| 00:00 | <annevk> | also, https://bugzilla.mozilla.org/show_bug.cgi?id=455070 |
| 00:04 | <Hixie> | IE doesn't seem to support accept-charset at all. |
| 00:05 | <Hixie> | webkit doesn't seem to support _charset_ |
| 00:05 | <Hixie> | which makes testing this a pain for webkit |
| 00:06 | <Hixie> | Mozilla supports it, only splits on spaces, and takes the first one |
| 00:06 | <othermaciej> | the charset mime parameter? |
| 00:06 | <annevk> | How is this: "There be dragons. Use UTF-8." |
| 00:06 | <Hixie> | (that it supports) |
| 00:06 | <othermaciej> | in what context? |
| 00:06 | <Hixie> | othermaciej: the magical <input type=hidden name=_charset_> value |
| 00:06 | <othermaciej> | oh |
| 00:07 | <Hixie> | opera supports both spaces and commas and picks the first one |
| 00:07 | <Hixie> | (that it supports, i assume, untested) |
| 00:09 | <Hixie> | hey, the tag has come across the vagueness of the html specs: http://www.w3.org/mid/OF08D6A3D0.C2A54890-ON852574C2.00793300-852574C2.007A5D9B⊙lc |
| 00:12 | <jgraham> | What happened to the guy who was going to redesign the WHATWG blog? |
| 00:12 | <jgraham> | Because the current theme makes me want to poke my eyes out... |
| 00:13 | <annevk> | oh, don't do that |
| 00:13 | <jgraham> | :) |
| 00:15 | <annevk> | here is that post: http://www.brandonfrohs.com/2008/08/whatwg-they-know-html-dont-they/ |
| 00:15 | <annevk> | seems he's been back for a few weeks |
| 00:15 | <annevk> | maybe we should chase him |
| 00:22 | <Hixie> | don't chase him :-) but do feel free to remind him politely :-) |
| 00:22 | <Lachy> | I'll contact him, if no-one else is |
| 00:24 | <Lachy> | I can't find an email or contact form for him, I'll just have to leave a comment |
| 00:28 | <Lachy> | done http://www.brandonfrohs.com/2008/08/whatwg-they-know-html-dont-they/#comment-38 |
| 00:35 | <Lachy> | oh crap, there must be something wrong with the .htaccess on the blog |
| 00:35 | <Lachy> | http://blog.whatwg.org/2022 returns 404 |
| 00:35 | <Dashiva> | You mean it's not 2022 yet? |
| 00:37 | <Lachy> | well, it can't be anything wrong with .htaccess, it must be a bug in the wordpress code |
| 00:38 | <Lachy> | probably because it thinks it's a year, not an article |
| 00:38 | <Lachy> | we'll have to move it |
| 00:39 | <annevk> | use two-thousand-twenty-two |
| 00:39 | <jgraham> | I must remmber this as a reason to make /2022/ and /2022 different |
| 00:39 | <Dashiva> | "Is it 2022 yet?" |
| 00:40 | <Lachy> | any suggestions for a new post slug? |
| 00:40 | <jgraham> | Dashiva: You've been asking that every five minutes since we left |
| 00:40 | <annevk> | Lachy, see above |
| 00:40 | <Dashiva> | See mine too |
| 00:40 | <jgraham> | Party like it's 2022 |
| 00:42 | <Lachy> | http://blog.whatwg.org/two-thousand-twenty-two |
| 00:42 | <Lachy> | I'll set up a redirect for /2022 so we don't break links from RSS feeds |
| 00:42 | <Dashiva> | But Lachy, then the blog won't be consistent in 2022! |
| 00:43 | <Dashiva> | Think of the children (somewhat literally ;) |
| 00:43 | <Lachy> | done |
| 00:44 | <Lachy> | I made it a temporary redirect |
| 00:44 | <Lachy> | so we can remove it by then |
| 00:44 | <annevk> | better make it permanent then |
| 00:44 | <annevk> | 302 means you should keep the old URI around as it might redirect to something else later |
| 00:44 | <annevk> | 301 means the old URI should not be used or cached |
| 00:44 | <Lachy> | in 14 years, it will |
| 00:45 | <annevk> | fair enough, I suppose |
| 00:45 | <Dashiva> | Interesting point, though |
| 00:46 | <Dashiva> | If the URLs current meaning has been permanently redirected, but you would like to put the URL into use again later, with a different meaning.... what redirect do you use? |
| 00:46 | <Lachy> | we should file a bug with wordpress so that they prevent post slugs like that which clash |
| 00:48 | <hdh> | I think 301 is better; in 2022 it will be 200, and there will be no record that it was redirection once |
| 00:49 | <Lachy> | hdh, you're assuming there aren't any tools or caches or whatever out there that keep permanent records of these things |
