| 01:33 | <Hixie> | othermaciej: i'm adding doc="" to <iframe> since abarth wants to implement it -- should this be in the version we're driving to LC at the w3c, or should I keep it out like <device>? |
| 01:34 | <othermaciej> | Hixie: I think since it is an important component of making an existing feature (sandboxed iframes) viable, I would suggest it should go in the W3C copy |
| 01:34 | <Hixie> | k |
| 01:34 | <othermaciej> | Hixie: it does not strike me as the kind of thing likely to cause huge controversy but you could send notice and an explanation to the list if you think it would help |
| 01:34 | <othermaciej> | "the list" being public-html |
| 01:35 | <Hixie> | sure |
| 02:07 | <tantek> | ICYMI: blog post re: HTML5 from web designer Andy Clarke: http://stuffandnonsense.co.uk/blog/about/keep_calm_and_carry_on_with_html5 |
| 02:08 | <tantek> | (n.b. I left a comment with my personal opinion/experience) |
| 02:10 | <othermaciej> | hi tantek |
| 02:11 | <othermaciej> | I like that the post title starts with "Keep calm" |
| 02:11 | <tantek> | hi othermaciej - indeed |
| 02:11 | <tantek> | would love to stay and chat but gotta run to tonight's microformats dinner |
| 02:11 | <othermaciej> | later |
| 02:11 | <tantek> | will try to be back on in a few hours later tonight |
| 02:15 | <jwalden> | DON'T PANIC |
| 02:24 | jwalden | 'hmm's at the conspicuous absence of Mozilla from the combatants list in that post |
| 02:24 | <jwalden> | and I suppose, to a much lesser degree, maybe Google |
| 02:40 | <othermaciej> | TabAtkins: are you around? |
| 02:40 | <TabAtkins> | Yo |
| 02:41 | <othermaciej> | TabAtkins: I'm thinking we should assemble some of the options from http://wiki.whatwg.org/index.php?title=Change_Proposal:_figure_and_details and submit them soon |
| 02:41 | <othermaciej> | TabAtkins: I'm gonna glue together the pieces for 3 in the w3c wiki but I would be happy if in terms of submitting on the mailing list you did 6 first |
| 02:42 | <TabAtkins> | Did 6 on the list, or did 6 on the w3c wiki? |
| 02:42 | <othermaciej> | I can also do 7 if you want, or you can feel free to steal my content from 3 and submit it on yourself |
| 02:42 | <othermaciej> | TabAtkins: I am expressing myself poorly and having a hard time rephrasing |
| 02:42 | <othermaciej> | let me try again |
| 02:42 | <TabAtkins> | Hehe. |
| 02:43 | <othermaciej> | - For actual submission to the mailing list, I'd prefer if you sent in 6 before I send in 3. |
| 02:43 | <othermaciej> | (I guess we should also ask Lachy if he wants to be listed as co-author) |
| 02:43 | <othermaciej> | - For assembling only the relevant pieces in the right order, I'm going to do that now, in the w3c wiki, so I can just mail a link to that when ready |
| 02:44 | <Hixie> | assuming the status quo and <legend> are non-options, from that list 2, 3, and 3b seem the best |
| 02:44 | <othermaciej> | I am going to include the names from 3b as alternatives in my 3 write-up |
| 02:44 | <TabAtkins> | Yeah, I think I like 3b the best too. |
| 02:45 | <TabAtkins> | I'd be willing to drop 6, since nobody else is really cheerleading it. |
| 02:45 | <othermaciej> | I could not decide which names are better (both sets seem good to me) |
| 02:45 | <othermaciej> | TabAtkins: I think it's good to have more alternatives on the table... maybe worth submitting since you did the writing and consider dropping later if no one likes it best |
| 02:45 | <TabAtkins> | Eh, kk. |
| 02:46 | <Hixie> | othermaciej: if it would save time, if the chairs can confirm that neither <dt>/<dd> nor <legend> are considered acceptable, i'd be happy to just do 3 or 3b |
| 02:46 | <Hixie> | or some hybrid thereof |
| 02:46 | <Hixie> | not sure which i prefer of <dlabel> and <dsummary> or if some other word would be better |
| 02:46 | <othermaciej> | Hixie: well, if no one writes a proposal defending them by the 16th, then neither will be acceptable |
| 02:47 | <othermaciej> | Hixie: I would rather not prejudge without seeing what arguments are mustered in their favor, if any |
| 02:47 | <othermaciej> | Hixie: but if you're willing to go along with one of the options presented then I do think that would save time |
| 02:49 | <TabAtkins> | So, want me to write up 6 tonight? I can do it while I'm letting the chinchillas out. |
| 02:49 | <TabAtkins> | Then I can write up 7 as a diff of 3 after you do it. |
| 02:50 | <othermaciej> | if you can do 6 that would be sweet |
| 02:50 | <othermaciej> | I am now assembling 3 at http://esw.w3.org/topic/HTML/ChangeProposals/DdDtFcaptionDlabel |
| 02:54 | <Hixie> | othermaciej: well, i can live with figcaption given that nobody seems to agree with me that we should use <legend>, and basically nobody at all seems to want <dt>/<dd>. I'm not a huge fan of <dlabel> and <dsummary>, but failing anything better I could live with one of those too. |
| 02:54 | <TabAtkins> | k, watching Chopped right now, will have by midnight my time (3 hours from now) |
| 02:55 | <Hixie> | maybe <summary>, though that would make the <table summary=""> people upset. (It would only be a problem if we kept summary="", which I don't think we should, but tying this to the summary="" issue seems like it would make the process complicated.) |
| 02:55 | <othermaciej> | Hixie: I'll probably list additional options (fcaption/figcaption/figurecaption and dlabel/detailslabel/dsummary/detailssummary/summary) |
| 02:55 | <TabAtkins> | Oh god, detailssummary is the worst possible name. Don't even suggest it. |
| 02:55 | <othermaciej> | TabAtkins: good point |
| 02:55 | <Hixie> | same with figurecaption :-) |
| 02:56 | <Hixie> | and detailslabel |
| 02:56 | <Hixie> | lslslslsl |
| 02:56 | <othermaciej> | all right |
| 02:56 | <Hixie> | :-P |
| 02:56 | <othermaciej> | fcaption/figcaption and dlabel/dsummary/summary as the options |
| 02:57 | <othermaciej> | and I may tweak which are presented as the default choice and/or leave picking one to a straw poll or editor's discretion |
| 02:57 | <TabAtkins> | I'm not sure what connection <summary> could possibly have to <table summary>, except the incidental name collision. There's nothing joining the two concepts at all, though. |
| 02:58 | <Hixie> | the name collision primarily |
| 02:58 | <Hixie> | also some people want a <summary> element for <caption> |
| 02:58 | <othermaciej> | Leif has also occasionally suggested a <summary> element as an extra way to give table accessibility descriptions |
| 02:58 | <Hixie> | though imho calling a table description a summary is misleading, but that's another story |
| 02:58 | <TabAtkins> | I don't see a problem, theoretically, with <caption> picking up a <summary> as well. It's not like it would be ambiguous. |
| 02:58 | <Hixie> | people didn't like that argument with <dt>/<dd> |
| 02:59 | <TabAtkins> | True. |
| 02:59 | <TabAtkins> | Eh. |
| 03:00 | <Hixie> | <* caption> is pretty funky |
| 03:00 | Hixie | fears he set a bad precedent with <time pubdate> |
| 03:00 | <TabAtkins> | Heh, that is indeed the precedent I was leaning on. |
| 03:12 | <othermaciej> | http://esw.w3.org/topic/HTML/ChangeProposals/DdDtFcaptionDlabel is done |
| 03:12 | <othermaciej> | by the way, whatever flavor of wiki that is, I like it a lot less than whatever flavor the whatwg wiki is |
| 03:12 | <Hixie> | whatwg is mediawiki, same as wikipedia |
| 03:12 | <Hixie> | esw is moinmoin |
| 03:13 | <othermaciej> | I like the MediaWiki syntax, styling, and set of easily accessible capabilities much better than MoinMoin |
| 03:13 | <othermaciej> | apparently |
| 03:25 | <othermaciej> | 231 bugs resolved so far this month, 77 still open |
| 03:25 | <othermaciej> | the influx is kinda crazy this month |
| 04:36 | <TabAtkins> | othermaciej: http://esw.w3.org/topic/HTML/ChangeProposals/DdDtCaptionAttribute |
| 04:38 | <othermaciej> | TabAtkins: neat |
| 04:38 | <TabAtkins> | Doing 7 right now |
| 04:38 | <othermaciej> | TabAtkins: you probably also want to copy over the general rationale against dt/dd |
| 04:38 | <TabAtkins> | The hugormous section you have in yours? |
| 04:40 | <othermaciej> | the one that I copied from Lachlan's page |
| 04:40 | <othermaciej> | (I assumed the point of his page was to share at least that part of the rationale) |
| 04:47 | <othermaciej> | TabAtkins: just updated http://esw.w3.org/topic/HTML/ChangeProposals/DdDtFcaptionDlabel to have the alternate names included (figcaption, dsummary, etc) |
| 04:47 | <TabAtkins> | kk. |
| 04:47 | <TabAtkins> | I'm doing 7 just as a diff on that one, so no problems here. |
| 04:48 | <othermaciej> | TabAtkins: if you wanted to copy it wholesale for 7, I believe all you need to change is the Details section (two more new elements, plus change the content models of <figure> and <details> appropriately), and reverse the argument in "Does Not Introduce Redundant Elements for the Content" |
| 04:48 | <othermaciej> | to something like "Introduces Convenient Optional Elements for the Content |
| 04:48 | <TabAtkins> | Yup, pretty much exactly what I'm doing. |
| 04:52 | <TabAtkins> | Hmm. Multiple <fbody>s in a <figure> - should they all be the figure content? Like how you can have multiple <tbody>s? |
| 04:53 | <othermaciej> | I think it should only be valid to have one - no thought on what should happen if you specify multiple |
| 05:18 | <TabAtkins> | othermaciej: http://esw.w3.org/topic/HTML/ChangeProposals/DdDtFbodyDbody, and I'm off to bed. |
| 05:18 | <othermaciej> | TabAtkins: right on |
| 05:19 | <erlehmann> | othermaciej, TabAtkins, i made a wordpress plugin that uses the figure / dd / dt construct: http://blog.dieweltistgarnichtso.net/html5-cc-plugin-fur-wordpress-version-0-6 |
| 05:19 | <erlehmann> | what would you suggest to use until that issue is cleared ? nested divs ? |
| 05:43 | <othermaciej> | erlehmann: if we're lucky we can have this one settled within a week or two |
| 05:43 | <othermaciej> | erlehmann: don't know what I would suggest in the meantime |
| 05:44 | <erlehmann> | wish me luck then |
| 06:01 | <erlehmann> | othermaciej, i hope that will go over the mailing list again ? i normally don't read #whatwg logs. |
| 06:01 | <othermaciej> | erlehmann: hope what will go over the mailing list? |
| 06:02 | <erlehmann> | the discussion about changing figure. |
| 06:02 | <erlehmann> | there wasn't anything in the last 2 weeks concerning that if i am right ? |
| 06:03 | <othermaciej> | there will definitely be mention on public-html |
| 06:03 | <othermaciej> | I guess it would be good to also mention it on the whatwg list |
| 06:04 | <erlehmann> | oh, i don't read public-html. whatever, ping me in chat if it changes or something. thanks for the info btw :) |
| 06:11 | <Hixie> | othermaciej, i'm guessing you posted that to the wrong list |
| 06:11 | <othermaciej> | Hixie: oops, yeah, bad habit |
| 06:12 | <othermaciej> | can't believe I didn't even notice |
| 06:19 | <othermaciej> | Hixie: whatever is causing that bug, though, it's really annoying! |
| 06:24 | <Hixie> | i get weird hangs in safari quite a lot of i leave html5 open for a lot |
| 06:24 | <Hixie> | s/lot/long time/ |
| 06:24 | <Hixie> | haven't filed a bug because i can't reliably reproduce it and the profiles i got from activity monitor are months old now |
| 06:24 | <Hixie> | and i haven't seen it for a while |
| 06:24 | <Hixie> | but that's mostly because i've been using chrome more recently |
| 06:25 | <Hixie> | gonna have to switch to firefox for a while in a few months, to keep all the browsers with equal time on my machine :-P |
| 06:26 | <othermaciej> | Hixie: I'm pretty sure it's not the same thing unless you started serving Flash ads on the html5 spec |
