| 02:14 | <MikeSmith> | http://lists.w3.org/Archives/Public/ietf-http-wg/2009JulSep/0397.html (Yngve on tab-level cookies) |
| 03:38 | <nathanhammond> | not to interrupt anybody's stellar evening, just wanting to stir up conversation with regards to an email I sent to the list: http://lists.whatwg.org/htdig.cgi/whatwg-whatwg.org/2009-July/021578.html |
| 04:13 | <bchoate> | will the html5 specification include requirements about the content generated for the 'contenteditable' attribute? i'd love to see an acid test for that one... |
| 04:19 | <MikeSmith> | bchoate: what kind of requirements? |
| 04:20 | <bchoate> | MikeSmith: that browsers produce valid html5 markup for instance |
| 04:20 | <MikeSmith> | bchoate: |
| 04:20 | <MikeSmith> | bchoate: ah |
| 04:21 | <bchoate> | and preferably not include garbage like class="Apple-style-span" or class="Mso-Normal", etc. |
| 04:21 | <MikeSmith> | interesting case.. I don't think it's currently a requirement that they produce valid html |
| 04:22 | <MikeSmith> | one man's garbage is another man's feature |
| 04:22 | <MikeSmith> | nathanhammond: seems like some of those are questions that Hixie needs to answer himself. And he seems particularly snowed under lately, so I'd guess you might have to wait for a bit before you get a reply from him. |
| 04:23 | <nathanhammond> | MikeSmith: I was getting that impression |
| 04:23 | <MikeSmith> | nathanhammond: I'd suspect it's possible that Hixie may not have even made time yet to read that message. |
| 04:23 | nathanhammond | nods |
| 04:24 | <nathanhammond> | I was guessing that may be the case. |
| 04:24 | <MikeSmith> | nathanhammond: but regardless, I'd recommend continuing to ask here |
| 04:24 | <MikeSmith> | the squeaky wheel gets the grease, and all that |
| 04:25 | <nathanhammond> | indeed, and I was curious if anybody else had comments |
| 04:25 | <nathanhammond> | but as is probably the case most of the time people are very interested in one subsection of the spec |
| 04:26 | <nathanhammond> | (just like me) |
| 04:26 | <MikeSmith> | yeah |
| 04:27 | <MikeSmith> | sometimes it seems to me like the blind men and the elephant |
| 04:27 | <MikeSmith> | nathanhammond: especially in context of some of the "HTML5 is a mess" comments |
| 04:28 | <nathanhammond> | and you just hope that all of the blind men get a chance to explain what their experience is |
| 04:28 | <MikeSmith> | what most people mean when they say that kind of stuff is, "This one particular part of the spec, which is the main/only part that I care about personally, is a mess." |
| 04:28 | <MikeSmith> | nathanhammond: yeah |
| 04:29 | <MikeSmith> | nathanhammond: so are you the owner for the history feature in Chrome? the UI and all? |
| 04:57 | <nathanhammond> | MikeSmith: I am not, I mostly stick to client-side code |
| 04:57 | <MikeSmith> | OK |
| 04:58 | <nathanhammond> | I wrote JSSM and am working on a HTML 5 History shim |
| 05:03 | <MikeSmith> | nathanhammond: I've never tried JSSM, but will take a look |
| 05:04 | <nathanhammond> | don't! |
| 05:04 | <MikeSmith> | ? |
| 05:04 | <MikeSmith> | my main history-management mechanism is to keep a bunch of tabs open |
| 05:04 | <MikeSmith> | I would really like to have something better and easier/natural to use |
| 05:04 | <nathanhammond> | hah, this is for websites |
| 05:04 | <MikeSmith> | ah |
| 05:05 | <MikeSmith> | my beef with most History features is that they just get in my way |
| 05:05 | <nathanhammond> | I agree, I do the tab thing |
| 05:05 | <nathanhammond> | with session saving on close |
| 05:05 | <nathanhammond> | (in the event that Flash pegs the cpu) |
| 05:08 | MikeSmith | reads more about JSSM and also http://nathanhammond.com/google-chrome-and-history-management |
| 05:08 | <nathanhammond> | out of date now |
| 05:08 | <nathanhammond> | *grin* |
| 05:09 | <nathanhammond> | but I still haven't been able to track down that issue. |
| 05:09 | <MikeSmith> | no resolution on the bug? |
| 05:12 | <nathanhammond> | the last time I tried to track it down I still could not successfully isolate it |
| 05:12 | <nathanhammond> | I hear rumor that the issue has disappeared in current versions of Chrome |
| 05:13 | <nathanhammond> | but without the unit tests that I've been writing for the updated version it really hasn't been a mature project |
| 05:28 | <Hixie> | nathanhammond: i'm about 10 days behind on the list |
| 05:28 | <Hixie> | bchoate: the spec says that the UA must generate valid content |
| 05:30 | <bchoate> | Hixie: oh that's good to hear -- where does it say that? i'm looking at http://www.w3.org/TR/html5/editing.html but maybe that isn't the latest draft? |
| 05:30 | <nathanhammond> | Hixie: I'll be patient. *grin* |
| 05:31 | <Hixie> | bchoate: for example, under "break block": "The exact behavior is UA-dependent, but user agents must not, in response to a request to break a paragraph, generate a DOM that is less conformant than the DOM prior to the request." |
| 05:33 | <bchoate> | Hixie: ok i see - thanks |
| 08:02 | <zcorpan> | Hixie: why is xmp equivalent to code rather than pre? |
| 08:52 | <Hixie> | zcorpan: is that wrong? |
| 08:52 | <Hixie> | zcorpan: i thought it was an inline element, no? |
| 08:52 | <Hixie> | i didn't look too closely |
| 08:52 | <Hixie> | so i may well have gotten it wrong |
| 09:13 | <takkaria> | hmm, how many windows-1252 documents are mislabelled as iso-* encodings? I'm betting it's more than .2% |
| 09:18 | <annevk2> | who made that claim? |
| 09:19 | <Lachy> | annevk2, Larry did when he made a sarcastic comment saying "The advice for the few "public browser" implementors who feel compelled to also deal with the increasing their compatibility with existing web sites from 67% to 67.2% of existing content by supporting odd, broken character transformations." |
| 09:21 | annevk2 | sighs |
| 09:21 | <Hixie> | Philip` can probably get you a list of iso-8859-1 pages that use the win1252 characters easily enough |
| 09:21 | <annevk2> | maybe all Adobe employees have mandatory "how to act like a troll" training |
| 09:22 | Hixie | notices it's less than 8 hours until his meeting, and decides to not go to his meeting |
| 09:22 | <annevk2> | I'm not really interested in proving him wrong |
| 09:22 | <annevk2> | Done that a lot before and he just continues to apply his rethoric to something completely different ignoring what you said |
| 09:23 | <Hixie> | apparently i'm working on a fork of html5 |
| 09:23 | <Hixie> | larry masinter never did reply to my e-mail asking him what he meant by that |
| 09:23 | <Philip`> | I don't trivially have that data (e.g. http://philip.html5.org/data/charsets.html says 100% of iso-8859-1 pages were considered valid by the decoder, probably because it doesn't forbid win-1252 characters) |
| 09:23 | <Hixie> | speaking of people who don't reply to e-mail... |
| 09:24 | <Philip`> | and would probably have to write more Java code to extract that data |
| 09:24 | <Philip`> | and I don't care enough |
| 09:24 | <Hixie> | Philip`: i just meant search for [\x80-\x9f] |
| 09:24 | <Philip`> | Hixie: I'd have to write code to search for that |
| 09:24 | <Hixie> | grep won't do it? |
