| 00:30 | <Lachy> | Now that we have the ability to generate a copy of the spec with static annotations, I wonder what it would take to combine that with the dynamic annotations in the WHATWG copy so that the script only needs to check for changes in the annotations since the last commit |
| 00:30 | <Lachy> | that would significantly reduce load time on the spec |
| 00:30 | <Hixie> | it would actually increase load time for the spec itself |
| 00:30 | <Lachy> | the bottle neck is the javascript performance, not the download time |
| 00:31 | <Hixie> | on another note, maybe we should have people pass a test to prove that they have read the html5 spec before they're allowed to propose text- or process- level changes to the spec or wg. |
| 00:31 | <Hixie> | Lachy: in your browser maybe :-) |
| 00:31 | <Hixie> | Lachy: in safari, the JS runs nice and fast |
| 00:32 | <Hixie> | crap i marked the last checkin editorial but it affected html5 parser implementations |
| 00:32 | <Lachy> | I don't like using safari for reading the spec because it lacks essential features I need for reading it efficiently |
| 00:32 | <othermaciej> | Hixie: that would seriously cut down on the number of proposals, if enforced |
| 00:33 | <Dashiva> | Hixie: What kind of test, given that people are unlikely to retain even a small part of the whole spec? |
| 00:33 | <Hixie> | Dashiva: open book |
| 00:34 | <Hixie> | Lachy: like what? |
| 00:34 | Hixie | wonders what he's missing! |
| 00:34 | <Dashiva> | So more like a "Read these sections" in practice then :) |
| 00:34 | <Lachy> | like Find As You Type |
| 00:34 | <Lachy> | without having to press cmd+F or '/' to start searching |
| 00:35 | <Hixie> | Dashiva: no, there are plenty of questions that require a clear understanding of many partso of the spec to answer. |
| 00:36 | <Lachy> | only Firefox has that feature. My repeated attempts to convince people that we need it Opera have failed :-( |
| 00:37 | <Lachy> | and Safari's find feature has an unacceptable delay |
| 00:38 | <Dashiva> | Lachy: putting a finger on a single key is too much? |
| 00:39 | <Dashiva> | I find it more than acceptable in return for being able to have single key shortcuts |
| 00:39 | <Lachy> | Dashiva, I hate single key shortcuts, and yes, being forced to press a single key I never remember to press is annoying |
| 00:45 | <Lachy> | Hixie, if I find some time this weekend, I'll see if I can take a look at the annotations script and merge it with the static version |
| 00:45 | <Hixie> | cool |
| 00:47 | <Lachy> | Hixie, how does status.cgi work? Is it backed by a database or static files? |
| 00:48 | <Hixie> | backed by mysql |
| 00:48 | <Lachy> | would it be possible to get a copy of that database, status.cgi and any other files used for the annotations? |
| 00:50 | <Lachy> | last time I tried to do anything with the annotations script, my attempts failed because I couldn't substitue status.cgi with static files and so the script wouldn't run properly |
| 00:52 | <Lachy> | oh, maybe I don't need to. I can just set up a local server side script that passes off all requests for status.cgi to your copy, thus bypassing the cross-origin restrictions |
| 00:53 | <Hixie> | http://damowmow.com/temp/status.tar.gz |
| 00:54 | <Lachy> | thanks |
| 01:05 | <Hixie> | can anyone remember why we block manifests to only contain URLs with the same scheme? |
| 07:33 | <othermaciej> | sorry about all the public-html spam |
| 07:34 | <othermaciej> | I was just following orders |
| 07:35 | <annevk42> | I was just about to say |
| 07:35 | <annevk42> | you dominate my inbox |
| 07:49 | <Hixie> | i like how manu decided that microdata -- which was introduced 3 months ago -- was controversial under his 4 month rule |
| 07:49 | <Hixie> | but let it not be suggested that he's just doing this because he wants to bring attention to rdfa! |
| 07:49 | <othermaciej> | annevk42: DanC and rubys very specifically asked me to send a fresh email per issue |
| 07:49 | <othermaciej> | Hixie: that's why I asked him for citatations... |
| 07:50 | <othermaciej> | I like how he didn't want to change his draft in the middle of the poll until it started looking like he was going to lose |
| 11:38 | <hsivonen> | <object> loading code look "fun" |
| 11:38 | <hsivonen> | *looks |
| 13:08 | <annevk42> | feedback so far on charsets |
| 13:10 | <annevk42> | giving Windows-31J a Windows-932 alias can be done, giving encodings an alias that starts with "x-" cannot be done and registering treating things such as ISO-8859-1 -> windows-1252 is considered unwise |
| 13:17 | <hsivonen> | annevk42: why unwise? |
| 13:17 | <hsivonen> | annevk42: I can understand the x- stuff on the level of principle but not on the level of usefulness |
| 13:19 | <annevk42> | not sure, I asked some follow-up questions |
| 13:19 | <annevk42> | this is the feedback in email form: http://mail.apps.ietf.org/ietf/charsets/msg01833.html |
| 13:20 | <hsivonen> | are people document.writing their flash objects in order to get around some Eolas-non-infringement silliness in IE? |
| 13:20 | <hsivonen> | or do people just love to document.write stuff? |
| 13:21 | <gsnedders|work> | hsivonen: Most offline applications do treat ISO-8859-1 as ISO-8859-1 |
