| 01:20 | <Yuhong> | <Hixie> anyone got IE handy? I need http://software.hixie.ch/utilities/js/live-dom-viewer/saved/1811's output |
| 01:21 | <Yuhong> | Using IE9 on Win7 in IE9 mode: Button, bamm, bamm2, bamm2 |
| 01:21 | <Hixie> | thanks |
| 01:22 | <Yuhong> | Using IE9 on Win7 in IE8 mode: Button, bamm, bamm2, bamm |
| 01:25 | <Yuhong> | Hixie: Did cwilson help you reverse engineer IE quirks, BTW? |
| 01:26 | <Yuhong> | I wish that MS could join the WHATWG, but in the meantime.... |
| 01:26 | <Hixie> | cwilson works for google now |
| 01:26 | <Yuhong> | And you work for Google too, right? |
| 01:27 | <Hixie> | yes, but i mean, he doesn't have any access that i don't... |
| 01:27 | <Yuhong> | No, but cwilson worked on IE at MS up to the IE8 era. |
| 01:27 | <Hixie> | i'm sure he's under NDA about that |
| 01:32 | <Yuhong> | Some other people that worked on IE before going to Google include jbeda, who worked on the DOM: http://news.ycombinator.com/item?id=3233935 |
| 01:40 | <MikeSmith> | can any Opera folk suggest somebody I can talk with at Opera about <ruby> ? |
| 06:09 | <hsivonen> | MikeSmith: a nice regexp you have sent for review… |
| 06:10 | <hsivonen> | this is why perl allows comments inside regexps |
| 06:35 | <hsivonen> | Hixie: in the syntaxt for the <time> element, why is there “A valid non-negative integer representing a year” without requiring a minimum of 4 digits? |
| 06:37 | <hsivonen> | what are the use cases for thing thing again… |
| 06:45 | <hsivonen> | *this thing |
| 07:09 | <[tm]> | hsivonen: "nice" is one way to describe it i guess :-) |
| 07:14 | <zcorpan> | matjas: https://www.w3.org/Bugs/Public/show_bug.cgi?id=18162 - html wg doesn't use LATER but instead changes product to HTML.next |
| 07:21 | <tantek> | zcorpan, I heard .next is now HTML 5.1 :) |
| 07:21 | <zcorpan> | tantek: there's no such thing in the Product drop-down :-P |
| 07:22 | <hsivonen> | [tm]: it took me about 50 minutes to dissect the regexp and compare it with the spec :-( |
| 07:22 | <zcorpan> | r-! |
| 07:22 | <tantek> | zcorpan - is that in bugzilla? I think there needs to be a product first before it can show up in the product drop-down. |
| 07:23 | <zcorpan> | tantek: yes. there is HTML.next in there which is being used |
| 07:23 | <tantek> | the nice thing about .next is that it's never .now |
| 07:24 | <zcorpan> | maybe they'll create a 5.1 product and move some bugs from .next to that |
| 07:24 | <zcorpan> | but that's just speculation |
| 07:34 | <othermaciej> | zcorpan: our tentative plan is to have products with names like "HTML spec" and "HTML spec (5.0 stable version)" |
| 07:34 | <othermaciej> | where "HTML spec" would cover all post-5.0 stuff |
| 07:34 | <othermaciej> | HTML.next product would be obsoleted |
| 07:34 | <othermaciej> | and someday there may be an "HTML spec (5.1 stable version)" component, but not yet |
| 07:35 | <hsivonen> | othermaciej: what’s the deal with publishing single-champion extension specs through the HTML WG? |
| 07:36 | <hsivonen> | othermaciej: how is that better than Hixie putting an unbaked idea in the WHATWG spec and making it look like it has more buy-in than it has? |
| 07:36 | <othermaciej> | hsivonen: are you looking to publish one, or are you worried that it will happen? |
| 07:37 | <othermaciej> | (or do you think there is an existing publication that is single-champion?) |
| 07:37 | <hsivonen> | othermaciej: Steve’s actions around hgroup confuse me |
| 07:37 | <othermaciej> | ah |
| 07:37 | <othermaciej> | so far he's got editor's drafts |
| 07:37 | <othermaciej> | perhaps he will take one or more to FPWD |
| 07:38 | <othermaciej> | I think it is unlikely all of them will be able to meet CR exit criteria, and I don't know if even one of them will (but if so, it probably deserves to replace hgroup) |
| 07:39 | <hsivonen> | othermaciej: so the path forward is to throw stuff on the wall and see if you can bait someone to implement so that you can exit CR? |
