| 01:37 | jwalden | finds an error in the web socket protocol draft |
| 02:08 | <jwalden> | or have I |
| 02:37 | jwalden | curses the mess that is RFC 2817 and its offhand definition of the CONNECT method |
| 02:46 | <jwalden> | no, seems to be an error |
| 03:22 | <jwalden> | email sent to that list, seems a bit of delay in its being visible in the web interface |
| 03:30 | <jwalden> | http://www.ietf.org/mail-archive/web/hybi/current/msg00312.html |
| 03:32 | <jwalden> | ...and fixed already :-) |
| 04:44 | <Hixie> | hey, we crossed the 1000 subscriber barrier! |
| 04:44 | <Hixie> | sweet |
| 04:50 | <takkaria> | Hixie: what big of the spec are you working on now? |
| 04:50 | <Hixie> | none at this very instant |
| 04:50 | <Hixie> | probably going to go through htmlwg bugs and try to get them down |
| 04:50 | <takkaria> | cool |
| 04:51 | <Hixie> | (i'll work on whatever people want me to work on, so if you have something important, let me know) |
| 04:51 | <ezyang> | Nice! |
| 04:52 | <Hixie> | this week has been hell in terms of getting anything done |
| 04:52 | <Hixie> | what with everyone and their dog talking about html5 because of Google I/O |
| 04:55 | <inimino> | I will admit to being mildly disappointed in the (ahem) wave of recent interest in HTML5 but I'm not quite sure why |
| 04:56 | <Hixie> | disappointed ho? |
| 04:56 | <Hixie> | how, rather |
| 04:56 | <othermaciej> | disappointed? |
| 04:56 | <othermaciej> | why? |
| 04:56 | <Hixie> | he said he wasn't sure why :-) |
| 04:56 | <Hixie> | that's why i asked how instead :-) |
| 04:56 | <othermaciej> | (I have been mildly irked by the garbling of the technical message.) |
| 04:56 | <othermaciej> | (But I like that the overall response is positive.) |
| 04:57 | <inimino> | yes most of it is probably just seeing "HTML5" become the next "Ajax" |
| 04:57 | <ezyang> | HTML5 is huge. (I mean that in a scope way) |
| 04:57 | <othermaciej> | HTML5 is huge, and people use it to refer to things that are not, technically, part of HTML5 |
| 04:57 | <othermaciej> | one thing that surprised me in the coverage was how in sync Apple and Google especially are with respect to what is most important in Web technology |
| 04:57 | <ezyang> | And it's building up the "eye-candy" factor that will sell it to the public |
| 04:57 | <othermaciej> | if I didn't know better, I would think it was a conspiracy |
| 04:58 | <Hixie> | othermaciej: if you think that message got garbled, just check out the lower-third text on http://cache.gawker.com/assets/images/gizmodo/2009/05/240ccede5360b093dbf298f8946025a5.png |
| 04:58 | <othermaciej> | Hixie: lol |
| 04:58 | <othermaciej> | (I actually laughed out loud) |
| 04:58 | <othermaciej> | so maybe it's just me |
| 04:58 | <inimino> | hehe |
| 04:59 | <jruderman> | i'm thinking of making a game |
| 04:59 | <othermaciej> | but I feel like back in the 80s and early 90s, technical reporters seemed to have more of a clue |
| 04:59 | <jruderman> | and if i do, i'll probably use javascript and <canvas> |
| 04:59 | <othermaciej> | I remember reading PC MAgazine back in the day |
| 04:59 | <othermaciej> | and it was full of technical facts and stuff |
| 04:59 | <jwalden> | othermaciej: did you ever see the coverage of the Morris worm? they didn't have a clue :-D |
| 04:59 | <othermaciej> | nowadays I look at CNet and I cry |
| 04:59 | <inimino> | well I think screen grabs of ridiculous Fox News tickers are practically a genre of their own by now... |
| 05:00 | <inimino> | I just don't look at CNet |
| 05:00 | <othermaciej> | ArsTechnica is one of the few beacons of relative technical clue, as far as tech news sites, and even they publish total claptrap at times |
| 05:00 | <inimino> | I think the technical audience has moved on and they are all reading each other's blogs instead of magazines |
| 05:01 | <jwalden> | demise of ddj and all that |
| 05:01 | <othermaciej> | with blogs it's hard to tell up front who is full of shit |
