| 01:02 | jgraham | connects just to mention that he to plans to work on html parsing over Christmas but doesn't think that has much bearning on whether it is considered sad |
| 01:04 | <jgraham> | Also BenMillard should consider redesigning the WHATWG blog because it is currently really ugly |
| 03:33 | <BenMillard> | jgraham, I'm not a designer but I could try applying the IRC log demos look-and-feel to it :) |
| 03:33 | <BenMillard> | or rather, create a demo of that |
| 05:13 | <BenMillard> | jgraham, how about this? http://projectcerbera.com/!dev/whatwg-blog/ |
| 05:15 | <BenMillard> | "entry" should have the sidebar same as any other page, imho |
| 05:16 | <BenMillard> | I removed a <style> block from the markup which was adding a background image, other than that the markup remains the same |
| 06:47 | <annevk> | BenMillard, if you read the archives (and I think you do), blog design looks great! |
| 06:47 | <annevk> | definitely a lot better than having the default |
| 06:56 | <annevk> | Hixie, setAttribute behaves differently in HTML and XML, dataset is defined in terms of setAttribute and does not do anything with the string beforehand (e.g. lowercasing) |
| 06:56 | <Hixie> | dataset isn't defined in terms of setAttribute. |
| 06:56 | <Hixie> | (well, except for the exception stuff) |
| 06:56 | <Hixie> | (but that's unrelated to what we're talking about here) |
| 06:57 | <annevk> | ah, that's confusing |
| 06:59 | <Hixie> | it is? |
| 06:59 | <Hixie> | why is it conusing |
| 06:59 | <Hixie> | what is confusing |
| 06:59 | <Hixie> | i'm confused. |
| 07:00 | <annevk> | the "If setAttribute() would have raised an exception when setting an attribute with the name name, then this must raise the same exception." made me think the algorithm was the same as for setAttribute... |
| 07:01 | <Hixie> | it would have said so if it was |
| 07:01 | <Hixie> | :-) |
| 07:01 | <annevk> | true |
| 07:02 | <annevk> | hopefully DOM core defines an abstract way of setting an attribute that we could use << zcorpan |
| 07:02 | <annevk> | that would also avoid potential confusion if setAttribute was prototyped or something like that |
| 07:03 | <Hixie> | that would be nice |
| 07:09 | <Hixie> | i worked out why that guy used such long paragraphs |
| 07:09 | <Hixie> | he in fact broke his paragraphs as much as he could |
| 07:09 | <Hixie> | it's just that some of his sentences were very long |
| 07:18 | <olliej> | Hixie: hmmm |
| 07:18 | <olliej> | Hixie: i'm not sure if hedral's technique counts as smart or stupid |
| 07:19 | <Hixie> | well it works |
| 07:19 | <Hixie> | stupid would be just bashing his head against the frame |
| 07:19 | <Hixie> | i.e. something that doesn't work |
| 07:19 | <Hixie> | it may not be optimal, though... |
| 07:19 | <olliej> | hehehe |
| 07:22 | <Hixie> | hey where in the spec does it say that <script> elements inserted by innerHTML don't execute? |
| 07:22 | <Hixie> | did i accidentally drop that at some point? |
| 07:23 | <Hixie> | oh nm found it |
| 07:26 | <olliej> | Hixie: hehe |
| 07:40 | <annevk> | olliej, is that thing from yesterday still important? |
| 07:41 | <olliej> | annevk: i was wondering what your/opera's opinion of drawImage(<video>,...) is? |
| 07:41 | <olliej> | on canvasrenderingcontext |
| 07:46 | <olliej> | annevk: the ffx nightlies already support it, and i have a prototype implementation for webkit |
| 07:48 | <annevk> | I think we're planning on doing it too |
| 07:49 | <annevk> | Maybe not when we first publicly release our <video> impl though |
| 07:50 | <olliej> | you haven't yet? |
