| 00:00 | <AryehGregor> | I have some thoughts about doing that, but it's kind of pointless unless I start writing more. |
| 00:00 | <AryehGregor> | gsnedders, me neither, I'd use Python. |
| 00:00 | <AryehGregor> | I don't know the Python names offhand, though. |
| 00:00 | <gsnedders> | me too. |
| 00:00 | <AryehGregor> | cgi.escape() or something? |
| 00:00 | <gsnedders> | I would just use a serializer, and not do that ;P |
| 00:00 | <Philip`> | I'd let people write arbitrary unfiltered HTML |
| 00:01 | AryehGregor | doesn't know what a serializer is, is probably too corrupted by PHP's weakness |
| 00:01 | <Philip`> | Much more fun that way |
| 00:01 | <gsnedders> | "When you do XSS attacks, it's considered highly impolite to make the resulting markup invalid" â Philip Taylor |
| 00:01 | <AryehGregor> | Philip`, maybe you should put in in <iframe sandbox>. That will encourage people to upgrade to browsers that support that feature, if they want to read your blog. |
| 00:01 | <annevk> | if <iframe sandbox doc=> was there and a couple of browsers supported it... |
| 00:02 | <annevk> | would be neat |
| 00:02 | <gsnedders> | I should probably add more to http://gsnedders.com/php-grievances |
| 00:02 | <AryehGregor> | gsnedders, oh, you have a page on PHP grievances? I've been meaning to write one too. |
| 00:02 | <Philip`> | There's still the problem of raising the minimum effort required to comment in order to stop spammers, so I'd probably forbid the letter 't' |
| 00:03 | <gsnedders> | AryehGregor: Yeah, I wrote it in around an hour one day after spending a while hacking on PHP code |
| 00:03 | <AryehGregor> | My objections seem to mostly be different, though. |
| 00:03 | <AryehGregor> | "((bool) '0' === false) is an endless source of bugs." |
| 00:03 | <AryehGregor> | YES. |
| 00:03 | <AryehGregor> | That is possibly my number one hated feature of PHP. |
| 00:03 | <gsnedders> | Also empty('0') === true |
| 00:03 | <AryehGregor> | Same principle, yeah. |
| 00:04 | <AryehGregor> | if (!$userinput) { # guess user didn't input anything } |
| 00:04 | <AryehGregor> | No, oops, guess the user input a literal '0'. |
| 00:04 | <AryehGregor> | How many retarded MediaWiki bugs have been caused by that, ugh. |
| 00:04 | <AryehGregor> | "XYZ fails for a user named '0'." |
| 00:04 | <AryehGregor> | "XYZ fails on the page [[0]]." |
| 00:04 | gsnedders | has for SimplePie test cases for more or less everything being equal to '0'. |
| 00:05 | <AryehGregor> | "XYZ fails when URL parameter W is equal to '0'." |
| 00:05 | <gsnedders> | Also !$userinput is bad anyway, what if it's not set? I don't want no stinkin' E_NOTICE! |
| 00:05 | <AryehGregor> | "Bugs reported to say this are bogus". You should put that as a separate objection, bugs are always marked bogus. |
| 00:05 | <AryehGregor> | gsnedders, I'm assuming it's already known to be set. |
| 00:06 | <AryehGregor> | Anyway, they mark everything bogus if they don't feel like fixing it right now. |
| 00:06 | <gsnedders> | AryehGregor: I had a fix for a thing committed today. |
| 00:06 | <AryehGregor> | I assume they have some policy where people get bonuses based on how many bugs they close or something. |
| 00:06 | <gsnedders> | (Namely, UTF-16 was broken with DOMDocument::loadXML() because it was using null-terminated strings) |
| 00:06 | <AryehGregor> | Or policies about having only a certain number of bugs open. |
| 00:06 | <gsnedders> | (and the XML spec says UTF-16 MUST be supported) |
| 00:06 | <AryehGregor> | Either that, or the PHP developers are just colossal morons. That seems like a plausible hypothesis too. |
| 00:06 | <AryehGregor> | gsnedders, yeah, sometimes they randomly accept things. |
| 00:07 | <gsnedders> | Well, the email to internals for PHP 5 bugs wasn't sent for a while because it had over 1000 things in it, and that broke the script! |
| 00:07 | <AryehGregor> | gsnedders, did you know, for instance, that if a __get is nested at any depth inside another __get, even for a totally different class (IIRC), it silently returns null without even raising a notice? |
| 00:07 | <gsnedders> | Most of the bugs I've reported in the past couple of years aren't bogus. |
| 00:07 | <AryehGregor> | And they closed it bogus. |
| 00:07 | <gsnedders> | No, they aren't. |
| 00:07 | <AryehGregor> | I was continuing my last sentence. |
| 00:07 | <AryehGregor> | Anyway. |
| 00:08 | <AryehGregor> | I'm sure they don't close *everything* bogus, but they close an unreasonable amount. |
| 00:08 | <gsnedders> | There again, several were crashes. They fixed those. |
| 00:08 | <gsnedders> | In PHP 5.3 CVS, before 5.3 shipped! |
| 00:09 | <AryehGregor> | The major problem with PHP is it's accreted, not designed. |
| 00:09 | <gsnedders> | Indeed. |
| 00:09 | <AryehGregor> | Do you know that f(func_get_args()) returns an error? |
| 00:09 | <AryehGregor> | Or doesn't work, anyway? |
| 00:09 | <gsnedders> | Yeah. |
| 00:09 | <gsnedders> | AryehGregor: And yes, I knew about __get. |
| 00:09 | <AryehGregor> | Why? Because some PHP developer said "Eh, it's not worth writing extra code for such a corner case, I'll just throw an error." |
| 00:10 | <AryehGregor> | Because they don't start from a specification and implement it. They just implement whatever they happen to feel like implementing at the time. |
| 00:10 | <gsnedders> | PHP really sucks. |
| 00:10 | <AryehGregor> | There are no coherent principles anywhere in the language. |
| 00:10 | <gsnedders> | Ruby didn't start from a spec. |
| 00:10 | <gsnedders> | I think the reason why Python and Ruby have done so much better is a gate-keeper with a coherent idea for the language. |
| 00:11 | <AryehGregor> | Well, Python does tend to spec things before implementing them, but I agree that's not strictly necessary if the people in charge are sane. |
| 00:11 | <gsnedders> | Right, Pthon does. |
| 00:11 | <gsnedders> | That's why I said Ruby and not Python :) |
| 00:11 | <gsnedders> | Anyhow, I need to pack. |
| 00:11 | <annevk> | feeling productive in the AM |
| 00:12 | <annevk> | fail |
| 00:12 | <Philip`> | I can't think of any successful modern languages that weren't accreted |
| 00:12 | <AryehGregor> | Philip`, is Python not successful, not modern, or accreted? |
| 00:13 | <annevk> | Python libraries are somewhat of a mess |
| 00:13 | <AryehGregor> | Being designed over a long period of time doesn't count as accretion in my books, either. Java or C# or whatever were designed, not accreted. |
| 00:13 | <annevk> | though they fixed some of it in 3.0 it seems |
| 00:15 | <Philip`> | I'm interpreting "accreted" as meaning the same as "evolved", and languages like Python and Java accreted with the guidance of some language design principles and a desire for self-consistency |
| 00:18 | <Philip`> | (The first versions were very different to the current versions - nobody sat down and wrote a complete spec for the current version before implementing any of it) |
| 00:25 | <AryehGregor> | Hmm, what branch cut does Python use for complex exponentiation? It looks like log(z) lies in (-pi, pi], I wonder if that's documented? |
| 00:26 | <AryehGregor> | >>> (-1-0.00001j)**(-1j) |
| 00:26 | <AryehGregor> | (0.043214350405115579-2.1607176989803302e-12j) |
| 00:26 | <AryehGregor> | >>> (-1)**(-1j) |
| 00:26 | <AryehGregor> | (23.140692632779263+0j) |
| 00:26 | <AryehGregor> | >>> (-1+0.00001j)**(-1j) |
| 00:26 | <AryehGregor> | (23.140461227009979-1.1570231570540908e-09j) |
| 00:27 | <AryehGregor> | It doesn't seem to say. :( http://docs.python.org/reference/expressions.html#power |
| 00:28 | <AryehGregor> | Complex numbers are fun! |
| 00:29 | <Hixie> | othermaciej: yeah, but I'm going through them at 50 a day, so 100 is not many |
| 00:30 | Hixie | comes off the phone with t-mobile technical support wondering wtf just happened (blogged it in case anyone can explain) |
| 00:30 | <Hixie> | time to go to work i guess |
| 00:33 | <annevk> | sounds broken |
| 00:34 | <annevk> | you should be able to do most of that through their online store as well btw |
| 00:34 | <annevk> | at least that's how I got my contract with t-mobile in the Netherlands... |
| 00:35 | <annevk> | take two at going to bed... |
| 00:35 | <annevk> | nn |
| 00:35 | AryehGregor | suddenly interpreted that as two nymphs standing in a row for some reason |
| 00:35 | AryehGregor | doesn't usually have NetHack flashbacks |
| 00:36 | <Hixie> | annevk: i did do it through their online store, it said there was a problem and i had to call them |
| 01:01 | <gsnedders> | Wait, you were on the phone!? :o |
| 01:02 | <gsnedders> | Hah. Fun. |
| 01:02 | gsnedders | sighs |
| 01:03 | <gsnedders> | I never knew wardrobes could be so painful to look inside. Oh well. |
| 02:02 | <Hixie> | othermaciej: should a reset button in a form reset the <progress> elements in that form? |
| 02:05 | <Hixie> | and should <progress> and <meter> elements appear in form.elements? |
| 02:08 | <deltab> | do they have defaultValue? |
| 02:19 | <Hixie> | deltab: no |
| 02:19 | <Hixie> | deltab: though i guess they could |
| 02:30 | <Hixie> | ok well lacking any reply, i'll go with no resetting and not in .elements |
| 03:14 | <Dashiva> | Imagine how much more accessible the @summary discussion would be if each post had a summary |
| 03:26 | karlcow | tries to imagine summary of summaries |
| 03:37 | <Dashiva> | Well, most posts would summarize to "I'm making argument X". The thread itself would be "People made arguments X, Y and Z" |
| 04:04 | <karlcow> | Dashiva: and everyone in the thread has tendency (and that's normal) to summarize (no pun) from his/her point of view or competences. |
| 04:06 | <karlcow> | implementability, implementation landscape, practice, society constraints (for accessibility legal frameworks), etc. are different flavors of implementations |
| 04:24 | <Hixie> | MikeSmithX: i'm stuck with cvs.bin commit: [04:30:17] waiting for www-data's lock in /sources/public/html5/spec |
| 04:24 | <Hixie> | for like an hour now |
| 04:25 | <MikeSmithX> | Hixie: OK, I clearly need to turn off my script |
| 04:25 | <MikeSmithX> | will do it right now |
| 04:37 | <othermaciej> | Hixie: does <output> have a defaultValue? |
| 04:52 | <MikeSmithX> | Hixie: it turns out the problem is not my script, but some problem on the cvs server.. I'm trying to figure it out now, but not clear to me how locking is handled in this cvs server instance |
| 05:22 | <zcorpan_> | "HTML5 has been an absolute pleasure to use. It’s the first HTML spec that has made complete sense to me." -- http://minimali.st/2010/01/letterpress/ |
| 05:32 | <Hixie> | MikeSmithX: yes |
| 05:33 | <Hixie> | MikeSmithX: k |
| 05:34 | <Hixie> | s/MikeSmithX: yes/othermaciej: yes/ |
| 05:34 | <Hixie> | zcorpan_: nice |
| 05:56 | <Hixie> | sigh, we don't even have interop on window.atob() |
| 05:57 | <zcorpan_> | that's why we need a spec :) |
| 05:59 | <cardona507> | Hixie - do you need any more graphic work for the spec? If so let me know - that was fun |
| 06:00 | <Hixie> | cardona507: i'll keep that in mind :-) |
| 06:00 | <Hixie> | cardona507: nothing springs to mind, but i'm sure we can find things |
| 06:00 | <zcorpan_> | cardona507: there's some ascii art in the spec currently i think |
| 06:01 | <cardona507> | zcorpan_ any idea which parts? |
| 06:02 | <zcorpan_> | 6.8.2.2 Sample user interface |
| 06:05 | <cardona507> | firing up adobe illustrator...... |
| 06:14 | <Hixie> | zcorpan_: ooh, good call |
| 06:14 | <Hixie> | cardona507: don't do the ascii art in the websockets section btw, that has to remain ascii art for the ietf |
| 06:15 | <cardona507> | Hixie - but 6.8.2.2 sample user interface is cool? |
| 06:15 | <Hixie> | yup |