| 13:22 | hsivonen | mumbles about *Internet* Assigned Name Authority vs. "offline" |
| 13:23 | <Lachy_> | hsivonen, yeah, the Eolas patent is a large part of the reason, and even though Microsoft have, I think, since entered into a licencing agreement, that problem doesn't exists as much. However, Opera still suffers from the issue cause we don't pay the licencing fees |
| 13:24 | <hsivonen> | Lachy_: Opera has code for working around Eolas? |
| 13:25 | <hsivonen> | gsnedders|work: I can see how it would suck if Java or Python APIs magically changed their meaning of ISO-8859-1 suddenly |
| 13:25 | <hsivonen> | it would still be good to have a Web-aware new decoder getter |
| 13:25 | <gsnedders|work> | hsivonen: Surely that's the inevitable result of getting them defined as aliases, though? |
| 13:25 | <Lachy_> | hsivonen, AIUI, we implement the same document.write workaround that IE did |
| 13:26 | <hsivonen> | Lachy_: I see. That's news to me. |
| 13:26 | <Lachy_> | but for flash embeded without document.write, we still require the click to activate nonsense |
| 13:43 | <krijnh> | Missing logs for 11th and 12th fixed, thanks Lachy_ |
| 13:44 | <hsivonen> | oh noes! all the secret cabal talk leaked. :-) |
| 13:44 | <hsivonen> | krijnh: thanks! |
| 13:44 | <krijnh> | Indeed :) |
| 13:44 | <krijnh> | I've only stripped some quit/join messages |
| 13:46 | <Lachy_> | krijnh, there may still be some missing. I think there were a few hours during which I wasn't connected |
| 13:46 | <Philip`> | Did you strip the bit where you admitted you were going to strip some secret discussions and then strip your admission and then deny all knowledge of it? |
| 13:46 | <krijnh> | Yeah |
| 13:46 | <krijnh> | But who cares |
| 13:46 | <Lachy_> | so if someone wants to fill in those hours between around midnight and 08:00, give or take a few hours, feel free to do so |
| 13:46 | <krijnh> | Philip`: I already said that two minutes ago, see the logs |
| 13:46 | <krijnh> | :) |
| 13:47 | <krijnh> | And in the meantime, don't flag any lines |
| 13:51 | <beowulf> | krijnh: i have logs if you need them |
| 13:52 | <krijnh> | I don't need them, do you think others do? |
| 13:53 | <beowulf> | iurro |
| 13:56 | <krijnh> | "<rubys2> I miss krijnhoetmer.nl" - a new one for the subtitles, yay |
| 13:57 | <krijnh> | Btw, the computer I'm logging from is stable, it's just that power inage isn't |
| 13:58 | <Lachy_> | krijnh, you need a UPS |
| 13:58 | <krijnh> | Nah |
| 13:58 | <krijnh> | I like being missed :> |
| 13:59 | gsnedders|work | hugs krijnh |
| 13:59 | <krijnh> | Also, when a bus drives by, sometimes the LAN cable (with only 4 wires, because the other 4 are used for phones) becomes unstable |
| 13:59 | <krijnh> | I don't think a UPS would solve that |
| 14:00 | <krijnh> | I you thought Dean Edwards hosting his stuff from his kitchen was weird, you should come take a look at my configuration :) |
| 14:00 | <krijnh> | *If |
| 14:01 | <krijnh> | And thanks for the hugs gsnedders|work! |
| 14:02 | <annevk42> | encodings are such a mess |
| 14:02 | <Philip`> | krijnh: You should build yourself a datacenter |
| 14:02 | <jgraham> | krijnh: Perhaps if your configuration is that weird you should post photos of it on slashdot |
| 14:02 | <Lachy> | we should find a more stable server to run IRC logging on |
| 14:03 | <Lachy> | I may have a server available, I just need to check the provider's policy about running IRC |
| 14:03 | <jgraham> | We should run logs on multiple independent machines and then have a cople of servers that pull logs from each logging node, merge the logs, and publish them |
| 14:03 | Philip` | logs IRC on a (fairly) stable server, but has no way of viewing the logs other than cat |
| 14:03 | <jgraham> | Redundancy is they key on such mission-crtitical applications |
| 14:03 | <krijnh> | jgraham: it's already producing smoke when Sam posts some links to it from his site :) |
| 14:03 | <Philip`> | krijnh: Sounds like you need caching :-) |
| 14:04 | <jgraham> | krijnh: Oh the /. suggestion was just to increase your downtime and/or set fire to your house |
| 14:04 | <krijnh> | That's three times 'you need' I don't really agree with :) |
| 14:05 | <krijnh> | Let's just not make HTML 5 a failure, so people don't need the logs anymore in a few years, okay? |
| 14:05 | <Lachy> | krijnh, you really need things you don't agree with |
| 14:05 | <Philip`> | krijnh: Also, you need to transfer the entire contents of your bank account to me |
| 14:06 | <krijnh> | (I think I once started making a backup on http://krijn.html5.org/irc-logs/) |
| 14:06 | <jgraham> | krijnh: I suggest you get a bank with a large overdraft, remove all the money to the overdraft limit as cash and then transfer the remaining debt to Philip` |
| 14:09 | <Philip`> | I'm afraid that's missing the point :-( |
| 14:12 | <Lachy> | looks like I can run a persistent process on my server, like IRC. I just need to find out what software to use to do it |
| 14:12 | <Philip`> | irssi! |
| 14:12 | <Lachy> | krijnh, could I get a copy of your logging software, so I can set up a mirror? |
| 14:12 | <annevk42> | hmm, krijnh is just hosting his IRC logs |
| 14:13 | <Lachy> | annevk42, yeah, but he has some nice scripts that format it as HTML that I don't want to rewrite if I don't have to |
