:37 | <TabAtkins> | Tell my wife that. ^_^ |
| 15:37 | <Lachy> | ok, is she in IRC? |
| 15:37 | <TabAtkins> | No. |
| 15:38 | <asmodai> | annevk: http://www.reddit.com/r/web_design/comments/b19ob/adobe_now_holding_up_publication_of_the_html5/ |
| 15:38 | <Lachy> | well, then pass on my message for me :-) |
| 15:38 | <TabAtkins> | I will direct you to her soonest. ^_^ |
| 15:41 | <a-ja> | anyone happen to have a xbl binding for details/summary, properly styled, that works with xbl2.js google code? |
| 15:41 | a-ja | is not holding his breath |
| 15:42 | <TabAtkins> | Nope. |
| 15:46 | <Philip`> | Maybe the idea is to split the HTML5 language into as many tiny pieces as possible, and then you can block every one of the pieces except the tiny core that still has the name "HTML5" attached to it, and then you can claim you're not blocking HTML5 |
| 15:46 | <a-ja> | tks anyway....(wasn't *even* gonna include "accessible" in that question) |
| 15:47 | Philip` | can't think of any reasonable and common meaning of "HTML5" by which the 2D canvas context is not part of HTML5 |
| 15:49 | <annevk> | othermaciej, you didn't finish your sentence on public-webapps |
| 15:50 | <annevk> | othermaciej, also, that test seems very pointless to optimize for |
| 15:51 | <othermaciej> | annevk: apparently Mozilla and Microsoft both thought it was sufficiently common to do getElementsByTagName in a loop that they optimized for it years before this test existed |
| 15:53 | <othermaciej> | (in fact, with caching it's not even that much of a perf hit to do it in a loop) |
| 15:53 | <annevk> | exposing the gc seems very wrong |
| 15:53 | <annevk> | also you can optimize by creating small js objects |
| 15:54 | <annevk> | and doing other tricks |
| 15:55 | <othermaciej> | the overhead is in doing a GC allocation at all |
| 15:55 | <othermaciej> | there is no trick you can use to get the speed back |
| 15:57 | <Philip`> | http://www.hotelgoal.com/city/vn/hotel-in-bai-chay.htm |
| 15:57 | <Philip`> | for(i=0; i< document.getElementById(frm).getElementsByTagName('SELECT').length; i++){ |
| 15:57 | <Philip`> | if(document.getElementById(frm).getElementsByTagName('SELECT')[i].value==0){ |
| 15:57 | <Philip`> | Is that what the optimisation is for? |
| 15:57 | <othermaciej> | Philip`: yes |
| 15:58 | <othermaciej> | Philip`: there's only 4 select elements in that page afaict so the effect in this case would be small |
| 15:58 | <Philip`> | It looks a very common pattern |
| 15:58 | <Philip`> | http://www.esiweb.org/index.php?lang=fr&id=280&portrait_ID=66&function=print |
| 15:59 | <Philip`> | while (document.getElementsByTagName("span")[i] != null) { |
| 15:59 | <Philip`> | if (document.getElementsByTagName("span")[i].className |
| 15:59 | <Philip`> | == "fn") { |
| 15:59 | <Philip`> | ... |
| 15:59 | <annevk> | othermaciej, you don't think the gc issue is a problem? |
| 16:00 | <othermaciej> | annevk: I dont think setting custom properties on NodeLists is at all common, so I don't think it's an issue in practice, even though it is theoretically nondeterministic behavior |
| 16:00 | <annevk> | we could forbid custom properties on nodelists |
| 16:00 | <othermaciej> | annevk: if we really wanted to avoid it, we could require returning the same object, which to the level of observational equivalence would require caching it so long as it is live, or if it has ever had a custom property set |
| 16:01 | <othermaciej> | but I don't think that is worth the effort |
| 16:01 | <annevk> | or pin objects |
| 16:01 | <annevk> | yeah |
| 16:01 | <Philip`> | Lots seem to be looping over 'a' and 'img', so I imagine there's going to be quite a few of those |
| 16:01 | <othermaciej> | Philip`: thanks for finding actual examples |
| 16:02 | <Philip`> | See also http://google.com/codesearch?q=getElementsByTagName%5C%28%5B%5E%29%5D%2B%5C%29%5C%5B%5Ba-zA-Z%5D |
| 16:02 | <annevk> | it seems somewhat dangerous to me |
| 16:04 | <othermaciej> | Philip`: clever search regexp |
| 16:04 | <annevk> | time for a beer |
| 16:05 | <othermaciej> | there is probably a small risk of depending on nondeterministic behavior in the unlikely case someone really does set custom properties |
| 16:33 | <openstandards> | Hi, have i came to the right place to ask about semantic mark up involving html 5? |
| 16:35 | <gsnedders> | Yeah, basically |
| 16:39 | <openstandards> | do you get any benefit from using input=search i'm aware of the other new fields which offer validation |
| 16:40 | <miketaylr> | you get a neat little x button to clear the field (in chrome/safari) |
| 16:43 | <openstandards> | ahhh ok thank you, these new ui widgets are much better for making web apps |
