| 07:04 | <theMadness> | Did a size attribute get dropped from the input type=file? |
| 08:12 | <hsivonen> | wow. running svn annotate on the spec is *slow* |
| 08:13 | hsivonen | wishes someone were caching a blame-annotated copy |
| 08:26 | <flox> | hello |
| 08:27 | <flox> | I retried to install html5 validator locally (Debian Lenny) |
| 08:27 | <flox> | still the same error http://paste.pocoo.org/show/222740/ |
| 08:30 | <MikeSmith> | flox: run the build script one more time |
| 08:30 | <MikeSmith> | that error is a know issue |
| 08:30 | <MikeSmith> | it always happens the first time you run the build |
| 08:30 | <MikeSmith> | you can fix it by running "build/build.py localent" |
| 08:31 | <MikeSmith> | or just be re-running "build/build.py all" |
| 08:31 | <MikeSmith> | or maybe even just re-running "build/build.py build" |
| 08:32 | <flox> | wow ... |
| 08:32 | <flox> | yesterday I retried it repeatedly, without success |
| 08:33 | <flox> | today it works on 2nd run |
| 08:34 | hsivonen | decides to fix the documentation right now |
| 08:35 | <MikeSmith> | flox: dunno what you were doing yesterday, but re-running the build twice with "all" should always work |
| 08:35 | <MikeSmith> | flox: but anyway, you should never need to do that again unless/until you do a fresh checkout of the sources |
| 08:35 | <Hixie> | hsivonen: there's an out of date one at http://www.whatwg.org/specs/web-apps/current-work/blame |
| 08:35 | <Hixie> | hsivonen: i regen it every now and then |
| 08:36 | <hsivonen> | Hixie: thanks |
| 08:36 | <flox> | MikeSmith: thank you |
| 08:39 | <hsivonen> | hmm. the commit message of http://html5.org/tools/web-apps-tracker?from=3194&to=3195&context= doesn't really give hints about why -- > commenent closing was deemed necessary |
| 08:39 | <MikeSmith> | flox: yw |
| 08:40 | <hsivonen> | flox, MikeSmith: edited http://about.validator.nu/#src |
| 08:40 | <Hixie> | not only is the group discussing the html5 license at the w3c a secret group, I can't even subscribe to their mailing list without being appointed a rep by google's AC |
| 08:40 | <Hixie> | anyone want to lay bets on the odds of a group that secretive coming up with an appropriate open license for the spec? |
| 08:41 | <hsivonen> | Hixie: do you remember the story behind -- > ? |
| 08:41 | <Hixie> | i think zcorpan was behind it, but that's just a guess |
| 08:41 | <MikeSmith> | hsivonen: doc update looks good. That makes it clear, anyway |
| 08:43 | <hsivonen> | Hixie: I'm willing to believe that without reparsing, we'd be worse off if -- > didn't close, but the bad news is that it breaks CNN Money |
| 08:43 | <MikeSmith> | hsivonen: would be nice to find somebody with ant or maven chops to rework the build |
| 08:44 | <hsivonen> | Whoever edits the source of CNN Money seems to think that putting spaces around -- makes them no longer act as commment delimites |
| 08:44 | <Hixie> | hsivonen: and that works in all other browsers? |
| 08:45 | <Hixie> | hsivonen: i think this is one of those cases where the three options are 1. x sites break, 2. x other sites break, or 3. we have a security vulnerability |
| 08:45 | <hsivonen> | Hixie: it seems that -- > doesn't terminate a comment in Opera 10.5x, IE8 or Chrome beta channel |
| 08:45 | <hsivonen> | MikeSmith: https://twitter.com/technomancy/status/10994115673 |
| 08:46 | <Hixie> | hsivonen: did you test safari? |
| 08:46 | <hsivonen> | Hixie: I didn't |
| 08:46 | <Hixie> | prolly the same as chrome then |
| 08:46 | <hsivonen> | these days, I tend to test WebKit as Chrome, since Chrome runs on Linux but Safari doesn't |
