2011-05-01 [18:21:00.0000] hello [19:05:00.0000] Hixie: OK, thanks [19:05:01.0000] I did already commit it that way yesterday [19:06:00.0000] the main reason I was asking was that the behavior's not totally consistent in that we don't currently do the same for some other cases where the spec provides some additional guidance [19:07:00.0000] for example, for Set-Cookie [19:08:00.0000] where we could emit "To set cookies, use HTTP headers instead." or whatever [19:11:00.0000] but anyway, it's a bigger priority to implement checks for all the actual document-conformance requirements in the spec [21:35:00.0000] abarth: yt? [21:47:00.0000] anyone got a webkit build handy? [21:48:00.0000] does http://software.hixie.ch/utilities/js/live-dom-viewer/saved/970 fail in safari/webkit too, or is it a chrome bug? [21:50:00.0000] Hixie, nothing changed colour for me in a semi-recent WebKit [21:50:01.0000] k thx [23:05:00.0000] Hixie: nothing changes in a fresh WebKit build - what's the bug in Chrome? [23:06:00.0000] /me is surprised that a Chrome-specific bug would even be possible here [23:06:01.0000] it focuses the [23:06:02.0000] yeah, me too [23:06:03.0000] hmm, let me try in full keyboard access mode in Safari [23:06:04.0000] (only happens if the output is display:block) [23:07:00.0000] it happens in Safari in Full Keyboard Access mode [23:07:01.0000] so it's just a bug in WebKit focusability rules [23:08:00.0000] it's hell weird that display: block on the matters [08:04:00.0000] AryehGregor: are you going to mention the things that were discussed in #whatwg earlier? <-- Which things that we discussed in #whatwg earlier? [08:04:01.0000] I plan to come up with as many specific examples of spec or spec organization forks that I can. [08:05:00.0000] The two that I know of offhand are the WHATWG splitting from the W3C, and the W3C being created because TBL had too many problems in the IETF, which are both excellent support for forking. [08:05:01.0000] Other people have mentioned an IETF fork of some version of HTML, what's the story on that? [08:05:02.0000] And does anyone know of other examples? [08:05:03.0000] One thing I wonder about is whether the W3C intends to take action against the ISO HTML, XHTML-MP, CHTML, XHTML Basic, XHTML 2, WML, WTVML, XHTML-Print, WHATWG HTML, or HDML forks of HTML [08:06:00.0000] Oh, it was ISO that forked it, not the IETF. [08:06:01.0000] Right. [08:06:02.0000] Surely XHTML 2 doesn't count, since that's a W3C-sponsored fork? [08:08:00.0000] I guess that depends on what the purpose of the ban on forks is [08:08:01.0000] CE-HTML [08:09:00.0000] and Open IPTV Forum "Declarative Application Environment V2.0" [08:09:01.0000] http://www.openiptvforum.org/specifications.html [08:09:02.0000] take a look a that spec some time - [08:09:03.0000] http://www.openiptvforum.org/docs/Release2/OIPF-T1-R2-Specification-Volume-5-Declarative-Application-Environment-v2_0-2010-09-07.pdf [08:10:00.0000] Heh, it references HTML5 [08:10:01.0000] Is anyone interested in helping with a wiki page to outline all the arguments for why forking is either harmless or good? [08:11:00.0000] That sounds like trying to prove a negative... [08:11:01.0000] No, you can cite all the existing forks and point out that they were all either harmless or good. [08:11:02.0000] That's evidence. [08:12:00.0000] Anecdotal evidence [08:12:01.0000] *-MP was not good. :P [08:12:02.0000] As opposed to no evidence, which is what the forking-is-bad crowd has. [08:12:03.0000] You can't expect randomized controlled double-blind trials for questions like "Is allowing spec forks a good idea?" You've got to work with what you have. [08:13:00.0000] I'll have only limited time to work on a real response to the survey unless Hixie decides it's part of my contract. [08:13:01.0000] (Why can't contracts just say stuff like "Do something useful and web standards-related"?) [08:13:02.0000] It's a mystery [08:13:03.0000] That seems to be the job description of some full-time employees. [08:14:00.0000] But really, several of the ones othermaciej listed are non-harmless in my opinion [08:14:01.0000] Interesting. Which ones? [08:14:02.0000] The ones building on XHTML, because they fragment the ecosystem and lead to a waste of time and effort when they fail [08:14:03.0000] They still probably don't strongly support the case of restrictive copyright licensing, given that a) they might not actually infringe copyright, and b) the W3C doesn't appear interested in suing even if they do. [08:15:00.0000] some I'm not sure totally count as forks [08:15:01.0000] So, like, XHTML 2? [08:15:02.0000] but XHTML-MP is pretty bad, as it sends content purporting to be XHTML but with different processing rules [08:15:03.0000] (really weird stuff too) [08:15:04.0000] CHTML is bad for similar reasons [08:16:00.0000] these are bad not only for browsers that implement them but also, when such content is posted, it creates non-interoperable worlds and makes it hard to successfully display all content [08:16:01.0000] WML is less bad in that regard since it has a distinct content type but I'm not totally sure if it is fair to call it a fork [08:17:00.0000] it certainly copies many of the elements in HTML [08:17:01.0000] But these are all basically unsuccessful, aren't they? [08:17:02.0000] They're only really harmful if they cause real problems, not if they would have caused problems had they been successful. [08:17:03.0000] there is a huge amount of XHTML-MP, CHTML and WML content out there [08:17:04.0000] Interesting. [08:17:05.0000] in east asia, there is huge market pressure for any mobile browser that gets released to support one or more of them [08:17:06.0000] I've never heard of any of them, I don't think. [08:17:07.0000] Oh, they're mobile things. [08:18:00.0000] Well, I'll briefly look at some of them, I guess. [08:18:01.0000] yes, mobile web stuff [08:18:02.0000] A lot of operators push for *-MP compliance, too. [08:18:03.0000] (Which sucks.) [08:18:04.0000] many of the forks from my list are W3C-sponsored but are nonetheless incompatible fors [08:18:05.0000] XHTML2 is much less damaging than the things I listed since no one uses it [08:18:06.0000] I don't think W3C forks are germane to the licensing question at hand. [08:19:00.0000] Since they'd be allowed under any license, presumably. [08:19:01.0000] none of these are germane to the licensing issue at hand [08:19:02.0000] Why not? [08:19:03.0000] none of them were prevented by the W3C Document License [08:19:04.0000] which does not allow forking [08:19:05.0000] That's evidence that the restrictive license is ineffective, which is germane. [08:20:00.0000] or to look at it another way, all of them are germane, since they show that anti-forking license provisions are not effective at preventing forking [08:20:01.0000] Right, exactly. [08:20:02.0000] /me has to go now, more later . . . if he can find the time [08:20:03.0000] the fact that the W3C itself forks is even more germane, since no license you could imagine could prevent that [08:21:00.0000] I'm glad we got those extra two licensing options up. I wonder how the results of the survey will be considered by the people whose business it is to consider them. [10:54:00.0000] Hixie: pong [10:54:01.0000] hmm, Canonical XML 2.0 [10:54:02.0000] dare I look [11:18:00.0000] AryehGregor: feel free to use a few of google's hours to work on a response that once and for all explains all the relevant issues here, but if you do then send me regular updates so i can give you more direct feedback [11:21:00.0000] abarth: if you are trying to hash passwords in a login db, then presumably you want to use MHAC with some algorithm like SHA-256 or whatever. but what key do you use? can you use the username as a key, or is that a Bad Idea? [11:22:00.0000] i bet there's some canonical explanation for how to do this [11:22:01.0000] but the general approach is to use a hash, not an HMAC [11:23:00.0000] with a public and a private salt [11:23:01.0000] k [11:23:02.0000] so there's no good way to avoid a private salt? [11:24:00.0000] i was hoping to be able to do a "SELECT ... WHERE ... AND password=?" instead of selecting the row then comparing the password [11:24:01.0000] so that if the password is wrong, the code never even gets the userid [11:24:02.0000] it's not a huge deal if it's not possible [11:24:03.0000] the private salt is just there to slow down the hash function [11:25:00.0000] so you can avoid it if you pick a slow hash function [11:25:01.0000] the public salt is technically supposed to be per-password though [11:25:02.0000] to prevent the attack from attacking all the passwords at once [11:25:03.0000] with a dictionary [11:25:04.0000] s/attack/attacker/ [11:26:00.0000] Hixie, if your DBMS supports the relevant functions, you can do something like "WHERE ... password = SHA256(salt + 'password')", if the user table has a salt column that contains the per-user salt. [11:27:00.0000] Of course, that's not indexable, so it won't scale to many users. [11:28:00.0000] Why would you want to look up users by their password anyway, though? [11:28:01.0000] Other than to detect weak passwords, maybe, in which case it's obviously going to be O(N) and indexing doesn't make much of a difference one way or the other. [11:29:00.0000] (It'd be indexable in Postgres, I think) [11:29:01.0000] Philip`, how? [11:29:02.0000] Postgres supports indexes on expressions, but what expression could you index here? [11:29:03.0000] (unless I'm missing understanding, which I probably am) [11:29:04.0000] In "WHERE ... password = SHA256(salt + 'password')", password and salt are both DB columns. [11:30:00.0000] And the string 'password' is provided literally by the application for the specific query. [11:31:00.0000] I think I was reading the literal and column the other way around [11:31:01.0000] (which would be silly) [11:39:00.0000] abarth: yeah, hence my question about using the username as the public salt [11:41:00.0000] AryehGregor: the actual query would be WHERE username=? AND password=? [11:41:01.0000] AryehGregor: so that it returns nothing when the password is wrong [11:41:02.0000] Hixie, oh, I see. [11:41:03.0000] that seems fine [11:41:04.0000] AryehGregor: let me see if i can use a SHA256() function [11:42:00.0000] but this isn't my area of expertise [11:42:01.0000] abarth: k [11:42:02.0000] Ideally your public salt should be long enough that you foil rainbow table attacks. [11:42:03.0000] For that purpose, it's probably best to make it a not-too-short randomish string. [11:42:04.0000] looks like sqlite3 doesn't have hash-related functions [11:43:00.0000] wonder if i can add custom functions from perl... [11:43:01.0000] Is this a WebSQL use-case? [11:43:02.0000] no it's just a toy i'm playing with at home [11:44:00.0000] I don't actually get the use-case at all, though. If the attacker can query the DB, why don't they just do "SELECT * FROM user"? Because they only have the right to execute prepared statements? [11:44:01.0000] ho ho ho, you _can_ add functions in perl [11:44:02.0000] that will make my life easier [11:44:03.0000] How does that even make sense with SQLite, though? If you can query an SQLite database, surely you have read access to the database file? [11:45:00.0000] AryehGregor: the concern is highly theoretical and mare an issue of having clear boundaries in code, but to the extent that it is a concern, the theoretical attack is someone who can cause the perl code to leak data from its environment [11:46:00.0000] AryehGregor: if someone is running their own arbitrary code, then yeah, they can read the db, at which point it's why we're hashing it [11:47:00.0000] So like someone who can read the contents of your Perl program's memory, but can't control it, and you're worried about them reading the userid in the brief window before it's garbage-collected or something . . . ? [11:47:01.0000] If so, why don't you first retrieve the hash and check if it matches, and only then query the id? [11:48:00.0000] Given that SQLite doesn't run in a separate process, if they can read the Perl program's memory then they can read the raw database data, probably [11:49:00.0000] That too. [11:49:01.0000] AryehGregor: my desire does not a priori stem from a security concern [11:50:00.0000] AryehGregor: it's about keeping clear boundaries in the code about what code gets to see what [11:50:01.0000] for readability [11:51:00.0000] k. [11:52:00.0000] and it gives me an opportunity to play with this function creation thing :-) [11:54:00.0000] Injecting custom Perl functions into SQLite so they can run inside SQL queries is meant to *increase* readability? :-) [11:55:00.0000] hehe [11:55:01.0000] actually in this case i think it will [11:55:02.0000] though i agree with your skepticism :-) [11:57:00.0000] /me wonders if there's a Perl module equivalent to http://apidoc.apsw.googlecode.com/hg/vtable.html (where you can redefine the whole table storage mechanism in a scripting language) [12:29:00.0000] Hixie et al., I started a page on forking on the WHATWG wiki: http://wiki.whatwg.org/wiki/Forking [12:29:01.0000] If anyone wants to fill it out some more, please do. [12:29:02.0000] I don't know if I'll work on it further. [12:30:00.0000] (I'll probably work on improving my personal survey response later on my own time) [13:07:00.0000] how long until someone forks the page on forking [16:52:00.0000] can someone explain why Lawrence Rosen keeps talking about patents when we're trying to discuss the copyright license? [16:52:01.0000] isn't he a lawyer? you'd think he'd know better that to try to confuse these unrelated matters. 2011-05-02 [17:11:00.0000] AryehGregor: i updated the wiki page with some recent arguments [17:11:01.0000] http://wiki.whatwg.org/wiki/Forking [18:30:00.0000] ohai [18:30:01.0000] about the cache manifest spec [18:32:00.0000] may I suggest that the manifest spec include another field for each vile, specifying the last-modified date of the file, so that the client can auto-update the file if it's been modified since the cache was stored? [21:26:00.0000] hmm [21:27:00.0000] the spec doesn't say anything about what the allowed value of the mediagroup attribute is [21:29:00.0000] clearly it's not meant to be a list of tokens [21:29:01.0000] so any string? [21:41:00.0000] hmm, I guess if no additional constraints are specified for any particular attribute, by default it's just text [21:41:01.0000] or text and character references [01:37:00.0000] "Whether the W3C allows forking or not, and even if we grant that a license could stop this, this scenario could already happen because the entire HTML spec is already under a liberal license at the W3C." http://wiki.whatwg.org/wiki/Forking - s/W3C./WHATWG./ ? [02:00:00.0000] I think Lawrence Rosen only likes the Option 1 and Option 3 licenses because he came up with them [02:10:00.0000] zcorpan: I think the right response is "it's a wiki". But I noticed the same mistake and didn't fix it so… [02:12:00.0000] fixed [02:18:00.0000] hmm [02:19:00.0000] I'm not sure it's right to call either CHTML or HDML "WC-hosted forks" [02:19:01.0000] A reasonable response to "there is nothing special about W3C sanctioned forks" might be that non-W3C forks aren't covered by the same IP commitments as W3C forks; so W3C forks are equivalent from a brand-dilution, confusion, fragmentation point of view, but not from a "people injecting stuff and charging for it later" point of view [02:19:02.0000] they were both originally developed outside the W3C, published as member submission notes, and then rejected and developed outside the W3C [02:19:03.0000] (afaik) [02:28:00.0000] (I don't know if there is a good response to that other than "the risk is outweight by the positive effects that the ability to fork brings") [02:29:00.0000] whaoh [02:29:01.0000] never heard of this before, HTML 4.0 Mobile: http://www.w3.org/TR/1999/NOTE-html40-mobile-19990315/ [02:29:02.0000] (has its own DTD and everything) [02:30:00.0000] also, XHTML+SMIL Profile: http://www.w3.org/TR/2002/NOTE-XHTMLplusSMIL-20020131/ [02:31:00.0000] (not entirely sure if that is an HTML/XHTML fork) [02:31:01.0000] Ah, another entry from the graveyard "mobile will never be able to browse the real web" [02:31:02.0000] +of [02:32:00.0000] it seems like the W3C forks HTML more than the rest of the world put together [02:32:01.0000] othermaciej: Interesting. A design goal of XHTML and XHTML2 was modularity so that people could take parts of specs and recombine them into a different form. That seems a lot like forking [02:32:02.0000] the whole XHTML Modularization thing? [02:33:00.0000] Yeah [02:33:01.0000] I do not know what that is all about [02:33:02.0000] I am not even sure whether XHTML 1.1 could reasonably be called an incompatible fork of 1.0 [02:34:00.0000] "This modularization provides a means for subsetting and extending XHTML, a feature needed for extending XHTML's reach onto emerging platforms" [02:34:01.0000] http://www.w3.org/TR/xhtml-modularization/ [02:34:02.0000] That sounds a lot like "forking" to me [02:35:00.0000] Albeit that you were probably not supposed to change the modules you reused. But you could add or subtract modules [02:35:01.0000] I guess advocates of this approach would distinguish reusability and extensibility from forking [02:35:02.0000] though forkability is really just reusability with arbitrarily fine granularity [02:35:03.0000] I don't see the practical difference [02:36:00.0000] In particular I don't see the difference from the point of view of IP [02:36:01.0000] restricting the granularity of reuse may encourage greater pockets of interoperability (in theory) [02:37:00.0000] it's funny to me that opinions on allowing forking per license, and on allowing free extension via "distributed extensibility" seem to be anticorrelated [02:37:01.0000] Yeah, that seems like a very theoretical outcome [02:37:02.0000] with the exception of notable individuals like Sam Ruby, who favors allowing both, which strikes me as a laudably consistent set of views [02:38:00.0000] (note: I mean that sincerely, not trying to be snarky in any way) [02:38:01.0000] jgraham: wow. indeed the whole Modularization thing is interesting in the context of being scared of licenses that allow forking [02:38:02.0000] jgraham: I had forgotten about Modularization [02:38:03.0000] hsivonen: Lucky you ;) [02:38:04.0000] jgraham: :-) [02:38:05.0000] othermaciej: I'd never hear of HTML 4.0 Mobile either [02:38:06.0000] I am amazed at how many different HTML forks have been published in some form by the W3C [02:38:07.0000] HTML 4.0 Mobile looks like a proto-"Mobile Web Best Practices" [02:39:00.0000] othermaciej: I don't think distributed extensibility and forking have much to do with each other really. At least the main arguments in favour of allowing forking aren't the same as the main arguments people use for DE [02:39:01.0000] that's not even counting the mainline successor specs, even though you could argue that XHTML1 is an incompatible fork of HTML4, and perhaps likewise for XHTML 1.1 [02:39:02.0000] hmm, the did actually publish a HTML 4.0 Mobile DTD - HTML 4.0 Mobile [02:39:03.0000] http://www.w3.org/TR/1999/NOTE-html40-mobile-19990315/DTD/html40-mobile.dtd [02:39:04.0000] HTML5 plausibly is a successor to HTML4 and XHTML1 so I'll give the W3C a pass on that [02:40:00.0000] MikeSmith: yeah, I thought it might be a "best practices" but it has its own DTD [02:40:01.0000] yeah [02:40:02.0000] othermaciej: btw, you are right about CHTML and HDML, afaik -- I don't think they ever had any organizational support from any standards body at all [02:41:00.0000] I mean, I was not