| 00:49 | <Lachy> | although, it's probably quite impractical to do so if people are careles and don't keep permanent redirects forever |
| 00:50 | <Dashiva> | It's not like people keep records of 410 URLs... |
| 00:51 | <Lachy> | 410 is different though |
| 00:51 | <Dashiva> | It's the same thing, just from the server's perspective |
| 00:51 | <Lachy> | browsers do keep records of 301's until the history is cleared |
| 00:51 | <Lachy> | so if someone never cleared their history, then it would affect them |
| 00:52 | <Lachy> | it's not the same thing. 410 means it's just gone, 301 means it's moved permanently elsewhere |
| 00:53 | <hdh> | ok, I got the cachable part wrong |
| 00:53 | <Dashiva> | Both require one of them to remember "There's nothing at this here URL." |
| 00:54 | <hdh> | so my whole sentence is backward, 302 is correct |
| 00:55 | <Dashiva> | What if we modify the scenario: Redirect for a few years, then 404/410 for a few years, then a new page is added at the same URL. |
| 01:11 | <Hixie> | how do i define form.reset()... hmm... |
| 01:12 | <Dashiva> | Along the line of this? for each input element in form.elements element.value = element.defaultValue |
| 01:17 | <Hixie> | the problem is <select> elements make this different |
| 01:17 | <Hixie> | and input checkbox and radio buttons too |
| 01:17 | <Hixie> | also i don't think it affects buttons |
| 01:17 | <Hixie> | lots of complications |
| 01:17 | <Hixie> | and i have to make sure i don't affect fieldsets |
| 01:17 | <Hixie> | but i do have to affect <output> elements... |
| 01:17 | <Dashiva> | Yeah, I bet there'll be a lot of interesting edge cases to explore :) |
| 01:18 | <Hixie> | i might just define an algorithm for each element and just invoke that for each one |
| 01:18 | <Hixie> | oop for specs |
| 01:18 | <Hixie> | polymorphic english |
| 01:18 | <Hixie> | bbiab |
| 01:19 | <Dashiva> | Hmm... might be able to get away with less |
| 01:21 | <Dashiva> | Oh wait, you have to do the selects recursively to get options, right. Then it gets messy anyway. |
| 06:46 | <GregHouston> | It's just a quick proof of concept, but for anyone interested I created a hack for using canvas/vml in a page viewed with IE8 beta 2 with the document mode set to IE8 Standards. http://greghoustondesign.com/examples/canvas-ie8/ |
| 08:28 | <annevk> | GregHouston, hmm, I thought IE8B2 fixed VML |
| 08:31 | <GregHouston> | With excanvas it works in browser mode IE7 with document mode in quirks mode. With moocanvas it works in all 6 of the 8 modes. VML does not work in the IE8 standards mode at all however. So without a workaround you either get VML or the new CSS features. |
| 08:32 | <GregHouston> | *6 of the 8 modes, not all 6 of the 8 modes |
| 08:34 | <othermaciej> | 8 modes? |
| 08:37 | <Hixie> | wait what? they removed VML? |
| 08:37 | <annevk> | I hope it's a known bug |
| 08:37 | <GregHouston> | Haha, yeah: So you have combinations of browser modes and document modes. Fun stuff. |
| 08:38 | <GregHouston> | Sorry, this is long: |
| 08:38 | <GregHouston> | - Browser Mode: IE7, Document Mode: Quirks Mode |
| 08:38 | <GregHouston> | - Browser Mode: IE7, Document Mode: IE7 Standards Mode |
| 08:38 | <GregHouston> | - Browser Mode: IE8, Document Mode: Quirks Mode |
| 08:38 | <GregHouston> | - Browser Mode: IE8, Document Mode: IE7 Standards Mode |
| 08:38 | <GregHouston> | - Browser Mode: IE8, Document Mode: IE8 Standards Mode |
| 08:38 | <annevk> | Hixie, Option is a NamedConstructor |
| 08:38 | <GregHouston> | - Browser Mode: IE8 Compatibility View, Document Mode: Quirks Mode |
| 08:38 | <GregHouston> | - Browser Mode: IE8 Compatibility View, Document Mode: IE7 Standards Mode |
| 08:38 | <GregHouston> | - Browser Mode: IE8 Compatibility View, Document Mode: IE8 Standards Mode |
| 08:38 | <GregHouston> | No, they closed the ticket on VML the same day IE8 beta 2 came out. It just says: Closed by design. |
| 08:39 | <annevk> | sigh |
| 08:39 | <Hixie> | wow, good times |
