| 05:40 | <annevk> | Hixie, yeah that's pretty great, but they haven't removed the init* methods yet |
| 05:40 | <annevk> | Hixie, I thought about writing a better spec for the remaining part of DOM3Events, but I don't really have the time |
| 05:41 | <annevk> | or well, I could make the time, but there are other things important too |
| 05:41 | <annevk> | Hixie, (WebKit hasn't removed init* methods yet, they did add the constructors) |
| 06:46 | <annevk> | Hixie, if you're still up, seems it is time to standardize <canvas> dashed strokes |
| 06:47 | <Hixie> | yeah, already noted that :-) |
| 06:47 | <Hixie> | annevk: did you file bugs on removing the init* methods? |
| 06:55 | <annevk> | Hixie, I asked the guy implementing the constructor methods, he said it was for compat... |
| 06:55 | <annevk> | Hixie, I guess I should file a bug at some point |
| 06:55 | <Hixie> | file a bug and cc me, the new stuff has no need for those methods even for compat. |
| 06:58 | <annevk> | btw, do you have plans to help spec out a new XBL or are you hoping someone else will? |
| 06:58 | <Hixie> | isn't that what dg and crew are doing? |
| 06:58 | <Hixie> | nobody seems interested in implementing xbl2 |
| 06:58 | <Hixie> | even with the html changes |
| 06:58 | <Hixie> | so... |
| 07:00 | <annevk> | dg said there would still be a part in HTML |
| 07:00 | <Hixie> | not sure what that means |
| 07:01 | <annevk> | me neither, the overall design is somewhat unclear to me |
| 07:03 | <hsivonen> | C++ is weirder than the Web: https://blog.mozilla.com/respindola/2011/09/25/interesting-c-snippet/ |
| 07:03 | <Hixie> | annevk: i haven't really been following it |
| 07:05 | <annevk> | k |
| 08:02 | <zcorpan> | you have to use URLs as identifiers when there's money involved? |
| 08:09 | <annevk> | ? |
| 08:12 | <zcorpan> | some email... |
| 10:02 | <annevk> | I tried to implement the new exception model in DOM4 |
| 10:02 | <annevk> | http://dvcs.w3.org/hg/domcore/raw-file/tip/Overview.html#concept-throw |
| 10:02 | <annevk> | Review please! |
| 10:03 | <annevk> | This is done roughly per the suggestion in http://www.w3.org/Bugs/Public/show_bug.cgi?id=10623#c14 and is what all specifications will end up using. |
| 10:35 | <hsivonen> | sad. it seems that getSVGDocument() lives in an interface all by itself. |
| 10:36 | <hsivonen> | as opposed to being part of one of the other interfaces e.g. iframes implement |
| 10:36 | <annevk> | that's prolly not required for legacy |
| 10:36 | <annevk> | it's kind of silly though given contentDocument |
| 10:37 | <annevk> | leftover from the Adobe SVG plugin that is now dead? |
| 11:27 | <hsivonen> | Why does the spec for insertAdjacentHTML describe WebKit's exception behavior instead of IE's? |
| 11:28 | <annevk> | I guess Hixie copied it from them maybe? |
| 11:28 | <annevk> | Nobody has access to IE anymore these days |
| 11:29 | <hsivonen> | shouldn't we be speccing IE's behavior considering that it's their API |
| 11:29 | <hsivonen> | and Opera doesn't throw, either, when the parent is the document object |
| 11:30 | <hsivonen> | (in fact, Opera doesn't throw even when IE throws) |
| 11:30 | <hsivonen> | (when the position string is bogus) |
| 11:31 | <hsivonen> | I shouldn't test other browsers when fixing bugs. that tends to reveal spec bugs, which prolongs my bug fixing time |
| 11:41 | <annevk> | hsivonen, revealing spec bugs is a bonus |
| 11:43 | <hsivonen> | maybe I should check if IE6 throws |
| 11:45 | <hsivonen> | hooray. IE6 creates a bogus DOM |
| 11:48 | <hsivonen> | yet more proof that IE9's modes aren't perfect copies of legacy IEs |
| 11:48 | <hsivonen> | IE7 creates a bogus DOM, too |
| 11:49 | <hsivonen> | as does IE8 |
| 11:50 | <hsivonen> | I guess I should try IE9 again and IE10... |
| 11:57 | <mpt> | Bogus DOM would be a great name for a band |
| 11:58 | <hsivonen> | we should keep a list of WHATWG band names. there are already several possibilities, IIRC |
| 11:59 | <hsivonen> | whoa. IE9 creates a bogus DOM, too. it just doesn't cause weird rendering as in IE6 |
| 11:59 | <jgraham> | mpt: In which genre? |
| 12:00 | <annevk> | big surprise that the IE versioning scheme doesn't work in practice |
| 12:00 | <hsivonen> | well, maybe it does, since IE9 creates a bogus DOM, too. I just failed to notice it before |
| 12:03 | <zcorpan> | matjas: B\&W\? and B\26 W\3F should be valid for css in http://mothereffingunquotedattributes.com/ |
| 12:03 | <mpt> | jgraham, folk or steampunk, perhaps, à la The Scarring Party |
| 12:06 | <mpt> | (IE6 could be classed as steampunk these days, no?) |
| 12:07 | <hsivonen> | what should one do when Windows 7 doesn't show the task bar? |
| 12:07 | <hsivonen> | move to another VM snapshot and forget about it? |
| 12:07 | <zcorpan> | run explorer.exe? |
| 12:08 | <zcorpan> | or kill it in taskmgr? |
| 12:10 | <hsivonen> | zcorpan: neither of those helped |
| 12:12 | <hsivonen> | huh? isn't the latest IE10 PP available for Windows 7? |
| 12:12 | jgraham | wonders if the Mozilla javascript team have ever considered forming a rock and roll band called Spidermonkey and the GC Jitters |
| 12:13 | <zcorpan> | hsivonen: then the standard answer is "restart your computer" |
| 12:15 | <hsivonen> | zcorpan: restarting the VM didn't help |
