| 01:17 | <Hixie> | hsivonen: http://identi.ca/notice/5639842 :-( ? |
| 04:45 | MikeSmith | smirks about "spelling errors occurred" |
| 04:49 | <MikeSmith> | description of r3304 says "Write some explanatory text around the HTML parser." but the actual change just looks like some additions to the embedded CSS stylesheet |
| 04:53 | <MikeSmith> | ah, I see now |
| 06:34 | <hsivonen> | Hixie: V.nu whines about cellspacing and cellpadding for me. Steps to reproduce? |
| 06:51 | <Hixie> | hsivonen: no idea, just saw the twitter |
| 08:40 | hsivonen | wishes cssquirrel studied the mailing lists enough to call other people on practicing the bible of handling people |
| 08:47 | <MikeSmith> | I wish cssquirrel comics has more sex and violence |
| 08:48 | <MikeSmith> | because there are few things in this world that are funnier than depictions of cartoon characters having sex with each other |
| 09:05 | <Dashiva> | I'm not sure I get the cssquirrel argument |
| 09:06 | <Dashiva> | It's been shown that these data studies can be independently verified, and his response is that we should throw away the useful data just because... it has to be verified? |
| 09:07 | <Dashiva> | Next up, dot.org is revealed to be the real mastermind behind this |
| 09:07 | <hsivonen> | Dashiva: it's the vast Google-wing conspiracy! independent verifications doesn't change that |
| 09:08 | <Dashiva> | It does seem to reduce to "Data is evil, let's just make up opinions instead" |
| 09:17 | <Philip`> | "[using non-Google data] helps take the wind out of any Google influence concerns that may arise (such as mine)" - I'm not sure that's necessarily the case, because we already do that stuff and he has those concerns anyway based on the misconception that we don't |
| 09:19 | <Philip`> | so I guess the way to alleviate those concerns is to make it clearer to concerned persons what data we use, rather than actually doing anything much differently |
| 09:24 | <Dashiva> | Data is evil, I tell you |
| 09:25 | <Philip`> | Only Google data is evil, because the editor works for Google |
| 09:26 | <Philip`> | Also, data collected by humans is evil, because the editor is a human |
| 09:26 | <Philip`> | My data's okay though |
| 09:26 | <hsivonen> | Philip`: why is your data okay? is it an expert opinion? |
| 09:26 | <Philip`> | even though all the actual processing is done in secret with non-public code and it could be full of bugs or I could be making it all up |
| 09:28 | <Philip`> | hsivonen: My expert opinion is that I'm not an expert and it would not generally be sensible to trust anything I say :-) |
| 09:30 | <Dashiva> | So you admit you're making it all up? |
| 09:30 | <hsivonen> | fortunately, real Web data is so crazy you can't make it up |
| 09:31 | <Philip`> | Dashiva: I can neither confirm nor deny such rumours |
| 09:32 | <Dashiva> | Google's data verifies that you're making it up |
| 09:33 | <Philip`> | http://www.google.com/search?q=Am+I+making+it+all+up%3F - apparently I am not making this up |
| 09:34 | <Dashiva> | That's a trope |
| 09:34 | <Dashiva> | You obviously wouldn't fall for that |
| 09:34 | <Dashiva> | So you are making it up! |
| 09:40 | <Philip`> | Dashiva: Hmph! Foiled again |
| 09:48 | <jgraham> | Philip`: It would be insane not to use Google data where it is useful. |
| 09:49 | <jgraham> | So it seems like the CSSquirrel guy should have said "oh, my whole article was based on the misconception that no one was trying to reproduce these results. Sorry!" |
| 09:49 | <jgraham> | Rather than saying "oh well you shouldn't use that readily avaliable source of data because it could lead to accusations like the one I just made" |
| 09:50 | <jgraham> | "even though people have taken reasonable care to apply rigour and reproducability to the analayses" |
| 10:13 | hsivonen | wonders what happened to markp's 'this week' posts |
| 10:25 | <Lachy> | regarding this test http://software.hixie.ch/utilities/js/live-dom-viewer/saved/150 it's not clear to me whether the UA can wait till the loop has finished executing before firing the events, or whether it can fire them synchronously each time setItem() is invoked |
| 10:26 | <Lachy> | Minefield seems to fire the events synchronously, but our internal Opera build waits until the loop is done and then fires all events together. |
| 10:27 | <Lachy> | (that test freezes the latest WebKit though) |
| 10:27 | <hsivonen> | does this mean Chrome 1.x will be gone as soon as instances autoupdate selves? http://googlechromereleases.blogspot.com/2009/06/stable-beta-update-security-fix.html |
| 10:29 | <annevk5> | Chrome 1.x was already gone |
| 10:29 | <abarth> | which is A Good Thing (tm) |
| 10:30 | <Lachy> | does Chrome automatically update itself without asking the user? |
| 10:30 | <annevk5> | hey abarth, I've been meaning to think you for taking care of the CORS thread |
| 10:30 | <abarth> | although i talked to someone the other day who still had Chrome 0.2 |
| 10:30 | <annevk5> | oops |
| 10:30 | <abarth> | he had hacked the omaha client to stop updates |