around at the time [02:41:01.0000] but from what I can tell, they were published in a totally unofficial way [02:42:00.0000] oh hey, another fork, the html4all "HTML 4.1" draft: http://html4all.org/HTMLDraft.html [02:44:00.0000] I think crockford has its own html5 fork [02:44:01.0000] FWIW, developers.whatwg.org is also technically a fork [02:44:02.0000] crockford's draft is vaporware [02:44:03.0000] or vapor-spec [02:44:04.0000] espadrine, where is crockford's draft? [02:45:00.0000] http://www.crockford.com/html/ [02:45:01.0000] Lachy: no idea, but he said in a talk I saw that he did have it! [02:45:02.0000] /me has CNN on in the background going on about Bin Laden [02:46:00.0000] MikeSmith: "That's it." Whaa, so soon? [02:46:01.0000] I think calling Crockford's effort a fork is insulting to all serious forkers everywhere [02:46:02.0000] "No more framesets, frames, or iframes. The security properties of these were problematic. Instead we'll have modules." [02:47:00.0000] "The only character encoding permitted in HTML 5 is UTF-8." <-- I like that part. [02:48:00.0000] so I added a section for "W3C-published but disavowed forks" [02:48:01.0000] I am not sure if XHTML2 should go in that section or in "W3C-hosted forks" [02:49:00.0000] did the ex-XHTML2 folks ever follow through on their plan to continue development outside the W3C? [02:50:00.0000] (FWIW I think that wiki page would be better without all the nonsense about governments creating incompatible forks to restrict access to the web. There is only ~1 person who can possibly seriously believe that) [02:50:01.0000] othermaciej: not that I've heard [02:51:00.0000] That wiki page you're working on seems to be using a strange definition of forking. Many of those specs listed are not forks of previous specs, but rather newly created versions. [02:52:00.0000] does JHTML count as an HTML fork: http://en.wikipedia.org/wiki/JHTML [02:52:01.0000] Lachy: mostly they fork the language without textually forking a previous spec document [02:52:02.0000] but then, I guess that just proves that the W3C's fear of forks that could be prevented by stricter copyright are a very limited subset. [02:53:00.0000] or how about FBML [02:53:01.0000] ok [02:54:00.0000] then I think it may be useful to more clearly distinguish between language forks and specification forks. The former would include ISO-HTML, EPUB, etc. The latter would include specs like developers.whatwg.org [02:55:00.0000] by that standard, you could claim the HTML5 author view is a spec fork [02:55:01.0000] it is [02:55:02.0000] but indeed, it is possible to fork the spec text without changing any normative requirements [02:55:03.0000] and it would be prevented by the W3C's strict licence [02:55:04.0000] Lachy: I think the point is that the supposed risks of forking are independent of whether one rewrites or one juct changes the existing text [02:56:00.0000] it's also an example of why the W3C should permit forking [02:56:01.0000] the dangerous part isn't about forking, it is the intent that the forker has [02:56:02.0000] if he just wants to confuse people, it will confuse people [02:56:03.0000] jgraham, yes, but it's also useful to show the examples of forking that are beneficial, in order to illustrate the risk of preventing forks [02:57:00.0000] I can't find a spec for FBML that makes clear whether it is actually an HTML fork or not [02:57:01.0000] othermaciej: FBML was abandonded so I think it might be hard to find documentation [02:57:02.0000] Wikipedia says it is a customized version of HTML with extra tags, but I can't verify that with facebook's docs [02:58:00.0000] They have now forked RDFa instead... [02:58:01.0000] (in the sense that their processing requirments are, by design, entirely different to those in the RDFa spec) [02:58:02.0000] fbml is still used [02:58:03.0000] e.g., for the like button [02:59:00.0000] Lachy: I agree [02:59:01.0000] http://developers.facebook.com/docs/reference/plugins/like/ [02:59:02.0000] [02:59:03.0000] Does it count as forking if you use colons? :) [03:00:00.0000] that i can't tell you [03:00:01.0000] but fbml uses lots of colons [03:00:02.0000] that looks more like "distributed extensibility" [03:00:03.0000] colons are definitely distributed [03:00:04.0000] it adds a bunch of tags with colons in them [03:00:05.0000] Although, having said the arguments in favour of the two are totally different, the arguments against forking also work against DE [03:00:06.0000] what I am not sure of is whether you mix those with HTML, or use only the funky fb: tags [03:00:07.0000] it also changes the semantics of existing tags [03:00:08.0000] like ? [15:15:00.0000] Hixie: hit upload [15:15:01.0000] done [15:15:02.0000] log: undefined [15:15:03.0000] thanks [15:16:00.0000] this was ie9 btw [15:18:00.0000] cool, thanks [15:19:00.0000] np [15:36:00.0000] Hmm, Opera 11.10's HTML serializer seems to be broken. When will it upgrade to an HTML5 serializer? [15:37:00.0000] (specifically, it looks like
    serializes as
      ) [15:37:01.0000] (which puts the
        inside the
      1. instead of as a sibling) [15:37:02.0000] doesn't ie do that as well? [15:37:03.0000] AryehGregor: I expect that will work once Ragnarok lands for real [15:39:00.0000] zcorpan, it does omit end tags in some cases, but it seems to get this case right (didn't test carefully). [15:39:01.0000] jgraham, . . . which means what? [15:40:00.0000] when it's ready, pretty much [15:40:01.0000] AryehGregor: I think this is the time that I pull the Apple line about not discussing future products or services [15:41:00.0000] Aw. [15:41:01.0000] Meanie. [15:41:02.0000] s/Apple/Google/ maybe [15:41:03.0000] AryehGregor: we'll serialize html correctly no sooner than when we parse html correctly (and likely not later either) [15:41:04.0000] AryehGregor: Anyway, basically what gsnedders said [15:41:05.0000] I was figuring that much, yeah. [15:43:00.0000] abarth: yt? [15:43:01.0000] (we couldn't do it sooner because it wouldn't work right with our current DOM tree lies about where elements end) [15:43:02.0000] The codename “Ragnarok” might give a clue about the time of its arrival. (c: [15:51:00.0000] Wow, we've gotten a huge number of eleventh-hour responses to the survey. [15:56:00.0000] Two different Microsoft employees responded [15:56:01.0000] (at least) [15:56:02.0000] which is unusual [15:59:00.0000] Oddly, it gives the e-mail addresses of non-respondents but not respondents. [16:01:00.0000] I'm counting Adrian Bateman, Eliot Graf, Kris Krueger, Frank Olivier, and Cynthia Shelly. [16:01:01.0000] Speculation: Adrian posted the official answer, and sent out a memo to other MS people to agree with the party line. [16:02:00.0000] (Did any of them disagree?) [16:02:01.0000] Are you implying a Microsoft Cabal? [16:02:02.0000] Unsurprising, since this is an individual vote rather than organizational. [16:02:03.0000] I don't think it's fair to call a publicly-traded corporation a "cabal". [16:03:00.0000] That carries some implications of secrecy that are unlikely to apply. [16:05:00.0000] But then there's no need to speculate about cabal-like activities [16:05:01.0000] Well, right, granted. [16:06:00.0000] heh in old opera 9.64 win.onmessage does not works but win.addEventListener('message' is ok [16:09:00.0000] MikeSmith: the copyright notice has a link to http://www.ercim.org/ which redirects to http://www.ercim.eu/ [16:09:01.0000] MikeSmith: should i change the link to http://www.ercim.eu/ ? [16:11:00.0000] AryehGregor: I guess the saving grace is that these "votes" are not actually votes [16:11:01.0000] There are tallies, though. [16:15:00.0000] In theory the analysis isn't "most votes wins" I think [16:16:00.0000] Not exactly, no, but if it has any impact at all, I expect the tallies to be a large part of it. [16:17:00.0000] I mean, almost no one *said* anything in their response other than "we like forking" or "we don't like forking". [16:17:01.0000] It's not like you can get much out of the survey other than the tallies. [16:18:00.0000] You could infer that there isn't consensus to adopt any of the PSIG licenses for any reasonable definition of consensus [16:19:00.0000] Well, you can get that out of the tallies. [16:21:00.0000] Yes, I'm not saying that the prose is very valuable [16:21:01.0000] http://www.w3.org/Bugs/Public/show_bug.cgi?id=11452#c9 is fun [16:22:00.0000] I doubt anyone said anything that will convince anyone else [16:22:01.0000] AryehGregor: did anyone say "we like forking"? Not "we like the ability to fork"? I doubt anyone actually likes forking per se... [16:22:02.0000] Hixie, I was using it as shorthand. [16:22:03.0000] oh ok [16:24:00.0000] is ★ a non-BMP character? I think we have problems with non-BMP characters in file names [16:26:00.0000] U+2605 - BMP [16:27:00.0000] 9733 [16:27:01.0000] definitely BMP [16:28:00.0000] ☃ is still BMP ... o_O [16:28:01.0000] Oh well it must be a differnet bug then [16:29:00.0000] opera failed hard on that test [16:29:01.0000] it let me pick the file and then said "what file? there's no file here" [16:32:00.0000] http://www.w3.org/Bugs/Public/show_bug.cgi?id=12611 - clearly html-differences needs a proper review [16:33:00.0000] Hixie: there is no spoon [16:38:00.0000] Philip`: let me know if you make changes to the splitter [16:56:00.0000] zcorpan: it's OK to change that ERCIM URL if you want, but you it's not a requirement [16:57:00.0000] zcorpan: as far as the content, there are a few other recent changes I can think of that you should list [16:58:00.0000] e.g., new mediagroup attribute [16:58:01.0000] MikeSmith: pls file bugs :) [16:58:02.0000] OK [16:58:03.0000] I was just going to ask if you wanted me to do that :) [16:59:00.0000] http://www.w3.org/Bugs/Public/show_bug.cgi?id=12223 - anyone care enough to get rel=help nuked? 2011-05-06 [17:05:00.0000] zcorpan: well, it's on the microformats "existing rel values" page, so it remains valid anyway [17:06:00.0000] which is true also for pretty much all the other link-relation names that have been dropped [17:06:01.0000] so from a user/author POV, it makes little difference in practice if it's dropped from the spec or not [17:06:02.0000] useless features should be moved to the invalid department imho [17:07:00.0000] but i don't feel strongly about it [17:07:01.0000] it's a wiki [17:07:02.0000] now time to sleep [17:07:03.0000] おやすみ [17:07:04.0000] don't make me translate that; i'm going to bed [17:22:00.0000] someone is wondering why there isn't a delta on onscroll events [17:25:00.0000] anyone know why? [17:28:00.0000] smaug____ may know [17:28:01.0000] jamesr: hmm. I'd say at least because of historical reasons [17:29:00.0000] jamesr: also, what kinds of values would you expect? [17:29:01.0000] and, scroll event happens after scrolling [17:30:00.0000] ...so I'm not quite sure what the delta would mean [17:30:01.0000] especially if the size of viewport changes while scrolling [17:30:02.0000] the difference in scroll offsets relative to the previous scroll event, i suppose [17:31:00.0000] the top/left offsets shouldn't depend on the viewport size [17:31:01.0000] just the possible max values for them [17:32:00.0000] at which point would you calculate offset to top/left? [17:33:00.0000] when user scrolls, or when the scroll event is dispatched? [17:33:01.0000] when the event dispatches [17:34:00.0000] if i scroll down 53px then the next scroll event would have scrollDeltaY = 53 (or -53, i can never remember which way the coord system goes) [17:36:00.0000] jamesr: and if you after that scroll say 10px, scrollDeltaY would be 63? [17:36:01.0000] or 10? [17:36:02.0000] the next scroll event would say 10 [17:37:00.0000] or i suppose each event could have the scrollTop/scrollLeft pair for the thing it was scrolling [17:37:01.0000] that would be tricky to implement, I think [17:37:02.0000] since layout may change during scrolling [17:37:03.0000] so how to decide that 10px was scrolled [17:37:04.0000] well, we have to do layout to scroll at all [17:38:00.0000] it would be easier to report offset to top/left [17:39:00.0000] but, we have that kinds of offset already elsewhere [17:39:01.0000] not sure i agree [17:39:02.0000] in webkit all the scrolls are internally done with deltas [17:40:00.0000] sure, in gecko, scrolling works with deltas too [17:42:00.0000] what would be the use case for adding offsets or deltas to scroll event? [17:49:00.0000] Seems like in gecko adding such delta would be quite easy. But since scroll event is async, I'm not quite sure how useful the delta would be. Perhaps there is some good use case ... [17:50:00.0000] some pages react to scroll events to lazy load parts of the DOM when they are close to scrolling into view, so currently they listen to scroll events and then in the handler have to figure out which way the scroll actually went [17:50:01.0000] there's buffering so it's ok if the event is later [17:52:00.0000] ah, that is a good use case [17:55:00.0000] i'm not totally sure why they want a delta instead of just querying scrollTop/scrollLeft in the handler - maybe that's just a PITA to keep track of [17:56:00.0000] is this some bug report where some web devs want a new feature? [17:58:00.0000] If such additional properties were to added to scroll event, I think cssom-view would be the right place to define that all [17:58:01.0000] an email, but yes [18:00:00.0000] jamesr: but don't they still need to check if the view port has been scrolled long enough before loading some new data to page? [18:00:01.0000] oh? do they cover dom events? [18:00:02.0000] yes, they internally use it to calculate their view offset thingy [18:00:03.0000] in that case delta might not be useful [18:01:00.0000] why's that? they listen to resize to detect viewport size changes [18:02:00.0000] I mean, perhaps having just some flag to indicate the direction might be enough [18:04:00.0000] Though, deltaX/Y are that kinds of flags with a bit more information [18:07:00.0000] so, the delta needs to be calculated when scrolling happens, not when the event is dispatched [18:16:00.0000] does firefox aggregate multiple user scrolls into one scroll event? [18:17:00.0000] or are they 1:!? [18:17:01.0000] 1:1* [18:19:00.0000] IIRC 1:1 [18:20:00.0000] /me tries to remember if there is some special case when handling Win7 touch screen panning [18:21:00.0000] ...there shouldn't be [18:25:00.0000] oh, smoothscrolling may change the default 1:1 [18:33:00.0000] TabAtkins, ap: who should i cc on bugs that affect for webkit's form controls implementation? [18:35:00.0000] (the bug in this case being http://www.w3.org/Bugs/Public/show_bug.cgi?id=11212 ) [18:41:00.0000] Kent Tamura [18:41:01.0000] maybe [18:41:02.0000] oh [18:42:00.0000] I meant if it's about HTML5 forms stuff [18:42:01.0000] /me looks at the bug [18:44:00.0000] it's about required="" on [19:02:00.0000] Hixie: yes, tkent is the right person for that. And I'll see the bug anyway, and might CC someone else [19:04:00.0000] oh, if it's a w3.org bug, I won't see it anyway, I as thinking of webkit.org [19:32:00.0000] On tables for layout: http://www.w3.org/Bugs/Public/show_bug.cgi?id=10963#c19 [00:38:00.0000] Hmm. is there some SEO scam that's based on translating Web pages into the mother tongue of the translator [00:39:00.0000] something like the translations actually being machine translations but the links to them give Google juice to a scam? [00:40:00.0000] so far, I've always permitted my pages to be translated and the translations seemed legitimate, but now I got a permission request that I don't know what to think of [00:40:01.0000] kinda sad, though, that I even have to consider the possibility of a scam [00:42:00.0000] hmm. googling around for the guy sure makes it look like the translations are all about building google juice for hosting-related domains [00:52:00.0000] hsivonen: there's some guy like that who has contributed translations of the CSS info hosted at W3C [01:03:00.0000] MikeSmith: I made this the first time I declined a translation permission [01:41:00.0000] hsivonen: if you have not read http://greggkellogg.net/2011/05/05/cme-and-the-semantic-web yet, it's worth reading [01:43:00.0000] [[ [01:43:01.0000] the fact is, though, that proprietary metadata formats are much simpler to implement and manage. According to a key opinion maker: “I bet the average developer can get a simple XML-based music metadata system up and running in less time than it would take to read the Music Ontology document. We can get most (all?) of the benefits of RDF through simpler means. [01:43:02.0000] ]] [01:44:00.0000] MikeSmith: thanks [01:45:00.0000] "a key opinion maker" sounds like weasel language, though [01:45:01.0000] heh [01:45:02.0000] indeed [01:45:03.0000] kind of suspicious [01:45:04.0000] but the piece overall seems pretty well-considered [01:48:00.0000] zcorpan: do you know if Opera has an public plan to implement CSP? [01:49:00.0000] i don't [01:49:01.0000] ok [02:02:00.0000] bet browsers aren't going to call all keyboard events LegacyKeyboardEvent http://dev.w3.org/2006/webapi/DOM-Level-3-Events/html/DOM3-Events.html#legacy-key-attributes [02:02:01.0000] /me will review more closely later [02:06:00.0000] MikeSmith: I commented on the music/SemWeb post. Got stuck in moderation [02:07:00.0000] MikeSmith: my comment was http://pastebin.mozilla.org/1219593 [02:07:01.0000] /me reads [02:09:00.0000] yeah [02:10:00.0000] "astounding uptake" does not seem the most apt description [13:43:00.0000] jgraham: yt? [13:43:01.0000] jgraham: http://www.w3.org/Bugs/Public/show_bug.cgi?id=12299#c9 [14:16:00.0000] Ms2ger: hah, i hadn't noticed it was you filing the bug in http://www.w3.org/Bugs/Public/show_bug.cgi?id=12187. I wouldn't have asked if it was a trick question if I had! :-P [14:17:00.0000] :) [14:18:00.0000] /me is happy to see bugs being fixed :) [14:41:00.0000] Has anyone thought about the fact that you can use the CSS cursor property to make it look like the cursor is anywhere you like on the page? Like, security-wise? [14:42:00.0000] Maybe there are easier ways to make the cursor look like it's pointing at the wrong thing, though. 2011-05-07 [03:53:00.0000] does document.hasFocus work in IE? [03:54:00.0000] nm [11:09:00.0000] hsivonen: yt? 2011-05-08 [18:44:00.0000] /me mumbles something about standardizing a tab as 8 spaces... [11:48:00.0000] Hixie, you said the stuff I'm doing for Google should be public-domain, right? I put a LICENSE file in my editcommands spec that says CC0, so tell me if that's not okay. [12:49:00.0000] AryehGregor: CC0 is fine. I just need to be able to reuse it in other specs. [12:49:01.0000] Hixie, k. 2011-05-09 [00:16:00.0000] Hixie: I'm back now [00:17:00.0000] i probably had a question about some bug or e-mail and replied there [00:20:00.0000] ok [00:32:00.0000] Hixie, I started documenting examples of how authors mark up subtitles and taglines today. http://wiki.whatwg.org/wiki/Hgroup_element [00:33:00.0000] my initial assessment is that the most common techniques are h1 with h2 for subtitle or h1 with p for tag line. [00:33:01.0000] yeah that was my assessment too when i looked at this a few years ago [00:34:00.0000] wow, even more
        wrappers around the h1/h2 or h1/p than i remember, too [00:34:01.0000] I'm really not getting the confusion surrounding hgroup though. It's quite easy to understand, although I would make a few minor adjustments to the spec to clarify it [00:35:00.0000] please file bugs for those, i'm happy to clear things up [00:35:01.0000] i don't really get the confusion either [00:35:02.0000] interestingly, the only example I found that matched Jame's suggestion using