| 05:01 | <othermaciej> | unless you are an expert in the particular subject are yourself, or know the person by reputation |
| 05:01 | <jruderman> | i'll help you out. my blog -- http://www.squarefree.com/ -- full of shit. |
| 05:01 | <othermaciej> | like Gizmodo is super popular and clearly appeals to tech enthusiasts but it is at best 50% true |
| 05:01 | <inimino> | true |
| 05:02 | <othermaciej> | jruderman: I've rarely known you to be full of shit |
| 05:02 | <othermaciej> | and... |
| 05:02 | <othermaciej> | depending on the nature of the game, JS + <canvas> is probably a decent choice |
| 05:02 | <othermaciej> | have you seen olliej's fluid dynamics demo? |
| 05:02 | <jruderman> | yep |
| 05:03 | <jruderman> | then i left it open in a background tab and wondered why my computer was hot |
| 05:03 | <othermaciej> | I don't know why but I find it endlessly fun to play with |
| 05:03 | <othermaciej> | it does manage to replicate that feature of Flash games |
| 05:03 | <jruderman> | hehe |
| 05:04 | <othermaciej> | whoah he managed to boost the frame rate |
| 05:04 | <inimino> | where is it? |
| 05:04 | ezyang | imagines the new cookie cutter industry: "Free, Addictive HTML5 Games!" |
| 05:04 | <othermaciej> | I'm getting nearly 60 fps here |
| 05:04 | <othermaciej> | http://nerget.com/pressure/pressure.html |
| 05:04 | <Hixie> | i just get a black screen :-( |
| 05:04 | <othermaciej> | it numerically approximates the solution to the Navier-Stokes equations of fluid dynamics |
| 05:04 | <Hixie> | oh i see |
| 05:04 | <Hixie> | you have to click |
| 05:04 | <othermaciej> | Hixie: you have to click or drag to make things happen |
| 05:05 | <othermaciej> | and if you toggle the view, it flips from pressure to velocity |
| 05:05 | <othermaciej> | I wanna put this up on webkit.org but I have to get rid of the layout table first |
| 05:05 | <othermaciej> | or people will laff at us |
| 05:05 | <jruderman> | lol |
| 05:05 | <inimino> | ooh, fun |
| 05:06 | <othermaciej> | Dear CSS Layout, Please learn how to be easy like tables. Love, The Internet. |
| 05:06 | <inimino> | now it just needs the audio equivalent of <canvas> and you can add splooshy sounds |
| 05:06 | <othermaciej> | realtime audio generation is very hard |
| 05:06 | <othermaciej> | actually |
| 05:06 | <othermaciej> | the latency requirements to get smooth realtime sound generation are ridiculous |
| 05:07 | <othermaciej> | so it's a much harder problem than canvas :-/ |
| 05:07 | ezyang | is mesmerized |
| 05:07 | <jruderman> | othermaciej: latency requirements? |
| 05:07 | <jruderman> | what makes it different from graphics? |
| 05:08 | <othermaciej> | jruderman: for audio to feel reactive, you need a fairly small buffer - audio delay is much more noticeable than video |
| 05:08 | <othermaciej> | jruderman: but on the flip side, if you don't fill the buffer in time, you get horrible sounding artifacts |
| 05:09 | <othermaciej> | this is why things like CoreAudio and OpenAL are so complicated |
| 05:10 | <jruderman> | games do fine choosing prerecorded sound effects and timing them to go with the video |
| 05:10 | <doublec> | that's what js8080 does for space invaders, etc, using <audio> |
| 05:11 | <jruderman> | what changes when you're creating the sound effects on the fly? |
| 05:11 | <doublec> | I would like a openal type api for audio however (in the browser) |
| 05:11 | <doublec> | you can generate wav data and use data url's to <audio> |
| 05:12 | <othermaciej> | with prerecorded sounds, you don't have the problem of the audio output buffer underflowing cause you couldn't fill it fast enough |
| 05:13 | <othermaciej> | having an ImageData like approach where you provided a large-ish buffer that gets mixed in might work |
| 05:13 | <othermaciej> | but doesn't have much of an obvious advantage over using <audio> |
| 05:15 | inimino | switches back to the fluid dynamics window and discovers The Blob |
| 05:19 | <othermaciej> | heh |
| 05:20 | <othermaciej> | I wish olliej let you control the viscosity |
| 05:43 | <jwalden> | um... http://twitter.com/WHATWG/status/1953544919 |