| 12:15 | <matjas> | zcorpan: thanks, filed https://github.com/mathiasbynens/mothereffingunquotedattributes/issues/3 |
| 12:19 | <zcorpan> | hsivonen: :( |
| 12:48 | <matjas> | zcorpan: is `B\0026 ` valid too? and `B\026 `? and `B\0000026 `? |
| 12:49 | <matjas> | it all seems to be |
| 12:58 | <hsivonen> | where does the "replacement enabled" navigation concept come from? that is, which browser is the spec describing there? |
| 13:18 | <zcorpan> | matjas: escapes are valid |
| 13:22 | <matjas> | needs moar value.replace(/\\([0-9A-F]{2,}) /g, '').replace(/\\./g, '') |
| 13:25 | <zcorpan> | i'm not sure that's entirely accurate |
| 13:26 | <zcorpan> | ecsapes have a max length of 6 chars i think |
| 13:26 | <zcorpan> | the space is optional and can be any css space char |
| 13:26 | <zcorpan> | also, a single escape is a valid identifier (but the empty string isn't) |
| 13:27 | <zcorpan> | also, \\ is a valid identifier |
| 13:42 | <matjas> | zcorpan: clear your appcache and go to http://mothereffingunquotedattributes.com/#B%5C%26W%5C%3F and http://mothereffingunquotedattributes.com/#B%5C26%20W%5C3F%20 |
| 13:42 | <matjas> | thanks again! |
| 13:43 | <matjas> | ooh just saw your latest messages |
| 13:43 | <matjas> | damn. |
| 13:45 | <zcorpan> | matjas: :) |
| 13:55 | <matjas> | zcorpan: not sure if it means anything, but `a[href=B\0000026] { }` (7-char escape) validates as per http://jigsaw.w3.org/css-validator/#validate_by_input |
| 13:59 | <zcorpan> | sure |
| 13:59 | <zcorpan> | it's equivalent to a[href=B\000002 6] |
| 14:00 | <matjas> | oh snap |
| 14:02 | <hsivonen> | what should I remember to test for HTML in XHR? |
| 14:02 | <hsivonen> | * character encoding |
| 14:02 | <hsivonen> | * noscript |
| 14:02 | <hsivonen> | * script |
| 14:03 | <hsivonen> | * that iframes don't load stuff |
| 14:03 | <matjas> | zcorpan: are you getting all that from `escape {unicode}|\\[^\n\r\f0-9a-f]` at http://www.w3.org/TR/CSS2/syndata.html#value-def-identifier? |
| 14:03 | <hsivonen> | * that imgs don't load stuff |
| 14:03 | <hsivonen> | anything else? |
| 14:03 | <zcorpan> | http://www.w3.org/TR/CSS21/syndata.html#escaped-characters "at most six hexadecimal digits" |
| 14:03 | <hsivonen> | * DOMContentLoaded and load relative to the XHR events |
| 14:05 | <zcorpan> | hsivonen: why don't imgs load stuff? |
| 14:05 | <zcorpan> | doesn't it with XHTML responseXML? |
| 14:05 | <zcorpan> | what about <video>? |
| 14:06 | <hsivonen> | zcorpan: oh, maybe they do and maybe I shouldn't bother to test that |
| 14:07 | <hsivonen> | hmm. actually I should test that |
| 14:08 | <annevk> | images load? |
| 14:08 | <annevk> | hmm |
| 14:08 | <hsivonen> | annevk: I don't know if they do |
| 14:08 | <annevk> | I'm not a 100% sure either |
| 14:09 | <hsivonen> | I should probably make sure I turn off speculative loading for scripts, style sheets and images |
| 14:09 | <annevk> | there's some stuff that depends on a browsing context being there, not sure if this is such a situation |
| 14:10 | <annevk> | matjas, you forgot use cases |
| 14:11 | <annevk> | matjas, you are saying "I want X, Y, Z", we want "I run into problem A, B, C" |
| 14:13 | <annevk> | matjas, fwiw, submit() used to do validation, but we made it submit because of deployed content |
| 14:14 | <annevk> | matjas, we didn't really have a good use case for programmatic submission that includes validation |
| 14:23 | <hsivonen> | annevk: why does XMLHttpRequestEventTarget exist as a separate interface? |
| 14:23 | <hsivonen> | oh. for uploads? |
| 14:23 | <annevk> | yes |
| 14:24 | <matjas> | annevk: i’m confused; was that @ me? |
| 14:24 | <hsivonen> | hmm. ProgressEvent isn't linkified. arbitrary link breakage when interface deps cross spec boundaries :-( |
| 14:25 | <annevk> | matjas, I guess Matias is not you lol |
| 14:26 | <annevk> | sorry |
| 14:26 | <matjas> | haha np |
| 14:26 | <annevk> | hsivonen, in http://dev.w3.org/2006/webapi/XMLHttpRequest-2/#interface-xmlhttprequest it is... |
| 14:28 | <matjas> | zcorpan: a[href=B\0]{} is valid according to the CSS validator even though the spec says “the backslash is followed by at most six hexadecimal digits (0..9A..F), which stand for the ISO 10646 ([ISO10646]) character with that number, which must not be zero. (It is undefined in CSS 2.1 what happens if a style sheet does contain a character with Unicode codepoint zero.)” |
| 14:29 | <matjas> | has this changed since CSS 2.1? |
| 14:29 | <zcorpan> | dunno |
| 14:29 | <zcorpan> | i guess that's a bug in the validator |
| 14:30 | <annevk> | that \0 is undefined does not make it invalid per se |
| 14:30 | <zcorpan> | TabAtkins: has handling of null chars been defined in css lately? |
| 14:30 | <zcorpan> | annevk: is says "must not be zero" |
| 14:34 | <annevk> | oh hmm |
| 14:43 | <hsivonen> | is there a way to set load and DOMContentLoaded event handlers on responseXML without racing with the network_ |
| 14:43 | <hsivonen> | ? |
| 14:44 | <annevk> | per spec you cannot see it dispatch |
| 14:45 | <annevk> | Document is returned when the fetch and parsing is done |
| 14:45 | <hsivonen> | annevk: what browser is that based on? |
| 14:46 | <annevk> | I think all of them? it's been a while |
| 14:46 | <hsivonen> | oh |
| 14:46 | <hsivonen> | maybe I misread the Gecko sources the other day |