        .. ...

        (or whatever it was), was the OED, that used h1/span [00:35:03.0000] *james' [00:35:04.0000] yeah but if you're right they also put some

        s inside

        so... [00:36:00.0000] yes, that too [00:36:01.0000] Lachy: wow. those examples really argue in favor of
        [00:36:02.0000] indeed. [00:37:00.0000] oh, the way this page uses hgroup doesn't http://www.apple.com/imac/features.html [00:37:01.0000] :-( [00:37:02.0000] i have been wondering about allowing

        in

        instead of [00:38:00.0000] mostly my plan is to wait and see if we can get away with not doing that though [00:38:01.0000] (i think if we allowed it authors might get more willing to use it for purposes that don't really fit its definition) [00:39:00.0000] that is, requiring the use of just leads people to think more about whether what they're doing is really
        or more of just
        [00:39:01.0000] anyway [00:39:02.0000] thanks for writing this up, data is always helpful [00:41:00.0000] Hixie, I found a lot of news papers were using h1/p, where others were using h1/h2, for their initial summary just after the article headline. I just didn't copy most of those into there yet [00:41:01.0000] yeah hx/p is pretty common in pre-html5 pages, because it's the only thing that html4 really allowed [00:41:02.0000] (notwithstanding that all w3c specs use h1/h2) [00:42:00.0000] but there seemed to be a fine line between what could be considered a subtitle and what was just an intoductory paragraph [00:42:01.0000] right [00:45:00.0000] Hixie, there doesn't seem to be too much point in allowing p inside hgroup, since authors can just do

        . They don't need hgroup at all in that case. [00:46:00.0000] indeed [00:52:00.0000] which variant

