| 11:31 | <annevk> | "i think we dropped +xml knowledge because it broke sites" sad state of affairs that is :( |
| 11:32 | <Ms2ger> | Welcome to the web |
| 12:57 | <zcorpan> | Hixie: maybe the status boxes could have a field for http://www.w3.org/Bugs/Public/show_bug.cgi?id=13853 |
| 15:52 | <jgraham> | TabAtkins: "Does web work add complex to web?" is clearly a question about whether it is possible to form multi-protein complexes on spider's webs by carefully controlled application of energy ("work") to the structure ("web") |
| 15:55 | <Ms2ger> | jgraham++ |
| 15:56 | <Ms2ger> | TabAtkins, I was mostly thinking about TR/, though |
| 16:40 | <annevk> | Hixie: content type is including parameters arguably so I am not sure that is the correct term and it seems MIME type is somewhat incorrect (at least, thus argue the IETF folk) |
| 16:52 | <dglazkov> | good morning, Whatwg! |
| 16:52 | <Ms2ger> | Evening |
| 16:53 | <Philip`> | May this day bless you and your family, dglazkov! |
| 16:54 | <dglazkov> | :) |
| 17:11 | <jgraham> | May your loins be blessed with many fruit? |
| 18:19 | <karlcow> | annevk: http://backstage.soundcloud.com/2011/08/soundcloud-mobile-proxies/ |
| 18:19 | <karlcow> | about CORS |
| 18:21 | <annevk> | someone should write a test suite |
| 18:22 | <Ms2ger> | We probably have one |
| 18:22 | <Ms2ger> | If you find it, I can probably get it submitted |
| 18:25 | <annevk> | I think the Gecko one uses a JavaScript server framework |
| 18:42 | <Ms2ger> | Probably |
| 18:49 | <Hixie> | annevk: well media type is just confusing |
| 18:49 | <Hixie> | annevk: maybe "contentType string"? |
| 18:49 | <Hixie> | annevk: it's not like this has any real meaning anyway |
| 18:49 | <Hixie> | annevk: it's not the type from the server, it's only barely the type the browser thinks the resource was, etc |
| 18:55 | <annevk> | "resource type"? |
| 18:55 | <timeless> | Ms2ger: ah, the hazards of writing long reviews |
| 18:55 | timeless | is reading heycam's response |
| 18:55 | <annevk> | could maybe rename MIME type to resource type |
| 18:56 | <timeless> | (for the second time actually, the first was using Gmail.java, which hides all quoted bits) |
| 18:59 | <Ms2ger> | timeless, mm, sending detailed feedback to D3E works better for that... You won't remember anything about your comments by the time you get a reply |
| 19:00 | <timeless> | Ms2ger: oh, heycam managed that too |
| 19:00 | <timeless> | i mean, i remembered vaguely being amused (when i read) about objects defying things |
| 19:00 | <timeless> | but i didn't really remember any of my comments in any useful level of detail |
| 19:01 | <timeless> | of course, the fact that my general buffer is probably at best 3 screens long, and each message i wrote was probably 30 didn't help |
| 19:01 | <timeless> | (and there were 5 messages0 |
| 19:01 | <timeless> | s/0/) |
| 19:02 | <Ms2ger> | Yeah, that reply was quite the wall of text |
| 19:02 | <timeless> | it was actually amazingly readable in Gmail.java when you consider the input |
| 19:03 | <timeless> | (there are minor issues w/ how Gmail.java strips quoted sections, so you can get unrelated paragraphs running into eachother) |
| 19:03 | <timeless> | (but really, all in all, amazingly readable) |
| 19:03 | <timeless> | but absolutely useless for trying to do DoC |
| 19:03 | timeless | is using notepad for that |
| 19:03 | timeless | used notepad to review the spec too |
| 19:04 | <Ms2ger> | Pff, DoC |
| 19:08 | <Ms2ger> | Oh, LC ends tomorrow |
| 19:09 | <timeless> | on WebIDL? |
| 19:09 | <timeless> | grr |
| 19:09 | <Hixie> | annevk: resource type is fine by me |
| 19:09 | Ms2ger | raises "I don't understand your grammar" |
| 19:10 | <Hixie> | annevk: MIME type is what everyone calls it, like URL. Don't listen to the Julians of the world, their pedancy is not useful. |
| 19:10 | <Hixie> | annevk: i agree that .contentType is not the document MIME type though, even though it looks like a MIME type |