| 14:47 | <hsivonen> | annevk: indeed. Thanks! |
| 14:47 | <hsivonen> | this makes stuff easier! |
| 14:59 | <annevk> | does HTML define there's a single Text node in http://software.hixie.ch/utilities/js/live-dom-viewer/saved/1149 ? |
| 15:08 | <zcorpan> | annevk: i thought the parser only kept track of parser-inserted text nodes |
| 15:09 | <hsivonen> | zcorpan: in Gecko, the parser doesn't track which text nodes are parser-inserted |
| 15:09 | <zcorpan> | oh |
| 15:09 | <zcorpan> | then nevermind me :) |
| 15:10 | <hsivonen> | well, that doesn't answer the question about what the spec says |
| 15:27 | <hsivonen> | I just realized I have to test this HTML in XHR stuff with sync XHR, too. :-( |
| 15:52 | <zewt> | if only we could find a way to usefully deprecate horrible legacy things like sync xhr in the main thread |
| 15:52 | <zewt> | like maybe browsers could play the macarana whenever a sync xhr call is made |
| 15:54 | <jgraham> | How else am I suppose to write tests that intentionally halt script execution? |
| 15:54 | <Philip`> | zewt: Yeah, it's terrible when authors use convenient APIs |
| 15:55 | <Philip`> | jgraham: You could run tests in a special browser mode that disables the macarena |
| 15:55 | <zcorpan> | hsivonen: can't the prescan be turned off also? |
| 15:55 | <zewt> | "convenient" isn't the word I'd use to describe synchronous APIs in the UI thread, heh |
| 15:56 | <zcorpan> | hsivonen: does BOM influence encoding choice for responseText? |
| 15:57 | <hsivonen> | zcorpan: technically, the prescan could be turned off. but why? |
| 15:57 | <hsivonen> | zcorpan: I haven't tested the BOM in existing browsers |
| 15:58 | <zewt> | dropping bombs on my poor plain text files :( |
| 15:58 | <zewt> | (would like to drop a bomb on whoever came up with the idea of stapling a header onto a headerless file format) |
| 15:58 | <Philip`> | zewt: Plain text is a horrible legacy thing |
| 15:59 | <pdr> | The spec (http://goo.gl/6UNDL) says selectionDirection of textarea/input is platform dependent, and effectively says on Macs it defaults to 'none' when selecting with the mouse. I think we should make the wording much more strict so selectionDirection is actually usable across platforms.. anyone have thoughts on that? |
| 16:00 | <GlitchMr> | I prefer "Content-type: text/plain;charset=utf-8"... |
| 16:00 | <GlitchMr> | This way, charset information isn't in file itself. |
| 16:01 | <Philip`> | pdr: That would presumably make it incompatible with platform conventions, which sounds bad for users |
| 16:01 | <zcorpan> | hsivonen: the meta doesn't work today when people do innerHTML = responseText |
| 16:01 | <pdr> | It appears there is no way to tell which way the user selects the text... for instance, where the cursor should be after a selection is made (whether it's at the beginning or end of the selection). AFAIK, that's all selectionDirection is good for. |
| 16:02 | <zewt> | gah |
| 16:02 | <zewt> | a web editor that catches shift-delete and turns it into delete instead of cut : | |
| 16:02 | <zcorpan> | hsivonen: i think responseText and responseXML should use the same encoding... but maybe the xml decl already broke that? |
| 16:06 | <hsivonen> | zcorpan: WebKit considers <meta> today when reading responseText |
| 16:06 | <hsivonen> | for text/html |
| 16:07 | <hsivonen> | zcorpan: Gecko considers the XML decl for XML types when reading responseText (based on reading Gecko source; I didn't test) |
| 16:09 | <zcorpan> | hsivonen: ok. then i guess it makes sense to do the scan (but also for responseText) |
| 16:11 | <hsivonen> | zcorpan: I intend to keep the charset for responseText and responseXML in sync for both HTML and XML |
| 16:12 | <zcorpan> | good :) |
| 16:13 | <hsivonen> | though there's a complication with streaming responseText |
| 16:13 | <hsivonen> | I guess I need to defer that until the HTML or XML parser has made its decision |
| 16:14 | <zcorpan> | does webkit scan 1024 bytes? |
| 16:16 | <hsivonen> | zcorpan: I haven't verified, but I'd be surprised if it didn't stop at 1024 bytes |
| 16:17 | <hsivonen> | though things can be surprising on the Web |
| 16:17 | <annevk> | I'm quite pleased with the HTML/XML TF |
| 16:18 | <annevk> | The document we ended up with is quite sensible. Having said that, whether anyone (apart from the TAG) still cares is another matter |
| 16:20 | <hsivonen> | hmm. I might not make it to the HTML/XML TF telecon tomorrow |
| 16:38 | <hsivonen> | http://lists.w3.org/Archives/Public/www-tag/2011Sep/0035.html |
| 16:41 | <annevk> | in case anyone cares: http://www.w3.org/TR/progress-events/ |
| 16:41 | <annevk> | CR now |
| 16:42 | <zewt> | hopefully FileReader doesn't make every spec that uses it evil, heh |
| 16:44 | <hsivonen> | annevk: congrats for CR |
| 16:45 | <annevk> | zewt, still not sure what the better behavior should be |
| 16:46 | <annevk> | we have some room to make changes here |
| 16:46 | <zewt> | well, i was assuming XHR was stuck and arguing that if necessary, FileAPI should break from XHR (if FileAPI follows XHR here, everything else forever will) |
| 16:46 | <zewt> | if XHR itself can be improved, all the better |
| 16:46 | <GlitchMr> | What tests? |