| 10:30 | <abarth> | annevk5: no problem. hopefully tyler isn't too mad at me |
| 10:31 | <annevk5> | Lachy, storage fires async |
| 10:31 | <annevk5> | Lachy, though maybe not yet in Gecko |
| 10:32 | <Lachy> | annevk5, ok |
| 10:33 | <abarth> | Lachy: yes. although you can control the updates with registry keys |
| 10:33 | <hsivonen> | it seems that Chrome makes it hard for Web authors to keep old versions around, which I guess is a good thing |
| 10:33 | <Lachy> | then I wonder what would happen if I let the loop run for a few million loops and let the task queue fill up |
| 10:34 | <abarth> | Lachy: http://code.google.com/p/omaha/wiki/UpdateControls |
| 10:34 | <abarth> | hsivonen: why do you want old versions? 99% of users are on the latest version |
| 10:35 | <hsivonen> | abarth: I don't. I said I guess it's a good thing that it's hard to keep old versions around. |
| 10:35 | <abarth> | hsivonen: ah, misread :) |
| 10:35 | annevk5 | wonders how this will work if Chrome is adopted by corporations |
| 10:35 | <Philip`> | Old versions are interesting as a historical curiosity |
| 10:36 | <abarth> | yeah |
| 10:36 | <abarth> | we tried running some super old version of navigator |
| 10:36 | <abarth> | they were super crashy |
| 10:36 | <abarth> | but fun |
| 10:36 | <hsivonen> | maybe I should subscribe my IE6 VM to Chrome stable, IE7 VM to Chrome beta and IE8 VM to Chrome dev |
| 10:36 | <abarth> | haha |
| 10:37 | <abarth> | the dev channel is too unstable for me |
| 10:37 | <abarth> | they're trying to make it better though |
| 10:37 | <Philip`> | (http://browsers.evolt.org/ is great, and lots of them still work in Windows or Wine) |
| 10:37 | <abarth> | with more automated testing |
| 10:37 | <Philip`> | (though some old versions of Netscape are presumably failing to send the Host header or something, and so are incapable of viewing almost every web site on the planet) |
| 10:38 | <abarth> | haha |
| 10:38 | <Philip`> | (or at least Google and Wikipedia, which I consider to be almost every web site on the planet) |
| 10:38 | <abarth> | we were looking at the lock icon |
| 10:38 | <jgraham> | Hmm it seems like the autoupdating + corporate users incompatibility would seriously damage the abaility of chrome to remove IE marketshare |
| 10:38 | <abarth> | i think the number of teeth on the lock represented the strength of the encryption |
| 10:39 | <abarth> | jgraham: chrome isn't optimized for enterprises yet |
| 10:39 | <jgraham> | (which I thought was the stated goal) |
| 10:39 | <abarth> | jgraham: the update controls are a first step in that direction |
| 10:39 | <gsnedders> | jgraham: I thought the stated goal was to help drive competition and hence development of all browsers |
| 10:40 | Philip` | once talked to someone who seemingly had a better-than-average understanding of web security, because he knew "if the thing at the top says something like HPTPS then it's safe" |
| 10:41 | <jgraham> | gsnedders: Maybe |
| 10:41 | Philip` | wouldn't really fancy having to explain how subtle changes in an icon represent encryption strength |
| 10:41 | <abarth> | yeah, it's just an interesting historical oddity |
| 10:42 | <abarth> | although it gives you some insight into their frame of mind at the time |
| 10:42 | <Philip`> | gsnedders: I thought the stated goal was world domination |
| 10:43 | <Philip`> | abarth: The userbase for web browsers was a bit different at that time too, so maybe it was not quite as unreasonable as it seems today |
| 10:44 | <gsnedders> | Philip`: What's the weather like there? |
| 10:45 | <Philip`> | gsnedders: I don't know, the weather monitoring station has been taken down for maintenance |
| 10:45 | <gsnedders> | :) |
| 10:45 | <jgraham> | gsnedders: It's like the siberian tundra with only slight changes to the temperature, precipitation and pressure |
| 10:46 | <annevk5> | Hixie, in the new Selector chapter you want to define "case-insensitive" as "ASCII case-insensitive" and link it and such |
| 11:28 | MikeSmith | wonders where zcorpan might be |
| 11:29 | <MikeSmith> | can't remember if Simon was around when we were discussing the [Reflect] thing yesterday |
| 11:30 | <hsivonen> | MikeSmith: he added a note to Web DOM Core later |
| 11:30 | <MikeSmith> | oh |
| 11:30 | <gsnedders> | MikeSmith: Linköping |
| 11:31 | <MikeSmith> | hsivonen: aha, I see it now |
| 11:31 | <jgraham> | gsnedders: He might be, he might not. It's hard to tell |
| 11:31 | <MikeSmith> | gsnedders: gesundheit |
| 11:34 | <gsnedders> | jgraham: Do you know whether you are right about where you thought I would be sitting? |
| 11:35 | <jgraham> | Yes. I know. |
| 11:37 | <gsnedders> | jgraham: Were you write? |
| 11:37 | <gsnedders> | *right |
| 11:38 | <takkaria> | oo, where? |
| 11:38 | <jgraham> | As I am handed |
| 11:38 | gsnedders | sighs |
| 11:39 | <gsnedders> | jgraham: Are the people I'll be around evil? :P |
| 11:41 | <jgraham> | Do you consider "overendowed with natural programming ability" to be a sign of evil? |
| 11:41 | <gsnedders> | jgraham: Possibly. |