| 14:13 | <krijnh> | Lachy: the logging software is just mIRC |
| 14:14 | <krijnh> | You don't want to receive my crappy scripts either :) |
| 14:14 | <krijnh> | Or put differently, I don't want to share my crappy scripts with the outside world, cause then nobody will ever take me serious anymore |
| 14:15 | <krijnh> | Self-handicapping ftw |
| 14:15 | <Lachy> | fine. I will have to find or write my own alternative. |
| 14:15 | <Lachy> | for now, I will try to figure out how I can install irissi on my server |
| 14:15 | <Philip`> | "apt-get install irssi" |
| 14:15 | <Philip`> | then run it in screen |
| 14:16 | <Philip`> | then configure it a tiny bit |
| 14:16 | <Lachy> | I have no experience using screen. How does that work? |
| 14:16 | <Lachy> | the apt-get command is not available |
| 14:17 | <Philip`> | Substitute the appropriate package manager for your Linux distro |
| 14:17 | <krijnh> | But I plan on rewriting some stuff |
| 14:17 | <krijnh> | And using Bens suggestions for styling as well |
| 14:17 | <Philip`> | You type "screen", and it runs a shell, and you run stuff in there; then you can press ctrl+A D to detach the screen, and "screen -r" to reattach to it later (possibly from a totally new ssh session or whatever) |
| 14:18 | <hsivonen> | you'll want to add -U to those to force UTF-8 sanity |
| 14:19 | <hsivonen> | unless your copy of screen is configured to be sane-by-default |
| 14:19 | <Lachy> | Philip`, I have no idea what linux distro is being used |
| 14:20 | <Philip`> | It seems to work for me without -U, with some combination of LANG=something.UTF-8 (both outside and inside screen) and suitable irssi configuration |
| 14:20 | <Philip`> | Lachy: Try 'cat /etc/lsb-release' perhaps |
| 14:20 | <Lachy> | # screen |
| 14:20 | <Lachy> | Cannot access '/dev/pts/1': No such file or directory |
| 14:21 | <Lachy> | # cat /etc/lsb-release |
| 14:21 | <Lachy> | cat: /etc/lsb-release: No such file or directory |
| 14:21 | <beowulf> | Lachy: if you are on dreamhost it's debian and there's a pre built irssi on the web somewhere |
| 14:21 | <Philip`> | Do you have root access to the server, or is it a silly share hosting thing? |
| 14:21 | <Lachy> | beowulf, no, this is asmallorange.com |
| 14:21 | <Philip`> | *shared |
| 14:21 | <Lachy> | it's a shared hosting thing |
| 14:21 | <Lachy> | so i might be out of luck |
| 14:21 | <Philip`> | Oh, package managers won't work then |
| 14:21 | <Philip`> | and you'd have to install stuff manually |
| 14:21 | <Philip`> | Does 'gcc' exist on there? |
| 14:22 | <krijnh> | Why can't mIRC just log to a remote server :/ |
| 14:22 | <krijnh> | That would solve a lot |
| 14:22 | <Philip`> | krijnh: I guess it can, if you set it to use a network drive |
| 14:23 | <krijnh> | True |
| 14:23 | <Philip`> | (which could be attached to Samba on a real OS) |
| 14:23 | <krijnh> | :) |
| 14:25 | <Philip`> | If someone has scripts to display IRC logs, I could probably run them on my server |
| 14:26 | <Lachy> | gcc is there |
| 14:26 | <Philip`> | Ah, so it's not totally hopeless |
| 14:27 | <Philip`> | If irssi isn't already installed you could probably just do ./configure && make to set it up without much pain |
| 14:27 | <jgraham> | You can install stuff like that on Dreamhost. At least I managed to build python |
| 14:27 | <Philip`> | You might be a bit stuck if screen doesn't work because it can't access the /dev stuff, though |
| 14:28 | <gsnedders|work> | Dreamhost TOC forbid IRC clients/bots |
| 14:28 | <Lachy> | that's why I'm not trying this on dreamhsot |
| 14:28 | <beowulf> | they are frowned upon, not forbidden, afaik |
| 14:28 | <Lachy> | but anyway, I wonder what it would cost to get a non-shared server for running all of my stuff on |
| 14:28 | <gsnedders|work> | beowulf: When I looked it up, it was forbidden. This was, however, around a year ago, though |
| 14:28 | <beowulf> | the irssi process gets killed every now and then |
| 14:29 | <Lachy> | http://wiki.dreamhost.com/KB_/_Unix_/_Cron_Jobs_and_Persistent_Processes#What_is_your_persistent_.28background.29_process_policy.3F |
| 14:29 | jgraham | could in principle run it but isn't actually going to |
| 14:29 | <Lachy> | "IRC-related persistent processes of any kind (including, but not limited to, bots, bouncers, etc.) are STRICTLY PROHIBITED, and are in violation of our TOS. " |
| 14:29 | <hsivonen> | not such a Dream after all :-( |
| 14:30 | <beowulf> | http://wiki.dreamhost.com/Irssi |
| 14:31 | <gsnedders|work> | Lachy: non-shared meaning dedicated? Or do you just mean VPS? |
| 14:31 | <Lachy> | gsnedders|work, not sure. Whatever will let me run anything I like without restriction |
| 14:31 | <gsnedders|work> | Lachy: http://www.cheapvps.co.uk/plans has unmanaged VPS from £5.00/month and is apparently good |
| 14:33 | <gsnedders|work> | http://www.intovps.com/ is apparently good too |
| 14:33 | <gsnedders|work> | http://slicehost.com is meant to be very good, but twice as much (i.e., $20/month) |
| 14:34 | <hsivonen> | what's the deal with http://www.w3.org/TR/2004/REC-DOM-Level-3-Core-20040407/namespaces-algorithms.html#lookupNamespaceURIAlgo not special casing "xml"? |
| 14:35 | <Lachy> | $120/year isn't too bad. Not sure if it's worth it though, as I would currently only need it for running an IRC logger, which isn't essential |