| 06:15 | <cardona507> | alrighty |
| 06:15 | <Hixie> | feel free to take artistic license too |
| 06:16 | <Hixie> | (e.g. let me know if i need to change the sample text around it to make it better fit whatever you come up with) |
| 06:17 | <cardona507> | ok - thanks |
| 06:24 | <Hixie> | wait, atob() is only in Gecko and Webkit! |
| 06:24 | <Hixie> | IE and Opera don't even support it |
| 06:25 | <zcorpan_> | Hixie: you need a newer opera |
| 06:25 | <Hixie> | Opera stopped autoupdating for me |
| 06:25 | <Hixie> | it says something about how it downloaded it, but won't update |
| 06:26 | <Hixie> | because it's not feeling well, or can't be bothered, or something |
| 06:26 | <zcorpan_> | weird |
| 06:26 | <zcorpan_> | try http://my.opera.com/desktopteam/blog/happy-new-year |
| 06:27 | <Hixie> | (and i can't find any ui anywhere to force an update) |
| 06:29 | <Hixie> | will that autoupdate? |
| 06:32 | <zcorpan_> | i guess it should... but it's the latest build currently and it has atob if you want to test it |
| 06:34 | <Hixie> | since IE doesn't have it and since it really should be in the JS spec i punted on it |
| 06:50 | <Hixie> | MikeSmithX: (i'm skipping cvs commits for now, since it's still blocking) |
| 07:05 | <cardona507> | Hixie, zcorpan_ - feedback? http://img685.imageshack.us/img685/7346/whatwg68ex101.png |
| 07:06 | <Hixie> | looks fantastic |
| 07:07 | <cardona507> | thanks |
| 07:12 | <Hixie> | tis checked in :-) |
| 07:14 | <cardona507> | that was quick - the image isn't working for me in safari... |
| 07:14 | <Hixie> | yeah i had a typo, fixing it already :-) |
| 07:31 | <zcorpan_> | Hixie: doesn't work in multipage |
| 07:31 | <Hixie> | yeah i accidentally killed the commit script before it called the multipage generator |
| 07:31 | <Hixie> | it'll be fixed when i do my next checkin |
| 07:32 | <zcorpan_> | ok |
| 07:56 | <Hixie> | i can't find any actual examples of <ms> use on the web (mathml element) |
| 07:56 | <Hixie> | i mean there are trivial examples, but nothing realistic |
| 07:59 | <Hixie> | cardona507: there's another bit of ascii art in that section, btw, in case you're still looking for things to do :-) |
| 08:00 | <cardona507> | but of course :) - what section? |
| 08:00 | <zcorpan_> | cardona507: same section |
| 08:01 | <Hixie> | same one, search for "||||||||||||||||" |
| 08:01 | <cardona507> | oh yeah - I am finished with that one and uploading it for feeback right now - |
| 08:01 | <cardona507> | got any more? |
| 08:01 | <Hixie> | sweet |
| 08:01 | <Hixie> | let's see |
| 08:04 | <Hixie> | if you want to try making stuff up, you could try making up examples of what some of hte form controls could look like |
| 08:04 | <cardona507> | http://img269.imageshack.us/img269/2720/whatwg68ex201.png - is that how you thought the "no application selected" area would look? or will it be an input? |
| 08:04 | <Hixie> | like the date controls and stuff |
| 08:04 | <cardona507> | from what section Hixie? |
| 08:04 | <Hixie> | (though that's somewhat of a rat's nest of stuff) |
| 08:05 | <Hixie> | the <input> element and its subsections |
| 08:05 | <cardona507> | ahhh - ok I'll check it out - |
| 08:06 | <Hixie> | looks good to me |
| 08:06 | <Hixie> | let me just finish the edit i'm doing now and i'll put that one in too :-) |
| 08:06 | <cardona507> | cool - thanks ian |
| 08:07 | <Hixie> | don't thank me, you're the one doing the work :-) |
| 08:08 | <zcorpan_> | hmm, should i add <device> to html5-elements? |
| 08:08 | <Hixie> | i wouldn't yet |
| 08:08 | <Hixie> | it's not even in the whatwg html spec's index yet |
| 08:08 | <zcorpan_> | ok |
| 08:12 | <zcorpan_> | /me notices a typo in html5-elements |
| 08:17 | <Hixie> | ok, checking in the new image cardona507 |
| 08:18 | <cardona507> | wow - 23 input types - sweet - much better than the old 7 |
| 08:19 | <Hixie> | hehe |
| 08:19 | <Hixie> | some have some images already if you need some inspiration |
| 08:21 | <Lachy> | Hixie, re your blog, what possessed you to finally give in and get yourself a phone? |
| 08:22 | <Hixie> | it's not for me |
| 08:22 | <Hixie> | :-) |
| 08:23 | <Hixie> | i was getting a sim for the nexus one that i'm giving my girlfriend |
| 08:23 | <Lachy> | ok |
| 08:23 | <Hixie> | to replace her horrific verizon phone |
| 08:24 | <cardona507> | how is the nexus one? android 2.0 is pretty HTML5 friendly isn't it? |
| 08:24 | <Lachy> | so, how come they wanted to charge you $70 for something that should have added up to $61.24? |
| 08:24 | <Hixie> | cardona507: seems pretty good to me, but i haven't played with it much and i hate phones in general, so... :-) |
| 08:25 | <Hixie> | lachy: i honestly didn't understand most of the phone call |
| 08:25 | <Hixie> | Lachy: i have not much clue what i'm actually going to receive |
| 08:25 | <Hixie> | we'll see |
| 08:28 | <cardona507> | Hixie - are they too super obvious for a graphic or do we need one for type="text" password etc? I don't mind creating them since they will be pretty simple |
| 08:28 | <Hixie> | personally i wouldn't bother with the simple ones |
| 08:28 | <cardona507> | ok |
| 08:29 | <Hixie> | but then personally i wouldn't bother with any of them because i'm lazy :-) |
| 08:29 | <cardona507> | ha |
| 08:29 | <Hixie> | my recommendation would be that you pick a few of the more interesting ones, and make some very different examples for them |
| 08:29 | <Hixie> | so e.g. type=range you could should a slider and a twisting knob thing |
| 08:30 | <Hixie> | oh another element that would be good to show examples of is <meter> -- you could show several different kinds of gauges |
| 08:30 | <cardona507> | good ideas |
| 08:30 | <Hixie> | whatever example images you come up with i can figure out some story around them to explain what the image is showing |
| 08:31 | <cardona507> | sounds good |
| 08:31 | <Hixie> | feel free to slide in some easter eggs btw -- a great many of the examples in the spec have secret messages or are referencing memes, or shows or books i like, or my friends, or whatever :-) |
| 08:31 | <Hixie> | gives the spec some levity |
| 08:33 | <cardona507> | hehe I will definitely take that advice- any examples that aren't too secret? |
| 08:34 | <Hixie> | search for "hedral" |
| 08:34 | <Hixie> | he's my cat :-) |
| 08:34 | <Hixie> | there's all kinds of examples with him in it |
| 08:34 | <Hixie> | you even made one :-P |
| 08:34 | <Hixie> | well the graphic you made didn't mention hedral, but it went with an example that did |
| 08:35 | <cardona507> | is he the inspiration from the validhtml5 button? hmmm - I didn't notice his name before |
| 08:35 | <Hixie> | nah the validhtml5 button is Astrophy http://hixie.ch/resources/images/astrophy |
| 08:35 | <hsivonen> | clearly, people who file bugs from the spec UI need more unique initials |
| 08:35 | <Hixie> | (not a real cat) |
| 08:37 | <zcorpan_> | maybe you can make a kettle with two gauges: one for amount of liquid and one for temperature |
| 08:38 | <zcorpan_> | dunno if an HTML kettle is useful for anything, but still |
| 08:38 | <cardona507> | ahhh I see now - from earlier - <option value="hedral⊙dc"> |
| 08:48 | <hsivonen> | Hixie: isn't http://html5.org/tools/web-apps-tracker?from=4532&to=4533 backwards-incompatible? |
| 08:48 | <Hixie> | with what? |
| 08:49 | <hsivonen> | generic rel munging |
| 08:49 | <hsivonen> | probably doesn't matter for any concrete piece of software |
| 08:49 | <Hixie> | if there are any cases that this breaks, let me know and i'll fix it |
| 08:49 | <Hixie> | the change is needed to make rdfa not break |
| 09:05 | <payman> | gsnedders: i know! |
| 09:17 | <Philip`> | cardona507: Graphical examples of canvas features (e.g. line caps and joins, and radial gradients) might be handy :-) |
| 09:17 | <cardona507> | Philip` - good to know - that will be my next project :) |
| 09:18 | <MikeSmith> | Hixie: as soon as I get resolution on the lock problem, I'll let you know |
| 09:21 | <zcorpan_> | cardona507: http://www.nihilogic.dk/labs/canvas_sheet/HTML5_Canvas_Cheat_Sheet.png has some canvas examples |
| 09:21 | <cardona507> | thanks zcorpan_ |
| 09:21 | <cardona507> | wow - nice link :) |
| 09:27 | <othermaciej> | Hixie: I think maybe bugs that look like total junk could be INVALID instead of NEEDSINFO (e.g. <http://www.w3.org/Bugs/Public/show_bug.cgi?id=8654>;) |
| 09:33 | <Hixie> | ok |
| 09:34 | <cardona507> | Hixie - how about this for a <input type="range"> between 100 and 700 http://img140.imageshack.us/img140/5747/whatwgrangeex101.png |
| 09:36 | <Hixie> | cardona507: i think i'd make that two examples -- the top one and the bottom one, separate, without the number field in the middle |
| 09:36 | <cardona507> | ok |
| 09:36 | <Hixie> | cardona507: (in practice browsers likely would not make a type=range be that big) |
| 09:36 | <cardona507> | gotcha |
| 09:36 | <zcorpan_> | cardona507: looks like the thumb is 100 steps too far to the right |
| 09:36 | <cardona507> | yeah -hehe - I just noticed |
| 09:36 | <zcorpan_> | cardona507: or the current value should say 607 instead |
| 09:37 | <MikeSmith> | Hixie: does the change for UAs to ignore charset in <style type="text/css; charset=xxx"> (r4520, bug 8556) not have any possible implications for conformance checkers? |
| 09:37 | <cardona507> | I am just gonna remove the value field and make it 2 examples |
| 09:37 | <Hixie> | MikeSmith: i guess it would, yeah, didn't really think of that |
| 09:38 | <Hixie> | afk |
| 09:39 | <cardona507> | potential range #1 http://img262.imageshack.us/img262/5747/whatwgrangeex101.png |
| 09:40 | <cardona507> | potential range #2 http://img267.imageshack.us/img267/9901/whatwgrangeex202.png |
| 09:40 | <zcorpan_> | looks good |
| 09:40 | <cardona507> | thanks :) |
| 09:40 | <zcorpan_> | should #2 indicate current value somehow? |
| 09:41 | <cardona507> | sounds good to me - what do you have in mind? |
| 09:41 | <cardona507> | a little down arrow along the top and center? |
| 09:42 | <zcorpan_> | yeah, maybe |
| 09:42 | <cardona507> | ahhh - I know - 1sec |
| 09:43 | <tametick> | greetings |
| 09:43 | <tametick> | how well supported is the audio api in current browsers? is it about comparable to canvas support? |
| 09:44 | <cardona507> | zcorpan_ http://img686.imageshack.us/img686/4175/whatwgrangeex20202.png better???? |
| 09:45 | <zcorpan_> | tametick: canvas came earlier |
| 09:45 | <zcorpan_> | cardona507: ooh nice |
| 09:45 | <cardona507> | Hixie - your opinion? |
| 09:46 | <tametick> | zcorpan_: http://en.wikipedia.org/wiki/Comparison_of_layout_engines_(HTML_5)#Elements -> seems to suggest that newer ff/chrome/safari/opera should support ogg & wav |
| 09:46 | <zcorpan_> | tametick: current opera has basic audio support based on old spec (before there was an <audio> element), only supporting wave pcm |
| 09:46 | <tametick> | hrm |
| 09:47 | <Hixie> | cardona507: looks good, i'll add them in tomorrow (or later tonight if i'm up to it) |
| 09:47 | <zcorpan_> | tametick: opera 10.50 will have updated support (including ogg/vorbis) |
| 09:47 | <cardona507> | tametick - firefox - wav & ogg - safari - mp3 - chrome - wave and mp3 (not sure on chrome) |
| 09:47 | <cardona507> | Hixie - no rush of course - I was just wondering if you had any change suggestions |
| 09:47 | <tametick> | cardona507: well that sucks... so there is no format that is supported on all 3? |
| 09:47 | <othermaciej> | cardona507: I'm pretty sure Safari supports wav as well |
| 09:48 | <othermaciej> | we support nearly every audio format that QuickTime supports |
| 09:48 | <Hixie> | cardona507: looks good to me |
| 09:48 | <Hixie> | gotta go, bbiab |
| 09:48 | <othermaciej> | so wav, mp3, AAC should all work |