| 11:42 | <Dashiva> | Depends on the language |
| 11:43 | <gsnedders> | If Perl, certainly. |
| 11:43 | <Philip`> | There's nothing natural about the ability to program Perl |
| 11:44 | <jgraham> | Hixie and Philip` both have a certian fondness for Perl. I guess that explains a lot |
| 11:46 | <jgraham> | Maybe the CSSquirrel is right |
| 11:46 | Philip` | hasn't written any Perl more complex than "while (($_ = <$authz>) =~ /\S/) {" today, and that's quite readable :-( |
| 11:46 | <hsivonen> | maybe it's a vast Perl-wing conspiracy instead of a Google conspiracy. Google uses Python. |
| 11:50 | <Philip`> | Perl users are too pathologically eclectic to form a conspiracy |
| 11:55 | <hsivonen> | looks like IBM now has an Orbeon-like product: http://www.ibm.com/developerworks/blogs/page/JohnBoyer?entry=xforms_xfdl_and_lotus_domino?ca=drs-bl |
| 12:00 | <hsivonen> | why isn't webapps killing mutation events altogether? |
| 12:46 | <takkaria> | I've not found anywhere in Sweden yet which sells fresh potatoes |
| 12:46 | <takkaria> | this saddens me |
| 12:46 | <takkaria> | I like potatoes |
| 12:48 | zcorpan_ | has eaten swedish fresh potatoes several times this year |
| 12:49 | <jgraham> | takkaria: Did you try a shop? |
| 12:49 | <takkaria> | I have |
| 12:50 | jgraham | has no idea what the difference is beteen fresh and non-fresh potatoes but has had no difficulty in aquiring potatoes |
| 12:50 | <takkaria> | well, I can find jars of salted potatoes |
| 12:50 | <takkaria> | I couldn't find any in Hemk�p, anyway |
| 12:51 | <jgraham> | Odd |
| 12:51 | <jgraham> | I am reasonably sure I have bought potatoes there |
| 12:52 | <jgraham> | I mean potatoes, in all their forms, seem to make up about 9/10 of the national diet |
| 12:52 | <takkaria> | maybe they've just not been then when I have |
| 12:53 | <takkaria> | s/then/there/ |
| 13:12 | <beowulf> | takkaria: ask gsnedders to bring some potatoes in his suitcase... |
| 13:24 | <Philip`> | He could smuggle them across the border disguised as Mr Potato Heads |
| 13:24 | <Philip`> | to avoid having to pay import duty |
| 13:51 | <zcorpan_> | Hixie: did you look at the emails when writing the latest checkin? |
| 14:32 | <drcode> | hi all |
| 15:19 | <zcorpan_> | Philip` says "I don’t remember Hixie quoting any private data about that issue." - http://www.cssquirrel.com/2009/06/22/comic-update-who-really-is-the-wizard-of-html5/ |
| 15:19 | <zcorpan_> | the response? "Philip, then Ian needs to use these external data sources from now on, rather than falling back on Google’s index." |
| 15:21 | <zcorpan_> | interesting that people complain about Hixie using google's index when he has only been looking at other people's data |
| 15:22 | <jgraham> | zcorpan_: I'm not sure interesting is quite the right word |
| 15:23 | Philip` | 's memory could easily be wrong, but he'll stick with it until someone provides evidence demonstrating otherwise |
| 15:26 | <MikeSmith> | that side of the discussion sorta seems to be missing the basic point that some data is at least better than no data |
| 15:27 | <zcorpan_> | MikeSmith: that's not what i see. i see complaints about hixie basing decisions on data from google's index |
| 15:28 | <zcorpan_> | MikeSmith: while in fact he's basing decisions on data provided by other people |
| 15:28 | <MikeSmith> | that is, if person/group P don't like the data that's been given by person/group H, then P is free to collect/publish other data |
| 15:28 | <MikeSmith> | zcorpan_: yeah, understood |
| 15:29 | <Philip`> | MikeSmith: Some data is no worse than no data, because you can always choose to ignore the data; but the complaints seem to be about people not ignoring the data |
| 15:29 | <jgraham> | I think that, at a more fundamental level, that side of the discussion is missing the fact that language design is fundamentally about HCI design. |
| 15:29 | <MikeSmith> | but my point about this is that it's perhaps not about the particular data, it's instead a kind of general fear of data for this case |
| 15:30 | <jgraham> | I really mean to write a blog post about that but I don't know how to get going |
| 15:30 | <MikeSmith> | that is, a realization that no matter where the data comes from for this case, it's going to support the conclusion that's been made so far after observing it |
| 15:31 | <MikeSmith> | so some choose to argue against the particular data-collection methods |
| 15:31 | <MikeSmith> | but they do so to get people to ignore the data |
| 15:32 | <takkaria> | quite a brilliant method, really |
| 15:32 | <takkaria> | I believe it's outlined in Hixie's Bible. :) |
| 15:32 | <MikeSmith> | so cssquirrel comic illustrates that pretty well |
| 15:32 | <MikeSmith> | ignore the data behind the curtain |
| 15:32 | <MikeSmith> | look over here where I'm waving my hands instead |
| 15:33 | <Philip`> | MikeSmith: The CSSquirrel thing doesn't say what specific case it's referring to, which makes it harder to argue against using specifics, but I guess it's really about accessibility again |