| 14:35 | <gsnedders|work> | I'll probably be getting an account on the 5GBP/10USD (the 5GBP being cheaper…) VPS in Sept. |
| 14:36 | <krijnh> | There's not really a problem with how it's currently done, right? |
| 14:36 | <Lachy> | krijnh, just the stability problem |
| 14:36 | <krijnh> | (Apart from my sarcasm wrt smoking computers) |
| 14:36 | <gsnedders|work> | krijnh: Apart from it going down half the time :P |
| 14:36 | <Philip`> | I think 128MB RAM isn't really enough if you want to run Apache and MySQL or anything like that |
| 14:37 | gsnedders|work | points out nobody ever visits his blog |
| 14:37 | jgraham | notes that WebFaction isn't a VPS but does seem to be pretty good |
| 14:37 | <krijnh> | A PII 450 mhz with 512MB is :) |
| 14:37 | <Philip`> | gsnedders|work: Processes take up RAM even with no visitors :-) |
| 14:38 | hsivonen | wonders if the lack of 'xml' and 'xmlns' special casing in the above algorithm is an error or by design |
| 14:38 | <Philip`> | and then one person loads your site and you get six Apache processes starting up |
| 14:38 | <gsnedders|work> | Philip`: I don't need many workers waiting, though :) |
| 14:38 | <Philip`> | if there's lots of stuff downloading in parallel |
| 14:38 | <hsivonen> | in other news, I'm in the Namespace drudgery land again |
| 14:38 | <hsivonen> | RDF Tax FTW! |
| 14:38 | <gsnedders|work> | w00t! namespaces! |
| 14:39 | gsnedders|work | still doesn't get how the tax is specific to RDF and not all XML with Namespaces |
| 14:39 | <Philip`> | gsnedders|work: ~10MB times not many is still a lot if you've only got 128MB in total (including for all the rest of the OS) :-p |
| 14:39 | <Lachy> | hsivonen, care to elaborate? |
| 14:39 | <hsivonen> | gsnedders|work: XML has Namespace because of RDF |
| 14:40 | <Lachy> | really? |
| 14:40 | <hsivonen> | Lachy: lookupNamespaceURI in Gecko trunk doesn't have interop with WebKit when applied to HTML DOMs |
| 14:40 | <hsivonen> | Lachy: yes |
| 14:41 | <gsnedders|work> | Philip`: If there's a will there's a way :P |
| 14:41 | <jgraham> | gsnedders|work: Demonstrably false |
| 14:41 | <Lachy> | so the designers of namespaces had RDF in mind as they were creating it, and not just the idea of creating multiple vocabularies that don't conflict? |
| 14:41 | <gsnedders|work> | jgraham: sssh |
| 14:42 | Philip` | uses http://www.gandi.net/hosting/ which seems about the same price as CheapVPS for 256MB RAM (~£12/month incl VAT) |
| 14:42 | <gsnedders|work> | Philip`: And who visits your site? :P |
| 14:42 | <gsnedders|work> | (Apart from that on Anne's hosting!) |
| 14:43 | <Philip`> | gsnedders|work: It's had 1.2TB of traffic in the past month :-p |
| 14:43 | <hsivonen> | http://lists.w3.org/Archives/Public/semantic-web/2007Dec/0116.html |
| 14:43 | <gsnedders|work> | Philip`: Dude, how much pr0n is that? :P |
| 14:44 | <Philip`> | gsnedders|work: sssh |
| 14:44 | <Philip`> | (Actually it's mostly SVN) |
| 14:44 | <gsnedders|work> | Binary files in SVN!? :\ |
| 14:44 | <Philip`> | Yes |
| 14:45 | <gsnedders|work> | Also: SVN!? :\ |
| 14:45 | <Philip`> | The repository is about 6GB, I think |
| 14:45 | <Philip`> | which really wouldn't work with DVCSes |
| 14:47 | <jgraham> | Philip`: Why? |
| 14:49 | <Philip`> | jgraham: Why what? |
| 14:50 | <jgraham> | Philip`: Why wouldn't having a 6Gb repo work with DVCS? |
| 14:50 | <Philip`> | jgraham: Because everyone who wants to check it out would have to download the 6GB repository, instead of the ~300MB subtree revision that they're usually interested in |
| 14:51 | <jgraham> | Philip`: http://git.or.cz/gitwiki/SubprojectSupport |
| 14:51 | <jgraham> | Sounds like it solves that problem |
| 14:52 | <jgraham> | Unless I misunderstood the problem |
| 14:52 | <gsnedders|work> | Why have everything in one repo is nobody cares about parts of it normally? |
| 14:53 | <annevk42> | hsivonen, RDF has brought us a great many wonderful thing |
| 14:53 | <annevk42> | s |
| 14:53 | <jgraham> | (I assume the problem is roughly "there is a whole load of artwork which people don't need to check out often and source code that they do") |
| 14:54 | <Philip`> | "This implementation has not been completed due to these problems. A newer implementation is in the works ..." - doesn't sound like the problem is actually solved |
| 14:55 | <jgraham> | Oh well I heard that Git had soloved these problems. Maybe the new thing is fixed |
| 14:55 | <Philip`> | (and it certainly wasn't solved several years ago when we started using SVN) |
| 14:55 | <Philip`> | (and migration would be a pain, even if it the new system was better) |
| 14:56 | <Philip`> | jgraham: It's approximately that problem, although there's some artwork which people do check out (in order to run the game) and lot more which they don't (which is mostly source material for artists, and isn't publicly visible at all because its copyright status is often questionable) |
| 14:58 | gsnedders|work | wonders what game |
| 14:58 | <Philip`> | Also there's lots of versions of precompiled executables in SVN, which is fine in SVN because everyone just downloads the latest revision and it's got everything they need to run the game, and we wouldn't want to use a DVCS where everyone has to download all historic revisions of those files too |