        or

        works better with existing AT? [00:53:00.0000] depends what you're looking for really [00:53:01.0000] but to a first approximation, they're as good as each other [02:25:00.0000] Hixie: Re bug 12299 the confusion is that the non-normative text implies that both .item(index) and [index] return null for out of range index. However [index] returns undefined because WebIDL makes that case work like normal property lookup [02:25:01.0000] oh, i should fix such an implication if there is one [02:25:02.0000] reopen the bug if i closed it [02:25:03.0000] Rather than e.g. amking it work as a catchall that always forwards to item() [02:25:04.0000] No, you didn't [02:25:05.0000] k [02:26:00.0000] i hadn't realised the bug was about the domintro text [02:26:01.0000] that makes more sense [02:26:02.0000] i'll add a comment [02:26:03.0000] thanks for the clarification [02:28:00.0000] right. time for me to head to bed. nn. [02:28:01.0000] gn [03:36:00.0000] /me wonders if he dare read enough mail to work out why the WebApps group would consider doing something patently insane like making a case-insensitive virtual filesystem API [03:37:00.0000] Presumably because that's what all widely used file systems are liked? [03:37:01.0000] *like [03:37:02.0000] Pretty sure that's not true [03:37:03.0000] Also, experience with FAT shows what a bad idea it is [03:39:00.0000] jgraham: NTFS and HFS+ both are. Certainly on OSes on which browsers are widely used, they cover pretty much 99% of the market. [03:41:00.0000] gsnedders: NTFS is case preserving although the API is case-insensitive. OSX has grown a case sensitive mode for HFS+ recently [03:41:01.0000] jgraham: I am aware of that. But in the modes in which both are used… [03:42:00.0000] Anyway the point is that this is a virtual filesystem and being case insenitive has precisely nothing to recommend it on its own merits [03:45:00.0000] http://dev.w3.org/html5/status/issue-status.html has been updated [03:45:01.0000] only open pre-LC issue is 131 (caret location API) [03:54:00.0000] hsivonen: if/when you're around and have a few minutes, I wanted to chat briefly about ARIA schema stuff [04:14:00.0000] is there a nice simple way can you think of for testing utf8 is supported by a browser inline? I.e. without checking page properties? [04:15:00.0000] MikeSmith: ok [04:16:00.0000] hendry: Data URIs? [04:16:01.0000] hendry: why do you want to test that? browsers that don't support UTF-8 are so old that why would you care about those_ [04:18:00.0000] I think there may be mobile browsers use in some parts of the world that don't support UTF-8 [04:18:01.0000] hsivonen: so I've been looking the existing ARIA integration schema with an eye towards updating it to bring it into conformance with the current HTML spec [04:19:00.0000] hsivonen: is it OK if I take a shot at putting a patch together, or is it something you've already been planning to work on? [04:20:00.0000] MikeSmith: I think most support UTF-8, just don't have non-ASCII fonts [04:20:01.0000] yeah, true about the fonts [04:21:00.0000] which is nuts [04:21:01.0000] MikeSmith: putting together a patch is OK. I'm still entagled in driving small Firefox 4 regression fixes into the tree [04:21:02.0000] OK [04:21:03.0000] MikeSmith: users of such browsers need some Opera Mini love ASAP [04:22:00.0000] I need a simple test to check off WS-1600 & WS-1610 in http://specs.wacapps.net/2.0/feb2011/core/spec.html hsivonen [04:22:01.0000] hsivonen: true, and not just in Asia. I know a few people with Android handsets in North America who avoid user the default browser and pretty much always use Opera Mini instead [04:23:00.0000] MikeSmith: well, at least the Android default browser supports UTF-8. [04:23:01.0000] yeah [04:23:02.0000] MikeSmith: still, it's a good idea to replace it with a Mozilla (or Opera) offering [04:23:03.0000] gsnedders: Data URIs look like a nice way of doing it [04:24:00.0000] hsivonen: I'd like to get a handset here that I could run Fennec on [04:25:00.0000] MikeSmith: did you read http://natalian.org/archives/2011/03/20/State_of_the_browser_2011/ ? When I tried fennec? ;) [04:25:01.0000] or Mobile Firefox is the official name now I guess [04:25:02.0000] hendry: nope, will read it now, though [04:26:00.0000] hendry: not sure what build you tried but I think there have been some significant improvements since then [04:28:00.0000] hsivonen: so, I'll start in earnest on updating the ARIA integration, and try to avoid bugging you, but if I get to a point where I'm stumped, I might ping you for some guidance [04:29:00.0000] mostly I guess I don't understand the ARIA spec very well [04:29:01.0000] but Steve Faulkner has been helping me a bit with that [04:29:02.0000] MikeSmith: I believe the name these days is Firefox for Android and Firefox for Maemo [04:29:03.0000] ok [04:29:04.0000] makes sense [04:30:00.0000] I hope they keep the fennec mascot, though [04:30:01.0000] MikeSmith: ok. re: ARIA. I really wish I'd be able to get to my pending hacking projects instead of fixing fallout from the HTML5 parser shipping [04:31:00.0000] OK [04:32:00.0000] I've been thinking it'd be useful to have in place by around the start of the Last Call timeframe, so that people can test out the document-conformance criteria of that part of the HTML5 spec [04:32:01.0000] yeah [04:33:00.0000] hsivonen: btw, as Steve pointed out on the public-html list, we have a case here were the criteria in the WHATWG spec are more restrictive than those in the W3C spec [04:33:01.0000] MikeSmith: Yeah, I need to look into that. [04:33:02.0000] MikeSmith: I'm not happy about normative parts of WHATWG and W3C specs diverging [04:33:03.0000] nope [04:33:04.0000] me neither [04:34:00.0000] we do also already have the case that ping is allowed in the WHATWG spec but not in the W3C one [04:34:01.0000] and Julian in fact raised a validator.nu bug for that [04:35:00.0000] hsivonen: if we end up having to address this, I think rather than providing an end-user option to choose what spec to validate against, it could instead by handled by a build-time option [04:35:01.0000] basically, just give an option to choose what driver file to use [04:36:00.0000] and add another toggle parameter to the schema [04:37:00.0000] MikeSmith: ping is a somewhat milder case, because it can be seen as an additional feature instead of surprising tweaks to a feature that is in both specs [04:37:01.0000] true [06:50:00.0000] seems like volkmar is closing in on the progress implementation [06:52:00.0000] MikeSmith: it doesn't have a11y (but patches are coming) nor vertical feature (but patches are ready) [06:52:01.0000] cool [06:53:00.0000] I can see there's quite a lot more work to implement that than most people would realize [06:54:00.0000] MikeSmith: native UI wasn't really easy ;) [06:54:01.0000] yeah, I bet [08:57:00.0000] hi [09:00:00.0000] shichuan: Hi [09:01:00.0000] jgraham, do you undestand X{15} from http header? [09:25:00.0000] https://twitter.com/#!/DanTonyBrown/status/67286248674426881 [09:26:00.0000] hsivonen: Is the right answer to that not "keep a ref to any audio objects you might want to pause"? [09:34:00.0000] jgraham: I don't know of a better answer [09:37:00.0000] Well having document.pauseAllAudio() would be a bit weird I think [09:39:00.0000] You would quickly want ways to pause all video and change volumes and, well pretty much everything [09:43:00.0000] http://natalian.org/archives/2011/03/20/State_of_the_browser_2011/ has annoyingly low contrast for the links [09:54:00.0000] zcorpan: So the state of the browser in 2011 is that it can't work out when the text you are interested in is illegible? [09:55:00.0000] hendry: any details about how Firefox for Android sucked? [09:56:00.0000] hsivonen: it's a known evidently, patrick from Mozilla followed up. It basically didn't work at all on a Nexus S [09:57:00.0000] hendry: oh. that [09:57:01.0000] hendry: I thought Google pushed a system update that fixed it [09:59:00.0000] it annoys me that dom3 events doesn't use webidl normatively [09:59:01.0000] hendry: Fixed in Android 2.3.3 says the wiki [10:01:00.0000] zcorpan: Is that all? [10:01:01.0000] no [12:16:00.0000] What's a good way to evade anti-right-click scripts? [12:18:00.0000] Disable scripting temporarily? [12:18:01.0000] Ctrl+Shift+J, and then do whatever I need directly in the Inspector instead. [12:18:02.0000] Ah, an Inspector shortcut. [12:18:03.0000] That works. [12:18:04.0000] Thanks. [12:18:05.0000] TabAtkins: Cmd+Opt+I, you mean :) [12:18:06.0000] Whatevs. [12:18:07.0000] You mac users and your crazy keyboard. [12:19:00.0000] Also, I think F12 is now a shortcut key in trunk? [12:19:01.0000] Is Opt the key with "alt" written on it? [12:19:02.0000] /me has never understood that [12:19:03.0000] It's the squiggly key. Looks kind of like a #. [12:20:00.0000] I thought that was Cmd [12:20:01.0000] I think it maps to Alt on real keyboards. [12:20:02.0000] Oh, you're right. [12:20:03.0000] Yes, Option also has "alt" written on it on my mac keyboard. [12:20:04.0000] Philip`: ⌥⌘I [12:22:00.0000] Why can't they just pick a name for each key that is pronounceable and written on the key and used consistently in all documentation etc? [12:22:01.0000] I guess that Command is usually what you map Ctrl-based chords to, which is easy to remember, but Option needs another helper to make it obvious that you map Alt-based chords to it. [12:23:00.0000] I wish they had kept Open Apple and Closed Apple. :) [12:24:00.0000] So tht was anti-right-click scripts to Apple keyboard layouts in 6 lines [12:24:01.0000] Well played [12:24:02.0000] I blame hober [12:24:03.0000] heh [13:39:00.0000] What's a good way to evade anti-right-click scripts? <- in Firefox, options -> content -> javascript advanced -> uncheck "disable or replace context menus"; not sure how to do it in Chrome [13:40:00.0000] That sounds like it would act weirdly if they actually had a replacement right-click menu. [13:40:01.0000] all browsers really need to not allow that by default; the current situation is a mess [13:40:02.0000] sites shouldn't do that; the context menu isn't theirs to override [13:41:00.0000] zewt-: Google Maps uses it. No browser will break that. [13:41:01.0000] google maps is the *only* site I've seen use it, and they need to stop doing that [13:41:02.0000] The default context menu is often not very useful in fancy web apps. There should be some way for authors to override the default menu, but users need to be able to get the default one easily. [13:41:03.0000] zewt-: Once enough browsers implement the context menu api, I'm sure Maps will switch over. [13:42:00.0000] that API to ... do what Tab just said--should help the situation [13:42:01.0000] YouTube uses it to suppress the context menu for HTML5 video. [13:42:02.0000] yeah, that's the sort of abuse that shouldn't be possible [13:42:03.0000] my browser features are not YouTube's to suppress [13:45:00.0000] I assume they've done something sneaky so that "Save As..." doesn't work right anyway. Or have they not figured out such a way? [13:45:01.0000] it's also just embarrassing that one of the single oldest, obnoxious JS abuses--alert("RIGHT CLICK NOT SUPPORTED")--still works [13:48:00.0000] if you can get to the context menu, I'm not sure how you could prevent the browser from saving to disk [13:48:01.0000] I'm not sure either, but I wouldn't be surprised if there was some way to mess it up. [14:05:00.0000] We should probably have a makeSaveSlightlyNonTrivial boolean attribute [14:06:00.0000] Anyone who wanted to save the video would have to learn to remove that attribute [14:06:01.0000] Just like now they learn to disable the context menu script and remove the transparent covering divs [14:29:00.0000] /me is sad when he nests
          s five deep in a spec [14:37:00.0000] AryehGregor: you should be glad that html doesn't have ol1 ol2 ol3 ol4 [14:37:01.0000] Heh, true. [14:38:00.0000] Ooh someone removed all the yellow from the logs [14:38:01.0000] what? [14:39:00.0000] has krijnh b0rked it? [14:40:00.0000] Dunno [14:42:00.0000] krijnh: ^ [15:09:00.0000] Can anyone confirm in Firefox 4.0 that going to , entering "
          foo
          bar
          [baz]
          quz
          " (no quotes) in the "New test input:" box under insertorderedlist, and hitting "Add test" causes a reproducible crash? [15:09:01.0000] Hixie: should we use the microformats wiki for as well? [15:09:02.0000] Also, if anyone has a more recent, unstable Firefox version, it would be nice to know if it crashed there too. [15:10:00.0000] /me tries in nightly [15:11:00.0000] doesn't crash [15:11:01.0000] AryehGregor: WFM in 4.0. [15:11:02.0000] Hmm. [15:11:03.0000] wilhelm_, what platform? [15:11:04.0000] Leenoks. [15:12:00.0000] /me mac [15:12:01.0000] Ubuntu, to be specific. [15:12:02.0000] I'm also Ubuntu. [15:12:03.0000] It seems like I don't have the latest point release, let me try that. [15:12:04.0000] “mv .mozilla .mozilla.old” and try again. [15:13:00.0000] Nope, still happens in 4.0.1. [15:13:01.0000] You mean try with a fresh profile. [15:13:02.0000] Yes. [15:13:03.0000] /me tries [15:14:00.0000] Still crashes. [15:15:00.0000] Hmm, now my old profile doesn't come back. That's bad. [15:15:01.0000] /me pokes [15:16:00.0000] I blame wilhelm_. [15:17:00.0000] See, that wouldn't have happed if you had mved. [15:18:00.0000] I did mv. [15:18:01.0000] Oh. [15:18:02.0000] I don't have any idea where it got these files from. [15:18:03.0000] When I look at my last daily backup, the same folder is totally different. [15:18:04.0000] Like with some things that actually have 2011 timestamps before today. [15:18:05.0000] It seems to have magicked-up a Firefox profile from 2009 or something. [15:19:00.0000] All I did was mv .mozilla{,~}, then test, close Firefox, rm -rf .mozilla, mv .mozilla{~,}. [15:19:01.0000] Extremely weird. [15:20:00.0000] I work for a different browser vendor, so sabotaging your Firefox install would have required a rather elaborate scheme. [15:20:01.0000] /me knew he shouldn't have rm -rf'd so fast [15:20:02.0000] Ah, but you have a clear motive. [15:21:00.0000] I mean, you don't hear him saying to run rm -rf ~/.opera do you? [15:21:01.0000] Oh. [15:21:02.0000] I see what happened. [15:21:03.0000] .mozilla~ already existed, so .mozilla is now a subdirectory. [15:21:04.0000] Nice. [15:21:05.0000] heh [15:21:06.0000] I absentmindedly assumed I'd get an error if it already existed, but of course mv interprets its last argument in an unpredictable fashion for directories. [15:23:00.0000] zcorpan: dunno, up to them i guess [15:24:00.0000] zcorpan: i think medium-term we should move to a model where the registry (wherever it is) is for reserving names that are being worked on; have them expire after a few years without use; and omve the ones that do get used to the spec [15:24:01.0000] zcorpan: rather than having some in the spec and some not, with no clear distinction [15:28:00.0000] joe: what's the conf #? 8606? [15:32:00.0000] http://html5.org/tools/web-apps-tracker?from=6030&to=6031 is an annoying diff for html-differences [15:34:00.0000] webvtt is still not in w3c html5 right? [15:35:00.0000] Hixie: what was the markers for? [15:39:00.0000] that diff should change nothing except move the TextTrack (TT) and MediaController (CONTROLLER) spec text into the w3c copy [15:40:00.0000] (look at the w3c equivalent diff) [15:43:00.0000] Hixie: but TextTrack was already in the w3c spec for the last two publications afaict [15:45:00.0000] yeah, [15:45:01.0000] but i wasn't sure when i added it that it was staying in [15:45:02.0000] so i hadn't removed the markers [15:46:00.0000] i'd just changed to which has no effect [15:46:01.0000] this diff removed all the KEEP-* markers [15:46:02.0000] ok [15:47:00.0000] is it standard? const select = document.createElement('select'); select.options[0] = new Option('a', 'a'); _log(select.options.length)? [15:49:00.0000] bga_: yeah, see the html spec [15:49:01.0000] bga_: except const isn't standard [15:49:02.0000] fantastic [15:51:00.0000] actually i dunno if options[0] = ... is supposed to work per spec [15:52:00.0000] does it work? [15:53:00.0000] yeah [15:53:01.0000] const select = document.createElement('select'); select.options[10] = new Option('a', 'a'); _log(select.options.length) // 11 [15:54:00.0000] i hope we will can works with other dom collections so easy [15:54:01.0000] wow [15:54:02.0000] %) [15:55:00.0000] i can't see in the spec that it should work [15:55:01.0000] so file a spec bug [15:55:02.0000] here http://www.whatwg.org/specs/web-apps/current-work/complete/common-dom-interfaces.html#htmloptionscollection [15:56:00.0000] firefox throws an exception for select.options[100000] = [15:57:00.0000] opera doesn't, and i daren't try another 0 :-) [15:58:00.0000] I believe that works in Opera. [15:58:01.0000] If you have enough memory. [15:58:02.0000] :) [15:58:03.0000] only webkit and opera [15:59:00.0000] but nice [16:00:00.0000] The fact that many things with select elements are O(n) in Opera may not help [16:00:01.0000] IIRC there are some things with select that are O(n^2) in Opera, but I don't remember what. [16:00:02.0000] way to make work with dom more easy w/o frameworks [16:01:00.0000] wrong [16:01:01.0000] ff and ie ok too [16:06:00.0000] gsnedders btw when opera will support Object.seal? do you know? :) [16:06:01.0000] bga_: In the future. [16:07:00.0000] i hope opera will throw exception in both modes [16:07:01.0000] as v8 [16:08:00.0000] because many ppl use {with} and other es3 stuff but want es5 api w/o silent errors [16:11:00.0000] gsnedders also can you node yourself to add proprietary api to throw error if code [[Get]] not existing property of object [16:11:01.0000] or Proxy api [16:12:00.0000] something like {var a = Object.makeStatic({a: 1}); a.b // error } [16:13:00.0000] bga_: Throw an exception in what case? [16:13:01.0000] i wonder why TC39 hasnt added something like that [16:14:00.0000] gsnedders to protect from typos [16:14:01.0000] /me shrugs [16:14:02.0000] bga_: I mean what you said originally, about V8 doing so? [16:14:03.0000] ah [16:15:00.0000] ok http://dev.w3.org/html5/html4-differences/ should be up-to-date now. please review (and file bugs) [16:15:01.0000] var a = Object.seal({}); a.b = 1 // errer even in nonstrict mode [16:16:00.0000] bga_: The spec should define that. [16:16:01.0000] :( [16:17:00.0000] many ppl uses es3 only features [16:17:01.0000] and can not use strict mode [16:18:00.0000] bga_: I don't know what the behaviour is off-hand, tbh [16:19:00.0000] Though I didn't think any of the Object stuff changed the [[Get]] behaviour in either strict or non-strict [16:20:00.0000] not [[Get]], [[Put]] [16:20:01.0000] Oh, [[Put]] [16:20:02.0000] /me can't read [16:20:03.0000] in sctict mode - error, in non script - nothing [16:20:04.0000] *strict [16:21:00.0000] typos [16:22:00.0000] bga_: Ah, right. It's the fact that whether [[Put]] can throw for any reason is controlled by strict mode [16:38:00.0000] bga_: did you file a bug on options[0]= ? [16:40:00.0000] sec [16:40:01.0000] requires js heh [16:49:00.0000] zcorpan http://www.whatwg.org/issues/ "E-mail is not valid or does not correspond to a user who has sent feedback that has not yet been dealt with." [16:50:00.0000] Hixie: ^ [16:50:01.0000] nn [16:50:02.0000] nn [16:51:00.0000] bga_: what's your e-mail address? [16:51:01.0000] bga.email⊙gc [16:51:02.0000] bga_: send me a mail at ian⊙hc [16:52:00.0000] bga_: my script will add your name to the list of accounts in the next hour or so [16:52:01.0000] bga_: you shoudn't need an account to file a bug though [16:52:02.0000] bga_: just use the box at the bottom right [16:52:03.0000] bga_: the account is only used for voting on issues and changing the status indicator thigies [16:53:00.0000] in other news, can we just drop aria altogether [16:53:01.0000] it's caused more problems in html than pretty much anything else so far [16:54:00.0000] lol [16:54:01.0000] i dont see any boxes at the bottom right [16:54:02.0000] Hixie: you should drop the ability to put block level elements inside of an anchor tag [16:55:00.0000] ah [16:55:01.0000] in spec page [16:55:02.0000] sorry [16:55:03.0000] boogyman: why? [16:55:04.0000] bga_: yeah 2011-05-10 [17:01:00.0000] Hixie: I've yet to come across a valid use-case for a block level element inside of an anchor. I've been given a multitude of suggestions, but I disagree with all of them... eg I'd like to make an entire table row clickable <---- lolwut??? [17:02:00.0000] makes me think of clientsfromhell.com every time someone says something as idiotic as that [17:02:01.0000] ok http://www.w3.org/Bugs/Public/show_bug.cgi?id=12639 [17:20:00.0000] boogyman, “don't allow block level elements inside anchors” seems like a bad case of “stop liking what i don't like” [17:21:00.0000] erlehmann: That is very much the case, however, I am open to the thought of a use-case that makes sense [17:24:00.0000] boogyman, browsers already handle links wrapped around block level elements. html5 is descriptive. deal. with. it. [17:24:01.0000] /me puts on glasses. [17:25:00.0000] erlehmann: previously it wouldn't validate [17:25:01.0000] boogyman, imagine a magazine style page with teasers … [17:26:00.0000] boogyman: look at whatwg.org. it has an example of block-like elements in links that i think is reasonable. [17:28:00.0000] Hixie: are you speaking about the
          bit? If so, I disagree with that as well. The title can be the link, what semantic value comes from making the entire section an anchor? [17:31:00.0000] erlehmann: I'm not following? if there are multiple teasers, use an anchor tag per teaser, with the "teaser" being an image and/or text contextual image Some description, then use CSS to properly determine layout [17:31:01.0000] boogyman, but my hypothetical teasers are , with complex inner workings! [17:32:00.0000] erlehmann: sounds like you have a non-semantic layout then [17:33:00.0000] the standards shouldn't be lenient or empathetic to un-semantic markup [17:33:01.0000] boogyman, well, should they be considered , then? [17:34:00.0000] the standards shouldn't be lenient! yellow screen of death and destruction for all! [17:37:00.0000] erlehmann: possibly an aside/section as exampled on whatwg.org, but again, I don't believe any semantic meaning value is gained by allowing the full "module" be a link [17:38:00.0000] boogyman, „the teaser links to the full-blown article“ is somewhat more in line with reality than „the teaser contains multiple links to the full-blown article“ [17:38:01.0000] boogyman: i mean the boxes, "HTML", "News", etc. look at the source. [17:38:02.0000] boogyman: the whole box each time is a link [17:38:03.0000] i'd argue it is “more semantic” based upon that [17:38:04.0000] boogyman: it makes no sense to put the link _in_ the box [17:39:00.0000] what the man said [17:40:00.0000] also sorry for my quotes [17:43:00.0000] Hixie: I'm not seeing those examples in the source [17:43:01.0000] erlehmann: This is some complex advert a span {...} [17:44:00.0000] I just don't see the semantic value :-s [17:45:00.0000] boogyman, then don't. others do and use it already. we like what you don't like. :) [17:46:00.0000] erlehmann: It's not a matter of emotional opinion; It's a matter of semantics [17:48:00.0000] boogyman: the second element child of the on http://whatwg.org/ [17:48:01.0000] boogyman, describe a scenario where you opine that wrapping a block in a link lacks semantic value opposed to a inline alternative [17:50:00.0000] boogyman: (there's not much else to that document, i'm not sure how you're missing it!) [18:58:00.0000] erlehmann: the page that Hixie is referring to is another example... http://jsfiddle.net/NXZLc/ is how I would write it up semantically [18:58:01.0000] there's no list there [18:59:00.0000] why do you have a list? [18:59:01.0000] because your markup is more aptly described as a list of options/links [18:59:02.0000] ew no, that's not at all a list [18:59:03.0000] it's just a bunch of links [19:00:00.0000] most*, could possibly even encapsulate it all within a "nav" too [19:00:01.0000] how would you handle it if the boxes had two paragraphs in them instead of the current one-paragraph solution? [19:01:00.0000] can you describe a scenario where multiple paragraphs are semantically appropriate for a single anchor? [19:07:00.0000] erlehmann: http://whatwg.com vs http://jsfiddle.net/NXZLc/ [19:07:01.0000] well the example on whatwg.org frankly i think the current markup (just one

