| 00:41 | <heycam> | we'll have catmull-rom curves in SVG2 btw |
| 05:02 | <Hixie> | do we have a way to detect when the user's mouse goes outside the window frame? |
| 05:05 | <Hixie> | oh duh i had a typo in my test |
| 05:05 | <Hixie> | nevermind |
| 05:18 | <Hixie> | how about composition effects... is there a way to make a CSS element's background be a semi-transparent blur? |
| 05:30 | <asmodai> | Hixie: semi-transparent, yes, but haven't seen a way to do a blur outside of pre-rendered images |
| 05:31 | <asmodai> | But that's just my meager knowledge. |
| 05:31 | <Hixie> | i'm trying to do the effect mac does with its context menus |
| 05:31 | <Hixie> | from css |
| 06:00 | <Von_Davidicus> | Question: How many websites use SSI these days? |
| 06:25 | <zcorpan> | how do i create a file with a LF in the file name on mac? |
| 07:54 | <zcorpan> | Hixie: css filters? |
| 08:01 | <wirepair> | anyone on who knows about Hixie's DOM testing page? |
| 08:01 | <wirepair> | can't seem to find it for some reason |
| 08:04 | <wirepair> | nm found it (ref: http://software.hixie.ch/utilities/js/live-dom-viewer/) |
| 09:00 | <jgraham> | Is there some way to see a squashed diff of multiple commits on github? |
| 09:02 | <Ms2ger> | I can tell you how to do it with hg? :) |
| 09:04 | <AryehGregor> | Ms2ger, on github? :) |
| 09:05 | <AryehGregor> | That would be pretty impressive! |
| 09:05 | <Ms2ger> | AryehGregor, "now you've got three problems" |
| 09:09 | <jgraham> | Ms2ger: I think "checkout the repo using hg-git and us hg diff" isn't the answer I'm looking for :p |
| 09:09 | <jgraham> | (obviously git diff *also* allows viewing a diff across multiple commits) |
| 09:11 | <SimonSapin> | Hi. I started a proposal for replacing the selector grammar with something based on css3-syntax: http://lists.w3.org/Archives/Public/www-style/2012Jun/0159.html |
| 09:11 | <SimonSapin> | What are the usual tools for editing specs? For cross-references in particular. |
| 09:12 | <annevk> | the CSS WG uses something quite close to Anolis |
| 09:12 | <Ms2ger> | The CSSWG uses Bert's W3C-Member-Only tool |
| 09:12 | <Ms2ger> | But I recommend Anolis :) |
| 09:12 | <annevk> | http://wiki.whatwg.org/wiki/Anolis has documentation on it's open sourced equivalent |
| 09:13 | <SimonSapin> | that, I’ll look into that |
| 09:15 | <annevk> | https://www.w3.org/Style/Group/css3-src/bin/postprocess (W3C Member-only) has information on the tool some in the CSS WG use fwiw |
| 09:16 | <SimonSapin> | I’m not a W3C member or in a member organization |
| 09:16 | <annevk> | k |
| 09:16 | <annevk> | it's kind of a sucky tool to be honest |
| 09:16 | <annevk> | that's why Anolis was created, which is not that much better, but at least you can "easily" run it locally |
| 09:18 | <annevk> | SimonSapin: looks cool btw |
| 09:18 | <SimonSapin> | I wanted to make it easy for the CSSWG to take this eventually if they want to. But maybe I should just use docutils and not worry about the format later |
| 09:18 | <SimonSapin> | annevk: thanks |
| 09:19 | <annevk> | SimonSapin: if you write it in HTML it should be easy enough to take in |
| 09:19 | <SimonSapin> | docutils / reStructuredText can produce HTML |
| 09:20 | <annevk> | SimonSapin: you can study the source of http://dev.w3.org/csswg/css3-syntax/Overview.src.html for what e.g. css3-syntax uses as input for its processor that produces the final document |
| 09:20 | <annevk> | s/e.g.// |
| 09:20 | <SimonSapin> | and reverse-engineer a tool to do the same conversion? |
| 09:20 | <annevk> | no :) |
| 09:21 | <annevk> | it's readable either way, it just misses the linking/table of contents |
| 09:21 | <annevk> | but that would make it trivial to integrate with the rest of CSS |
| 09:24 | <zcorpan> | SimonSapin: nice |
| 09:24 | <SimonSapin> | but I want the links in the meantime :) |
| 09:24 | <annevk> | SimonSapin: then use Anolis :) |
| 09:25 | <annevk> | SimonSapin: you can use an online interface: http://pimpmyspec.net/ |
| 09:25 | <annevk> | SimonSapin: it takes the same input as what that css3-syntax source document uses |
| 09:25 | <annevk> | SimonSapin: or you can install it locally |