| 14:59 | <Philip`> | (and if we don't include the executables in the VCS then it's more of a pain for anyone who wants to run it) |
| 14:59 | <Philip`> | gsnedders|work: http://os.wildfiregames.com/ |
| 15:00 | <annevk42> | next suggestion from the encoding front: do your own registry |
| 15:00 | <annevk42> | I kind of like that solution to be honest |
| 15:01 | <annevk42> | that way we can just specify what is required for "web encodings" lock it down and be done with it |
| 15:01 | <gsnedders|work> | Philip`: Is it fun? |
| 15:04 | <Philip`> | gsnedders|work: Not as a game to actually play, because it's nowhere near finished |
| 15:17 | <annevk42> | http://www.rockmelt.com/ -- lots of talk, no walk? |
| 15:18 | <hsivonen> | annevk42: what is it about? |
| 15:18 | <hsivonen> | I don't even noticed any talk :-) |
| 15:18 | <annevk42> | supposedly some new browser backed by Andreessen |
| 15:19 | <hsivonen> | which engine? |
| 15:19 | <jgraham> | It may or may not have some social networking features like facebook integration |
| 15:20 | <hsivonen> | jgraham: wasn't Flock supposed to be that browser? |
| 15:20 | jgraham | bets on a WebKit backed browser with similar ideas to flock but more usable |
| 15:20 | <jgraham> | AFAICT Flock is a usability disaster |
| 15:22 | <hsivonen> | hmm. so if a telecon isn't recorded, WCAG doesn't require it to be accessible, but if it is recorded, WCAG requires the recording to be accessible |
| 15:22 | <hsivonen> | did I understand WCAG right? |
| 15:22 | <hsivonen> | WCAG 2 that is |
| 15:23 | jgraham | assumes that the minutes would count as an accessible version albeit with some information loss |
| 15:23 | <hsivonen> | the entire premise here is that the minutes aren't really equivalent to the audio data |
| 15:24 | <annevk42> | hsivonen, could be true, I believe it's also that if you publish a video rather than keep it to yourself it needs to be accessible |
| 15:24 | <jgraham> | In any case I am stunned that Laura thinks the situation shouldn't be improved for some people if it can't be improved for anyone |
| 15:24 | <annevk42> | hsivonen, see also the Flickr debate |
| 15:24 | <jgraham> | s/everyone/anyone/ |
| 15:24 | <annevk42> | jgraham, she hasn't said that |
| 15:25 | <jgraham> | annevk42: Well true. But she did bring it up as a "social/policy question" |
| 15:26 | <jgraham> | Which kind of implies that she wouldn't see audio+minutes as an improvement over minutes alone |
| 15:37 | <hsivonen> | does Opera implement lookupNamespaceURI per spec? |
| 15:40 | <hsivonen> | I just realized that the spec design badness sicking mentioned in https://bugzilla.mozilla.org/show_bug.cgi?id=505178#c10 is yet another reason why the DOM specs make dynamic CURIE-using content suck even more than one would expect |
| 15:47 | <gsnedders|work> | Hixie: http://www.hixie.ch/tests/adhoc/css/text/white-space/normal/003.html — learn to spell normal in the CSS! |
| 15:50 | <hsivonen> | aaargh. there's document.createNSResolver |
| 15:50 | <hsivonen> | grr. |
| 15:55 | <hsivonen> | whew. maybe createNSResolver isn't as crazy as I thought |
| 15:58 | <Lachy> | if we have audio recordings, then in theory, someone could take the time revise the minutes to be more accurate than they are when they're done live |
| 16:00 | <annevk42> | yup |
| 16:09 | <annevk42> | http://mail.apps.ietf.org/ietf/charsets/msg01781.html is interesting |
| 16:09 | <annevk42> | maybe we should just standardize the IE mapping |
| 16:50 | <annevk42> | hmm, old guard doesn't want recordings :/ |
| 16:52 | <jgraham> | It seems to me that a big advantage of recordings is that it allows you to check how something was presented in a telecon e.g. you can see if things were (accidentially) misrepresented |
| 16:53 | <jgraham> | Not to mention that it expands the scope of people to whom explainations are avaliable |
| 16:53 | <Lachy> | jgraham, I agree. The current system relies on people reviewing the minutes for themselves to see if what they said was accurately represented |
| 16:54 | <jgraham> | Since rubys seems to be suggesting that he plans to use telecons for more liason work in the future |
| 16:54 | <jgraham> | Lachy: In general the minutes are horrible to read. I often try to follow along on IRC a bit but generally it is impossible to pick up the substance of what anyone is saying |
| 16:55 | <Lachy> | I know |
| 16:55 | <annevk42> | yeah, DanC says they're not too important since it's only for people who want it while rubys seems to be pushing more people to attend |
| 19:00 | <aboodman> | Hixie: want to clarify something about workers and cross-origin requests: As I read the current spec, the behavior is simple: workers can only be created from same-origin resources, importScript() allows cross origin resources. |
| 19:00 | <aboodman> | is that correect? |
| 19:00 | <aboodman> | (I got confused by some older messages I saw on the list that indicated it was more subtle) |
| 19:07 | <dave_levin> | aboodman: Yep. importscript is support to be like <script> iirc |
| 19:10 | <hober> | hmm, html4all.org's dns registration expired |
| 19:14 | <dave_levin> | aboodman: http://lists.whatwg.org/pipermail/whatwg-whatwg.org/2009-June/020355.html |
| 19:15 | <aboodman_mac> | dave_levin: can you state whether you think that matches my parsing of it above? |