| 09:25 | <Philip`> | grep won't restrict it to iso-8859-1-declared pages |
| 09:25 | gsnedders | wonders what "his meeting" is |
| 09:25 | <Hixie> | i'd search for it myself, but unfortunately my corpus is post-conversion-to-utf-8 |
| 09:27 | <othermaciej> | takkaria: I was going to comment on that, but I lost heart before I finished my email reply |
| 09:28 | <Hixie> | MikeSmith: are we going to have a survey on who is attending the meeting in november? |
| 09:28 | <gsnedders> | What should typeof HTMLElement give? |
| 09:29 | <annevk2> | the alternative helding it at Google proposal is not going through btw? |
| 09:29 | <Lachy> | Hixie, I assume Larry simply used the term fork incorrectly to refer to your's and Manu's specs as competing alternatives, rather than Manu's as simply being a supplimentary section to be considered for inclusion. |
| 09:30 | <othermaciej> | isn't Manu's spec in fact a full-text alternative? |
| 09:30 | <gsnedders> | othermaciej: There are both forms |
| 09:31 | <othermaciej> | should be easy to recast as a standalone spec, then, if the author were so inclined |
| 09:31 | <Lachy> | othermaciej, AIUI, when he generates it, he takes the trunk at the time and inserts his new section into it, and then publishes both as a standalone and incorporated version |
| 09:31 | <othermaciej> | I do note though that he didn't remove the microdata section, at least not yet |
| 09:32 | <MikeSmith> | Hixie: yeah, I set up the survey already |
| 09:32 | <Lachy> | MikeSmith, what survey? |
| 09:32 | <MikeSmith> | I'll get the URL |
| 09:32 | <Lachy> | oh, I guess Hixie meant the TPAC survey |
| 09:32 | <MikeSmith> | Lachy: yeah |
| 09:32 | <othermaciej> | which makes me wonder even more why Sam felt compelled to mix microdata into my suggestion about RDFa |
| 09:33 | <othermaciej> | I hope it's not too late to register for TPAC, as I have been too lazy to do so |
| 09:33 | <zcorpan> | Hixie: xmp a block element |
| 09:33 | <Hixie> | zcorpan: huh go figure |
| 09:33 | <Hixie> | zcorpan: i'll fix it, thanks |
| 09:36 | <Lachy> | othermaciej, I don't believe it is |
| 09:37 | <Lachy> | Otherwise, no-one at Opera will be attended, as we're all waiting on Chaals to make up his mind about who to send |
| 09:37 | <Hixie> | zcorpan: done |
| 09:37 | <Hixie> | when does the price go up? |
| 09:37 | <Hixie> | it's $250 each right now right? |
| 09:37 | <Hixie> | for the week? |
| 09:38 | <Lachy> | really? That's a lot. Didn't it used to be $50? Did it go up once already, or have I just remembered incorrectly? |
| 09:38 | <gsnedders> | Lachy: $50 per day |
| 09:39 | <Lachy> | ah, ok. |
| 09:39 | <gsnedders> | Did someone here not link to something by Steven Pemberton yesterday? |
| 09:40 | <gsnedders> | s/yesterday/two days ago/ |
| 09:40 | <gsnedders> | http://www.pemberton.nl/vandf/2009/07/xhtml2-not-dead.html |
| 09:42 | <Philip`> | gsnedders: Many people did not link to it |
| 09:42 | <gsnedders> | Many did not, but annevk did |
| 09:54 | <zcorpan> | "Ian Hickson sings the praise of those modules elsewhere." |
| 09:56 | <tkent_> | Why the return type of HTMLInputElement::list() is HTMLElement instead of HTMLDataListElement? |
| 09:56 | <MikeSmith> | othermaciej: it's not too late. I'll try to get the green light on announcing the registration survey by beginning of next week |
| 09:57 | <MikeSmith> | othermaciej: please cajole hyatt into going, if you can |
| 09:57 | <othermaciej> | MikeSmith: it's less convenient for hyatt than for me, but I'll ask if he's interested |
| 09:57 | <MikeSmith> | and would be great to meet other webkit folks f2f |
| 09:57 | <othermaciej> | Apple people based in Cupertino would be more likely to be available |
| 09:58 | <MikeSmith> | yeah, understood |
| 09:58 | <annevk2> | zcorpan, I thought that was funny too; the post he points to is 7y old or so :) |
| 09:58 | <gsnedders> | Was Dean Jackson not the only Apple guy last year? |
| 09:58 | <MikeSmith> | Dave Singer was there too |
| 09:58 | <MikeSmith> | I seem to remember at least |
| 09:58 | <gsnedders> | I don't remember meeting him… |
| 09:59 | <Hixie> | if othermaciej can get hyatt to go, i'll be impressed |
| 09:59 | <MikeSmith> | tkent_: ping Hixie about that |
| 09:59 | <othermaciej> | hyatt hates standards meetings |
| 09:59 | gsnedders | on the whole doubts he'll be going |
| 09:59 | <tkent_> | Hixie, how about that? |
| 10:01 | Philip` | wonders if it's worth pointing out that the 'decentralised' identifier problem can't be solved with DNS, because DNS relies on centralised authorities, and all you're doing is pushing the centralisation away by a few levels of abstraction |
| 10:01 | <heycam> | gsnedders, "object" |
| 10:01 | <heycam> | (for typeof HTMLElement) |
| 10:01 | <gsnedders> | heycam: k, thx |
| 10:02 | <Hixie> | tkent: i think at some point <select> was allowed as a target also |
| 10:02 | <othermaciej> | I am mainly interested in going for the chance to meet people in person that I might not have met at all, or don't get to talk to regularly |
| 10:02 | <othermaciej> | not really expecting valuable technical breakthroughs |
| 10:02 | <Hixie> | Philip`: in practice, dns is decentralised enough |
| 10:03 | <tkent> | Hixie: ok, I understand. Thanks. |
| 10:04 | <annevk2> | gsnedders, while you're reporting things like HTMLUnknownElement maybe you should investigate XMLDocument and such too |
| 10:04 | <gsnedders> | annevk2: I'm really working on other stuff now though |
| 10:04 | <annevk2> | gsnedders, never mind then |
| 10:04 | <Hixie> | hey i don't suppose any of you feel like writing the scripts that would autogenerate the index tables would you? |
| 10:04 | <Hixie> | i recently figured out what the columns should be |
| 10:05 | <annevk2> | it looked pretty complex |
| 10:05 | <othermaciej> | we've never had HTMLUnknownElement |
| 10:05 | <othermaciej> | I wonder if Yahoo Mail Beta fails for us, or does something different |
| 10:05 | <gsnedders> | A generic index of all dfns I have a plan for |
| 10:05 | <gsnedders> | othermaciej: You get WebKit specific JS |
| 10:05 | <gsnedders> | (Likewise IE gets its own too) |
| 10:06 | <othermaciej> | does Opera get the same as Firefox? |
| 10:06 | <othermaciej> | HTMLUnknownElement seems like a really goofy idea, but I guess it's far from the dumbest thing in the web platform |
| 10:06 | <Philip`> | Hixie: The "enough" is just a consequence of various factors (like the ease of registering non-conflicting names, and robustness to failures, etc), rather than there being a fundamental binary difference to e.g. a centralised wiki page, so there might be lots of other options that are somewhere in the middle |
| 10:07 | <gsnedders> | othermaciej: I haven't checked exactly, but from the brief look I took it seems so |
| 10:07 | <Hixie> | othermaciej: if you send an e-mail saying we shouldn't have it, that's pretty much all it would take to convince me to go back to using HTMLElement |
| 10:07 | <Hixie> | Philip`: preaching to the choir :-) |
| 10:07 | <Philip`> | so I think it's a bit misleading to think of decentralised vs centralised, because it's all just different degrees of where the centralisation occurs |