| 16:43 | <NickRiviera> | Hixie, Hi. |
| 16:57 | <AryehGregor> | openstandards, http://diveintohtml5.org/forms.html |
| 16:57 | <AryehGregor> | That mentions how most browsers implement the various input types. |
| 16:57 | <AryehGregor> | type="search" mainly just looks more like a native search box on Macs. |
| 16:58 | <AryehGregor> | But it's conceivable that browsers will treat it cleverly in other ways in the future, so why not. |
| 16:59 | <paul_irish> | and these are proposals for others, but not implemented quite yet http://docs.google.com/View?id=dch3zh37_0cf8kc8c4 |
| 16:59 | <AryehGregor> | (Actually, it looks wrong on Chrome on Windows/Linux, IMO: http://code.google.com/p/chromium/issues/detail?id=21560 ) |
| 16:59 | <AryehGregor> | paul_irish, well, those are proposals for WebKit, yeah. |
| 16:59 | <AryehGregor> | Opera and iPhone already implement some of that. |
| 17:00 | <paul_irish> | aye |
| 17:00 | <openstandards> | AryehGregor, i've been playing about with the html 5 spec for sometime opera seems to be leading the way |
| 17:00 | <AryehGregor> | openstandards, in some parts yes, some parts no. |
| 17:00 | <paul_irish> | with webforms definitely |
| 17:00 | <AryehGregor> | For instance, it's only just getting localStorage support in 10.50, IIRC. |
| 17:00 | <AryehGregor> | But yeah, it's had a pretty complete Web Forms implementation for a long time now, it's neat. |
| 17:01 | <openstandards> | i do like the datepicker function thats been added to the spec, thats extremely neat |
| 17:06 | <openstandards> | i haven't seen much info on <menu> yet hows that treated? |
| 17:11 | <openstandards> | aroben, thank you for that rather interesting article never knew the iphone's browser was that good |
| 17:11 | <aroben> | openstandards: I think you might have the wrong person |
| 17:12 | <openstandards> | grrrrr sorry i never tabbed enough |
| 17:13 | <openstandards> | tab complete very useful function when used correctly |
| 17:20 | <Philip`> | With the html5lib serialiser, how can I make it not output doctypes? |
| 17:22 | <AryehGregor> | tail -n +2 |
| 17:23 | <Philip`> | In the token serialiser loop: if text != '<!DOCTYPE html>' |
| 17:23 | <Philip`> | That works well enough for me |
| 17:28 | <AryehGregor> | w3c-archive is accessible to who? Not people who are just HTMLWG Invited Experts, right? |
| 17:29 | <Philip`> | No |
| 17:29 | <AryehGregor> | Hurrah. |
| 17:29 | <Philip`> | Members only, I think |
| 17:29 | <AryehGregor> | What's it used for? |
| 17:29 | <AryehGregor> | Well, I'm a "member" of the HTMLWG, just I only have public access, I think. Or something. |
| 17:29 | <Philip`> | For hiding things from mere invited experts, I guess |
| 17:29 | <Philip`> | but I don't know since I can't see it |
| 17:30 | <AryehGregor> | I can't even find a public list description. |
| 17:30 | <Philip`> | AryehGregor: But you're not a Member |
| 17:30 | <AryehGregor> | Oh, maybe I'm a Participant but not a Member? http://www.w3.org/2000/09/dbwg/details?group=40318&public=1 |
| 17:31 | <AryehGregor> | It says "(public)" next to my name, I guess that means I don't get to see the cool secret stuff. |
| 17:31 | <Philip`> | I think Members are only employees of http://www.w3.org/Consortium/Member/List |
| 17:32 | <AryehGregor> | For some people it says "(public) Invited expert", and for some it says "W3C Invited Experts". |
| 17:32 | <AryehGregor> | Like Tantek is a W3C Invited Expert. |
| 17:32 | <AryehGregor> | Maybe I should bug someone to get Wikimedia joining the W3C and make me their representative. |
| 17:33 | <AryehGregor> | Doesn't seem worth it. I guess I'll have to if I become a full-time employee at any point, though. |
| 17:33 | <Philip`> | Would that be a good use of $8000 of Wikimedia's money? |
| 17:34 | <AryehGregor> | Probably, it has "advancing open standards" as part of its mission. |
| 17:34 | <AryehGregor> | http://wikimediafoundation.org/wiki/Values "We believe that this mission requires thriving open formats and open standards on the web to allow the creation of content not subject to restrictions on creation, use, and reuse." |
| 17:34 | <Philip`> | Would W3C membership help advance open standards? |
| 17:34 | <AryehGregor> | Heh. |
| 17:34 | <othermaciej> | AryehGregor: if you get invited as an Invited Expert to any group but HTML WG, you'd probably become a real Invited Expert |
| 17:34 | <Philip`> | more than spending the money/resources on other things |
| 17:35 | <AryehGregor> | othermaciej, because in most groups, Invited Experts are actually invited experts? :) |