| 07:39 | <hsivonen> | and in the meantime, authors out there are confused about what they are supposed to use |
| 07:39 | <hsivonen> | and the W3C validator is supposed to support everything? |
| 07:39 | <othermaciej> | my personal opinion is that <hgroup> is not a very good syntax for subheadings but I don't particularly love any of Steve's alternatives |
| 07:40 | <othermaciej> | I doubt the W3C validator will support each of the alternatives |
| 07:40 | <othermaciej> | I agree there are potential downsides to "let the market decide", at least in this specific case |
| 07:41 | <othermaciej> | for stuff like longdesc or pubdate, it's less likely we'd have competing alternatives for the same use case |
| 07:41 | <hsivonen> | I’m generally unhappy about moving to FPWD without implementor buy-in and revisiting buy-in at CR exit |
| 07:41 | <othermaciej> | historically HTML WG has let you FPWD just about anything but we nominally ask for three independent supporters |
| 07:41 | <hsivonen> | If bad stuff is published like that if one implementor gets tricked into implementing, we might all have to implement bad stuff |
| 07:42 | <hsivonen> | (applies to Hixie’s unbaked ideas, too, but I sort of expected the W3C to improve this aspect) |
| 07:42 | <othermaciej> | I personally think that Microdata vs. RDFa was a case where "let the market decide" worked out ok and in the meantime ended the painful permathread flamewars |
| 07:43 | <tantek> | hsivonen - I think there's quite a bit of that going on: "bad stuff is published like that ... one implementor gets tricked into implementing, we might all have to implement bad stuff" |
| 07:44 | <danbri> | do we have any Microdata users (in terms of vocabulary) beyond schema.org yet? |
| 07:44 | <othermaciej> | granted, hgroup vs hsub vs outlineMask vs the other form of hsub vs sublime vs "don't even address the use case" is not quite the same |
| 07:45 | <tantek> | othermaciej - you mean worked out well for the market of folks (no pun intended) who like to learn handfuls of new attributes for adding semantics? |
| 07:45 | <othermaciej> | in that it is both lower stakes, and more confusing |
| 07:45 | <tantek> | hgroup always seemed fined to me (except for being potentially too rare a use-case to be worthy of inclusion in the core - despite the fact that I personally find it often handy) |
| 07:46 | <tantek> | it's the "outline algorithm" that never made any sense to anyone that I know of |
| 07:46 | <othermaciej> | the outline algorithm doesn't make much sense to me and I doubt it will make it out of the w3c's CR process |
| 07:46 | <tantek> | yeah, I don't know of anyone with any intent to implement it |
| 07:47 | <tantek> | might as well propose punting it to 5.1 now |
| 07:47 | <hsivonen> | othermaciej: I’m rather unhappy the Chairs just didn’t tell the RDFa folks to publish through their WG. after all, they have a WG, so they don’t need another WG to publish their stuff |
| 07:47 | <othermaciej> | I think one of the forms of <hsub> seems more logical to me than <hgroup>(though I would probably call it <subhead> or something) |
| 07:47 | <hsivonen> | (whereas the WHATWG didn’t have a WG) |
| 07:47 | <othermaciej> | hsivonen: we did now tell the RDFa folks to publish through their WG, and they will |
| 07:47 | <tantek> | hsivonen - it's because in order to be marketing buzzword compliant, you need to have an "HTML5 …" prefix at the start of your spec |
| 07:47 | <hsivonen> | tantek: yes |
| 07:48 | <tantek> | I think it shows insecurity in your own WG if you feel you must publish in another WG. |
| 07:48 | <othermaciej> | http://www.w3.org/2012/09/rdfa-wg-charter |
| 07:48 | <tantek> | I think it also shows insecurity in your spec if you feel you must hardcode it into a more popular spec in order to get adoption. |
| 07:48 | <danbri> | For a long while the two groups were too far apart |