| 10:08 | <othermaciej> | (it seems bogus because it can stop working if an element stops being "unknown", and also because many actual recognizedelements are just vanilla HTMLElement, so it seems weird to cater more to unrecognized ones) |
| 10:08 | <annevk2> | gsnedders, why did we not want the WebKit codepath? |
| 10:08 | <Hixie> | no argument from me there |
| 10:08 | <gsnedders> | As long as websites break if we remove it (esp. high profile ones like Y! Mail beta), it's unlikely we'll remove it |
| 10:08 | <Philip`> | Hixie: Good, that must mean I'm not totally crazy (at least relative to this environment) :-) |
| 10:08 | <Hixie> | othermaciej: send mail :-) |
| 10:08 | <gsnedders> | annevk2: I tried UA sniffing, and it never finished loading |
| 10:08 | <gsnedders> | *spoofing |
| 10:09 | <annevk2> | http://lists.w3.org/Archives/Public/www-dom/2009JulSep/0125.html touche othermaciej! |
| 10:09 | <othermaciej> | I'll consider it tomorrow and maybe send mail |
| 10:09 | <Philip`> | gsnedders: Make the browser detect if the user works for Yahoo, and if so then do the sane thing that breaks Yahoo Mail, so they notice and fix the problem, without it affecting many real users |
| 10:10 | <gsnedders> | :P |
| 10:10 | <Hixie> | what's a good thing to use as an example for how to use microdata |
| 10:10 | <othermaciej> | annevk2: yeah I wasn't sure if I should send that, but I have forgiven myself this minor sin |
| 10:11 | <gsnedders> | Hixie: http://quotes.burntelectrons.org/4288 |
| 10:11 | <othermaciej> | Hixie: semantic markup for photos of a model train collection? |
| 10:11 | <annevk2> | Hixie, markup for your StarGate DVD collection? |
| 10:11 | <othermaciej> | or has that example been overused? |
| 10:11 | <Hixie> | let me rephrase my question |
| 10:11 | <Philip`> | It seems better to have examples that would be slightly more relevant for normal people |
| 10:11 | <Hixie> | what's a good thing to use as an example for how to use microdata *that someone would actually care about and do something with* |
| 10:11 | gsnedders | ignores othermaciej for his rudeness in that email |
| 10:12 | <Hixie> | rather than your typical semantic web examples of "mark everything up and maybe it'll be useful one day" |
| 10:12 | <gsnedders> | Hixie: Do we actually have examples like that in HTML 5? :\ |
| 10:12 | <takkaria> | are there any such examples? |
| 10:12 | <othermaciej> | gsnedders: it wasn't chronic! |
| 10:12 | <gsnedders> | How dull. |
| 10:12 | <othermaciej> | (can't say as to insufferable) |
| 10:13 | <Philip`> | Hixie: Something from the list of use cases that motivated the inclusion of microdata in the spec |
| 10:13 | <annevk2> | Hixie, markup for a TV guide? |
| 10:13 | <mookid> | ahah ..? |
| 10:13 | <othermaciej> | Hixie: what kind of "do something with" - are you interested in examples with prior agreement between producer and consumer, or does it have to be data that unknown unrelated consumers may want? |
| 10:13 | <othermaciej> | one example for the prior case would be bugzilla |
| 10:13 | <annevk2> | nm that suggestion |
| 10:14 | <Hixie> | the main use cases were all vcard and vevent related, but i want something where you make up your own vocab, for this example |
| 10:14 | <othermaciej> | bugzilla HTML markup changes between different versions, so a script can't reliably screen-scrape |
| 10:14 | <Hixie> | othermaciej: i guess either is fine |
| 10:14 | <Hixie> | ooh, bugzilla's a good one |
| 10:14 | <othermaciej> | the XML version is more reliable but completely separate |
| 10:14 | <mookid> | =) |
| 10:14 | <othermaciej> | microdata would let you parse the HTML and reliably extract data in the face of markup changes |
| 10:15 | <Philip`> | Microdata would make the extracted data extremely fragile because it would keep breaking whenever someone changed the markup and forgot to check the microdata was identical in all cases |
| 10:15 | <gsnedders> | othermaciej: But why not use the XML interface? |
| 10:15 | <othermaciej> | though in general consumers will intend to work with bugzilla specifically and don't need a generic central "bug database" vocabulary |
| 10:15 | <mookid> | or you could negotiate content and use a more appropriate format than HTML for data interchange? :D |
| 10:15 | <othermaciej> | gsnedders: less risk of inconsistency if there's only one interface |
| 10:16 | <gsnedders> | othermaciej: Indeed. But there are already two interfaces. |
| 10:16 | <othermaciej> | mookid: if using a separate non-HTML formatted value is an appropriate solution, then clearly microdata (or RDFa) won't help |
| 10:17 | <othermaciej> | gsnedders: well, two responses on that |
| 10:17 | <othermaciej> | (1) but would you design it like that from scratch? we shouldn't assume bugzilla is the only bugzilla-like piece of software ever to be made |
| 10:17 | <mookid> | do you have to have a head node in an html document for it to be valid? |
| 10:18 | <gsnedders> | mookid: Yes, and title within it. |
| 10:18 | <mookid> | is there a reason for that? |
| 10:18 | <othermaciej> | (2) we're intending to customize bugzilla a fair bit for WebKit, and reliable semantic markup in the HTML would be easier to keep in sync with changes than a whole separate XML version |
| 10:18 | <Philip`> | mookid: (The head *tags* are optional, though the element is required) |
| 10:18 | <mookid> | what's the logic behind that descision? |
| 10:18 | <takkaria> | page titles are an essential part of a page |
| 10:19 | <gsnedders> | othermaciej: (1) No; (2) But it seems likely that things already rely upon Bugzilla's XML interface ;( |
| 10:19 | <gsnedders> | * :( |
| 10:19 | <mookid> | how are they essential? :/ |
| 10:19 | <othermaciej> | Hixie: search results would be another good example, if search engines wanted to enable screen scraping instead of being kind of hostile to it |
| 10:19 | <mookid> | <head><title></title></head> - valid? |
| 10:20 | <annevk2> | you forgot a doctype |
| 10:20 | <Hixie> | othermaciej: the bug example is fine |
| 10:20 | <mookid> | -_- |
| 10:20 | <gsnedders> | mookid: No, you need a DOCTYPE and a body element |
| 10:20 | <takkaria> | mookid: <!DOCTYPE html><title</title> I believe is the shortest valid bit of HTML |
| 10:20 | <mookid> | yeah yeah I just mean the head element |
| 10:20 | <othermaciej> | any example I can think of is vulnerable to the criticism that you could just provide a completely separate data interface |
| 10:20 | <Hixie> | othermaciej: yeah, well, there's a reason i'm not a fan of this whole technology concept |
| 10:20 | <mookid> | why do you require the title though - I don't understand why you would insist on that |
| 10:21 | <othermaciej> | however, in the case of bugzilla, having the machine-readable data combined with the human-readable presentation seems intuitively appealing |
| 10:21 | <othermaciej> | to me anyway |
| 10:21 | <annevk2> | mookid: whatwg⊙wo or public-html-comments⊙wo |
| 10:21 | <mookid> | :/ |
| 10:22 | <Philip`> | takkaria: I think you might want more >s in there |
| 10:22 | <mookid> | I don't want to change anything I just want to understand the rationale |
| 10:22 | <annevk2> | fair enough, I don't really have any :) |
| 10:22 | <takkaria> | Philip`: yeah, a couple more won't hurt |