| 09:49 | <tametick> | othermaciej: how about ogg? |
| 09:49 | <othermaciej> | by default no, but if you install the Ogg codec for QuickTime it will work |
| 09:49 | <cardona507> | nope - you need to do <audio controls><source src=".wav"><source src=".ogg"></audio> - tametick |
| 09:49 | <othermaciej> | (most users don't have that though) |
| 09:49 | <hsivonen> | tametick: http://xiph.org/quicktime/ |
| 09:49 | <othermaciej> | I would hope MP3 will be universally available soon since all remaining patents on it expire in at most a few years |
| 09:50 | <doublec> | cardona507: don't forget the "type=.." in the source |
| 09:50 | <cardona507> | that would be soooo nice - wav and ogg are great and everything but.... |
| 09:50 | <hsivonen> | kinda sad to go back to MP3 when Vorbis offers better compression |
| 09:50 | <cardona507> | doublec - I had only used type with video - |
| 09:51 | <zcorpan_> | which browser does not support wave? |
| 09:51 | <tametick> | cardona507: so in that case it will play wav if it can and ogg if it can't? |
| 09:51 | <cardona507> | but yeah - it would make sense |
| 09:51 | <cardona507> | tametick - yep |
| 09:51 | <tametick> | ok thanks |
| 09:51 | <othermaciej> | as does AAC; but there's probably a lot more MP3s out there than any other format |
| 09:51 | <cardona507> | if you put type like doublec recommended above then the browser will only download it if it is the correct type |
| 09:52 | <cardona507> | hsivonen - does vorbis offer better? |
| 09:52 | <zcorpan_> | i thought almost all lossy audio formars offer better compression than mp3 |
| 09:53 | <Philip`> | There weren't many lossy audio formats before MP3, as far as I'm aware |
| 09:53 | <Philip`> | and nobody would have bothered making ones afterwards that were worse |
| 09:53 | <zcorpan_> | indeed |
| 09:54 | <hsivonen> | cardona507: I don't know about the relative technical merits of Vorbis and AAC. However, no one is collecting royalties for Vorbis. |
| 09:54 | <othermaciej> | I think all the MP3 US patents expire on 2012 except for non-essential ones on optional ways to improve encoding quality |
| 09:54 | <tametick> | cardona507: actually i only need to use the audio api for sound effects (that need to play fast with as little latency as possible), i guess i could use that wav/ogg solution for that and just use plain old <embed> for background music, since latency is a lot less important for it |
| 09:54 | <tametick> | even midis would do |
| 09:54 | <tametick> | whatever is most supported |
| 09:54 | <othermaciej> | *in 2012 |
| 09:55 | <cardona507> | hsivonen - I agree completely |
| 09:55 | <othermaciej> | hopefully Mozilla and Opera will no longer have a problem with it at that point |
| 09:56 | <cardona507> | isn't there a whisper in the wind that google will release On2's codec as open? or is that only video - and strictly a rumor of course |
| 09:56 | <othermaciej> | (in fact I'm not sure there's any live patents on MP3 *de*coding any more, although IANAL) |
| 09:56 | <othermaciej> | no one really knows what Google will do with On2 |
| 09:56 | <cardona507> | yeah I figured - but it was worth a shot :) |
| 09:56 | <Philip`> | cardona507: Only video |
| 09:57 | <othermaciej> | but On2 has supposedly already granted whatever patents they have on their codecs (or rather made a non-assertion prmise) |
| 09:57 | <othermaciej> | well, at least as to VP3.2 anyway |
| 09:57 | <othermaciej> | VP3.2 being the ancestor of Ogg Vorbis |
| 09:57 | <othermaciej> | the worry about those is that other parties may have patents that are infringed |
| 09:58 | <hsivonen> | othermaciej: ITYM Theora |
| 09:58 | <othermaciej> | er |
| 09:58 | <othermaciej> | yeah |
| 09:58 | <hsivonen> | (it seems plausible that On2 has patented stuff in the VP series after 3.2) |
| 09:58 | <othermaciej> | Theora |
| 09:58 | <hsivonen> | s/has/may have/ |
| 09:58 | <cardona507> | perhaps google has something in house already that they needed On2 brains to dial in... |
| 09:59 | <Philip`> | VP3 was somewhat obsolete before they opened it |
| 09:59 | <hsivonen> | kinda useless to speculate now, though, since the shareholder vote is in February |
| 09:59 | <Philip`> | and that was eight years ago |
| 09:59 | hsivonen | wonders if they vote to adjourn again |
| 09:59 | <Philip`> | and they seem to have done a lot of new codecs since then |
| 10:00 | <othermaciej> | I wonder if it is a standard feature of corporate bylaws that merger requires a majority of shares outstanding, not just shares voted |
| 10:00 | <othermaciej> | I guess for a major company with lots of institutional investors it is unlikely to be a problem, but it still seems like poor design |
| 10:03 | <zcorpan_> | looks like most bugs are editorial, looking at the past 100 checkins |
| 10:04 | <othermaciej> | which is good for the stage we are at |
| 10:04 | <Philip`> | I'd expect there's a significant bias because editorial bugs are far easier to spot than serious bugs |
| 10:05 | <othermaciej> | if you're looking at checkins, another possible source of bias is that a greater proportion of non-editorial bugs get rejected |
| 10:06 | <zcorpan_> | and maybe also easier for Hixie to fix |
| 10:06 | <zcorpan_> | or postponed |
| 10:07 | <Philip`> | Maybe it's a sign that there's a need for test cases now, to find errors in the spec that aren't obvious from simply reading it |
| 10:08 | <zcorpan_> | and implementations |
| 10:09 | <hsivonen> | https://bugzilla.mozilla.org/show_bug.cgi?id=537948 might turn out to be a spec bug (could be an implementation bug, though) |
| 10:11 | <roc> | speaking for myself I'd like to support MP3 the day the patents run out |
| 10:12 | <othermaciej> | roc: perhaps you should have your legal dept do the search on which patents are applicable to decoding and not facially invalid due to the Mp3 spec being prior art |
| 10:12 | <othermaciej> | roc: the longest-running ones go to 2017 but it looks like all the ones that are not expired already are non-essential encoder tricks |
| 10:13 | <othermaciej> | and in theory anything covered in the spec itself shouldn't be able to have a patent beyond 2012 |
| 10:14 | <annevk> | has anyone implemented the document.domain blocking for storage? |
| 10:14 | <annevk> | we don't really like it apparently |
| 10:14 | <annevk> | too much additional security checks needed all over the place |
| 10:15 | <annevk> | also the spec seems somewhat wrong for the case where you do |
| 10:15 | <annevk> | var store = localStorage; document.domain = "example.org"; store.item = "foo" // does not throw |
| 10:16 | <Hixie> | othermaciej: givn the eolas patent, prior art doesn't seem to hold much weight with the courts these days |
| 10:17 | <Hixie> | annevk: please file a bug |
| 10:25 | <annevk> | I will after I somewhat more carefully check our position |
| 10:25 | <annevk> | well "our" |
| 10:30 | <Hixie> | well there are various choices to solve the problem |
| 10:30 | <Hixie> | we could make a storage area for the "wildcard" port |
| 10:30 | <Hixie> | but that seemed like a high level of complexity given that we're trying to discourage document.domain |
| 10:31 | <annevk> | "wildcard" port? |
| 10:31 | <annevk> | what's the problem with multiple storage areas? |
| 10:32 | <jgraham> | Hixie: BTW I have no confidence whatsoever that the ES people will want to specify btoa and atob given their typical attriude towards the legacy |
| 10:34 | <othermaciej> | they do have a bad attitude |
| 10:34 | <othermaciej> | but if IE doesn't have them, then they might not really be needed for Web compat |
| 10:34 | <annevk> | maybe IE has equivalent methods people use |
| 10:35 | <annevk> | and the methods are somewhat useful |
| 10:37 | <Hixie> | annevk: once you set document.domain, the origin becomes something with a wildcard port |
| 10:37 | <Hixie> | ok seriously microsoft, wtf. i need to install silverlight to watch the CES keynote video?! |
| 10:37 | <jgraham> | othermaciej: It is not clear if they are really needed but at least Webkit and Opera have felt it useful to add them when they didn't already exist |
| 10:38 | <Hixie> | jgraham: well if they refuse, reopen the bug and i'll do it |
| 10:38 | <othermaciej> | jgraham: we added a lot of Firefox JS extensions kind of reflexively back in the day |
| 10:39 | <annevk> | Hixie, but we also still know the original origin |
| 10:39 | <annevk> | Hixie, could just continue using that |
| 10:39 | <Hixie> | annevk: no, that would result in a mutex deadlock |
| 10:40 | <annevk> | Hixie, and not let document.domain have any effect whatsoever |
| 10:40 | <Hixie> | annevk: since changing document.domain lets you call scripts cross-origin synchronously |
| 10:40 | <annevk> | aah, that was the reason |
| 10:40 | <annevk> | I see |
| 10:43 | <Philip`> | http://www.tagesspiegel.de/magazin/werbinich/wir-muessen-reden/ - if(window.pkcs11){ if(document.ids){ window.location = "/errorpages/ns6.html"; } else if(!window.atob) window.location = "http://www.tagesspiegel.de/errorpages/ns6.html";; } |
| 10:45 | <Hixie> | o_O |
| 10:49 | <Hixie> | anyone know of a good example of where a disclosure triangle is visible in stock OSX? |
| 10:50 | <othermaciej> | yes |
| 10:50 | <othermaciej> | Finer Get Info window |
| 10:50 | <Hixie> | ooh good idea |
| 10:50 | <othermaciej> | hit Cmd-I with a file selected and you can see a bunch of them in a row |
| 10:50 | <Philip`> | http://www.hfxmitterer.de/go.to/modix/3,1,rxnw3mz/fahrzeugsuche.html?sort=&ftyp=>yp=&fart=&version=320 - if(document.defaultCharset && !window.atob) BrowserIsIE = true; |
| 10:51 | <Philip`> | So far, it looks like the implementation of the function is irrelevant, all that matters is whether it exists or not |
| 10:54 | <Hixie> | wtf, my desktop seems to be hung |
| 10:55 | <Hixie> | finder works fine except it won't react to clicks on the desktop |
| 10:55 | <Hixie> | and new files aren't showing up |
| 10:57 | <Philip`> | Those two sites are the only ones that use atob/btoa, out of 425K pages |
| 11:31 | <Hixie> | there. now pillar can no longer be jealous that hedral's the only one with all the examples. |
| 11:40 | <Hixie> | othermaciej: how should i handle bugs that are just questions, e.g. http://www.w3.org/Bugs/Public/show_bug.cgi?id=8661 ? (I mean, other than answering the question, which I will do, obviously.) |
| 11:40 | <Hixie> | reject? partial accept? "i don't understand"? |
| 11:41 | <othermaciej> | Hixie: not really sure |
| 11:41 | <othermaciej> | Hixie: this one has an implied spec change based on a faulty premise in addition to the question |
| 11:41 | <othermaciej> | i.e. it says there is a contradiction which actually there is not |
| 11:42 | <othermaciej> | so I'd handle this one on that basis (in addition to answering the question) |
| 11:43 | <othermaciej> | if a bug really just asked a question without any report of a problem or request for a change (even implied), then I would consider that INVALID, same as garbage text or a completely off-topic comment |
| 11:43 | <othermaciej> | or maybe NEEDSINFO if it seemed like the person was getting at something but failed to express themselves |
| 11:43 | <Hixie> | k |
| 11:50 | <Hixie> | man i'm not even keeping up with the incoming bugs today |
| 11:50 | <Philip`> | Clearly we need more editors, since you can't cope with the volume |
| 11:55 | <Hixie> | i know we need more editors |
| 11:55 | <Hixie> | we have piles of specs that are languishing |
| 11:56 | <Hixie> | you don't happen to be looking for a job or anything are you? :-) |
| 11:57 | jgraham | has lost track of how much PhD Philip` has left to do |
| 11:58 | <jgraham> | Is it supposed to be finished this September or next September)? |