| 19:15 | <aboodman_mac> | all the talk about redirects is freaking me out. but this does basically say that you can only create workers from same-origin resources, right? |
| 19:15 | <dave_levin> | aboodman_mac: Yes. |
| 19:16 | <aboodman_mac> | Ok. I don't think we have test coverage for this in webkit. |
| 19:16 | <aboodman_mac> | i will make a patch. |
| 19:16 | <dave_levin> | aboodman_mac: Possibly not. |
| 19:16 | <dave_levin> | aboodman_mac: thx. |
| 20:02 | <mpilgrim> | annevk42: the fact that an accessibility "expert" refuses to discuss accessibility in an accessible medium is ENTIRELY YOUR FAULT |
| 20:03 | <mpilgrim> | (c.f. http://lists.w3.org/Archives/Public/public-html/2009Aug/0747.html ) |
| 20:03 | gsnedders | facepalms |
| 20:10 | Philip` | wonders if any HTML WG members would actually require transcriptions of audio |
| 20:11 | <mpilgrim> | at the risk of agreeing with john foliot, that's really beside the point |
| 20:11 | <mpilgrim> | even if we ignore deaf people, there are hearing impaired and hard of hearing people who could have trouble keeping up with a live telephone conversation |
| 20:12 | <mpilgrim> | even if we ignore those people, there are people whose shaky command of the english language would preclude them from keeping up with a live telephone conversation |
| 20:12 | <mpilgrim> | even if we ignore those people, there are people who would be asleep during the live telephone conversation, regardless of when it was held |
| 20:13 | <Philip`> | I thought the issue was audio recordings, which aren't live |
| 20:13 | <mpilgrim> | even if we ignore those people, there are people who were sick yesterday (cough me cough) and were in no condition to attend a live telephone conversation |
| 20:14 | <Philip`> | (Transcriptions would certainly be more useful regardless of accessibility, because you can skim-read and copy-and-paste etc) |
| 20:14 | <mpilgrim> | even if we ignore those people, there are people who would like to be able to read, search, copy-and-paste, and generally interact with written text |
| 20:14 | <mpilgrim> | jinx |
| 20:16 | <mpilgrim> | anyway, it boggles my mind that the w3c charges companies a million dollars a year to participate, insists on weekly teleconferences, puts people in leadership positions who use their position to harass people into attending those teleconferences, then says they don't have the money to transcribe the teleconference |
| 20:16 | <Dashiva> | Surely you aren't claiming the w3c is inaccessible? |
| 20:17 | <mpilgrim> | as far as i know, the WHATWG has never held a teleconference |
| 20:17 | <Dashiva> | Sorry, forgot my </sarcasm> |
| 20:17 | <mpilgrim> | never insisted that people attend a teleconference |
| 20:17 | <annevk42> | it boggles my mind that not publishing recordings is ok, but publishing them but not publishing an additional transcript that was not part of the conversation in the first place would not be acceptable |
| 20:17 | <mpilgrim> | yes, i was getting to that part too |
| 20:18 | <Dashiva> | There's no lobby for people who don't attend telcons |
| 20:18 | <mpilgrim> | apparently, the w3c has a policy -- and we all know how much work goes into their policies -- that inaccessible mediums are fine as long as you don't share them |
| 20:19 | <mpilgrim> | and john foliot takes this opportunity to... blame anne! |
| 20:19 | <mpilgrim> | so awesome |
| 20:20 | <annevk42> | I asked for clarification just in case I misunderstood |
| 20:20 | <Dashiva> | Does anyone know when JF went from against wcag2 to being for it? |
| 20:20 | <mpilgrim> | no, you understood perfectly |
| 20:21 | <annevk42> | I was afraid of that |
| 20:21 | <mpilgrim> | dashiva: as soon as he decided that hixie was against it (which he's not, but that's irrelevant) |
| 20:21 | <annevk42> | Dashiva, somewhere between the ALA article and now? :) |
| 20:22 | <mpilgrim> | meanwhile, i would STILL like to know what Richard Schwerdtfeger and maciej talked about |
| 20:25 | <Dashiva> | mpilgrim: Actually, the only reason they were talking was so maciej could proxy for Hixie. It doesn't matter for the rest of us. |
| 20:25 | <mpilgrim> | i'm actually interested in the roles and states module (aka ARIA) |
| 20:26 | <mpilgrim> | i was on richard's team at ibm while he was working on getting it into xhtml2 |
| 20:27 | <mpilgrim> | i've written code -- which as far as i know is still being used -- to allow you to use roles and states in text/html (by abusing the class attribute) |
| 20:27 | <mpilgrim> | i even, shocker of shockers, think that much of the roles and states module is a good idea |
| 20:28 | <mpilgrim> | hampered only by the albatross of xhtml2 |
| 20:28 | <annevk42> | I just unsubscribed from your twitter account and blog :p |
| 20:28 | <mpilgrim> | which people are finally rectifying |
| 20:28 | annevk42 | -> food |
| 20:30 | <Dashiva> | I'm curious to see how this will end up being retold |
| 20:32 | <Dashiva> | annevk42: The current telecons aren't in conflit with WCAG because WCAG only applies to the web |
| 20:33 | <Dashiva> | Maybe s/conflict with WCAG/inaccessible/ or somesuch |
| 20:50 | <Philip`> | When people talk about .tags on HTMLCollections, is that stuff like document.all.tags("blink") or something else? |