| 15:33 | <MikeSmith> | Philip`: any time nothing is mentioned specifically, it's usually about accessibility |
| 15:33 | jgraham | is still clinging on to the hope that people aren't malicious but misguided |
| 15:34 | <MikeSmith> | because unfortunately many people don't want to talk about it explicitly |
| 15:34 | <MikeSmith> | not talk about it in any kind of negative way, at least |
| 15:35 | <gsnedders> | Is there any way to resubmit a form in Opera? |
| 15:36 | <gsnedders> | Or do I really need to re-write what I just wrote? |
| 15:36 | <Dashiva> | ? |
| 15:36 | <MikeSmith> | jgraham: I don't think anybody involved in this is actually maliciously or even that they are particularly misguided |
| 15:36 | Philip` | thinks it's hard to avoid being misguided, because there's so much information about HTML5 development that it's pretty much impossible to keep up with it all, and so most people only notice the few specific areas where someone shouts loudly |
| 15:36 | <Dashiva> | gsnedders: When you go back, the form contents should still be there? |
| 15:36 | <gsnedders> | Dashiva: It isn't. It reloads the page with the original content. |
| 15:36 | <Philip`> | gsnedders: The back button works for me, except when the site is stupid and tries to do something clever |
| 15:37 | <MikeSmith> | jgraham: seems like most everybody has an investment/agenda they want to protect, even when they don't admit it or call it that |
| 15:38 | <jgraham> | MikeSmith: Maybe |
| 15:39 | <MikeSmith> | jgraham: and I guess some people are more easily able to cut their losses on a what turns out might have been a not-so-great investment, and some people find it a lot harder |
| 15:40 | <MikeSmith> | (how's that for talking around things?) |
| 15:40 | <MikeSmith> | I should shut up now |
| 15:41 | <beowulf> | i don't see how it helps that people are scared to talk about accessibility, if i can put it that way |
| 15:41 | <beowulf> | but what do i know? |
| 15:41 | <Philip`> | You shouldn't shut up - you should try saying more and more outrageous things, and see how long people keep thinking everything you say in IRC should be taken seriously |
| 15:41 | <MikeSmith> | heh |
| 15:41 | <MikeSmith> | I may be past that point already |
| 15:42 | <MikeSmith> | Philip`: and I've already succeeded in tainting our bugzilla with that kind of ambiquity and confusing now to |
| 15:42 | gsnedders_ | contemplates going back to using Safari to write blog posts, because he never lost data there |
| 15:42 | gsnedders_ | sighs |
| 15:43 | <Philip`> | gsnedders_: You should write them in Notepad and then paste them into the browser |
| 15:43 | <beowulf> | gsnedders_: use vim for that :) |
| 15:43 | <zcorpan_> | takkaria: do you have your irc client set to utf-8? |
| 15:43 | <takkaria> | zcorpan_: I think so |
| 15:43 | <takkaria> | gsnedders_: text editor first, paste into form later? |
| 15:43 | <gsnedders_> | Reloading the submitted form page (which was just "connection failed" gives me a blank page) |
| 15:43 | <zcorpan_> | takkaria: http://krijnhoetmer.nl/irc-logs/whatwg/20090623#l-321 |
| 15:44 | <takkaria> | oh, damn, not one but two people got there first |
| 15:44 | <gsnedders_> | My browser should not cause me to lose data. |
| 15:44 | <takkaria> | zcorpan_: hm, I'll tweak |
| 15:44 | gsnedders_ | wonders if there is anyway to get at the data I lost |
| 15:44 | <MikeSmith> | beowulf: on the subject of humor, you riff about gsnedders_ suitcase yesterday was solid gold |
| 15:44 | <MikeSmith> | "carved out of single Scottish oak" or something like that |
| 15:45 | <Philip`> | Solid gold? That would explain why it was so heavy |
| 15:45 | <beowulf> | finally i make a contribution, where's my cv... |
| 15:45 | <takkaria> | gsnedders_: depends |
| 15:45 | <takkaria> | gsnedders_: do a memory dump to disk and grep |
| 15:45 | <takkaria> | worked for me once |
| 15:45 | <MikeSmith> | the helium and stuff from Philip` and others after that that was gold-ish too |
| 15:45 | <takkaria> | took me about an hour |
| 15:45 | <jgraham> | Anyway, like I said, the big philosopical divide I see is beteen people who view writing a language as basically an engineering problem (we have to provide features x y and z so we will have markup corresponding to x y and z) and people wwho view it as a HCI problem (we have to maximise the chance that users are exposed to the features x y and z and that they are used in the right way) |
| 15:47 | gsnedders_ | wonders how to get a memory dump |
| 15:48 | <Philip`> | 'cat /dev/mem' if you have a sensible OS |
| 15:48 | <takkaria> | as root |
| 15:50 | <gsnedders_> | There is no /dev/mem |
| 15:50 | <Philip`> | Install a sensible OS first, then |
| 15:50 | <Philip`> | being careful not to reboot as it may disrupt your RAM content |
| 16:01 | <MikeSmith> | heh |
| 16:02 | <MikeSmith> | (about "being careful not to reboot as it may disrupt your RAM content") |
| 16:03 | <takkaria> | another way would be to write a program which malloc()s progressively more memory without freeing it and saving each chunk todisk |