| 17:35 | <Philip`> | and contributing from a non-Member perspective |
| 17:39 | <AryehGregor> | I dunno, that's not my decision. |
| 17:39 | <AryehGregor> | I think that if I get employed by them, though, they have to become a member if I'm to remain part of the HTMLWG, or something? |
| 17:41 | <othermaciej> | AryehGregor: in most groups, yeah, you have to be actually invited |
| 17:41 | <othermaciej> | the $8k fee for nonprofits seems kinda steep |
| 17:42 | <openstandards> | othermaciej, that it does |
| 17:43 | <AryehGregor> | Wikimedia has about $8.6m in revenue for FY2008, so I guess they'd be able to afford it if necessary. Dunno. |
| 17:43 | <AryehGregor> | They plan to make $10.6m in 2009-10. |
| 17:43 | <Philip`> | AryehGregor: I don't believe they do - you can be an Invited Expert as long as your employer is not a Member, though it seems they don't like IE status being used as a way to circumvent membership fees |
| 17:43 | <Philip`> | s/they do/they do have to become a member/ |
| 17:44 | <AryehGregor> | Philip`, when I signed up, I was asked if my employer had anything to do with the web, or something like that. |
| 17:44 | <othermaciej> | I don't know how strict they are about nonprofits |
| 17:44 | <AryehGregor> | That's a point, maybe they'd be nice. Dunno. |
| 17:44 | <Philip`> | I'd expect that if you're contributing as an individual then that's okay |
| 17:44 | <AryehGregor> | I'll deal with that when I come to it. |
| 17:44 | <Philip`> | rather than contributing as effectively an official representative of your employer |
| 17:45 | <AryehGregor> | See, this is what WF2 is good for: http://img.thedailywtf.com/images/201002/errord/linkedinwtf.png |
| 17:45 | <othermaciej> | Wikimedia being a W3C Member would probably be a good thing for marginally improving the quality of AC decisions |
| 17:45 | <AryehGregor> | othermaciej, yeah, I was thinking it would be a good thing to have more members whose interests actually lie with a free web. |
| 17:47 | <othermaciej> | I was thinking from the POV of caring about open source issues, and also just a large content publisher that cares about advancing the web platform |
| 17:47 | <AryehGregor> | Right. |
| 17:48 | <othermaciej> | I wonder if the list at <http://www.w3.org/Consortium/Member/List>; is actually current |
| 17:48 | <AryehGregor> | You'd think $7,900 would be enough to pay someone to update the list. |
| 17:49 | <othermaciej> | I thought I'd heard that Yahoo! is not a member any more |
| 17:49 | <AryehGregor> | That would be pretty interesting. |
| 17:51 | <othermaciej> | It is sad though that there aren't more large content publishers represented (and ideally active in the HTML WG) |
| 17:52 | <othermaciej> | it would be nice to hear direct from e.g. Facebook or Twitter or Amazon what they'd like to see |
| 17:52 | <AryehGregor> | Yeah, kind of funny that none of them are part of the W3C. |
| 17:53 | <AryehGregor> | Does the HTMLWG have any private media of communication? Ian's latest blog post seems to suggest Adobe was raising objections in secret. Do these count for anything until they're filed as bugs or whatnot? |
| 17:53 | <othermaciej> | w3c could probably improve their financial situation a lot if they could convince big Web content companies that W3C membership is relevant to them |
| 17:53 | <othermaciej> | the HTML WG does not have any private mailing lists or the like |
| 17:54 | <AryehGregor> | So what is w3c-archive? |
| 17:54 | <Philip`> | The WHATWG has $0 membership fees and still doesn't seem to attract any people from big Web content companies |
| 17:54 | <Philip`> | so why would the W3C be more successful at getting them involved? |
| 17:54 | <othermaciej> | w3c-archive is like www-archive, but restricted to member-only access |
| 17:54 | <othermaciej> | Philip`: I'm surprised the WHATWG hasn't been successful at it either |
| 17:54 | <AryehGregor> | It is kind of weird. |
| 17:55 | <othermaciej> | Philip`: maybe getting improvements to specs to ultimately get improvements to browsers is beyond their time horizon |
| 17:55 | <AryehGregor> | Whereas I guess implementers have a more short-term interest. |
| 17:55 | <othermaciej> | AryehGregor: the objection I am aware of was first mailed to some members of the W3C Team, and subsequently some related emails were sent to the Member-only w3c-archive list |
| 17:56 | <othermaciej> | why that person thought it was appropriate to raise an objection about a fully public Working Group through private channels, I don't know |
| 17:56 | <AryehGregor> | But it will have to be posted publicly before anyone is required to address it or it can impede the spec's progress in any way, I guess? |