| 08:39 | <Hixie> | competing with IE is gonna be easy if they just shoot themselvs in the foot the whole time |
| 08:41 | <hsivonen> | what will devs do? not support IE8? switch to IE 7 mode? use silverlight? Use flash? use Renesis? Use vlad's canvas control? |
| 08:41 | <Hixie> | GregHouston: so what is Browser Mode: IE7? |
| 08:42 | <annevk> | Hixie, https://bugzilla.mozilla.org/show_bug.cgi?id=444222 is relevant for Window.name |
| 08:42 | <Hixie> | hsivonen: VML was never a serious solution for canvas anyway |
| 08:42 | <Hixie> | annevk: please send feedback by e-mail or bug report, i am not in a position to take notes right now |
| 08:43 | <hsivonen> | what's MS's refusal to implement canvas about? Apple's patents before HTML5 is a REC? Silverlight team? something else? |
| 08:44 | <Hixie> | hsivonen: they haven't refused to implement it have they? |
| 08:44 | <hsivonen> | Hixie: oops. s/refusal/failure to prioritize to coincide with VML removal/ |
| 08:45 | <annevk> | I heard Apple patents again yesterday when petele replied to a question regarding <canvas> |
| 08:46 | <GregHouston> | Hixie, truthfully, it is still all rather confusing to me. With the developer bar in IE8 you can see what modes a page defaults to. How the browser mode is set is still unclear to me. You can set a page to quirks supposedly by not using a doctype. You can set it to IE7 Standars with the X-UA-Compatible meta tag. |
| 08:46 | <GregHouston> | With the toolbar you can test pages in all 8 modes to see what works. :D |
| 08:48 | <GregHouston> | Here is the first VML example I get on Google. It defaults to browser mode IE8, and document Quirks Mode. If you try to view it in IE8 Standards the VML dissappears. http://web.syr.edu/~prai/ |
| 08:50 | <hsivonen> | isn't VML more standard than canvas or SVG now that VML is in an ISO standard? |
| 08:53 | <Hixie> | hsivonen: i doubt it is anything but lack of resources |
| 08:58 | <Hixie> | VML is an ISO standard? |
| 08:58 | <Hixie> | i really shouldn't ask these questions |
| 08:58 | <GregHouston> | hsivonen: How do I parse this: s/refusal/failure to prioritize to coincide with VML removal/ |
| 09:01 | <annevk> | Hixie, part of OOXML iirc |
| 09:01 | <hsivonen> | GregHouston: substitute "refusal" with "/failure to prioritize to coincide with VML removal |
| 09:01 | <hsivonen> | " |
| 09:01 | <GregHouston> | Thanks. |
| 09:02 | <annevk> | dfn.js is nice |
| 09:13 | <GregHouston> | Just one last comment on this subject. I created a ticket for canvas in IE8 back in March. It was originally set to "Closed by design". At some point however they changed it to Closed (Postponed). So that is somewhat encouraging. https://connect.microsoft.com/IE/feedback/ViewFeedback.aspx?FeedbackID=334060 |
| 09:17 | <Philip`> | hsivonen: http://www.microsoft.com/windowsxp/expertzone/chats/transcripts/08_0515_ez_ie8.mspx says of <canvas>, "In part because the specification still is in flux, this will not be in the IE8 product." |
| 09:17 | <othermaciej> | what kind of a crazy bug state is Closed (Postponed) |
| 09:18 | <othermaciej> | either the bug is invalid or it isn't |
| 09:18 | <GregHouston> | Haha |
| 09:19 | <Philip`> | GregHouston: That http://web.syr.edu/~prai/ page breaks because it uses non-standard CSS, which IE8 doesn't support since it tries to do CSS properly - see https://connect.microsoft.com/IE/feedback/ViewFeedback.aspx?FeedbackID=333905 |
| 09:20 | <Philip`> | It should still work if you use <?import ...?> instead, as far as I'm aware |
| 09:20 | <hsivonen> | othermaciej: "REMIND and LATER considered harmful" |
| 09:21 | <othermaciej> | hsivonen: I never understood how REMIND or LATER were valid resolutions |
| 09:21 | <othermaciej> | those are maybe milestones, not bug states |
| 09:26 | <GregHouston> | Philip: interesting. |
| 09:30 | <Philip`> | This isn't a bug tracker that Microsoft uses to track bugs and feature requests - it's just telling customers that Microsoft has acknowledged their issue but is not going to fix it in this release |
