| 10:37 | <annevk> | hey |
| 10:38 | <annevk> | what's would be a good license? |
| 10:38 | <annevk> | for xml5? |
| 10:38 | <virtuelv> | annevk: the parser, I presume? |
| 10:38 | <virtuelv> | tokenizer* |
| 10:39 | <annevk> | the whole project as it stands now :) |
| 10:40 | <annevk> | I'll go with the Apache 2.0 for now |
| 10:40 | <annevk> | and maybe make it MIT later or something |
| 10:54 | <hsivonen> | annevk: Apache 2.0 is otherwise good except it isn't GPLv2-compatible |
| 11:11 | <annevk> | hmm ok |
| 11:34 | <annevk> | created http://code.google.com/p/xml5/ |
| 11:34 | <annevk> | took some time to get all the SVN to play nicely :( |
| 11:41 | <MikeSmith> | question: Other than annevk, hsivonen, Hixie, Lachy, and WHATWG blog, is there anybody else who's blogging semi-regularly about the the topic of HTML5? |
| 11:42 | <MikeSmith> | (I ask because I'm setting up a planet aggregator and want to know what ought to aggregate) |
| 11:44 | <hsivonen> | MikeSmith: should I create an HTML5-specific feed for my site? |
| 11:57 | <hendry> | http://natalian.org/archives/2007/05/26/mobile-web-3/ # i'm having fun with the .mobi guys |
| 12:01 | <hsivonen> | Hixie: do you know how many bytes Gecko usually inspects using chardet? it seems to inspect the first buffer that is passed to the parser but I have no idea of the size of the buffer |
| 12:17 | <hsivonen> | hendry: apparently there are still mobile folks who believe supporting mobiles means catering to crappy browsers instead of making sure that Presto or WebKit or Gecko runs on the device |
| 12:22 | <hsivonen> | hendry: at XTech there was still one guy talking about "Mobile Ajax" on Pocket IE |
| 12:29 | <MikeSmith> | hendry - question: I wonder if you asked James Pearce whether it would be be OK to publish his private e-mail to you on your weblog |
| 12:32 | <hendry> | MikeSmith: I can't recall. This was a while back. |
| 18:37 | <gsnedders> | how does IE6's relative URL resolution work? scheme/authority/path+query/fragment? |
| 18:42 | <gsnedders> | and can anyone from Apple say whether Safari pays any attention to <ttl> in RSS? |
| 19:32 | <MikeSmith> | I mentioned this over on #public-html, but will give a heads-up about it here too - |
| 19:32 | <MikeSmith> | I've set up beginnings of a simple "Planet HTML5" aggregator - |
| 19:32 | <MikeSmith> | http://people.w3.org/mike/planet/html5/ |
| 19:32 | <MikeSmith> | if anybody else on the channel has been blogging about (or planning to blog about) HTML5/WHATWG/HTMLWG and wants me to add your feed to the aggregator, just let me know |
| 19:33 | <MikeSmith> | either just paste in a URL here or /msg me or e-mail me at mike⊙wo |
| 19:34 | <MikeSmith> | part of the intent of setting up the aggregator is for people are not HTML WG members and/or who may not have time to read and follow public-html discussions |
| 23:16 | <SimonW> | does anyone know the reason that HTTP doesn't include a way of specifying the character set used in a POST? |
| 23:16 | <SimonW> | Or does it have a method that I don't know about? |
| 23:17 | <zcorpan_> | SimonW: you mean character encoding? |
| 23:17 | <SimonW> | yes, character encoding |
| 23:17 | <SimonW> | in fact, I can't find the bit of the spec that says "browsers shall submit forms using the same character encoding as the form was served up in" |
| 23:17 | <SimonW> | is that HTML spec instead? |
| 23:17 | SimonW | is pretty bad at specs |
| 23:18 | <othermaciej> | SimonW: you can't specify Content-Type on a request? |
| 23:18 | <othermaciej> | SimonW: I'd imagine Content-Type w/ a charset parameter would be it, but perhaps that is only allowed as a response header |
| 23:19 | <SimonW> | othermaciej: as far as I can tell you can't |
| 23:19 | <SimonW> | http://www.w3.org/Protocols/rfc2616/rfc2616-sec5.html#sec5 |
| 23:19 | <SimonW> | you can say "Accept-Charset" |
| 23:19 | <SimonW> | common knowledge appears to be that you can't either |
| 23:19 | <SimonW> | which is ludicrous - it means that server side scripts have to either guess the character set |