| 08:46 | <Hixie> | the webkit guys are implementing the spec too, if they get there soon enough maybe we can just bully our way through under the banner of HTML5 or something |
| 08:48 | <Matjas> | Is using the solidus ('/>') required when using the XHTML serialization of HTML5 (by sending the document with Content-Type: application/xhtml+xml)? |
| 08:48 | <Hixie> | Matjas: only if you omit the end tag |
| 08:48 | <Hixie> | Matjas: in xml you can do <foo></foo> or <foo/>, they mean the same |
| 08:48 | <Hixie> | e.g. <img></img> or <img/> |
| 08:48 | <MikeSmith> | hsivonen: :) I guess I had assumed ant and maven were actually preferable to however else to put together a build for java apps.. but perhaps now |
| 08:48 | <MikeSmith> | *not |
| 08:48 | <Matjas> | Hixie: Thanks. |
| 08:49 | <Matjas> | Hixie: So <img></img> would be valid HTML5 (in XML mode)? |
| 08:49 | <hsivonen> | MikeSmith: well, obviously build.py isn't so nice, either |
| 08:50 | <Hixie> | Matjas: well you need src="" and alt="" also, but yeah, from a syntax perspective it's correct in XML |
| 08:50 | <hsivonen> | Matjas: <img alt='...' src='...'></img> is valid XHTML5, yes |
| 08:50 | <Hixie> | it's just like XHTML1 :-) |
| 08:51 | <Matjas> | I knew it was syntactically equivalent, but I didn't know it was actually confirming/valid. Mind = blown. Thanks guys! |
| 08:55 | <hsivonen> | Python md5 module deprecation warning is another piece of gratuitous change that humanity could have done without |
| 08:57 | <hsivonen> | MikeSmith: speaking of Maven, I need to learn how to make Maven run some Java code supplied by me as part of the build process... |
| 08:57 | <hsivonen> | also, the piece of code I have modified a .java file in place instead of having distinct input and output |
| 08:57 | <hsivonen> | which is probably considered wrong |
| 08:58 | hsivonen | gestures angrily in HotSpot's general direction |
| 08:59 | <MikeSmith> | I seem to run into the "considered wrong" cases often when I write anything at all in Java |
| 08:59 | <MikeSmith> | Java ecosystem seems to consider a lot of things wrong |
| 08:59 | <hsivonen> | hah |
| 09:00 | <Matjas> | Why is it that html5.validator.nu chokes on </img>? http://html5.validator.nu/?doc=http://mathiasbynens.be/edu/test.html&showsource=yes Note that this is in HTML serialization mode. |
| 09:01 | <Hixie> | HTML != XHTML |
| 09:01 | <Hixie> | </img> is only valid in XHTML |
| 09:01 | <hsivonen> | (The reason why I made it modify the file in place is that I don't want to rewrite the identity of the class just to apply a workaround for release binaries |
| 09:01 | <hsivonen> | ) |
| 09:01 | <MikeSmith> | hsivonen: ah |
| 09:02 | <MikeSmith> | yeah, that makes sense |
| 09:02 | <Matjas> | Fair enough. Now, to trigger XML mode, I just need to serve the file with a header like Content-Type: application/xhtml+xml;charset=UTF-8, right? |
| 09:03 | <Matjas> | Or do I need to make other changes to the document as well? |
| 09:03 | <Matjas> | http://mathiasbynens.be/edu/test.html (content-type header added) seems to break, I must be doing something wrong. |
| 09:03 | <Hixie> | you need xmlns="http://www.w3.org/1999/xhtml"; on the <html> |
| 09:04 | <Hixie> | you don't need the doctype in xhtml mode |
| 09:04 | <Hixie> | but if you have it it has to be uppercase |
| 09:04 | <Hixie> | <!DOCTYPE html> |
| 09:04 | <Hixie> | frankly though i wouldn't bother using XHTML unless you have a really good reason to :-) |
| 09:05 | hsivonen | wishes the last argument of addEventListener were optional |
| 09:05 | <Matjas> | Don't worry, I'm never gonna use XHTML. Just trying to understand how the XML serialization works :) |
| 09:05 | <Hixie> | :-) |
| 09:31 | <jgraham> | I think the -- > thing caused a bug in Opera or so, but zcorpan will remember |