| 11:58 | <Dashiva> | I can hardly keep up with the bugmail :) |
| 11:58 | <Philip`> | I've lost track of it too |
| 11:59 | <Hixie> | hah |
| 11:59 | <Hixie> | Philip` is firmly ensconced in academia i see |
| 11:59 | <Philip`> | I started a bit over two years ago, so it'll be however long it takes |
| 12:00 | <jgraham> | So I guess it is, in theory, supposed to be finished this September |
| 12:00 | <jgraham> | Or that's when they stop giving you money or something |
| 12:00 | <jgraham> | (and then a while after that they get really annoyed) |
| 12:00 | <Philip`> | The initial funding was for 3 years |
| 12:01 | <Philip`> | though it seems pretty rare for people to finish in that time |
| 12:01 | <Philip`> | They get really annoyed if you take more than 4 years, I think |
| 12:01 | <Philip`> | (or was it 5?) |
| 12:01 | <jgraham> | I think it's 4 |
| 12:02 | <jgraham> | (I got 4 years of funding but I really really had to finish in 4 even so) |
| 12:02 | Philip` | should probably know more about this than he does |
| 12:03 | <Hixie> | ok i give up, going to bed! |
| 12:05 | <Philip`> | Anyway, I suppose it should be finished in a finite time, and then I'll need to work out what to do next :-) |
| 12:05 | <Hixie> | nn |
| 12:06 | <jgraham> | gn |
| 12:06 | <jgraham> | Philip`: "A finite time" isn't much of a constraint |
| 12:07 | <Dashiva> | Given that it's somewhat impossible to finish something in infinite time :P |
| 12:07 | <jgraham> | Dashiva: You could finish it in the limit as t->inf |
| 12:08 | <Dashiva> | Once you finish it, t stops approaching inf :) |
| 12:36 | <workmad3> | Dashiva: working out the decimal value of pi takes infinite time |
| 12:37 | <Dashiva> | And you won't finish |
| 12:37 | <workmad3> | sure you will, once t reaches infinity :P |
| 12:38 | <Philip`> | How do you know you won't reach the end of pi and get nothing but 0 digits after it? |
| 12:39 | <workmad3> | Philip`: pretty sure it's been proven that pi doesn't end |
| 12:39 | <workmad3> | something to do with it being a transcendental number or something |
| 12:39 | <workmad3> | same with e |
| 12:39 | <Philip`> | Bah, that's just a proof |
| 12:40 | <workmad3> | yeah, what do those mathematicians that spent their life working on such things know? :P |
| 12:40 | <Philip`> | The mathematicians will be sorry when somebody bothers to print out the end of pi and realises it's all 0s |
| 12:41 | <workmad3> | Philip`: yeah, the mathematicians will probably come up with some excuse, along the lines of a computer only has finite precision or something :P |
| 12:41 | <Dashiva> | But they'll have to print an infinite number of 0s still, so it doesn't change much :P |
| 12:41 | <Philip`> | They could compress it |
| 12:44 | <Dashiva> | But then they'd have to check all the digits first to make sure they were zero |
| 13:42 | <hsivonen> | Hixie: instead of LATER, shouldn't bugs remain open in a HTML6 or somesuch bugzilla component? |
| 13:45 | <zcorpan_> | there's an "HTML future versions" component |
| 14:08 | <zcorpan_> | is there an ogg video that i can use for a demo? |
| 14:08 | <zcorpan_> | preferably a movie trailer or similar |
| 14:09 | <annevk> | http://www.bluishcoder.co.nz/2008/10/html-5-video-element-examples.html links a bunch |
| 14:11 | <zcorpan_> | thanks |
| 14:16 | <annevk> | has anyone actually implemented the storage mutex? |
| 14:16 | <annevk> | or the document.domain localStorage thing? |
| 14:16 | <annevk> | it seems like it's all #fail |
| 14:21 | <hsivonen> | does Safari have localStorage? |
| 14:22 | <annevk> | yes |
| 14:22 | <hsivonen> | what about Opera? |
| 14:22 | <annevk> | 10.5 |
| 14:22 | <hsivonen> | so it's pretty much everywhere--and without the mutex? |
| 14:23 | <annevk> | yes |
| 14:23 | <annevk> | or the document.domain stuff |
| 14:23 | <annevk> | afaict |
| 14:24 | <annevk> | (based on very limited testing and inspection) |
| 14:25 | <hsivonen> | on a totally different topic: if I want IMAP+Webmail that doesn't suck, what should I be evaluating besides Zimbra and Gmail? |
| 14:26 | annevk | only knows about Gmail being nice for Webmail |
| 14:27 | <hsivonen> | I'm concerned that if something goes wrong with my account and I lose access, I'm paying Google nothing and I'm one among millions, so that Google doesn't have enough of a reason to care and I have no recourse |
| 14:28 | <hsivonen> | also, I'm not too thrilled about having all my email analyzed |
| 14:28 | <hsivonen> | moreover, I care about the Freedom to Leave, so I should have at least existence proof of where to I could leave |
| 14:30 | <hsivonen> | self-hosting Zimbra seems like a total overkill |
| 14:31 | <annevk> | you have some kind of mail server with your host right? |
| 14:31 | <annevk> | you could always move there |
| 14:32 | <Lachy> | hsivonen, if you don't like self-hosting, are there any problems with the Zimbra accounts they host for you? |
| 14:32 | <hsivonen> | annevk: only if quota there is large enough and Webmail there doesn't suck |
| 14:32 | <Philip`> | hsivonen: Since you're one among millions, a problem will either affect lots of people and therefore be important to fix quickly, or it will affect one person and there's a million-to-one chance it won't be you |
| 14:32 | <hsivonen> | Lachy: who's "they"? |
| 14:33 | <Lachy> | I guess, what they call "ZCS hosted email by a Zimbra partner" - http://www.zimbra.com/learn/ |
| 14:34 | <annevk> | hsivonen, so what do you do if competition is not good enough? |
| 14:35 | <hsivonen> | Lachy: the list of partners looked very sad in terms of choice of jurisdiction and descriptions of services on the Web |
| 14:35 | <hsivonen> | annevk: I haven't yet decided what to do |
| 14:36 | <hsivonen> | it bothers me that I haven't found an offering for email&Jabber hosting that were competitive with Google Apps for Domains |
| 14:36 | <hsivonen> | open protocols aren't much good if Google doesn't have other providers to federate with |
| 14:37 | <hsivonen> | if Wave takes off, it'll be even harder to get non-Google email+Jabber+Wave hosting without being an enterprise with staff for figuring it out |
| 14:38 | <annevk> | lol |
| 14:38 | <zcorpan_> | /me is reminded about the google toilet video |
| 14:38 | <annevk> | http://www.brucelawson.co.uk/2010/html5-it-is-a-changin/ is funny |
| 14:38 | <hsivonen> | it not only bothers me that I don't know how to solve this for me |
| 14:39 | <hsivonen> | but I also don't have a non-Google solution to point family members to |
| 14:39 | <annevk> | seems sort of natural for Mozilla to enter this space |
| 14:39 | <annevk> | especially with their statement on the homepage |
| 14:40 | <annevk> | "We believe that the internet should be public, open and accessible." |
| 14:40 | <annevk> | hsivonen, though it's a huge fail, I guess you could look into Mobile Me |
| 14:40 | <annevk> | from Apple |
| 14:41 | Lachy | feels compelled to mention OperaMail |
| 14:41 | Lachy | runs... |
| 14:41 | <jgraham> | Is that the one with 2Mb of space? |
| 14:41 | <Lachy> | no. |
| 14:41 | <Lachy> | 3MB |
| 14:41 | <Philip`> | How about Outlook? |
| 14:42 | <annevk> | does that still exist? |
| 14:42 | Philip` | shrugs |
| 14:42 | <annevk> | right :p |
| 14:42 | <Philip`> | All the other browser vendors got mentioned so Microsoft shouldn't be left out :-p |
| 14:42 | <Lachy> | Philip`, don't you mean ExchangeServer? Outlook is the client |
| 14:44 | <hsivonen> | Philip`: I don't trust Microsoft's commitment to open specs enough to use their hosted services. I don't want to find that one day IMAP has been replaced by Exchange protocols and Webmail by Silverlightmail |
| 14:44 | <hsivonen> | and I don't trust Yahoo!'s corporate longevity enough |
| 14:44 | <Lachy> | Microsoft have already tried to replace IMAP with their own protocols |
| 14:45 | <Lachy> | and they've largely succeeded, given how pervasive Exchange is in the enterprise market, and why Outlook is the most popular client because there isn't much else that's compatible with it |
| 14:45 | <jgraham> | s/tried to/succeeded in/ for the purposes of many large intranets |
| 14:45 | <Lachy> | though Apple Mail is now I think they've licensed it |
| 14:45 | <jgraham> | Oh you said that |
| 14:47 | <karlcow> | hsivonen: I host my own (for mail servers) but didn't install any webmail, which I found suboptimal. But I may be a very particular case. |
| 14:47 | <hsivonen> | also, as a matter of principle, I think one should try to minimize the probability that a communication host provides a feed to foreign three-letter agencies |
| 14:47 | <hsivonen> | otoh, hosting in Finland means that even the regular police can take your servers at a whim |
| 14:47 | <annevk> | or four :) |
| 14:48 | <TabAtkins> | To be honest, having the TLAs be homegrown doesn't make it any better. |
| 14:48 | <Lachy> | karlcow, do you have your own servers, or do you mean you rent a server from some company like Dreamhost or whatever, who have mail set up on them for you? |
| 14:48 | <karlcow> | I have my own physical machine (1U) in a bay. |
| 14:48 | <TabAtkins> | Though I suppose we Americans can at least pretend we can change something about their operation. |
| 14:49 | <Lachy> | nice. Is that stored at some hosting facility, or run it from home? |
| 14:49 | Lachy | wanted to do that one day, but too expensive right now |
| 14:49 | <karlcow> | I have had a machine for now… hmmm… more than 10 years. not from home, and had different hosting facilities. |
| 14:50 | <karlcow> | including friends, organization, etc. |
| 14:50 | <jgraham> | annevk: If we're being picky MI5 is only two letters :) |
| 14:50 | <karlcow> | I know in Montreal one of the cheapest hosting services is http://iweb.com/ |
| 14:51 | <annevk> | jgraham, it's all falling apart |
| 14:51 | <annevk> | hsivonen and his fancy words :p |
| 14:56 | <hsivonen> | karlcow: I'm considering a lifestyle change to more cloud apps. for that to work, Webmail must not suck. |
| 14:57 | karlcow | is trying to make the link between the cloud and webmail |
| 14:58 | <hsivonen> | also, even if I end up continuing to use IMAP clients, I want server-side filtering to keep my mail filters the same regardless of client machine |
| 14:58 | <karlcow> | hsivonen: yes, I use procmail on the server side which gives me some filtering like classifying my mail in a dated space. |
| 14:59 | <zcorpan_> | http://simon.html5.org/sandbox/html/video/ - a little experiment |
| 14:59 | <hsivonen> | I'm using cloud app in the sense that the app core is on a server somewhere and the UI runs in a browser |
| 14:59 | <karlcow> | ah ok. understood. |
| 15:00 | <daedb> | zcorpan_: Neat :) |
| 15:00 | <zcorpan_> | daedb: thanks |
| 15:02 | <TabAtkins> | zcorpan_: I like it! |
| 15:02 | <TabAtkins> | ^_^ |
| 15:02 | <annevk> | what's up with the script being there twice? |
| 15:03 | <TabAtkins> | Dammit, annevk, I was just about to ask that. |
| 15:04 | <zcorpan_> | what script? |
| 15:04 | <annevk> | the one at the end? |
| 15:04 | <annevk> | well, the two... |
| 15:05 | <zcorpan_> | i only see one script |
| 15:05 | <zcorpan_> | oh wait |
| 15:06 | <zcorpan_> | wtf |
| 15:08 | <Lachy> | zcorpan_, I just see the video centred in the page, with seemingly no special effect, and get script errors |
| 15:08 | <Lachy> | Using Minefield. |
| 15:09 | <zcorpan_> | seems like a bug in chrome's view source the extra script |
| 15:09 | jgraham | finally finds a browser it works in, at the third try |
| 15:09 | <TabAtkins> | Over here in current public chrome, I'm not seeing the <canvas> stretched across the whole viewport, like the CSS suggests it should be. |