| 06:26 | <Hixie> | indeed, different issue |
| 06:28 | <othermaciej> | I have also never seen it, though opening multiple copies of the spec can make things slow for quite some time |
| 06:29 | <Hixie> | it might be some leopard thing (i'm still on leopard on my google laptop) |
| 06:36 | <othermaciej> | ah, my bug would be SnowLeopard-specific |
| 07:09 | <roc> | mmm, Flash_EnforceLocalSecurity, my favouritest symbol |
| 07:32 | <othermaciej> | roc: I know, right? |
| 07:32 | <othermaciej> | roc: someone reminded me of the Flash bug that I'm likely experiencing but there's clearly also a bug in WebKit's out-of-process plugin support in that we let the browser get hung on a Flash hang |
| 07:33 | <jtbandes> | Oh, interesting, I frequently get hangs/crashes on that in Safari. I've never bothered to investigate. |
| 07:41 | <othermaciej> | Flash_EnforceLocalSecurity is just the only function in Flash that appears to have a symbol |
| 07:42 | <othermaciej> | so all Flash crashes or hangs appear to be in that function |
| 07:42 | <jtbandes> | Oh, I see |
| 07:42 | <othermaciej> | so it's kind of a joke among browser developers |
| 07:42 | <jtbandes> | Hehe. |
| 07:44 | <roc> | they should rename it to HTML5Parser or something |
| 07:45 | <othermaciej> | lol |
| 07:45 | <hsivonen> | at least the real HTML5 parser is less crashy than Flash |
| 07:46 | <hsivonen> | (would be pretty sad if that weren't the case) |
| 07:47 | <hsivonen> | (maybe the HTML5 parser on the 1.9.2 branch should be named Flash) |
| 08:02 | <zcorpan__> | "Andy, you are advocating Tag Soup. Yes, nobody will die from this. However, this practice makes it more difficult for tool vendors to process HTML which increases their costs and creates an uncompetitive environment. " - http://stuffandnonsense.co.uk/blog/about/keep_calm_and_carry_on_with_html5#r4080 |
| 08:03 | <hsivonen> | confusing use of quotation marks there |
| 08:03 | <zcorpan__> | seems vlad has not understood that tool vendors can use off-the-shelf html5 parsers to process html, which reduces teh cost and creates a competetive environment |
| 08:03 | <hsivonen> | that seems odd |
| 08:04 | <hsivonen> | considering that Vlad is a tool vendor |
| 08:04 | <zcorpan__> | indeed |
| 08:04 | <hsivonen> | and, therefore, one could expect him to be familiar with the serializer APIs |
| 08:05 | <tantekc> | greetings |
| 08:05 | <othermaciej> | good evening |
| 08:06 | <zcorpan__> | i also don't see where andy is advocating tag soup |
| 08:06 | <othermaciej> | depends on what you mean "tag soup" |
| 08:07 | <othermaciej> | he's clearly advocating early use of not-quite-yet-standard features |
| 08:07 | <zcorpan__> | unless vlad considers <a><div></div></a> to be tag soup |
| 08:07 | <Hixie> | that's valid in HTML5 |
| 08:07 | <othermaciej> | maybe even advocating not worrying much about whether you follow required content models |
| 08:08 | <othermaciej> | he doesn't seem to be calling for writing stuff like <i>foo <b>bar</i> baz</b> |
| 08:08 | <othermaciej> | or in general for writing things that only get parsed sorta by accident |
| 08:08 | <zcorpan__> | my interpretation is "use cutting-edge if it works and don't worry so much about it" |
| 08:09 | <othermaciej> | "tag soup" isn't a very enlightening term |
| 08:09 | <hsivonen> | well, I learned something new by reading Lachy's comment on Vlad's blog |
| 08:09 | hsivonen | was unaware of http://en.wikipedia.org/wiki/Australia-Asia_debate |
| 08:10 | <zcorpan__> | hsivonen: comment where? |
| 08:11 | <hsivonen> | (relative to British Parliamentary Style) |
| 08:13 | Philip` | thought that page would be about a debate on whether Australia is a part of Asia or not |
| 08:13 | <hsivonen> | zcorpan__: http://rebuildingtheweb.com/en/invitation-html5-team/#c20100107043000 |
| 08:15 | <othermaciej> | that s strangely fascinating |
| 08:16 | <othermaciej> | the kind of debate that I believe is most common in america is "forensic debate" or something like that |
| 08:16 | <hsivonen> | I read "traditional style 3 vs. 3 approach" and thought "what? isn't *traditional* 4 vs. 4" |
| 08:17 | <zcorpan__> | ooh debate software |
| 08:17 | <hsivonen> | and then checked Wikipedia to see if Australia deviates from the British tradition |
| 08:17 | <zcorpan__> | never heard of that before |
| 08:17 | <hsivonen> | zcorpan__: that was interesting, too |
| 08:21 | <othermaciej> | maybe the W3C should adopt parliamentary rules of debate |
| 08:21 | <othermaciej> | (and/or procedure in general) |
| 08:21 | <othermaciej> | although I would not enjoy Prime Minister's Question Hour |
| 08:22 | <Hixie> | i thought the kind of debate that the US usually was called "the media tries to convince the candidates that they won't be made to look foolish so that they can make money on ads" |
| 08:22 | <Hixie> | usually has, i should say |
| 08:23 | <othermaciej> | I don't think anyone considers the type of verbal challenge events that candidates for political office engage in to be a serious form of debate |
| 08:23 | <Hixie> | :-) |
| 08:24 | <hsivonen> | othermaciej: sometimes I wish debates at the W3C were adjudicated so that one would lose for making randomly self-contradicting statements |
| 08:24 | <hsivonen> | but otherwise, I really don't welcome BP kind of debate to the W3C |
| 08:24 | <othermaciej> | hsivonen: that would be unfair to people with poor logical reasoning skills |
| 08:25 | <hsivonen> | since one of the aspects of BP is boldly making up bullshit in the hope you don't get caught |
| 08:25 | <othermaciej> | then I think we might already be on that system :-) |
| 08:26 | <Hixie> | hsivonen: w3c debates are supposed to be adjudicated, that's what the chairs are supposed to be doing |
| 08:26 | <Hixie> | they rarely do, for some reason that i don't understand |
| 08:29 | <othermaciej> | in the HTML WG, the chairs try to intervene to prevent abusive behavior (granted we don't always succeed) but generally prefer to avoid ruling on validity of arguments or debating tactics |
| 08:29 | <Hixie> | picking whatever argument is convenient for supporting some agenda, even when those arguments contradict each other, is abusive behaviour. |
| 08:30 | <Hixie> | at least imho. |
| 08:30 | <hsivonen> | in BP, the adjudicator is not supposed to rule on the vecity of bullshit but on the internal consistency of the bullshit |
| 08:31 | <Lachy> | hsivonen, when I posted that comment, I just assumed it was a common debating technique everywhere, since that's basically the only style I group up with |
| 08:31 | <Lachy> | and it roughly matched the style I'd seen on American TV shows |
| 08:31 | <hsivonen> | even in the W3C WGs it would be useful if the chairs called blatant logical contradictions that are detectable without judging the veracity of the arguments |
| 08:32 | <Hixie> | yeah i'm not suggesting that subtle argumental contradictions should be called out, they are often the result of people changing their mind over time or making honest mistakes |
| 08:32 | <Hixie> | i just mean blatent contradictions from day to day |
| 08:33 | <hsivonen> | changing one mind is different from spinning stuff on the spot |
| 08:33 | <hsivonen> | one's |
| 08:33 | <Hixie> | right |
| 08:33 | <othermaciej> | I think it would be ok for people to just point out each other's contradictions |
| 08:34 | <othermaciej> | I am not sure contradicting yourself is per se abusive, unless used to create drama or filibuster a topic or the like |
| 08:35 | <Hixie> | hard to say in the hypothetical |
| 08:35 | <othermaciej> | I think ad hominem is the one fallacy that is clearly nearly always abusive regardless of context |
| 08:36 | <tantekc> | indeed |
| 08:36 | <Hixie> | it's certainly more psychologically harmful |
| 08:36 | <othermaciej> | but even an argument from authority is generally merely invalid as opposed to abusive, and informal debate citing an authority may not even be invalid |
| 08:37 | <othermaciej> | (if it is a piece of supporting evidence rather than your whole case, say) |
| 08:37 | <tantekc> | there is a rhetorical device used by pundits to "challenge to a one-one debate" in political circles etc. as a method of claiming victory (by lack of answer to the challenge) - I don't remember what this specific rhetorical device is called though - I'm pretty sure it has a name |
| 08:38 | <Hixie> | othermaciej: i was refering more to someone saying "i think we should do a, not b, because z" one day and "i think we should do c, not d, because not-z" the next |
| 08:38 | <tantekc> | othermaciej - yes, argument from authority is perhaps a weak/indirect form of evidence ("experience" or "expertise") but is both better than nothing / novice opinion, and certainly better than ad hominem or other logically flawed attacks. |
| 08:39 | <Hixie> | othermaciej: though "i don't have to defend my position, but i will object to the hilt your suggestion that i may be wrong, and i think your arguments are irrelevant" is equally abusive, imho |
| 08:39 | <othermaciej> | however saying something like "I don't have to answer your reasoned arguments because I am the top expert in the field" would be a bad form of argument from authority |
| 08:40 | <Hixie> | othermaciej: since it precludes compromises from being derived |
| 08:40 | <Hixie> | yeah |
| 08:40 | <othermaciej> | Hixie: being constantly contradictory without an supporting arguments would likely be abusive |
| 08:40 | <zcorpan__> | i don't see the need for <fbody> and <dbody> |
| 08:41 | <othermaciej> | refusal to compromise - that is borderline |
| 08:41 | <othermaciej> | sometimes to compromise would be unprincipled |
| 08:41 | <othermaciej> | refusal to even hear others out - that would certainly be bad form |
| 08:41 | <Hixie> | refusal to compromise isn't abusive, but not providing any reasons such that it is impossible to even try to find other solutions that might satisfy all requirements is, imho |
| 08:42 | <Hixie> | that's why in the whatwg i only look at the arguments, and not the volume of opinions, and occasionally come up with completely novel solutions that nobody was looking for but which people are surprisingly happy with |
| 08:42 | <Hixie> | and why i find the polarising argument style in the htmlwg to be unproductive |
| 08:42 | <othermaciej> | sure, disagreeing without any supporting arguments is not productive |
| 08:42 | <othermaciej> | repeating the same arguments is also often unproductive |
| 08:42 | <Hixie> | yeah |
| 08:42 | <Hixie> | or ignoring other people's |
| 08:43 | <othermaciej> | complaining about things that are bad without specific and practical suggestions on how to make things better, also not super productive |
| 08:43 | <tantekc> | polarizing (artificial dichotomy) only makes sense when you're appealing to emotion rather than reason ;) |
| 08:43 | <zcorpan__> | i like <fcaption> and <dlabel> without explicit bodies best |
| 08:43 | <othermaciej> | zcorpan__: me too (or one of the various similar names) |
| 08:43 | <zcorpan__> | not <figcaption> because it's different from <thead> |
| 08:43 | <zcorpan__> | and longer |
| 08:44 | <Hixie> | i'm not a fan of <fcaption>, but i can't work out why. |
| 08:44 | <tantekc> | if you abbreviated further, you could bust a <dcap> |
| 08:44 | <tantekc> | or an <fcap> |
| 08:44 | <Hixie> | i could live with <fcaption> and <dlabel> if there's really no chance of the imho far superior <legend> :-) |
| 08:45 | <Hixie> | (which i presume there isn't) |