| 16:03 | <takkaria> | eventually it crashes and you have a dump of free memoryspace |
| 16:04 | <Philip`> | Except that the OS will zero any pages before giving them to your process |
| 16:04 | <Philip`> | (otherwise it'd be a horrendous security hole) |
| 16:04 | <takkaria> | yeah, that's a point, really |
| 16:04 | <takkaria> | it would have worked on RISC OS |
| 16:05 | <takkaria> | that is all |
| 16:06 | <Philip`> | Would you need to malloc at all in that case, rather than just doing fwrite((void*)0, 1, 0xffffffff, f)"? |
| 16:06 | <Philip`> | s/"// |
| 16:08 | <takkaria> | hmm |
| 16:08 | <takkaria> | I don't know if RISC OS would let you do that |
| 16:23 | <MikeSmith> | hmm, speaking of fear, I predict the jd-style practice of people putting scare quotes around particular words, an indicator that they fear them, will spread |
| 16:23 | <MikeSmith> | or has already |
| 16:27 | <beowulf> | i'd really like a dummies guide to the @summary issue |
| 16:28 | <Philip`> | MikeSmith: Hmm, I wish more people published indexes from their web "spiders" |
| 16:36 | <Philip`> | beowulf: That would be a nice thing to have, but I'm not aware of any that have been written with biases I agree with :-( |
| 16:44 | <Midler> | whats is it called again, the specs that make it possible to save data localy? |
| 16:44 | <Midler> | (i really sucks when typing in english) |
| 16:46 | <Philip`> | Midler: Do you mean http://dev.w3.org/html5/webstorage/ ? |
| 16:46 | <takkaria> | I don't know if RISC OS would let you do that |
| 16:46 | takkaria | rms that last line |
| 16:47 | <Philip`> | Someone needs to make an IRC client with an "undo send" feature |
| 16:48 | jgraham | wonders what rms has to do with it |
| 16:48 | <takkaria> | i guess like gmail's but with a shorter gap before sending |
| 16:49 | <Midler> | Philip`: yup, thars the one. Localstorage |
| 16:54 | <beowulf> | Philip`: then that means people will make their minds up based on which personality they happen to like, or which football team they support, i guess |
| 16:59 | <Philip`> | beowulf: I don't think that's a great solution, since I don't support any football team |
| 17:00 | <Philip`> | (but I don't care enough about accessibility to try to find or write a suitable document describing the issues) |
| 17:00 | <Philip`> | (so I'll let other people worry about it instead) |
| 17:01 | beowulf | will be more careful in choosing his metaphors in future |
| 17:03 | <takkaria> | yes, Shelley, "data is ... used to undercut those with experience and expertise in accessibility" |
| 17:03 | <Philip`> | You should choose a metaphor like you choose a wine |
| 17:03 | <takkaria> | that's the only reason people use data |
| 17:03 | <takkaria> | is to undercut other people's expertise |
| 17:03 | <takkaria> | not because it's an acutal grounds for making decisions on or anything |
| 17:04 | takkaria | rages a bit |
| 17:04 | <Midler> | How does firefox save localstore data? |
| 17:04 | <Midler> | in localstore.rdf? |
| 17:04 | <takkaria> | it uses sqlite |
| 17:04 | <jgraham> | Philip`: I don't drink wine. Does that mean I can't use metaphors? |
| 17:04 | <Midler> | ok |
| 17:04 | <takkaria> | Philip`: what if you just buy cheap red all the time? |
| 17:05 | <beowulf> | Philip`: I don't drink wine, you don't support a football team, we're doomed, doomed I say |
| 17:06 | <Philip`> | This all proves my point about metaphors! |
| 17:06 | <Philip`> | I don't know what my point was, but it's proved nonetheless |
| 17:07 | <takkaria> | in much the same way that bread is proved, I'm assuming |
| 17:08 | <jgraham> | By covering it in clingfilm and leaving it in a warm place? Or are we talking sourdough? |
| 17:08 | <takkaria> | I did that with some curry by accident today |
| 17:09 | <Philip`> | 10g yeast, 500g flour, 295ml warm water, QED? |
| 17:09 | <takkaria> | forgot to put it in the fridge |
| 17:10 | <jgraham> | Philip`: Depends on the type of yeast and the type of flour |
| 17:12 | <beowulf> | are we still talking in code about accessibility? |
| 17:12 | <beowulf> | i'm only slightly less lost than the actual threads on the wg list, if that's the case |
| 17:12 | <jgraham> | beowulf: I was just talking about bread making. What were you talking about? |
| 17:12 | <jgraham> | :) |
| 17:12 | <jgraham> | s/:/;/ |
| 17:13 | <Philip`> | http://www09.wolframalpha.com/input/?i=1+hogshead+of+bread - 2705% of the recommended daily amount of fat? That can't be good |
| 17:13 | <beowulf> | mmm, 2705% rda of fat |
| 17:16 | Philip` | continues his quest to avoid productive discussion |
| 17:20 | <takkaria> | 2705% rdfa of fat would make some people happier |
| 17:31 | Philip` | finds a program which appears to have a self-defeating anti-debugging mechanism |
| 17:31 | <Philip`> | If you run an expired trial version, it (presumably) checks some files or registry entries (on Windows) to decide it's expired, and then complains |