| 07:51 | olliej | is surprised |
| 07:51 | <olliej> | does that mean safari is still the only release browser with <video> support? |
| 08:02 | <annevk> | olliej, I think so |
| 08:02 | <annevk> | well, I'm pretty sure, unless Firefox 3.1 is final now... |
| 08:03 | <olliej> | annevk: looks like 3.1 is the release for <video> -- i guess that means that opera needs an additional 2.1 revisions ;) |
| 08:05 | olliej | hides from the opera engineer who has to deal with UA shenanigans |
| 08:05 | <annevk> | heh |
| 08:11 | <MikeSmith> | annevk: any clues if the Opera 10 "Auto Update" feature actually works? Particularly "Download All Snapshots"? .. It doesn't seem to be working for me. |
| 08:11 | <MikeSmith> | or maybe I'm doing soemthing wrong |
| 08:13 | <Hixie> | opera 10 actually finally has an autoupdate? |
| 08:14 | <annevk> | MikeSmith, I don't, sorry |
| 08:14 | <annevk> | Hixie, that's the idea |
| 08:14 | <Hixie> | holy crap |
| 08:14 | <Hixie> | took y'all long enough :-P |
| 08:14 | <MikeSmith> | Hixie: in theory.. |
| 08:14 | <MikeSmith> | in practice, it ain't actually working for me at least |
| 08:14 | <Hixie> | man, i might actually upgrade opera soon then |
| 08:15 | <Hixie> | is there a "beta channel" like for other browsers? |
| 08:15 | <Hixie> | or a dev channel or something? |
| 08:17 | <MikeSmith> | Hixie: you mean IRC? |
| 08:17 | <annevk> | haha: http://websocket.org/ |
| 08:17 | <Hixie> | no i mean like with firefox and chrome you can have it autoupdate along different channels |
| 08:17 | <Hixie> | depending on how bleeding edge you want to be |
| 08:17 | <annevk> | Hixie, yeah, there's a setting |
| 08:17 | <Hixie> | cool |
| 08:18 | MikeSmith | sees that http://my.opera.com/desktopteam/blog/2008/12/19/the-christmas-edition has been updated |
| 08:18 | <MikeSmith> | "Update 2008-12-23: The builds haves not been added to the autoupdate server yet due to people going on vacation. Sorry for the false start, but we'll get back to you with automatic updates as soon as possible (depending on when people return from vacation)." |
| 08:19 | annevk | was just about to quote that |
| 08:19 | <MikeSmith> | annevk: those dudes do love web sockets |
| 09:06 | <annevk> | <noframes> can be present without <frameset> ? |
| 09:06 | <annevk> | so it seems |
| 09:24 | <jgraham> | BenMillard: That looks much better than the current look. Personally I would make it even more minimal, in the style of http://tomayko.com/ or |
| 09:28 | <annevk> | BenMillard, I think you should blow up the font size a bit though |
| 09:28 | <annevk> | BenMillard, just realized I had zoom set to 150 when I first viewed it |
| 09:37 | <Hixie> | <noframes> is always non-conforming |
| 09:37 | <Hixie> | as is <frameset> |
| 09:37 | <annevk> | oops, I meant in HTML4 |
| 09:37 | <Hixie> | ah |
| 09:40 | <Philip`> | http://www.crummy.com/2008/12/23/1 - "... html5lib (not as slow as I'd thought) ..." - that sounds encouraging |
| 09:45 | <annevk> | I'm a bit confused as to how compliant he's planning to be with HTML5 |
| 10:44 | <gsnedders> | ooo… xkcd Christmas special! |
| 11:13 | Philip` | gets horridly broken quotes when trying to reply to Hixie's latest public-html message |
| 11:14 | <Hixie> | yeah my bad |
| 11:14 | <Hixie> | pine doesn't do encodings well |
| 11:14 | <Hixie> | so anything non-ascii gets screwd up when i reply |
| 11:15 | gsnedders | struggles on with Javascript |
| 11:19 | gsnedders | wonders what he's doing wrong now |
| 11:19 | <gsnedders> | oh, I see |
| 11:19 | <gsnedders> | .length and not .len |