| 16:47 | <GlitchMr> | There will be at least two interoperable implementations passing all test cases in the test suite for this specification. An implementation is to be available (i.e. for download), shipping (i.e. not private), and not experimental (i.e. intended for a wide audience). The working group will decide when the test suite is of sufficient quality to test interoperability and will produce implementation reports (hosted together with the test sui |
| 16:47 | <GlitchMr> | te). |
| 16:47 | <GlitchMr> | Makes sense... |
| 16:47 | <zewt> | the ideal behavior, IMO, would be to forbid send() until the progress sequence (and any readystatechanges) have finished being dispatched, which is what I recommended for FileAPI |
| 16:48 | <zewt> | (if you want to reuse the object to send another request/read another block, stuff it in a setTimeout(0)) |
| 16:49 | <zewt> | that way, no matter how you listen to the events, loadstart/loadend are always cleanly paired with no overlap |
| 16:50 | <zewt> | (modulo stopPropagation or changes to the tree, if any, of course) |
| 16:51 | <zewt> | (neither of which apply to these APIs) |
| 16:51 | <annevk> | GlitchMr, isn't there a link? |
| 16:51 | <GlitchMr> | Yeah, but I fail at finding those tests in directory structure... |
| 16:51 | <zewt> | (well, with these APIs, there's still stopImmediatePropagation, but anyway) |
| 16:51 | <GlitchMr> | Is it /ProgressEvents/tests/submissions/Ms2ger/? |
| 16:52 | <GlitchMr> | That's only thing I've found in directory structure which somewhat reminds tests... |
| 16:52 | <annevk> | yeah |
| 16:52 | <GlitchMr> | But then, where is <script src=/resources/testharness.js></script> |
| 16:53 | <annevk> | GlitchMr, might be more useful to look here: http://w3c-test.org/webapps/ProgressEvents/tests/submissions/Ms2ger/ |
| 16:53 | <GlitchMr> | My browser fails at them. |
| 16:53 | <GlitchMr> | I'm not surprised. |
| 16:54 | <GlitchMr> | It got CR not long time ago. |
| 16:54 | <GlitchMr> | http://testcases.glitchmr.pl/html/tag-nesting-003.html |
| 16:54 | <GlitchMr> | I find this behavior weird... |
| 16:54 | <annevk> | most browsers support ProgressEvent to some extent, but constructing them not really... and that's mostly what is being tested |
| 16:55 | <GlitchMr> | Pass - Interface objects properties should not be Enumerable |
| 16:55 | <GlitchMr> | actually, yeah... |
| 16:55 | <GlitchMr> | oh wait... that's expected if browser doesn't support it :P. |
| 17:00 | <GlitchMr> | http://hixie.ch/tests/adhoc/http/content-type/css/001.html |
| 17:00 | <GlitchMr> | ... for me this test is invalid |
| 17:00 | <GlitchMr> | It uses Quirks Mode... |
| 17:00 | <GlitchMr> | Yeah, it's expected that text/plain shouldn't be interpreted as CSS, but that page used Quirks Mode... |
| 17:50 | <smaug____> | is 'attribute' reserved word in webidl? |
| 18:21 | <jarek> | Hi |
| 18:21 | <jarek> | when using HTML5 application cache |
| 18:21 | <Jusan> | hi |
| 18:21 | <jarek> | why both Firefox and Chrome are caching the manifest file? (thus preventing the app from getting any further updates) |
| 18:21 | <jarek> | is this a bug in the browsers or is it by design? |
| 18:30 | <Jusan> | hi |
| 18:58 | <timeless> | hsivonen: ping |
| 19:01 | <timeless> | annevk: ping (html/xml tf) |
| 19:58 | <Ms2ger> | smaug____, I assume 'attribute' would be reserved, yes |
| 19:58 | <Ms2ger> | You can escape as '_attribute' |
| 20:00 | <smaug____> | Ms2ger: do you mean _attribute would be interpret as attribute? |
| 20:00 | <Ms2ger> | Right |
| 20:00 | <smaug____> | Ms2ger: where is such escaping defined? |
| 20:00 | <Ms2ger> | "For all of these constructs, the identifier is the value of the identifier token with any single leading U+005F LOW LINE ("_") character (underscore) removed." |
| 20:01 | <Ms2ger> | "A leading "_" is used to escape an identifier from looking like a reserved word so that, for example, an interface named interface can be defined. The leading "_" is dropped to unescape the identifier." |
| 20:01 | <Ms2ger> | http://dev.w3.org/2006/webapi/WebIDL/#idl-names |
| 20:05 | <smaug____> | Ms2ger: thanks |
| 20:06 | <Ms2ger> | Np |
| 20:22 | <annevk> | got to love how John Cowan keeps playing ignorant about lenient XML |
| 20:23 | <annevk> | timeless, seems you ran out of time |
| 20:23 | <annevk> | GlitchMr, quirks mode is slowly becoming standardized, testing it is fine |
| 20:24 | <GlitchMr> | annevk, but as far I know, CSS could use text/plain in Quirks Mode... |
| 20:24 | <GlitchMr> | Or it was changed? |
| 20:35 | <Hixie> | TabAtkins: what's the status with @scoped? is it ok to start speccing? |
| 20:36 | <GlitchMr> | <style scoped>? |
| 20:38 | <annevk> | GlitchMr, that page is not in quirks mode |
| 20:39 | <GlitchMr> | "CSS1Compat" |
| 20:39 | <GlitchMr> | ... |
| 20:40 | <GlitchMr> | I through that "HTML 4.0" DOCTYPE activated Quirks Mode... |
| 20:40 | <GlitchMr> | Oh right, Strict |
| 20:40 | <GlitchMr> | fail |
| 20:40 | <annevk> | Hixie, why make TTML work? |
| 20:41 | <annevk> | lets not optimize for TTML |
| 20:41 | <GlitchMr> | That was case in Transitional... |
| 20:42 | <GlitchMr> | OK, but I think it's in Almost Standards Mode, but I think that only change it has is img {display: block}... |
| 20:42 | <GlitchMr> | Unless there is more changes? |