| 19:10 | <timeless> | interface TimelessParsingWhatWG { DOMString onWebIDL?; }; |
| 19:10 | <timeless> | grr, no, that fails |
| 19:10 | <Ms2ger> | DOMString? onWebIDL |
| 19:10 | <timeless> | yeah |
| 19:11 | <Ms2ger> | No |
| 19:11 | <Ms2ger> | attribute DOMString? onWebIDL |
| 19:11 | <timeless> | well, hrm, that's awkward |
| 19:12 | <Ms2ger> | Did anybody raise an issue to drop the chapter about Java? |
| 19:12 | <timeless> | interface on { attribute DOMString? discussion; }; interface TimelessParsingWhatWg { attribute on WebIDL; }; |
| 19:13 | <timeless> | we'll skip the discussion about using a the timeless module, because we wouldn't want any other groups to think there's a ghost of a chance of this supporting them |
| 19:13 | <timeless> | Ms2ger: but, that' |
| 19:13 | <timeless> | s parsable, right? :) |
| 19:14 | <Ms2ger> | I... think so |
| 19:14 | <Ms2ger> | I mean, I understand the code, but I don't understand the grammar |
| 19:15 | timeless | gets lost |
| 19:21 | <timeless> | ooh |
| 19:21 | timeless | reaches part 4 |
| 19:22 | <annevk> | Hixie, it's not the specified one, but it's the determined one, no? |
| 19:51 | timeless | reaches part 5! |
| 19:51 | <timeless> | ooh, it only took me half an hour to read through the response to one section |
| 19:52 | <timeless> | that's not bad |
| 20:02 | <timeless> | so, wordpad claims my reply is 5 pages |
| 20:04 | <timeless> | outlook (which i unfortunately trust a bit more here) claims it's 15 |
| 20:16 | <Ms2ger> | Who uploaded a contenteditable test to livedom? |
| 20:27 | <annevk> | WHATWG Weekly will have to wait till tomorrow |
| 20:47 | <Ms2ger> | Hixie, yt? |
| 20:52 | <AryehGregor> | Ms2ger, probably me. |
| 20:55 | <Ms2ger> | AryehGregor, so do you know about insertAdjacentHTML? |
| 20:55 | <AryehGregor> | Ms2ger, not really. |
| 21:05 | <timeless> | Ms2ger: ping |
| 21:06 | <Ms2ger> | Hi |
| 21:06 | <Ms2ger> | Do *you* know about insertAdjacentHTML? :) |
| 21:07 | <timeless> | i could |
| 21:07 | <timeless> | i'll trade |
| 21:09 | <timeless> | right, so, what about it? |
| 21:17 | <Hixie> | annevk: vaguely the determined one, but only vaguely |
| 21:17 | <Hixie> | annevk: e.g. createDocument() has one |
| 21:17 | <Hixie> | annevk: as do error messages |
| 21:17 | <Hixie> | Ms2ger: here |
| 21:19 | <Ms2ger> | http://www.whatwg.org/specs/web-apps/current-work/multipage/content-models.html#insertadjacenthtml |
| 21:19 | <Ms2ger> | Did you consider the case where the parent of target is a DocumentFragment? |
| 21:19 | <Hixie> | probably not |
| 21:20 | <Hixie> | doesn't seem to break in that case though |
| 21:20 | <Ms2ger> | ...HTML fragment parsing algorithm. The algorithm optionally takes as input an Element node, referred to as the context element... |
| 21:21 | <Ms2ger> | So in that case, the context element isn't actually an element, right? |
| 21:26 | <Hixie> | holy cow, the ietf changes to the websocket protocol have introduced a massive security hole |
| 21:26 | <Hixie> | the server no longer explicitly lists the origin! |
| 21:31 | <AryehGregor> | SIGH: http://lists.w3.org/Archives/Public/spec-prod/2011JulSep/0035.html |
| 21:31 | <AryehGregor> | This is what happens when you try to make reasonable suggestions at the W3C. |
| 21:31 | <AryehGregor> | Well, let's see how it goes. |
| 21:32 | <Ms2ger> | I think HTML4 has a test suite, actually |
| 21:32 | <AryehGregor> | It would be hard for it to be a useful one. |
| 21:33 | <Ms2ger> | Well, it sure isn't useful |
| 21:34 | <Ms2ger> | http://www.w3.org/MarkUp/Group/Test/HTML401/tests/index.html |
| 21:34 | <AryehGregor> | This looks promising: https://lists.webkit.org/pipermail/webkit-dev/2011-August/017738.html |
| 21:34 | <AryehGregor> | Nice, requires W3C login? |
| 21:34 | <AryehGregor> | To be fair, HTML5 has practically no useful tests either, other than in a few tiny areas. |
| 21:35 | <timeless> | Hixie: why is that shocking? |
| 21:35 | <Ms2ger> | I think there's a public place as well, but I can't find it |