| 09:30 | <Philip`> | I would assume their internal bug tracker is better at handling not-going-to-fix-it-yet-but-maybe-in-the-future issues |
| 09:33 | hsivonen | wonders if Chrome is going to silently autoupgrade to major releases--not just security patches... |
| 09:35 | Philip` | wonders if there will ever be a major release, or if will always be 0.2 Beta |
| 09:35 | <othermaciej> | I believe their position is they will upgrade you when they feel like it |
| 09:35 | <othermaciej> | you don't get a version, you get a "channel" |
| 09:36 | <othermaciej> | they want to update it "as often as a Web app" which to me sounds a little naiive but maybe they are smarter |
| 09:37 | <hsivonen> | ok. It'll be interesting to see the competitive effect of that kind of "channel" |
| 09:43 | <ronny> | hi |
| 09:45 | <annevk> | hi |
| 09:45 | <ronny> | what are the reasons for inventing an own format for event streams instead of using an xml event stream? |
| 09:45 | <annevk> | simpler |
| 09:47 | <ronny> | annevk: from my point of view its more complex, basically all browsers already got some kind of sax parser, and it allows to just drop strucutred data to the client, while the new thing is some weird obscure text format that needs a new parser |
| 09:48 | <othermaciej> | as a browser engine implementer, I'd rather write a parser for a trivial custom format than to use an XML format |
| 09:49 | <Hixie> | ronny: XML streaming doesn't make sense really (XMPP notwithstanding) |
| 09:50 | <ronny> | Hixie: hu? |
| 09:51 | <ronny> | xmpp is basically just an evetn-stream, i dont see why browsers shouldnt get something semilar |
| 09:51 | <Hixie> | XML doesn't define where a well-formedness error occurs exactly, so it's unclear where a stream should halt if there is a well-formedness error, for instance |
| 09:51 | <Hixie> | XMPP is not a good design, imho |
| 09:52 | <Hixie> | also, XML isn't really appropriate here. What we have is a streaming series of name/value pair dictionaries |
| 09:52 | <Hixie> | XML describes a single tree structure |
| 09:53 | <Hixie> | mapping a streaming series of name/value pair dictionaries onto a tree structure is an unnatural mapping |
| 09:54 | <ronny> | how is that so much different to something like <name>value</name> |
| 09:55 | <Hixie> | "value" can contain other elements, and can contain attributes |
| 09:55 | <Hixie> | er, and "name" can contain attributes |
| 09:55 | <Hixie> | XML describes an annotated tree |
| 09:55 | <ronny> | yeah, and how is that currently relevant? |
| 09:55 | <othermaciej> | ooh yeah I forgot about the streaming aspect |
| 09:56 | <Hixie> | annotated trees are a very different data structure than name/value pair dictionary streams |
| 09:56 | <othermaciej> | streaming xml is nuts |
| 09:57 | <ronny> | why? |
| 09:57 | <ronny> | its basically just a series of annotated events |
| 09:57 | <othermaciej> | because it requires draconian error handling, but you can't tell whether an error occurred for sure until you get to the end of the document |
| 09:57 | <othermaciej> | so streaming it violates a fundamental premise of XML |
| 09:58 | <ronny> | brb, called away |
| 09:58 | <othermaciej> | (unless you send whole separate well-formed documents, but I don't think anyone considers that wise) |
| 09:58 | <othermaciej> | XMPP is a dancing bear trick |
| 10:00 | <Philip`> | It's fun making ejabberd send namespace-ill-formed output to arbitrary clients and causing them to drop their connection |
| 10:10 | <hsivonen> | othermaciej: you can do useful things without ever reaching the end of an XML document |
| 10:11 | <hsivonen> | in practice, streaming XML is tacitly well-defined except for the exact error byte within a tag |
| 10:11 | <othermaciej> | hsivonen: it just seems sloppy to me |
| 10:12 | <othermaciej> | maybe a notion of "well defined fragment" and then sending a series of those would somehow make me feel better |
| 10:12 | <othermaciej> | but ideally you want error recovery so a single ill-formed fragment doesn't poison your connection |