| 15:09 | <Lachy> | jgraham, which browser worked for you? |
| 15:09 | <TabAtkins> | It fills perhaps the center 3/4. |
| 15:09 | <jgraham> | chrome |
| 15:10 | <annevk> | works in 10.5 too |
| 15:10 | <annevk> | but looks different |
| 15:10 | <jgraham> | annevk: On linux? |
| 15:10 | <annevk> | yes |
| 15:10 | <jgraham> | Hmm, not for me |
| 15:10 | <annevk> | maybe it does not like you |
| 15:10 | <jgraham> | No it does not; it just crahsed my 10.50 |
| 15:11 | <zcorpan_> | chrome has blocky canvas rendering, it's nicer in opera |
| 15:11 | <Lachy> | latest internal mac build of 10.50 isn't working for me at all. |
| 15:11 | <zcorpan_> | Lachy: dunno why it doesn't work in firefox |
| 15:11 | <jgraham> | when it works (/me sulks) |
| 15:11 | <annevk> | parsing |
| 15:11 | <annevk> | prolly |
| 15:11 | <annevk> | there's no content |
| 15:12 | <zcorpan_> | it still doesn't work with html5.enable for me |
| 15:12 | <annevk> | oh |
| 15:13 | <annevk> | well, with html5.enable the error does change |
| 15:13 | <annevk> | the <canvas> is found for instance |
| 15:13 | <annevk> | Error: uncaught exception: [Exception... "Component returned failure code: 0x80040111 (NS_ERROR_NOT_AVAILABLE) [nsIDOMCanvasRenderingContext2D.drawImage]" nsresult: "0x80040111 (NS_ERROR_NOT_AVAILABLE)" location: "JS frame :: http://simon.html5.org/sandbox/html/video/ :: anonymous :: line 16" data: no] |
| 15:13 | <annevk> | is what I get now |
| 15:13 | <TabAtkins> | Ah, wait, nm. I get why it doens't stretch to the whole screen. Because the drawImage is only 1,1,4,4, duh. |
| 15:13 | <TabAtkins> | Still, pretty awesome. |
| 15:14 | <annevk> | so in Opera it's not blocky because? |
| 15:14 | <annevk> | must be because we have a high-res backing store suddenly... |
| 15:14 | <annevk> | thought we didn't want that? |
| 15:15 | <zcorpan_> | because it's smooth instead :) |
| 15:15 | <annevk> | jaja |
| 15:15 | <Lachy> | zcorpan_, why don't you remove the thick white border from the video? |
| 15:16 | <zcorpan_> | annevk: i guess chrome uses nearest neighbour and we use bicubic or something |
| 15:16 | <zcorpan_> | annevk: i don't think it has anything to do with highres backing store |
| 15:17 | <zcorpan_> | Lachy: because i wanted a thick white border |
| 15:17 | <annevk> | well, the result is stored in 16-canvas pixels |
| 15:17 | <annevk> | that's not a lot |
| 15:17 | <zcorpan_> | 36 |
| 15:17 | <annevk> | i thought the graphic was only drawn at a 4*4 space? |
| 15:18 | <annevk> | regardless, you cannot get that smooth in such a small area, afaict |
| 15:19 | <TabAtkins> | Hrm, it's not wanting to work when I download the video and run the page from my desktop... |
| 15:20 | <zcorpan_> | TabAtkins: with opera? |
| 15:20 | <TabAtkins> | With Chrome. |
| 15:21 | <TabAtkins> | Does video run with origin restrictions by default? Do those act weird with file:// urls? |
| 15:21 | <TabAtkins> | I'll try opening it through localhost... |
| 15:21 | <Lachy> | I found a version of Opera 10.5 that isn't completely broken, but still doesn't work for me on mac. annevk, which build are you using? |
| 15:21 | <annevk> | not a Mac |
| 15:22 | <annevk> | 6177 |
| 15:22 | <annevk> | for Ubuntu 64-bit |
| 15:22 | <annevk> | it is teh awesome |
| 15:22 | <annevk> | apart from all the bugs |
| 15:22 | <zcorpan_> | Lachy: video doesn't work on mac yet |
| 15:23 | <Lachy> | ah, ok. |
| 15:24 | <TabAtkins> | Okay, localhost worked, once I restarted apache. |
| 15:26 | <TabAtkins> | zcorpan_: What's with the extraneous <div>? Remnant of previous hacking? |
| 15:28 | <zcorpan_> | TabAtkins: yeah. removed |
| 15:29 | <zcorpan_> | wonder if it starts working in firefox if i add a try/catch around drawimage |
| 15:31 | <zcorpan_> | nope |
| 16:47 | <annevk> | http://www.reuters.com/article/idUSTRE6062Z720100107 |
| 16:50 | <cardona507> | wow - they must really want it |
| 16:51 | <cardona507> | let's hope it brings us a video codec :) |
| 16:51 | <annevk> | oh yes |
| 16:51 | <TabAtkins> | Ooh, cool. |
| 16:51 | <TabAtkins> | Damn On2. Just give up already. >_< |
| 16:52 | <Philip`> | Why would they give up, when they can wait to get more money from Google? |
| 16:52 | <TabAtkins> | Damn your game-theoretic logic! |
| 16:52 | <cardona507> | It looks like their waiting is PAYING off.... (pun intended) |
| 17:47 | <Lachy> | oh crap, microdata will be split :-(. I can't believe the obviously flawed arguments for that position won out over the technical superiority of the arguments against. |
| 17:58 | <AryehGregor> | Lachy, are you really surprised? |
| 17:58 | <AryehGregor> | I expected it. |
| 17:59 | <AryehGregor> | As I'm reading the procedure, TBL has to personally sign off on it to go to Last Call, right? And the TAG, which is headed by TBL, strongly requested that it be split? |
| 17:59 | <TabAtkins> | Sigh. RDFa doesn't have major implementations. The major cited implementations are supporting a language vaguely similar to RDFa in syntax, but different in significant ways. |
| 18:00 | <AryehGregor> | Better for the chairs to save their energy for more important things, if they have to fight against the W3C administration over anything. |
| 18:00 | <AryehGregor> | I imagine there's a limit to how much they can put in the spec against the TAG's wishes and still get the spec approved for progression. |
| 18:00 | <AryehGregor> | Although I don't actually know anything about how the W3C works, so I'm just guessing here. |
| 18:01 | <Lachy> | I don't think TimBL could really overrule a WG decision if it had been decided to keep it, regardless of Tim's or the TAG's position on the issue |
| 18:01 | <AryehGregor> | Doesn't the W3C procedure say that the Director needs to personally approve specs for them to progress? |
| 18:02 | <Philip`> | TabAtkins: Yahoo's is fairly close to RDFa, if I remember correctly |
| 18:02 | <Lachy> | that's basically to confirm that proper W3C procedures have been followed during it's development, rather than having the director say whether or not he agrees with everything in the spec |
| 18:02 | <Philip`> | (though Google's isn't) |
| 18:02 | <Philip`> | (but Google's doesn't even do HTML or XML, just something vaguely similar to them in syntax) |
| 18:03 | <TabAtkins> | Philip`: Ah, okay. |
| 18:03 | <AryehGregor> | Well, anyway, it makes no technical difference, just a PR difference. |
| 18:03 | <TabAtkins> | AryehGregor: Yeah. |
| 18:03 | <AryehGregor> | I'd prefer for us to lose this fight than a lot of other fights that are coming up. |
| 18:04 | <Lachy> | I'm confident that Microdata will ultimately win in the market place anyway, especially given the support I've seen from some in the microformats community |
| 18:04 | <AryehGregor> | I'm glad to hear that. |
| 18:05 | <TabAtkins> | Well, yeah. Microformats generally map *directly* to a Microdata vocabulary in an extremely obvious way, nearly to the point of just being able to search/replace the current class-based syntax with a Microdata-based one. |
| 18:05 | <Lachy> | but, you're right, there are bigger fights coming up that are more important than this |
| 18:05 | <AryehGregor> | But I'm in an awkward position as a web developer, with one well-established spec that a W3C Recommendation and everyone's heard of, and a better spec that's not in any W3C spec anymore, as soon as it's removed from the HTML5 draft. |
| 18:05 | <AryehGregor> | Also, CC license metadata is currently only de facto standard in RDFa, IIRC. |
| 18:05 | <Lachy> | the next step for this issue is just to take Microdata to FPWD |
| 18:06 | <AryehGregor> | I assume that will probably happen, given the enthusiasm from the RDFa camp about equal footings and all. |
| 18:06 | <AryehGregor> | Does anyone know someone at the CC who they could convince to have an official microdata vocabulary for CC licenses and encourage people to use that instead of RDFa? |
| 18:07 | <AryehGregor> | We now have a few major parties that are consuming some form of RDFa, none for microdata that I see. |
| 18:07 | <AryehGregor> | Oh well. |
| 18:08 | <Lachy> | I'm sure there are people in the RDFa camp who would personally rather see microdata just die, but who've held a public position of leveling the playing field |
| 18:09 | <AryehGregor> | Yes, which is why I think we'll probably get a FPWD. |
| 18:09 | <Lachy> | othermaciej, the mail you sent seems to be incomplete |
| 18:09 | <AryehGregor> | Because most of them will be at least reluctant to admit that. |
| 18:09 | <Lachy> | it's chopped off after "functionality such as" |
| 18:09 | <othermaciej> | Lachy: looks ok to me when I open the attachment |
| 18:10 | <Lachy> | and there's a mistake in it where it says "Also, the partial reliance on HTML5-specific features, such as or , does not" |
| 18:10 | <othermaciej> | attachment also looks ok in the archives: http://lists.w3.org/Archives/Public/public-html/2010Jan/att-0218/issue-76-decision.html |
| 18:10 | <othermaciej> | I also do not see that error in the archive |
| 18:10 | <Lachy> | oh, I didn't realise it was an attachment. I was just reading it directly in my mail client |
| 18:10 | <othermaciej> | "Also, the partial reliance on HTML5-specific features, such as <time> or <meta>" |
| 18:11 | <Lachy> | it just shows up as plain text for me, so it looked the same |
| 18:11 | <othermaciej> | I did notice some other typosand formatting errors though |
| 18:11 | <Lachy> | so it's a bug in my mail client that it doesn't show HTML attachments properly |
| 18:12 | <othermaciej> | I was hesitant to send it as HTML at all but worse case people can read it in the archives I guess |
| 18:15 | <Philip`> | What's this newfangled "HTML" you're trying to make us read? |
| 18:15 | <othermaciej> | I know, right? |
| 18:16 | <Philip`> | Plain text was good enough for typewriter users, it should be good enough for us |
| 18:16 | <Lachy> | my only regret in this issue is that I never got around to debunking the claim about how splitting it would somehow level the playing field, but that's just cause I went on holidays just before the poll was announced |
| 18:17 | <Philip`> | I like how Gmail gives me three different ways to access the attachment, without even attempting to explain the difference between them |
| 18:17 | <AryehGregor> | That's the problem with adding new features. |
| 18:19 | <Lachy> | Philip`, I'm guessing there's options to download and another to view. What's the third? |
| 18:20 | <GarethAdams|Work> | Lachy: create a google doc |
| 18:20 | <Philip`> | Open as a Google Document |
| 18:20 | <Philip`> | (and View and Download) |
| 18:20 | <Philip`> | "We're sorry, but Google Docs has encountered an unexpected error." - hmm, that narrows down the (working) choices |
| 18:22 | <Philip`> | errMsg=invalid+URI%3A+http%3A%2F%2Fwww.w3.org%2FTR%2Fhtml5%2Fmicrodata.html%23microdata%0A - odd |
| 18:22 | <Philip`> | (in the URL of the error page) |
| 18:41 | <AryehGregor> | "There's nothing special about boolean attributes in this case. In the |
| 18:42 | <AryehGregor> | HTML5 text/html syntax the attribute value can always be omitted to say |
| 18:42 | <AryehGregor> | that the value is the empty string." |
| 18:42 | <AryehGregor> | Huh, I never knew that. |
| 18:42 | <AryehGregor> | That's pretty neat. |
| 18:46 | <Philip`> | It couldn't really do anything else |
| 18:46 | <Philip`> | given how the DOM works |
| 18:47 | <Philip`> | (i.e. given that attributes have a name and value and nothing much else) |
| 18:51 | <AryehGregor> | You could say it's invalid, though, and/or specify some other way of parsing it. |
| 18:51 | <AryehGregor> | I mean, different for boolean attributes and others. |
| 18:51 | <AryehGregor> | But this makes sense. |