| 20:42 | <Hixie> | annevk: if we're explicitly deciding to not make TTML work at all, that's fine by me, but then we should make sure that's clear to everyone and we don't end up with, e.g., IE implementing some crazy sniffing for that case. |
| 20:43 | <GlitchMr> | They have already removed support for text/plain in websites... |
| 20:44 | <annevk> | IE implementing crazy sniffing and nobody interoperating because nobody else wants TTML seems like an okay outcome to me |
| 20:45 | <Hixie> | not to me |
| 20:47 | <Hixie> | abarth: what's the url for the mimesniff stuff that is most likely to remain up to date? |
| 20:48 | <Hixie> | (http://www.w3.org/Bugs/Public/show_bug.cgi?id=12929) |
| 20:48 | <GlitchMr> | What this thing actually is? Because specification isn't exactly clean for me? Something like CSS with XML-like markup? |
| 20:51 | <GlitchMr> | Oh, I see... It's something like subtitles format. |
| 20:53 | <GlitchMr> | Maybe it won't happen, but probably (at least for me), nearly nobody will use it... |
| 20:54 | <jwalden> | GlitchMr: square one reference? |
| 20:55 | <GlitchMr> | jwalden, I don't know what is square one :P |
| 20:55 | <jwalden> | GlitchMr: guess not, then |
| 20:55 | <jwalden> | GlitchMr: there was a TV series, Square One, that had a Mr. Glitch in it |
| 20:55 | <jwalden> | http://en.wikipedia.org/wiki/Mathman |
| 20:55 | <GlitchMr> | Nope |
| 20:55 | <annevk> | Hixie, why do you need interoperability for a proprietary feature? |
| 20:56 | <GlitchMr> | But I find it amusing that I made similar nick without knowing about that... |
| 20:56 | <jwalden> | indeed |
| 20:56 | <Hixie> | annevk: TTML is hardly a "proprietary feature", however we may personally dislike it |
| 20:57 | <Hixie> | annevk: nor is TTML the only type that would be affected by this -- what if a browser wants to support SRT natively? |
| 20:57 | <annevk> | you cannot implement SRT because there is no sane spec |
| 20:57 | <GlitchMr> | That nick is just because I tried inserting some random stuff to get account on Gmail. It has so many accounts... |
| 20:57 | <annevk> | and if one vendor implements a feature it's proprietary imo |
| 20:57 | <GlitchMr> | lol |
| 20:57 | <Hixie> | there was no sane spec for HTML For 15 years, that didn't stop anyone |
| 20:57 | <annevk> | e.g. I would consider Opera's support for DOM Level 3 Load & Save proprietary |
| 20:58 | <timeless> | annevk: ping html-xml |
| 20:58 | <timeless> | Hixie: do you know anyone from google involved in gmail/calendar? |
| 20:58 | <annevk> | timeless, hey you're back; pong |
| 20:58 | <timeless> | yeah, sorry, i was rewiring my office |
| 20:58 | <timeless> | hixie: ... my webkit based browser won't let me click on the gmail account chooser, and i can't type into calendar fields :( |
| 20:59 | <timeless> | annevk: so, any idea what happened to http://lists.w3.org/Archives/Public/public-html-xml/2011Aug/0049.html ? |
| 20:59 | <Hixie> | timeless: not personally, no |
| 20:59 | <Hixie> | timeless: sounds like a bug in the webkit you're using |
| 20:59 | <timeless> | Hixie: i don't suppose google has a "test site" for implementers? |
| 21:00 | <timeless> | some easy way to test the features that they require to have working.. |
| 21:00 | <AryehGregor> | timeless, I get the impression they use browser-specific hacks and just ignore unsupported browsers. |
| 21:00 | <AryehGregor> | (who doesn't?) |
| 21:00 | <timeless> | AryehGregor: yeah well |
| 21:00 | <timeless> | um... |
| 21:00 | <GlitchMr> | I find it interesting how browsers tried to ignore errors. Usually SGML parsers fail at errors. |
| 21:00 | <AryehGregor> | (for complicated sites, anyway) |
| 21:00 | <Hixie> | timeless: not to my knowledge, but this isn't my area of expertise |
| 21:01 | <annevk> | timeless, not sure, I asked Norm |
| 21:01 | <timeless> | annevk: thanks |
| 21:01 | <timeless> | annevk: of course, i won't be around thursday/friday.. |
| 21:01 | <timeless> | enjoy oslo btw :) |
| 21:01 | <annevk> | timeless, they're in the queue |
| 21:01 | <annevk> | heh thanks |
| 21:01 | <timeless> | annevk: is the queue public? |
| 21:01 | <annevk> | nope |
| 21:02 | <annevk> | I reckon it's how Norm refers to his backlog :) |
| 21:02 | <timeless> | what confused me is that the latest update http://lists.w3.org/Archives/Public/public-html-xml/2011Sep/0005.html seemed to be based on a september 1 meeting |
| 21:02 | <abarth> | Hixie: http://mimesniff.spec.whatwg.org/ |
| 21:03 | <timeless> | which by my reckoning would be after my post :) |
| 21:03 | <timeless> | GlitchMr: so... browsers evolved to make users happy |
| 21:04 | <GlitchMr> | I think that IE6 is good browser through :P. |
| 21:04 | <timeless> | http://en.wikipedia.org/wiki/File:Yellow_screen_of_death.png |
| 21:04 | <GlitchMr> | (compared to browsers made in 2001) |
| 21:04 | <timeless> | yellow screens of death did not make users happy |
| 21:04 | <GlitchMr> | timeless, that's XHTML |
| 21:04 | <timeless> | so? |
| 21:05 | <timeless> | html was implemented for normal humans |
| 21:05 | <timeless> | and partially implemented by people who did not have access to the $$$ based SGML spec |
| 21:05 | <timeless> | (whose printing would kill rather large trees..) |
| 21:05 | <GlitchMr> | Makes sense... |
| 21:06 | <Hixie> | abarth: k |
| 21:06 | <Hixie> | oooh, prettified html |