| 05:43 | <jwalden> | way to go twitter |
| 05:44 | <ezyang> | Wow. Twitter doesn't escape ampersands |
| 05:44 | <jwalden> | gotta be an XSS waiting to happen in that somehow, I'd expect |
| 05:45 | <ezyang> | Mm, not really |
| 05:46 | jwalden | notes that account has 2**10 followers |
| 05:50 | mgrdcm | +1 |
| 06:32 | jwalden | wonders what specification would have the most claim to being responsible for describing what must happen when DOM APIs are provided with non-UTF-16 strings, like "foo\ud863\u000abar" -- webidl? |
| 06:32 | jwalden | is looking again at acid3 test 68, for a little context |
| 06:33 | <heycam`> | webidl doesn't say anything about non-UTF-16 strings in DOMStrings, at the moment |
| 06:33 | <heycam`> | is acid test 68 the svg string methods? |
| 06:34 | <heycam`> | ah no, ecmascript i see |
| 06:35 | <jwalden> | no, it's |document.createElement("input").value = highSurrogate + asciiData| and whether that throws an exception, sets the identical value, or sets a value with the high surrogate replaced with U+FFFD |
| 06:35 | <jwalden> | or rather utf16Data, I guess |
| 06:35 | jwalden | thinks he enumerated all the permitted behaviors |
| 06:36 | <heycam`> | yeah i'm not sure what layer that should be handled on |
| 06:36 | <jwalden> | ah, missed one, could set to utf16Data as well |
| 06:36 | <jwalden> | DOM 3 Core seems to not even entertain the prospect |
| 06:36 | <heycam`> | heh |
| 06:37 | <jwalden> | webidl says "(to be interpreted as UTF-16 encoded strings [RFC2781])" |
| 06:37 | <heycam`> | taboo, please don't mention it in mixed company |
| 06:37 | <jwalden> | rfc2781 of course says "Error recovery is not specified by this document." |
| 06:38 | <heycam`> | if there's consistent behaviour across attributes / operation arguments, then specific behaviour could be mandated in webidl |
| 06:38 | <heycam`> | i haven't tested that though |
| 06:43 | <jwalden> | gecko sets a value with the high surrogate replaced with U+FFFD; webkit preserves the code units exactly; opera does the same; ie does the same |
| 06:44 | <jwalden> | this is rather unfortunate, as gecko's behavior is rather deeply embedded, as best as I recall |
| 06:44 | <heycam`> | the latter seems cleaner to me |
| 06:45 | <jwalden> | testing with javascript: var e = document.createElement("input"); e.value = "\ud863test"; alert(e.value.split("").map(function(v) { return v.charCodeAt(0); })) in non-IE and javascript: var e = document.createElement("input"); e.value = "\ud863test"; alert(e.value.split("")) in IE |
| 06:47 | <jwalden> | although looking slightly, this may end up being specific to input.value, not sure yet |
| 06:48 | <heycam`> | k. i'll put a note in webidl and see if anyone has an opinion on it being dealt with in that spec. |
| 06:50 | <jwalden> | cool |
| 06:53 | <jwalden> | hm, looks like it turns into a ghastly mess partly because attribute values are optimized into particular encodings, if I'm reading right |
| 06:53 | <jwalden> | and even beyond that things like numbers are stored as such |
| 07:01 | <jwalden> | hm, or maybe no, but now I'm diving into template metaprogramming of string manipulation |
| 07:03 | <jwalden> | stupid traits antipattern |
| 07:05 | <jwalden> | heycam`: another fun one: javascript: var e = document.createElement("input"); e.value = "\ud863\0\0test"; alert(e.value.split("")) |
| 07:06 | <jwalden> | seems all the browsers I can test truncate that |
| 07:06 | <jwalden> | gah, what a mess |
| 07:08 | <heycam`> | ouch |
| 07:10 | <jwalden> | based on what I know of gecko, I bet we truncate to U+FFFD while the others truncate to a lone high surrogate |
| 07:11 | <heycam`> | differing behaviour between javascript string classes and dom string classes doesn't seem good |
| 07:16 | <jwalden> | none of this is any good |
| 07:18 | <jwalden> | particularly since encoding mistakes have a knack for becoming security problems if you're unlucky |
| 09:46 | <Hixie> | i'm starting to need to keep track of e-mails for which i'm awaiting replies |