| 09:26 | <SimonSapin> | annevk: are http://hg.hoppipolla.co.uk/hgwebdir.cgi/anolis/ and https://bitbucket.org/ms2ger/anolis the same? |
| 09:26 | <Ms2ger> | The former is what runs pimpmyspec, the latter is maintained :) |
| 09:26 | <SimonSapin> | ok |
| 09:28 | <kennyluck> | SimonSapin, hey hi. I was sending a mail to you and then I just noticed you talking here. |
| 09:28 | <SimonSapin> | kennyluck: yes, I just answered your email |
| 09:29 | <SimonSapin> | oh, the previous one |
| 09:31 | <annevk> | SimonSapin: if you install it locally use the latter |
| 09:32 | <SimonSapin> | annevk: I’ll do that |
| 09:32 | <annevk> | I've been trying to get jgraham to do something about this situation but he's busy :( |
| 09:33 | <Von_Davidicus> | Question: How many websites still use Server Side Includes? |
| 09:34 | <annevk> | Von_Davidicus: how would you measure that? |
| 09:36 | <Von_Davidicus> | Well, by how many websites that have webpages that have the .shtml extension. |
| 09:40 | <Von_Davidicus> | I was just wondering if SSI is still relatively common, considering the proliferation of other server-side languages that can do a lot more. |
| 10:14 | <AryehGregor> | I suspect it's common in absolute terms but rare in relative terms. Probably a very small percentage of pages use it (<5%), but a very large number (>1,000,000). |
| 10:19 | <jgraham> | FWIW I think looking for .shtml isn't going to help you |
| 10:20 | <jgraham> | (random fact: until quite recently bbc.co.uk ran on a complex mess of SSI because it was so hard to get new software approved) |
| 10:32 | <annevk> | you also don't need a particular extension to activate it, depends mostly on how you configure your Apache installation |
| 11:31 | <annevk> | Velmont: http://dvcs.w3.org/hg/cors/rev/9ac9da9be418 |
| 11:55 | <jarek> | why there are so many factory methods in SVG DOM? |
| 11:56 | <jarek> | this makes it impossible to inherit from most of the standard DOM objects |
| 11:57 | <jarek> | I would expect at least something like "point = new SVGPoint" to work |
| 11:58 | <jgraham> | Java |
| 12:02 | <annevk> | jarek: SVG was designed with Java in mind |
| 12:02 | <annevk> | jarek: no longer true these days, but you know, legacy |
| 12:03 | <jarek> | what's even worse, there doesn't seem to be an easy way to abstract this away |
| 12:04 | <jarek> | I mean if I create wrapper for SVGMatrix then I will not be able to pass it as argument to other SVG DOM methods |
| 12:05 | <jarek> | s/wrapper/wrapper object |
| 12:12 | <annevk> | okay |
| 12:12 | <annevk> | I'm going to do this CORS security considerations thing one more time |
| 12:13 | <annevk> | and then it better be fricking perfect or someone else can start maintaining it |
| 12:13 | <annevk> | at least until I get bored enough to write "Fetch" |
| 12:13 | <jarek> | wouldn't it make more sense if factory methods were specified on SVGPathElement instead of SVGPathElement.prototype |
| 12:13 | <annevk> | (and enough years have passed) |
| 12:14 | <annevk> | jarek: it's better to discuss SVG elsewhere probably |
| 12:14 | <annevk> | there's hardly any SVG guys around here |
| 12:14 | <annevk> | jarek: that is, we're happy with you discussing SVG here, but you might not make meaningful progress |
| 12:14 | <smaug____> | shepazu and heycam|away might be around |
| 12:15 | <smaug____> | but w3 #svg would be better place |
| 12:25 | <kennyluck> | or just www-svg |
| 12:25 | <jarek> | there are 8 persons on #svg@w3c, I have also tried looking for most common SVG methods on Github and Google and it looks like barely anyone is using them |
| 12:41 | <annevk> | so guess CORS text has to wait a bit; not convinced this new text provided via PDF/Word (?!) is an improvement |
| 12:43 | <Velmont> | annevk: Bounce it back. Ask for a real format. |
| 12:43 | <annevk> | asked for better text too |
| 13:29 | <annevk> | heh http://www.ietf.org/proceedings/42/slides/notify-bof-scenarios-slides-98aug/sld016.htm |
| 13:30 | <annevk> | maybe ours should be called "Platform notifications" |
| 13:31 | <annevk> | or "End-user notifications" possibly |
| 13:56 | Ms2ger | shakes his fist at annevk |
| 13:56 | <annevk> | what did I do? |