          ) is a bit dubious. I was thinking of changing it to something and a

          , just not sure what the first one should be [19:07:02.0000] matbe a

          [19:07:03.0000] maybe [19:08:00.0000] boogyman, PS1="\w > rm \w" [19:08:01.0000] eh? [19:08:02.0000] :D [19:08:03.0000] /me puts on his troll hat. [19:09:00.0000] boogyman, you realize this *will* degrade less gracefully then just using block elements? [19:10:00.0000] li a span {display:block} … and nothing of value was gained [19:11:00.0000] erlehmann: it's cosmetic, therefore, by definition its to be done via css [19:12:00.0000] boogyman, tell that to the guy with spec edit rights, not me. %) [19:12:01.0000] it's not cosmetic [19:12:02.0000] the whole thing is a link, and the whole thing is a paragraph [19:12:03.0000] arguably two paragraphs [19:13:00.0000] “the whole thing is a link”, there is the point. boogyman, do you think that the whole thing is not a link? [19:13:01.0000] I'd think of it more of a dl, or section with a heading and description [19:14:00.0000] it's clearly not a bunch of name-value pairs :-P [19:14:01.0000] you seem to have drunk some semantic koolade :-P [19:15:00.0000] koolade is for the birds, but erm, i think it's a very loose fit. But anyway, this is the first example I've seen that could potentially be semantically appropriate. I just disagree with the principle [19:17:00.0000] but, it's still a list of links [19:17:01.0000] it's not more a list of links than a book is a list of paragraphs [19:21:00.0000] correct, but a paragraph has semantic meaning in both a single and group setting. This is a group of links that navigate the user to specific pieces of the website [19:23:00.0000] boogyman, a web site is a list of links. [19:23:01.0000] :3 [19:23:02.0000] (thanks, I'll be here all week) [19:23:03.0000] badum, chee [19:26:00.0000] Hixie: please tell me you're not considering