| 09:33 | <hsivonen> | jgraham: caused a bug which way? |
| 09:33 | <jgraham> | Not supporting it |
| 09:33 | <hsivonen> | jgraham: in an internal build that no longer reparses? |
| 09:33 | <jgraham> | Umm, I might be thinking of the wrong thing |
| 09:34 | <jgraham> | zcorpan really has a better memory than me |
| 09:34 | <hsivonen> | Well, I CCed zcorpan on the Mozilla bug. |
| 09:45 | <henrikl> | Quick question: HTML5 + Live streaming video. What options do I have? |
| 09:53 | <doublec> | henrikl, browsers that support theora will play live streams using icecast |
| 09:57 | <henrikl> | doublec: Ok. Do you know how the lay of the land is with regards to H.264? |
| 09:57 | <doublec> | henrikl, unfortuntately not. I believe safari can do it via quicktime but I don't know the details. |
| 09:58 | <henrikl> | doublec: Ok, thanks! I`ll just have to set up some tests then. :) |
| 10:11 | <nessy> | darwin streaming server for quicktime |
| 10:11 | <nessy> | flumotion for theora and webm |
| 10:13 | <nessy> | several companies also offer h.264 streaming services using Adobe's Server |
| 13:53 | <hsivonen> | http://stackoverflow.com/questions/2972817/what-reasons-could-justify-apples-html5-demos-use-of-html4-doctypes |
| 13:59 | <Lachy> | It's good that people are recognising those "HTML5 demos" for what they are as just a marketing gimmick, rather than any serious attempt to showcasing standards. |
| 14:00 | <Phae> | the apple ones? |
| 14:00 | <Lachy> | yes |
| 14:00 | <Phae> | mm.. agreed. |
| 14:05 | <hsivonen> | Lachy: some people are, some people aren't |
| 14:07 | <Philip`> | Marketing gimmicks don't need to convince everybody in order to be successful |
| 14:07 | <Philip`> | I guess they just need to convince more people than the number who react adversely to the realisation that it's a gimmick |
| 14:08 | <hsivonen> | Lachy: e.g. Gruber's initial post seemed to take them at face value |
| 14:09 | <Lachy> | sure, we can't expect everyone to realise these things. |
| 14:32 | <jgraham> | You need to weight by the value of the person e.g. there is no point in convincing lots of people who can't afforrd Apple products that Apple are awesome. But there may be disproportionate weight in convincing percieved tech leaders |
| 18:25 | <Dashiva> | "In common usage, HTML5 may also refer to the additional use of CSS3, as both technologies are under development in parallel." |
| 18:25 | <AryehGregor> | In that case, presumably it can also refer to the use of SSDs or unmanned attack drones, since those are also under development in parallel. |
| 18:26 | <Dashiva> | Well, are you saying wikipedia is imperfect? |
| 18:26 | <AryehGregor> | No, that sentence accurately reflects popular use of "HTML5" as a buzzword. |
| 18:27 | <Dashiva> | "HTML5 is also used for any web technology that entered the public consciousness after 2005" |
| 18:27 | <Philip`> | HTML5 makes my computer faster and kills terrorists? |
| 18:27 | <Philip`> | Sounds like good marketing to me |
| 18:27 | <AryehGregor> | Well, maybe some unmanned attack drones have a web interface. I don't know, it's probably classified. |
| 18:50 | <zcorpan_> | Philip`: yo |
| 18:51 | <zcorpan_> | Philip`: do you have newer dotbot data than 2009-04? |
| 18:53 | <zcorpan_> | seems dotbot doesn't have newer data |
| 18:54 | <Philip`> | zcorpan_: I'm not aware of any newer data than the year-old release |
| 18:55 | <zcorpan_> | i was considering the possibility of running a new research for -- > for https://bugzilla.mozilla.org/show_bug.cgi?id=570309 |
| 18:58 | <gsnedders> | Philip`: Can you take a look in a similar manner to what you did for script for style with comments? |