| 17:56 | <othermaciej> | I am also not entirely sure what said person is objecting to, whether it's a Formal Objection, or what |
| 17:57 | <othermaciej> | since the objection was addressed to the W3C Team and not to the Chairs at all, I think it's for the W3C to take care of it |
| 17:57 | <othermaciej> | and the Chairs are going to proceed as if it doesn't exist until we hear otherwise |
| 17:57 | <AryehGregor> | k. |
| 18:05 | <TabAtkins> | >_< I hate the member-only lists. Is there anything *actually* binding me to 'secrecy' about them? Did I agree to something when I became an IE that I'm not aware of? Or is it just a cultural thing? |
| 18:07 | <othermaciej> | http://www.w3.org/2005/10/Process-20051014/comm.html#confidentiality-levels |
| 18:09 | <TabAtkins> | Annoying, but ok. This is frankly an abuse of the member-only list, and I'm disappointed in the author. |
| 18:12 | <Philip`> | It's fun watching arguments on public lists about what was said on private lists, where the participants know what was said but can't actually say what it was |
| 18:12 | <othermaciej> | TabAtkins: there's a particular person who has started quite a few html-wg-related threads on w3c-archive recently, and I must say I do not fully understand why |
| 18:13 | <Philip`> | (...watching from the perspective of someone who doesn't know what was said, and can just reconstruct a vague outline and apply imagination to guess the rest) |
| 18:13 | <othermaciej> | A Formal Objection has to be publicly available per the Process: http://www.w3.org/2005/10/Process-20051014/policies.html#WGArchiveMinorityViews |
| 18:14 | <AryehGregor> | Maybe they want to see if they can apply enough pressure to get something changed without anyone being able to pin it on them? |
| 18:14 | <AryehGregor> | Or they want to test reactions privately before committing to it publicly? |
| 18:14 | <AryehGregor> | Like for PR sort of reasons. |
| 18:14 | <othermaciej> | however, there is also the option to appeal a chair's decision if a member feels they are not getting due process: http://www.w3.org/2005/10/Process-20051014/policies.html#WGAppeals |
| 18:15 | <AryehGregor> | I really haven't noticed anything obnoxious from Adobe on the public lists, or at least not related to Flash or related technologies that HTML5 threatens. |
| 18:15 | <AryehGregor> | There were some posts about how Flash interacts with sandboxes, but nothing terribly unreasonable. |
| 18:15 | <AryehGregor> | (sorry, I mean "plugins", not "Flash") |
| 18:15 | <AryehGregor> | (to be fair, it would be Acrobat as much as Flash) |
| 18:17 | <othermaciej> | AC reps can also appeal certain decisions in a way that is only Member-visible, but I am not sure that applies here: http://www.w3.org/2005/10/Process-20051014/acreview.html#ACAppeal |
| 18:18 | <TabAtkins> | Well, I've registered my disapproval in any case. |
| 18:21 | <TabAtkins> | Hm. Why don't I see w3c-archive on the lists.w3.org page? |
| 18:21 | <othermaciej> | there's a separate page of Member-only lists |
| 18:22 | <othermaciej> | http://lists.w3.org/Archives/Member/ |
| 18:22 | <Philip`> | Sadly that list of Member-only lists is Member-only |
| 18:24 | AryehGregor | is reminded a bit of The Castle. |
| 18:24 | <othermaciej> | those of you who are permitted Member-only access may wish to view this Member-only post where someone reminds me of the importance of transparency: http://lists.w3.org/Archives/Member/w3c-archive/2010Feb/0082.html |
| 18:25 | <AryehGregor> | We should all just follow the example of, I think, David Baron and refuse to post on private lists (unless absolutely required). |
| 18:28 | <AryehGregor> | (I'm currently following that policy with the magnanimous assistance of the W3C, which doesn't let me post on private lists to begin with.) |
| 18:34 | <TabAtkins> | othermaciej: Heehee. |
| 18:35 | <TabAtkins> | AryehGregor: The only relevant member-only list I participate in is the CSSWG's one, and all we use it for is posting agendas and coordinating meetups. This is the maximum extent to which I think private lists should be used. |
| 18:36 | <AryehGregor> | Why does that need to be private? |
| 18:37 | <TabAtkins> | Because there's no particular need for it to be public; it just clutters up the public list with useless stuff. No actual discussion takes place over there. |
| 18:37 | <Dashiva> | That's an answer for "why is it a separate list?", not for why it's private |
| 18:37 | <AryehGregor> | Then you could have it on a separate public list. |
| 18:37 | <TabAtkins> | Sure, but no one would read it. Shrug. |