| 18:51 | AryehGregor | will have to add that to MediaWiki's Html class |
| 18:52 | <Philip`> | If you could parse it some other way, what would you do? |
| 18:52 | <Philip`> | I suppose you could drop the entire attribute, but I can't think of anything else sane |
| 18:52 | <AryehGregor> | I had assumed it was invalid, I guess. |
| 18:53 | <AryehGregor> | I'd have probably expected it to be parsed as having an empty string if it was actually present, though. |
| 18:53 | <Philip`> | Ah, okay |
| 18:53 | <AryehGregor> | But that's implementer stuff, I don't think about that. |
| 18:53 | <AryehGregor> | I'm an author. |
| 18:53 | <Philip`> | (Maybe it should be invalid) |
| 18:53 | <AryehGregor> | It does seem somewhat confusing. But hey, I don't mind the extra three bytes. |
| 19:09 | <foolip> | so who wants to gamble on how Microdata will be split over (a) W3C HTML5 (b) WHATWG HTML5 and (c) WHATWG HTML ? |
| 19:13 | foolip | hopes for no split at all in (b) and (c), at the very least (c) |
| 19:14 | Philip` | hopes for (a) and (b) to be equivalent |
| 19:14 | <Philip`> | (unless they've already diverged?) |
| 19:14 | <foolip> | not to my knowledge |
| 19:15 | AryehGregor | bets on a separate spec in the W3C, same spec at WHATWG |
| 19:25 | foolip | facepalms over dads insitence to live in happy happy quirks mode land instead of fixing his CSS bugs |
| 19:28 | <AryehGregor> | Quirks mode works about as consistently as standards mode, so if people want to use it, why not? |
| 19:28 | <AryehGregor> | I mean, I won't, but it's not worth going to much effort to rewrite your site for. |
| 19:28 | <othermaciej> | quirks mode is much less consistent between different browsers |
| 19:28 | <foolip> | what about IE's box model? |
| 19:29 | <AryehGregor> | Well, HTML5 wants to make quirks mode consistent between browsers, at least, so your dad can just wait a few years. |
| 19:29 | <AryehGregor> | Unless IE sticks with legacy modes forever. |
| 19:29 | <AryehGregor> | Anyway, if it makes him happy, I don't care, let him stick with non-standard stuff. |
| 19:29 | <AryehGregor> | Not all bugs are critical, he probably has better things to do. |
| 19:30 | <foolip> | it started out as a IE-only bug |
| 19:30 | <foolip> | I don't suggest rewriting just because it's fun |
| 20:06 | <Lachy> | Philip`, (a) and (b) have already diverged |
| 20:07 | <Lachy> | e.g. the new device element in the WHATWG spec, which is not in W3C HTML5 |
| 20:07 | <Lachy> | that's why the WHATWG copy is now called "WHATWG HTML (Including HTML5)" |
| 20:08 | <Lachy> | so I would assume the split will only affect the W3C copy |
| 20:08 | <Lachy> | but, at the very least, it will stay in complete.html |
| 20:09 | <Philip`> | Lachy: I thought that was the difference between foolip's (b) and (c) |
| 20:10 | <Philip`> | and (b) was the version that matched (a) |
| 20:35 | <Lachy> | Philip`, huh? My understanding was that foolip was referring to (a) w3.org copy, (b) whatwg /current-work/ and (c) whatwg /current-work/complete.html |
| 20:37 | <Lachy> | in this case, (a) and (b) are mostly identical, except for the header (basically everything above the TOC) and the recent introduction of the device element. |
| 20:37 | <Philip`> | /current-work/ is HTML, not HTML5 |
| 20:37 | <Lachy> | and (b) is a subset of (c), which incorporates several other specs. |
| 20:37 | <Philip`> | /html5/ is HTML5 |
| 20:38 | <Philip`> | complete.html is not any kind of HTML at all |
| 20:39 | <Philip`> | (where "/html5/" means http://www.whatwg.org/specs/web-apps/html5/, not http://www.whatwg.org/html5/) |
| 20:39 | <Lachy> | oh. I wasn't aware of that copy |
| 20:41 | <Philip`> | Lachy: http://wiki.whatwg.org/wiki/FAQ#What_are_the_various_versions_of_the_spec.3F should make it all wonderfully clear |
| 20:41 | <Lachy> | I can't find anywhere on whatwg.org that actually links to the web-apps/html5/ copy |
| 20:48 | jgraham | wonders how the crieria used to spilt out microdata apply to other parts of the specification |
| 20:48 | <cardona507> | what is a shadow dom? |
| 20:48 | <jgraham> | In particular "maturity" and "market success" |
| 20:49 | <jgraham> | cardona507: In what contex? XBL defines an officil meaning but it has been used in a different sense to talk about <canvas> accessibility |
| 20:50 | <TabAtkins> | I'm not sure precisely how the <canvas> folks are using the term, but when you hear it in reference to XBL2, it means a special transformed DOM that is only accessible to CSS. |
| 20:50 | <TabAtkins> | Javascript still sees the original, untransformed DOM. |
| 20:50 | <cardona507> | I was just reading a canvas accessibility email from HTMLwg and I am unclear by "shadow dom" - tabatkins - only accessible to css? what are the benefits? |
| 20:51 | <jgraham> | cardona507: http://www.w3.org/TR/xbl/#shadow is what xbl means |
| 20:51 | <cardona507> | cool - thanks jgraham |
| 20:51 | <Philip`> | In the canvas accessibility discussion, it just means the DOM subtree of the <canvas> element (which doesn't get rendered in graphical browsers), as far as I'm aware |
| 20:51 | <jgraham> | in the <canvas> case they just mean "desendants of the canvas element" as I understand it |
| 20:51 | <TabAtkins> | cardona507: The benefit is that it allows you to do transformations that don't actually make sense from a content perspective, but are necessary for styling. Like, say, wrapping a box in multiple <div>s so that you can style each one. |
| 20:52 | <othermaciej> | in the case of <canvas> it's just the descendants of canvas which are not rendered, but are exposed to accessibility tools, per the spec |
| 20:52 | <cardona507> | ahhh - I see - thanks othermaciej |
| 20:52 | <TabAtkins> | The fact that your styling happens to be easier to apply when you have a structure like <div><div><div><div>foo</div></div></div></div> shouldn't have any effect on your actual content, which just uses <div>foo</div>. |
| 20:53 | <othermaciej> | jgraham: most of the individual criteria were not enough to decide the matter by themselves, and it's not clear if any other part of the spec is precisely in the somewhat unusual position that Microdata is in |
| 20:54 | <TabAtkins> | What precisely *is* the unusual position Microdata is in? |
| 20:54 | <othermaciej> | jgraham: plus, we're not going to rule on splitting anything that no one actually objects to |
| 20:54 | <othermaciej> | read the long review of arguments |
| 20:54 | <TabAtkins> | othermaciej: That still leaves you with, iirc, something like half a dozen requests to remove things. ^_^ |
| 20:54 | <Philip`> | Isn't the XBL thing going to be a nightmare for debugging and for understanding what tree structure your document represents? |
| 20:54 | <othermaciej> | and see if you think those same arguments all apply equally to any other part of the spec |
| 20:55 | <TabAtkins> | Philip: I don't think so. The kind of reorganizing you can do is relatively limited. |
| 20:55 | <othermaciej> | anyway, I am interested in moving on to the next issue on deck which is the dt/dd/details/figure thing |
| 20:55 | <othermaciej> | I hope someone else writes an alternate proposal for that because I am not really keen on <fltcap> |
| 20:56 | <othermaciej> | I might have to write one myself, or more than one |
| 20:56 | <TabAtkins> | I still have *no* clue what fltcap is supposed to stand for. I see that "cap" part is for "caption", but what is "flt"? |
| 20:56 | <othermaciej> | it's supposed to stand for "floating caption" |
| 20:56 | <jgraham> | Oh I forot about that issue |
| 20:56 | <Lachy> | the right way to deal with that is, since Hixie won't introduce new elements, is to drop both figure and details from the spec and reintroduce them with <legend> in the future when the parsing issues are resolved with it. |
| 20:56 | <Lachy> | ah, rending issues. |
| 20:56 | <jgraham> | Yeah there needs to be another chage proposal at least |
| 20:56 | <Lachy> | *rendering |
| 20:57 | <Lachy> | ok, someone should write down what I just said as a change proposal |
| 20:57 | <othermaciej> | as far as I am aware, that is not an ordinary term of art for the caption for a figure, let alone the label of a disclosure control |
| 20:57 | <othermaciej> | I am considering writing Change Proposals for one or more of the following: |
| 20:57 | <othermaciej> | 1) Use <caption> for <figure> and <label> for <details>, with a slight tweak to parsing so that <caption> inside <figure> does not break tables |
| 20:58 | <Lachy> | that would be nice if you think the legacy issues with <caption> aren't insurmountable |
| 20:58 | <othermaciej> | 2) Introduce new elements <fcaption> and <dlabel> because it's totally a standard design pattern for HTML elements with compound structure to have specialized helper elements that exist only for purposes of their structure |
| 20:59 | <Lachy> | I like <c> as an element name, which I've suggested before |
| 20:59 | <othermaciej> | despite Hixie's distaste for the idea, we have tbody, thead, tfoot, tr, td and caption which are (currently) only used for tables, li only for lists, dt/dd previously only for dl, param only for object |
| 20:59 | <TabAtkins> | The issues with <caption> only arise when you put a <figure> inside of a <table> anyway. If you can avoid that you're golden. |
| 20:59 | <Lachy> | I think I've also suggested <description> |
| 21:00 | <othermaciej> | TabAtkins: and parsing can be changed in the future to make it not a problem in the long run, so I think it's better than <legend>, since both are fixable and in the short term <caption> only affects <figure> in <table> |
| 21:00 | <othermaciej> | 3) Suggest using <h1> (or some other <hn> element) for figure captions and the label of <details> |
| 21:00 | <Lachy> | TabAtkins, given how many legacy pages there are using layout tables around the whole page, I don't think it's unrealistic for a <figure> to end up being added to an old template like that |
| 21:00 | <jgraham> | I think my general opinion is that a) parsing issues should not be underestimated and b) adding more elements is not a problem |
| 21:00 | <TabAtkins> | Then they just wait for parsing changes before doing so, Lachy. |
| 21:01 | <othermaciej> | I asked Hixie what the problem with <h1> would be and he said it might interfere with the content model of <figure> |
| 21:01 | <jgraham> | Seriously people "just wait for parsing changes" is a really bad strategy |
| 21:01 | <Lachy> | I don't like the idea of reusing heading elements, since a caption is not a heading. It's a very different concept, and I think the default styling of h1 will make it very unappealing to authors |
| 21:01 | <jgraham> | If we can't do something that can be deployed in legacy browsers we should wait for the next version of HTML |
| 21:01 | <othermaciej> | my reaction is that this is not a problem for <details>, and in the case of <figure> in the unusual case where you need headers inside the figure content, you can wrap it in a <div> or <section> |
| 21:02 | <TabAtkins> | If the parsing changes are required to use it at *all*, like <legend>, it's a problem. If they are more minor, it's less of a problem. |
| 21:02 | <jgraham> | "Needed to use it in a table based layout" is a serious problem |
| 21:02 | <othermaciej> | I think <details><label> can be deployed in legacy browsers with essentially no problem, and <figure><caption> with problems only for figures in tables, but making it usable everywhere else |
| 21:02 | <Lachy> | also, reusing h1 to h6 for this will make stylesheets overly complicated, since authors wanting to apply styles to headings will have to do some magic with their CSS to undo the styles for figure>h1 |