| 19:00 | <Philip`> | What did I do for script for style with comments? |
| 19:00 | <gsnedders> | Parse Error. |
| 19:03 | gsnedders | tries to find that wiki page about comments in script |
| 19:05 | flox | was just reading this part :) http://is.gd/cGaXS |
| 19:40 | <jwalden> | "the first shipped implementation may have been developed in stealth mode without peer review during the development process" hah |
| 19:41 | <svl> | Hrm, H.264 for this Apple "FaceTime" video calling thing which they say will become "an open industry standard". |
| 19:43 | <jgraham> | Y'know I really wish the status markers in the W3C version of the spec weren't quite so ugly |
| 19:44 | <jgraham> | I feel somehow responsible |
| 19:44 | <jgraham> | flox: No need for URL shorteners here, this isn't twitter |
| 19:45 | <flox> | jgraham: it's my IRC client which is configured to shorten URLs |
| 19:48 | <AryehGregor> | What a broken IRC client. |
| 19:48 | AryehGregor | is wary of following shortened URLs from people he doesn't know |
| 19:49 | <jgraham> | Indeed |
| 19:50 | <Philip`> | Why are unshortened URLs any safer than shortened URLs? |
| 19:50 | <Philip`> | Unless you know and trust the domain name, it could be anything |
| 19:50 | <jgraham> | Philip`: Well that seems like a big win already |
| 19:51 | <jgraham> | You might also know and mistrust the domain name |
| 19:51 | <jwalden> | twitter will auto-shorten some URLs, in some cases; not sure why they don't just always do it, given the arbitrary length limits and all |
| 19:53 | <jgraham> | It also seems to be possible to totally break twitter posts by retweeting them so that the link at the end (i.e. the whole useful contents) gets totally stripped off |
| 19:53 | <jgraham> | The success of such a broken service is some kind of lesson |
| 19:53 | <AryehGregor> | Philip`, 1) Often I do know and trust the domain name. 2) Even if I'm not worried about a malicious URL, the name often gives a hint to the content. 3) Trolls are more likely to use a stock URL shortener for goatse than to register a whole domain. |
| 19:53 | <jgraham> | But I don't know what |
| 19:54 | <jgraham> | On the subject of which; |
| 19:54 | <jgraham> | gsnedders: You can't |
| 19:54 | <jwalden> | e.g. https://twitter.com/jswalden/status/5179465673 and https://twitter.com/jswalden/status/5179505187 demonstrating such a forced-shortening |
| 19:55 | <gsnedders> | jgraham: What? Number three? |
| 19:55 | <gsnedders> | jgraham: How evil do you think I am, bitch. |
| 19:55 | <gsnedders> | Oh, wait, that joke was too late on Saturday. |
| 19:55 | <gsnedders> | s/, bitch// |
| 19:55 | <jgraham> | gsnedders: Counting from the top down? Yes |
| 19:56 | <jgraham> | (one based) |
| 19:56 | <jgraham> | And no, I have no idea why you are calling me a bitch |
| 19:56 | <gsnedders> | jgraham: I mean the third item in AryehGregor's list. |
| 19:56 | jgraham | is confused |
| 19:56 | <jgraham> | But that is OK |
| 19:56 | <jgraham> | It's fun |
| 19:56 | AryehGregor | is too. |
| 19:57 | <gsnedders> | Uh, so Miss Squeaky was in car giving directions to their flat as if a GPS. I started appending "bitch" on to the end of all the instructions. This may have remained a joke for the rest of the day. |
| 19:57 | <gsnedders> | jgraham: I don't entirely understand what I'm being told I can't do either |
| 19:58 | <jgraham> | Well that's all good |
| 20:00 | <AryehGregor> | Are spaces prohibited in href? If so, why? Do browsers not encode them consistently when making the request? |
| 20:04 | <Philip`> | AryehGregor: They're prohibited in the URL syntax, I think |
| 20:04 | <AryehGregor> | The validator says so. Do you know why? |