| 10:22 | <mookid> | :) |
| 10:23 | <othermaciej> | I imagine the rationale would be that many systems present HTML documents by title, thus lacking one would be considered an error |
| 10:23 | <mookid> | doesn't enforcing a title/head element kind of damage HTML as a decent 'fragment' format? |
| 10:23 | <othermaciej> | (examples of places where the title shows up: browser titlebar, browser history, browser bookmarks, search engine search results) |
| 10:24 | <annevk2> | if you just want a fragment you do not need a full document (or title) |
| 10:24 | <othermaciej> | an HTML fragment doesn't need to be a conforming HTML document |
| 10:24 | <Philip`> | (Browser tabs too) |
| 10:24 | <othermaciej> | although I'm not sure there is a conformance class for it |
| 10:24 | <mookid> | if(!title) title = 'Zomg the end of the world!1!1'; |
| 10:25 | <mookid> | you're welcome. |
| 10:25 | <othermaciej> | presumably the fragment context would have to define what kinds of fragments are allowed, and the processing model |
| 10:25 | <mookid> | annevk2: that's not true if I want my fragments to be crawlable and valid html |
| 10:27 | <mookid> | think that's a fair requirement, no? |
| 10:27 | <annevk2> | I need some more context for your use case |
| 10:29 | <mookid> | website that uses html fragments/Ajax also wants individual fragments to be crawlable and valid markup |
| 10:29 | <mookid> | for SEO etc. |
| 10:33 | <MikeSmith> | what happened to onbeforeunload? |
| 10:33 | MikeSmith | is auditing the list of event-handler attributes that v.nu knows about against the list in the spec |
| 10:33 | <Hixie> | btw Lachy the blog is still suffering from the effects of a hacker |
| 10:34 | <Hixie> | MikeSmith: it's in, isn't it? |
| 10:34 | <MikeSmith> | Hixie: it's not in the list in the Global Attributes section |
| 10:34 | <MikeSmith> | I don't think |
| 10:34 | MikeSmith | doublechecks |
| 10:34 | <Hixie> | it's not a global attribute |
| 10:35 | <MikeSmith> | ah |
| 10:35 | <Hixie> | it's only on <body> (and <frameset>, though since that element is obsolete, that doesn't matter much) |
| 10:35 | <MikeSmith> | OK |
| 10:35 | <MikeSmith> | yeah |
| 10:35 | MikeSmith | hits himself with clue stick |
| 10:37 | <Hixie> | um, if i try to preview a blog post on blog.whatwg.org i get asked for a Magic password? |
| 10:39 | <Lachy> | Hixie, I don't get asked for a password |
| 10:39 | <annevk2> | mookid, titles are important for SEO |
| 10:40 | <Hixie> | Lachy: "To view this page, you must log in to area âMagicâ on blog.whatwg.org:80." |
| 10:40 | <Hixie> | that's for http://blog.whatwg.org/?p=924&preview=true |
| 10:41 | <Hixie> | it worked when i reloaded |
| 10:41 | <Hixie> | wtf |
| 10:41 | <Lachy> | works for me |
| 10:41 | <annevk2> | interesting, WordPress does not lowercase href attributes |
| 10:41 | <othermaciej> | darn, I got goaded into writing an anti-namespace jeremiad |
| 10:41 | <Lachy> | Is there anything else to appears to be a result of a hacker? |
| 10:42 | <annevk2> | I almost started flaming back at Larry, but I'll instead pull a Larry and not reply |
| 10:42 | <Hixie> | Lachy: do you see the spam on the front page? |
| 10:42 | <annevk2> | trying to correct all his misconceptions about how the Web works is not worth it |
| 10:43 | <othermaciej> | annevk2: or you could pull a Larry by sarcastically misrepresenting his point, and then accusing him of using a strawman argument |
| 10:43 | <Hixie> | Lachy: oh, looks like you fixed it |
| 10:43 | <Hixie> | you guys are killing me |
| 10:43 | <Lachy> | no |
| 10:43 | Hixie | schedules 2pm-3pm tomorrow for responding to e-mails from people reading the chat log and complaining to me for not telling you off for being rude about people who are being rude |
| 10:44 | <othermaciej> | Hixie: usually takes the log trawlers more than 24 hours to embark on their fishing expeditions |
| 10:44 | <Lachy> | Hixie, were you looking at a cached copy of the page? |
| 10:44 | <Hixie> | Lachy: yeah, maybe |
| 10:44 | <Lachy> | ok |
| 10:45 | <Hixie> | clearly i don't know how to use web browsers |
| 10:45 | <othermaciej> | whoah, MLW is now trolling in near real-time |
| 10:45 | <Hixie> | soon i'll be ready to work for w3c |
| 10:45 | <Hixie> | i am happy to say that i haven't looked at MLW in at least 6 months now, possibly more |
| 10:46 | <Lachy> | othermaciej, yeah, I have a strong suspicion that MLW is someone who is in this channel |
| 10:46 | <Lachy> | I just haven't figured out who |
| 10:47 | <othermaciej> | I hadn't looked in a while until you mentioned having to deal with people reading the chat log |
| 10:48 | <annevk2> | I was subscribed to that blog for a while, but it got really boring |
| 10:49 | <othermaciej> | the sarcastic posts can be amusing but lately it seems he's descended to just stringing together swear words |
| 10:53 | <Lachy> | MLW has also claimed to affiliated with the W3C in some way http://twitter.com/MrLastWeek/status/2530468490 |
| 10:53 | <Lachy> | though, admittedly, the message is vague and it isn't clear what "activities of an official W3C nature" may be. |
| 10:54 | <othermaciej> | I think we just gave him more attention than he deserves |
| 10:54 | <othermaciej> | mea culpa |
| 10:57 | <Hixie> | ok, blogged about microdata on the whatwg blog |
| 10:58 | <mookid> | link! |
| 10:58 | <mookid> | make sure it's got a .html on the end of it please |
| 10:58 | <Lachy> | mookid, why? |
| 10:59 | <Lachy> | http://blog.whatwg.org/microdata1 |
| 11:00 | <mookid> | just being an idiot, ignore me |
| 11:00 | <mookid> | :) |
| 11:02 | <Philip`> | Hixie: I think your examples would be less misleading if they used a real domain, instead of something unnaturally short like example.org, e.g. it's going to be more like <body item="org.mozilla.bugzilla.bug"> |
| 11:02 | <Lachy> | Philip`, it could just as easily be org.mozilla.bug |
| 11:03 | <Hixie> | it would actually just be org.bugzilla.bug |
| 11:03 | <Hixie> | which is only one character longer |
| 11:04 | <Hixie> | i considered using ch.hixie.bug, since that's what i would use |
| 11:04 | <Hixie> | (and that's even shorter!) |
| 11:05 | <Hixie> | anyway. bed time. nn. |
| 11:05 | <annevk2> | I saw a 4 character domain the other day, something like a.gh or some such |
| 11:05 | <gavin_> | http://e.dk/ |
| 11:05 | annevk2 | thought the minimum was 4 |
| 11:05 | <annevk2> | 5, doh |
| 11:06 | <annevk2> | Lachy, so we fixed the spam problem on the blog we found the other day? |
| 11:06 | <annevk2> | Lachy, also, can we upload a new skin that at least enables people to see the name of the person who posted? |
| 11:06 | <Lachy> | annevk2, yeah. Some hacker had somehow modified the templates. That's now been removed and prevented |
| 11:06 | <annevk2> | Lachy, it would be even better if turned on the skin Ben made |
| 11:07 | <Lachy> | which skin did Ben make? |
| 11:07 | <annevk2> | a nice green one |
| 11:07 | <Lachy> | Has it been uploaded to blog.whatwg.org? |
| 11:07 | <annevk2> | I don't think so |
| 11:07 | <Lachy> | or is it available from somewhere? |
| 11:08 | <annevk2> | actually, it was not a skin, it was a design done in HTML/CSS |
| 11:08 | <annevk2> | but it would be easy to make the blog use it I think, maybe email him? |