| 21:35 | <Hixie> | timeless: i was told that the ietf would bring expert review to the protocol |
| 21:36 | <Hixie> | timeless: i knew that that wasn't true, but i didn't realise it would be so bad that they would actually introduce vulnerabilities |
| 21:36 | <Ms2ger> | Ah, http://www.w3.org/MarkUp/Test/HTML401/current/tests/index.html |
| 21:37 | <Hixie> | Ms2ger: that test suite makes unwarranted assumptions about the conformance requirements in html4 |
| 21:37 | <Ms2ger> | Well sure |
| 21:38 | <Ms2ger> | I only pointed out its existence, I was expecting it to suck |
| 21:39 | <Hixie> | i could have sworn i subscribed to member-webapps-cvs or whatever it's called |
| 21:39 | <Hixie> | but i don't seem to get e-mail from it |
| 21:39 | <Hixie> | wtf |
| 21:46 | <timeless> | so, wrt html4 test suite |
| 21:46 | <timeless> | if you want to blame someone |
| 21:46 | <timeless> | http://www.w3.org/MarkUp/Test/HTML401/current/tests/acknowledgements.html |
| 21:46 | <timeless> | "Tantek Çelik, who acted as senior mentor for this effort." |
| 21:46 | <timeless> | :) |
| 21:53 | <Hixie> | no blame to be assigned |
| 21:54 | <Hixie> | tantek's the only one even tried (well, tantek, me, and a few others -- there have been several such efforts over the years) |
| 21:54 | <Hixie> | the problem is the spec sucks :-) |
| 22:07 | <Hixie> | reminder to anyone who might be interested in giving talks about HTML5 to add yourself to http://wiki.whatwg.org/wiki/Presentations#Speakers |
| 22:07 | <Hixie> | i point people to that page pretty regularly |
| 22:37 | <Hixie> | what's the right terminology for saying that a list should be sorted first by some key A, then breaking ties on some key B, and breaking further ties on some key C? |
| 22:39 | <timeless> | secondary/tertiary sorting? |
| 22:39 | <Hixie> | complete this sentence: "Sort the list L " |
| 22:39 | <jcranmer> | it should be as if sorted by C then B then A using stable sorts |
| 22:40 | <Hixie> | there's no better terminology? bummer. |
| 22:40 | <timeless> | yeah, jcranmer 's text is correct |
| 22:40 | <timeless> | there's text for hashing |
| 22:41 | <jcranmer> | it's more precise but I don't like it that much |
| 22:41 | <timeless> | in hashing, it's double hashing where you specify the second hash function |
| 22:43 | <karlcow> | AryehGregor: I didn't reply to Liam, because I didn't think it would move forward the discussion. |
| 22:43 | <timeless> | Hixie: is your sorting going to specify how to handle unicode garbage? :) |
| 22:43 | timeless | goes to look up the technical term |
| 22:43 | <Hixie> | no |
| 22:43 | <Hixie> | it's sorting the tasks that get queued during the processing of captions |
| 22:43 | <Hixie> | for video |
| 22:46 | <timeless> | oh right |
| 22:46 | <timeless> | opera mobile *sucks* at dealing w/ large <textarea>s |
| 22:49 | <Philip`> | Hixie: Sort by a lexicographic order on (A, B, C) |
| 22:49 | <Hixie> | it wouldn't be lexicographic, but more importantly, where is it defined what that means? |
| 22:49 | <Philip`> | Why not lexicographic? |
| 22:49 | <Hixie> | well A is a time, B is a position in another list, and C is creation order :-) |
| 22:50 | <Hixie> | no graphs to lexic, as it were |
| 22:50 | <Philip`> | http://mathworld.wolfram.com/LexicographicOrder.html |
| 22:50 | <Philip`> | (Doesn't need to be anything to do with alphabets) |
| 22:50 | <Hixie> | i have no idea what that page is saying |
| 22:51 | <Hixie> | ok the wikipedia page clued me in |
| 22:51 | <Hixie> | what a confusing term to reuse |
| 22:53 | <timeless> | hi heycam |
| 22:53 | <heycam> | morning timeless |
| 22:53 | <timeless> | i have a longer response which i'm trying to paste |
| 22:53 | <timeless> | but i'm on the phone now |
| 22:54 | <timeless> | and i'll send another 1-8 messages |
| 22:54 | <heycam> | oh dear 8 messages? :) |
| 22:54 | <Philip`> | Hixie: Algebraists aren't necessarily known for non-confusingness |