| 20:04 | <Philip`> | and it's not HTML's job to redefine the valid URL syntax |
| 20:06 | <AryehGregor> | Well, it looks like the browser encodes the spaces as %20. Something must tell it to do so, unless that's unspecced. |
| 20:07 | <AryehGregor> | Anyway, it's HTML's job to say what markup is valid. It could provide a preprocessing algorithm and then say the URL is valid HTML if the result of the preprocessing is a valid URL. |
| 20:09 | <jgraham> | AryehGregor: It does |
| 20:09 | <jgraham> | http://www.whatwg.org/specs/web-apps/current-work/multipage/urls.html#valid-url |
| 20:10 | <jgraham> | The %20 thing comes from the "resolve an address" algorithm |
| 20:10 | <AryehGregor> | Okay, so can anyone actually answer my original question rather than giving me procedural reasons for why the question doesn't make sense? |
| 20:11 | <jgraham> | Which is, I think, part of the spec-hot-potato that is the URL specification |
| 20:11 | <AryehGregor> | Is there some reason to ban spaces in href/src/etc.? |
| 20:12 | <jgraham> | AryehGregor: From HTML's point of view it is because RFC3987 says so |
| 20:12 | <jgraham> | I imagine the RFC says tht so they people can split lists of URLs on spaces to get distinct URLs |
| 20:12 | <AryehGregor> | Why does HTML have to require that URLs in HTML match RFC3987? |
| 20:12 | <jgraham> | for example |
| 20:12 | <AryehGregor> | That's incompatible with URL processing in HTML already. |
| 20:12 | <gsnedders> | It doesn't |
| 20:13 | <gsnedders> | It requires RFCs match RFC3986 or if the document is encoded in UTF-8 or UTF-16 it matches RFC3987 |
| 20:13 | <gsnedders> | (IIRC) |
| 20:13 | <AryehGregor> | s/RFCs/URLs/ I assume? |
| 20:13 | <gsnedders> | Yeah |
| 20:15 | <jgraham> | gsnedders: That is mising the point somewhat |
| 20:16 | <gsnedders> | Of course, it's me. |
| 20:16 | <gsnedders> | What do you expect? |
| 20:16 | <jgraham> | AryehGregor: I assume because changing the definition of valid HTML would be a huge fight for no particular purpose |
| 20:16 | <jgraham> | You don't really gain much by allowing literal spaces |
| 20:16 | gsnedders | doesn't think it was a very common validation error |
| 20:16 | <Philip`> | Also you lose compatibility with tools that expect syntactically valid URL |
| 20:16 | <Philip`> | s |
| 20:17 | <AryehGregor> | I was asking because I noticed the error when validating a site (not one I control). |
| 20:17 | <Philip`> | ...when they're hooked into an HTML parser and a dumb attribute-value extractor |
| 20:18 | <Philip`> | although actually you get the same problem even when not involving HTML, because some sites send spaces in URLs in HTTP redirect headers |
| 20:18 | <Philip`> | so I guess those tools should just be fixed |
| 20:18 | <gsnedders> | (within the URL, leading/trailing whitepace is allowed there) |
| 20:19 | <gsnedders> | (I presume Philip` means) |
| 20:20 | <Philip`> | I don't mean leading/trailing whitespace |
| 20:55 | <gsnedders> | (Really, how I felt didn't change my behaviour around her that much. The fact that both of us are in general really physically has far more to do with it than anything else.) |
| 20:56 | <Philip`> | (You are really physically?) |
| 20:57 | <AryehGregor> | I'm not sure if I'd prefer to be physically or mentally. Maybe spiritually? Virtually? |
| 20:57 | <gsnedders> | Definately mentally. |
| 20:58 | <gsnedders> | And where the hell did that come from? |
| 21:00 | <gsnedders> | Oh, there. |
| 21:01 | <gsnedders> | Heh. IRC script searching through logs going back five years, one sent to the wrong window. Interesting. |
| 21:01 | gsnedders | wonders where that even comes fro |
| 21:01 | <gsnedders> | *from |