| 08:45 | <othermaciej> | when I think a thread is off the rails enough to need active moderation, the main things I ask for are that people not overly repeat arguments that are already on the record, and that people focus on concrete use cases or proposals |
| 08:45 | <othermaciej> | fcaption when you pronounce it sounds vaguely like a word that means something |
| 08:45 | <Lachy> | Hixie, a new element is superior to legend. But reusing legend is preferable to reusing any other element. |
| 08:45 | <othermaciej> | that might be what seems weird about it |
| 08:45 | <othermaciej> | it could be the name of a perfume |
| 08:46 | <othermaciej> | Effcaption... by Calvin Klein |
| 08:46 | <hsivonen> | the main problem with not compromising is that under a "can you live with it" regime, other people who normally would compromise have to also adopt an uncompromising line, because otherwise those who don't budge get their way over and over again |
| 08:46 | <Hixie> | hsivonen: i'm not talking about compromising really but about finding solutions that are equally good for you, but also good for others |
| 08:46 | <othermaciej> | hsivonen: I don't entirely buy into "can you live with it" as the standard for anything but truly bikeshed-type issues |
| 08:47 | <Hixie> | hsivonen: e.g. if someone wants a rail bridge and someone else wants a car ferry, it may be that a car/rail bridge would equally please both camps -- depends on their reasons |
| 08:47 | <Hixie> | hsivonen: or it could be that a tunnel would be a solution |
| 08:47 | <othermaciej> | the W3C process on managing disagreement stresses strength of objections, and as often I like to say, I believe "strength" refers to quality of their supporting arguments, not the vehemence with which they are expressed |
| 08:48 | <Hixie> | hsivonen: depends if the reasons are "has to carry cars" or "has to not be ugly" or "has to support freight" or "has to support tourist passenger rail cars" |
| 08:48 | <othermaciej> | in fact the W3C process specifically says you can't halt progress just by saying you cannot live with a decision |
| 08:49 | <Philip`> | http://rebuildingtheweb.com/en/invitation-html5-team/ - hmm, I think I made all the comments disappear :-( |
| 08:49 | <Philip`> | All I did was post one with a U+FFFF |
| 08:49 | <othermaciej> | indeed, sometimes there are better solutions outside of the box, and they can be hard to find if people are not clear about their requirements and motivations |
| 08:49 | <zcorpan__> | it would be pretty bad if someone killed himself because of a decision in w3c |
| 08:49 | <hsivonen> | Philip`: Awesome. XML FTW! |
| 08:49 | <othermaciej> | Philip`: harsh |
| 08:50 | <zcorpan__> | at lease i don't see an error message |
| 08:50 | <Philip`> | The preview page just gave a well-formedness error, I didn't realise it'd break the main page like that |
| 08:50 | <jgraham> | zcorpan__: <dbody> seems rather useful because it seems rather common to want to e.g. give some extra left margin to the details expansion |
| 08:51 | <Hixie> | why can't <div> do that |
| 08:51 | <zcorpan__> | jgraham: you can have a negative margin on dlabel |
| 08:51 | <zcorpan__> | yeah or div |
| 08:51 | <jgraham> | Hixie: It can be Why make the pattern <details><dlabel></><div class="dbody"> |
| 08:52 | <jgraham> | s/be/but/ |
| 08:52 | <zcorpan__> | there are lots of useful things we could add to help styling |
| 08:52 | <Hixie> | because the real solution is power in css |
| 08:52 | <Hixie> | don't fix css by adding elements to html |
| 08:53 | <jgraham> | since power in CSS generally implies slowness, I guess implementors might disagree :) |
| 08:54 | <jgraham> | (anyway it seems somewhat reasonable to have a grouping element in a situation in which we know authors will often want to deal with just the children of that element) |
| 08:55 | <MikeSmith> | Hixie: a follow-up about the cvs locking problem: I'm told a change has now been made the should prevent the problem from ever recurring. But if/when you ever notice it hanging again on commits, please let me know |
| 08:55 | <Hixie> | k |
| 08:55 | <Hixie> | thanks |
| 08:56 | <Hixie> | jgraham: we have a grouping element here -- <div> |
| 08:56 | <Hixie> | that's what it's for |
| 08:56 | <Hixie> | grouping for CSS |
| 08:56 | <Hixie> | and script |
| 08:57 | <othermaciej> | I think <dbody> would serve no semantic or implementation-level purpose, which would make it purely a styling hook, so overall I am not sold, but I can see how it can be convenient |
| 08:57 | <zcorpan__> | i'd add <di> before <dbody> |
| 08:58 | <Hixie> | even i would add <di> before <dbody> :-) |
| 08:58 | <Lachy> | yeah, just use <div> |
| 08:59 | <othermaciej> | it seems like "details > div" would be a fine way to add on your styling |
| 08:59 | <othermaciej> | and if dbody is optional it will be no more effective at truly catching all your details bodies than div |
| 08:59 | <hsivonen> | I want to see how pageshow fires when an iframe is inserted to the doc by the parser |
| 08:59 | <hsivonen> | should I expect to be able to inject a listered via the prototype of HTMLIframeElement somehow? |
| 09:00 | <hsivonen> | *listener |
| 09:00 | <hsivonen> | also, are Gecko and WebKit the only engines that support pageshow/pagehide? |
| 09:01 | <zcorpan__> | pageshow fires on iframes? |
| 09:01 | <zcorpan__> | hsivonen: yes, afaik |
| 09:01 | othermaciej | still hates the way pageshow is defined |
| 09:01 | <roc> | Philip`: brilliant! |
| 09:01 | <othermaciej> | it would be so much better if it fired only on returning via cache and not on initial page load |
| 09:02 | <Hixie> | othermaciej: in the spec or in UAs? |
| 09:02 | <hsivonen> | zcorpan__: I want to test if and how it fires. |
| 09:02 | <Hixie> | i think the spec just matches what browsers do |
| 09:02 | <othermaciej> | it does |
| 09:02 | <roc> | I know the name Vlad Alexander from somewhere |
| 09:02 | <othermaciej> | I think Gecko's longstanding implementation makes it impossible to change now |
| 09:02 | <tantek> | Philip` - that comment trick was harsh but effective. |
| 09:02 | <othermaciej> | but it's still not so good |
| 09:03 | <roc> | ah yes, he works on an XHTML editor |
| 09:03 | <othermaciej> | it sucks that it fires on initial load because: |
| 09:03 | <Philip`> | tantek: It wasn't effective since it didn't cause error messages in browsers |
| 09:03 | <jgraham> | I think my primary objection to <div> is that it does nothing for readability. It seems like a shame to design in a patern where a <div> will likely be used more often than not |
| 09:03 | <zcorpan__> | othermaciej: it's not impossible to change if changing it doesn't break sites |
| 09:04 | <tantek> | only effective in making a point about the ridiculousness of draconianism |
| 09:04 | <othermaciej> | a) it makes it awkward to do reinitialization that should only be done when restoring from page cache; you have to record manually whether it fired once before |
| 09:04 | <othermaciej> | b) you can't use pageshow as your sole event even for first-time init anyway; you have to use load for browsers that don't (yet) have pageshow |
| 09:04 | <Hixie> | jgraham: if that becomes the case we can always add a different kind of <div> later |
| 09:04 | <Hixie> | jgraham: as we did with <Section> |
| 09:04 | <Philip`> | tantek: I think the point is more about the problems of output validation being stricter than input validation |
| 09:05 | <Philip`> | It's okay if they're both equally strict |
| 09:05 | <othermaciej> | and if first-time init and restore share some but not all code, it's much easier to call one function from two event handlers than to make one event handler behave differently depending on if it has been invoked before |
| 09:05 | <Philip`> | but XML is stricter than everybody's input validation |
| 09:05 | <othermaciej> | zcorpan__: sites use it - I don't know how much they depend on the Gecko semantic |
| 09:05 | <hsivonen> | tantek: I think it demonstrates that it's silly to ban arbitrary characters |
| 09:05 | <zcorpan__> | i see a .presisted attribute on pageshow and pagehide events in http://webkit.org/blog/516/webkit-page-cache-ii-the-unload-event/ |
| 09:05 | <othermaciej> | it's a pet peeve, bit I am not sure it is worth fighting the battle at this point |
| 09:05 | <tantek> | I always figured the proper rhetorical response to draconialists was to simply stop reading their arguments at the first grammatical or spelling error that they made, especially in speech. |
| 09:06 | <hsivonen> | tantek: the Name production in XML is an even worse example of spec authors banning stuff Draconianly based on aesthetics |
| 09:07 | <hsivonen> | zcorpan__: I think I've see .persisted in Gecko, too |
| 09:07 | <hsivonen> | *seen |
| 09:07 | <Philip`> | hsivonen: Do you think it's also silly to ban things that are not really characters at all, like U+D800 or like malformed UTF-8 sequences? |
| 09:07 | <othermaciej> | zcorpan__: I guess you can just avoid ever putting code in the !evt.persisted case |
| 09:07 | <othermaciej> | zcorpan__: so it's not too terrible, just slightly suboptimal |
| 09:08 | <zcorpan__> | the event's interface is PageTransitionEvent in chrome |
| 09:08 | <zcorpan__> | html5 just uses Event |
| 09:09 | <hsivonen> | Philip`: I think it's reasonable to turn lone surrogates or malformed UTF-8 into REPLACEMENT CHARACTER |
| 09:09 | <zcorpan__> | firefox same as chrome |
| 09:14 | <hsivonen> | hmm. so how can I add listeners via the prototype? |
| 09:15 | <Lachy> | hsivonen, you can't. |
| 09:15 | <Philip`> | Ah, good, looks like he fixed my comment |
| 09:15 | <Lachy> | event listeners can only be added to instances of elements |
| 09:16 | <Lachy> | but you can create an element and add event listeners before you add it to the DOM |
| 09:17 | <hsivonen> | I was thinking of detecting pageshow for parser-inserted iframes |
| 09:19 | <zcorpan__> | hsivonen: you could use onpageshow="" |
| 09:20 | <hsivonen> | zcorpan__: is it actually supported by all browsers that fire the event? |
| 09:20 | <Hixie> | where are you looking for this event? |
| 09:21 | <zcorpan__> | hsivonen: it's supported in firefox, not in chrome |
| 09:21 | <Hixie> | in the document with the iframe, or the doc in the iframe? |
| 09:21 | <hsivonen> | Hixie: in the parent doc |
| 09:21 | <zcorpan__> | hsivonen: but it bubbles |
| 09:21 | <Hixie> | just set a capture listener |
| 09:21 | <zcorpan__> | hsivonen: doesn't fire on iframes |
| 09:21 | <Hixie> | but why would pageshow fire on the iframe parent doc? |
| 09:22 | <hsivonen> | Hixie: I want to capture the events for iframe navigation from the parent |
| 09:22 | <hsivonen> | zcorpan__: hmm. that complicates testing |
| 09:22 | <Hixie> | i'm not aware of a way to do that |
| 09:22 | hsivonen | thought an iframe was event-wise equivalent to a XUL browser |
| 09:23 | <hsivonen> | if pages can't listen to iframe pageshow from the outside, perhaps I shouldn't care about it |
| 09:23 | <hsivonen> | since it can't be compat-sensitive for about:blank if it isn't observable |
| 09:23 | <zcorpan__> | pageshow has the same weirdness as the load event in being fired on window but has target document |
| 09:24 | hsivonen | knows way too little about the quirks of DOM events to be properly testing this stuff |
| 09:26 | <Lachy> | I wonder what Larry's motives are for his attempt at twisting the meaning of the charter beyond any reasonable interpretation? http://lists.w3.org/Archives/Public/public-html/2010Jan/0507.html |
| 09:27 | <Hixie> | he probably wants namespaces in html |
| 09:27 | <Lachy> | yeah, probably |