| 18:37 | <TabAtkins> | It's really irrelevant where that sort of stuff goes. |
| 18:37 | <AryehGregor> | You should err on the side of transparency. |
| 18:38 | <Dashiva> | The thing is, nobody on the outside knows what happens on a private list |
| 18:38 | <Dashiva> | So you _might_ be discussing all kinds of things there |
| 18:38 | <AryehGregor> | That too. |
| 18:38 | <AryehGregor> | It creates the appearance of hiding things. |
| 18:38 | <TabAtkins> | We do giggle over pictures of you, Dashiva. |
| 18:38 | <AryehGregor> | I've heard that it used to be used more often, no? |
| 18:38 | <Dashiva> | There's more than one picture of me? |
| 18:38 | <TabAtkins> | Yeah, it did. And that was a bad idea. |
| 18:38 | <AryehGregor> | I'm sure I heard that some CSSWG members used to never post on the public list at all. |
| 18:39 | <AryehGregor> | From reading othermaciej's link, though, it looks like the W3C treats Members-only areas as a perk of membership. I wonder what the procedure would be to convert a WG to all-public like the HTMLWG? |
| 18:39 | <othermaciej> | Web Apps WG is almost all-public |
| 18:39 | <TabAtkins> | I suppose something like "Hey, could you kill our private list? We don't need it any more." |
| 18:40 | <othermaciej> | there is a member-only list that is only occasionally used for administrative matters |
| 18:40 | <othermaciej> | confidentiality level for a group is defined by the charter |
| 18:40 | <AryehGregor> | Which you just said is really hard to change. |
| 18:40 | <AryehGregor> | Or said recently. |
| 18:41 | <Dashiva> | I don't think the two ancestors of webapps were as open as webapps is |
| 18:54 | <othermaciej> | Dashiva: the webapps charter was intentionally written to be more open |
| 18:54 | <othermaciej> | AryehGregor: it's a pain in the ass to change it, but charters do expire periodically |
| 19:17 | <karlushi> | [12:54] <othermaciej> I wonder if the list at <http://www.w3.org/Consortium/Member/List>; is actually current |
| 19:17 | <karlushi> | the list is dynamic |
| 19:17 | <wycats> | Hixie: what's all this about Adobe :P |
| 19:17 | <AryehGregor> | wycats, he can't say, it's secret. |
| 19:19 | <wycats> | AryehGregor: I know :P |
| 19:20 | <wycats> | I'm on one of the sekrit lists -- /me looks |
| 19:21 | <wycats> | AryehGregor: I was just ribbing |
| 19:23 | <karlushi> | Member lists sometimes give the possibility to share information which is private but need to be archived. It could be phone numbers, addresses, etc. |
| 19:23 | <karlushi> | The Member lists come from a time where closed doors helped to solve conflicts that participants failed to resolve in public (transition ietf to w3c) |
| 19:23 | <wycats> | karlushi: I know I know |
| 19:23 | karlushi | replying to previous comments on irc. not yours wycats |
| 19:24 | <wycats> | karlushi: :) |
| 19:25 | <TabAtkins> | karlushi: Pretty sure all our public information exists on the web anyway. Anyone could find out where I live and what my phone number is without too much difficult, frex. |
| 19:25 | <AryehGregor> | Also, you could just e-mail it to a bunch of people directly without a mailing list. You're not going to be communicating that kind of stuff very often. |
| 19:25 | <karlushi> | TabAtkins, well if you published it |
| 19:26 | <TabAtkins> | I mean that that sort of info tends to get published *anyway*. It's usually public info, even if you *do* have to dig for it. |
| 19:27 | <wycats> | TabAtkins: this brush-up over private lists reminds of the "omg Pelosi and Reid are having a private conversation not on CSPAN" |
| 19:27 | <wycats> | the idea that there will be private communications period is not unreasonable |
| 19:27 | <wycats> | are you going to ban W3C members from any email that is not through the public lists? |
| 19:28 | <AryehGregor> | No, private communications period are fine, but there shouldn't be official W3C fora for them. |
| 19:28 | <AryehGregor> | Official business should all be in public. |
| 19:28 | <TabAtkins> | Of course not. I would like to discourage w3c members from making important arguments about w3c products on private lists. |
| 19:29 | <TabAtkins> | We're developing open standards. Doing anything like this behind closed doors is inappropriate. |
| 19:32 | <wycats> | AryehGregor, TabAtkins: all the "official fora" does is provide an alternative to reply-all |
| 19:33 | karlushi | notes that he has no access to member lists |
| 19:33 | <Dashiva> | wycats: They also legitimize it |
| 19:33 | <TabAtkins> | wycats: What? It also provides publicly-available archives, and an obvious way to both read and write to the relevant group. |