| 11:22 | <gsnedders> | What's the easiest to get all elements in the current dt/dd group when you have a dd element? |
| 11:26 | <MikeSmith> | Hixie: maybe Alpine does better with encodings |
| 11:27 | <Hixie> | MikeSmith: ? |
| 11:27 | <MikeSmith> | Hixie: Alpine is a fork of Pine, under a free license and with some enhancements, I think |
| 11:28 | <gsnedders> | Hixie: Replacement of Pine |
| 11:28 | <Hixie> | oh, i see |
| 11:28 | <Hixie> | yeah maybe |
| 11:28 | <gsnedders> | Hixie: One of the main points is Unicode support :P |
| 11:28 | <Lachy> | Hixie, doesn't Pine support UTF-8? |
| 11:28 | <Hixie> | i don't really care too much :-) |
| 11:28 | gsnedders | makes sure to send Hixie as much email as possible that he can't read because it's Unicode :) |
| 11:28 | <MikeSmith> | Doug McIlroy uses Alpine, so it must be good :) |
| 11:33 | <MikeSmith> | actually, fwiw, I remember now that it's Brian Kernighan that uses Alpine, not McIlroy |
| 11:35 | <gsnedders> | Does <http://pastebin.com/m11e9d4f2>; seem sane for getting a dl group? |
| 11:35 | <gsnedders> | Linus Torvalds uses Alpine, IIRC |
| 11:37 | <Philip`> | gsnedders: Write all your emails using Unicode's full-width letters, claiming it's so that readers will get the proper monospaced rendering even if they've got a silly proportional font |
| 11:37 | <gsnedders> | Philip`: :) |
| 11:41 | <gsnedders> | Now, what about that code? :P |
| 11:45 | <Lachy> | surely no one attempts to read plain text email using a proportional font?! That's crazy cause it makes ASCII art render poorly |
| 11:45 | <gsnedders> | Do Javascript arrays have any equiv of Python's list.extend? |
| 11:45 | <Lachy> | gsnedders, what does list.extend do? |
| 11:46 | <gsnedders> | it takes one parameter, another list, and adds every member of that list to the list on which you call it |
| 11:46 | <Lachy> | I think there's a join() method for that |
| 11:47 | gsnedders | finally finds in MDC what he wants |
| 11:47 | <gsnedders> | concat |
| 11:47 | <gsnedders> | join turns it into a string |
| 11:48 | <Lachy> | https://developer.mozilla.org/en/Core_JavaScript_1.5_Reference/Global_Objects/Array/concat |
| 11:48 | <Philip`> | Lachy: Gmail uses a proportional font |
| 11:49 | <Lachy> | doesn't it have an option to use a monospace font? |
| 11:49 | <gsnedders> | Why is that not listed in <https://developer.mozilla.org/en/Core_JavaScript_1.5_Reference/Global_Objects/Array>;!? |
| 11:49 | <Lachy> | gsnedders, because they moved them all to https://developer.mozilla.org/en/Core_JavaScript_1.5_Reference/Global_Objects/Array/prototype |
| 11:53 | <annevk> | Hixie, you call it both storage and storageArea... |
| 11:53 | <Hixie> | oops |
| 11:54 | <Hixie> | fixed |
| 11:54 | <Hixie> | thanks |
| 11:56 | <annevk> | the title attribute in the IDL is still wrong... |
| 11:56 | <annevk> | Hixie, ^ |
| 11:57 | <Hixie> | crap |
| 11:57 | <Hixie> | ok will fix shortly, hold on |
| 11:57 | <Hixie> | :-) |
| 12:07 | gsnedders | sighs |
| 12:07 | gsnedders | wishes toggling display of an element was easier in JS |
| 12:08 | <Philip`> | element.style.display = (element.style.display == 'none' ? 'inherit' : 'none'); I think |
| 12:09 | <Philip`> | or $(element).toggle() |
| 12:09 | <gsnedders> | That doesn't work when I need to explicitly set a special display value |
| 12:15 | gsnedders | writes a more complex function that caches the display value |
| 12:16 | Philip` | has mostly given up doing everything by hand and uses jQuery nowadays |
| 12:16 | <gsnedders> | Is it bad I've written more code for my English dissertation than for my computing project? |