| 13:57 | <annevk> | or maybe I should strike first, then ask questions... |
| 13:58 | <Ms2ger> | This Gecko patch you wrote in 2005 |
| 13:59 | <annevk> | heh |
| 13:59 | <Velmont> | lul |
| 13:59 | <annevk> | it's still haunting Bugzilla? |
| 13:59 | <Ms2ger> | I just noticed you forgot to change some layout code |
| 14:00 | <Ms2ger> | (Can't have been terribly important...) |
| 14:02 | <annevk> | Ms2ger: oh so I actually managed to get my patch landed? |
| 14:02 | <Ms2ger> | timeless landed it |
| 14:03 | <annevk> | oh I'm in about:credits |
| 14:03 | <Velmont> | annevk: boom |
| 14:03 | <annevk> | sweet |
| 14:03 | <annevk> | but under the wrong letter :( |
| 14:03 | <Ms2ger> | You sort under K? |
| 14:03 | <annevk> | yeah |
| 14:03 | <Velmont> | It's like "the" |
| 14:03 | <Ms2ger> | Silly Dutchies |
| 14:03 | <Velmont> | Dutchess |
| 14:03 | <Velmont> | :P |
| 14:03 | <Ms2ger> | More like "from", Velmont |
| 14:04 | <annevk> | silly or not, bug in Firefox :p |
| 14:04 | <Velmont> | Ms2ger: I meant in the sorting sense. |
| 14:04 | <Ms2ger> | Bah |
| 14:04 | <Ms2ger> | Hah, even |
| 14:05 | <Velmont> | Dutchess Anne, ruler of the Lands in the Nether. |
| 14:07 | <jgraham> | So if we submit a patch to fix annevk's position in about:credits, will we be able to get into about:credits? |
| 14:08 | <Velmont> | jgraham: If you do something wrong there, I can fix your fix. about:credits inception :] |
| 14:09 | <Ms2ger> | 'This is a list of people who have "made a significant investment of time, with useful results, into Mozilla project-governed activities".' |
| 14:09 | <jgraham> | I wonderif there is anyone who has got a patch into Opera/Gecko/WebKit/Trident |
| 14:09 | <zewt> | Ms2ger: that sure sounds like it was written by committee |
| 14:09 | <jgraham> | Ms2ger: If you don't sort annevk's name correctly he might invade Belgium |
| 14:10 | <jgraham> | Averting that is surely a useful result |
| 14:10 | <Ms2ger> | jgraham, he wouldn't, because we'd sort him under V here :) |
| 14:10 | <jgraham> | Yeah, that's why he'd invade |
| 14:10 | <jgraham> | To correct your sorting algorithm |
| 14:12 | <zewt> | wars have been started for lesser causes |
| 14:13 | <zewt> | they use emacs? BOMB THEM |
| 14:15 | <annevk> | :) |
| 14:18 | karlcow | thinks that if annevk wants to cope with the real world, he has to change sex |
| 14:18 | <karlcow> | anne5 |
| 14:19 | <annevk> | nah, the Netherlands once ruled a large part of the world |
| 14:19 | <annevk> | surely we can do that again |
| 14:19 | <Ms2ger> | Once |
| 14:21 | <karlcow> | heh |
| 14:22 | <Ms2ger> | jgraham, I'd guess fantasai is most likely to have patches in all four |
| 14:23 | <Ms2ger> | Unless Hixie sneaked something into Trident at some point |
| 14:23 | <jgraham> | Oooh, she worked for Microsoft for a bit, didn't she |
| 14:24 | <Ms2ger> | She contracted with everyone and HP at one point, IIRC |
| 14:36 | <timeless> | heh |
| 14:36 | <timeless> | Chrome Canary is showing me "timeless -2 of 5" |
| 14:36 | <timeless> | before it was showing me "timeless 30 of 5" |
| 14:39 | <timeless> | annevk: i don't think i'm responsible for your missorting |
| 14:40 | <annevk> | timeless: no worries :) |
| 14:41 | <timeless> | annevk: speaking of interesting quirks in sorting and such |
| 14:41 | <timeless> | for your preferred language, are there any fun gotchas? |
| 14:41 | timeless | is filing localization bugs |
| 14:41 | <timeless> | gotchas must be easy to search for :) |
| 14:42 | <timeless> | oh, on the subject of getting features into systems, i've had people add features to windbg for me :) |
| 14:43 | <Ms2ger> | Duh, you're its only user ;) |
| 14:44 | <timeless> | =b |
| 14:54 | <annevk> | oh god |
| 14:54 | <annevk> | URLs are teh pain |
| 14:54 | <annevk> | a = <a href=/test/></a>; a.protocol = "mailto"; alert(a.href) |
| 14:54 | <annevk> | now change "mailto" to "ma<ilto" or some such |
| 14:55 | <Ms2ger> | No thanks :) |
| 15:09 | <annevk> | MikeSmith: hey man |
| 15:10 | <annevk> | MikeSmith: so what are the notes you have on URLs? |
| 15:10 | <annevk> | MikeSmith: if you could put tests and other links on http://wiki.whatwg.org/wiki/URL that'd be nice |