| 10:12 | <othermaciej> | but also XML is a poor format for simple key-value pair data |
| 10:12 | <hsivonen> | sure |
| 10:15 | <Philip`> | XMPP says "An XML stanza exists at the direct child level of the root <stream/> element and is said to be well-balanced if it matches the production [43] content of [XML].", which sounds like that notion |
| 10:15 | <Philip`> | s/XMPP/RFC 3920 (as used by XMPP)/ |
| 10:16 | <Hixie> | if i have a cd in a mac, and it somehow got unmounted, how do i remount it again without ejecting it and reinserting it? |
| 10:16 | <Hixie> | i am ssh'ed into the machine and have remote desktop but i don't want to have to walk over to the living room to reinsert the disk... |
| 10:18 | <Lachy> | Hixie, try Disk Utility |
| 10:19 | <hsivonen> | Hixie: using remote desktop, you should be able to mount it with Disk Utility |
| 10:25 | Hixie | tries |
| 10:25 | <Hixie> | sweet that worked |
| 10:25 | <Hixie> | i wonder what the equivalent command line command would be |
| 10:26 | <Hixie> | thanks, btw |
| 10:35 | <Lachy> | Hixie, it's probably the mount command, but I don't know how to use it |
| 10:35 | <ronny> | re |
| 10:36 | <ronny> | othermaciej: since the stream is supposed to be valid i dont see whats wrong in considering a invalid fragment a critical error |
| 10:37 | <othermaciej> | ronny: it's a needlessly brittle way to design a high-level protocol IMO |
| 10:37 | <ronny> | othermaciej: these days generating invalid xml has only 2 causes - being malicious or being incompetent |
| 10:37 | <Hixie> | hah |
| 10:37 | <othermaciej> | it is silly to have the inefficiency of a text-based protocol and the brittleness of a binary protocol |
| 10:37 | <othermaciej> | ronny: saying that in this channel will earn you derisive laughter |
| 10:37 | <Hixie> | ronny: do you have any software that generates xml? |
| 10:37 | hsivonen | laughs |
| 10:38 | Hixie | prepares the Philip` Gun (tm) |
| 10:38 | <ronny> | Hixie: yeah, im very keen on keeping it valid, i dont want to be as incompetent as the other oart of the 80:20 rule |
| 10:39 | <othermaciej> | ronny: sorry for being blunt, but experience seems to show that even world-class XML experts cannot avoid generating invalid XML with their software |
| 10:39 | <othermaciej> | there has been much discussion of that phenomenon in these circles and it is part of why we think the HTML classic syntax of HTML5, and the well-defined error handling, is important |
| 10:40 | <Hixie> | ronny: as far as i am aware Philip` has never found any XML generating software that can't be tricked into outputting non-wellformed XML, so if you have a candidate for his testing skills, i'm certainly interested in a URL to give him :-) |
| 10:40 | <othermaciej> | it's not just a legacy thing, it is a feature to have a relatively error-tolerant delivery format when dealing with heterogenous producers and consumers |
| 10:40 | <Hixie> | "be tolerant in what you receive", as the saying goes, though tolerant in the very specific way defined in the spec, as i say |
| 10:41 | <othermaciej> | and it seems to me that chat is a case where the same benefits could accrue |
| 10:41 | <ronny> | hmm |
| 10:41 | <othermaciej> | it's just not mission-critical enough for anything that serious people want to get in there and redesign the protocols |
| 10:41 | <ronny> | ok |
| 10:41 | <hsivonen> | Hixie: Philip` hasn't found holes in validator.nu XML output since I got rid of third-party XML serialization code |
| 10:42 | <Hixie> | hsivonen: it hardly counts if you consider software that has reaped the benefits of his testing already |
| 10:43 | <ronny> | i yet have to dig something up that requires disasterous testing |
| 10:44 | <othermaciej> | ronny: anyway, I hope that even if you still disagree, that you can at least see that the custom event stream protocol was not random or thoughtless but actually based on giving the issues a fair amount of consideration |
| 10:44 | <ronny> | well, i dont have to like it |
| 10:44 | <Hixie> | the event stream protocol was in fact designed to make it impossible to output invalid streams |