| 09:27 | <Hixie> | but i think he's missing the fact that both rdfa and microdata can be used to create entire vocabularies in text/html |
| 09:27 | <Hixie> | just like namespaces in html |
| 09:27 | <Hixie> | so... |
| 09:28 | <Hixie> | maybe someone should ask him what kind of vocabulary can't be expressed in rdfa |
| 09:29 | <Lachy> | I think the problem is that the charter has not been written clearly enough. |
| 09:29 | <Lachy> | and I hope we can get these issues fixed when this group recharters next year |
| 09:29 | <Hixie> | that's usually intentional, since otherwise the groups get never get chartered |
| 09:29 | <Hixie> | you have to write them in a way that everyone sees what they want to see so they vote "yes" |
| 09:30 | Hixie | isn't a fan of charters for wgs |
| 09:30 | <othermaciej> | rechartering will be a world of hurt, I expect |
| 09:30 | <Lachy> | if it's really intentional, then it sucks cause it just creates an environment for charter lawyers to waste time |
| 09:32 | <Lachy> | othermaciej, will WG members be given the opportunity to give feedback on what the new charter should say when the time comes? |
| 09:32 | Philip` | hopes not |
| 09:32 | <othermaciej> | Lachy: I'm actually not very clear on how rechartering mechanics go |
| 09:32 | <Philip`> | since it'll cause everything to get discussed again |
| 09:32 | <zcorpan__> | maybe the w3c decides to not renew the charter |
| 09:32 | <othermaciej> | also I need to get to bed |
| 09:33 | <Lachy> | Philip`, why? Otherwise, we'll end up with a mess like we have now, which was written by people who ignored outside input, especially the input from Hixie |
| 09:35 | <Hixie> | Lachy: i don't really see the point in charters in general. People are going to work on what they want to work on, what's the point of having them go through long complicated processes to get a mailing list to work on them? |
| 09:35 | <Philip`> | Lachy: Because it's a mess we can generally ignore, vs weeks/months of effort wasted on discussions of what should and shouldn't be in HTML and everybody trying to get the charter to say things that will support all their future arguments about everything |
| 09:35 | <webben> | does the parsing definition + deference to other applicable specs constitute "a" mechanism. |
| 09:35 | <Lachy> | Hixie, personally, I agree, but the W3C requires charters and it seems to me the only way to avoid having crazy people try to reinterpret the charter to suit their agenda is to make sure it's written clearly |
| 09:36 | <Hixie> | Lachy: yeah but then those crazy people would vote it down |
| 09:36 | webben | thought chartering was so companies could decide whether they risked any ip by engaging with a wg. |
| 09:37 | <othermaciej> | that's really the main practical effect of chartering |
| 09:37 | <othermaciej> | knowing the scope of IP that could be entangled |
| 09:38 | <othermaciej> | it also seems like at least a nonbinding statement of purpose for a group is useful to decide if you want to participate, regardless of IP issues |
| 09:39 | <webben> | othermaciej: Agreed. |
| 09:41 | <Hixie> | othermaciej: oh certainly a statment of purpose is useful, yes |
| 09:41 | <zcorpan__> | does the <source> error event bubble per spec yet? |
| 09:41 | <Hixie> | othermaciej: in practice i think the patent thing is only an issue for patent-hostile companies (i.e. companies that might want to use their patents to pro-actively sue someone rather than just defensively) |
| 09:42 | <zcorpan__> | seems not |
| 09:43 | <Hixie> | othermaciej: but even if a company is partially that way, it isn't like they're going to have a problem with what's in scope and what is slightly out of scope |
| 09:43 | <MikeSmith> | Hixie: I'm relooking at the script element text-content BNFs, and wanted to ask if it's intended that the following be conformant: |
| 09:43 | <MikeSmith> | <script type="text/javascript" src="url"> |
| 09:43 | <MikeSmith> | / foo |
| 09:43 | <MikeSmith> | <!-- --> |
| 09:43 | <MikeSmith> | / bar |
| 09:43 | <MikeSmith> | </script> |
| 09:43 | <MikeSmith> | oops |
| 09:44 | <MikeSmith> | (with double slashes instead) |
| 09:45 | <Hixie> | why would it not be? |
| 09:45 | <Hixie> | oh wait that's the documentation thing |
| 09:45 | <MikeSmith> | yeah |
| 09:45 | <Hixie> | no, that wouldn't be conforming |
| 09:45 | <Hixie> | since not everything is commented out |
| 09:46 | <MikeSmith> | OK |
| 09:46 | <zcorpan__> | hmm, <img> error doesn't bubble in firefox or webkit |
| 09:46 | <zcorpan__> | does it bubble in ie? |
| 09:50 | <MikeSmith> | Hixie: so I regard to the note in section 4.3.1.3 (the section about the script@src case) that says, "This requirement is in addition to the earlier restrictions on the syntax of contents of script elements." -- I'm trying to think of a case of script@src text contents that would conform to the 4.3.1.3 BNF on its own, but not conform to the 4.3.1.2 BNF |
| 09:50 | <MikeSmith> | *in regard to |
| 09:51 | <MikeSmith> | Hixie: <script src=url>/* <-- */</script> maybe? |
| 09:52 | <Hixie> | <script src=''>//<!--<script></script> |
| 09:55 | <zcorpan__> | Hixie: there's a missing end tag... |
| 09:56 | <Hixie> | hm? |
| 09:57 | <zcorpan__> | Hixie: in the above snippet, my brain finds EOF in script data escaped |
| 09:57 | <MikeSmith> | yeah, seems like |
| 09:58 | <Hixie> | that's the idea |
| 09:58 | <zcorpan__> | and textContent is "//<!--<script></script>" |
| 09:58 | <Hixie> | er |
| 09:58 | <Hixie> | oh i see what you're saying |
| 09:58 | <Hixie> | yes, you couldn't do this in text/html |
| 09:58 | <Hixie> | i was giving a DOM serialisation |
| 09:58 | <MikeSmith> | OK |
| 09:58 | <Hixie> | not sure if it's possible to do it in text/html |
| 09:59 | <Hixie> | it certainly is in XML |
| 09:59 | <Hixie> | or via script |
| 10:00 | <MikeSmith> | the context for my question is that I'm trying to figure out if there's any checking that could be added that would need to check that a document conforms both to the script@src BNF and the the script (with not src) BNF |
| 10:00 | <MikeSmith> | it seems like maybe there's not |
| 10:00 | <MikeSmith> | as far as conformance checking of serialized documents goes |
| 10:01 | <MikeSmith> | because seems like anything that would not be conformant is going to cause a parse error first |
| 10:01 | <MikeSmith> | for that case |
| 10:01 | <zcorpan__> | MikeSmith: not in xhtml5 |
| 10:02 | <MikeSmith> | ah |
| 10:02 | <MikeSmith> | wonderful |
| 10:03 | <Hixie> | btw if whatwg.org and co are down in the coming days it's likely because of a server move |
| 10:04 | <Hixie> | i'm upgrading to the newer private virtual servers |
| 10:04 | <MikeSmith> | cool |
| 10:04 | <MikeSmith> | so you will be running on a VM that you have root access to? |
| 10:04 | <Hixie> | hopefully |
| 10:04 | <MikeSmith> | nice |
| 10:08 | <MikeSmith> | <script src=''>//<!--<script></script> still gives a parse error, because of the "--" |
| 10:08 | <hsivonen> | I see Gecko firing pagehide and pageshow inside an iframe but WebKit doesn't appear to do that |
| 10:08 | <MikeSmith> | in XML |
| 10:09 | <Hixie> | just put it all in a <![CDATA[ ... ]]> escape |
| 10:09 | <hsivonen> | is navigating an iframe supposed to clear the event listeners on its contentWindow? |
| 10:10 | <hsivonen> | pardon my cluelessness here |
| 10:12 | <hsivonen> | hmm. apparently not |
| 10:12 | <hsivonen> | I misinterpreted something |
| 10:12 | <zcorpan__> | hmm seems webkit fires two error events on <source> and firefox fires a bubbling error event on <video> :( |
| 10:21 | <MikeSmith> | Hixie, zcorpan__ : so that following should be non-conformant in XHTML? |
| 10:21 | <MikeSmith> | <script src="url">//<![CDATA[<script src=''><!--<script>]]></script> |
| 10:21 | <MikeSmith> | oops |
| 10:21 | <Hixie> | it wouldn't be, no, but you can make simpler examples :-) |
| 10:22 | <Hixie> | hsivonen: you get a whole new contentWindow if you navigate an iframe |
| 10:22 | <zcorpan__> | <script src=''>//<!--<script></script> |
| 10:22 | <Hixie> | hsivonen: thought it would be === to the previous one due to the magic of WindowProxy |
| 10:22 | <zcorpan__> | the web platform is insane |
| 10:47 | <Hixie> | hm, jgraham isn't here |
| 10:47 | <Hixie> | http://pimpmyspec.net/ is down |
| 10:54 | foolip | would like to see the issue of <source> error bubbling addressed ASAP |
| 10:54 | <foolip> | (I want to let it bubble, like Firefox) |
| 10:56 | <gsnedder1> | Hixie: Known bug. |
| 10:57 | <gsnedders> | gsnedders: Server is down, which is why he isn't here. |
| 10:57 | <gsnedders> | Hixie: ^^ |
| 10:57 | <zcorpan__> | foolip: but <img> error doesn't bubble |
| 11:00 | <foolip> | zcorpan__: yes it does, in Opera at least |
| 11:00 | <foolip> | what does the spec say? |
| 11:00 | <zcorpan__> | foolip: not in firefox or webkit |
| 11:00 | <zcorpan__> | spec says not to bubble |
| 11:00 | <zcorpan__> | i filed a bug about that today |
| 11:01 | <zcorpan__> | CORE-26763 |
| 11:01 | <zcorpan__> | haven't tested ie |
| 11:04 | <zcorpan__> | Hixie: shouldn't sandbox also block openIndexedDatabase() or whatever it's called? |
| 11:06 | <Hixie> | probably, but that's up to whoever is editing that these days |
| 11:06 | <Hixie> | gsnedders: k |
| 11:07 | gsnedders | haz magic move-mouth communication method still available with jgraham |
| 11:07 | <foolip> | ok, but IIRC error bubbles from <source> in FF |
| 11:08 | <foolip> | so, getting this nailed downed would be very nice |
| 11:08 | <zcorpan__> | gsnedders: is that the next generation in communication technology? |
| 11:08 | <zcorpan__> | foolip: firefox fires on <video>, not <source> |
| 11:08 | <gsnedders> | zcorpan__: Yeah. You need to get with the times and use it more often. ;P |
| 11:09 | <foolip> | oh, well that explains it |
| 11:10 | <zcorpan__> | (but it does bubble in firefox) |
| 11:11 | <zcorpan__> | i don't know when firefox decides to fire error, though |
| 11:11 | <foolip> | notwithstanding actual implementation, what makes sense? |
| 11:11 | <zcorpan__> | i think the spec is good as it's now |
| 11:12 | <foolip> | bubbling is slightly useful to capture any error on <video>, but not terribly so |
| 11:12 | <Hixie> | ok bed time |
| 11:12 | <Hixie> | nn |
| 11:12 | <zcorpan__> | just use a listener on the last source. or if you want to listen to all, use a captureing listener |
| 11:13 | <zcorpan__> | making error bubble screws up window.onerror |
| 11:13 | <foolip> | I suppose so |
| 11:14 | <zcorpan__> | /me heads over to the office |
| 11:14 | <hsivonen> | Hixie: so if I add pageshow/hide listeners from outside the iframe, they get added to WindowProxy and don't change with navigation and if I add them from within the iframe they are added to Window and change with navigation? |
| 11:19 | <hsivonen> | oh, joy. Gecko has a thingy that pushes an stream to the parser synchronously |
| 11:20 | <hsivonen> | and with the old parser, if the stream has no external scripts, this causes the whole thing to parse synchronously |
| 11:20 | hsivonen | is saddened by the implications to the HTML5 parser |
| 11:37 | <hsivonen> | even worse |
| 11:37 | <hsivonen> | Gecko has two thingies that push a stream to the parser synchronously |
| 11:37 | <hsivonen> | one that just *happens* to do it for buffer-backed streams like about:blank |
| 11:37 | <hsivonen> | and another for forcing it |
| 11:40 | hsivonen | wonders if HTTP streams become sync in WebKit when loaded from cache |
| 11:41 | <hsivonen> | apparently not |
| 11:41 | <hsivonen> | whew |