| 15:10 | <annevk> | MikeSmith: I'll go ahead and create that page to link to the spec as a placeholder |
| 15:50 | <timeless> | annevk: is it safe to assume that Teune van Steeg should sort under S? |
| 15:51 | <timeless> | Michiel van Leeuwen is sorted under L |
| 15:51 | <annevk> | timeless: if he's Dutch :) |
| 15:51 | <timeless> | Peter van der Woude is probably wrong too |
| 15:51 | <timeless> | Gert-Paul van der Beek is under B |
| 15:51 | <timeless> | and Tim Maks van den Broek |
| 15:52 | <timeless> | Frank van der Linden is under L |
| 15:52 | <timeless> | Erik van der Poel is under P |
| 15:52 | <Ms2ger> | peterv would sort under V, though :) |
| 15:52 | <timeless> | HJ van Rantwijk is under R |
| 15:52 | <timeless> | Ms2ger: why? |
| 15:52 | <Ms2ger> | He's Belgian |
| 15:52 | <annevk> | timeless: Google search for Teune suggests he's Dutch |
| 15:52 | <annevk> | got to go |
| 15:53 | timeless | sighs |
| 15:53 | <timeless> | you guys are collectively confusing |
| 15:54 | <Ms2ger> | Just sort by IRC handle :) |
| 15:55 | <timeless> | :) |
| 15:55 | <timeless> | i currently sort after Trần Nguyễn Sơn |
| 16:03 | <jgraham> | Tried to sort a list by first name here once and generated a non-trivial amount of confusion from people who are better known by irc handle, nickname, or surname |
| 16:05 | <jgraham> | Hixie: Why does step 4 in "Navigating across documents" only apply if the navigation is cross-origin? |
| 16:21 | <Hixie> | jgraham: so that a site can't stop you from leaving a site |
| 16:26 | <jgraham> | Hixie: So what stops onunload handlers being called recursively in general? |
| 16:28 | <zewt> | how is it 2012 and popunders are still possible |
| 16:31 | <jgraham> | Hixie: (the motivating question is "given a site that has a navigation triggered in an unload handler, will you end up at the original destination, or at the destination of the navigation triggered in unload) |
| 16:31 | <jgraham> | s/)/")/ |
| 16:38 | <sm0lm> | Hi... What's the rational behind killing <br /> style tags off, and the more general break with XMLism that seems to be happening? |
| 16:39 | <jgraham> | sm0lm: The don't and, for historical reasons, can't, do anything special in the HTML parser |
| 16:39 | <sm0lm> | jgraham, "the don't..."? |
| 16:39 | <sm0lm> | "they"? |
| 16:40 | <sm0lm> | (not trying to be nitpicky, just double checkin what you mean.) |
| 16:57 | <Hixie> | jgraham: why would they be called recursively? |
| 16:58 | <Hixie> | sm0lm: we didn't "break" with xmlisms, we just never had xmlisms in html |
| 16:59 | <sm0lm> | oh, hi there Hixie. |
| 16:59 | <Hixie> | sm0lm: actually in some ways we added some (e.g. we made /> on <br> be allowed, though it does the same as not having it) |
| 17:00 | <sm0lm> | Right - but there's all this stuff about how it's dangerous to send XHTML as html, and all that. I'm having some trouble sorting through what the situation actually is. |
| 17:01 | <Hixie> | simplest way of looking at it is that there's just text/html |
| 17:01 | <sm0lm> | Hixie, Why is there all of this backwards-compatibility business of having a different version of the spec for browser vendors and webdevs? |
| 17:02 | <Hixie> | there's only one spec, the webdev version is just the spec with a few things removed (the browser-specific stuff) |
| 17:02 | <sm0lm> | But why is it seen as a constraint to keep things compatible with older web content? |
| 17:03 | <Hixie> | browsers wouldn't get any market share if they implemented a spec that made cnn.com not work |
| 17:04 | <Hixie> | so if a spec said to do something that made cnn.com not work, the spec would be ignored |
| 17:04 | <sm0lm> | It's an entirely separate document namespace within which HTML5 markup resides... |
| 17:04 | <Hixie> | well, that's why it's important that we don't work on "HTML5" but just on "HTML" |
| 17:04 | <sm0lm> | In other words (the way I think it should work is that) if you're opening with a <!doctype html> declaration, then you better just adhere to the spec you're claiming to use. |
| 17:05 | <zewt> | (that's not what the html doctype means) |
| 17:05 | sm0lm | is fully in favor of any spec which breaks such an insidious corporate propaganda machine as CNN.com |