| 21:06 | <GlitchMr> | Technically, if something would go differently we could have values between `` and tags between {}... |
| 21:06 | <timeless> | GlitchMr: you mean rtf? |
| 21:07 | <GlitchMr> | What RTF does? |
| 21:07 | <timeless> | abarth: ooh, indeed, shiny +1 |
| 21:07 | <GlitchMr> | Oh I see, {}... |
| 21:07 | <GlitchMr> | That's just coincidence. |
| 21:07 | <timeless> | abarth: > Otherwise, if the octets in s starting at pos match any of the sequences of octets in the first column of the following table, then the user agent MUST follow the steps given in the corresponding cell in the second column of the same row. | |
| 21:07 | <GlitchMr> | There is a lot of such pairs on keyboard. |
| 21:07 | <timeless> | what's the stray character at the end of that doing? |
| 21:08 | <timeless> | GlitchMr: please note that not all non en-us keyboards have easy access to things like {} |
| 21:08 | <GlitchMr> | There is also () and []... |
| 21:08 | <GlitchMr> | timeless, oh right |
| 21:08 | <timeless> | http://stackoverflow.com/questions/311244/keyboard-layout-for-international-programmers |
| 21:08 | <abarth> | timeless: probably a holdover from the IETF XML |
| 21:08 | <abarth> | timeless: will fix |
| 21:09 | <GlitchMr> | I'm using polish keyboard. Polish keyboard is like English keyboard with additional characters accessed using Alt Gr + letter... |
| 21:09 | <timeless> | abarth: gimme 5-10 mins to read the actual document.. |
| 21:09 | <GlitchMr> | But yeah, I'm aware of that |
| 21:09 | <GlitchMr> | Right... |
| 21:09 | <timeless> | GlitchMr: i've spent too much time dealing w/ various intl issues |
| 21:09 | <timeless> | Hixie / abarth : the ToC feels double spaced, is that normal? |
| 21:10 | <abarth> | timeless: fixenated |
| 21:10 | <timeless> | abarth: also, would you mind quoting your attributes in source? |
| 21:10 | <abarth> | timeless: i think normally there are subsections |
| 21:10 | <timeless> | things like class=no-num or href=#web-data scare me |
| 21:10 | <timeless> | it's easier if you just quote all attributes :) |
| 21:11 | <abarth> | i'm not sure its such a big deal |
| 21:11 | <abarth> | HTML can handleit |
| 21:11 | <timeless> | also, i generally recommend `<span ...>x</span> ` over `<span ...>x </span>` <- i.e. trailing space outside of span (see toc) |
| 21:11 | <timeless> | <p>Many web servers supply incorrect Content-Type header fields with their HTTP |
| 21:11 | <abarth> | timeless: i'm just emulating the HTML spec in that regard |
| 21:12 | <timeless> | can you mark up `Content-Type` in something which results in roughly "typewriter" font? |
| 21:12 | timeless | doesn't care if it's <tt> or <pre> or something magically styled |
| 21:12 | <GlitchMr> | For my private sites, I ignore stuff like direction rtl. I don't care about Arabic :P. |
| 21:12 | <timeless> | GlitchMr: i care about Hebrew |
| 21:12 | <GlitchMr> | I can find it annoying. |
| 21:12 | <timeless> | you're free to not care about Arabic, you still need to get RTL right :) |
| 21:13 | <GlitchMr> | RTL is annoying when making CSS layouts |
| 21:13 | <abarth> | timeless: done |
| 21:13 | <timeless> | abarth: possibly ... s/user agents/User Agents/ as in: |
| 21:13 | <timeless> | responses. In order to be compatible with these servers, user agents consider |
| 21:14 | <GlitchMr> | Personally, I find Polish keyboard very easy to use for programming. It doesn't annoy while programming and I could easily use Polish letters when I want :P. |
| 21:14 | <timeless> | > Without a clear specification of how to "sniff" the media type, each user agent implementor was forced to reverse engineer the behavior of the other user agents and to develop |
| 21:14 | <timeless> | abarth: s/the other/other/ -- there are some UAs who were ignored when the sniffing of a given UA was developed :) |
| 21:15 | <timeless> | > their own algorithm |
| 21:15 | <timeless> | i'm not sure if `algorithm` here belongs in singular or plural, i got distracted :) |
| 21:15 | <GlitchMr> | The only thing which annoys me is ~ which does something completely useless for me in Windows... |
| 21:15 | <timeless> | > an HTTP response to be interpreted as one media type but some user agents interpret the responses as another media type. |
| 21:16 | <timeless> | s/responses/response/ |
| 21:16 | <timeless> | (agreement with first part) |
| 21:16 | <abarth> | all should be fixed |
| 21:16 | <abarth> | pls reload to see if you like |
| 21:16 | <timeless> | > However, if a user agent does interpret a low-privilege media type, such as image/gif, as a high-privilege media type, such as text/html, the user agent has created a privilege escalation vulnerability in the server. |
| 21:16 | <timeless> | abarth: i'll reload at the end... |
| 21:17 | <timeless> | s/, the user agent/, then the user agent/ |
| 21:17 | <timeless> | > This document describes a content sniffing algorithm that carefully balances the compatibility needs of user agent implementors with the security constraints. |
| 21:17 | <timeless> | `the security constraints` is problematic, i don't think `the` references anything |
| 21:17 | <timeless> | so either drop `the`, or provide a reference :/ |
| 21:17 | <abarth> | timeless: would you be willing to send me these in email? that seems more efficient |