| 17:32 | <Philip`> | but if you run Process Monitor, to attempt to see which files/entries it's checking, it appears to skip the checks entirely and acts as if it's non-expired |
| 17:32 | <Philip`> | (presumably to prevent you easily reverse-engineering it and working out what it's checking) |
| 17:33 | <Philip`> | so all you need to do is keeping Process Monitor running in the background all the time |
| 17:34 | <MikeSmith> | x |
| 17:34 | <MikeSmith> | sweet |
| 17:37 | <Midler> | Or, like my friend did with one app.. Just changed the date for the computer |
| 17:37 | <Midler> | And if u are wondering, its still 1994 for him ;) |
| 17:39 | <zcorpan_> | hsivonen, MikeSmith: "Warning: Attribute with the local name href is not serializable as XML 1.0." - http://html5.validator.nu/?doc=http%3A%2F%2Fhacks.mozilla.org%2F2009%2F06%2Fhtml5-video-fallbacks-markup%2F |
| 17:40 | <zcorpan_> | also, "Error: The internal character encoding declaration must be the first child of the head element." |
| 17:42 | <zcorpan_> | it seems wordpress inserts <br /> for each new line in the <video> markup |
| 17:43 | <MikeSmith> | great |
| 17:43 | <zcorpan_> | the algorithm to select a <source> copes, though |
| 17:43 | <zcorpan_> | if it would insert a <p>, then it would not work |
| 17:46 | <zcorpan_> | i think "Error: Attribute href not allowed here." is supposed to say xmlns:dc, too |
| 17:51 | <zcorpan_> | http://bugzilla.validator.nu/show_bug.cgi?id=584 |
| 17:55 | MikeSmith | reads through hsivonen treebuilder code, feels sense of intimidation setting in |
| 17:56 | <MikeSmith> | zcorpan_: I don't get the "Error: The internal character encoding declaration must be the first child of the head element." message |
| 17:57 | <MikeSmith> | zcorpan_: I'm hoping the xmlns:dc problem is just an easy-to-fix off-by-one error |
| 18:02 | <MikeSmith> | zcorpan_: oh, actually, I do |
| 18:02 | <MikeSmith> | but only if I use html5.validator.nu |
| 18:02 | <MikeSmith> | not if I use validator.nu |
| 18:04 | <MikeSmith> | d'oh! |
| 18:05 | <MikeSmith> | (hadn't manually set validator.nu select option to html5) |
| 18:13 | zcorpan_ | filed a bug on wordpress http://core.trac.wordpress.org/ticket/10247 |
| 18:33 | <beowulf> | would <span> be allowed in a summary element? |
| 18:34 | <beowulf> | or <p>? |
| 18:35 | <MikeSmith> | zcorpan_: fix for "encoding declaration must be the first child" problem checked in |
| 19:29 | <zcorpan_> | MikeSmith: nice |
| 19:29 | zcorpan_ | notes that there would be much less resistance to just allow summary="" in html5 |
| 19:31 | <Dashiva> | zcorpan_: What design principle is that? :) |
| 19:32 | <zcorpan_> | Dashiva: it's not a design principle |
| 19:33 | <zcorpan_> | "Seems to me that this working group's underlying practice is more to take the path of least resistance," -- http://www.w3.org/mid/643cc0270906230930j1c37cf7dv46e58637d5dda46⊙mgc |
| 19:57 | <Hixie> | zcorpan_: there would be less resistance from one particular group of people, yes |
| 19:57 | <Hixie> | zcorpan_: my goal though is not to make those peopel happy but to actually measurably improve accessibility |
| 20:08 | <jgraham> | Expertise in the HTMLWG is like penises. Everyone likes to brag about how much they have but when it comes time to do some measurements they become remarkably reluctant to put it on the table. |
| 20:08 | <Hixie> | lol |
| 20:09 | takkaria | grins |
| 20:43 | <annevk2> | so summary="" is by default not revealed to end users? |
| 20:45 | <Hixie> | ? |
| 20:45 | <annevk2> | Hixie, at some point earlier today I pointed out that case-insensitive in the new Selectors section was not linked to ASCII case-insensitive |
| 20:46 | <Hixie> | yeah that was somewhat intentional... i was hoping css defined it |
| 20:46 | <annevk2> | Hixie, (the summary="" was about http://lists.w3.org/Archives/Public/public-html/2009Jun/0604.html and zcorpan's reply) |
| 20:46 | <annevk2> | Hixie, CSS wants the markup language to define it |
| 20:46 | <Hixie> | oh i'll get to public-html in a few fours |
| 20:46 | <Hixie> | swamped with www-style right now |
| 20:47 | <annevk2> | "The case sensitivity of document language element names, attribute names, and attribute values in selectors depends on the document language." -- http://www.w3.org/TR/css3-selectors/#casesens |
| 20:49 | <Hixie> | ah, bummer |
| 20:49 | <Hixie> | ok |
| 20:53 | <Hixie> | zcorpan_: technically <video hidden> means "this isn't relevant" and so we should probably make it mute the audio as well |
| 20:53 | <annevk2> | it's sort of nice as it means we could make all attribute values case-sensitive (which I think we should do) |
| 20:54 | <Hixie> | well we can't make all of them case-sensitive |
| 20:54 | <Hixie> | the ones in the spec are case-insensitive already in at least two browsers |
| 20:54 | <annevk2> | IE until IE7 had everything case-sensitive |
| 20:54 | <Hixie> | IE until IE7 didn't have attribute selectors, did it? |
| 20:54 | <annevk2> | including IE7, I should say |
| 21:02 | <Hixie> | shelley thinks i'm an expert? even _i_ don't think i'm an expert. |