| 22:55 | <Philip`> | but at least that particular term seems to only have a single definition in mathematics, so it's only unobvious and not ambiguous |
| 22:55 | <timeless> | main reply, 6 spun off items, and a review of the actual diff commit |
| 22:56 | <timeless> | plus an optional review of the current spec |
| 22:56 | <Hixie> | Philip`: well it would be ambiguous if one of the keys was a number |
| 22:56 | <timeless> | ok, so, maybe 2-9? :) |
| 22:56 | <Hixie> | Philip`: since lexical order for numbers (string definition) is not the same as lexical order for numbers (mathematical definition) |
| 23:00 | <Philip`> | If it's a lexicographic order over a product of types, you have to say what order to use for each of those types (or assume it's obvious to readers, like numbers are in increasing order) - the lexicographicness doesn't matter to those types |
| 23:26 | <timeless> | grr |
| 23:27 | <timeless> | Opera is trying to parse mail.google.com/mail/u/0/x/*/ as xml |
| 23:27 | <timeless> | and gives me an error w/ a link to reparse as html |
| 23:27 | <timeless> | for each link i click through in gmail |
| 23:29 | <timeless> | heycam: ok, the big email is outbound |
| 23:30 | <heycam> | timeless, thanks |
| 23:30 | <timeless> | that just leaves me w/ the diff, the 6 subissues, and re-reviewing the entire document |
| 23:34 | <timeless> | heycam: a note on VCS commit blame churn... |
| 23:34 | <timeless> | </x:codeblock> |
| 23:34 | <timeless> | and similar tags should really be on their own line ;-) |
| 23:35 | <heycam> | timeless, that might result in some extra white space in the published document |
| 23:35 | <timeless> | really |
| 23:35 | <timeless> | ? |
| 23:35 | <heycam> | since it gets turned into a <pre> |
| 23:35 | <timeless> | yuck |
| 23:35 | <heycam> | maybe |
| 23:35 | <heycam> | can't remember exactly what my xslt does |
| 23:35 | <timeless> | get a better preprocessor :) |
| 23:36 | <timeless> | http://dev.w3.org/cvsweb/2006/webapi/WebIDL/Overview.xml.diff?r1=1.341;r2=1.342;f=h |
| 23:36 | <timeless> | near Line 5765 results.numbers.length; <span clas |
| 23:37 | <timeless> | i feel bad |
| 23:37 | <timeless> | the <must> in there ruins how <take(s)> should be used :( |
| 23:38 | <heycam> | timeless, yeah I noticed that and fixed it up in a following change |
| 23:38 | <timeless> | oh good |
| 23:39 | <timeless> | sorry about that (it happens in a bunch of spots) |
| 23:39 | <timeless> | this is where a dvcs is useful |
| 23:39 | <timeless> | commit changes for person |
| 23:39 | <timeless> | commit fixes for changes for person |
| 23:39 | <timeless> | commit further fixes for changes for changes... |
| 23:39 | <timeless> | merge |
| 23:39 | <timeless> | if you need even more changes, you go back to the premerged version and make them there |
| 23:40 | <timeless> | and when you want to show someone what you did for them, you can send either the leaf, or the delta between their origin point and their leaf |
| 23:40 | <timeless> | (no one does this) |
| 23:44 | <timeless> | isn't this illegal: |
| 23:44 | <timeless> | [TreatUndefinedAs=Null] attribute DOMString? owner; |
| 23:44 | <timeless> | oh, no, that's TreatNulllAs |
| 23:44 | timeless | sighs |
| 23:44 | <timeless> | terribly confusing :) |
| 23:44 | timeless | should go home and sleep or something |
| 23:45 | <heycam> | :) |
| 23:45 | <timeless> | (when you get online, it's a hint that i've been at work too long) |
| 23:50 | <timeless> | ooh, i'm in the credits! |
| 23:51 | <timeless> | right, i've reached the end of the diff |
| 23:53 | <timeless> | ok, i'll send my comments about the diff, and then i'm giving up for the day |
| 23:53 | <timeless> | that leaves [x1..x6] + rereview |
| 23:53 | <timeless> | there's no way i can possibly do the rereview today, or really tomorrow |
| 23:54 | <timeless> | given it took me an entire week to do the initial review, and most of today just to review your reply to my review.. |
| 23:55 | <timeless> | alright, sent |
| 23:55 | <timeless> | have a good day |