| 21:18 | <abarth> | i'm happy to fix all these things ) |
| 21:18 | <abarth> | :) |
| 21:18 | <timeless> | abarth: hrm, i can copy what i write here into an email, it's easier for me to write here and then copy out |
| 21:18 | <abarth> | ok |
| 21:18 | <timeless> | i'm nowhere near as productive writing into an email |
| 21:18 | timeless | can't explain why |
| 21:18 | <abarth> | i've got everything done through "s/, the user agent/, then the user agent/" |
| 21:18 | <timeless> | > and metrics collected from implementations deployed to a sizable number of users . |
| 21:18 | <timeless> | s/ ././ |
| 21:19 | <timeless> | > s (such as "strip any leading space characters" or "return false and abort these steps") are to be interpreted with the meaning of the key word ("MUST", "SHOULD", "MAY", etc) |
| 21:20 | <timeless> | s/etc/etc./g |
| 21:20 | timeless | will have to figure out what list this feedback belongs on and what summary magic is required for it |
| 21:21 | <timeless> | ooh, the content-type change is nice, much easier to read :) |
| 21:21 | <timeless> | "official-type" should probably be given some styling |
| 21:21 | <timeless> | preferably not the same styling as "Content-Type" |
| 21:22 | <timeless> | > (Such messages are invalid according to RFC2616. |
| 21:22 | <timeless> | s/./.)/ |
| 21:22 | <timeless> | the rfcs should be href references of some sort btw :) |
| 21:22 | <timeless> | > If an HTTP response contains multiple Content-Type header fields, the User Agent MUST use the textually last Content-Type header field to the official-type. For example, if the last Content-Type header field contains the value "foo", then there is no official media type because "foo" cannot be interpreted as a media type (even if the HTTP response contains another Content-Type header field that could be interpreted as a media type). |
| 21:23 | <timeless> | the for example part here applies to the previous paragraph, the sentence needs to be moved to the paragraph before the instruction for multiple header fields |
| 21:24 | <timeless> | > FTP RFC0959 |
| 21:24 | <timeless> | is there a reason for the leading 0? |
| 21:24 | <timeless> | > Comparisons between media types, as defined by MIME specifications, are done in an ASCII case-insensitive manner. [RFC2046] |
| 21:24 | <timeless> | -- gecko historically had some case sensitive rules in its sniffing :) |
| 21:25 | <timeless> | > If the official-type ends in "+xml", or if it is either "text/xml" or "application/xml", then let the sniffed-type be the official-type and abort these steps. |
| 21:26 | <timeless> | please mark up `sniffed-type` and `official-type` |
| 21:26 | <timeless> | > If the official-type is an image type supported by the User Agent (e.g., "image/png", "image/gif", "image/jpeg", etc), then jump to the "images" section below. |
| 21:26 | <timeless> | s/etc// |
| 21:27 | <timeless> | > If none of the first n octets are binary data octets then let the sniffed-type be "text/plain" and abort these steps. |
| 21:28 | <timeless> | > Binary Data Byte Ranges |
| 21:28 | <timeless> | you don't actually define a `binary data octet` as any item within the ranges defined in the `binary data byte ranges` |
| 21:29 | <timeless> | > If the first octets match one of the octet sequences in the "pattern" column of the table in the "unknown type" section below, ignoring any rows whose cell in the "security" column says "scriptable" (or "n/a"), then let the sniffed-type be the type given in the corresponding cell in the "sniffed type" column on that row and abort these steps. |
| 21:29 | <timeless> | if you could make `"unknown type" section` a link to the section, that would be helpful |
| 21:30 | <timeless> | > For each row in the table below: |
| 21:30 | <timeless> | > If the row has no "WS" octets: |
| 21:31 | <timeless> | i know that "WS" appears in the table below, but it isn't defined, and i don't want to guess what it means (whitespace?) |
| 21:31 | <timeless> | > If the octets of the masked-data matches the given pattern octets exactly, then let the sniffed-type be the type given in the cell of the third column in that row and abort these steps. |
| 21:31 | <timeless> | s/matches/match/ |
| 21:31 | <timeless> | > If the row has a "WS" octet or a "_>" octet: |
| 21:32 | <timeless> | i'd probably write `an "_>" octet` (reading _ as underscore) |
| 21:32 | <timeless> | > LOOP: If index-stream points beyond the end of the octet stream, then this row doesn't match and skip this row. |
| 21:32 | <timeless> | please style `LOOP` |
| 21:32 | <timeless> | > If the index-pattern-th octet of the pattern is a normal hexadecimal octet and not a "WS" octet or a "_>" octet: |
| 21:32 | <timeless> | s/or a/nor a/ |
| 21:33 | <timeless> | s/not/neither/ |
| 21:33 | <timeless> | > "WS" means "whitespace", and allows insignificant whitespace to be skipped when sniffing for a type signature. |
| 21:33 | <timeless> | oh, so that's where you hid the definition -- way too late :) |
| 21:34 | <timeless> | > If the index-stream-th octet of the stream is one of 0x09 (ASCII TAB), 0x0A (ASCII LF), 0x0C (ASCII FF), 0x0D (ASCII CR), or 0x20 (ASCII space), then increment only the index-stream to the next octet in the octet stream. |
| 21:34 | <timeless> | if you could style the 0xXX items in something <tt>-ish, that'd be appreciated |