| 07:48 | <hsivonen> | othermaciej: I think debating subheadings is premature until we have someone implementing a section level selector and has a good idea about the perf impact of various ways of masking stuff from the outline |
| 07:49 | <tantek> | danbri - which two groups? any two groups? ;) |
| 07:49 | <danbri> | at least they're not heading in much closer directions (rdfa more clearly detached from xhtml-isms, etc) |
| 07:49 | <othermaciej> | yeah, in retrospect it was probably a mistake to implement support for <hgroup> without implementing any of its relevant semantics |
| 07:49 | <danbri> | rdfa/html |
| 07:49 | <danbri> | re marketingisms, webgl seems to be perceived as 'part of' html5 by a lot of people; even if much of the speccing happened elsewhere |
| 07:49 | <tantek> | danbri, wasn't that just a side effect of staying on the XHTML2 bandwagon too long? |
| 07:50 | <tantek> | and btw - good to see you here in #whatwg |
| 07:50 | <danbri> | yes, but also understimating how much work converging w/ html would be |
| 07:51 | <tantek> | I think another large forcing function has been the massive developer preference for consuming JSON over all other syntaxes. |
| 07:51 | <danbri> | thanks :) I lurk in a lot of irc channels but usually just read |
| 07:53 | <hsivonen> | Speaking of extension specs: glazou asked about supporting epub:type in the validator. epub:type is like ARIA landmarks but from a different group and, hence, under a different (namespaced) name to comply with Conway's Law. I wonder what the benefit/harm considerations would look like for an XHTML5+EPUB option on Validator.nu. |
| 07:55 | <hsivonen> | epub:type degrades gracefully if it leaks onto the Web, but there's other EPUB stuff that won't degrade gracefully if it leaks to the Web. |
| 07:57 | <hsivonen> | Also, namespacing the attributes means they won’t have DOM Consistency if/when IDPF decides to adopt the text/html serialization of HTML |
| 07:57 | <hsivonen> | as for implementations, so far I’m only aware of iBooks supporting one particular pair of epub:type values |
| 07:58 | <hsivonen> | I have no idea if the rest are supported by anyone |
| 07:59 | <hsivonen> | EPUB really needs its equivalent of caniuse.com. I have no idea how real any of the EPUB extensions to either XHTML or CSS are |
| 08:06 | <othermaciej> | I wish IPDF used epub-type instead of epub:type and so forth; or better yet managed to avoid extending [X]HTML5 |
| 08:10 | <hsivonen> | or had folded their landmarks into ARIA landmarks as I, IIRC, suggested at TPAC 2 years ago |
| 08:12 | <Ms2ger> | He. He. He. |
| 08:14 | <annevk> | hsivonen: does iBooks and such enforce XHTML? |
| 08:14 | <hsivonen> | annevk: no idea |
| 08:14 | <annevk> | cause you know, people told us mobile was XHTML too |
| 08:14 | <annevk> | they lied |
| 08:16 | <jgraham> | I really don't think I should find the word "landmarks" as funny as I do. |
| 08:16 | <annevk> | jgraham: think less, enjoy more? |
| 08:16 | <jgraham> | But in common usage it's not really used in the sense of "things you would navigate by" |
| 08:16 | <jgraham> | More "tourist attractions" |
| 08:17 | <jgraham> | Or, rather, in its commenest usage |
| 08:17 | <jgraham> | *comnmonest |
| 08:17 | <jgraham> | Oh for fucks sake |
| 08:19 | <tantek> | epub:type huh - what's its use case? |
| 08:21 | <hsivonen> | tantek: Someone who should know claims on the Internet that it can be used in iBooks to make links to footnotes causes a footnote to appear as a pop-up instead of causing navigation to the footnote. As for the other values, I can make guesses about use cases but I don't really know. |
| 08:22 | <hsivonen> | s/causes/cause/ |
| 08:22 | <tantek> | oh it's how epub makes up their own element names and shoe horns them into HTML - http://idpf.org/epub/vocab/structure/ |
| 08:23 | <hsivonen> | the more landmarkish values seem redundant with OPF <guide> |
| 08:24 | <tantek> | here's a nice one: http://idpf.org/epub/vocab/structure/#pagebreak |
| 08:24 | <tantek> | because that's semantic after all and has nothing to do with presentation ;) |