| 10:45 | <Hixie> | (at the syntax level) |
| 10:45 | <othermaciej> | I am merely asking you to understand at this point, not to like :-) |
| 10:45 | <ronny> | bit given your input it seems much more nice |
| 10:45 | <Hixie> | i believe there isn't a single byte combination that can't be unambiguously parsed without error |
| 10:45 | <Hixie> | in that format |
| 10:45 | <ronny> | hmm |
| 10:45 | <Hixie> | so in a way it does have draconian error handling like xml... there just aren't any possible errors :-) |
| 10:46 | <othermaciej> | gotta quit for some testing, I'll be back |
| 10:46 | <othermaciej> | Hixie: that does sort of finesse the issue |
| 10:46 | <hsivonen> | Hixie: in a way, yyeah, ut the new code is independent of the Apache/Lotus-originating code |
| 10:46 | <hsivonen> | but |
| 10:46 | <ronny> | btw, i think its wrong to give programmers any silent error-tolerant savety net for their own errors |
| 10:47 | <ronny> | they arent users, they should know better, and they should be punished for not knowing |
| 10:47 | <Hixie> | ronny: i agree that programmers shouldn't get their errors ignored, but the users shouldn't be exposed to the errors if errors slip through |
| 10:47 | <ronny> | Hixie: well, will browsers have a way to disable html5 error recovery? |
| 10:48 | <ronny> | (i mean a well defined way that allows to dump my own stupid erros just into my face) |
| 10:48 | <Hixie> | most already do, just open the error console in e.g. firefox, safari, chrome, or opera |
| 10:50 | <Hixie> | i suppose browsers could have a way to just abort as soon as an error is output on the error console, but that would make debugging hard since you couldn't do anything until you'd fixed all the errors before the one you were debugging |
| 10:51 | <annevk> | CSS extensions would make that really annoying |
| 10:51 | hsivonen | wonders what dumping HTML5 parse errors to console would do to Firefox perf |
| 10:51 | <hsivonen> | or browser perf in general |
| 10:53 | <ronny> | browsers should indicate how incappable the creator of the site is (ie show how stupid he is at making valid html5 that doesnt need any recovery) |
| 10:53 | <jgraham> | indicate to whom? |
| 10:53 | <jgraham> | And why? |
| 10:54 | <annevk> | someone requested a Web Workers diff, lots of people have requested html5 diffs meanwhile |
| 10:54 | <ronny> | well, people would care to fix if the browsers shows every lillle stupid thing they did |
| 10:54 | <ronny> | after all it is bad for reputation |
| 10:54 | <annevk> | based on the little amount of != 10 context diffs I suspect hsivonen to use context 50 |
| 10:56 | <hsivonen> | annevk: IIRC, I've used 200 |
| 10:56 | <annevk> | ronny, it would just be annoying for end users and they would switch to some browser that was simpler |
| 10:56 | <jgraham> | You mean like an icon that was end-user visible? Since that would show up on >90% of web sites people wouldn't really associate it with badness in the site |
| 10:56 | <annevk> | hsivonen, really? that's not in my cache list, hmm |
| 10:56 | <hsivonen> | annevk: I haven't diffed in a recent days, though |
| 10:56 | <Hixie> | ronny: based on studies we've done looking at literally billions of pages on the Web, somewhere in the region of more than 90% of Web sites have at least one error, so users would just see the indicator on most of hte Web and it would lose meaning |
| 10:57 | <annevk> | hsivonen, ok |
| 10:58 | <ronny> | Hixie: well, one error isnt really stupid, it should be an icon that indicates the error rate, ie the more errors, the more incappable is the site-creator |
| 10:58 | <annevk> | seems rather rude to me |
| 10:59 | <Hixie> | ronny: there are hundreds of errors on common sites like cnn.com, i don't think most users care and i doubt cnn.com cares whether a browser says they're any good or not |
| 10:59 | <ronny> | annevk: thats the point, 1 error is most likely just overlooking something, but when it goes into the hundreds its bah |