| 21:34 | <timeless> | ... and if you could style the names (ASCII TAB, etc.) in something, that'd also be appreciated |
| 21:35 | <timeless> | oh _ doesn't mean underscore |
| 21:35 | <timeless> | grr, please put those definitions before their use, not way below their use :( |
| 21:36 | <timeless> | > "_>" means "space-or-bracket", and allows HTML tag names to terminate with either a space or a greater than sign. |
| 21:37 | <timeless> | > If the first n octets match the signature for MP4 (as define in ), then let the sniffed-type be video/mp4 and abort these steps. |
| 21:37 | <timeless> | s/define/defined/ |
| 21:37 | <timeless> | -- the markup you're using failed to generate a reference, could you get the tool to generate an XXX when it fails? :) |
| 21:37 | <timeless> | [reference/visible-reference] |
| 21:39 | <timeless> | > FF FF FF FF FF FF WS 3C 3f 78 6d 6c text/xml Scriptable <?xml (Note the case sensitivity and lack of trailing _>) |
| 21:40 | <timeless> | s/sensitivity/sensitivity [mask = FF instead of DF]/ |
| 21:41 | <timeless> | > A JPEG SOI marker followed by a octet of another marker. |
| 21:41 | <timeless> | s/a octet/an octet/ |
| 21:41 | <timeless> | -- the table doesn't currently handle .SWF |
| 21:42 | <timeless> | in the past, that has been a problem |
| 21:42 | <timeless> | http://www.digitalpreservation.gov/formats/fdd/fdd000130.shtml |
| 21:44 | <timeless> | > If n is less than 4, then the sequence does not match the signature for MP4 and abort these steps. |
| 21:44 | <timeless> | `and` doesn't work; s/ and/;/ ? |
| 21:45 | <timeless> | oh, hrm, you're consistent in that style... oh wel |
| 21:45 | <timeless> | l... it feels wrong |
| 21:46 | <timeless> | hrm, no, in all previous cases, the form was `let foo and abort these steps` |
| 21:47 | <timeless> | here it's `then <statement of truth> and` |
| 21:47 | <timeless> | > If n is less than box-size or if box-size is not evenly divisible by 4, then the sequence does not match the signature for MP4 and abort these steps. |
| 21:47 | <timeless> | -- i stand by my complaint :) |
| 21:48 | <timeless> | (the fix is probably to move to "return TRUTH/FALSE value and abort these steps") |
| 21:48 | <timeless> | > For each i from 2 to box-size/4 - 1 (inclusive): |
| 21:49 | <timeless> | if you could put `box-size/4 - 1` into some markup to indicate that it's a math section, that'd be helpful |
| 21:49 | <timeless> | > If octets 4*i through 4*i + 2 (inclusive) of the sequence are 0x6D 0x70 0x34 (the ASCII string "mp4"), then the sequence does match the signature for MP4 and abort these steps. |
| 21:49 | <timeless> | and here for `4*i` and `4*i + 2` |
| 21:50 | <timeless> | i think you need s/If octets/If any octets/ |
| 21:50 | <timeless> | otherwise, it's ambiguous between `any` and `all` |
| 21:51 | <timeless> | > 7 Images |
| 21:51 | <timeless> | > Otherwise, let the sniffed-type be the official-type and abort these steps. |
| 21:51 | <timeless> | i'd rather otherwise be step 3 instead of part of the bulleted list inside step 2 |
| 21:53 | <timeless> | hsivonen: what happens if a document contains multiple BOMs? :) |
| 21:53 | <timeless> | (compatible or incompatible, i don't really care, although i'd be amused to hear how things respond) |
| 21:54 | <timeless> | > If the octets with positions pos to pos+2 in s are exactly equal to 0x2D, 0x2D, 0x3E respectively (ASCII for "-->"), then increase pos by 3 and jump back to the previous step (the step labeled loop start) in the overall algorithm in this section. |
| 21:55 | <timeless> | `loop start` should be a link to the LOOP label and preferably have the same case as the LOOP label |
| 21:55 | <timeless> | > Return to step 2 in these substeps. |
| 21:55 | <timeless> | it'd be nice if this was a link to an anchor in the right part of the steps |
| 21:57 | <timeless> | > If RDF-flag is 1 and RSS-flag is 1, then let the sniffed-type be "application/rss+xml" and abort these steps. |
| 21:57 | <timeless> | s/and/or/ ?? |
| 21:58 | <timeless> | ok, i've reached end of file |
| 21:58 | timeless | goes to convert this to an email |
| 21:58 | <timeless> | anyone know where the email should go? :) |
| 21:58 | <Ms2ger> | whatwg works, I think |
| 22:31 | <ojan> | TabAtkins: yt? just want to confirm something... |
| 22:31 | <ojan> | TabAtkins: row-reverse should not affect things like margin-start, right? |
| 22:31 | <ojan> | TabAtkins: as in, direction:ltr, row-reverse and margin-start should set the margin-left, right? |
| 22:32 | <timeless> | abarth: http://lists.whatwg.org/htdig.cgi/whatwg-whatwg.org/2011-September/033328.html |
| 22:32 | <abarth> | timeless: many thanks. I should have all your comments addressed within 14 hurs |
| 22:32 | <abarth> | hrs |
| 22:32 | <abarth> | 24 hrs |
| 22:32 | <abarth> | (man, I can't type today) |
| 22:32 | timeless | knows the feeling |
| 22:33 | <timeless> | i have a busy end of week, so i could easily not see it wednesday, in which case i won't touch it until next tuesday |
| 23:15 | <Hixie> | AryehGregor: http://www.w3.org/Bugs/Public/show_bug.cgi?id=13425#c6 |
| 23:58 | <Hixie> | anyone know how i can test http://www.w3.org/Bugs/Public/show_bug.cgi?id=12199 ? |
| 23:58 | <Hixie> | I don't really know ISO-2022-JP :-/ |