| 11:08 | <MikeSmith> | http://a.gd/ |
| 11:08 | <Lachy> | well, if someone converts it to a WP template, I can install it easily |
| 11:08 | <annevk2> | MikeSmith, cheerio |
| 11:09 | <annevk2> | what's his email? |
| 11:09 | <Philip`> | ➡.ws is only four characters too |
| 11:10 | <annevk2> | not in ASCII |
| 11:10 | <Philip`> | Who cares about ASCII? |
| 11:10 | <annevk2> | HTTP |
| 11:11 | <mookid> | who cares about HTTP? |
| 11:11 | <Philip`> | Who cares that HTTP cares about ASCII? |
| 11:11 | <Lachy> | doesn't HTTP use ISO-8859-1? |
| 11:12 | <Lachy> | (I don't care that it's a superset of ASCII) |
| 11:12 | <annevk2> | (I don't care about your question) |
| 11:12 | <Lachy> | I don't care. |
| 11:12 | <annevk2> | great :) |
| 11:14 | <Lachy> | wow, another lawsuit against TPB http://torrentfreak.com/italian-riaa-suesthe-pirate-bay-for-1-million-euros-090731/ |
| 11:14 | <Lachy> | I guess the MAFIAA are just going to keep suing them in every country they can |
| 11:21 | <takkaria> | what better things have they go to do with themselves, really? |
| 11:21 | <Lachy> | takkaria, they could spend their time and money on developing better and more sustainable business models |
| 11:22 | <Philip`> | Suing people for a million dollars seems like a good business model |
| 11:22 | <takkaria> | what, rather than operating under the assumption that one day the Internet will be un-invented? |
| 11:22 | <Lachy> | not when they know the defendants won't pay |
| 11:23 | annevk2 | emailed Ben |
| 11:23 | <Lachy> | takkaria, very few things have been uninvented |
| 11:24 | <Philip`> | I guess lots of things invented in ancient Egypt and ancient Greece etc have been uninvented |
| 11:25 | <Lachy> | if you consider uninvented to mean technology that we have lost all knowledge of through history |
| 11:26 | <Philip`> | Yes - inventing is the gaining of knowledge of how to do something, and uninventing is the losing of that knowledge |
| 11:28 | <takkaria> | Lachy: I think my preferred approach for them would be just to keep on doing what they're doing except keep the money they're paying lawyers for themselves |
| 11:29 | <takkaria> | that way they'll retire earlier and eventually new people will start managing the music companies who have a clue |
| 11:30 | Philip` | wonders why everyone focuses on TPB when there's dozens of other large torrent sites |
| 11:32 | <takkaria> | well, mininova and isohunt have started talking to the movie industry I believe to start self-censoring |
| 11:34 | Philip` | doesn't remember it taking long for TPB to gain prominence when Suprnova was shut down |
| 11:34 | <Lachy> | TPB will probably die in a month anyway, and openbittorrent.org and publicbt.org will take over as the word's main trackers, and torrent files will start being distributed through other means, raising the bar for the MAFIAA again |
| 11:34 | <Philip`> | so I suppose I'm assuming someone else will take over in the near future |
| 11:36 | <Philip`> | Lachy: Trackers aren't really interesting - the important thing is a decent-quality searchable index of torrents, which is an orthogonal problem |
| 11:36 | <Lachy> | it wouldn't surprise me if someone set up a torrent site that is run using technoloy like Freenet, so that they don't host the material themselves and anyone is free to anonymously add torrents to the network |
| 11:36 | <Philip`> | How do you ensure quality in that situation? |
| 11:37 | <Lachy> | that would be a challenge |
| 11:38 | <Philip`> | Indeed |
| 11:38 | <Lachy> | but even TPB didn't do much to ensure quality, since they never removed torrents |
| 11:39 | <Philip`> | As far as I'm aware they managed to avoid being overwhelmed by mislabelled torrents and virus-infected programs |
| 11:39 | <Philip`> | (It's particularly a problem when people are being paid to intentionally disrupt the system and reduce its value to users) |
| 11:40 | <takkaria> | I think the most likely end result is most people using a wider spread of trackers |
| 11:41 | <takkaria> | and tools may emerge to make that really much easier |
| 11:41 | <takkaria> | e.g. trackers supporting OpenSearch and a site/app searching a lot of trackers for you |
| 11:42 | <takkaria> | could become built-in to torrent apps themselves |
| 11:45 | <Philip`> | Why would the people running trackers want to cooperate with the people who are running aggregators/search engines and taking all the ad revenue away from the trackers? |
| 11:46 | <takkaria> | non-profit trackers? |
| 11:46 | <takkaria> | / other private trackers |
| 11:47 | <Lachy> | Philip`, trackers like openbittorrent.org and publicbt.org aren't ad supported. But are you referring to sites that actually host the torrents? |
| 11:48 | <Lachy> | it would be nice to have a service dedicated to hosting the torrents files, and only providing an API for indexing sites |
| 11:49 | <Lachy> | that would make the search sites largely immune from lawsuits since, like Google and other search engines, they don't host or have any control over the torrents themselves |
| 12:01 | <mookid> | look what I found! |
| 12:01 | <mookid> | http://whitepapers.theregister.co.uk/paper/view/814/oracle-814.pdf (not a pdf!) |
| 12:02 | <mookid> | I guess they didn't get the "URI's have special endings because that's what the most cleverest(tm) developers do" memo |
| 12:03 | <annevk2> | maybe they got the memo that stated that URIs are opaque strings |
| 12:04 | <mookid> | huhuhu :) |
| 12:04 | <mookid> | couldn't agree more. |
| 12:05 | gsnedders | wonders if IE copes |
| 12:05 | <gsnedders> | (As don't IE still do sniffing based upon extension?) |
| 12:07 | <Lachy> | http://www.theage.com.au/technology/biz-tech/shock-threat-to-shut-skype-20090731-e3qe.html |
| 12:07 | <mookid> | [13:37] <Philip`> If you encode the desired content-type in the URI, it's not possible for a general-purpose HTTP-processing device to understand what's the resource identifier and what's the content-type selector, because there's no standard for encoding that stuff in URIs |
| 12:08 | <mookid> | [13:38] <Hixie> Philip`: such a standard or convention could be easily established |
| 12:08 | <mookid> | :) |
| 12:08 | <mookid> | annevk2 ^ |
| 12:08 | <takkaria> | Opera does some sniffing by file extension... |
| 12:09 | <mookid> | where is the spec that suggests that is a good idea? |
| 12:09 | <takkaria> | I imagine it's stuffed up someone's arse |
| 12:09 | <mookid> | if it's not in a spec that sounds like a very BAD idea |
| 12:09 | <takkaria> | it's a bad idea |
| 12:10 | <takkaria> | mostly because sniffing based on the first few bytes is a far more effective streategy |
| 12:10 | <mookid> | or Content-Type header? -_- |
| 12:10 | <gsnedders> | mookid: Content-Type sniffing in general is a bad idea, but needed for web compat., and until recently there was nowhere near any consensus on how to do it |
| 12:10 | <takkaria> | gsnedders: actually, sniffing is more effective |
| 12:10 | <annevk2> | in hindsight, content-type and content negotation were pretty bad ideas |
| 12:10 | <gsnedders> | takkaria: I never questioned that :P |
| 12:11 | <mookid> | this is the exact reason that HTTP has headers for. |