| 19:34 | <wycats> | TabAtkins: I'm not in favor of conducting public business in private, but the idea that private communication shouldn't occur is just asking for problems |
| 19:34 | <wycats> | not everything can be done under the glare of the cameras |
| 19:34 | <wycats> | sometimes people need to be able to make candid arguments to each other |
| 19:34 | <TabAtkins> | I didn't say that private communication shouldn't occur. |
| 19:34 | <karlushi> | TabAtkins, that is true that open is good, but it also creates challenges sometimes. There are issues on both styles. I have seen both at work and none are perfect. |
| 19:34 | <TabAtkins> | I said that public brusiness shouldn't occur in private. |
| 19:34 | <AryehGregor> | Pretty much everything can be done in public. |
| 19:35 | <AryehGregor> | That's how many open-source projects operate, and it works fine. |
| 19:35 | <TabAtkins> | In this case, public business *is* occurring in private. That's inappropriate. |
| 19:35 | <AryehGregor> | People are just candid in public. |
| 19:35 | <Dashiva> | There's also a difference between "private" and "only for people who pay" |
| 19:35 | <karlushi> | Dashiva, here it is also a question of history |
| 19:36 | <karlushi> | moving from one financing model to another one is very very very hard |
| 19:36 | <Dashiva> | That seems unrelated |
| 19:36 | <karlushi> | Dashiva, nope. The full structure of the organization was based on a model which is now challenged by the new trend and the new culture. |
| 19:37 | <karlushi> | But the employees, aka *people*, are still a reality you have to deal with |
| 19:37 | <karlushi> | and W3C has not a big budget |
| 19:38 | <karlushi> | You can decide to open everything, and make a donation plan, but this is a full reorganization. |
| 19:38 | <Dashiva> | That seems unrelated still |
| 19:38 | <karlushi> | switching from one model to the other takes time, a lot of time |
| 19:38 | <Dashiva> | It's possible to have members without letting them run everything in hidden fora |
| 19:38 | <karlushi> | Dashiva, but some members who pay may not want to do it in public. |
| 19:39 | <karlushi> | and if they say, public = no money |
| 19:39 | <karlushi> | how do you deal with it? aka the reality |
| 19:39 | <TabAtkins> | Has anyone said that? |
| 19:39 | <karlushi> | same difficulty for Mozilla-Google |
| 19:39 | <Dashiva> | You respond by making their claims public? |
| 19:39 | <karlushi> | nope |
| 19:40 | TabAtkins | has to bow out and go get his license renewed. bbs |
| 19:40 | <AryehGregor> | karlushi, has any of this been a problem for the HTMLWG? |
| 19:40 | <AryehGregor> | Because that's completely open. |
| 19:40 | <AryehGregor> | Some other WGs are almost completely open at this point, reportedly, like CSSWG or Web Apps. |
| 19:40 | <Dashiva> | AryehGregor: Well, except the w3c-member stuff... |
| 19:40 | <AryehGregor> | Doesn't seem like it needs "a full reorganization". |
| 19:40 | <karlushi> | just being logical. Make a concert in an open area draw a circle line, inside the circle line, you pay, outside you don't pay. |
| 19:40 | <AryehGregor> | Dashiva, that's outside the WG. |
| 19:40 | <karlushi> | I bet you will not get a lot of money |
| 19:41 | <AryehGregor> | karlushi, the W3C can require membership for organizations whose employees post to the list, as I think is the case for the HTMLWG. |
| 19:41 | <karlushi> | AryehGregor, the htmlwg is part of an organization which has 15 years of existence. |
| 19:41 | <AryehGregor> | So you can read, but not influence unless you pay. |
| 19:41 | <karlushi> | history history |
| 19:42 | <Dashiva> | karlushi: Well, actually, the winter sports arena in Oslo had a hill just outside called "the free hill" |
| 19:42 | <karlushi> | AryehGregor, that doesn't work very well. People can hide behind email addresses. Having been the first staff contact for the new htmlwg, I can tell you that it's almost impossible to detect |
| 19:42 | <Dashiva> | And for some strange reason, people still paid to get inside |
| 19:42 | <AryehGregor> | karlushi, sure. Now how many members does the HTMLWG have? |
| 19:43 | <AryehGregor> | Quite a few. |
| 19:43 | <AryehGregor> | So a few cheat. |
| 19:43 | <karlushi> | they were more when it started |
| 19:43 | <AryehGregor> | This reminds me of arguments about DRM. |
| 19:44 | <karlushi> | htmlwg is public. And I'm very happy that it is public and open. It has been a long discussion for it to be. |
| 19:44 | <karlushi> | do not shoot the messenger. |
| 19:44 | <karlushi> | I advocate for it. |
| 19:44 | <karlushi> | BUT there is financial reality of a structure too. |