| 21:01 | gsnedders | thinks his Perl is wrong |
| 21:02 | <vlad_> | hey folks, WebIDL question -- if an interface 'operation void foo(in long a);' is called as x.foo() from JS |
| 21:02 | <vlad_> | should that throw an exception? since 'a' does not have [Optional], or will the argument be 'undefined' in JS, and then converted to a long (0?)? |
| 21:03 | <gsnedders> | Treating it as undefined doesn't apply because it's a host object |
| 21:03 | <gsnedders> | I think throw TypeError |
| 21:05 | <vlad_> | hmm, ok |
| 21:05 | <vlad_> | the WebIDL draft doesn't seem to mention this case at all |
| 21:06 | <gsnedders> | It does |
| 21:06 | <gsnedders> | http://dev.w3.org/2006/webapi/WebIDL/#call |
| 21:06 | <gsnedders> | It falls into step 3 of that |
| 21:07 | <vlad_> | well, that depends on what happens in the overload resolution |
| 21:07 | <vlad_> | there's even an ed note that says to test how implementations actually behave with too few or too many args :) |
| 21:08 | <vlad_> | the question is I guess whether the type list is "undefined" in that case, or is the wrong length |
| 21:08 | <gsnedders> | I think it is |
| 21:11 | <jgraham> | iirc (whithout reading any specs or anyhting) this ase is underdefined |
| 21:12 | <jgraham> | I think the simplest reading is that it throws TypeError, but I don't really trust the spec to be right about that |
| 21:15 | <jgraham> | gsnedders is really physically in the sense that he had corporeal existence |
| 21:15 | <jgraham> | *has |
| 21:15 | <jgraham> | Well I guess had last time I saw him |
| 21:16 | <jgraham> | I couldn't say what happened after that |
| 21:17 | <gsnedders> | You have evidence of corporeal existence then, proof I wasn't just a ghost? |
| 21:17 | <gsnedders> | s/have/had/ |
| 21:22 | <jgraham> | gsnedders: """Limits of the diaphane. But he adds: in bodies. Then he was aware of them bodies before of them coloured. How? By knocking his sconce against them, sure.""" |
| 21:23 | <gsnedders> | Hey, I've never finished that book. There again, nor have you. :P |
| 21:24 | <jgraham> | Indeed not |
| 21:24 | jwalden | wonders what book that is, and assumes it's some sort of "high-class" literature that is also thoroughly unreadable |
| 21:24 | <gsnedders> | jwalden: Ulysses, Joyce |
| 21:24 | <jgraham> | jwalden: Yes |
| 21:24 | <jwalden> | ah |
| 21:25 | <jwalden> | I read some of dubliners, it was interesting but not remarkable unless you were particularly omphaloskeptic in your reading |
| 21:25 | <gsnedders> | Do you really expect people to understand omphaloskeptic? |
| 21:26 | <jwalden> | c'mon, it's a fun word to use! :-D |
| 21:26 | <gsnedders> | True. |
| 21:26 | jwalden | demands an interesting story first and foremost; if the author wants to layer symbolism underneath that, that's fine, but it shouldn't be necessary to enjoy the book |
| 21:26 | <jgraham> | A Portrait of the Artist as a Young Man is excellent and accessible |
| 21:26 | Philip` | demands aliens and lasers and spaceships first and foremost |
| 21:27 | <gsnedders> | jwalden: Better yet is something where the symbolism is underneath a good story, and where the implied story is as good as the main one |
| 21:27 | <Philip`> | (or dragons) |
| 21:27 | <jgraham> | Philip`: But what about qualities yoou like in a book? |
| 21:27 | <jwalden> | gsnedders: true |
| 21:27 | <Philip`> | Actually, mostly I look for thickness |
| 21:27 | <Philip`> | All books seem to be about the same price so I might as well get one with twice as many pages |
| 21:28 | <gsnedders> | jgraham: Wow, that photo you uploaded of me on Flcikr makes me look really tired and otherwise terrible |
| 21:28 | <gsnedders> | *Flickr |