| 09:47 | <hsivonen> | Hixie: it seems to me that the comments at http://www.w3.org/QA/2009/05/_watching_the_google_io.html#comments don't deal with the main "it's not done yet" point of the post |
| 09:49 | <Hixie> | as i just mentioned on the owf list, html5 is more done than html4 ever was |
| 09:49 | <Hixie> | so that seems like a meaningless point :-) |
| 09:50 | <hsivonen> | Hixie: maybe that point would have been more useful to make in the Q&A/QA blog |
| 09:51 | <Hixie> | be my guest |
| 09:56 | <hsivonen> | I wonder when we'll see a log post warning about premature deployment of RDFa in text/html |
| 09:56 | <hsivonen> | *blog |
| 09:59 | <theMadness> | I think Crockford accidentally made the point for html5 a while ago. |
| 09:59 | <hsivonen> | theMadness: which point and where? |
| 09:59 | <theMadness> | http://blog.360.yahoo.com/blog-TBPekxc1dLNy5DOloPfzVvFIVOWMB0li?p=978 here |
| 10:00 | <theMadness> | He advocated for a new html that is not very new, just a fix of the 4.01. |
| 10:01 | <theMadness> | But if you take away webforms and the new elements (which are all more specific and semantic heavy versions of div), html5 is just what he calls html 4.2 |
| 10:03 | <theMadness> | But maybe he is considering the DOM changes too. |
| 10:05 | <tantek> | theMadness - this is not a new idea, I advocated for a new html that is not very new, just a fix of 4.01, back in 2003: http://tantek.com/log/2003/01.html#L20030114 |
| 10:05 | <hsivonen> | making "just a fix" *now* would be rather disappointing |
| 10:05 | <theMadness> | tantek, I see, but I recalled that post as when I read it in my mind it sounded "html5 is too much, let's have html5 instead". |
| 10:06 | <tantek> | hsivonen, I don't disagree. those were my opinions as of 6+ years ago. |
| 10:07 | <theMadness> | Which probably was the right time for them. |
| 10:08 | <theMadness> | But right now we're too high on RIA, stepping back will shoot us back to the plugin/applet/proprietary extensions era. :P |
| 10:08 | <tantek> | however, note that many folks here have been writing up well thought out criticisms of XHTML2 / focus-of-W3C for 6+ years. how little has changed in that regard is IMHO what is rather disappointing. |
| 10:08 | <Philip`> | Let me be the first to advocate for a new HTML that is not very new, just a fix of 5, so I can point back to this IRC log entry several years from now |
| 10:09 | <theMadness> | Philip`, you mean 4? |
| 10:10 | <Philip`> | theMadness: No - I'd be at least six years too late to be the first to advocate that :-) |
| 10:10 | <tantek> | Philip`, I think you have to wait at least until HTML5 is a REC (projected 2022?) and the first post-REC issue is discovered/acknowledged before advocating a new HTML that is just a fix of 5. |
| 10:11 | <theMadness> | What he said. I didn't have the courage to say it myself. |
| 10:11 | Hixie | mumbles that he doesn't think specs should be numbered in this way and that we should just have rolling versions the way we're doing HTML5 today and the way most websites work |
| 10:12 | hsivonen | I always feel uncomfortable when I see "RIA" in the context of HTML5. |
| 10:12 | <Hixie> | why? |
| 10:12 | <tantek> | Anyone seen mpilgrim recently? |
| 10:12 | <tantek> | This link appears to have broken: http://diveintomark.org/archives/2003/01/13.html#semantic_obsolescence |
| 10:12 | <tantek> | It was a criticism of XHTML2 IIRC. |
| 10:13 | <hsivonen> | Hixie: "RIA" seems to be a Flash/Silverlight/JavaFX not-the-bowser-platform term |
| 10:13 | <theMadness> | hsivonen, exactly the reason why we must "take it back". :> |
| 10:14 | <Philip`> | tantek: Why would post- vs pre-REC make a difference to the need to create a new version to fix issues? |
| 10:14 | <tantek> | because pre-REC, issues can be fixed without creating a new version. |
| 10:14 | <othermaciej> | I'm happy to say "Web application" |
| 10:14 | <othermaciej> | but WA is not a TLA |
| 10:14 | <othermaciej> | Standard-based Web Application? |