| 19:44 | <AryehGregor> | Maybe the W3C could manage with less money if it cut down on bureaucracy, too. The WHATWG doesn't have to charge membership fees, and it's arguably producing a superior spec to the W3C version of HTML5. |
| 19:45 | karlushi | is not replying to that. |
| 19:45 | <AryehGregor> | :) |
| 19:46 | <Dashiva> | You could argue that the reason WHATWG is so efficient is because W3C serves as a pressure release valve :) |
| 19:46 | <karlushi> | I do not think whatwg is efficient. period. |
| 19:47 | <AryehGregor> | Dashiva, no, people don't do things like argue about process here because there is no process. If they disagree with a decision, they don't argue incessantly because it's pointless, they're just going to get ignored. |
| 19:48 | <AryehGregor> | I can see why some perceive it as unequitable, but I don't see how it could be considered inefficient at its goal -- namely, coordinating and standardizing implementation of new web features. |
| 19:48 | <Dashiva> | AryehGregor: Yes, but you also have an alternate forum more open to your complaints. If that didn't exist, you'd have to choose between silence or trying harder |
| 19:49 | <AryehGregor> | Maybe. But I suspect that kind of behavior wouldn't be tolerated if it started to make the list less useful. A lot like most open-source projects' communications media -- you can make whatever arguments you like until you get in the way. |
| 19:49 | <karlushi> | Shoutocracy |
| 19:50 | <karlushi> | it is very hard to find the right equilibrium |
| 19:50 | <karlushi> | in terms of listening people, and making the project healthy without people derailing it |
| 19:51 | <AryehGregor> | Well, it's impossible to find something that everyone will agree is the right equilibrium. For a given opinion on what constitutes the right equilibrium, it can be pretty easy to find the right one. |
| 19:52 | <karlushi> | whatever the chosen system, there will be benefits and issues. And some projects tend to ignore some voices for more velocity, but at the same time avoiding certain realities. But this is a discussion that philosophers have discussed for ages. I don't think we'll find the solutions today ;) |
| 19:55 | <AryehGregor> | Nope, not likely. Everyone has to do whatever they think will be best. |
| 19:56 | <karlushi> | AryehGregor, first reasonable sentence of this chat. |
| 19:56 | <karlushi> | ;) |
| 19:57 | <AryehGregor> | Unfortunately, some people (including sometimes me) are not as considerate as they could be of other people's opinions on what course is best, and that's exacerbated by the WHATWG/W3C split. But we are where we are. |
| 20:09 | <othermaciej> | look, our little spat made Ajaxian: http://ajaxian.com/archives/adobe-html5-standards-blocking-and-the-evil-of-the-private-backroom |
| 20:12 | <AryehGregor> | Wait, so was the objection that canvas wasn't in scope? |
| 20:12 | <AryehGregor> | I mean, that was the formal objection being referred to? |
| 20:12 | <AryehGregor> | (et al.) |
| 20:15 | <othermaciej> | honestly I am not entirely sure of the content or form of the objection |
| 20:25 | <wycats> | to be honest, I hope that's the argument |
| 20:26 | <wycats> | because it will expose Adobe to serious hurt |
| 20:26 | <AryehGregor> | Well, Larry has made a lot of objections personally that seem unrelated to Adobe's interests. |
| 20:26 | <AryehGregor> | Mostly they seem to be procedural, like this. |
| 20:27 | <AryehGregor> | If the real idea was to try throwing a spanner into canvas standardization to avoid competition with Flash or something, it seems pretty well disguised. |
| 20:27 | <wycats> | AryehGregor: "canvas is out of scope" is not procedural |
| 20:27 | AryehGregor | will continue to assume good faith. |
| 20:27 | <AryehGregor> | I'd call it procedural. It's not saying there's anything *technically* wrong with the spec, just that its existence doesn't conform to some policy somewhere. |
| 20:28 | <wycats> | AryehGregor: it's procedural like saying HCR doesn't conform to reconciliation rules |
| 20:28 | <wycats> | it's TRUE |
| 20:28 | <AryehGregor> | Er, what's HCR? |
| 20:28 | <wycats> | but you don't bring it up unless you want to block it for some other reason |
| 20:28 | <wycats> | health care reform |
| 20:28 | <wycats> | sorry |
| 20:28 | <wycats> | political analogy ;) |
| 20:28 | <AryehGregor> | Oh. |
| 20:28 | <AryehGregor> | . . .What are reconciliation rules? |
| 20:29 | <wycats> | ha |
| 20:29 | <wycats> | there are rules in the Senate that allow you to bypass the 60-vote filibuster rules for items related to the budget |