| 17:05 | <sm0lm> | =) |
| 17:05 | <zewt> | it pretty much means "this is html" and "don't dump me in quirks mode", i think |
| 17:06 | <Hixie> | sm0lm: how many versions of html would you want browsers to implement? |
| 17:07 | <Hixie> | sm0lm: (note that each such version would basically be an entirely new browser, so each one would require a new dev team, new testing team, etc) |
| 17:07 | <zewt> | (if it was possible to get into that mode without a doctype at all, then that's probably what would be recommended) |
| 17:07 | <sm0lm> | This seems to be a fairly widely limiting factor in moving fwd with web standards... ES6 is facing it, and so is the new HTML draft. |
| 17:08 | <sm0lm> | BTW, how does one refer to post-HTML5 HTML without also referring to pre-html5 html? |
| 17:08 | <sm0lm> | Hixie, I don't follow you... :( |
| 17:09 | <sm0lm> | I mean, the ES6 proposal is running into syntax clashes and so they're having to use suboptimal syntaxes so as not to break older ES scripts. |
| 17:09 | <Hixie> | i refer to "contemporary HTML" to refer to the HTML from the last decade or so rather than the HTML from more than 10 years ago (which had versions) |
| 17:09 | <Hixie> | backwards compatibility isn't a problem, imho |
| 17:09 | <sm0lm> | Hixie, "HTML5" is much more concise. |
| 17:10 | <Hixie> | "HTML5" makes people think there are versions |
| 17:10 | <Ms2ger> | Meh |
| 17:10 | <sm0lm> | I feel like we should just do a full "reset", start from scratch, building a spec from the ground up without consideration for older material not rendering... |
| 17:11 | <Hixie> | sm0lm: that's what xhtml2 tried |
| 17:11 | <Hixie> | sm0lm: it failed |
| 17:11 | <Hixie> | sm0lm: the only way it would succeed is if the new thing was so radically better that it was worth the huge cost |
| 17:11 | <Hixie> | sm0lm: good luck :-) |
| 17:11 | <Hixie> | sm0lm: while you're doing that i'll keep maintaining html :-) |
| 17:11 | <sm0lm> | IE, dev'mt would halt on the rendering support for older content -- ie, the implementations would still get shipped, just not a whole lot of improvements. |
| 17:11 | <sm0lm> | Wait, please hear me out. |
| 17:12 | <sm0lm> | I was in here with another name asking about <u>s and <i>s and <b>s |
| 17:12 | <sm0lm> | a little while back. |
| 17:13 | <sm0lm> | I mean as long as the handling of legacy content doesn't become worse than it was at the time of said content's authoring, nothing will break. |
| 17:14 | <Hixie> | in each browser, things are already broken |
| 17:14 | <Hixie> | and the browser vendors really want to fix what's broken |
| 17:14 | <Hixie> | so it's not like we could convince them to just stop working on it |
| 17:14 | <sm0lm> | Then things that get delivered with a Content-Type of text/n[ew]html will have to adhere to a new standard that, while it may be loosely based on the old stuff, isn't bound to remaining compaitble with it. |
| 17:14 | <sm0lm> | It could have DRYer, HAML-like syntax. |
| 17:14 | <Ms2ger> | Hixie, (unrelatedly, I bet you're glad to hear that we still use tests bearing your name for XML perf) |
| 17:15 | <Hixie> | switching has a _huge_ cost. to make it worth it, the new tech would have to be orders of magnitude better. |
| 17:15 | <Hixie> | nobody has come up with anything better enough yet. |
| 17:15 | <sm0lm> | Then the ever-evolving nhtml spec would not conflict with previous versions of *itself* |
| 17:16 | <Hixie> | how would we avoid this new html spec from running into the same problems with itself? |
| 17:16 | <sm0lm> | Well Hixie, look at it this way: 5 years from now, "contemporary HTML" will be orders of magnitude than it was 5 years /ago/ |
| 17:17 | <Hixie> | (in developing HTML in the last ten years we've made mistakes that we've had to be backwards compatible with, it's not just an issue of being compatible with HTML4) |
| 17:17 | <Hixie> | sm0lm: right. but you have to be better than _that_ in order to make it worth switching. |
| 17:17 | <sm0lm> | If we change the trajectory now so that 5 years from now we aren't still constrained by shit from 20 years BACK, we could go much further. |
| 17:17 | <Hixie> | sm0lm: no we couldn't, because we'd have to start over, so we'd be 5 years behind to start with. |