| 12:27 | <hsivonen> | i hate it when windowserver goes crazy because i left xcode with minefield stopped on a breakpoint in the background |
| 12:28 | <hsivonen> | can't just reboot due to unsaved data in other apps |
| 12:29 | <hsivonen> | i wonder how to get xcode killed when the ui is mostly inoperable |
| 12:31 | <Philip`> | ssh, kill -9? |
| 12:32 | <hsivonen> | does a debugged app cause some kind of windowserver-wide event craziness? |
| 12:32 | <hsivonen> | firewall |
| 12:32 | <Philip`> | If it was like Linux you could do ctrl-alt-f1 to switch to a usable terminal |
| 12:33 | Philip` | guesses OS X doesn't have anything similar, though |
| 12:33 | <hsivonen> | i managed to click the close button of the parallels vm list window and windowserver recovered |
| 12:34 | <hsivonen> | ctrl-alt-f1 FTW, indeed |
| 12:34 | <hsivonen> | too bad OS X doesn't have a working panic button for killing stuff |
| 12:36 | <Philip`> | What would you do if you wanted to kill the panic button? |
| 12:36 | <Philip`> | I guess you'd really need two panic buttons |
| 12:41 | <hsivonen> | without knowing the OS X internals, I'm guessing that holding a Cocoa app on a breakpoint for long periods of time causes event queue badness on the window server level |
| 12:41 | <hsivonen> | because of the weirdness of other Cocoa apps when the debugged app is killed |
| 12:58 | <Philip`> | Pausing programs on Windows is fun, because some applications send DDE messages which are broadcast synchronously to every window on the system, and so they hang if some other program's window is not responding |
| 12:58 | <Philip`> | (or something a bit like that) |
| 13:13 | <gsnedders> | So don't use Windows? :P |
| 13:14 | <Philip`> | That's not really an option when I'm writing programs for people to use, since people use Windows |
| 13:15 | <hsivonen> | whew. I finally got my about:blank email researched and sent. |
| 13:18 | <Philip`> | If empty pages are that complicated, I can't imagine what it's like for pages that have content |
| 13:18 | <hsivonen> | hah |
| 13:58 | <hsivonen> | Does XBL2 require the binding files to be loaded synchronously? |
| 14:10 | <Lachy> | hsivonen, from memory, I don't believe so. |
| 14:10 | <Lachy> | oh, wait. Maybe it does. |
| 14:10 | <hsivonen> | XBL1 seems to be implemented with a synchronous load, which seems undesirable |
| 14:11 | <Lachy> | if it wasn't syncronous, then that would make it difficult to bind events like load and pageshow to iframes using it, since you couldn't depend on the XBL being loaded before those events fired |
| 14:12 | <zcorpan> | maybe you could have an event fired when bindings are applied |
| 14:12 | <zcorpan> | or maybe xbl should delay the load event |
| 14:12 | <zcorpan> | or both |
| 14:35 | <Lachy> | the other problem with async techniques is that since load time is often faster in local testing environments, authors may depend on bindings being loaded before something else in the page, even though that may not turn out to be the case in real world conditions. |
| 14:35 | <Lachy> | this issue occurred with stylesheets, where authors have scripts depending on element styles being set by stylesheets earlier in the page |
| 14:37 | <Lachy> | althogh, syncronous processing sucks for incremental loading. Maybe the best solution would be to delay events until after bindings have been loaded |
| 14:51 | <Philip`> | Lachy: Like how the load event gets delayed by things, so everyone uses domcontentloaded or runs directly from a <script> at the end of a document to avoid the delay? |
| 14:54 | <Lachy> | yeah |
| 14:54 | <hsivonen> | http://lists.w3.org/Archives/Public/www-validator/2010Jan/att-0018/00-part |
| 15:03 | <MikeSmith> | yeah, it's a real shame that "w3c management" isn't interested in further updates to DTDs |
| 15:04 | <MikeSmith> | and w3c clearly missed a huge opportunity by not just deciding to put Shane in charge of all decisions |
| 15:04 | <MikeSmith> | so that's all water under the bridge now, and all that's left is to just try to figure out how to deal with the shambles that we're left with |
| 15:09 | <AryehGregor> | "Please Note: While we will copy over channel modes and topics for registered channels (there will be no changes to the services database, all nick and channel settings with services will stay the same) we are unable to do so for channels NOT registered with ChanServ. If your project utilises non-registered channels for whatever reason, please make note of the topics and modes so you can make a manual transfer of these yourselves." http://ann |
| 15:09 | <AryehGregor> | ounce.freenode.net/ |
| 15:19 | hsivonen | notes "detractor performant" |
| 15:19 | <hsivonen> | http://stuffandnonsense.co.uk/blog/about/keep_calm_and_carry_on_with_html5#r4092 |
| 15:20 | <hsivonen> | great comment by tantek |
| 15:21 | <MikeSmith> | AryehGregor: from yesterday, any way with MediaWiki API to get contents of a section of a page/ |
| 15:21 | <AryehGregor> | MikeSmith, you didn't get my response? |
| 15:21 | <MikeSmith> | ... |
| 15:22 | <MikeSmith> | I think my network connection maybe dropped |
| 15:22 | <MikeSmith> | I'm using a "pocket wifi" thing |
| 15:22 | AryehGregor | gets log links |
| 15:23 | <AryehGregor> | MikeSmith, http://krijnhoetmer.nl/irc-logs/whatwg/20100112#l-752 |
| 15:23 | gsnedders | wonders how Direct2D deals with accessibility |
| 15:24 | <hsivonen> | gsnedders: how do you mean? |
| 15:25 | hsivonen | assumes it deals similarly to how Cairo or Quartz 2D deals |
| 15:25 | <gsnedders> | hsivonen: It's another 2D non-retained mode graphics API, like the canvas 2D context. |
| 15:26 | <MikeSmith> | AryehGregor: thanks |
| 15:27 | <Philip`> | http://blogs.technet.com/thomasolsen/archive/2008/10/29/introducing-the-microsoft-direct2d-api.aspx - ""Also, are there any accessibility issues at this point?" Can you be more specific? We are a 2D rendering API. Other than text, we have no policy regarding layout." |
| 15:29 | <Philip`> | Fortunately people using the D2D API can make use of Windows' normal programmable accessibility API |
| 15:29 | <Philip`> | whereas HTML/JS does not have a programmable accessibility API |
| 15:29 | <hsivonen> | Philip`: the programmable accessibility API for HTML/JS is called ARIA |
| 15:33 | <Philip`> | Hmm, I suppose that's true |
| 15:34 | <Philip`> | I guess there isn't a huge difference between that and what OSes provide, since in both cases I guess you need to construct some kind of annotated tree representing the UI |
| 16:00 | <othermaciej> | hsivonen: the term "detractor performant" sounds so medical |
| 16:02 | <TabAtkins> | I'm really not sure why he tried to avoid the term troll. It's infinitely better than "detractor performant". |
| 16:03 | <TabAtkins> | Especially since it's hard for me to read "performant" as anything other than an adjective. |
| 16:06 | <karlcow> | maybe because troll is also often used for getting rids of comments, which are just not understandable from one community to the others. (Different views of the world). |
| 16:09 | <Philip`> | Maybe because (as I understand it) "troll" implies intent, whereas some people are unintentionally productivity sinks |
| 16:10 | <TabAtkins> | I like that explanation. |
| 16:10 | <Philip`> | But I don't understand what the phrase "detractor performant" means, or whether it's even meant to be English, so I don't think that's a good alternative |
| 16:10 | <TabAtkins> | "detractor performant" is the worst phrase to use in replacement ever, though. |
| 16:10 | <karlcow> | Philip`: yes |
| 16:10 | <AryehGregor> | I think this is the first time I've seen anyone from Adobe mention plugins on public-html. |
| 16:13 | <Philip`> | AryehGregor: Hmm, I was going to give a counter-example but Gmail says that's the only message matching "from:adobe plugin" :-( |
| 16:13 | <AryehGregor> | Larry Masinter has posted a lot, but I've never seen him mention anything even tangentially relating to Flash or PDF. |
| 16:15 | <AryehGregor> | Now we get three in a row? |
| 16:15 | <AryehGregor> | "Why couldn't the iFrame be pointing to an SVG image, for example, or a PDF?" |
| 16:15 | AryehGregor | didn't even realize you could put non-HTML in <iframe>s. |
| 16:16 | <TabAtkins> | Yeah, putting pdfs in iframes is nice if the user has a browser-embedded viewer. |
| 16:16 | <TabAtkins> | But the question is completely nonsense as stated, since it's talking about @doc, which isn't a url. |
| 16:16 | <AryehGregor> | Right, I caught that part. |
| 16:16 | <tantek> | Philip`, TabAtkins - "detractor performant" means the person is a detractor (disagrees with a technology, which in itself isn't necessarily a bad thing, if they're constructive), and is a performant (is "acting out", trying to gain attention, etc.) - both of which specific types of trolls do |
| 16:16 | daedb | hates in-browsers PDFs |
| 16:17 | <TabAtkins> | I have never heard the word performant being used in that sense ever. I gathered it's meaning from structure and context, it's just definitely, um, strange. |
| 16:17 | <tantek> | TabAtkins, novel rather than strange |
| 16:18 | <workmad3> | esoteric even |
| 16:18 | <tantek> | yes I paired the two specifically for that blog comment - I'd never seen them used like that either |
| 16:18 | <othermaciej> | I have heard "performant" used as a noun meaning "one who performs" |
| 16:18 | <tantek> | othermaciej, right |
| 16:18 | <othermaciej> | though it sounds more medical / psychological than "performer" |
| 16:18 | <tantek> | a lot of trolling is done for what they might call an "art" of "performance" |
| 16:18 | <TabAtkins> | I've only ever heard it as an adjective, describing something which has good performance, in the technology sense. |
| 16:19 | <tantek> | TabAtkins - that's another meaning, yes. |
| 16:19 | <workmad3> | stupid, ambiguous, context sensitive english :P |
| 16:19 | <Philip`> | My dictionary says "A person who performs a duty, ceremony, etc., a performer." |
| 16:19 | <Philip`> | which is not the same as an acting kind of performer |
| 16:20 | <tantek> | the point is that such individuals, whether intentionally or not, are more destructive than constructive toward the development of an open community based technology, either intentionally so, or unintentionally, but likely with disregard for such, and often paired with craving more attention/recognition/ego-boosting. |
| 16:20 | <AryehGregor> | Where was this strange and/or novel phrase used? |
| 16:20 | <tantek> | if you can recognize the pattern, and recognize that it is harmful to the development of the technology (and the community), then you can take action to prevent further damange |
| 16:21 | <TabAtkins> | I'd still have just used "troll". Unintentional trolls are no less trolls. |
| 16:21 | <tantek> | AryehGregor - in my comment on Andy Clarke's blog post. |
| 16:22 | <TabAtkins> | AryehGregor: Check your scrollback, just before MikeSmith got your attention. |
| 16:22 | karlcow | wonders if it is more a question of homogeneity of a group (shared values and goals). |
| 16:23 | <Philip`> | No, it's a question of some people wasting huge amounts of time |
| 16:23 | <Philip`> | It's fine when people disagree efficiently |
| 16:23 | <AryehGregor> | I think the problem is that the whole Change Proposal thing is too lengthy and complicated. |
| 16:24 | <AryehGregor> | Mainly because it tries to appear equitable. |
| 16:24 | <karlcow> | Philip`: I do not believe into that. it is an attempt at Universalism |
| 16:24 | <AryehGregor> | Getting things done is more important than being equitable here. |