| 21:04 | <Philip`> | An expert is just a retired pert |
| 21:14 | <atwilson> | I have a question about built-in constructors, and what the behavior should be if you pass a constructor to a different window or frame. |
| 21:16 | <atwilson> | Let's say I pass the window.XMLHttpRequest constructor into a child frame. That child frame uses the constructor to instantiate an XHR object. Does the XHR object get the origin from the originating window (the window that the XHR constructor came from) or from the calling scope (the child frame)? |
| 21:16 | <atwilson> | I have a similar question about Worker objects, but I figure there's probably a general principle at work here but I don't know where in the spec I might look for it. |
| 21:18 | <ap> | atwilson: I don't remember the rules, but the current behavior is a result of many tweaks and fixes - so there's chance that it's correct |
| 21:19 | <ap> | atwilson: sorry, didn't notice where we are :) |
| 21:19 | <ap> | atwilson: I was talking about WebKit |
| 21:19 | <atwilson> | ap: The reason I ask is the WebKit behavior for XHR differs from the WebKit behavior for Workers. |
| 21:19 | <atwilson> | ap: so I wanted to see the actual spec. |
| 21:20 | <atwilson> | ap: (per the recent webkit-dev thread) |
| 21:21 | <ap> | atwilson: like <http://dev.w3.org/2006/webapi/XMLHttpRequest-2/#origin-and-base-url>;? |
| 21:21 | <ap> | atwilson: I think that annevk2 was going to update that though |
| 21:22 | <atwilson> | ap: yes, but I'm not sure how to parse that spec language. |
| 21:23 | <atwilson> | ap: how does one determine 'which is the Document object associated with the Window object for which the XMLHttpRequest interface was created'? |
| 21:24 | <ap> | atwilson: I think we should wait for Anne to weigh in on whether this text in in flux indeed |
| 21:24 | <atwilson> | Just look up the scope chain to find the global scope, and use the Window object associated with that? Because if that's the spec, then we aren't doing that in WebKit. |
| 21:24 | <atwilson> | Agreed. And I'm only using XHR as an example - the same issue applies to Workers. |
| 21:25 | <atwilson> | ap: Worker lifecycle depends on what document a given Worker is associated with, so making sure we associate with the correct Window is important. |
| 21:32 | <othermaciej_> | atwilson: XMLHttpRequest is totally clear on what to do |
| 21:33 | <othermaciej_> | (in my opinion) |
| 21:33 | <othermaciej> | and I am quite sure WebKit matches the intended spec behavior |
| 21:33 | <othermaciej> | Web Workers is a bit less clear |
| 21:33 | <othermaciej> | the only use of an original Window or Document would be for resolving a relative script URL against a base |
| 21:33 | <othermaciej> | it says to resolve the URL but it's not very clear about the base |
| 21:34 | <othermaciej> | my assumption is it should work like Location, which would mean the calling function's global object's current document is right |
| 21:35 | <othermaciej> | atwilson: the intent of the spec (as I understand it) is that whenever the XMLHttpRequest object is first instantiated by retrieving it from a particular Window object, it should record that Window object's current Document |
| 21:36 | <othermaciej> | WebKit records the Window instead of the Document |
| 21:36 | <othermaciej> | I can't remember if this makes an observable difference |
| 21:36 | <othermaciej> | there is definitely no looking up the scope chain involved |
| 21:40 | <ap> | othermaciej: I think the difference is that we record the window in constructor, not just in XMLHttpRequest object |
| 21:40 | <ap> | othermaciej: an I think that Anne was going to update the spec in this regard |
| 21:47 | <othermaciej> | ap: I see, the spec is probably wrong then |
| 21:47 | <othermaciej> | ap: I find its use of "the XMLHttpRequest object" a little confusing |
| 21:47 | <othermaciej> | hard to tell if it means an instance, or the constructor |
| 21:48 | <othermaciej> | it's "Window object for which the XMLHttpRequest interface object was created" that is unclear I guess |
| 22:00 | <atwilson> | othermaciej: yeah, if the spec is updated to reflect that the Window is bound to the constructor rather than to any given instance, then the spec is clear (and matches WebKit). |
| 22:01 | <zcorpan_> | aren't interface objects and constructors different things? |
| 22:01 | <zcorpan_> | window.XMLHttpRequest - interface object |
| 22:01 | <zcorpan_> | new XMLHttpRequest() - constructor |
| 22:02 | <atwilson> | Hixie: Do you have an opinion about how WebWorkers should work (if I pass window.Worker into a child frame, and then the child frame instantiates a Worker using that passed-in constructor, which document object ends up in that Worker's document set)? |
| 22:04 | <atwilson> | zcorpan_: Quite possibly. I'm trying to find out where 'interface object' is defined in the spec now. I can't find that string in the HTML5 spec when I search for it, although it is referenced in the XHR spec. |
| 22:04 | <zcorpan_> | atwilson: it's defined in webidl |
| 22:04 | <zcorpan_> | constructors too |