| 20:51 | <Dashiva> | It's that |
| 20:52 | <Dashiva> | But it's leaked outside of just document.all |
| 20:52 | <gsnedders> | At least it's not document.body.all |
| 20:52 | <Philip`> | Like fld.children.tags('td')[0].style.borderColor='#FF9966' etc? |
| 20:52 | <Dashiva> | My eyes |
| 20:52 | <gsnedders> | My kingdom for my eyes? |
| 20:54 | <Philip`> | Hmm, most seem to be children.tags or all.tags |
| 20:54 | <Philip`> | (where "most" means "somewhere between most and all (inclusive)") |
| 20:57 | <gsnedders> | Does .all exist on all HTMLElements in IE? |
| 21:04 | gsnedders | wonders why he doesn't get everything to ietf-charsets |
| 21:09 | <annevk42> | you mean s/to/on/? |
| 21:09 | <annevk42> | and by that do you mean emails or understanding? |
| 21:10 | <annevk42> | actually, why do I bother, it's Friday and I'm in France, ttyl |
| 21:10 | <gsnedders> | emails |
| 21:10 | <gsnedders> | s/to/sent to/ |
| 21:14 | <Philip`> | Ooh, I found one that's not all or children |
| 21:14 | <Philip`> | www.cartonservice.fr/public_site/info/contact.php?g_contact=2 var coll = document.dmd_cata.tags("input"); |
| 21:18 | <Philip`> | Maybe there are more, but this regexp is incredibly slow and I don't have any progress indicators |
| 21:25 | Philip` | does it a faster way |
| 21:26 | <Dashiva> | Madness |
| 21:26 | <Philip`> | Ah, there's just that one and |
| 21:26 | <Philip`> | www.mums.ac.ir/erc/fa/cornea2 for (var R = 0; R < document.forms[S].elements.tags("SELECT").length; R++) { |
| 21:26 | <Philip`> | www.mums.ac.ir/erc/fa/cornea2 document.forms[S].elements.tags("SELECT")[R].style.visibility = action; |
| 21:26 | <Philip`> | matching (?<!all|children)\.tags\( |
| 21:45 | <Dashiva> | Lastweek is so last week |
| 21:45 | <Dashiva> | Or at least earlier this week |
| 21:47 | <gsnedders> | Is there anyway to get the html parser to create an element in no namespace? |
| 21:50 | <jgraham> | gsnedders: No |
| 21:51 | <jgraham> | Why would you want that? |
| 21:51 | <gsnedders> | jgraham: Because html5lib has code for that case |
| 21:51 | <gsnedders> | jgraham: And using a conforming DOM Level 3 Core impl causes the DOM impl to break |
| 21:51 | <Dashiva> | Is it possible to create with any namespace that's no html, svg or mathml? |
| 21:52 | <Lachy> | Dashiva, only via createElementNS() |
| 21:52 | <gsnedders> | jgraham: As it creates element in no namespace using createElement(), and that means localName, prefix, and namespaceURI are None |
| 21:52 | <jgraham> | gsnedders: Oh well html5lib will, eventually, allow you to put html elements in no namespace thater than in the HTML namespace |
| 21:52 | <jgraham> | At the moment it ought to put everything into a namespace |
| 21:52 | <gsnedders> | (I get around 800 failures due to this bug) |
| 21:52 | <gsnedders> | It doesn't. |
| 21:53 | <Lachy> | jgraham, why will it allow you to put html elements in no namespace? |
| 21:53 | <gsnedders> | DOM is badly broken |
| 21:54 | <jgraham> | Lachy: Because for non-browser cases there is really no good reason to pay the namespace tax in the common case |
| 21:54 | <Lachy> | fair enough |
| 21:54 | <gsnedders> | Adding "assert namespace is not None" causes 1800 failures! |
| 21:54 | <jgraham> | Since there is no interoperability concern and namespaces make many otherwise good apis suck |
| 21:54 | <jgraham> | gsnedders: Well clearly something is going wrong |
| 21:55 | <jgraham> | I con't really look at it right now though |
| 21:55 | <jgraham> | Fell free too make a patch or keep complaining at me untill I do |
| 21:55 | gsnedders | is looking at it right now |
| 21:56 | <gsnedders> | token.get("namespace", self.defaultNamespace) returns None |
| 21:56 | <gsnedders> | Ah, the TreeBuilder is being created with namespaceHTMLElements=False |
| 21:59 | <gsnedders> | Because it defaults to not namespacing them |
| 22:03 | <gsnedders> | We default to breaking the spec. Sounds great. :\ |
| 22:05 | <jgraham> | We default to sane behaviour for most python applications |
| 22:05 | <jgraham> | But the default cn be changed if you are really bothered |
| 22:08 | <hsivonen> | can one get transcripts for $1.35 a minute by people who reside in North America and have health insurance? |
| 22:11 | <Philip`> | Do it on Mechanical Turk for $0.01 a minute |
| 22:13 | <Hixie> | play it back to a google voice account and have google voice voicemail transcription do it |
| 22:14 | <hsivonen> | Hixie: does Google have voice transription that works and doesn't require per-speaker training? |
| 22:14 | <Hixie> | yes |
| 22:15 | <hsivonen> | awesomeness! |
| 22:15 | <Hixie> | (for some definition of "works", anyway) |
| 22:15 | <hsivonen> | is it available for ed users for dictation? |
| 22:15 | <hsivonen> | *end users |
| 22:15 | <Hixie> | i don't think it's available anywhere other than google voice voicemail transcriptions right now |
| 22:15 | <gsnedders> | jgraham: I mind more than anything else because False is currently _really_ broken, and I'd rather we defaulted to following the spec |
| 22:15 | <hsivonen> | :-( |
| 22:16 | <gsnedders> | jgraham: Do you mind if I add a warning for the False case currently, warning that it is broken? |
| 22:16 | <Hixie> | there are people _against_ making the working group more transparent??? |