| 21:03 | <othermaciej> | jgraham: I agree with you in general that adding new elements is not a problem; Hixie seems to be almost alone in very strongly opposing new elements |
| 21:03 | <othermaciej> | out of my 3 proposals I most prefer <fcaption>+<dlabel> |
| 21:03 | <Lachy> | I think Hixie's stance on new elements is not supported by sound rationale |
| 21:03 | <othermaciej> | because, like I said, specialized child elements for structured containers are totally normal |
| 21:04 | <cardona507> | When does the HTML5wg charter end? when the spec is finished? |
| 21:04 | <othermaciej> | heck, <datagrid> introduced half a dozen specialized children |
| 21:04 | <jgraham> | (the reason I think that deploying things that cannot work in the legacy is bad is that authours read a bunch of tutorials saying "well you can't actually use this because of legacy concerns" and then ten years later when the legacy concerns are gone, people still cargo-cult whatever workaround was developed in the inierim) |
| 21:04 | <othermaciej> | cardona507: it expires at the end of this year, but will hopefully be renewed |
| 21:04 | <cardona507> | is there really any doubt??? |
| 21:04 | <Lachy> | nice. When we renew it, can we fix the mistakes that sadly ended up in the current one? |
| 21:04 | <othermaciej> | anyway, if anyone wants to write up any of my 3 ideas as a Change Proposal, please volunteer on the list (or just do it) |
| 21:04 | <jgraham> | othermaciej: Yes, Hixie's position seems to be based mainly on aesthetics |
| 21:04 | <cardona507> | what would happen if it wasn't renewed? who owns the copyright? w3c or whatwg? |
| 21:05 | <TabAtkins> | othermaciej: I'll do so. |
| 21:05 | <othermaciej> | or any other ideas |
| 21:05 | <TabAtkins> | I'll have some for you by Sunday. |
| 21:05 | <othermaciej> | if no one proposes an idea I like by the day before the deadline, I will volunteer myself and ask for an extensio |
| 21:05 | <othermaciej> | *extension |
| 21:05 | <Lachy> | cardona507, the WHATWG maintains dual-copyright on the spec |
| 21:06 | <Lachy> | when is the deadline? |
| 21:06 | <othermaciej> | another idea that was raised before was adding a caption global attribute so you could use <p caption> or <h1 caption> or whatever |
| 21:06 | <cardona507> | when will we know if HTMLwg charter is gonna be renewed? |
| 21:06 | <TabAtkins> | That one's definitely going in, since I think I'm the one that originated it. ^_^ |
| 21:06 | <Lachy> | yeah, that was originally my idea. I'm not a big fan of that. But I think TabAtkins was |
| 21:06 | <TabAtkins> | Ok, maybe Lachy did. |
| 21:06 | <Lachy> | yeah, you stole it from me :-) |
| 21:06 | <TabAtkins> | Be flattered. ^_^ |
| 21:07 | <Lachy> | yeah, I don't mind at all |
| 21:07 | <jgraham> | TabAtkins: FWIW I think that proposal sucks :) |
| 21:07 | <daedb> | I'm for anything that keeps figure and details in the spec, dropping (or postponing to a future version) them is the only thing I'm very strongly against. |
| 21:07 | <othermaciej> | I don't like that one very much |
| 21:07 | <Lachy> | this is going to be confusing if we get 5 change proposals to choose between |
| 21:07 | <TabAtkins> | jgraham: fwiw I think YOU suck. |
| 21:07 | <othermaciej> | I guess it's marginally better than <fltcap> but not much |
| 21:07 | <jgraham> | TabAtkins: A fir assessment |
| 21:07 | <jgraham> | +a |
| 21:08 | <othermaciej> | Lachy: if we get many proposals perhaps some end up merging or getting withdrawn based on discussion |
| 21:08 | <Lachy> | what does "fltcap" stand for? |
| 21:08 | <jgraham> | heh |
| 21:08 | <TabAtkins> | "floating caption". |
| 21:08 | <othermaciej> | "floating caption" |
| 21:08 | <TabAtkins> | For some reason. |
| 21:08 | <Lachy> | oh. wtf? |
| 21:08 | <Philip`> | Flute cape |
| 21:08 | <jgraham> | I think it might be a term from print media |
| 21:08 | <Philip`> | It's a kind of musical garment |
| 21:09 | <Lachy> | ok, that idea sucks big time. We can't have a name that will have authors asking "wtf?" all the time. |
| 21:09 | <Philip`> | I don't think the proposal was particularly fixated on the name, and said it could be easily changed |
| 21:10 | <jgraham> | (google suggets I m talking rubbish) |
| 21:10 | <Lachy> | othermaciej, I think we should merge the change proposals to introduce a new element together into one, and just list the possible alternatives. Then when people make their case for/against the change proposals, they can state the name preference |
| 21:10 | <jgraham> | Philip`: The problem as I understand it is that we would have to adopt the proposal wholesale. So we would need the name at least temporarily |
| 21:10 | daedb | likes the <fcaption>, <dlabel> and <c> ideas |
| 21:11 | <jgraham> | And fixing up something that has already gone through the WG sasuage machine might get messy |
| 21:11 | <Lachy> | daedb, the problem with <c> is that it's only for <figure>. It doesn't have a matching alternative for details |
| 21:11 | <othermaciej> | <c> could be used for <details> too, though |
| 21:11 | <othermaciej> | I don't think it's a great semantic fit |
| 21:11 | <Philip`> | jgraham: Did you mean "the WG assuage machine"? |
| 21:12 | <Philip`> | (Sounds like an appropriate term) |
| 21:12 | <jgraham> | Lachy: I think there are reasonably strong differences between a proposal that has one element for both and seperate elements |
| 21:12 | <othermaciej> | ultimately if this ends up in a decision it has to be a concrete proposal |
| 21:12 | <Lachy> | I also expect the usual objections about wasting single-letter element names, which came up so often in the whatwg when we were arguing about <m> and <x>, etc. |
| 21:12 | <jgraham> | And also for proposals that reuse elements vs ones that mint new ones |
| 21:12 | <othermaciej> | and since many of the concerns about this issue are aesthetic, the specific name of the element or elements will probably be important to people who care in the first place |
| 21:12 | <TabAtkins> | I vaguely remember <m> and <x> being discussed. What was that about? |
| 21:13 | <daedb> | Lachy: I don't particularly care about <details> (though I want to keep it), <figure> is very important to me though. I just like <c> because it's short. |
| 21:13 | <othermaciej> | <m> was the old name of <mark> |
| 21:13 | <Lachy> | jgraham, fair point. I suppose we should wait and see what change proposals we end up with before we figure out merging |
| 21:13 | <jgraham> | plus I think a bunch of the justification in the existing proposal is bogus even though I sort of agree with the conclusion |
| 21:13 | <Lachy> | I haven't seen the existing change proposal yet |
| 21:13 | <Lachy> | got a pointer? |
| 21:13 | <daedb> | othermaciej: When did <m> change to <mark>? |
| 21:13 | <othermaciej> | don't remember |
| 21:14 | <othermaciej> | http://esw.w3.org/topic/HTML/ChangeProposals/DdDt |
| 21:14 | <TabAtkins> | So we have four proposals: (1) Leave it as it is. (2) Add a new element (for one or both). (3) Add an attribute (for one or both). |
| 21:14 | <othermaciej> | everything relevant is listed on http://dev.w3.org/html5/status/issue-status.html |
| 21:14 | <daedb> | I just noticed that a few days ago or something... <m> was fine me. |
| 21:15 | <othermaciej> | I think just the three different things I proposed are all materially different from each other to the point that people will have different opinions |
| 21:15 | <TabAtkins> | s/four/three/ |
| 21:15 | <othermaciej> | even though they all end up changing the element used for each |
| 21:15 | <othermaciej> | and would thus be in your #2 |
| 21:15 | <othermaciej> | I don't know if anyone but Hixie is willing to defend the status quo on this |
| 21:16 | <TabAtkins> | Nah, <caption> and <h1> should be #4. |
| 21:16 | <othermaciej> | (though I must admit I dislike dd/dt less than fltcap) |
| 21:16 | <TabAtkins> | I even *like* dt/dd on details. |
| 21:16 | <othermaciej> | I think it reads bad on <figure> |
| 21:16 | <TabAtkins> | it mirrors how I'll often be coding it for styling, anyway |
| 21:16 | <TabAtkins> | I agree. |
| 21:17 | <othermaciej> | it's ok on <details> but not enough of a win to be better than a new element |
| 21:17 | <TabAtkins> | The IE problems are pretty weird. |
| 21:17 | Philip` | dislikes dt/dd on details since he can barely remember which way around they go in dl |
| 21:18 | <Philip`> | s/details/figure/ or whatever |
| 21:18 | <TabAtkins> | <fcaption>, <fbody>(optional), <dlabel>,<dbody>(optional) |
| 21:18 | <jgraham> | Philip`'s argument is my number one objection |
| 21:18 | <jgraham> | I think the confusion value is too high because the elements are so poorly named |
| 21:18 | <TabAtkins> | Those sort of mirror the <table> naming method. |
| 21:18 | <Philip`> | Also, it seems entirely non-obvious when you've got it the wrong way round, because it won't render differently |
| 21:19 | AryehGregor | agrees, dt/dd on anything other than dl is really confusing |
| 21:19 | <TabAtkins> | Huh? Yeah it will. <dd> has a margin-left by default, <dt> doesn't. |
| 21:19 | <othermaciej> | there is a limit to how many very short element names you can get away with, and dt and dd don't make the short list |
| 21:19 | <othermaciej> | TabAtkins: I don't think it would render that way inside <details> |
| 21:19 | <TabAtkins> | Oh, right. No it wouldn't. |
| 21:19 | <Philip`> | TabAtkins: I want neither my image nor my caption to be indented |
| 21:20 | <Lachy> | I think dt/dd sort of works for details, but the compat problems are annoying for authors to have to deal with, especially since the workaround isn't all that intuitive, at least without understanding how IE's crazy parsing works |
| 21:20 | <daedb> | I prefer a new element (though I dislike the ftlcap name) for <figure> instead of dt/dd just because dt/dd feels like unnecessary wrapping most of the time. |
| 21:20 | <Lachy> | but I really hate dt/dd for figure. Worst. Idea. Ever. |
| 21:21 | <othermaciej> | so it sounds like out of my 3 ideas, fcaption/dlabel is probably the most popular in this crowd |
| 21:22 | <jgraham> | I would prefer <figcaption> over <fcaption> I think |
| 21:22 | <jgraham> | </bikeshed> |
| 21:22 | <Lachy> | I suppose the rationale against using dt/dd would be the same for virtually all change proposals, so we should have that clearly documented once, and then let the rest of the change proposals just argue which alternative is best |
| 21:22 | <othermaciej> | that's it, we should name the new element <bikeshed> |
| 21:23 | <daedb> | othermaciej: too long, let's use the abbreviated <bs> :) |
| 21:23 | <Philip`> | Why give this element such special status, compared to all the other bikesheds we have? |
| 21:23 | <jgraham> | (because it seems odd to spell out <figure> and spell out caption> but have a random f hanging around) |
| 21:23 | <othermaciej> | sure, everyone making their own rationale might not be the best use of time, I think it would be ok to cite another Change Proposal's rationale wholesale and the new rationale just explains why this proposal seems best |
| 21:23 | <othermaciej> | ) |
| 21:23 | <Philip`> | jgraham: It matches <table>/<head>/<thead> |
| 21:23 | <othermaciej> | jgraham: there is <tbody> for precedent |
| 21:24 | <Lachy> | the bikehed element needs two attributes <bikeshed color="blue" colour="blue"> to keep everyone happy |
| 21:24 | <webben> | othermaciej: I think if it needs to be an abbreviation those are good names ... following the pattern of tbody and iframe. |
| 21:24 | <jgraham> | Yeah but no one uses <thead> or <tbody> :) |
| 21:24 | <jgraham> | and noone cares what the i in <iframe> stands for |
| 21:24 | <TabAtkins> | I use <thead> *constantly*. But I guess I'm weird. |