| 22:05 | <atwilson> | awesome, I'll look there. Thanks for the pointer. |
| 22:06 | <zcorpan_> | (though maybe a constructor is always also an interface object in implementations?) |
| 22:08 | <zcorpan_> | "An ECMAScript implementation supporting these interfaces would have a [[Construct]] property on the Circle interface object which would return a new object that implements the interface." |
| 22:09 | <atwilson> | The XHR spec is much clearer now, since it's clear that the constructor is bound to the originating window (i.e. var f = window.XMLHttpRequest; var xhr = new f(); yields an XHR object bound to the original window, regardless of what context new f() was invoked in) |
| 22:12 | <atwilson> | It sounds like Workers operate differently, though - the owning document seems to be tied to a "script's global scope". |
| 22:12 | <atwilson> | (section 4.5 of the Web Workers spec) |
| 22:17 | <zcorpan_> | it seems webidl allows [NoInterfaceObject] [Constructor] interface Foo {} |
| 22:17 | <zcorpan_> | which makes no sense in ecmascript |
| 22:35 | <Hixie> | atwilson: looking at "if I pass window.Worker into a child frame, and then the child frame instantiates a Worker using that passed-in constructor, which document object ends up in that Worker's document set" now... |
| 22:38 | <Hixie> | atwilson: web workers doesn't say how to resolve the argument to the constructor. will fix that now. |
| 22:54 | <annevk2> | so I did have a pass at XHR for those wanting to know |
| 22:54 | <annevk2> | currently base URL and origin are fixed when invoking open() |
| 22:54 | <annevk2> | I'm wondering if that's entirely correct or not |
| 22:55 | <annevk2> | it seems implementations differ a bit on the details and I've to look at all relevant testcases again (plus maybe make some more) to figure out if anything needs to change |
| 23:02 | <othermaciej_> | zcorpan_: that's not how I understood "interface object" but I could be confused |
| 23:02 | <othermaciej_> | zcorpan_: the object used for window.XMLHttpRequest is the same as the one used when invoing new XMLHttpRequest() |
| 23:02 | <othermaciej_> | zcorpan_: but different from the object *returned* by XMLHttpRequest |
| 23:04 | <Hixie> | atwilson: ok, made it so that the url is resolved relative to the /first script/'s /script's base URL/ (see HTML5 for meanings of those terms), so all Worker constructors work identically |
| 23:05 | <annevk2> | othermaciej_, window.XMLHttpRequest is the interface object |
| 23:06 | <annevk2> | othermaciej_, though what you're saying there is very interesting and something I'd like to learn about |
| 23:10 | <annevk2> | specifically, the relation between the various objects |
| 23:10 | <annevk2> | but I should go to bed |
| 23:11 | <annevk2> | Hixie, in Web Workers you also need to define the relevant bits for XMLHttpRequest |
| 23:12 | <annevk2> | Hixie, I introduced the concepts "XMLHttpRequest base URL" and "XMLHttpRequest origin" |
| 23:12 | <annevk2> | Hixie, for when there's no such thing as "XMLHttpRequest Document" |
| 23:15 | <atwilson> | Hixie: OK, so it sounds like your intent is, indeed, to use the current script's global scope to determine the Document to insert in the set. So if I pass window.Worker to another frame, it doesn't carry with it any intrinsic relationship to that original window or its document. Correct? |
| 23:16 | <zcorpan_> | othermaciej: yeah, the constructor is the interface object |
| 23:17 | <othermaciej> | in that case the XHR spec is fairly clear and correct, I think |
| 23:18 | <Hixie> | atwilson: correct, all Worker constructors are created equal |
| 23:18 | <Hixie> | annevk2: can you drop me a mail about that? |
| 23:18 | <atwilson> | Hixie: thx. |
| 23:25 | jgraham | thought he had got over his "write a long essay" phase of email composition. Sadly not |
| 23:26 | <jgraham> | Also we should totally have a HTMLWG metup in Linkoping. Me + zcorpan_ + gsedders + takkaria for starters :) |
| 23:28 | <annevk2> | I have this plan to come over to Sweden |
| 23:28 | <annevk2> | maybe end of July? |
| 23:28 | annevk2 | is really going to bed now; has a flight around 10ish |
| 23:30 | <jgraham> | For the record of the logs I am away at the end of July |
| 23:34 | <othermaciej> | jgraham: yeah, that was kind of a wall of text |
| 23:43 | <zcorpan_> | http://twitter.com/tieleman/statuses/2292450418 |
| 23:49 | <Hixie> | is there an up-to-date implementation of the <th scope> algorithm in a web tool i can poke at? |
| 23:50 | <Hixie> | is http://james.html5.org/tables/table_inspector.html up to date? |
| 23:50 | <Hixie> | (jgraham?) |
| 23:52 | <zcorpan_> | Hixie: it's up to jan 2009 |
| 23:55 | <Hixie> | k |
| 23:57 | <zcorpan_> | "Did you know that HTML5 editor Ian Hickson stated that HTML5 won’t fully be implemented in all browsers until 2022?" - http://www.davidtucker.net/2009/06/19/an-honest-open-discussion-on-web-standards-and-html-5/ |
| 23:57 | <zcorpan_> | i can't remember Hixie having stated that |
| 23:57 | <Hixie> | i already commented |