| 10:59 | <Hixie> | ronny: but feel free to ask browser vendors to do it :-) (i hear iCab does it, you might want to use that) |
| 11:00 | <annevk> | ronny, doesn't mean we should bother end users with silly icons |
| 11:00 | <ronny> | annevk: well, how else would one spread awareness? |
| 11:01 | <jgraham> | ronny: If the browser displays "500 errors" (or whatever) on cnn.com it makes the browser, nt the site look worse |
| 11:01 | <takkaria> | someone needs to tell John Foliot to stop the +1 posts |
| 11:01 | <annevk> | ronny, education, advocatism, Web Standards Project, etc. |
| 11:01 | <jgraham> | ronny: Why do you care about spreading awareness? |
| 11:01 | <othermaciej> | I think browser developer tools should include built-in support for validation |
| 11:01 | <jgraham> | takkaria: +1 |
| 11:02 | <ronny> | annevk: that doesnt make normal users aware |
| 11:02 | <othermaciej> | I think it would be cool if Safari's Web Inspector let you validate your CSS and HTML |
| 11:02 | <annevk> | ronny, my mom will never care |
| 11:02 | <othermaciej> | maybe even lint your JavaScript using someone's tool |
| 11:02 | <GregHouston> | The Opera error console already shows errors for html, as well as an assortment of other things, xml, css, javascript, svg. Plus you can filter them. |
| 11:02 | <othermaciej> | normal users don't care and shouldn't have to |
| 11:02 | <annevk> | ronny, normal users shouldn't be bothered with mistakes of others, would just annoy them |
| 11:03 | <othermaciej> | imagine if your television pointed out every continuity error or instance of poor writing in the TV shows you watch |
| 11:03 | <othermaciej> | that would be pretty annoying |
| 11:03 | <othermaciej> | most of us here are Web technology geeks and care about things like content conformance and what the errors are |
| 11:04 | <othermaciej> | but normal people just don't care |
| 11:04 | <othermaciej> | any more than we care to hear about awkward camera angles or bad lighting when watching tv |
| 11:04 | <othermaciej> | or poor quality paper or ink when reading a magazine |
| 11:04 | hsivonen | points to the bookmarklets on http://about.validator.nu |
| 11:05 | <othermaciej> | most people use the Web for the content, not the technical details of the medium |
| 11:05 | <othermaciej> | and we have to be humble enough to realize that |
| 11:06 | <othermaciej> | to draw a more technical analogy, what if your iPhone or Mac or PC or whatever told you about every compiler warning each time you launch an app |
| 11:07 | <othermaciej> | when I run Firefox I don't really care that "warning: enumeration value ‘LIR64’ not handled in switch" |
| 11:07 | <othermaciej> | I would guess you do not either |
| 11:07 | <GregHouston> | If I had to click through error messages for every fault in my home I would never make it through the front door. |
| 11:08 | <othermaciej> | even I as a well versed C++ hacker would have no idea what to do with the knowledge that on line 1050, "warning: pointer targets in passing argument 2 of ‘strcmp’ differ in signedness" |
| 11:09 | <othermaciej> | (these are real error messages from building the JS parts of Gecko, btw, and I am sure they are useful to the developers of Gecko, but would not be to a Firefox user) |
| 11:09 | <othermaciej> | (but to most normal people html validation errors would sound about as comprehensible as those C++ warnings) |
| 11:09 | <othermaciej> | sorry for ranting |
| 11:09 | <Hixie> | it would be more like your TV reporting every teletext formatting error, every cable programme guide data syntax error, every tuning error |
| 11:09 | <othermaciej> | it's just such a pet peeve of mine |
| 11:10 | <roc> | every time the browser detects a conformance error, the site operator is taxed 1c |
| 11:10 | <othermaciej> | oh man if I could get a penny for every warning my compiler finds while building mozilla... |
| 11:10 | <jgraham> | ronny: As you may have noticed you are not the first person to suggest such a thing :) |
| 11:11 | <ronny> | yeah |
| 11:11 | <ronny> | well, it should be irelevant what errors there are |