| 17:18 | <Hixie> | sm0lm: and we'd still have to deal with backwards compatibility with the new stuff |
| 17:18 | <Hixie> | sm0lm: so 10 years later we'd be back where we are now |
| 17:18 | <Hixie> | sm0lm: in terms of being constrained |
| 17:18 | <sm0lm> | Oh, c'mon... We've learned SO many lessons from the early days of the web. |
| 17:19 | <Hixie> | we are constrained by things we did just two or three years ago |
| 17:19 | <Hixie> | not just 10, 20 years ago |
| 17:19 | <sm0lm> | I don't think we'll learn keep learning them at the same rate.. I think the mistakes that get made and the lessons that get learned diminish asymptoptically. |
| 17:19 | <sm0lm> | Hixie, example? |
| 17:19 | <Ms2ger> | Not to count the number of things we're constrained by that Hixie did |
| 17:20 | <Hixie> | pushState()'s second argument, for example |
| 17:20 | <Hixie> | localStorage's locking |
| 17:20 | <Hixie> | tons of stuff |
| 17:20 | <sm0lm> | Isn't that all the more reason to do what I'm suggesting? |
| 17:20 | <Hixie> | how would what you're suggesting change this? |
| 17:20 | <Hixie> | we'd invent your new language |
| 17:21 | <Hixie> | promptly make a mistake |
| 17:21 | <Hixie> | and be constrained again |
| 17:21 | <sm0lm> | If 2 or 3 years time yields so many conflicts, why would you EVER want to have 20 or 30 years' worth's on your shoulders? |
| 17:21 | <Hixie> | because every time you start over you increase your workload 100% |
| 17:22 | <Hixie> | because you effectively now have to maintain two browsers |
| 17:22 | <Hixie> | then three |
| 17:22 | <Hixie> | then four |
| 17:22 | <sm0lm> | Perhaps a periodic "reset" is in order then. |
| 17:22 | <Hixie> | then five |
| 17:22 | <Hixie> | etc |
| 17:22 | <sm0lm> | Once every 10 years. |
| 17:22 | <sm0lm> | Why do you need to maintain them? |
| 17:22 | <Hixie> | because there's 2 trillion existing web pages that people will still want to look at |
| 17:22 | <sm0lm> | They'll just be for legacy content and so as I outlined above, they simply shouldn't regress. |
| 17:22 | <Hixie> | so you have to fix bugs that make them not work |
| 17:23 | <sm0lm> | If they displayed in a satisfactory way to the web designer authoring them during testing, then they'll continue to display at least that well for ever and ever. |
| 17:24 | <Hixie> | browser vendors wouldn't agree to not fix bugs |
| 17:25 | sm0lm | thinks that an awful lot of those 2 trillion - perhaps a great majority - are not static files and are generated by programs onthefly based on textual content. Like all the articles on CNN.com, for instance. When the CMS gets updated which it will if it isactively maintained, your 2T figure diminishes greatly. |
| 17:25 | <Hixie> | because all it takes is one of the vendors to fix a few more bugs, and the others will be forced to do it as well to compete |
| 17:25 | <zewt> | i hope nobody would encourage that, either :) |
| 17:25 | <Hixie> | are you going to pay cnn.com to update to the new tech? :_) |
| 17:25 | <Hixie> | what benefit does cnn get for updating? |
| 17:25 | <sm0lm> | What's :_) mean? (sorry.) |
| 17:26 | <Hixie> | other than it not working with any legacy browsers? |
| 17:26 | <sm0lm> | tear dripping down the right cheek..? |
| 17:26 | <Hixie> | :_) is :-) with the shift key down by mistake |
| 17:26 | <sm0lm> | ah, ok. |
| 17:26 | <sm0lm> | ;_) |
| 17:27 | <sm0lm> | Easier maintenance of their view templates. |
| 17:27 | <sm0lm> | A MUCH DRYer syntax if the HAML thing is implemented. |
| 17:27 | <Hixie> | rewriting their entire site isn't easier than doing nothing |
| 17:27 | <sm0lm> | The entire site doesn't need to be rewritten --- only the MVC views. |
| 17:28 | <Hixie> | and presumably they also have to keep maintaining the old site as well? since no browser in the hands of users would support this new technology at first? |
| 17:28 | <zewt> | a fundamental concept on the web is that when you write a webpage, you don't have to keep coming back to it periodically to make it work again; you write it once, it works, and it works forever |