| 10:15 | <Hixie> | tantek: they can post-rec too |
| 10:15 | <tantek> | Hixie, indeed, that whole "2nd edition" trick |
| 10:15 | <Philip`> | tantek: If the spec is ever going to reach REC, at some point it will have to stop responding to issues (e.g. stop adding features) so it can stabilise and get test suites and interoperable implementations, long before it can become REC, and during that time period it won't be able to respond adequately to all issues |
| 10:15 | <Hixie> | anyway i'm sure html6 will have started long before html5 is in rec |
| 10:15 | <annevk2> | Hixie, btw, apparently you have some action item assigned to you |
| 10:16 | <Hixie> | annevk2: no, i don't |
| 10:16 | <theMadness> | By the way, I've seen complaints about versioning and the doctype, how will a browser be able to tell a html6 page? |
| 10:16 | <Philip`> | theMadness: It won't be able to |
| 10:16 | <Hixie> | theMadness: why would it need to? |
| 10:16 | <Philip`> | theMadness: (and it won't need to) |
| 10:17 | <theMadness> | To trigger legacy/edge support? |
| 10:17 | <tantek> | Philip`: sure, amend previous statement to pre-CR then :) |
| 10:17 | <annevk2> | tantek, http://diveintomark.org/archives/2003/01/13/semantic_obsolescence |
| 10:17 | <Hixie> | theMadness: we've never had to do that before, why would we do it now? :) |
| 10:18 | <annevk2> | Hixie, I guess you never accepted it, but all indications are that you have |
| 10:18 | <theMadness> | Wait, we did, what about quirks/standards? |
| 10:18 | <Hixie> | annevk2: check the tracker |
| 10:18 | <annevk2> | ah |
| 10:18 | <Hixie> | theMadness: quirks vs standards isn't done along version lines |
| 10:18 | <tantek> | annevk2, note that the blog post at that URL links to http://diveintomark.org/archives/2003/01/14.html#eddies_in_the_spacetime_continuum which is also 404 |
| 10:19 | <Hixie> | theMadness: or to put it another way, why would we want to add more quirks modes? |
| 10:19 | <othermaciej> | I think the experience of quirks/standards has made many of us decide we never want to do that sort of thing again |
| 10:19 | <theMadness> | So basically html6 will have to be compatible with 5. |
| 10:19 | <Hixie> | (except microsoft, who decided that it was teh awesome) |
| 10:19 | <Philip`> | tantek: Okay, fair enough :-) |
| 10:19 | <Hixie> | (and added 3 or 4 new modes just in IE8) |
| 10:19 | <othermaciej> | it will have to be compatible with the corpus of deployed HTML content |
| 10:19 | <othermaciej> | which by that time will hopefully include lots of HTML5 |
| 10:20 | <annevk2> | tantek, do some URL hacking, http://diveintomark.org/archives/2003/01/14/eddies_in_the_spacetime_continuum :) |
| 10:20 | <othermaciej> | HTML5 is also trying to be compatible with deployed Web content as opposed to with prior specs |
| 10:20 | <othermaciej> | hopefully it will get less things wrong than HTML 4.01 did |
| 10:20 | <annevk2> | tantek, they used to redirect though, Mark dropped the ball with regards to cool URLs |
| 10:20 | <Philip`> | tantek: (...and amend the statement from 2022 to 2012) |
| 10:20 | <tantek> | did he drop his daily archives? |
| 10:21 | <annevk2> | tantek, no, they're just under a different URL |
| 10:21 | <Philip`> | tantek: But also, the people maintaining HTML5 a decade from now will not be the same people who are working on it today, and the new people who are innovating on the web will get fed up with the stuffy old HTML5 guys and will want to change it to go in a different direction |
| 10:21 | <annevk2> | tantek, i.e. without the .html |
| 10:21 | <tantek> | Philip`, indeed, need to get ahead of the Mayans (or was it Aztecs?) ;) |
| 10:22 | <othermaciej> | Philip`: does that mean we can all aspire to be the Roy Fielding of the future? |
| 10:22 | <Philip`> | othermaciej: It's not an aspiration, it's an inevitability |
| 10:23 | <tantek> | Philip`: I kind of doubt it - unless another Great Web Schism occurs. |
| 10:23 | <tantek> | speaking of which, is anyone organizing a celebration of the 5th anniversary of the Great Web Standards Schism? |