          Category

          foo description

          [19:29:00.0000] s/strong/h1/ge [20:10:00.0000] very cool to see that doublec is working on a media fragments implementation [20:10:01.0000] hopefully the questions and feedback will drive some changes to the spec [20:14:00.0000] thanks MikeSmith! [20:16:00.0000] doublec; was also reading foolip comments in the bug, and your reply [20:17:00.0000] whatever subset you end up supporting, hopefully other implementors can implement that subset also [20:17:01.0000] yes, I'm hoping we can agree. foolip and I seem to be in agreement from what I can tell. [20:18:00.0000] At the Foundations of Open Media workshop in 2010 we (firefox media devs) gave similar feedback to those working on the fragment spec [20:29:00.0000] doublec: I guess the feedback sometimes doesn't get treated seriously enough until the implementation work starts in earnest [20:30:00.0000] hey jamesr - I think you mentioned something about needing mercurial help here a while back [20:30:01.0000] I just wanted to say if it was in the context of using the W3C mercurial repo, I am always glad to help with that [20:44:00.0000] MikeSmith: hi! i had some trouble using the wrong acct name, but sorted that [20:44:01.0000] MikeSmith: i don't understand how properly to handle merges [20:44:02.0000] but i have to run and eat dinner now. i might ask you about best practices for multiple editors/merges when using the w3c mercurial repo at some point in the future if you're willing to explain things [20:45:00.0000] i think i grok the git model, but i'm not sure if that is helpful or harmful to understanding the hg model [20:45:01.0000] ok [20:45:02.0000] feel free to ping me any time [20:45:03.0000] I'm always on this channel when I'm online [21:01:00.0000] Hello, just curious as to why the TextMetrics object only contains a `width` property [21:02:00.0000] why it's been decided that way [21:30:00.0000] eboyjr: we'll add more in due course [21:33:00.0000] Hixie: okay, cool. what's due course? [21:33:01.0000] once the browsers have caught up with what the spec already says [21:33:02.0000] no point adding features when they still haven't implemented the last bunch :-) [21:35:00.0000] hrm i'da figured it would be good to come up with a full set of features and let browsers implement it at its own will and discretion [21:38:00.0000] bounding-box metrics would be more valuable, imo [22:43:00.0000] hmm, the dfn for term "space character" is marked up with class=impl [22:43:01.0000] so it only appears in the full version of the spec [22:43:02.0000] but not in the non-implementor view [22:44:00.0000] so non-implementors don't get to know what the space characters actually are [01:04:00.0000] /me #whatwg modes... [02:07:00.0000] zcorpan: about the links in the developer versin [02:07:01.0000] *version [02:07:02.0000] I think there are some references that it's acceptable to have be links back to the full spec [02:08:00.0000] in the case of the W3C Web Author Edition, the build process I set up to create that takes any broken links it finds, and rewrites them to point to the full spec [02:09:00.0000] MikeSmith: oh they're not just broken links? [02:09:01.0000] aha [02:09:02.0000] yeah, would be broken if it weren't for the rewriting [02:09:03.0000] Ben uses a different build setup than the one I do [02:10:00.0000] so I'm not sure if he is doing the fixup thing or not [02:11:00.0000] anyway, for example, there are a lot of links in the IDLs that reference stuff which is only in the impl view [02:11:01.0000] well then there are other ways to find which links are "broken" [02:11:02.0000] true [02:11:03.0000] maybe a user style sheet and casual browsing is effective [02:11:04.0000] but my point is, a lot of them are human-judgement cases, I think [02:11:05.0000] yeah [02:11:06.0000] could help [02:12:00.0000] as far as finding the ones that need fixing [02:12:01.0000] yes [02:12:02.0000] the way the stylesheet is for the author view now, those links to the full spec do show up differently on hover [02:12:03.0000] but not normally [02:13:00.0000] but of course it'd be possible to have them show up even not on hover [02:13:01.0000] I think for my build at least they may all have class=full_spec or something on them [02:14:00.0000] so wouldn't even have to use any fancy selector stuff to select them [02:35:00.0000] it's impossible to draw eillipses with canvas? surely it isn't? [02:39:00.0000] describe it using beziers, or stretch an arc with a transform [03:09:00.0000] Philip`, still waiting for http://www.whatwg.org/specs/web-apps/current-work/multipage/the-iframe-element.html#getting-media-metadata to become http://www.whatwg.org/specs/web-apps/current-work/multipage/the-video-element.html#getting-media-metadata ;) [03:11:00.0000] zcorpan: about the u change [03:11:01.0000] in the Changed Elements section [03:11:02.0000] minor formatting nit [03:11:03.0000] you put no around the "u" [03:12:00.0000] so it's not rendered the same as the other element names in that section [03:15:00.0000] oops [03:16:00.0000] fixed [03:16:01.0000] /me checks [03:16:02.0000] cool [03:16:03.0000] thanks [03:17:00.0000] what we really need is a specific element for marking up Chinese proper names [03:17:01.0000] just as with the proposed element for markup up ship names [07:49:00.0000] ugh - you'd think browsers would be able to implement alpha transparency based word-wrapping :) [07:51:00.0000] what? [07:52:00.0000] well, we have nice block wrapping of text, but nothing more advanced, eg. having an image of a black circle with transparent background [07:52:01.0000] then have browser detect parts that are 'empty' and allow text flow around it [07:53:00.0000] oh, right. I think there might be some CSS proposals for that somewhere, and I'm pretty sure it can be done in SVG by specifying the path. I'm not sure of the details though. [07:54:00.0000] looks like I get to write a fun article if it's possible via svg [07:54:01.0000] alystair, is this what you're looking for? http://dev.w3.org/csswg/css3-exclusions/ [07:55:00.0000] People will presumably complain about security implications, so it'd have to be restricted to word-wrapping around same-origin images [07:55:01.0000] neat! [07:56:00.0000] security implications? [07:56:01.0000] alystair, being able to detect the content of the image at all is a security risk for non-same-origin images [07:57:00.0000] alystair: using the feature to probe the alpha channel of a confidential image [07:57:01.0000] .... [07:57:02.0000] the program would be doing that, not a human being [07:57:03.0000] but that CSS proposal works by specifying an explicit path, rather than any image heuristics or edge detection [07:57:04.0000] oh. [07:58:00.0000] with clever SVG filters applied, it might be possible to move each of R, G and B to alpha channel, too [07:58:01.0000] so what you're saying is that by using automatic edge detection one could externally somehow identify it's content by the edges? :P [07:58:02.0000] alystair: yeah [07:58:03.0000] depending on what the image is, yes [08:00:00.0000] it's a fun concept to think of using 2px fontsize, but realistically what sort of image with transparency would have some sort of secret information [08:01:00.0000] an email address perhaps? [08:01:01.0000] http://top-secret-project-intranet-server/logo-containing-secret-project-name.gif [08:01:02.0000] you'd then have to OCR the edges... [08:01:03.0000] and not the entire shape [08:01:04.0000] Which you could maybe do [08:02:00.0000] somehow I doubt OCR has advanced that much :P [08:02:01.0000] You don't need OCR, you just send the raw outline data back to the attacker's server [08:02:02.0000] as hsivonen said, you could use SVG filters, clipping regions or other effects to adjust which colours are rendered transparently, which means you can probe the image to find edges [08:02:03.0000] where they can read it manually [08:02:04.0000] Right, that's basically what I meant [08:03:00.0000] talk about an edge case :P [08:04:00.0000] alystair, these are similar reasons to why canvases become tainted, and restrict the ability to obtain pixel data, when a non-same-origin image is added [08:04:01.0000] Security is all about edge cases, since it only takes one to break the security model [08:05:00.0000] realistically can't the wrap script just have a certain limit on scan resolution of the image? [08:05:01.0000] eg. 12px blocks as a minimum [08:06:00.0000] (the "edge case" thing scored 1 comedy drumroll btw) [08:07:00.0000] You could try making a limit like that but it seems hard to justify [08:07:01.0000] (But with images it seems pretty futile to try squashing all attacks, because there's various (hypothetical?) timing-related attacks and presumably the performance cost of protecting against them is unacceptable in a benchmark-driven browser marketing environment) [08:07:02.0000] alystair, that all depends on the target image and what level of detail is really necessary to obtain the information. Even a crude outline might be enough in some cases [08:07:03.0000] /me wonders if anyone has actually got such attacks working in practice [08:08:00.0000] I find it silly that things like this even need to be discussed because we're trying to shelter even the dumbest programmer from an attack like that [08:08:01.0000] whereas disallowing this kind of functionality for cross domain images seems like a quite reasonable solution that has the normal properties of the web security model [08:08:02.0000] Uh [08:08:03.0000] if someone is able to inject something that sends data back to a 3rd party server then they have much much larger issues [08:08:04.0000] It's not about injecting something [08:08:05.0000] There are no programmers involved [08:09:00.0000] It's about a user visiting http://attacker.com/ which runs scripts on the user's machine (i.e. behind their firewall) [08:09:01.0000] alystair, [08:10:00.0000] Evil site A convinces a user to visit their page and loads a resource from secret IP-restricted server B into A. It then uses some technique to extract information from that resource [08:10:01.0000] Allowing them to discover secrets held on B [08:13:00.0000] like a flash uploader since they happen to know the EXACT address of the secret image url :P [08:42:00.0000] http://www.google.com/events/io/2011/index.html <- pretty :) [08:45:00.0000] /me notes that the person who designed that never tried it on a laptop screen [08:46:00.0000] jgraham, looks perfect on my laptop screen. [08:48:00.0000] perhaps they are running the assumption that many of the people visiting the page would have high resolution screens [08:48:01.0000] Lachy: You have the world's biggest laptop screen [08:48:02.0000] Roughly [08:48:03.0000] On a 13" screen it doesn't work [08:48:04.0000] not quite. I only have 17". Dell makes a 19" [08:49:00.0000] It looks alright on mine, which is only 1280x800 [08:49:01.0000] It might depend on how much other stuff you have [08:49:02.0000] (as long as I don't want to read the first two digits) [08:49:03.0000] On a pretty standard OSX setup the other day the QR code blocked the time [08:49:04.0000] The QR code is on top of the counter [08:50:00.0000] Ah, yes, quite broken [08:50:01.0000] hehe I see that [08:50:02.0000] I thought that was intentional at first [08:50:03.0000] if you drag it up enough it overlaps everything else :) [08:50:04.0000] houla and the CPU is screaming [08:51:00.0000] do most of the people in the whatwg participate on the mailinglist or do many contribute/edit the spec without talking on it? [08:53:00.0000] COMMAND %CPU [08:53:01.0000] Opera 88.6 [08:53:02.0000] The WHATWG is basically the mailing list [08:53:03.0000] Plus this channel [08:53:04.0000] Plus an evil cabal [08:53:05.0000] Who evilly do nothing [08:54:00.0000] The only person who directly edits the spec is Hixie [08:58:00.0000] (the "evil cabal" is actually just a mailing list of some people who can, theoretically, do things like change the editor. Apparently they have hardly had any email on that list ever) [08:59:00.0000] (and they are likely unneeded anyway since one could change the editor by forking the spec. Of course it would only work if people wanted to follow the fork rather than the original. But that is basically an isomorphic problem to getting the people on that list to agree to change the editor) [09:02:00.0000] haha alright [09:02:01.0000] there's actually an easteregg on the google.io site [09:02:02.0000] when you drag the logo it allows you to move the buttons and stuff around as well [09:02:03.0000] which work as blockers for the balls :) [09:32:00.0000] Why doesn't input type=email allow one to set the IDL attribute to a unicode string and have it converted to punycode, or something? [09:34:00.0000] Or even better, have it stored as unicode and converted to punycode for submission [10:08:00.0000] http://www.google.com/events/io/2011/index.html <- live streaming from event now [10:16:00.0000] alystair: can't get it to work from France [10:16:01.0000] streaming movie rental [10:16:02.0000] really? lame [10:16:03.0000] neither live stream? [10:18:00.0000] http://www.youtube.com/watch?v=RhmWg7Lp0i0 direct link :S [10:19:00.0000] yeah, it says the video is not available [11:40:00.0000] Reproducible crash for me in Opera 11.10 on Ubuntu 10.10: go to http://aryeh.name/spec/editcommands/autoimplementation.html#insertunorderedlist, click "Run tests" under "insertorderedlist". Can anyone else reproduce? [11:40:01.0000] (It seems I have an amazing talent for finding crash bugs, especially in Opera) [11:41:00.0000] (it really makes me appreciate Chrome and IE, where you can just reload the tab) [11:45:00.0000] AryehGregor: I was just about to say that the test worked for me, but then Opera crashed. :P [11:45:01.0000] AryehGregor: So, reproduced. [11:52:00.0000] AryehGregor, reggna: submit a crash log? what email? [11:52:01.0000] gsnedders, I didn't get prompted to submit a crash log. [11:52:02.0000] Can you reproduce? [11:52:03.0000] AryehGregor: Not tried [11:56:00.0000] AryehGregor: Yeah, repro with no crash dialog [11:56:01.0000] k. [11:57:00.0000] Please tell me if you figure out what the problem is so I can update the page. [11:59:00.0000] /me is downloading more recent debug build… [12:00:00.0000] /me wants quicker internets [12:03:00.0000] AryehGregor: Trying to clear a selection is failing to remove an element not in the range. [12:03:01.0000] What do you mean? Selection.removeAllRanges() is what's crashing? [12:03:02.0000] Yeah. [12:04:00.0000] Because there's something in the range that isn't in the range. [12:04:01.0000] Oh. [12:05:00.0000] /me accidentally opens link in his main Opera install [12:05:01.0000] woops. [12:09:00.0000] heh, the only time I've ever really found browser themes to be useful is when debugging a browser side-by-side with my real browser--to make it easy to remember which is which [12:10:00.0000] I just idlely right clicked -> open link [12:10:01.0000] It doesn't crash when you just open the link, though, right? [12:10:02.0000] There's a reason why I know quite a lot of browser people who have a browser they don't work on as their main browser [12:10:03.0000] AryehGregor: Right, but close enough to not be nice. [12:10:04.0000] Ah, okay. [12:11:00.0000] Of course, browsers usually do pretty well these days on restoring after a crash. [12:11:01.0000] at least session restoration has made browser crashes marginally less annoying than they used to be :) [12:11:02.0000] let's say the same thing; ready, go [12:12:00.0000] AryehGregor: Yeah, it's pretty much fine, just means restarting browser, which when you have far too many tabs open can take a bit [12:14:00.0000] i have something like 100 right now :| [12:17:00.0000] zewt, gsnedders, on firefox i use bar tab, it only loads tabs that are accessed. [12:18:00.0000] but still i once had a crash where the restart would crash again … the session restore tab would restore a session restore tab! [12:18:01.0000] it was session restore tabs all the way down. [12:19:00.0000] ff4 seems to try to preferentially restore tabs when they're accessed when loading a session, though i find the end result is just slower [12:19:01.0000] bar tab then [12:20:00.0000] like, built-in? [12:20:01.0000] standard ff with a vertical tab addon (doesn't change behavior) [12:20:02.0000] Built-in, yes [12:21:00.0000] fine :) [12:32:00.0000] AryehGregor: http://pastebin.com/yjDymymq is a more minimal copy of the inline script [12:33:00.0000] gsnedders, okay, thanks. [12:33:01.0000] /me doesn't have time to take it further [12:35:00.0000] Meaning you