| 17:30 | <sm0lm> | Hixie, ok, ok. I see how this works. it's really a shame though and it's upsetting to accept that we just have to live with all of these limitations holding the web back and there's really no way a round it. |
| 17:30 | <sm0lm> | Thank you all for your patience with my naivete. |
| 17:31 | <zewt> | what's being held back, exactly? |
| 17:31 | <zewt> | web-compat is an annoyance, but not a fatal one |
| 17:32 | <sm0lm> | Would something like HAML syntax be possible to implement clientside interpretation for? |
| 17:33 | <sm0lm> | I mean, given the current iterative evolutionary development paradigm of the web? |
| 17:36 | <sm0lm> | Hixie, What's your stance on Dart? |
| 17:38 | Ms2ger | suggests Hixie do work instead :) |
| 17:39 | <sm0lm> | What is anyone in here's stance on it then? i mean its an example of a revolutionary vs evolutionary approach... |
| 17:39 | <Hixie> | Ms2ger: work? dude, it's wwdc keynote time, who's doing work |
| 17:40 | <Ms2ger> | sm0lm, yeah, and it sucks :) |
| 17:40 | <Ms2ger> | What's wwdc? ;) |
| 17:40 | <sm0lm> | Hixie, hahaha. |
| 17:41 | <sm0lm> | Hixie, do you hold a critical view of Dart in the same way as what I just proposed? |
| 17:41 | <Hixie> | if i did, it would probably be unwise for me to say so publicly since i work for the company that's pushing it |
| 17:43 | <Hixie> | from testing, it looks like http://www.w3.org/mid/4F62C516.2000707⊙me isn't limited to the current page load |
| 17:45 | <Hixie> | this is some crazy stuff |
| 17:45 | <Hixie> | i have to say "if it's locally cached"? |
| 17:45 | <Hixie> | man |
| 17:45 | <Ms2ger> | Welcome to the web :) |
| 17:45 | <Hixie> | this exposes the web history so blatently |
| 17:45 | <Hixie> | it's not even a timing attack anymore |
| 17:46 | <zewt> | isn't that pretty much impossible to spec interoperably? (because it'll never happen on a browser with no cache) |
| 17:46 | <zewt> | (putting aside data:) |
| 17:46 | <Hixie> | it already doesn't happen on a browser with no cache |
| 17:47 | <zewt> | i'd suspect (hope?) that web-compatibility only actually requires it to be synchronous if the image is loaded in the current page load |
| 17:47 | <zewt> | (no idea, but that should be enough for pages that preload thumbs, etc. by loading a bunch of HTMLImageElements in the background) |
| 17:48 | <Hixie> | i guess i could spec a "list of available images" |
| 17:48 | <Hixie> | and allow the UA to prefill it with same-origin images |
| 17:48 | <zewt> | i guess the bigger question is whether it needs to be "all images loaded in this page load", or simply "all images that are currently loaded" |
| 17:49 | <zewt> | since the former implies a (minor) memory leak |
| 17:53 | <Yuhong> | <Hixie> "HTML5" makes people think there are versions |
| 17:54 | <Yuhong> | <Hixie> there's no advantage to changing the doctype line unless you are changing it from something that used to trigger quirks mode |
| 17:54 | <Yuhong> | Yea, I am thinking that "HTML5" even as a buzzword is a misnomer for these reasons. |
| 17:55 | <Yuhong> | Talking in terms of individual features is much better. |
| 17:56 | <Yuhong> | For example, as I said before, canvas dates back to 2005. |
| 17:57 | <Hixie> | ho ho, it's not cross-tab |
| 17:58 | <Hixie> | might even be only within the current top-level browsing context |
| 17:58 | <Yuhong> | Just tweeted to EricLaw: |
| 17:58 | <Yuhong> | https://twitter.com/yuhong2/status/212242605625446400 |
| 17:59 | <Yuhong> | Hixie: Now, there are people who formerly worked on IE at MS in Google too. |
| 17:59 | <Yuhong> | Hixie: And you work there too. |
| 17:59 | <sm0lm> | Yuhong, sorry did i miss anything since you started chiming in...? my connection blipped out and i got your first message after zewt's about memory leaks. |
| 18:00 | <Yuhong> | sm0lm: No, you didn't |
| 18:01 | <Yuhong> | Hixie: for example, jbeda who worked on IE's DOM: |
| 18:01 | <Yuhong> | http://news.ycombinator.com/item?id=3233935 |
| 18:06 | <Yuhong> | Someone suggested versioning authoring but not implementation requirements. |
| 18:07 | <sm0lm> | I think I'd be in favor of that. |
| 18:08 | <Yuhong> | Hixie: And there is also Chris Wilson. Have you talked with them? |
| 18:10 | <Hixie> | Yuhong: i've spoken with chris and others, sure |