| 10:23 | <tantek> | http://www.w3.org/2004/04/webapps-cdf-ws/minutes-20040601.html |
| 10:23 | <othermaciej> | the people I work with who have age, experience and accomplishments similar to Roy do not have his kind of attitude |
| 10:24 | <theMadness> | Standards: the 21st century version of protestantism. |
| 10:24 | <Philip`> | othermaciej: Oh, okay then |
| 10:24 | <othermaciej> | (for instance Darin Adler has never told me to STFU because he used to create software by toggling in the 1s and 0s with a magnet) |
| 10:24 | <tantek> | that meeting is what really caused the momentum to get behind HTML5, and for that matter microformats. |
| 10:24 | <Philip`> | othermaciej: Hopefully we can all aspire to be more like him :-) |
| 10:24 | <othermaciej> | yes, the Workshop of the Bastille |
| 10:25 | <othermaciej> | Philip`: he only tells me to STFU when I actually need to and am not aware enough to realize it |
| 10:25 | <othermaciej> | (which is more rarely these days) |
| 10:25 | <tantek> | it was quite an epic workshop, othermaciej. at the time I didn't quite realize how historic an event it was going to be. |
| 10:25 | <othermaciej> | so some people mellow with age, like wine, instead of becoming sharper, like cheddar |
| 10:26 | <othermaciej> | I know a surprising number of people on the attendees list for non-web-standards-related reasons |
| 10:26 | <othermaciej> | (and of course many for web standards related reasons too) |
| 10:27 | <tantek> | re-reading the minutes, note that namespaces in data on the web are just as dead as they were back then, 5 years ago. |
| 10:28 | <othermaciej> | Creative Commons is doing its best to spread them via copypasta |
| 10:28 | <othermaciej> | does that count? |
| 10:29 | <tantek> | not when typical implementation (not going to name names) simply uses regexes to scrap NS prefixes, no, that doesn't count. ;) |
| 10:29 | <tantek> | and yes, we've critiqued ccREL already in this channel |
| 10:30 | <tantek> | its very design promotes combinatorial license proliferation which is actually a stated non-goal of both Creative Commons (numerous presentations etc.) and many experienced folks in the open source communities. |
| 10:31 | <othermaciej> | was it obvious at the time they were invented how hard namespaces would be to use and understand? |
| 10:31 | <Hixie> | othermaciej: i'm sure gramps has told you to STFU :-P |
| 10:31 | <tantek> | othermaciej, yes, namespaces in content have been ugly since day 1 |
| 10:31 | <othermaciej> | Hixie: he has, but (a) only when I actually needed to and (b) he did not justify it by giving a lengthy list of his dinosaur accomplishments |
| 10:31 | <tantek> | I think I heard the term somewhere else, but namespaces in content is an excellent example of "syntactic vinnegar" |
| 10:31 | <Hixie> | othermaciej: :-P |
| 10:32 | <othermaciej> | gramps is not even 5% as crusty as Roy |
| 10:32 | <Hixie> | othermaciej: i did not mean to imply gramps was like roy :-P |
| 10:32 | <othermaciej> | I've heard it as "syntactic salt", but yes |
| 10:33 | <tantek> | syntactic vinegar as in the opposite of syntactic sugar |
| 10:33 | <theMadness> | I'm officially hungry. |
| 10:33 | <othermaciej> | yeah, I get the reference |
| 10:33 | <tantek> | it makes the syntax uglier / less friendly without actually providing a benefit (and in many cases causing harm - e.g. silos) |
| 10:34 | <Philip`> | Syntactic salt as in raising one's blood pressure? |
| 10:34 | <theMadness> | Also, being from Modena, I'm not taking the disrespect for vinegar. :P |
| 10:34 | <tantek> | it's not balsamic |
| 10:34 | <theMadness> | Ah, someone did his homework. :) |
| 10:35 | <tantek> | I have some Mediterranean heritage ;) |
| 10:35 | <othermaciej> | "syntactic salt" is just a different joke way of saying "the opposite of syntactic sugar" |
| 10:36 | <tantek> | othermaciej - it's not the same thing |
| 10:36 | <Hixie> | http://www.flickr.com/photos/glutnix/3325234350/ lol |