| 12:17 | <Philip`> | <html xmlns:spaces xmlns:Web class="ClientFramework" web:culture="en-GB"> is amusing in far too many ways |
| 12:17 | gsnedders | would need so much custom code for this project that using JQuery for one or two functions wouldn't be worthwhile |
| 12:17 | <Philip`> | gsnedders: Yes, it is |
| 12:31 | Philip` | prefers the term "correct" rather than "pedantic" |
| 12:32 | gsnedders | wants to add a youN00b exception to Anolis |
| 12:35 | <Hixie> | gsnedders: just toggle the 'hidden' attribute and have [hidden] { display: none } in the CSS |
| 12:35 | <Hixie> | it's Compliant HTML5! (tm) |
| 12:35 | <gsnedders> | Hixie: Does that work in IE6? |
| 12:35 | <Hixie> | if you're trying to make something work in IE6, you have bigger problems |
| 12:36 | <gsnedders> | :) |
| 12:36 | <Hixie> | in ie6, just use a class instead |
| 12:36 | <gsnedders> | Thankfully, I'm not trying to do very much |
| 12:36 | <Philip`> | In IE6, use *{display:expr(this.getAttribute('hidden')?'none':'inherit'} |
| 12:38 | <Hixie> | Philip`: thanks for the data, i don't see anything there that argues for spec changes, so i won't reply |
| 12:44 | gsnedders | wonders why "quotes.parentNode.insertBefore(document.createElement("h3"), quotes);" won't work |
| 12:44 | <gsnedders> | NOT_FOUND_ERR :\ |
| 12:44 | <Hixie> | nn |
| 12:44 | <gsnedders> | nighty |
| 12:47 | <Hixie> | hm |
| 12:47 | Hixie | sees "SVG Tiny 1.2 is a W3C Recommendation" while brushing his teeth |
| 12:48 | <Hixie> | i wonder where the test cases are |
| 12:49 | gsnedders | wonders whether Hixie typed that while brushing his teeth |
| 12:49 | <Hixie> | i can only find about 600 tests |
| 12:49 | <Hixie> | which doesn't seem right |
| 12:50 | <Hixie> | how did they test a 434 page spec with only 600 tests? |
| 12:51 | <takkaria> | the spec is obviously very verbose and repetetive :) |
| 12:51 | <gsnedders> | Quite easily and quickly I expect. |
| 12:53 | <Hixie> | ok bed time for real |
| 12:53 | <Hixie> | nn |
| 12:54 | <gsnedders> | How does "magicStringSets[magicStrings[i]] = getMagicStringSet(magicStrings[i] + ":");" add one to i? |
| 12:55 | <Lachy> | I guess other groups simply don't think they need 20,000 TCs for a spec that size, or the spec just doesn't define all that much that is testable. |
| 12:57 | <Philip`> | Maybe each test case tests many features of the spec |
| 14:09 | <Philip`> | Hixie: The <script language> data is kind of suggesting a spec change |
| 16:02 | <gsnedders> | TRs aren't being served :\ |
| 16:28 | <gsnedders> | How can you see if something is in a JS array? |
| 16:31 | <gavin> | ar.indexOf(something) != -1 |
| 16:31 | <gavin> | though it doesn't work in IE |
| 16:31 | <gavin> | otherwise, there's no good way other than iterating |
| 16:31 | <gsnedders> | ergh. Yay. :\ |
| 16:32 | <gsnedders> | I need to find all the items in one array that aren't in another n arrays |
| 16:33 | <gsnedders> | But if I'm going to have to iterate this much, it may be quicker to just hide everything and then show it again. |
| 16:37 | <gsnedders> | Actually, it'll take the some number of iterations just more re-drawing |
| 16:57 | <Philip`> | gsnedders: x={}; for each (i in one_array) x[i]=1; for each (i in the other n arrays) delete x[i]; for (i in x) do something with i |
| 17:34 | <BenMillard> | annevk, the main text should be around 12px which is actually towards the larger end of what's common on the web |
| 17:35 | <BenMillard> | annevk, maybe your using a laptop with a small screen and very high pixel density? |