| 12:11 | <takkaria> | yeah, that was a mental typo. s/gsnedders/mookid/ |
| 12:11 | <gsnedders> | It's pure mental, like |
| 12:11 | gsnedders | notes that joke will probably be lost on everyone outside of the UK |
| 12:12 | <takkaria> | and indeed, me |
| 12:12 | <gsnedders> | s/the UK/Scotland/ then maybe |
| 12:12 | <mookid> | annevk2: you keep saying that about content negotiation |
| 12:12 | <mookid> | there is nothing to support that assertion whatsoever |
| 12:12 | <mookid> | other than the *pathetic* support for it in browsers/html which tend to dictate how web mechanisms are leveraged in practice |
| 12:13 | <mookid> | if you don't have html or browsers in your web architecture there is no problem with conneg at all |
| 12:13 | <takkaria> | there's also very little web. :) |
| 12:13 | <mookid> | ... |
| 12:13 | <mookid> | HTTP is the web |
| 12:13 | <mookid> | URIs |
| 12:14 | <mookid> | html/browsers are a necessary evil |
| 12:14 | <gsnedders> | mookid: Quite frankly, we've been over this several times here already. We've given plenty reasons to support that assertion, though they seemingly have fallen on deaf ears. You aren't going to get a different response by rehashing the same arguments. If you want to get a greater diversity of responses, email whatwg⊙wo/public-html-comments⊙wo |
| 12:15 | <annevk2> | I'm not really sure how to put my argument |
| 12:15 | <mookid> | annevk2: maybe you don't have one then? :p |
| 12:15 | <gsnedders> | mookid: Well, I'm sure if you ask timbl he'll tell you originally HTTP _was_ HTML and URLs, and that everything else was added on later. |
| 12:15 | <annevk2> | I think what I believe is that if an idea does not get a good implementation over a decade it is simply not worth it and should be dropped (and is a bad idea in hindsight) |
| 12:16 | <mookid> | what tbl thinks is irrelevant realy |
| 12:16 | <mookid> | the web is HTTP/URIs |
| 12:16 | <mookid> | Services on HTTP and RIA's are the web |
| 12:18 | <mookid> | browsers and html are important to the web on a practical level but they don't actually have anything to do with the underlying architecture |
| 12:18 | <gsnedders> | Hixie: You still awake? |
| 12:19 | <annevk2> | gsnedders was just talking about the original idea of the Web |
| 12:19 | <mookid> | so what the original implementation of HTTP was pretty poorly concieved anyway |
| 12:19 | <annevk2> | and the Web and the fundamental architecture of it are separate concepts in my mind |
| 12:19 | <mookid> | the fundemental driver of success is the architecture |
| 12:20 | <gsnedders> | mookid: It's a legacy you won't escape from, though, the poorly concieved original |
| 12:20 | <mookid> | not if we have standards bodies populated by individuals with the predisposition to continue this insanty, no. |
| 12:21 | <annevk2> | so do something about it |
| 12:21 | <gsnedders> | mookid: Well you seem quite uninterested to actually do anything about it, and prefer to just complain on IRC. |
| 12:21 | <annevk2> | talking here won't really further your cause |
| 12:22 | <annevk2> | well, I suppose it might, but given the length of this conversation I'm pretty sure nobody is reading |
| 12:22 | <mookid> | :) |
| 12:22 | <gsnedders> | http://lists.w3.org/Archives/Public/public-html/2009Jul/0807.html |
| 12:22 | <gsnedders> | Where in the select section? |
| 12:22 | <mookid> | Poor governance basically http://en.wikipedia.org/wiki/Government_failure |
| 12:22 | <annevk2> | also, I have the feeling that each time I chat with you you're saying the same things... |
| 12:23 | <annevk2> | (and I'm repeating myself too) |
| 12:23 | <mookid> | right - that's mainly because you repeat stuff that doesn't make sense |
| 12:23 | <mookid> | like |
| 12:23 | <mookid> | "conneg is broken" |
| 12:23 | <gsnedders> | mookid: We've given plenty of reasons, too, but you've ignored them all. |
| 12:24 | <mookid> | mainly because you're rationale is based entirely on insufficiencies on markup/browser insufficiencies |
| 12:24 | <mookid> | and nothing to do with the actually Web |
| 12:24 | <mookid> | actual^ |
| 12:24 | <gsnedders> | ah-ha! "If the multiple attribute is absent, whenever an option element in the select element's list of options has its selectedness set to true, and whenever an option element with its selectedness set to true is added to the select element's list of options, the user agent must set the selectedness of all the other option element in its list of options to false." |
| 12:24 | <annevk2> | ok, i'll just ignore you now until you have something constructive |
| 12:24 | <gsnedders> | mookid: We've given plenty of other reasons too. |
| 12:24 | <mookid> | you haven't actually mate I wouldn't waste my time if you had |
| 12:26 | <mookid> | if you point me to the part of the spec where Conneg is flawed, fair enough |
| 12:26 | <mookid> | otherwise you're just holding someone's head under the water and having a go at them for not being able to swim properly |
| 12:27 | <annevk2> | everyone can write a spec that is technically sound |
| 12:27 | <annevk2> | that does not mean it's good |
| 12:27 | <mookid> | ok so maybe next time the HTTP guys should do it with phonetically spelled words and nice pictures? |
| 12:28 | <mookid> | and special gold star stickers if you manage to actually read the damn spec properly |
| 12:29 | <gsnedders> | I guess the employees of W3C who wrote HTML 4.0 are equally incompetent in your view? |
| 12:29 | <gsnedders> | If the W3C has been so incompetent for the past twelve years, what makes you think it will change? |
| 12:29 | <mookid> | I think they have way more excuses than you lot do. |
| 12:30 | <othermaciej> | mookid: I think your continual repetition of the same positions is not constructive and borders on trolling |
| 12:30 | <mookid> | terrific |
| 12:30 | <mookid> | ironically unconstructive comment |
| 12:30 | <othermaciej> | I will admit to having fed the troll before, but it's kind of getting old |
| 12:30 | <mookid> | thanks for sharing |
| 12:30 | <gsnedders> | What excuses do they have that we don't? Regardless, you're just giving the same arguments over and over again, and ignoring ours totally. If you want to try and get anywhere at all, mail the list. Bitching in IRC will make _absolutely no difference whatsoever_. |
| 12:31 | <mookid> | I'm giving the same arguments because ever 'response' I get contains invalid and irrelevant points that don't address what I actually said |
| 12:32 | <mookid> | I'll say this again - there is *nothing* wrong with HTTP conneg, and a lot wrong with browsers/html |
| 12:32 | <mookid> | if I avoid both of those I can leverage HTTP conneg fine |
| 12:33 | <othermaciej> | mookid: if you want to summarize your position in email (either whatwg⊙wo or public-html⊙wo, your choice), that would be useful |
| 12:33 | <othermaciej> | I think repeating the same points over and over on IRC, and getting more or less the same responses, is not very useful |
| 12:34 | <othermaciej> | or you can file a bug at <http://www.w3.org/Bugs/Public/>; in the HTML5 spec proposal component, with your concrete proposal |