don't have time to reduce it further, but you'll still make sure that the crash bug is fixed, because this might be happening on real sites? [12:36:00.0000] /me isn't sure whether implementers prioritize crash bugs as all super-high priority or what [12:39:00.0000] AryehGregor: many crashes like this tend to have security implications and are prioritized pretty high because of that [12:39:01.0000] speaking of, if you find any more in webkit i'd appreciate if you'd file a security bug about it first and not post the repro in public IRC channels :). i think other vendors would appreciate the same [12:40:00.0000] jamesr, hmm, really? Okay. Didn't realize they were that likely to have security implications. I don't think I've found any crashes in WebKit lately. [12:40:01.0000] not all will, but many will [12:40:02.0000] I found one in Firefox a day or two ago, but no one else could reproduce it, so I figured it wasn't worth reporting. [12:41:00.0000] I've found a few in Opera that I mentioned here, and no one from Opera said they'd prefer it privately. [12:41:01.0000] if it's something like a double free or use after free then the crash might not be reliable [12:41:02.0000] i'm not familiar with opera's policies [12:44:00.0000] jamesr, isn't full disclosure standard? [12:44:01.0000] except with webkit? [12:44:02.0000] what? no [12:44:03.0000] you don't just 0-day everyone [12:44:04.0000] I don't think anyone appreciates full disclosure without first at least informing the vendor. [12:45:00.0000] people argue about details but the IMO reasonable behavior is to disclose to vendor, give them a chance to patch users, then disclose publicly [12:46:00.0000] I once published a short script crashing a chat client and it made people angry and they fixed the bug. But, it made people angry :> [12:46:01.0000] Otherwise, I have no idea how bugs should be handled. [12:51:00.0000] The annoying thing is it's considerably more effort for me to actually file a bug report, and the response is likely to be considerably longer, compared to just saying it here. But I guess I'll be more careful in the future. [12:51:01.0000] why is it considerably more effort? [12:52:00.0000] jamesr: this almost certainly isn't exploitable, at least [12:52:01.0000] Because over here I can just type one line in a chat, and someone will probably answer in a few minutes. To file a bug I have to make a permanent link, go through a bunch of forms, and then wait who knows how long for a response. [12:53:00.0000] My normal approach is to file a bug and poke people on IRC [12:54:00.0000] AryehGregor: Basically the policy for things that might be security issues is to treat them as security issues until proven otherwise. [12:55:00.0000] Whether you treat all crash bugs like that or not is a good question [12:55:01.0000] I've been assuming they don't need to be treated like that. [12:56:00.0000] I have no idea what our policy is :) [12:59:00.0000] All the range crash bugs you've found have just been null-pointer dereferences, nothing interesting from a security POV [13:01:00.0000] Generally the bugs I find are in pages I'm rapidly changing anyway, and in a few hours it won't be reproducible from the instructions I gave in the chat. [13:02:00.0000] (since it's not very useful for me to have my tests crash) [13:02:01.0000] So I'm not going to worry too much. [13:04:00.0000] AryehGregor: Uh, I think there may well be more than one crash bug there. [13:04:01.0000] Interesting. [13:05:00.0000] AryehGregor: Like that reduced script gives an entirely different crash [13:06:00.0000] Kind of weird for me to hit two crashes at once. [13:20:00.0000] Not really if that code happens to be buggy [13:30:00.0000] AryehGregor: Back to looking at it after all [13:42:00.0000] AryehGregor: http://pastebin.com/Uxv2Nczw [13:43:00.0000] Interesting. [13:55:00.0000] "* Philip` wonders if anyone has actually got such attacks working in practice" - turns out they have - http://www.contextis.com/resources/blog/webgl/ [13:56:00.0000] /me wonders if it's possible without using WebGL, based on the performance of 2D canvas operations [13:58:00.0000] Philip`: that'd be pretty tricky [13:59:00.0000] "pretty tricky" sounds different to "impossible" :-) [13:59:01.0000] well, side channel attacks are always crazy [14:00:00.0000] but to make this attack work you need the draw time to vary as much as possible based on the value of the pixels [14:00:01.0000] and that's hard to do with the canvas2d ops [14:01:00.0000] much easier if you can write your own shader [14:01:01.0000] this attack also depends a lot on how the graphics driver+card evaluate shaders [14:07:00.0000] Ugh. br elements are a huge pain in the neck. [14:07:01.0000] br { display: none; } [14:07:02.0000] :3 [15:19:00.0000] Will those who cannot live with the upvoted licenses now commit seppuku? [15:21:00.0000] If only they were so honorable. [15:22:00.0000] I love writing about character references in HTML. &amp; all over the place. [15:37:00.0000] wilhelm_, why should they? Is the W3C going to adopt them? [15:37:01.0000] I mean, it was a non-binding survey, wasn't it? [15:40:00.0000] So you aren't bound by your statement that you couldn't live with it? [15:40:01.0000] Well, then a different subset of the surveyed should commit seppuku. Unless they were lying, and can live with it. [15:41:00.0000] No, I'm saying you should only expect anyone to commit seppuku once a particular license is actually adopted. [15:41:01.0000] True. [15:43:00.0000] Presumably people who can't live with any non-forking licences also can't live with the spec's current non-forking licence, in which case they wouldn't be alive to object to it even if it'll change later [16:22:00.0000] The per-organization breakdown of the license survey is interesting. [16:23:00.0000] Implementers and unaffiliated people are overwhelmingly in favor of some type of free license. [16:24:00.0000] But if you do a count by organization and ignore unaffiliated people, non-free licenses would win, because the non-implementer organizations lean strongly against free licenses. [16:25:00.0000] In fact, only one non-implementer organization (Intel) said that it couldn't live with option 3, while eight non-implementer organizations said they preferred or could live with it. [16:25:01.0000] This pattern is interesting, because the AC is made up almost entirely of non-implementer organizations, and unaffiliated people get no say. [16:26:00.0000] And, of course, it voted against free licenses. [16:26:01.0000] I always find it interseting that there is such a disconnect between the orgs that actually have a stake in the tech and those that don't. [16:27:00.0000] It suggests to me that perhaps implementers and individuals are involved in the W3C because they have a stake in the issue and don't care so much about the W3C itself, while non-implementer organizations are more likely to be interested in influencing the standards through the W3C, and so would be unhappy with the W3C losing power. [16:27:01.0000] Implementers will control the specs anyway, and unaffiliated individuals have no say anyway, so it doesn't make a big difference to them whether the specs are edited at the W3C or someplace else. [16:28:00.0000] Except Microsoft. Who knows who made Microsoft's decision and why. [16:28:01.0000] Of course, the W3C is dependent on having lots of members that don't have a big stake in things, because that's the only way it can get enough membership dues to support its bureaucracy. [16:29:00.0000] So I don't think the W3C is likely to be fixable, in the end. [16:29:01.0000] To fix it, you'd have to give the power to the people who have an actual stake in things, and that would disenfranchise the other members. [16:29:02.0000] Who would either not permit it, or leave and take away the W3C's funding. [16:30:00.0000] I guess the implementers aren't going to actually leave the W3C anytime soon, though, since the job it does isn't sufficiently bad to warrant the trouble, so we're stuck with the status quo. [16:45:00.0000] jgraham: yt? (or any other opera peeps) [16:47:00.0000] \o 2011-05-11 [17:08:00.0000] AryehGregor: I didn't understand the breakdown by organizations. Because basically it is not really the voice of the organization but the voice of people inside the organization which is slightly different. [17:08:01.0000] I guess the AC vote is the real breakdown by organizations [18:29:00.0000] AryehGregor, any rationale why non-implementors do not like free licenses in general? [18:30:00.0000] /me just doesn't get it [18:34:00.0000] erlehmann: Based on the responses that were given, it's mostly because they dont' understand copyright law or history. [18:34:01.0000] It's possible that some of the people who voted but didn't give a response have some reasonable objections, but nobody who actually gave a response did. [18:35:00.0000] how unfortunate. but based on my experience regarding voting on technical issues, it seems plausible. [18:37:00.0000] For example, several people seem to believe that by restricting forking with copyright they can actually prevent forking, which is obviously incorrect (it just makes it more difficult, as we saw with HTML5). [18:38:00.0000] Others seem to believe that the mere presence of a fork is automatically harmful, despite the fact that forks are only as strong as the implementors who support them, as evidenced by the multitude of existing forks that are completely irrelevant. [18:38:01.0000] Finally, one crazy person believes that forking the HTML spec may allow dictators to oppress their populace more effectively. I... I just don't know what that one's about. [18:41:00.0000] Because, obviously, autocrats will respect copyright. [18:41:01.0000] Yes, that's certainly part of the reason why that objection is crazy. [18:43:00.0000] There should be some “civilized” form to say that something is wrong in an entirely wrong way. [18:43:01.0000] "You're not even wrong."\ [18:43:02.0000] Oh, I may use that :3 [18:43:03.0000] ah, pauli [18:44:00.0000] Yus. [18:44:01.0000] not entirely apropos here, but yeah [18:44:02.0000] i prefer "fractally wrong" for this kind of thing [18:44:03.0000] wrong at every level [18:48:00.0000] those with Member access should look at the survey the Team actually posted to the AC [18:50:00.0000] Ooh, where? [18:51:00.0000] Unrelated: Yo, people in Japan, I've lost the super-convenient site for getting info/booking for the shinkansen. Any help? [18:59:00.0000] http://www.youtube.com/watch?v=TXtOv0QlYao [19:01:00.0000] Japanese people: found it, it was hyperdia.com [01:34:00.0000] does anyone happen to have a demo with a a long WebM file for testing if seeking works over HTTP Range requests? [01:34:01.0000] it's supposed to work, right? [01:39:00.0000] testing with http://lachy.id.au/log/2010/05/webm seems to work if Firefox and Opera [01:40:00.0000] also Chrome 12 [01:40:01.0000] not in Chromium 11 [01:56:00.0000] what's the deal with it appearing that Range request-based seeking is in Chrome 12 but not in Chromium 11. Were they really this late to add it, am I testing wrong or do Chrome and Chromium have very different video code paths? [01:56:01.0000] I tested Google-provided Chrome and Canonical-provided Chromium [01:58:00.0000] maybe ask over in #chrome channel? [01:58:01.0000] nessy: does the result that Firefox and Opera support Range-based seeking seem correct to you? [01:59:00.0000] I know they do [01:59:01.0000] nessy: ok. thanks [01:59:02.0000] and yes, from what you are describing it seems chromium doesn't [01:59:03.0000] which I find strange [01:59:04.0000] like you :-) [02:00:00.0000] I just tried Chrome 11 on Mac [02:01:00.0000] it seems to support range-based seeking [02:05:00.0000] (ups, channel is #chromium) [02:36:00.0000] nessy: thanks for offending people about what is "Web content". Now I'm not the only one. :-) See also http://hsivonen.iki.fi/web-stack/ [02:37:00.0000] hsivonen: I was hoping there were some people that support my view :-) [02:37:01.0000] I was very amused at that discussion [02:39:00.0000] folks at the W3C can be pretty sensitive about getting called on doing stuff that's not Webby in the sense of being part of the interoperable platform implemented in browsers [02:43:00.0000] I'd be surprised if chromium didn't support range requests with video [02:44:00.0000] nessy: seen this? http://schepers.cc/webmandering [02:45:00.0000] doublec: I'm just guessing here, but it's plausible that getting the index data out of WebM into the browser requires some ffmpeg API that's not in the system ffmpeg [02:45:01.0000] hsivonen: ah,right. possibly. [02:45:02.0000] doublec: based on how the Chromium packaging in Ubuntu is structured, it seems that ffmpeg is still involved [02:46:00.0000] I wonder if they'll drop ffmpeg and use the Xiph libs and libvpx directly when they drop H.264 [02:47:00.0000] I think ffmpeg's webm decoder is faster than libvpx [02:47:01.0000] or at least, that was the claim at one point [02:47:02.0000] http://x264dev.multimedia.cx/archives/499 [02:47:03.0000] doublec: is Chrome using the ffmpeg VP8 decoder or libvpx-wrapped-in-ffmpeg? [02:49:00.0000] wow. Dark Shikari's blog has been really quiet for a year now [02:49:01.0000] hsivonen: there's a post in april [02:50:00.0000] hsivonen: or february - depending on what date format he's using [02:50:01.0000] doublec: yeah, but that one doesn't say much [02:50:02.0000] true [02:50:03.0000] maybe he's working on some super secret project [03:02:00.0000] doublec: I was wishing Dark Shikari was working on a VP8 encoder [03:02:01.0000] doublec: but, curiously, he broke the news that another person was [03:13:00.0000] I wonder why the git and release activity around the WebM QuickTime Component has ceased [03:23:00.0000] If Microsoft really does by Skype, then I wonder what will happen to the support for VP8 currently in Skype 5? [03:26:00.0000] It will go the same way as the Linux client? [03:27:00.0000] jgraham: which way is that? [03:27:01.0000] Well, I can't imagine that Microsoft buying skype is going to *improve* the already third-rate linux support [03:28:00.0000] (the latest download for linux is a version of 2 labelled beta) [03:29:00.0000] (although cynical people might point out that 2-anything is a win over 5-anything) [03:29:01.0000] jgraham: I thought the 2.x Mac version was first-rate, Linux was second-rate and Windows was third-rate [03:29:02.0000] the Mac Skype 5 sucks. 