| 08:24 | <hsivonen> | tantek: is that an idea, a draft or a spec? I haven’t seen that doc before when navigating from the front page |
| 08:24 | <tantek> | hsivonen - found referenced from here: http://matt.garrish.ca/epub3/guidelines/content/semantics/epub-type.php |
| 08:25 | <tantek> | hsivonen I'd say neither draft nor spec but yet another semantic registry |
| 08:25 | <hsivonen> | tantek: the page break thing is for correlating epub positions to print page numbers at a book club or in class |
| 08:26 | <hsivonen> | so it’s not crazy. Amazon automates the generation of that data for Kindle files. |
| 08:27 | <tantek> | nah, it's pretty crazy |
| 08:27 | <tantek> | reminds me of the Palm WebOS extensions to HTML |
| 08:27 | <tantek> | back when that was happening |
| 08:28 | <tantek> | hsivonen, "make links to footnotes causes a footnote to appear as a pop-up instead of causing navigation to the footnote" sounds presentational and something that could/should be added to CSS. |
| 08:29 | <hsivonen> | I find it remarkable that Amazon refuses to ship an EPUB reader for their devices, but one of the main paths for publishing for Kindle is preparing an EPUB file an running it through an Amazon-supplied converter. |
| 08:29 | <hsivonen> | tantek: has any iBooks stuff resulted in proposals to the CSS WG? |
| 08:29 | <tantek> | good question for hober |
| 08:30 | <tantek> | why bother with an EPUB reader for Kindles when Kindles already have a web browser of sorts |
| 08:30 | <hsivonen> | to avoid device/store lock-in |
| 08:31 | <jgraham> | hsivonen: I presume that's about funneling as many sales to kindle devices via amazon.com as possible |
| 08:32 | <jgraham> | +book |
| 08:32 | <hsivonen> | of course. Giving the converter to publishers but not automating conversion on the device for users is Amazon showing their corporate middle finger to their users. |
| 08:33 | <hsivonen> | why doesn’t the DOJ look into *that*? |
| 08:36 | <jgraham> | Is it illegal? I mean I agree it is obnoxious, but no more so than - for example - Apple's "no competing software" policy or Microsoft's "Microsoft-only API" policies (on windows 8/ARM) |
| 08:38 | <zcorpan> | so the html wg still hasn't published eh? the drafts were supposed to be ready and done on 15th last month |
| 08:38 | <zcorpan> | not that i'm surprised |
| 08:39 | <tantek> | publishing is hard, let's create process |
| 08:40 | <zcorpan> | publishing should be automated |
| 08:41 | <zcorpan> | speaking of which, what's the plan with FSAs? are we waiting on something? |
| 08:41 | <zcorpan> | Hixie: ^ |
| 08:42 | <jgraham> | Finite State Automotons? |
| 08:42 | <tantek> | you have to spend your FSA by the end of the calendar year or else it disappears |
| 08:42 | <jgraham> | Free Software Allowance? |
| 08:43 | <tantek> | :) |
| 08:44 | <zcorpan> | Fish Spawning Aggregation |
| 09:41 | <MikeSmith> | hsivonen: it took me at least 50 minutes to dissect that part of the spec to write that regexp :-) |
| 09:42 | <MikeSmith> | hsivonen: if you think it'd be useful for ease of maintenance to split it up, I can do that |
| 09:45 | <MikeSmith> | hsivonen: also fwiw I don't know that the use case is for allowing years of less than 4 digits there |
| 09:46 | <MikeSmith> | nor do I know what the use case is for allowing bare timezone offsets |
| 09:47 | <MikeSmith> | but the spec allows those |
| 09:47 | <MikeSmith> | "A valid time-zone offset string" is item 6 in the list at http://www.whatwg.org/specs/web-apps/current-work/multipage/text-level-semantics.html#datetime-value |
| 09:49 | <MikeSmith> | hsivonen: about the epub stuff, you know they are maintaining their own validator |
| 09:50 | <MikeSmith> | and epub books are not Web content |
| 09:50 | <MikeSmith> | oh maybe I'm confused about what epub extensions are |
| 09:51 | <MikeSmith> | if they are attributes that are meant to be added to web content instead of epub books, I guess I'd really wonder what the use case is |