| 16:24 | <tantek> | AryehGregor - AFAICT, the whole Change Proposal thing is designed to slow down and discourage spec feedback filibustering |
| 16:24 | <AryehGregor> | If we had, say, a quick vote of a small group of actual members in the event of dispute, like I guess happens in most WGs, there would be no problem. |
| 16:25 | <AryehGregor> | tantek, it makes it harder to raise formal complaints, but also harder to address them. |
| 16:25 | <tantek> | it's a mechanism designed to cope with the aforementioned "trolling" |
| 16:25 | <othermaciej> | it prevents you from spamming the group with blockers |
| 16:25 | <tantek> | AryehGregor - agreed |
| 16:25 | <AryehGregor> | The problem is when they're being raised by people who seem to have unlimited time to pursue their agenda, and addressed by people who have actual work to do. |
| 16:25 | <othermaciej> | to get something to be a blocker you have to do a bunch of work |
| 16:25 | <othermaciej> | then the group has to do work in response |
| 16:25 | <TabAtkins> | Hrm. These google results seem likely virus-filled. Anyone got a suggestion for a reliable swf-to-flv converter? |
| 16:25 | <othermaciej> | the way it is set up has enough credibility that decisions can stand and remove blockers |
| 16:26 | <AryehGregor> | But the ones doing work in response are mostly people like browser developers who could more productively be, e.g., implementing the spec and providing sound feedback. |
| 16:26 | <tantek> | IMHO much better to admit that negative contributors exist, and kick them out per a community guidelines policy, than introduce more bureaucracy to deal with them slowly and painful with everyone else involved. |
| 16:26 | <Philip`> | TabAtkins: Is that even possible? |
| 16:26 | <othermaciej> | it's true that voting is another potential decision mechanism that could have perceived legitimacy |
| 16:26 | <TabAtkins> | Philip`: There are plenty of programs claiming that it is on the web. |
| 16:26 | <tantek> | AryehGregor: "The problem is when they're being raised by people who seem to have unlimited time to pursue their agenda, and addressed by people who have actual work to do." <---- exactly |
| 16:26 | <tantek> | hence, filibustering |
| 16:26 | <Philip`> | TabAtkins: (I thought SWF was a lossy optimised format, so you couldn't reverse it decently in general) |
| 16:26 | <othermaciej> | however, voting in a group of 500 people, of whom maybe 20-30 are active posters and who knows how many others are active readers, is hard to define |
| 16:27 | <TabAtkins> | Some loss is acceptable for my purposes. |
| 16:27 | <AryehGregor> | Perceived legitimacy shouldn't be the primary goal, though. Most WGs don't even let members of the public view the real mailing list, why do we have to bend over backwards to address all their complaints? (Where by "we" I don't include myself, oddly, since I'm just part of the general public.) |
| 16:28 | <Philip`> | AryehGregor: Most WGs produce specs that are often considered poor quality by people here |
| 16:28 | <AryehGregor> | Well, yes. |
| 16:28 | <Philip`> | so that's not clearly a good model to follow |
| 16:29 | <AryehGregor> | What I really mean, obviously, is just to be like the WHATWG and have everything decided by a small fixed group of people that's open to new members by invitation only. |
| 16:29 | <tantek> | AryehGregor - in order to address the problem of "people who seem to have unlimited time to pursue their agenda" - we've had to create more and more guidelines to forbid specific disruptive and community-damaging behaviors in the microformats forums: http://microformats.org/wiki/mailing-lists#General_guidelines |
| 16:29 | <AryehGregor> | A small fixed group of the *right* people. |
| 16:29 | <AryehGregor> | Like, say, the WHATWG members. |
| 16:29 | <tantek> | AryehGregor - better to encourage a growing group of *right-behaved* people |
| 16:29 | <tantek> | it will change over time |
| 16:29 | <tantek> | and that's ok |
| 16:29 | <Philip`> | The WHATWG Member thing seems to be basically fiction |
| 16:29 | <tantek> | new folks should be encouraged to participate, but only constructively so |
| 16:30 | <Philip`> | since the only thing they can do is kick Hixie, and they won't do that because the WHATWG would collapse |
| 16:30 | <Philip`> | *kick out |
| 16:30 | <othermaciej> | the WHATWG Members / Steering Committee in practice have not done much of anything |
| 16:30 | <gsnedders> | Philip`: They could in theory tell Hixie do x or we drop you. |
| 16:30 | <AryehGregor> | Well, yes, but that's because they agree with Hixie, generally. |
| 16:30 | <othermaciej> | I could imagine making that an escalation path but right now it isn't really in practice |
| 16:30 | <Philip`> | gsnedders: And he could ignore them |
| 16:31 | <AryehGregor> | Philip`, the charter says he can't publish anything as a Last Call or Rec if they don't all agree with or something, I think. |
| 16:31 | <othermaciej> | managing a W3C Working Group has some constraints that don't necessarily apply to an independent community |
| 16:31 | <othermaciej> | you have to do things that will be seen as legitimate by W3C Process |
| 16:31 | <AryehGregor> | But it's moot, since they don't disagree with him strongly on anything AFAICT, so it all works great. |
| 16:31 | <Philip`> | Then he could create the WHATWG2 and move his server to run on the whatwg2.org domain |
| 16:31 | <othermaciej> | that by itself removes much trolling and avoids creation of additional trolling |
| 16:32 | <AryehGregor> | Philip`, and then who would implement his spec? Chrome would implement WHATWG2, Firefox/Safari/Opera would implement WHATWG1, IE would go with the W3C? |
| 16:32 | <AryehGregor> | As they say, the great thing about standards is how many there are to choose from. |
| 16:32 | <othermaciej> | most of the WHATWG Steering Committee is browser vendors, and Hixie is motivated to meet the needs of browser vendors to the point that disputes would stop well short of kicking him |
| 16:32 | <AryehGregor> | Right, exactly. |
| 16:32 | <AryehGregor> | Everyone is working with the same goals, so there's very little conflict. |
| 16:32 | <AryehGregor> | The problem in the W3C is people have conflicting goals. |
| 16:33 | <AryehGregor> | If actual decisions are only made by people who share the same goals, you eliminate a lot of conflict. |
| 16:33 | <tantek> | AryehGregor - more specifically when people in your community have goals of *disruption* and/or *being noticed / gaining notoriety*, your community is kinda screwed |
| 16:33 | <AryehGregor> | If you have to accommodate widely different sets of goals, you get lengthy arguments that end in bad compromises. |
| 16:34 | <tantek> | so you have to make a choice |
| 16:34 | <Philip`> | So it seems to me that it's best to pretend the steering committee doesn't exist, when considering the organisation of the WHATWG, because it has no effect and there are other forces that prevent the hypothetical situations where it could have an effect |
| 16:34 | <Philip`> | and so "like the WHATWG and have everything decided by a small fixed group of people" seems like an inaccurate characterisation |
| 16:34 | <AryehGregor> | tantek, I don't honestly think that applies in the HTMLWG. It's just that a lot of people there don't share the goal of "Write something authors will want now, that implementers can ship now, and to heck with purity or compatibility with barely-deployed specs like RDFa." |
| 16:35 | <AryehGregor> | Philip`, okay, sure. Hixie listens to the browser vendors for other reasons anyway. So just say it's a consortium of browser implementers where they work out what they want to implement, and Hixie helps by coordinating. |
| 16:35 | <tantek> | AryehGregor - it's not difficult to find numerous messages on the w3c list from specific contributors that back up the disruption/attention-grabbing point I made. |
| 16:36 | <tantek> | Who knows about Gresham's law? |
| 16:36 | <AryehGregor> | tantek, I find that I assume good faith to a much greater degree than most people. I doubt I would agree with your judgment. |
| 16:36 | <tantek> | AryehGregor - when you see outright insults/namecalling - it's hard to defend that in any interpretation of good faith. |
| 16:36 | <Philip`> | AryehGregor: It seems we're very fortunate that the browser implementors almost always agree on everything |
| 16:36 | <tantek> | AryehGregor - it's also hard to defend people spamming the list with numerous emails in a row |
| 16:36 | <tantek> | (= performers) |
| 16:37 | <Philip`> | AryehGregor: (so I can't imagine that model working when implementors have more diverging wishes) |
| 16:37 | tantek | recommends reading http://adactio.tumblr.com/post/27789278/theres-a-sort-of-greshams-law-of-trolls-trolls |
| 16:37 | <AryehGregor> | tantek, insults happen when people acting in good faith get frustrated. Look at some of the things that are said in this chat. Spamming multiple e-mails in a row is just poor netiquette, not indicative of bad faith. |
| 16:37 | <AryehGregor> | Philip`, if they diverge, then you either have no standard (e.g., video format) or you hash out a compromise through sweat, blood, and tears (e.g., font format). |
| 16:38 | <AryehGregor> | But in either case you're still just reflecting vendor consensus, wherever it exists. |
| 16:38 | <cardona507> | http://video.google.com/videoplay?docid=-4216011961522818645# - How Open Source Projects Survive Poisonous People (And You Can Too) |
| 16:38 | <tantek> | AryehGregor - when the behaviors cause those acting in good faith who can remain civil to leave, you have to make a decision - let the insulters stay, or encourage those who contribute constructively to stay. |
| 16:38 | <AryehGregor> | You can't do better than that unless you like speccing idyllic fantasy worlds, like the XHTMLWG did. |
| 16:38 | <AryehGregor> | tantek, oh, sure. But I wouldn't say anyone has to be acting in bad faith for that decision to come up. |
| 16:39 | <tantek> | Because in all experience, they typically don't - the problem of people who have real work to do, rather than nearly infinite time to filibuster |
| 16:39 | <tantek> | cardona507's reference is another important one |
| 16:40 | <tantek> | sometimes an individual's behavior is so poisonous to the community that it doesn't matter what positive contributions them make - the health of the community (and continuing to grow it) is much more important than a single individual's performances (whether good or bad) |
| 16:41 | <Philip`> | You can justify anything if it's for the greater good |
| 16:42 | <tantek> | Philip` - a hypothetical statement which is not often true in practice |
| 16:43 | <tantek> | when you are watching knowledgable professionals leave a community because a novice hobbyist/pundit/advocate is being abusive - you have to make a choice about what kind of community you want. |
| 16:43 | <TabAtkins> | Or rather, a relatively useless statement - even if true, it says nothing about whether a particular statement can be justified usefully with other methods as well. |
| 16:43 | <tantek> | and that's not a theoretical situation - it's a real situation we've had to confront several times in the microformats lists/wiki etc. |
| 16:44 | <tantek> | one which, frankly, we (admins) erred too much in the direction of assumption of good faith, and lost many good people while trying to take time to be "fair" to a disruptor |
| 16:44 | <tantek> | we know better now (per the poisonous people video posted above) |
| 16:44 | Philip` | was mostly thinking of Hot Fuzz in relation to that statement |