| 22:16 | <Hixie> | what kind of crazy world do we live in here |
| 22:16 | <gsnedders> | Hixie: Earth |
| 22:18 | <Dashiva> | Hixie: One with lawyers |
| 22:18 | <Dashiva> | Maybe maciej could make a statement on Apple's behalf (even if it's only that he'll have to talk to the lawyers) to defuse it :) |
| 22:20 | <Lachy> | new draft for about: URI scheme will be published soon. http://www.ietf.org/staging/draft-holsten-about-uri-scheme-02.txt |
| 22:20 | <Lachy> | I got mailed about it from the submission tool when joseph submitted it, so I hadn't had a chance to review it beforehand |
| 22:21 | <Lachy> | anyway, he's made a little bit of a mess out of some sections, but hopefully I can get those fixed and have an updated draft soon |
| 22:22 | <krijnh> | Is http://twitpic.com/dw92r the new Opera logo? |
| 22:23 | <Lachy> | krijnh, no |
| 22:23 | <krijnh> | I've never seen it before |
| 22:24 | <Lachy> | at least, I hope not. It looks like whoever made that did some weird stuff with all the logos |
| 22:24 | gsnedders | wonders if anyone in here knows what the new Opera logo looks like |
| 22:24 | <krijnh> | 50% in here works for Opera, no? :) |
| 22:24 | <Dashiva> | Is there a new Opera logo at all? |
| 22:25 | <Lachy> | Dashiva, no comment |
| 22:25 | <Dashiva> | Exactly |
| 22:25 | <gsnedders> | Dashiva: Yes. |
| 22:25 | <Lachy> | gsnedders, don't spoil the surprise |
| 22:25 | <gsnedders> | Lachy: But it's already been announced! |
| 22:25 | <Lachy> | where? |
| 22:26 | <krijnh> | On Hicks' site |
| 22:26 | <gsnedders> | Yeah |
| 22:26 | <krijnh> | And in his Sitepoint interview |
| 22:27 | <Dashiva> | It doesn't exist until it's unveiled |
| 22:27 | <Dashiva> | Rumors, which may be lies |
| 22:27 | <krijnh> | Damnit :) |
| 22:27 | <gsnedders> | krijnh: Just because 50% of people in here are from Opera doesn't mean 100% of people from Opera know :) |
| 22:27 | <Lachy> | krijnh, gsnedders, can you post a link? I can't see it anywhere on his site |
| 22:27 | <gsnedders> | It was on Twitter, IIRC |
| 22:28 | gsnedders | got asked when he got home from work one day by a friend whether he knew anything about it, and didn't even know about the tweet at the time |
| 22:28 | <krijnh> | http://hicksdesign.co.uk/journal/and-now-opera |
| 22:28 | <krijnh> | http://www.sitepoint.com/blogs/2009/08/02/podcast-21-jon-hicks/ |
| 22:29 | <krijnh> | "The answer is yes, I have." |
| 22:30 | <Lachy> | Don't quote mine! "And if the next question was have those designs been accepted and are they going to be used, the answer is no." |
| 22:31 | <krijnh> | Who cares, the answer is yes! :) |
| 22:31 | <Lachy> | there have definitely been proposals for a new logo |
| 22:31 | <Dashiva> | Everyone and their dog has redesigned the Opera logo, pssh :P |
| 22:31 | <krijnh> | :D |
| 22:31 | <Lachy> | there's been many, from others even before Hicks joined us |
| 22:31 | <Dashiva> | I mean, I interned there three summers in a row, and it was a discussion topic every time |
| 22:32 | <Dashiva> | "It sucks" "I agree" "Let's talk about it while nothing happens" |
| 22:32 | <krijnh> | Like now :) |
| 22:34 | <Hixie> | the entire thread on the recording of teleconferences is ludicrous |
| 22:37 | <hsivonen> | does WebApps actively maintain DOM Level 3 Core errata? |
| 22:37 | <gsnedders> | No |
| 22:37 | <hsivonen> | ok |
| 22:59 | <scherkus> | silly question, but I've never thought about it before |
| 22:59 | <scherkus> | is there anywhere in the spec that says _not_ to apply default styles to an element? |
| 22:59 | <scherkus> | someone was wondering if <video> could have border: 1px solid black; applied by default |
| 23:00 | <gsnedders> | It could. The spec says nothing normatively about rendering. |
| 23:01 | <scherkus> | I suppose it's one of those "user agents reach consensus" sort of thing |
| 23:02 | <gsnedders> | The informative advice in the spec for video is… |
| 23:02 | gsnedders | looks |
| 23:03 | <gsnedders> | video {} |
| 23:03 | <gsnedders> | So no style :P |
| 23:03 | <scherkus> | :P |
| 23:03 | <scherkus> | where is that btw? |
| 23:04 | <scherkus> | ...or are you referring to how nothing is mentioned at all :) |
| 23:05 | <gsnedders> | Nothing is mentioned, so implicitly that is what the rendering section says for it :) |
| 23:24 | <jacobolus> | http://developer.yahoo.com/yui/theater/video.php?v=crockford-json |
| 23:24 | <jacobolus> | skip to 40 minutes |
| 23:24 | <jacobolus> | that ~3 min of the talk made my day, a couple weeks ago |
| 23:31 | <Xanthir> | Bwahahaha |
| 23:42 | <Philip`> | jacobolus: Do you have a transcript of that? |
| 23:42 | <Philip`> | (or a summary) |
| 23:42 | Philip` | is currently incapable of viewing video |
| 23:43 | <jacobolus> | Philip`: scroll down |
| 23:43 | <jacobolus> | Philip`: search for "needed to put a software license" |
| 23:43 | <Philip`> | jacobolus: Oh, okay, I didn't actually open the link |
| 23:43 | <jacobolus> | and read the next few ¶s |
| 23:44 | <jacobolus> | I'm the "Audience member" in the transcript; my shouted question is inaudible in the video ;) |
| 23:48 | <inimino> | haha |
| 23:56 | <Hixie> | annevk42: why does XHR use Origin and not Sec-From? What's the difference? |
| 23:56 | <Hixie> | (trying to work out which websocket should use) |