| 20:29 | <AryehGregor> | Ah. |
| 20:29 | <wycats> | there was some talk of trying to pass healthcare through those rules |
| 20:29 | <wycats> | but too much of it was unrelated to budget |
| 20:29 | <wycats> | but if someone objected on those grounds, they were really objecting to the bill |
| 21:06 | <karlushi> | [proposed] Web Fonts Working Group Charter |
| 21:06 | <karlushi> | http://www.w3.org/2009/08/WebFonts/charter.html |
| 21:11 | <TabAtkins> | karlushi: Man, where did *that* come from? I haven't seen any activity on the www-font list for some time. |
| 21:13 | <TabAtkins> | Looks good, though. Just what we discussed at the planning meeting. |
| 21:14 | <Philip`> | 2009/08? Sounds obsolete already |
| 21:14 | <TabAtkins> | That's definitely a strange date, since we didn't actually pin down any details until 09/11 |
| 21:14 | <TabAtkins> | 2009/11, that is. |
| 21:21 | <karlushi> | Philip`, dated space are identifiers not date. |
| 21:22 | <karlushi> | look at the bottom of the document $Date: 2010/02/11 19:43:34 $ |
| 21:26 | <Dashiva> | Wasn't there talk about finally getting rid of dates in the URLs? |
| 21:27 | <Dashiva> | Since it doesn't actually do anything except confuse people... |
| 21:27 | <Dashiva> | (same with dates in namespaces) |
| 21:27 | <othermaciej> | looks like that Ajaxian article made reddit |
| 21:27 | <othermaciej> | "are we having fun yet?" |
| 21:28 | <karlushi> | Dashiva, what do you replace them with? |
| 21:28 | <Philip`> | Nothing |
| 21:28 | <Philip`> | Just choose names that are unique within all names forever, rather than names that only have to be unique within all names chosen in that month |
| 21:28 | <Dashiva> | Or a generic group name, e.g. /ns/ for namespaces |
| 21:29 | <karlushi> | Philip`, http://w3.org/Nothing you mean ? |
| 21:29 | <wycats> | othermaciej: of course |
| 21:29 | <karlushi> | Philip`, on a long term no name is forever |
| 21:29 | <Philip`> | http://www.w3.org/WebFonts/charter.html etc |
| 21:30 | <karlushi> | then in 2014, a new WebFonts WG. what do you do you had on top ? you call it WebFonts2/charter or Webfonts/charters2 ? |
| 21:30 | <Philip`> | You don't call it the WebFonts WG |
| 21:31 | <karlushi> | hehe |
| 21:31 | <karlushi> | ok |
| 21:31 | <karlushi> | next one is WWWFonts |
| 21:31 | <karlushi> | then next one is |
| 21:31 | <karlushi> | WWWPolice |
| 21:31 | <karlushi> | and then… :) |
| 21:32 | <karlushi> | btw, http://lists.w3.org/Archives/Public/www-archive/2010Feb/0002 |
| 21:36 | <Dashiva> | Or you just reuse the name, and say that the group has had several incarnations |
| 21:36 | <Dashiva> | Sort of like /html works today |
| 21:43 | <karlushi> | html didn't reuse the name |
| 21:43 | <karlushi> | it was previously http://www.w3.org/MarkUp/ ;) |
| 21:46 | <Philip`> | A demonstration of how easy it is to find non-conflicting names that make sense :-) |
| 21:47 | <Dashiva> | karlushi: Well, the current /html manages the multi-capacity role just fine |
| 21:51 | Philip` | discovers that const char* foo = "<bar baz=""/>"; doesn't do quite what he expected, even though the compiler is perfectly happy with it |
| 21:52 | <othermaciej> | Philip`: it's a feature! |
| 21:53 | <jgraham> | me wonders what that does |
| 21:53 | <ment> | it really is :) |
| 21:53 | <jgraham> | s//// |
| 21:54 | <ment> | 0 == strcmp("a" "b", "ab") |
| 21:54 | <othermaciej> | two string literals that are adjacent to each other get concatenated |
| 21:54 | <othermaciej> | it's useful because of macros |
| 21:54 | <othermaciej> | but it is kind of evil |
| 21:58 | <Philip`> | It's useful for splitting long strings over multiple lines too |
| 22:04 | <Dashiva> | I'd think a line-terminating \ is a lesser evil there, or a heredoc syntax |
| 22:05 | <Lachy> | The line-terminating \ should, ideally, be unnecessary. It's really inconvenient |
| 22:06 | <Lachy> | but I think the only reason it exists is to ensure the author really meant for the string to continue, rather than an accidentally omitted quote |
| 22:06 | <Dashiva> | Makes error messages nicer :) |
| 22:06 | <Dashiva> | But I agree it's horrible to use |
| 22:08 | <Philip`> | Maybe it'd be better to have a compile-time-string-concatenation operator |
| 22:12 | <Hixie> | which "big content companies" are involved in public-html but not whatwg? |
| 23:29 | <rektide> | cross document messaging is only for nested documents ? |
| 23:29 | <rektide> | it all seems to hinge around having a window to address |
| 23:30 | <rektide> | i'm trying to come up with a solution for http://stackoverflow.com/questions/2236828/javascript-getting-two-sibling-windows-communicate |