| 21:24 | <Philip`> | TabAtkins: You are |
| 21:25 | webben | would prefer figurecaption and detailslabel. |
| 21:25 | <TabAtkins> | It's so good for styling, though! |
| 21:25 | <Philip`> | TabAtkins: http://philip.html5.org/data/tag-count-pages.txt - nobody uses thead |
| 21:25 | <Lachy> | why don't we go with <details><dhead><dbody></details> |
| 21:25 | <TabAtkins> | I don't see how you can go without it, when you can easily do "thead th" versus "tbody th" for your row and column headers. |
| 21:25 | <Philip`> | Lots use tbody but they're probably from serializers |
| 21:25 | Lachy | runs before people start making jokes about why dhead is bad... |
| 21:25 | <jgraham> | TabAtkins: Based on having looked at a lot of table markup when all the table accessibility stuff was still interesting, I can confirm that you are indeed weird |
| 21:26 | <TabAtkins> | Man… |
| 21:26 | <Philip`> | TabAtkins: Easier to do <td class="rowhead"> |
| 21:26 | <TabAtkins> | Lachy: I don't like <dhead> or <fhead> because neither are intuitive. |
| 21:26 | <TabAtkins> | Philip`: … Or just do <th scope=row> |
| 21:27 | <webben> | what is "o:p" from Philip`'s list? |
| 21:27 | <webben> | some Microsoft Office thing? |
| 21:27 | <Philip`> | webben: Microsoft Word output |
| 21:27 | <TabAtkins> | An MS element. |
| 21:27 | <Lachy> | TabAtkins, my suggestion was more of a joke, in case you haven't figured out the obvious alternative interpretation of what dhead stands for yet |
| 21:27 | <TabAtkins> | Lachy: Ah. I have never, ever seen it abbreviated to dhead before, so I didn't think of it. |
| 21:28 | <Lachy> | it's not usually. That's just the obvious expansion |
| 21:28 | <TabAtkins> | Maybe to you, with your dirty mind. |
| 21:28 | <TabAtkins> | So, so dirty… |
| 21:28 | <Lachy> | hah |
| 21:28 | <Philip`> | Doody head? |
| 21:28 | <Lachy> | close. |
| 21:29 | <Philip`> | Dunderhead? |
| 21:29 | Philip` | cannot think of any other possibilities |
| 21:31 | <Lachy> | that's fine. We'll spare your young mind from learning about more profanities |
| 21:34 | <jgraham> | Philip`: You need too allow for the incorrect capilatisation. It is clearly supposed to be DhEAd which likely stands for Dharmic Epiphany Advent |
| 21:35 | <jgraham> | i.e. the suddent coming of a feeling that one must do ones righteous duty |
| 21:53 | <cardona507> | hsivonen - your opinion - http://img510.imageshack.us/img510/5561/whatwgcanvasexbutt0202.png for butt ----------- http://img22.imageshack.us/img22/760/whatwgcanvasexround.png for round ----------- & http://img207.imageshack.us/img207/8403/whatwgcanvasexsquare.png for square |
| 21:53 | <cardona507> | or anyone else's opinion of course |
| 21:56 | <TabAtkins> | cardona507: "canva sex butt"? |
| 21:56 | <cardona507> | haha! canvas ex butt :) |
| 21:57 | <Lachy> | TabAtkins, who's the one with the filthy mind now, huh? |
| 21:57 | <TabAtkins> | Nope, canva sex butt, now and forever. You can't ever put "ex" after "canvas" without some sort of separator. |
| 21:57 | <TabAtkins> | The butt is just icing on the cake. |
| 21:57 | <cardona507> | :D - I can't now |
| 21:57 | <TabAtkins> | Lachy: Still you. |
| 22:00 | <cardona507> | "The butt is just icing on the cake" - nice |
| 22:05 | <AryehGregor> | How about we use <rubric> for the new elements, define it to work the same as <caption>, <legend>, and <label> in the appropriate contexts, and eventually make all of the latter nonconforming? |
| 22:05 | <Lachy> | AryehGregor, no. |
| 22:05 | <AryehGregor> | :( |
| 22:05 | <othermaciej> | if the concern is aesthetics, then I can't say <rubric> is a win |
| 22:05 | <Lachy> | if you look up the actual definition of rubric, rather than just checking a thesaurus, you'll understand why |
| 22:06 | <Lachy> | also, it's an aesthetically sucky name |
| 22:07 | <AryehGregor> | Lachy, you don't think "red ocher" is an appropriate characterization? |
| 22:07 | <AryehGregor> | Anyway, the dictionary gives one definition as "A title; a name". |
| 22:07 | <AryehGregor> | Which is what I meant it as, except it also conveys the idea of being sort of abbreviated, like legend/caption. |
| 22:07 | <AryehGregor> | (although <label> is actually a sort of separate concept) |
| 22:10 | <TabAtkins> | I don't even know what rubric *means*. |
| 22:10 | <TabAtkins> | And I'm a word geek. |
| 22:13 | <cardona507> | yeah - rubric isn't too common to us dumbed down Amairicans |
| 22:15 | <Lachy> | I don't think rubric is too common among people who aren't 14th century priests |
| 22:16 | <Philip`> | cardona507: The round one is the wrong shape |
| 22:16 | <Philip`> | (It should be a semicircle on the end) |
| 22:16 | <Philip`> | The rectangles look odd since they're antialiased on the right and not the left |
| 22:17 | <cardona507> | Philip`: - fixing round - I don't see the antialias - on all 3 or which one? and are they too small? should I scale them up slightly? |
| 22:18 | <cardona507> | I kinda see the anti alias now - wierd |
| 22:19 | <daedb> | AryehGregor: <rubric> is too close to the swedish word rubrik for my taste |
| 22:19 | <Philip`> | I've heard "rubric" quite frequently, for the instructions on the front of exam papers |
| 22:19 | <AryehGregor> | What does the Swedish word "rubrik" mean? |
| 22:19 | <Philip`> | cardona507: Could make them SVG so they're scalable :-) |
| 22:19 | <AryehGregor> | Anyway, "rubric" isn't too obscure, not like "copacetic" or something, but hardly a common word, no. |
| 22:20 | <daedb> | rubrik = headline |
| 22:20 | <AryehGregor> | That's pretty similar to the English meaning. |
| 22:20 | <TabAtkins> | I can assure you that "rubric" is too obscure. |
| 22:21 | <AryehGregor> | Yes, yes, I wasn't really serious. |
| 22:21 | <AryehGregor> | Although people can just memorize it as gibberish, you know. |
| 22:21 | <TabAtkins> | I'd prefer to minimize the amount of gibberish in the language, especially gibberish that looks like it should be a word. |
| 22:21 | <AryehGregor> | Works well for some existing elements that are abbreviations that most HTML authors probably don't remember. |
| 22:21 | <AryehGregor> | <div> and <span> are more or less gibberish. |
| 22:22 | <TabAtkins> | Less gibberish than "rubric". |
| 22:22 | <Lachy> | rubric is only slightly less gibberish than fltcap |
| 22:22 | <cardona507> | Philip` - I will gladly try :) - I haven't tried SVG before - will it be in code or is it something I do in Illustrator? |
| 22:23 | <TabAtkins> | cardona507: It can be either! SVG is fun to code by hand. |
| 22:23 | <daedb> | div = various crap |
| 22:30 | <jgraham> | Ah it should have been called <vcrap> |
| 22:39 | <webster42> | Hi there folks, anybody on? |
| 22:41 | <TabAtkins> | Nope, we're all gone. |
| 22:43 | <Philip`> | I'm not here either |
| 22:43 | <webster42> | :P |
| 22:44 | <webster42> | So, I've been following along casually for years and I've decided I want to really get into HTML5 stuff and push the envelope |
| 22:44 | <webster42> | I have a feeling in 2010 it's going to start finally becoming more and more popular |
| 22:45 | <webster42> | am I off-base/too-early? |
| 22:45 | <webster42> | do you guys know a couple of impressive examples (real sites) that use the new functionality? |
| 22:50 | <cardona507> | wow - svg is a trip |
| 22:52 | <cardona507> | webster42 - google.com uses the HTML5 doctype - if you consider that and impressive site |
| 22:53 | <webster42> | cardona507: I don't. It's an impressive company however. :) |
| 22:53 | <cardona507> | hehe - true |
| 22:53 | <webster42> | what I mean is, have there been any cool webapps yet that take advantage of all of this stuff? (and don't care about IE) |
| 22:55 | <cardona507> | Google wave uses some html5 - it's pretty cool - but the spec isn't done yet - so there aren't any killer apps that I am aware of yet |
| 22:55 | <webster42> | I suppose it's pointless to ask about "when will the spec be done-er |
| 22:55 | <webster42> | " |
| 22:55 | <cardona507> | but quite a bit of the spec is implemented in modern browsers - so if you are creative you can cook up a nice html5 app |
| 22:56 | <cardona507> | done.... good question - who wants to field that one? |
| 22:56 | <webster42> | alright, well I have to run right now but I'll come back soon and explore this further |
| 22:56 | <webster42> | bye for now |
| 23:09 | <cardona507> | Philip`: or tabatkins - now that I have my svg code - how do I view it? Of course I can look at it in AI - but when I make it code and put it online it only looks like code in my browser http://cardonadesigns.com/butt.svg (yeah it says butt tab) |
| 23:10 | <TabAtkins> | cardona507: Your webserver is sending it with a content-type of text/plain. |
| 23:11 | <cardona507> | so I need to change htaccess to allow the mime type? |
| 23:11 | <TabAtkins> | Make it send as image/svg+xml and it'll work. |
| 23:15 | <cardona507> | cool - TabAtkins - it worked |
| 23:15 | <cardona507> | it doesnt scale very gracefully though - |
| 23:18 | <TabAtkins> | It seems to scale fine. It's just that your viewport isn't scaling. |
| 23:20 | <cardona507> | so how does it look in to relation to canvas line cap butt? |
| 23:25 | <TabAtkins> | cardona507: It looks similar. Is there supposed to be some fuzzing on the edges, or is that just an artifact of me scaling up the png? |
| 23:26 | <cardona507> | there aren't supposed to be fuzzy edges |
| 23:27 | <TabAtkins> | Okay, then they look roughly the same, modulo some slight color differences and positioning changes. |
| 23:27 | <cardona507> | yeah - but the first was a png and the second was svg - |
| 23:27 | <TabAtkins> | Yes? |
| 23:27 | <cardona507> | nm - Philip` asked me to redo my png as svg and I was asking for feedback |
| 23:28 | <TabAtkins> | What I meant is, I see a black rectangle partially overlapping a bluish rectangle in both. |
| 23:29 | <cardona507> | and do you think that is an appropriate graphic for the part about linecap="butt" ? |
| 23:32 | <cardona507> | section 4.8.11.1.5 |
| 23:36 | <Philip`> | cardona507: Looks reasonable to me :-) |
| 23:36 | <TabAtkins> | Yeah, seems fine. |
| 23:37 | <Philip`> | I think it's nice to have these things in SVG partly since scalability is good (particularly when printing) and partly because it's good to have in an editable format |
| 23:37 | <cardona507> | cool - I"ll make the other 2 correct and in svg - do you recommend that I do any future graphics for the spec in svg? |
| 23:37 | <Philip`> | so someone can make further changes without having to start from scratch |
| 23:38 | <Philip`> | (though I think the spec still needs a PNG fallback for obsolete browsers, like IE8) |
| 23:39 | <Philip`> | I'd prefer SVG, but my opinion has no particular weight so there's no reason to take it into account :-) |
| 23:42 | <AryehGregor> | Philip`, when I pointed out a Unicode character that didn't render on Firefox/Chrome on the latest Ubuntu, Ian told me to get better fonts and didn't change the text. So to heck with fallbacks? :) |
| 23:44 | <TabAtkins> | You *should* get better fonts, though. |
| 23:46 | <Philip`> | AryehGregor: Given that Microsoft people have asked questions about canvas, and assuming that Microsoft people use IE, there would be practical benefits to having fallbacks for those diagrams |
| 23:46 | <TabAtkins> | That's not a good assumption. |
| 23:46 | <AryehGregor> | Yeah, probably. |
| 23:46 | <TabAtkins> | From what I know. |
| 23:46 | <AryehGregor> | TabAtkins, seriously? |
| 23:46 | <AryehGregor> | You know people on the IE team who don't use IE as their primary browser? |
| 23:46 | <TabAtkins> | Yes. |
| 23:47 | <AryehGregor> | Can you name names, or is this secret? |