| 09:54 | <MikeSmith> | annevk, hsivonen : HTML files in epub 3.0 books are required to be well-formed XML XHTML |
| 09:55 | <MikeSmith> | I briefly tried to get that epub wg to not make that a requirement, but I lost |
| 09:55 | <MikeSmith> | some people in that group were insisting on it as a requirement because they said there was no way to validate them otherwise |
| 09:56 | <MikeSmith> | which is wrong of course but oh well |
| 09:57 | <hsivonen> | MikeSmith: I’m OK with a single regexp if you can track the capture groups right. :-) the code for that is still to be written, right? |
| 09:58 | <MikeSmith> | well I saw that question in your e-mail message but I have to admit I don't understand why we need to track the capture groups |
| 09:58 | <hsivonen> | MikeSmith: I’m aware that there is such a thing as epubcheck, but I don’t know what exactly it checks |
| 09:58 | <hsivonen> | MikeSmith: as I understood it, glazou didn’t suggest using epub:type on the Web |
| 09:59 | <MikeSmith> | we already had a monster regexp in that code but were not tracking the capture groups in other parts of the code as far as I can see |
| 09:59 | <MikeSmith> | hsivonen: OK |
| 09:59 | <hsivonen> | MikeSmith: the superclass is trying to track the capture groups |
| 10:00 | <MikeSmith> | hsivonen: I don't know what epubcheck is checking either but I know they forked the vnu schema and are using that -- without the HTML5 datatype library (the replaced all of them with regexp patterns in the schema itself) |
| 10:00 | <MikeSmith> | hsivonen: ah OK |
| 10:00 | <MikeSmith> | OK, will address that then |
| 10:01 | <MikeSmith> | hsivonen: fwiw, epub schema is at http://epubcheck.googlecode.com/svn/branches/epub3/com.adobe.epubcheck/src/com/adobe/epubcheck/schema/30/mod/html5/ |
| 10:01 | <hsivonen> | The other day, I was trying to find some test EPUB files and got worried because I didn't find them. So either I'm bad at searching, EPUB test cases are secret or, worst case, there isn't a proper test suite. |
| 10:01 | <MikeSmith> | of course that schema can't recognize data-* attributes |
| 10:02 | <MikeSmith> | I don't think they have a proper test suite |
| 10:02 | <MikeSmith> | I've not seen it anywhere at least |
| 10:02 | <hsivonen> | they broke fantasai’s indent style :-( |
| 10:03 | <MikeSmith> | they broke a lot of things in that schema… |
| 10:04 | <MikeSmith> | I met with Murata-san to talk about it before they forked it. He didn't want to do it without talking about it first. But I just told him it's free software and they're welcome to do what they please with it |
| 10:07 | <hsivonen> | oh, breaking the indent style is permitted by the license, sure |
| 10:08 | <hsivonen> | MikeSmith: anyway, given that the IDPF schema is relaxed, I think it’s not unreasonable for glazou to be interested in Validator.nu + the EPUB stuff |
| 10:09 | <MikeSmith> | yah true |
| 10:10 | <hsivonen> | some day, I’m going to make some test books to learn the truth about EPUB readers |
| 10:11 | <MikeSmith> | I wish the whole thing would move to actually putting books on the Web at URLs |
| 10:12 | <hsivonen> | yeah, though I kinda sympathize with the desire to make it clear what the bounds of a book are for caching |
| 10:12 | <hsivonen> | and I sympathize with the desire to let users hold copies of their books forever |
| 10:13 | <hsivonen> | though DRM pretty much reverses that benefit |
| 10:13 | <hsivonen> | also, to read it private |
| 10:13 | <hsivonen> | again, foiled by DRM |
| 10:20 | <MikeSmith> | the DRM issue muffs up everything about the user experience |
| 10:21 | <MikeSmith> | it effectively restricts you to only being able to read your purchased books in the reader provided by the vendor that sold you the book |
| 10:22 | <MikeSmith> | and it gives the vendors almost no incentive to compete for quality in their readers |
| 10:49 | <hsivonen> | MikeSmith: it appears that almost everyone except Apple and Amazon is Adobe’s OEM basically |