2.8 is still superior for usability. [03:29:03.0000] Well I haven't really used it on windows much [03:30:00.0000] jgraham: at least the Mac and Linux versions don't stick junk into Firefox [03:30:01.0000] Well that kind of behaviour seems to be more tolerated on Windows in general [03:30:02.0000] jgraham: unfortunately, yes [03:41:00.0000] jgraham: if they bought Skype for the userbase, it would be silly to drop support for various platforms, since availability on pretty much all platforms is what drives the network effects [03:41:01.0000] hsivonen: They would be insane to drop Mac for that reason [03:42:00.0000] But Linux is already pretty much unsupported [03:42:01.0000] and likely has a negligible fration of the users [03:44:00.0000] jgraham: hard to tell what effect dropping Linux support would have on userbase [03:47:00.0000] jgraham: http://www.omgubuntu.co.uk/2011/05/microsoft-will-invest-and-support-skype-on-linux/ [03:48:00.0000] anyway, it's pretty clear that the world needs VOIP that's not tied to a particular app vendor [03:48:01.0000] too bad SIP sucks [03:48:02.0000] what happened to libjingle? why didn't it go anywhere? [03:48:03.0000] Hard to tell what "non-Microsoft platforms" menas [03:49:00.0000] I mean they won't drop OSX support [03:49:01.0000] That would be insanity [03:51:00.0000] one of the big problems with getting cross-vendor VOIP is that if you leave e.g. the mobile part to traditional vendors in that space, they will want to use a proprietary codec and will let carriers impose ridiculous restrictions [03:52:00.0000] the other big problem is addressing [03:53:00.0000] do you require a phone number (what about desktops?), do you require an XMPP ID (how you make people get one)? [04:30:00.0000] hsivonen: those graphics from before make for a fun read :-) not that I really care: for every purpose you can draw something else [07:59:00.0000] Why is HTMLPropertiesCollection.namedItem a caller and a getter? [08:06:00.0000] erlehmann, although you aren't here anymore: I'd hazard a guess that implementers like free licenses because it means the W3C has no control over the specs in the end, which means the implementers will have de facto control. Non-implementers will have no control over the spec outside the W3C. Also, one or two of the implementers are ideologically committed to open-source type stuff, which practically no companies are. [08:07:00.0000] What's this survey that the Team submitted to the AC that Maciej mentioned? [08:08:00.0000] Hmm, visited links on this W3C page are lighter than unvisited links. How incredibly confusing. [08:09:00.0000] Oh, is this it? http://www.w3.org/2002/09/wbs/33280/doclic201105/ [08:09:01.0000] yes [08:12:00.0000] Oh, the Intel AC rep is Wayne Carr? Like, the guy who said we shouldn't adopt free licenses because it will allow dictators to suppress their citizens' Internet access? That's a shame. [08:12:01.0000] wtf? [08:13:00.0000] oh, are you referring to his comments about governements forking the spec and stuff? [08:13:01.0000] Yeah. [08:14:00.0000] oh, yeah. Perhaps he doesn't realise that governements can and do pass crazy laws all the time without forking specs [08:15:00.0000] That is one possibility [08:15:01.0000] Seriously, what stake does Intel have in HTML5 anyway? [08:15:02.0000] Pentiums make the web faster, if I remember the adverts [08:17:00.0000] Then they should have left the W3C when they stopped making Pentiums, surely? [08:17:01.0000] There are a large number of Members with no obvious stake in the web [08:17:02.0000] ]I assume that is why W3C does so many non-Web things [08:18:00.0000] Which is why it's kind of a puzzle how they get to run the W3C. Or really it's not, it's because they give the W3C money. [08:18:01.0000] /me keeps the AC survey results open in a tab to look at occasionally as they progress, although being unable to talk about them publicly is kind of annoying [08:19:00.0000] AryehGregor, it's useful to keep such members in the W3C even if they have no obvious stake in current specifications, since a) the W3C gets membership fees from them, and b) such companies with large patent portfolios continue to be bound by the patent policy [08:19:01.0000] They're only bound by the patent policy for WGs that they're members of. [08:20:00.0000] But even non-HTMLWG members are voting on the HTML5 license. [08:20:01.0000] So that part is weak. [08:20:02.0000] Intel is a WG member [08:20:03.0000] I know. [08:20:04.0000] That's how Wayne Carr answered the HTMLWG survey. [08:20:05.0000] As for (a), maybe if the W3C were less bureaucratic, it wouldn't need so much money. [08:21:00.0000] Or maybe implementers would be willing to pay more of the bill if they were the ones that controlled the organization. [08:21:01.0000] they need to cover operational costs somehow. [08:22:00.0000] What's the W3C's budget? A few tens of millions of dollars, probably? That would be a pretty small burden if split between Microsoft, Apple, and Google. [08:22:01.0000] Plus they could get rid of a lot of the operational costs if they didn't have to have decision-making structures independent of implementer consensus. [08:24:00.0000] All companies have a stake in the web to the extent that they are heavy users of it and benefit from improvements to the platform - I assume that's largely why Google started Chrome (they're not doing it to sell an OS, they're doing it to drive progress in the technology their business depends on) but other companies that haven't bothered doing their own browser development may have a similar desire for improvements [08:24:01.0000] To be fair giving even more control to those with the deepest pockets doesn't sound like a winning move [08:26:00.0000] It's not about deep pockets, it's about market share. Mozilla has very shallow pockets compared to Apple or Google, but it has more clout in web standards. [08:27:00.0000] Look at WebSQL. Apple+Google+Opera = more money than Microsoft+Mozilla, but nowhere close to enough market share to make it tenable to ignore them, so they lose. [10:55:00.0000] lazyirc, what's the correct behavior when a [21:38:00.0000] note that in text/x-foo is equivalent to document.write(variable here) [21:38:01.0000] and that the text/x-foo language does not have comments [23:12:00.0000] On http://ln.hixie.ch/?start=1137799947&count=1 [23:12:01.0000] Arjun Ray has complained about it long before 1999. [23:12:02.0000] Arjun Ray has complained about SGML comments long before 1999. [23:14:00.0000] Since at least 1995, I think. [23:17:00.0000] Here is an old test page: http://www.nyct.net/~aray/junk/hide.html [23:17:01.0000] (Sadly it uses the HTML 2.0 DOCTYPE which has never triggered standard mode. [10:04:00.0000] Hey folks! [10:05:00.0000] Yall were so helpful with my websocket questions yeterday, I thought I would come back and ask more. [10:05:01.0000] http://herofort.blogspot.com/2011/05/websocket-server-blitz-max.html [10:07:00.0000] So, I think I am going to call it a day, but I will idle here. [10:07:01.0000] Thanks! [15:07:00.0000] Remember when tesco.com used application/xhtml+xml? [15:07:01.0000] Since they changed to text/html, it has developed an error: [15:07:02.0000] http://validator.w3.org/check?uri=http%3A%2F%2Fwww.tesco.com%2F&charset=%28detect+automatically%29&doctype=Inline&group=0 2011-05-22 [21:16:00.0000] back [07:52:00.0000] hmm, what happened to onformchange and onforminput ? [07:52:01.0000] Ms2ger: ↑ [07:53:00.0000] ah [07:53:01.0000] http://html5.org/tools/web-apps-tracker?from=5927&to=5928 [07:53:02.0000] Wasn't me :) [07:54:00.0000] :) [07:54:01.0000] just figured you might know [07:54:02.0000] anyway, it appears they are no more [07:54:03.0000] "Drop forminput and formchange events, and their associated machinery, since you can now put oninput and onchange everywhere which makes them far less compelling." [11:11:00.0000] How are you supposed to link to sections in anolis? Figure out the generated anchor and make an by hand? [11:12:00.0000] You probably want to give the section an id yourself [11:12:01.0000] Hmm. [11:45:00.0000] yeah just give an id and then link to it [11:45:01.0000] anolis won't override your ids 2011-05-23 [01:22:00.0000] did the LC poll reach quorum? [01:32:00.0000] hsivonen: i think i've lost my ability to vote [01:32:01.0000] abarth: the poll closed yesterday [01:33:00.0000] (i was more explaining why I didn't participate) [01:39:00.0000] hsivonen: do you have a clue as to how a questionmark appeared as favicon for your feed in my feedreader? [01:47:00.0000] zcorpan: is your browser your feed reader? [01:55:00.0000] hsivonen: yep, opera [02:09:00.0000] zcorpan: could it be that you looked at my rel="shortcut icon" test cases and Opera decided to use the favicon declared in those files for any other URLs from hsivonen.iki.fi that don't have an explicit favicon? [02:10:00.0000] hsivonen: it's indeed that icon. but opera doesn't show that icon when browsing your site, just the feed. weird [02:10:01.0000] /me files a bug [02:42:00.0000] hsivonen: the poll did reach quorum, yeah [03:02:00.0000] MikeSmith: for all documents? [03:04:00.0000] heycam|away: wait what happened with sequence and MessagePortArray? [03:47:00.0000] zcorpan: yeah, for all documents [03:47:01.0000] as far as I can tell [04:18:00.0000] any opinions on whether it would be a good idea to set up a mailbot that would automatically create HTML5 spec bugs from new messages posted to the public-html-comments list? [04:18:01.0000] that is, non-reply messages [04:18:02.0000] Hmm [04:19:00.0000] I can't see that it would be much of a problem as far as risk of malicious spam [04:19:01.0000] certainly not more than the current comment form [04:20:00.0000] because to post to that list, users basically need to first do the archive-approval step [04:21:00.0000] which essentially also works as an authentication mechanism to verify that they aren't spoofing an e-mail address [04:22:00.0000] the main problem I'd see is that some messages posted to that list may not actually be intended as comments [04:22:01.0000] but for those, it'd be easy enough to just move them to resolved-invalid [04:23:00.0000] so I guess it's a question of how much of them would be of that type [04:23:01.0000] what the signal-to-noise ratio would be [04:24:00.0000] I think there's little enough traffic to try it [04:24:01.0000] OK [04:25:00.0000] I can mess around with setting something up [04:26:00.0000] hmm [04:26:01.0000] as long as I'm not misunderstanding you [04:26:02.0000] Ms2ger: did you mean you think it's worth trying, or not worth trying? [04:27:00.0000] Worth trying [04:27:01.0000] ok [04:27:02.0000] thanks [04:29:00.0000] http://vkontakte.ru/photo63802_263023722?all=1 [04:30:00.0000] I do! [04:32:00.0000] heh [04:33:00.0000] I speak pidgin Javascript [05:40:00.0000] <_bga> gsnedders are you here? [05:46:00.0000] _bga: yes [05:47:00.0000] <_bga> what proseccors is supported by carakan [05:47:01.0000] <_bga> i mean jit [05:47:02.0000] <_bga> only x86? [05:49:00.0000] x86, x86_64, and ARM. [05:50:00.0000] <_bga> what about jvm bytecode? [05:50:01.0000] <_bga> as rhino [05:50:02.0000] (Well, x86 with x87 (though this has never been shipped), x86 with SSE2, x86_64, ARM with/without VFP) [05:50:03.0000] _bga: Why would we? [05:51:00.0000] <_bga> opera mobile [05:51:01.0000] _bga: Opera Mobile is C++ [05:53:00.0000] <_bga> i thought its java [05:53:01.0000] <_bga> ok [05:53:02.0000] _bga: We don't have two separate web browsers here. [05:54:00.0000] <_bga> only opera mini is java based? [05:54:01.0000] (So on x86 without SSE2 it'll disable the JIT entirely?) [05:55:00.0000] Philip`: (Yeah) [05:55:01.0000] _bga: Mini is just a thin-client, Presto runs on server. [05:55:02.0000] <_bga> understand [05:56:00.0000] (That seems unfortunate for some percents of users) [05:57:00.0000] Philip`: (Everyone, except Chrome which just refuses to run, does that.) [05:58:00.0000] Philip`: (And x86/x87 would effectively be yet another arch to support, maybe it'll be finished some day…) [05:59:00.0000] /me knows SpiderMonkey has x87 JIT but has no idea if/how it's exposed via the browser [06:00:00.0000] Philip`: It does? [06:00:01.0000] /me knows because it had bugs that broke his SpiderMonkey-embedding game for non-SSE2 users :-( [06:00:02.0000] Philip`: Are you sure it was doing that and not just using x87 for doubles in the interpreter? [06:01:00.0000] <_bga> i see superh is another popular processor [06:01:01.0000] <_bga> jscore and sm support it [06:01:02.0000] nanojit/Nativei386.cpp has paths for "if (!_config.i386_sse2)" etc [06:02:00.0000] Philip`: Heh, I thought it didn't. JSC certainly doesn't. [06:03:00.0000] _bga: SuperH seems to be being used less and less [08:03:00.0000] TabAtkins: you wondered about double line breaks when copying from View Source [08:03:01.0000] TabAtkins: double line breaks appear where a pre ends and a new one starts [08:04:00.0000] TabAtkins: since pre serializes as a block, there's an extra line break [08:05:00.0000] TabAtkins: there are multiple pre elements, because that way the bidi algorithm runs on smaller chunks [08:05:01.0000] TabAtkins: https://bugzilla.mozilla.org/show_bug.cgi?id=649613 [09:22:00.0000] hsivonen: Ah, that makes sense. Cool. [09:35:00.0000] /me wonders who exactly the people demanding more ways to submit LC feedback expect to build the system they are demanding [09:36:00.0000] /me suggests that the WG would be rather happy to adopt a system they build that meets their requirements [09:36:01.0000] In fact it really wouldn't require anyone to adopt anything [09:36:02.0000] It would just exist [09:37:00.0000] It could be linked to from useful places of course [10:24:00.0000] anyone any ideas on how I could determine the zoom level (window.innerWidth/screen.width) of a webpage on a handheld device that wasn't reporting innerWidth correctly? the device is a samsung galaxy s [10:29:00.0000] beowulf a trick? [10:30:00.0000] bga_: will try that [10:50:00.0000] hsivonen: I'm trying to figure out what part of the validator.nu schema allows data-* attributes [11:31:00.0000] What's a good reference for using @font-face in practice, like all the gotchas you need to be aware of? People are discussing it on wikitech-l. [11:33:00.0000] paul_irish has a post or two [11:33:01.0000] Ooh, Firefox 4.0 has an XSS filter that relies on a modal prompt to determine whether you can continue. :( [11:34:00.0000] "Load contents passed in the query part? The contents may contain malicious scripts. [11:34:01.0000] Continue only if you're aware that this may trigger a XSS attack and the author of this page does not take any responsibility for loading this contents" [11:34:02.0000] (I assume that's part of Firefox, dunno where else it could have come from) [11:34:03.0000] http://paulirish.com/2010/font-face-gotchas/ [11:34:04.0000] AryehGregor: I don't think that's part of Firefox 4 [11:35:00.0000] AryehGregor: or maybe it's CSP in action ? [11:35:01.0000] Then where could it have come from? [11:35:02.0000] /me can't give the name of the site, since it seems like it's supposed to be private) [11:35:03.0000] s/\)$// [11:35:04.0000] An XSS protection script in the page itself? [11:36:00.0000] /me doubts Firefox would have UI with such bad grammar, or that expects users to understand what "XSS" means [11:37:00.0000] MikeSmith: data-* is a pre-schema filter, IIRC [11:37:01.0000] hsivonen: ah [11:37:02.0000] It only triggered in Firefox. [11:37:03.0000] Maybe some kind of extension or something. [11:37:04.0000] A non-portable XSS protection script in the page itself? [11:38:00.0000] Maybe. :P [11:38:01.0000] NoScript? [11:40:00.0000] In IE6-8, using createStyleSheet and then setting styleElem.styleSheet.cssText to a text value that includes a @font-face declaration going into will crash IE6-8. [11:40:01.0000] sweet [11:40:02.0000] The only extension I have active is Firebug 1.7.1. [11:41:00.0000] hope they un-break ctrl+shift+c soon [11:41:01.0000] /me finds util/classes/nu/validator/xml/dataattributes/DataAttributeDroppingContentHandlerWrapper.java [11:43:00.0000] hsivonen: wondered how you did that. :) nice [11:44:00.0000] good thing there are no special restrictions on the values of data-* attributes [12:08:00.0000] Does Opera not support ? data:text/html,
          abc
          [13:29:00.0000] What's a normal sort of resolution for a high-res image used in printing? [13:29:01.0000] (I want something reasonable for this example I'm writing.) [13:34:00.0000] 300dpi? [13:34:01.0000] That's pretty much what I was thinking. kk. [13:35:00.0000] 300dpi is the low end, FWIW [13:37:00.0000] Seems to depend somewhat on how the printer works - if it's doing colours by dithering then it probably needs higher print resolution than the original image resolution [13:38:00.0000] 300dpi works for the example, which is just to demonstrate how you can use 'image-resolution' to embed a high-res image into a page so it looks good both on-screen and in print. [13:47:00.0000] zcorpan, it sounds like MessagePortArray should be a read only array (http://dev.w3.org/2006/webapi/WebIDL/#dfn-read-only-array) instead of a sequence, since you don't want a new object being created every time you access the property [13:47:01.0000] zcorpan, but I'll reply on the list later this morning once I've woken up [13:51:00.0000] heycam: ok [13:52:00.0000] didn't we already talk about that and didn't i already fix it? [13:52:01.0000] or was that something else that i fixed [13:52:02.0000] hm [13:53:00.0000] If only you had some sort of history of your revisions that you could search for this sort of thing. [13:53:01.0000] it's quite possible MessagePortArray was the reason that bug was filed in the first place -- at least, there was something that was in html you suggested removing sequence attributes because of [13:54:00.0000] i think it was some microdata thing [13:54:01.0000] i guess i missed this one [14:37:00.0000] TabAtkins, wouldnt some CC licensed icons work? [14:38:00.0000] well, you can always just go with triangle, square, dot, circle :> [14:38:01.0000] aho: For my twitter post? Sure. But I'm just building my own right now. [14:38:02.0000] http://kaioa.com/k/test/svgsprites/index.html [14:38:03.0000] works for me ;> [14:41:00.0000] http://youtu.be/IPwCH8WQ8Qo <- webkit's input type=color thingy [14:41:01.0000] interesting that they added a color picker [14:42:00.0000] eh.. some pipette thingy [14:45:00.0000] ok i fixed that CORS thing from last week [14:45:01.0000] anything else urgent for me to work on? [14:45:02.0000] or should i go back to the regular pile of feedback [14:46:00.0000] dglazkov: yt? [14:46:01.0000] Hixie: I certainly am [14:46:02.0000] dglazkov: re the scoped thing, keep in mind that there's more to scoping