| 17:36 | <BenMillard> | jgraham, any more minimal and you can't tell things apart...but feel free to download the files and create your own demo |
| 17:37 | <BenMillard> | ultimately the style of the WHATWG Blog is determined by who does the work on it |
| 17:37 | <annevk> | BenMillard, I just like larger fonts, e.g. simply the browser default of 16px |
| 17:37 | <annevk> | (my laptop has a large widescreen with very high resolution) |
| 17:37 | <annevk> | "large" and "very high" |
| 17:38 | <annevk> | just trying to influence you :) |
| 17:38 | <BenMillard> | annevk, then you should use a browser which respects your DPI setting, imho, then the whole web will get scaled as you want it |
| 17:39 | <BenMillard> | does anything break when you view it at 150% zoom? |
| 17:39 | <annevk> | nope, but I rather not change settings |
| 17:40 | <BenMillard> | well, if you apply the theme to the blog then you can use as much or as little as you like from the demo :) |
| 17:40 | <BenMillard> | oh, family dinner is starting, cya round |
| 18:46 | <gsnedders> | Hmmm |
| 18:46 | <gsnedders> | alert(element); |
| 18:46 | <gsnedders> | alert(element.style); |
| 18:46 | <gsnedders> | [object HTMLElement] |
| 18:46 | <gsnedders> | undefined |
| 18:49 | <gsnedders> | Philip`: Does that change it to a string? |
| 19:02 | gsnedders | doesn't get why this isn't working |
| 19:03 | <gsnedders> | oh dud. |
| 19:03 | <gsnedders> | *duh |
| 19:03 | gsnedders | realizes |
| 19:27 | <gsnedders> | anyone have IE6? |
| 19:35 | gsnedders | gets an error in IE7 |
| 19:39 | <gsnedders> | "object doesn't support this property or method" on a line/character combination which is LF |
| 19:43 | <gsnedders> | It seems to be failing on quotes = document.getElementById("quotes"); |
| 19:58 | <gsnedders> | now "textcontent is null or not an object" |
| 19:58 | <gsnedders> | yay :\ |
| 20:00 | <jgraham> | IE doesn't/didn't support .textContent |
| 20:00 | <jgraham> | (assuming that's what you meant) |
| 20:02 | <gsnedders> | as of what version? |
| 20:02 | <jgraham> | gsnedders: No idea. I'm pretty sure 6 doesn't support it |
| 20:03 | gsnedders | is currently trying 7 |
| 20:03 | <gsnedders> | Ah, it simply doesn't support it at all |
| 20:03 | <jgraham> | (it looks like you could use something called .innerText for IE) |
| 20:06 | <gsnedders> | Is it not the case that DOM prototypes don't exist in IE? |
| 20:13 | <gsnedders> | Yeah, they don't. |
| 20:15 | <gsnedders> | Now failing on for (item in magicStringSets[magicStrings[k]]) |
| 20:16 | <gsnedders> | Can't use item as a variable |
| 20:21 | <gsnedders> | .checked on input@type=checkbox is reset on adding it to a document in IE |
| 21:07 | <Hixie> | Philip`: i couldn't work out how to decide what strings would be downplayed if we allow javascript1.x. |
| 21:14 | <ap> | hi Hixie - do you have a moment for an AppCache question? |
| 21:14 | <Hixie> | sure |
| 21:15 | <ap> | Hixie: I can't understand what this note means: "The Document with which the ApplicationCache object is associated can only itself be associated with an application cache if it is in a top-level browsing context." |
| 21:15 | <Hixie> | it's a lie |
| 21:16 | <ap> | Hixie: is it that appcache used to be tied to a top-level context, but it no longer is? |
| 21:16 | <Hixie> | yeah |
| 21:16 | <ap> | Hixie: thanks! would you like me to send an e-mail about this? |
| 21:17 | <Hixie> | i've just removed the note |
| 21:17 | <ap> | Hixie: one more question, about application cache selection algorithm - what are the examples of "f the resource is not being loaded as part of navigation of a browsing context" there? |