| 10:58 | <MikeSmith> | hsivonen: yeah I guess it's like that |
| 11:41 | <annevk> | MikeSmith: well mobile was required to be XHTML too |
| 11:41 | <annevk> | MikeSmith: doesn't mean a) it is and b) that clients actually enforce |
| 11:55 | <MikeSmith> | annevk: yeah |
| 12:00 | <darobin> | I strongly doubt that it's enforced |
| 12:01 | <darobin> | book readers are slow machines, you don't want to do anything other than feed the content straight to the rendering engine |
| 12:01 | <darobin> | which is normally an HTML rendering engine |
| 12:01 | <annevk> | no man, XML is faster!!!! |
| 12:02 | <annevk> | how they mighty have fallen |
| 12:02 | <annevk> | the* |
| 12:02 | <MikeSmith> | well their loss -- they're missing out on the benefits of catch fire and fail |
| 12:02 | <MikeSmith> | ask Tim Bray |
| 12:02 | <hsivonen> | darobin: is Adobe’s SDK an HTML engine? |
| 12:02 | <MikeSmith> | if I have an e-book reader in my pacemaker I certainly don't want it to be using loosy-goosey unlclean tag soup parsing |
| 12:03 | <hsivonen> | AFAICT, pretty much every non-Amazon e-Ink device embeds Adobe’s engine |
| 12:03 | <darobin> | hsivonen: I haven't looked closely, but last I heard they had somethin WebKit based, no? |
| 12:03 | <hsivonen> | darobin: I have no idea what Adobe’s SDK is based on |
| 12:03 | <darobin> | my understanding from talking to book people is that it's all webkit |
| 12:04 | <darobin> | Amazon used to use NetFront but they thankfully ditched that |
| 12:04 | <darobin> | annevk: they should really use EXI — now *that* is faster :) |
| 12:05 | darobin | only half joking, it would actually make some sense to use exified HTML on that level of dumb device |
| 12:05 | <hsivonen> | Linux on Freescale ARM SoC is not that dumb |
| 12:06 | <darobin> | hsivonen: I reckon that a good way of assessing what works in epub world would be to grab a bunch of books from one of those book pirate teams |
| 12:06 | <darobin> | because they tend to produce stuff that has been properly corrected to actually be interoperable |
| 12:06 | <darobin> | hsivonen: the ipad isn't so dumb either, but many of those devices are pretty slow |
| 12:07 | <darobin> | the K3 can barely keep up with its screen :) |
| 12:44 | <jgraham> | Isn't "as soon as someone writes a js library for it" to obvious answer to "when will we see a filesystem library on top of IndexedDB" |
| 12:44 | <jgraham> | ? |
| 13:39 | <darobin> | does whatwg.org have some kind of protection against spidering? |
| 13:41 | <darobin> | same for html5.org |
| 13:41 | <darobin> | I'm running link checking on the spec, and those both seem to just drop the connection |
| 15:10 | <odinho> | There's an Opera survey and they want different people to answer, anyone in this channel is obviously qualified :P -> http://fluidsurveys.com/s/operasurvey/ |
| 15:27 | <Lachy> | odinho, where was that survey announced? I don't see it mentioned in any internal mail. |
| 15:28 | <odinho> | Lachy: http://my.opera.com/chooseopera/blog/2012/10/04/help-opera-improve-its-products-and-services |
| 15:28 | <odinho> | Lachy: To be honest I just talked with the guy who made it randomly on the bus. |
| 15:29 | <Lachy> | ok |
| 15:29 | <Lachy> | might be more useful to have it posted on Opera's twitter account |
| 15:31 | <odinho> | Lachy: Ah, they prolly will. I would've just thought that whatwg would probably be forgotten, so I took it upon myself to fix that :P |
| 15:32 | <Lachy> | The question about industry partners is asking a bit too much for any useful responses, as it includes too many vastly different groups, of which Opera's need and ability to establish connections with varies greatly. |
| 15:33 | <odinho> | Lachy: Uh, you're an employee of Opera, you have a different survey. |
| 15:33 | <Lachy> | I'm not completing the survey. I just went through it with blank answers to take a look |
| 16:09 | <dglazkov> | good morning, Whatwg! |