| 21:28 | <gsnedders> | <http://www.flickr.com/photos/jgraham/4679370417/>; is the photo in question. |
| 21:29 | <jgraham> | gsnedders: Slep more and you won't look so tired |
| 21:29 | <jgraham> | </helpful-advice> |
| 21:29 | <jgraham> | *Sleep |
| 21:29 | <gsnedders> | jgraham: "In the zoo, the mighty zoo/Geoffrey sleeps today" |
| 21:30 | <gsnedders> | There again, songs like that sung by people like me are probably not what you want mentioned :P |
| 21:30 | <gsnedders> | jgraham: Also, I slept 19 hours last nigh.t |
| 21:30 | <gsnedders> | *night |
| 21:30 | <gsnedders> | (and yesterday evening, and this morning) |
| 21:31 | <jgraham> | gsnedders: Well I guess you look less tired now |
| 21:31 | <gsnedders> | I looked less tired having slept in the car on the way back |
| 21:31 | <gsnedders> | Or at least was half asleep with my eyes shut for most of the time |
| 21:31 | <jgraham> | Although you should try to reduce the standard deviation |
| 21:31 | <gsnedders> | Yeah, totally :( |
| 21:32 | <jgraham> | A nice gaussian distribution centred on 8 or so |
| 21:32 | jgraham | realises this doesn't work in the limit of small t |
| 21:33 | <jgraham> | (since you can't slep for a negative amount of time) |
| 21:33 | <gsnedders> | jgraham: Also, damn you for taking way better photos than me |
| 21:33 | <jgraham> | *sleep |
| 21:33 | <jgraham> | dammit |
| 21:34 | <jgraham> | Really not very good |
| 21:34 | <gsnedders> | I didn't say they were very good, I just said way better than mine :P |
| 21:35 | <jgraham> | I know, that was just for the benefit of those following along at home |
| 21:38 | <AryehGregor> | Where does it say that CSS prefixes are dropped at CR? |
| 21:39 | <gsnedders> | Unwritten CSS WG policy, AFAIK |
| 21:40 | <AryehGregor> | I'm sure I've seen it written down somewhere. |
| 21:43 | <AryehGregor> | http://www.w3.org/TR/css-beijing/ |
| 21:44 | gsnedders | gets confused by documents like that being a TR |
| 21:45 | jgraham | just gets confused by documents like that |
| 21:45 | <gsnedders> | jgraham: You're always confused. |
| 21:46 | gsnedders | wonders whether it's reasonable to reply to emails saying, "At least you didn't top-quote." |
| 21:46 | <gsnedders> | Or even s/At/I might not care, but at/ |
| 21:47 | <jgraham> | gsnedders: Do you get many emails saying "At least you didn't top-quote"? |
| 21:47 | <gsnedders> | jgraham: No. |
| 21:47 | <gsnedders> | But I'm unconventional. |
| 21:47 | <jgraham> | Well it is quite unreasonable to reply to emails you don't get |
| 21:48 | <gsnedders> | I find it far more interesting to do that. |
| 21:49 | <jgraham> | I suppose you could create a fictional email to reply to |
| 21:49 | <jgraham> | People might be offended hough |
| 21:49 | <gsnedders> | I often do. |
| 21:49 | <jgraham> | +t |
| 21:50 | <gsnedders> | It's fun to see how many people believe the fictional email's authorship. |
| 21:50 | <jgraham> | gsnedders: I think you need to get out more |
| 21:50 | <gsnedders> | jgraham: You got any plans for Sunday? |
| 21:51 | <jgraham> | gsnedders: Not yet |
| 21:51 | <jgraham> | (afaik, usual disclimers apply) |
| 21:51 | gsnedders | will send email about Sunday tomorrow/Wednesday, unless beaten to it |
| 21:53 | <gsnedders> | (in which case I likely will send an email about it, just a reply) |
| 21:55 | <jgraham> | gsnedders: You should spend the time you spend talking about sending email sending email and the time you save doing something fun |
| 21:55 | <jgraham> | like making cupcakes |
| 21:56 | <gsnedders> | Not bagels, as suggested to me earlier? |
| 21:56 | <jgraham> | hmm? |
| 21:57 | <gsnedders> | (No, I haven't been online all day.) |