| 21:18 | <ap> | Hixie: I'm having hard time fully understanding this section, as "the resource being downloaded" is not defined in it |
| 21:19 | <ap> | s/downloaded/loaded/ |
| 21:21 | <Hixie> | e.g. a document downloaded by XHR |
| 21:21 | <Hixie> | there's always a resource being downloaded when the algorithm is invoked |
| 21:21 | <ap> | Hixie: will application cache selection algorithm even be invoked for XHR? |
| 21:21 | <Hixie> | sure, it's invoked by the HTML parser |
| 21:22 | <ap> | Hixie: ah well, it's just that we don't parse HTML there, only XML |
| 21:22 | <Hixie> | with xhr2 you will |
| 21:23 | <ap> | Hixie: am I right that document.write cannot be used to make an html tag with a manifest? |
| 21:25 | <ap> | Hixie: as for "there's always a resource being downloaded when the algorithm is invoked" - what about XMLParser? |
| 21:26 | <Hixie> | document.write() could be used to generate a manifest if the docment was document.open()ed first, iirc |
| 21:26 | <Hixie> | as in, if it all came from document.write() |
| 21:26 | <Hixie> | but if you see a <script> tag then you can no longer document.write() the <html> tag |
| 21:26 | <Hixie> | what about XMLParser? |
| 21:27 | <ap> | Hixie: that's a way to invoke parser (which will invoke cache selection algorithm) without downloading any resource |
| 21:28 | <ap> | Hixie: same for document.write, as you just explained |
| 21:28 | <Hixie> | (actually come to think of if document.open() wouldn't be part of navigation so it wouldn't trigger this) |
| 21:29 | <Hixie> | (it would trigger the If clause you asked about -- the "do nothing" one) |
| 21:29 | <Hixie> | in all the cases where you hit the mentions of "resource being loaded", there is a resource being loaded |
| 21:29 | <Hixie> | because if there wasn't, you would hit the first If clause and bail out |
| 21:30 | <ap> | Hixie: hmm, maybe it's good enough for a spec, but formally a clause that says "If the resource is not being loaded as part of navigation of a browsing context" doesn't match if there is no resource |
| 21:35 | <Hixie> | how can there be no resource? |
| 21:35 | <Hixie> | there's always a resource |
| 21:35 | <Hixie> | even if it's just a Document created by createDocument() |
| 21:38 | <ap> | Hixie: is "resource" formally defined somewhere? that's not how I understand that |
| 21:40 | <gsnedders> | What's the best way to loop over an object (à la for…in) but sorted in JS? |
| 21:45 | <Hixie> | ap: i'm just using it in the dictionary sense of "asset" |
| 21:46 | <ap> | Hixie: so, if a Document created with createDocument is a resource - is any Element a resource, too? |
| 21:46 | <ap> | Hixie: to me, a resource is something that can be loaded |
| 21:46 | <Hixie> | ap: *shrug* what terminology would you rather we use here? |
| 21:47 | <ap> | Hixie: to me, it would have been much easier if the resource was an explicit input parameter of the algorithm, just like manifest URL is, and it were specified at call sites |
| 21:48 | <Hixie> | i guess i can do that |
| 21:48 | <ap> | Hixie: cool, thanks! |
| 21:56 | <Hixie> | done |
| 22:00 | <ap> | looks good |
| 22:14 | <Philip`> | gsnedders: You might need to be careful to do "var quotes = ...", since window.quotes is the element with id="quotes" or something like that |
| 23:10 | <Hixie> | hsivonen: yt? |
| 23:11 | <gsnedders> | jgraham: yt? |
| 23:12 | <Hixie> | wow, svg doesn't define the content model of <foreignObject> at all |
| 23:33 | <deltab> | gsnedders: sorted property names? |