| 16:44 | gsnedders | has never seen that |
| 16:45 | gsnedders | meant to, but never did. |
| 16:45 | TabAtkins | too. |
| 16:47 | <Philip`> | It's never too late |
| 16:48 | <tantek> | Philip` - the analogy of an open community (with archived discussions etc.) to such films/media is flawed because they all assume/use closed/secret methods. |
| 16:49 | <tantek> | Wikipedia is a better example - open guidelines, discussions, and bannings. |
| 16:50 | <AryehGregor> | Is Wikipedia meant to be an example of a successful or unsuccessful community? :) |
| 16:51 | <Philip`> | I'm not sure why the methods of the film are relevant, I was just thinking of the murders and the people chanting around a table |
| 16:52 | <Philip`> | Poor living statue :-( |
| 16:53 | <tantek> | AryehGregor - here is the URL to my aforementioned comment on Andy Clarke's post (took me a min to dig it up) where I used the phrase "detractor performants": http://stuffandnonsense.co.uk/blog/about/keep_calm_and_carry_on_with_html5/#r4092 |
| 16:54 | <AryehGregor> | tantek, I already saw it in backscroll, thanks. |
| 16:54 | tantek | also has bills to pay. |
| 16:54 | AryehGregor | doesn't, yay |
| 16:54 | <tantek> | (per Foliot's comment) |
| 16:54 | <AryehGregor> | http://www.google.com/search?q="detractor+performants"; |
| 16:54 | <AryehGregor> | :) |
| 16:54 | <TabAtkins> | Ah man, a google whack. |
| 16:55 | <TabAtkins> | Temporarily, at least, until the logs get spidered again. |
| 16:55 | <shepazu> | I wish I knew better how to deal with people exhibiting poisonous behavior |
| 16:55 | <tantek> | shepazu - you and me both |
| 16:55 | <jcranmer> | a canister of chloroform? |
| 16:55 | <TabAtkins> | Violence doesn't work so well over the internet, sadly. |
| 16:56 | <shepazu> | there was a person on www-dom and public-webapps that I tried to reason with, who was being a royal pain, insulting and belittling people |
| 16:56 | <shepazu> | but trying to talk with him offline resulted in him being more abusive |
| 16:56 | <tantek> | shepazu - here's what microformats.org has come up with so far: http://microformats.org/wiki/mailing-lists - if you have additional suggestions, please /join #microformats and feel free to discuss |
| 16:57 | <tantek> | my experience with trying to talk offline with disruptive individuals got the same response (profanity based personal insults) |
| 16:57 | <othermaciej> | shepazu: IMO you should have banned him sooner |
| 16:57 | <othermaciej> | shepazu: at least a temporary suspension |
| 16:57 | <othermaciej> | assuming this is the person I think you mean |
| 16:57 | <shepazu> | ultimately, much to my disappointment, we had to ban him for 2 weeks... surprisingly, that seems to have worked, since he's actually behaved fairly well since then |
| 16:58 | <shepazu> | othermaciej: yes, I agree, we should have... but since I was one of the ones he was insulting, I tried to be objective |
| 16:59 | tantek | passes the community management / troll-discussion baton to hober |
| 16:59 | <shepazu> | perhaps we should have put it to a WG vote, but we were also concerned where that would lead |
| 16:59 | <othermaciej> | I do strongly prefer to deal with people through coaching and discussion |
| 17:00 | <AryehGregor> | Openly insulting people crosses a line. The worst we've seen in public-html is thinly-veiled insinuations, for the most part. |
| 17:00 | <othermaciej> | but in the WebKit community there is one person that we ultimately had to completely ban from our mailing lists and bugzilla |
| 17:00 | <shepazu> | othermaciej: and regardless of his behavior, he was raising genuinely good points fairly often... it's just that ultimately the signal/noise ratio was off |
| 17:00 | <Philip`> | Voting people out of the group? Sounds like a new kind of reality TV show |
| 17:00 | <Philip`> | although admittedly a really really boring one |
| 17:00 | <AryehGregor> | I'm an admin of a gaming forum, so I know all about having to ban people. :P |
| 17:00 | <othermaciej> | I think voting people off the island would not be a healthy dynamic |
| 17:01 | <shepazu> | Philip`: yeah, we really didn't want to go there, and ultimately didn't |
| 17:01 | <shepazu> | othermaciej: exactly |
| 17:01 | <othermaciej> | better to have the chairs and team contacts determine when transgressions are enough to merit a suspension |
| 17:01 | <othermaciej> | the hardest thing to deal with though is continuous low-level disruption over a long period where no one incident is clearly over the line |
| 17:02 | <othermaciej> | that's kind of what happened with the person we had to ban from WebKit development |
| 17:02 | <AryehGregor> | That sort of thing has happened on Wikipedia too. |
| 17:03 | <othermaciej> | TabAtkins: so are you gonna post your proposal for 6 to public-html today? |
| 17:03 | <othermaciej> | TabAtkins: if you do I'll follow up with 3 |
| 17:03 | <TabAtkins> | othermaciej: Sure. |
| 17:03 | <shepazu> | othermaciej: to be frank, we were a little concerned about the criticism that might bring upon W3C, which is one thing that delayed our decision... if we felt like the majority of people would see and respect our reasons, and wouldn't try to score political points off W3C "banning" someone (even temporarily, even when well deserved), we woudl probably have acted sooner |
| 17:03 | <shepazu> | I do appreciate the system of checks and balances on authority, though |
| 17:04 | <TabAtkins> | othermaciej: I'll do it over lunch, in an hour or two. |
| 17:04 | <othermaciej> | shepazu: as someone who was not faced with the decision himself, I think you guys may have been too cautious, but I can also understand that these are hard decisions |
| 17:05 | <shepazu> | in the case of the HTML group, at this point, I would favor warning the individual that their behavior is seen as disruptive by a wide variety of people, and ask them to step back and closely examine their actions before they make them |
| 17:05 | <shepazu> | or face a temporary ban |
| 17:05 | <JonathanNeal> | Love! |
| 17:06 | <gsnedders> | shepazu: But the question is would the W3C follow through with that threat? |
| 17:06 | <shepazu> | but I'm afraid that would evoke a strongly negative reaction |
| 17:07 | Philip` | doesn't remember seeing much unreasonably disruptive behaviour recently, compared to the earlier years in public-html |
| 17:07 | <shepazu> | gsnedders: absolutely, just as we reluctantly did in WebApps WG...but not "W3C the Team", "W3C the WG Chairs and Team Contact" |
| 17:07 | <JonathanNeal> | w3c follow through with WHAT threat? |
| 17:08 | <Philip`> | JonathanNeal: The W3C is threatening to bomb China |
| 17:08 | <JonathanNeal> | Philip`, damnit W3C, have you learned nothing! |
| 17:08 | <shepazu> | Philip`: and we don't see as much violent crime now as we did in the 1800s, but we punish it when we do, even more so... community standards for behavior change |
| 17:09 | <shepazu> | JonathanNeal: you have to admit, they've been asking for it! they are threatening one our our member states^H companies, Google |
| 17:10 | <Philip`> | shepazu: Well, we don't hang people for it any more |
| 17:10 | <Philip`> | (at least in my country) |
| 17:10 | <shepazu> | I must admit the threat is a bit weak, since we only have M80s and roman candles, but they could get a nasty burn |
| 17:11 | <shepazu> | Philip`: yes, community standards for punishment have changed accordingly... we tolerate less, but punish less |
| 17:13 | <shepazu> | and people tend to adapt to that as normative behavior... that's the danger of trolls, they change the perceived normative behavior of the community... and to be frank, I think the editor is also doing that currently, with (let's say) "whimsical" decisions |
| 17:14 | <JonathanNeal> | question, will aria-role="navigation" and <nav> be treated the same way? |
| 17:15 | <othermaciej> | JonathanNeal: for purposs of assistive technologies, they should be |
| 17:16 | <othermaciej> | shepazu: I think some of his decisions were a little... edgy, but we seemed to have restored normalcy, and now everyone in the group has a newfound appreciation for making sure major changes get due discussion |
| 17:16 | <shepazu> | JonathanNeal: interesting question... are there any differences in described behavior? they should probably be brought into line, if so, to have a unified model |
| 17:16 | <shepazu> | othermaciej: yes, I think you handled that well |
| 17:17 | <othermaciej> | ARIA attributes are technically not supposed to have any effect other than for assistive technology |
| 17:17 | <othermaciej> | I could imagine you may have a browser feature that is supposed to skip over or highlight the navigation section which is not intended as an assistive technology |
| 17:21 | <Dashiva> | If the authorities don't stop certain types of behavior, it can easily seem like they are condoning it, which may lead other people who wouldn't normally act that way to do start |
| 17:26 | Dashiva | reads a comment that claims using pre-PR CSS3 is 'tag soup' |
| 17:26 | <JonathanNeal> | hmmm |
| 17:26 | <JonathanNeal> | Dashiva, grab your sword, we travel at once! |
| 17:27 | <Dashiva> | My sword is a plowshare nowadays |
| 17:27 | <Dashiva> | I poke people with it sometimes, but it's a horrible weapon |
| 17:36 | <AryehGregor> | A plowshare can do a lot of damage if you get in a good swing. |
| 17:39 | <JonathanNeal> | No doubt. |
| 17:41 | <Dashiva> | Yeah, but the sword will be in your torso before you finish the swing |
| 17:41 | <TabAtkins> | Sneak attack. |
| 17:42 | <Dashiva> | "Better still, why can’t this be done as an open incubator group at the W3C?" |
| 17:47 | <JonathanNeal> | So, the W3C is mad at you guys for messing with the spec? |
| 17:47 | <JonathanNeal> | And threatening Google in someway because of it? |
| 17:49 | <danbri> | W3C is an organization, how can it be mad? |
| 17:52 | <JonathanNeal> | The NAZIs where an organization too! What say you now, Godwin? |
| 17:53 | JonathanNeal | appreciates how he mispelled and everything ... he would make a wonderful troll. |
| 17:55 | <JonathanNeal> | TabAtkins, I didn't see you, hiiiiiia! |
| 17:56 | <TabAtkins> | Yo. |
| 17:56 | <TabAtkins> | Or, since we're trolling, |
| 17:56 | <TabAtkins> | OH HAI |
| 17:56 | <TabAtkins> | I C U THEIR |
| 17:56 | <JonathanNeal> | Trolling aside, your help was invaluable when I was building our new company site. It supposedly launches this Friday. |
| 17:56 | <TabAtkins> | Woo! |
| 17:56 | <JonathanNeal> | Fully HTML5, totally outliner friendly. |
| 17:57 | <JonathanNeal> | I would do it differently if I was doing it again, but I think it totally works. I'll be sure to show you once it's up. Okay back to troll. |
| 17:57 | <JonathanNeal> | WAAAAAA FIX MY IE6 BUG U MAD N UR PRODUCT! |
| 17:57 | <gsnedders> | teh html5 outlinor ftw! |
| 17:58 | <TabAtkins> | OMG U GUIZE |
| 17:59 | <JonathanNeal> | WATS DAT TAATKINS? |
| 17:59 | <TabAtkins> | I LOLD |
| 18:00 | <JonathanNeal> | OMFROTFG ME2 |
| 18:00 | <TabAtkins> | ^_^ |
| 18:03 | <gsnedders> | othermaciej: You wouldn't happen to be anywhere near olliej? |
| 18:03 | <othermaciej> | gsnedders: no, but I will be in an hour or two |
| 18:03 | <gsnedders> | Ah, no use |
| 19:40 | <AryehGregor> | I find it both ironic and annoying that Analytics doesn't fully work for me in Chrome. Because, of course, it uses Flash. |
| 20:38 | <TabAtkins> | Which issue is the dt/dd thing associated with? |
| 20:41 | <othermaciej> | TabAtkins: http://dev.w3.org/html5/status/issue-status.html#ISSUE-083 |
| 20:42 | <TabAtkins> | danke, othermaciej. I always forget where that list is. |