| 10:36 | <tantek> | see http://en.wikipedia.org/wiki/Syntactic_sugar |
| 10:36 | <othermaciej> | that's not the way I have heard it used |
| 10:36 | <theMadness> | Complicated versus descriptive. |
| 10:36 | <othermaciej> | (the wikipedia note) |
| 10:37 | <theMadness> | Also head is not what he means by head. |
| 10:37 | <othermaciej> | I'll agree with the "a hoop the programmer must jump through" part, but not "designed to make it harder to write bad code" |
| 10:37 | <othermaciej> | but anyway |
| 10:37 | <Hixie> | right well lest i start actually editing the spec, i think i should go to bed now |
| 10:37 | <Hixie> | nn |
| 10:38 | <othermaciej> | I think we are on the same page with respect to </3 namespaces |
| 10:38 | <othermaciej> | g'night |
| 10:38 | <tantek> | right. syntactic vinegar is both a hoop the programmer/author must jump through, and is designed (perhaps unintentionally) to make it harder to write easy code. |
| 10:38 | <tantek> | now, who's willing to add that description of syntactic vinegar to http://en.wikipedia.org/wiki/Syntactic_sugar ? |
| 10:41 | <tantek> | ok, well if no one else is planning a 5 year remembrance/celebration of the Great Web Standards Schism - we can certainly do it at the microformats meetup dinner in SF on Tuesday 2009-153 at 21st Amendment: http://microformats.org/wiki/events/2009-06-02-weekly-dinner-sf |
| 10:41 | Hixie | is too busy looking forwards to look back :-P |
| 10:42 | Philip` | is too busy looking at his navel to look either backwards or forwards |
| 12:01 | <othermaciej> | the Schism is ancient history in internet time |
| 12:05 | <annevk2> | pre-historic, even :) |
| 12:07 | <annevk2> | I like the name "The Workshop" better though (coined by hsivonen iirc) |
| 12:13 | heycam | finds it interesting that he presented at The Workshop even though he didn't know of its significance at the time |
| 16:21 | <annevk42> | http://twitpic.com/69dic #contentsniffingfail |
| 16:23 | <myakura> | :( |
| 16:27 | myakura | regrets that he should've test that with http://tools.ietf.org/id/draft-abarth-mime-sniff-00.txt |
| 16:28 | <annevk42> | heh |
| 16:29 | <annevk42> | meanwhile I'm getting pretty close to the point where I can check in XMLHttpRequest again |
| 16:30 | <annevk42> | and I guess then I've to port everything to XMLHttpRequest Level 2 -- sigh |
| 16:47 | <gsnedders> | The sun is a great anti-productivity tool. |
| 16:50 | <annevk42> | weekends are too |
| 16:51 | gsnedders | has sand in his mouth from the beath |
| 16:51 | <gsnedders> | *beach |
| 19:44 | <gsnedders> | Is there any way to force Opera to reload from server, ignoring cache? |
| 19:46 | <Philip`> | gsnedders: Click the reload button |
| 19:46 | <Philip`> | (I think that tells it to ignore caches) |
| 19:47 | <Philip`> | (though I could be wrong) |
| 19:48 | <Dashiva> | Yes |
| 19:55 | gsnedders | is slowly adjusting to this alien browser :P |
| 19:59 | <annevk42> | http://jszen.blogspot.com/2009/05/html-5.html -- I'm guessing he's mostly talking about the parts splitted out of HTML5 but it's still nice to hear :) |
| 20:00 | <Hixie> | this press has certainly been very encouraging |
| 20:00 | <Hixie> | it's almost all positive |
| 20:51 | <Philip`> | Hixie: "return to the top of step labeled <i>loop</i>" - should say "*the* step" |
| 20:52 | <Hixie> | fixed |
| 20:53 | gsnedders | implemented that, yet didn't notice that. |
| 20:53 | <gsnedders> | Oh well, I'm obviously a brilliant reviewer :P |
| 20:57 | gsnedders | wonders why that section is #history-0 |
| 20:57 | <gsnedders> | I think I do say the order in which ids are generated is undefined. |
| 21:00 | <theMadness> | Sigh, people is blind. |
| 21:01 | <theMadness> | They didn't get excited about html5 until google told them to. |
| 21:01 | <theMadness> | How sad is that? |
| 21:01 | <Hixie> | pretty much par for the course |
| 21:01 | gsnedders | wonders if anyone in standards land is the slightest bit surprised by that |
| 21:02 | <theMadness> | Heh, gsnedders, I'm allowed to, I'm a lowly developer :D |