| 23:20 | <SimonW> | or use statistical analysis to figure it out |
| 23:20 | <SimonW> | or just apply a heuristic: "try to decode as UTF 8, if that fails assume ISO-885-1" |
| 23:20 | <SimonW> | 8859-1 |
| 23:20 | <SimonW> | so I thought I'd pop in here and check my facts before ranting about it on my blag |
| 23:21 | <SimonW> | unless there's a more appropriate forum? |
| 23:22 | <zcorpan_> | dunno. i know that trackbacks usually don't leave hints about encoding, and come in different encodings |
| 23:23 | <webben> | Do you mean the transfer encoding or the content encoding? |
| 23:23 | <zcorpan_> | webben: content encoding |
| 23:24 | <SimonW> | content encoding - utf-8 v.s. iso-8859-1 etc |
| 23:24 | <SimonW> | it appears to be a huge gaping hole in HTTP which everyone has been ignoring for the past 15 years |
| 23:25 | <zcorpan_> | then we have anothing thing to fix for HTTP5... ;) |
| 23:25 | <webben> | Weird at http://www.w3.org/Protocols/rfc2616/rfc2616-sec14.html#sec14.11 |
| 23:25 | <webben> | it says: "If the content-coding of an entity in a request message is not acceptable to the origin server, the server SHOULD respond with a status code of 415 (Unsupported Media Type)." |
| 23:26 | <webben> | which would seem to imply that request messages must be able to specify content-coding. |
| 23:26 | <zcorpan_> | webben: yeah, but that's about gzip &c, not utf-8 &c |
| 23:27 | <zcorpan_> | a different level of "encoding" |
| 23:27 | <webben> | oh, a third level |
| 23:28 | <zcorpan_> | character encoding (utf-8) -> content encoding (gzip) -> transfer encoding (7bit) |
| 23:29 | <SimonW> | aah, so I care about character encodin |
| 23:29 | <zcorpan_> | SimonW: yeah |
| 23:29 | <SimonW> | I'm going to e-mail Roy Fielding |
| 23:33 | <othermaciej> | SimonW: I guess that's a bug in HTTP, though de facto you have to post in the page encoding and the server assumes that |
| 23:33 | <SimonW> | I've been trying to find where that de facto rule is written down |
| 23:33 | <othermaciej> | SimonW: I don't think it is |
| 23:34 | <othermaciej> | if you mean in the form of documentation or normative reference |
| 23:34 | <zcorpan_> | SimonW: a de facto rule is that iso-8859-1 pages will submit form content as windows-1252 |
| 23:34 | <othermaciej> | there are no iso-8859-1 pages on the web |
| 23:35 | <othermaciej> | (the ones that claim to be are windows-1252) |
| 23:35 | <zcorpan_> | right |
| 23:35 | <Dashiva> | There was that "bug" earlier about webpage not "supporting" pure ASCII |
| 23:35 | <SimonW> | aha, http://www.w3.org/Protocols/rfc2616/rfc2616-sec3.html#sec3.7.1 looks relevant |
| 23:37 | <zcorpan_> | annevk: hey |
| 23:38 | <zcorpan_> | annevk: found a blog software written in turbogears that i wanted to try out |
| 23:39 | <SimonW> | pah, turbogears :P |
| 23:39 | <annevk> | zcorpan_, don't talk about that with SimonW around :p |
| 23:39 | <zcorpan_> | why not? |
| 23:39 | annevk | is kidding |
| 23:39 | <zcorpan_> | SimonW: what's wrong with turbogears? |
| 23:39 | <SimonW> | it's not Django :) |
| 23:40 | SimonW | co-founded Django |
| 23:40 | <SimonW> | on a more practical note, it seems to be being eclipsed by Pylons these days |
| 23:40 | <SimonW> | the TurboGears team made a bunch of decisions at the start which were very sane then but haven't really held out over the long run |
| 23:41 | <SimonW> | their choices for the core components have mostly been replaced by better alternatives |
| 23:41 | <othermaciej> | just rename Django to Python in Pails or something |
| 23:41 | <SimonW> | which seems to have left the project in an awkward position |
| 23:41 | <SimonW> | personally I'm hoping TurboGears and Pylons will merge, and Django and Pylons will work towards a common API for core features |
| 23:42 | <Dashiva> | Pythons on Planes |
| 23:42 | <othermaciej> | heh :-) |
| 23:59 | <annevk> | SimonW, first step of http://www.whatwg.org/specs/web-forms/current-work/#x-www-form-urlencoded defines what you're looking for |