| 20:44 | <roc> | sometimes I wonder whether working on Web browsers is a risk to my mental health |
| 20:46 | <othermaciej> | roc: how could it not be? |
| 20:46 | <Dashiva> | Anything is a risk |
| 20:50 | <TabAtkins> | othermaciej: Sent proposal 6 to the list. If you'll send 3, I'll then send 7 (since it's just a diff, it shoudl come afterwards). |
| 20:51 | <othermaciej> | TabAtkins: roger that |
| 20:52 | <AryehGregor> | Hmm, is there no rel="edit" defined in the spec? |
| 20:53 | <AryehGregor> | It's used on Wikipedia and elsewhere, it would be nice if it were standard. Relevant: http://universaleditbutton.org/Registered_MIME_type |
| 20:54 | <AryehGregor> | I could swear this was in the spec at some point. |
| 20:55 | AryehGregor | realizes this isn't the first time he's wanted to run decent history examination tools on the spec . . . time to break out git-svn. |
| 20:58 | <AryehGregor> | Hurrah for tiny repos, this looks like it won't take more than an hour. |
| 21:05 | Philip` | has been using hgsvn (with mixed success in its bridging), and whenever he moves from the Hg side to the SVN side he's surprised by how incredibly slow everything is |
| 21:05 | <AryehGregor> | Does hgsvn also have really slow initial checkout? |
| 21:05 | <Philip`> | Yes |
| 21:06 | <AryehGregor> | git-svn takes a few days for large repositories like MediaWiki's. |
| 21:06 | <AryehGregor> | I guess it really is SVN's fault, then. |
| 21:06 | <Philip`> | As far as I can tell, it basically checks out one revision, commits it to Hg, does an svn up to the next revision, commits, repeat |
| 21:07 | <AryehGregor> | Dunno how git-svn works, all I know is it's horrifyingly slow. |
| 21:07 | <AryehGregor> | But that sounds about right. |
| 21:07 | <Philip`> | I suppose it doesn't help that I've had to disable various features on the SVN server (like compression, and bulk updates) to stop it from using up all my RAM and bringing the machine down |
| 21:08 | <Philip`> | ...when somebody accesses it with a certain version of TortoiseSVN, at least |
| 21:17 | <AryehGregor> | Load is one nice thing about DVCSes, you only have to worry about bandwidth. :) |
| 21:17 | <AryehGregor> | (which is still probably much less than non-DVCSes) |
| 21:21 | <Philip`> | Bandwidth was a problem long before load for me, using SVN |
| 21:21 | <Philip`> | since I had something like 100GB/day and <10% CPU load |
| 21:21 | <Philip`> | though that was almost entirely checkouts, not logs or annotates or anything |
| 21:22 | <Philip`> | (or at least it was <10% once I'd figured out how to configure Apache so it wasn't constantly killing and respawning processes) |
| 21:23 | <AryehGregor> | Why not use svn: instead of http:? |
| 21:24 | <Philip`> | I needed per-directory access control |
| 21:24 | <Philip`> | (which is also a problem with DVCSs :-( ) |
| 21:30 | <othermaciej> | TabAtkins: done |
| 21:30 | <TabAtkins> | kk |
| 21:53 | <TabAtkins> | othermaciej: Done, obviously. |
| 21:54 | <othermaciej> | TabAtkins: w00t |
| 21:54 | <TabAtkins> | So I think that's all the change proposals for 83 that people are willing to write something on. |
| 21:54 | <othermaciej> | seems like no one believes in any of the other options enough to put them forward |
| 21:54 | <TabAtkins> | Yeah. |
| 21:54 | <othermaciej> | even though they could cut & paste for most of what they need... |
| 21:55 | <TabAtkins> | And that's fine with me. I see no reason to put forward anything that *someone* doesn't like enough to champion. |
| 21:55 | <Philip`> | That's unfair discrimination against the preferences of lazy people |
| 21:56 | <TabAtkins> | Yeah, but what are they gonna do about it? |
| 21:57 | <TabAtkins> | Sitting on the couch and eating cheetos doesn't exactly scare me. |
| 22:12 | <Lachy> | is issue 83 the dt/dd issue? |
| 22:13 | <TabAtkins> | Yeah. |
| 22:13 | <Lachy> | I haven't had time to do the other proposals. But I need to think about which one I like, if any, better than figcaption |
| 22:16 | <othermaciej> | Lachy: you've got 2 days still |
| 22:18 | <Lachy> | good |
| 22:18 | <Lachy> | I will do it before then, at least for <c> |
| 22:18 | <othermaciej> | Lachy: do you want me to list you as a co-author for mine? ( a lot of the text in it is the rationale you wrote) |
| 22:18 | <Lachy> | that won't take long. I can just reuse most of what you wrote and add a little extra |
| 22:18 | <Lachy> | sure |
| 22:19 | <Lachy> | how many change proposals do we have done so far? |
| 22:19 | <TabAtkins> | 3 |
| 22:19 | <TabAtkins> | we've done 3, 6, and 7 |
| 22:19 | <Lachy> | plus that other fltcap one |
| 22:20 | <TabAtkins> | Right, so four. |
| 22:20 | <TabAtkins> | I still read that as "flit-cap" every time. |
| 22:21 | <Lachy> | do we have 3b as a separate proposal too? |
| 22:21 | <TabAtkins> | Nah, it's just an option on 3 |
| 22:22 | <Lachy> | I'm surprised someone bothered to do 7. We really don't need the fbody/dbody elements |
| 22:22 | <Lachy> | was that you? |
| 22:22 | <TabAtkins> | Yes. |
| 22:22 | <Lachy> | ok |
| 22:23 | <TabAtkins> | We certainly don't *need* them, but if it's going to be common to style the contents anyway (and it will be for me, at least), having a blessed element is better than using a classed div every time. |
| 22:25 | <Hixie> | why class it? |
| 22:25 | <Hixie> | just use <div>, no need for classes |
| 22:25 | <TabAtkins> | To distinguish it from times when I have a <details> that just happens to have <div> children. |
| 22:26 | <Hixie> | just be consistent |
| 22:26 | <TabAtkins> | div-as-details-content-wrapper would be styled differently than div-as-wrapper-of-some-content-in-details |
| 22:28 | <TabAtkins> | I already know that I can't just "be consistent" with <section>s - I use <div class=sbody> when I need to style the body of the section, because I sometimes use <div>s in the content too. |
| 22:28 | <Hixie> | see, css has huge holes |
| 22:29 | <Hixie> | every time you use <div>, god kills a kitten |
| 22:29 | <Hixie> | the csswg is responsible for the death of millions of kittens! |
| 22:29 | <TabAtkins> | Good. I can't get my dwarves to kill them fast enough. |
| 22:29 | <TabAtkins> | And then they just tantrum anyway. |
| 22:29 | <othermaciej> | it does seem like an artificial pseudo-element for "everything inside this element except some particular part" would be useful for a lot of things |
| 22:30 | <TabAtkins> | Yeah, I'll be writing up a proposal to www-style soon, once I work out how to deal with overlap problems. |
| 22:30 | <TabAtkins> | If you guys have any suggestions, by the way, they'd be appreciated. Quick overview: |
| 22:33 | <TabAtkins> | ::wrap(selector1[, selector2]?) would create a wrapping pseudo with its pseudo-start-tag just before each element matched by selector1, and its pseudo-end-tag just before the next sibling of the selector1-element that matches selector2. If there is no such sibling, or if selector2 isn't specified, the pseudo-end-tag is inserted just before the end of the selector1-element's parent. |
| 22:33 | <TabAtkins> | So, problem: given the structure <b>a</b> <i>b</i> <b>c</b> <i>d</i> and the css ::wrap(b,b) {} ::wrap(i,i) {}, what is the pseudo-structure? |
| 22:34 | <TabAtkins> | The ::wrap(b,b) wants to produce a structure like <::wrap><b>a</b> <i>b</i></::wrap><::wrap><b>c</b> <i>d</d></::wrap> |
| 22:35 | <TabAtkins> | But ::wrap(i,i) wants to produce a structure like <b>a</b> <::wrap><i>b</i> <b>c</b></::wrap> <::wrap><i>d</i></::wrap> |
| 22:35 | <TabAtkins> | Naively applying both causes misnesting. |
| 22:38 | <TabAtkins> | In context, I'd be able to style figure bodies with ::wrap(figure > :first-child, figcaption),::wrap(figcaption + *) {}, and details bodies with ::wrap(dsummary + *) {} |
| 22:39 | <Lachy> | is it a problem if pseudo-elements overlap? |
| 22:40 | <TabAtkins> | Yes. Very much so. |
| 22:40 | <TabAtkins> | Imagine you specified borders on both. |
| 22:42 | <TabAtkins> | Basically you want to guarantee that the CSS box structure is a tree for many of the same reasons you want to guarantee that the DOM is a tree. |
| 22:46 | <Lachy> | how does that work for ::first-line, and elements that span multiple lines? |
| 22:47 | <TabAtkins> | ::first-line is extremely limited in what properties one can apply, precisely so that it can be done halfway-sanely. It breaks itself up to avoid misnesting. |
| 22:47 | <Lachy> | hmm, why does border not apply to ::first-line? |
| 22:47 | <Lachy> | oh, is that why? |
| 22:47 | <TabAtkins> | Because of the conflicts. <p>foo foo foo <span>bar [linebreak] bar bar</span></p> p::first-line{ border: 1px solid black; } p span { border: 1px dotted red; } |
| 22:48 | <TabAtkins> | What happens there? |
| 22:48 | <Lachy> | that's basically the same test I just wrote |
| 22:48 | <TabAtkins> | The answer is that the pseudostructure is actually <p><::first-line>foo foo foo </::first-line><span><::first-line>bar</::first-line> bar bar</span></p> |
| 22:48 | <Lachy> | but where is it defined in CSS which properties apply to ::first-line? |
| 22:48 | <Lachy> | yeah, I get that |
| 22:49 | <TabAtkins> | In 2.1, under the defifnition of first-line. |
| 22:49 | <Lachy> | that's where I'm looking |
| 22:50 | <TabAtkins> | Very last line paragraph of that section. |
| 22:50 | <TabAtkins> | Just above the ::first-letter heading. |
| 22:50 | <Lachy> | ah, yeah. I missed that when I first looked |
| 22:53 | <TabAtkins> | For some reason 2.1 suggests that UAs should break up the span to keep the ::first-line intact, but that's obviously stupid and nobody does it that way afaik. |
| 23:11 | <AryehGregor> | Nice, git-svn finished in less than an hour. |
| 23:12 | <othermaciej> | TabAtkins: is ::wrap() allowed to take arbitrarily complex selectors? |
| 23:12 | <TabAtkins> | Yes. |
| 23:13 | <othermaciej> | does it take only a selector or group of selectors? |
| 23:14 | <TabAtkins> | only a selector, so that it can use the , to separate the start-selector and end-selector, but by using the :matches() pseudoclass you can combine multiple selectors. |
| 23:14 | <TabAtkins> | :matches() doens't exist yet, but it's just a matter of time. It'll be implemeneted long before ::wrap |
| 23:15 | <TabAtkins> | So, frex, you can create section wrappers in a page that uses only implied sections with ::wrap(h1,h1), ::wrap(h2,:matches(h1,h2)), etc. |
| 23:17 | <TabAtkins> | Though to be fully general you'd have to deal with <hgroup> as well, and you can't do that until the :has() pseudoclass is supported. |
| 23:18 | <TabAtkins> | ::wrap(:matches(h1,hgroup:has(h1)), :matches(h1,hgroup:has(h1))) {} |
| 23:18 | <TabAtkins> | And :has isn't going to be implementable for a while yet, apparently. |
| 23:21 | <TabAtkins> | Though, that could be re-simplified with a :heading() pseudoclass. ::wrap(:heading(1),:heading(1)) {}, ::wrap(:heading(2),:heading(2 - n)) {}, etc. |
| 23:22 | <TabAtkins> | There are a number of pseudos that I want access to yesterday, obviously. ^_^ And they should be cheaply implementable, too. |
| 23:51 | <AryehGregor> | Man, the -S option to git log is slow. |
| 23:51 | <AryehGregor> | Totally CPU-bound. |
| 23:51 | <AryehGregor> | (Of course, SVN doesn't even offer it. I shudder to think what it would be like trying to do that client-side with SVN.) |
| 23:52 | AryehGregor | realizes that for his purposes he can just use git bisect |
| 23:52 | AryehGregor | hates how you need one good rev and one bad for that to work, though . . . |
| 23:56 | <AryehGregor> | Hmm, git isn't so happy with multi-MB files either, probably. |