| 21:57 | <gsnedders> | jgraham: Steal you partner's Facebook account. |
| 21:57 | <gsnedders> | *your |
| 21:57 | <gsnedders> | (or, alternatively, get your own) |
| 21:59 | gsnedders | expects he was expected to put burgers inside the bagels |
| 21:59 | <gsnedders> | (Just because it alliterates) |
| 22:03 | <jgraham> | cupcakes can cunningly conceal cyanaide, causing curtains for consumers |
| 22:13 | <AryehGregor> | You couldn't figure out a way around "for"? |
| 22:14 | <AryehGregor> | Cupcakes can cunningly conceal cyanide, causing certain consumer cessation. |
| 23:33 | <Hixie> | christ |
| 23:33 | <Hixie> | the a11y community sure write bad alt text |
| 23:41 | <TabAtkins_> | Hixie: I find that too. Which is very surprising. Good alt text is ridiculously easy if you know wtf you're doing. |
| 23:42 | TabAtkins_ | doesn't get why anyone ever has issues with it, but they do for some reason, so shrug. |
| 23:43 | <AryehGregor> | Because normal people have no reason to learn how to write good alt text, because they have so few blind users that it makes no practical difference to them. |
| 23:43 | <AryehGregor> | HTML cannot mandate solutions contrary to economics. |
| 23:43 | <TabAtkins_> | I've found that good alt text has been helpful for me personally. I spent quite a while with a broken image once, frex, on a page that I read fairly regularly, because the alt-text was so good that I didn't notice it was broken. |
| 23:44 | TabAtkins_ | supposes it's kind of weird that he considers it helpful that it hid the existence of a broken image. |
| 23:44 | <TabAtkins_> | I only found out about it when I made my 404 page start emailing me. |
| 23:45 | <AryehGregor> | Just because there are isolated cases where it worked well doesn't mean that it would have been worth the effort to learn how, if you weren't involved in standards anyway. |
| 23:45 | <AryehGregor> | (so that you need the knowledge for non-authoring purposes) |
| 23:46 | <AryehGregor> | TabAtkins_, oh, right, I meant to ask you -- actually, I did ask you -- could you look at this, since you're a Chromium person and all? http://code.google.com/p/chromium/issues/detail?id=45804 |
| 23:46 | <TabAtkins_> | True, I acquired the knowledge because of involvement in standards, but my point is that it's *very easy* knowledge. It takes less than 5 minutes to learn. This makes it even worse that some of the a11y people can't do it right in their examples, at least. |
| 23:49 | <AryehGregor> | It might take less than five minutes to learn how to do it, but people won't learn it anyway unless there's obvious and immediate benefit. |
| 23:49 | <AryehGregor> | Most people don't formally learn HTML at all, they just imitate existing use and maybe read some tutorials or skim a book. |
| 23:50 | <AryehGregor> | If it looks right to them, then it's good enough. |
| 23:51 | <TabAtkins_> | Yeah, I know that. I'm specifically complaining about a11y people, though. |
| 23:51 | <TabAtkins_> | And their bad examples. |
| 23:52 | <AryehGregor> | Probably people who do use alt don't bother to learn the correct way because they think they already know it. |
| 23:52 | <AryehGregor> | Then it doesn't matter how easy it is to learn. |
| 23:53 | <AryehGregor> | I'm wondering who the a11y people at the W3C are. Are they mostly people who have professional expertise, like working with blind people on a regular basis (or being blind)? |
| 23:53 | <AryehGregor> | I suppose I could try Google-stalking a bunch of them, but I'm too lazy. |
| 23:53 | <AryehGregor> | Also, I don't really care. |
| 23:55 | <AryehGregor> | Hi, dglazkov! Do you have any thoughts on <http://code.google.com/p/chromium/issues/detail?id=45804>;? |