| 21:02 | gsnedders | stands up on high |
| 21:02 | <gsnedders> | :D |
| 21:07 | <jruderman> | "Mozilla executives also promised HTML5 support inside the forthcoming Firefox 3.5 browser." orly |
| 21:07 | <theMadness> | I wonder if that includes webforms. |
| 21:08 | <theMadness> | or "the part of html5 formerly known as webforms". |
| 21:08 | <Hixie> | it's still known as webforms :-P |
| 21:08 | <gsnedders> | Alternatively, "HTML 5 forms" |
| 21:08 | <Hixie> | or "forms" for short :-P |
| 21:08 | <gsnedders> | Or: "Bob's your uncle." |
| 21:08 | <theMadness> | Formshizzle. |
| 21:45 | gsnedders | is sure ezyang said he had got tests2.dat working |
| 21:59 | gsnedders | knows he's gonna end up checking over the tree constructor |
| 22:30 | <annevk42> | afaik only Opera does HTML5 forms |
| 22:31 | <theMadness> | Most of it, yep. |
| 22:31 | <theMadness> | Except stuff like placeholder. |
| 22:31 | <annevk42> | though WebKit does some I think and I read on IRC that someone from Google was looking into implementing more |
| 22:32 | <jgraham> | hmm. I think the mosty recent change to tests2.dat is wrong |
| 22:33 | <jgraham> | <head></html><meta> should put <meta> as a child of <head> because </html> in <head> puts you in the after-head mode (not the in-body mode) and in after-head headish elements get appended to the head |
| 22:34 | <jgraham> | Anyone fancy confirming that for me before I revert the change? |
| 22:35 | <annevk42> | let me check |
| 22:37 | <annevk42> | you're wrong |
| 22:38 | <jgraham> | Oh |
| 22:39 | <jgraham> | Really? Why? |
| 22:43 | <Hixie> | you seem correct to me |
| 22:44 | <Hixie> | oh no |
| 22:44 | <Hixie> | the </html> implies <body> first |
| 22:44 | <Hixie> | so you fly through the after head mode, then into in body mode, then into after body, then after after body |
| 22:44 | <Hixie> | then you consume the </html> |
| 22:44 | <Hixie> | and then the <meta> goes only back as far as in body mode |
| 22:46 | <jgraham> | Ah, I forgot to reprocess the </html> in after-head |
| 22:46 | <jgraham> | Thanks :) |
| 22:56 | <Dashiva> | jgraham: I made the same mistake yesterday, when the fix was discussed |
| 22:56 | <Dashiva> | You end up in "in head", but the stack has body. Sneaky! |
| 23:00 | jgraham | thinks he has done his merge the wrong way around |
| 23:01 | <jgraham> | So I have ended up with the tip on a branch called svgmathml |
| 23:12 | <gsnedders> | Oh, if only HTML were logical… |
| 23:15 | <Hixie> | > For example, perhaps the document should advised following |
| 23:15 | <Hixie> | > the algorithms only when it is clearly necessary |
| 23:15 | <Hixie> | > (SHOULD NOT perform content sniffing EXCEPT when |
| 23:15 | <Hixie> | > necessary because of continued misconfiguration of |
| 23:15 | <Hixie> | > HTTP servers), |
| 23:15 | Hixie | wonders how that would work |
| 23:16 | <Philip`> | You could add a X-Not-Misconfigured header to servers that aren't misconfigured |
| 23:16 | <takkaria> | ha |
| 23:17 | <Hixie> | lol |
| 23:26 | <gsnedders> | Content-Type-But-I-Really-Mean-It-This-Time? |
| 23:28 | <Philip`> | Quite a few people use X-Content-Type-Options: nosniff already |
| 23:28 | <Dashiva> | Recreate the universe so that all http servers ship with an automatic legacy: true header? :) |
| 23:28 | <Philip`> | (including most Google sites) |
| 23:31 | <Philip`> | You don't need to recreate the universe, you just need to shift into a worldtrack in which that's already true |
| 23:37 | <Hixie> | hsivonen! curse you for filing bug 6766! |
| 23:38 | <Hixie> | ok which browser should i base document.all on |
| 23:38 | <Hixie> | any browser vendors here want to bribe me to make theirs the compliant one? :-P |
| 23:46 | <annevk42> | I'm not sure I'm in favor of Opera's one. Can I get money from you for the reverse favor? |
| 23:54 | <Hixie> | who should i use instead? |
| 23:57 | <annevk42> | It seems best to be informed by all crappy impl of it |
| 23:58 | <Hixie> | heycam: yt? |