| 12:35 | <mookid> | well I'm pretty sure there's some fundamental problems in my communication and/or your comprehension which are unlikely the magically dissapear in a 'conclusion' |
| 12:35 | <annevk2> | mookid, have you considered that you might not get our arguments? |
| 12:36 | <mookid> | well I'm pretty well versed in HTTP and working with it on a day to day basis |
| 12:36 | <othermaciej> | mookid: I agree there is likely a communication and/or comprehension problem |
| 12:36 | <mookid> | so I have a good chance of 'getting' what you're saying |
| 12:36 | <annevk2> | the problems I'm pointing out have not so much to do with HTTP technicalities |
| 12:36 | <othermaciej> | I also think repeating the same things more times is not going to foster better communication |
| 12:36 | <annevk2> | not at all, I think |
| 12:37 | <mookid> | othermaciej: not if the people I'm talking to are interested in winning an argument rather than constructive discussion |
| 12:38 | <mookid> | annevk2: what are these problems then? |
| 12:38 | <mookid> | where do they come from.. browsers/html ? :P |
| 12:38 | <gsnedders> | mookid: There are others here who are well versed in HTTP and work with it every day |
| 12:38 | <mookid> | sure, they have different design philosophies |
| 12:39 | <mookid> | I'm just telling you that I can't use HTTP conneg |
| 12:39 | <mookid> | it's there to be used and I can't practically use it because there's no browser/html mechanisms for it |
| 12:40 | <mookid> | both approaches (URI vs protocol) conneg would work fine along side each other |
| 12:40 | <othermaciej> | I think you've made your position on the matter abundantly clear |
| 12:40 | <mookid> | do you mind? |
| 12:40 | <othermaciej> | yes, I do |
| 12:40 | <gsnedders> | Also, characterising it as people who are interested in winning an argument rather than constructive discussion is _not_ going to help. |
| 12:40 | <othermaciej> | because you've been going on about it every day for weeks |
| 12:41 | <othermaciej> | and every day you repeat yourself |
| 12:41 | <othermaciej> | and this isn't #groundhogday |
| 12:41 | <mookid> | yeah that gandhi guy sure was irritating |
| 12:41 | <othermaciej> | so bring some new information or take it elsewhere |
| 12:41 | <mookid> | I beg your pardon? |
| 12:47 | <mookid> | 'what about xyz?' - 'yeah well nevermind about xyz because I can juggle with fruit' - '......... what about xyz?' |
| 12:47 | <mookid> | 'hey stop repeating yourself!' |
| 12:49 | <mookid> | a simple "its too much like hard work" would be fine if that's the reality of the situation |
| 13:09 | <takkaria> | mookid: I don't understand your link to "Government failure" earlier |
| 13:15 | <mookid> | :P |
| 13:18 | <mookid> | you don't understand why I brought that up or the content of the page? |
| 13:19 | <mookid> | Government === 'a body who govern' |
| 13:35 | <annevk> | Philip`, btw, dreamhost worked for me again when I tried it later last night |
| 13:37 | <Philip`> | annevk: Thanks, works for me now too |
| 14:15 | <takkaria> | mookid: I don't grok how it's relevant |
| 14:33 | <mookid> | takkaria: governing bodies often miss-allocate resources |
| 14:45 | <takkaria> | right, that's what the article says, but why is it relevant? :) |
| 14:52 | <mookid> | it would be nice to think it isn't relevant at all |
| 14:53 | <mookid> | 'Regulatory capture - The co-opting of regulatory agencies by members of or the entire regulated industry. Rent seeking and rational ignorance are two of the mechanisms which allow this to happen.' |
| 14:55 | <takkaria> | mookid: ok, I'm obviously not smart enough to see whatever you're inferring here |
| 14:55 | <Philip`> | Maybe he's calling us ignorant |
| 14:56 | <mookid> | I'm not really inferring anything to be honest, it just is what it is |
| 14:58 | <Philip`> | So you're saying that a page exists on Wikipedia and describes something, and that's it? |
| 14:59 | <mookid> | more the concept of the content on that page phil.. :P |
| 15:00 | <mookid> | I don't really have a solid opinion one way or another because I don't actually know the details of how you guys operate |
| 15:01 | <mookid> | i've read the rants, though |
| 15:01 | <mookid> | seems like that's what most of them are getting at |
| 15:02 | <mookid> | the OPEC of the internet |
| 15:08 | <mookid> | It doesn't matter that much anyway - as long as we get video tags I'll be happy; can't wait to have video in web pages |
| 15:11 | <mookid> | then we'll get websites dedicated to serving video content and everything |
| 15:11 | <mookid> | it's going to be a revolution |
| 15:12 | <mookid> | an attractively tagged revolution |
| 15:14 | <taf2> | is there anyway to use the canvas tag to extract pixel information... e.g. loop over the matrix of pixels in a canvas tag... allowing one to do analysis on the color matrix? |
| 15:15 | <Philip`> | taf2: getImageData() |
| 15:15 | <taf2> | cool |
| 15:15 | <Philip`> | taf2: http://www.whatwg.org/specs/web-apps/current-work/multipage/the-canvas-element.html#pixel-manipulation etc |
| 15:40 | <annevk> | sweet |
| 15:40 | <annevk> | people are actually going to fix the cookie issue |
| 15:40 | <annevk> | https://www.ietf.org/mailman/listinfo/http-state |
| 15:40 | <annevk> | awesomeness |
| 15:41 | <jacobolus> | what mime type should I use to serve javascript? |
| 15:42 | <annevk> | doesn't matter |
| 15:42 | <jacobolus> | well, if I have to pick something... |
| 15:42 | <annevk> | you could not include a mime type |
| 15:42 | <jacobolus> | haha |
| 15:42 | <annevk> | IETF prefers applicatoin/ecmascript or application/javascript; HTML5 prefers text/javascript |
| 15:43 | <jacobolus> | text/javascript it is :) |
| 15:43 | <zcorpan> | apache defaults to application/x-javascript i think |
| 15:43 | <jacobolus> | zcorpan: actually, application/javascript |
| 15:43 | <jacobolus> | at least on my mac |
| 15:43 | <zcorpan> | oh they changed |
| 15:44 | <jacobolus> | zcorpan: they also have an entry for "text/javascript" with no extension |
| 15:44 | <jacobolus> | http://svn.apache.org/repos/asf/httpd/httpd/trunk/docs/conf/mime.types |
| 15:44 | <zcorpan> | is that the latest mime.types for apache? |
| 15:44 | <jacobolus> | zcorpan: for some reason the python mimetypes module overrides javascript to do application/x-javascript |
| 15:44 | <jacobolus> | zcorpan: I assume it's the latest |
| 15:44 | <jacobolus> | it's in their svn |
| 15:46 | zcorpan | updates http://wiki.whatwg.org/wiki/Video_type_parameters |
| 15:48 | <mookid> | annevk: they're finally getting rid of cookies? |
| 15:49 | <mookid> | cool! |
| 15:49 | <mookid> | =) |
| 15:52 | <mookid> | a working group developing new ways to abuse HTTP |
| 15:53 | <mookid> | love it |
| 15:53 | <mookid> | isn't HTTP just the wrong protocol if you need stateful client/server interaction? |
| 15:54 | <Philip`> | mookid: I thought you were in favour of adding features to the language just because somebody thinks they're useful and the language designers shouldn't get in their way, so you ought to be happy that they're working on cookies which lots of people find useful |
| 15:54 | <mookid> | no I'm in favour of adding features which provide mechanisms to leverage components of the underlying architecture of the web |