| 18:51 | <sarspazam> | hi guys. I've got a dissertation to start thinking about in the coming months. looking to make some kinda web app with html5 and asp.net mvc |
| 18:51 | <sarspazam> | be interested to hear any ideas of something I could make |
| 18:53 | <annevk> | cool game |
| 19:11 | <jgraham> | Hixie: I presume it would be recursive (naively) because starting a navigation from unload would eventually call unload again, which would start a navigation... |
| 19:12 | <Hixie> | it is possible to have an infinite sequence, but it doesn't recurse |
| 19:12 | <Hixie> | that is, they're not nested |
| 19:12 | <Hixie> | just sequential |
| 19:12 | <Hixie> | iirc |
| 19:14 | <jgraham> | Hixie: But that doesn't seem to match browsers afaict. That is if you have onload="location='data:text/html,1'" and onunload="location='data:text/html,2'" you will eventually end up at 1 or 2 |
| 19:14 | <jgraham> | s/eventually// |
| 19:14 | <Hixie> | sure |
| 19:14 | <Hixie> | 2, specifically |
| 19:14 | <Hixie> | iirc |
| 19:14 | <Hixie> | that's fine |
| 19:15 | <Hixie> | where's the confusion there? |
| 19:15 | <Hixie> | maybe i'm missing something |
| 19:16 | <jgraham> | The confusion is that you end up at 1 |
| 19:16 | <jgraham> | So something must cancel the navigation in the unload |
| 19:16 | <Hixie> | send mail and i'll look into it in more detail |
| 19:17 | <Hixie> | ideally with convenient test cases i can just load :-) |
| 19:17 | <jgraham> | If nothing cancelled the navigation, I don't see why you wouldn't just end up calling the unload handler over and over |
| 19:17 | <jgraham> | Sure |
| 19:18 | <Hixie> | (you end up at 1? really?) weird) |
| 19:19 | <jgraham> | data:text/html;charset=utf-8,%3C!DOCTYPE%20html%3E%0A%3Cbody%20onload%3D%22location%3D'data%3Atext%2Fhtml%2C1'%22%20onunload%3D%22location%3D'data%3Atext%2Fhtml%2C2'%22%3E in gecko + webkit at least |
| 19:19 | <Hixie> | can't look right now, sorry |
| 19:19 | <Hixie> | but send mail and i'll look this afternoon |
| 19:20 | <jgraham> | I guess I will send mail in the morning :) |
| 19:38 | <annevk> | so now all MacBook's are without optical drive? |
| 19:41 | <Hixie> | only the new macbook pros and the small macbook airs |
| 19:42 | <annevk> | which other MacBook's are there? |
| 19:42 | <othermaciej> | the "legacy" non-retina MacBook Pros |
| 19:42 | <Hixie> | the new non-retina macbook pros have dvd drives |
| 19:42 | <annevk> | it does not say so on the web pages |
| 19:43 | <annevk> | http://www.apple.com/macbook-pro/specs/13-and-15-inch/ does not mention cd, dvd, or optical |
| 19:43 | <annevk> | and the view from the side does not show it (although only one side is shown) |
| 19:46 | <Hixie> | if you try to buy one you'll see it in the specs |
| 19:47 | <annevk> | had not tried that |
| 19:48 | <annevk> | I guess per othermaciej that one is prolly correct |
| 20:37 | <moo-_-> | annevk: they have no optical AFAIK |
| 20:37 | <moo-_-> | that would cost... extra! |
| 20:37 | <moo-_-> | ;) |
| 20:39 | <paul_irish> | annevk: whyd you ping me about h5bp's nonvalidation? relatedly. can X-meta tags _not_ throw a validation error? (a warning seems a bit much even) |
| 20:40 | <annevk> | paul_irish: thought you worked on that |
| 20:41 | <annevk> | paul_irish: for the latter you want to bug hsivonen and MikeSmith |
| 20:41 | <sm0lm> | annevk, try grepping for "super" |
| 20:41 | <paul_irish> | annevk: cuz of issues like that i dont really aim for 100% valid. |
| 20:41 | <paul_irish> | aiight |
| 20:42 | <annevk> | paul_irish: fair enough, though the doctype could be somewhat better ;) |
| 20:42 | <paul_irish> | "the doctype is amazing" are the words i think you're looking for |
| 20:42 | <paul_irish> | :D |
| 21:03 | <jgraham> | Also, what is bz doing sending useful technical feedback to public-html |
| 21:03 | <jgraham> | How can I ignore it if he insists on throwing in curveballs like that |
| 23:11 | <Hixie> | MikeSmith: cvs being slow again |
| 23:32 | <Hixie> | wait, hober, your CP is actually just reverting the revert? i.e. it makes the whatwg spec and the w3c spec converge? |
| 23:32 | <Hixie> | that makes this thread even more ridiculous |