| 15:55 | <mookid> | no augment bullshit ontop of it like cookies |
| 15:55 | <mookid> | which exist because people try to do stupid things with HTTP |
| 15:55 | <Philip`> | They exist because people try to solve practical problems using HTTP |
| 15:56 | <mookid> | stateful interaction is mostly bad design |
| 15:57 | <mookid> | particularly on HTTP |
| 15:57 | <Philip`> | Or HTTP is badly designed for stateful interaction :-) |
| 15:58 | <mookid> | stateful interaction is a horrific way to build an anarchic scalable distributed system the size of the web |
| 16:02 | <Philip`> | Most people aren't building the web, they're just building a single centralised web site with no need for scalability and it works fine for them |
| 16:04 | <mookid> | I'm talking about the why HTTP was designed the way it was |
| 16:04 | <mookid> | and why it was successful |
| 16:04 | <mookid> | that platform is 'the web' |
| 16:05 | <mookid> | I don't really see the problem with people creating systems that operate the same way the web does |
| 16:06 | <mookid> | not really hard to see why that might be a good idea in the long run as your domain evolves |
| 16:06 | <mookid> | the solution isn't cookies it's education and tooling |
| 16:07 | <cgriego> | I think reusing the legend element, given the issues, is a contradiction of the goal of backwards compatability. |
| 16:12 | <annevk> | cgriego, but that's not the only goal we have |
| 16:12 | <annevk> | cgriego, so it's not a simple as that ;) |
| 16:12 | <jacobolus> | huh. apache's list has the 'wav' extension twice, for audio/wav and audio/x-wav |
| 16:13 | <jacobolus> | I wonder how they decide precedence. That's probably a bug |
| 16:13 | <jacobolus> | /typo |
| 16:15 | <MikeSmith> | jacobolus: judging by past experiences, if you file a bug now, they might get around to fixing it 5 years from now |
| 16:15 | <jacobolus> | heh |
| 16:16 | <MikeSmith> | and you will get a lot of entertaining comments posted to the bug during that time, from a wide variety of genuine jackasses |
| 16:16 | <MikeSmith> | great entertainment value |
| 16:18 | <jacobolus> | python's mimetypes module overrides rdf and wsdl to be application/xml instead of application/rdf+xml & application/wsdl+xml |
| 16:18 | <jacobolus> | is there a good reason to send application/xml instead of the others? |
| 16:19 | <cgriego> | annevk: I think legend would be an excellent element to use with figure and details, except every major browser has issues with it, and not because they did anything wrong. HTML 4 defines legend, as a form element, to be a replaced element. Undoing that is a radical change in definition. |
| 16:26 | <annevk> | I don't think HTML4 defined it as such |
| 16:26 | <adactio> | cgriego: I agree. I tried to some DOM Scripting with <details> and it proved to be impossible because of the way that browsers interpret the presence of the <legend> element (Safari throws it away, Firefox generates a <fieldset>). |
| 16:26 | <adactio> | The problem is ...what's the alternative? |
| 16:27 | <annevk> | (in theory it should be treated as an unknown element outside of <fieldset>) |
| 16:27 | <annevk> | to elaborate a little bit, one of the other goals we have is keeping the language simple |
| 16:27 | <annevk> | and as browsers fix their bugs <legend> will become useful |
| 16:28 | <annevk> | it might take a little longer, but not having to introduce a new element seems worth it |
| 16:28 | <adactio> | annevk: I agree and I think that Safari and Firefox will probably fix this pretty quickly but IE (again) is going to really slow things down. |
| 16:29 | <adactio> | I've read through the arguments for and against <legend>: http://lists.whatwg.org/htdig.cgi/whatwg-whatwg.org/2008-February/014038.html |
| 16:29 | <adactio> | And I agree, in principle, that it's far more preferable to "recycle" an existing element than to create a new one. |
| 16:30 | <adactio> | But I'd really like to be able to use the new HTML5 elements and as long as some of them use <legend>, they're effectively out of bounds. :-( |
| 16:32 | <adactio> | Remy has a more detailed breakdown of the issues: http://html5doctor.com/legend-not-such-a-legend-anymore/ |
| 16:32 | <annevk> | yeah, it's dilemma |
| 16:32 | <taf2> | can a Worker be used to run canvas operations? |
| 16:32 | <annevk> | I guess so far we decided not to cater to the early adopter crowd :) |
| 16:32 | <takkaria> | taf2: sure |
| 16:33 | <annevk> | taf2, it can only process ImageData objects |
| 16:33 | <annevk> | takkaria, not really, there's no DOM in a Worker |
| 16:34 | <taf2> | annevk: but that is okay i think... for me... if i get the image data pass to a worker and mainipulate the data postMessage back to the main thread and putImageData... |
| 16:34 | <adactio> | annevk: Yeah, and that's fair enough. It's just a shame that so much doesn't need catering to (e.g. <section>, <article>, <header>); we can just go ahead and use them without waiting for browsers. On the face of it, <details> and <figure> should fall into the same category ...but that damned <legend> throws a spanner in the works. |
| 16:34 | <takkaria> | annevk: I meant you can do ImageData stuff, but you were more precise. :) |
| 16:35 | <annevk> | <div class="legend"> :/ |
| 16:46 | <adactio> | annevk: yeah, that's what I'm doing for now with the new structural elements anyway (using class names rather than elements, as an interim step) so I can do that for anything requiring a <legend> too. |
| 16:47 | <cgriego> | Has anyone looked to see if similar issues arise with the caption element? |
| 16:48 | <annevk> | yes, it does not work because of <table> parsing |
| 16:48 | <adactio> | cgriego: I believe <caption> has its own issues because it is expected inside a table. |
| 17:03 | <cgriego> | How about <label>? |
| 17:04 | <annevk> | maybe you should read the email adactio pointed to |
| 17:05 | <cgriego> | good point |
| 19:18 | <nathanhammond> | As a conversation starter: http://lists.whatwg.org/htdig.cgi/whatwg-whatwg.org/2009-July/021578.html ... anybody have any thoughts regarding what I propose? |
| 19:23 | <annevk> | the order is already defined |
| 19:25 | <annevk> | not sure what to think of the rest |
| 19:25 | <annevk> | does that help? :) |
| 19:26 | <annevk> | (the order is defined because in step 5 popstate is dispatched (if needed) and in step 6 hashchange is dispatched (if needed) |
| 19:26 | <annevk> | ) |
| 20:52 | <nathanhammond> | annevk: good to know about the ordering, and better that it matches what I'd hoped |
| 20:54 | <nathanhammond> | however, with the discussion I've read, it looks like popstate is only fired on history traversal, not on input (reasonable) |
| 21:44 | gsnedders_ | wonders whether to give birthday wishes before or after going to bed |
| 22:57 | <gsnedders_> | annevk: Happy birthday (provided Facebook is honest)! |
| 23:05 | <virtuelv> | gsnedders_: http://annevankesteren.nl/2004/08/birthday |
| 23:05 | <gsnedders_> | Then I guess Facebook is honest. |
| 23:06 | <ezyang> | Whoo, happy birthday annevk! |
| 23:11 | <hober> | http://pbfcomics.com/archive_b/PBF032-Todays_My_Birthday.gif |
| 23:55 | <Darxus> | Div's suck for page layout. |