| 11:11 | <ronny> | just tell the user the people making the site are incappable of doing it right |
| 11:11 | <othermaciej> | would you like to be told what native apps on your OS of choice had compiler warnings every time you launch? |
| 11:11 | <Lachy> | http://www.brandonfrohs.com/2008/08/whatwg-they-know-html-dont-they/#comment-39 |
| 11:12 | <othermaciej> | "warning, Adium was written by silly people who left in 52 compiler warnings!" |
| 11:12 | <othermaciej> | (no offense to any Adium developers present and I am sure it is in fact a wonderful piece of software) |
| 11:12 | <ronny> | othermaciej: that would be nice, cause im pretty sure that would get fixed |
| 11:12 | <Lachy> | Adium certainly is buggy |
| 11:13 | <othermaciej> | it would get whatever OS is doing it uninstalled |
| 11:13 | <othermaciej> | users hate useless warnings and error messages and such |
| 11:13 | <othermaciej> | for a lot of people, it just makes them fearful of using the computer |
| 11:14 | <othermaciej> | you should never display an error message to the user when you have successfully recovered and there is no choice the user needs to make |
| 11:14 | <othermaciej> | that's human interface design 101 |
| 11:14 | <ronny> | othermaciej: uh, wouldnt that got vista uninstalled like AGES ago? |
| 11:14 | <Hixie> | vista isn't exactly having a roaring success |
| 11:14 | <Lachy> | as a real example, I really hate how VLC reports error messages almost every time I watch a DVD with it |
| 11:14 | <ronny> | well, its stil around |
| 11:14 | <Hixie> | it may in fact be the best example of othermaciej's point yet raised in this discussion |
| 11:14 | <jgraham> | Vista displays all he compile time warnings on startup? |
| 11:15 | <jgraham> | I guess the security stuff? |
| 11:15 | <Lachy> | every time, the errors are meaningless to me, yet there's no way I've found to disable them completely |
| 11:15 | <ronny> | hmm, if the error notification is modal, there is something wrong |
| 11:15 | <othermaciej> | ronny: I hate to break it to you, but: <http://blog.seattlepi.nwsource.com/microsoft/archives/130626.asp>; |
| 11:16 | <Hixie> | <form> is nearly done! |
| 11:16 | <othermaciej> | getting vista uninstalled is a service many people want to pay for |
| 11:16 | jgraham | actually has no idea what warnings vista does display having almost entirely avoided it |
| 11:16 | <Hixie> | right, bed time |
| 11:16 | <Hixie> | nn |
| 11:17 | <ronny> | othermaciej: well, if those messages are modal dialogs, its done wrong from the beginnings |
| 11:18 | <ronny> | it should be more like the "oh, i blocked populs" notifications of firefox that appears, then disappears without much hassle and without blocking anything |
| 11:19 | <jgraham> | ronny: Most warnings may as well read "Two Jabberwockys have been froobed". There is no point in displaying that ever |
| 11:21 | <ronny> | k |
| 13:16 | <hsivonen> | Hixie: I take it that the XML submission format got killed? |
| 13:37 | <annevk> | http://search.twitter.com/search?q=%40whatwg |
| 14:15 | gsnedders | smacks .ac.uk sites for needing www. |
| 14:20 | <jgraham> | gsnedders: That annoys me _so_ much |
| 14:21 | <jgraham> | cam.ac.uk has/had a policy that all public facing webserver subdomains need to start with www leading to craziness like www-xray.ast.cam.ac.uk |
| 14:21 | gsnedders | facedesks |
| 14:23 | hsivonen | finds http://openandclosed.org/topics/fileformats/ |
| 14:26 | <hsivonen> | aaronlev: Do you know what Joe Clark's non-profit is up to with file formats? |
| 14:27 | <aaronlev> | no |
| 14:27 | <webben> | hsivonen: JC has some proposals as to what the file format should do. |
| 14:27 | <webben> | I don't think he's gotten the funds to actually spec it though. |
| 14:29 | <webben> | oh, I think that page used to have more details |
| 14:29 | <webben> | ah ha |
| 14:29 | <webben> | hsivonen: http://openandclosed.org/docs/AccessibilityExchange.html |
| 14:29 | <hsivonen> | webben: tanks |
| 14:58 | <jgraham> | Argh. I have exactly one bibtex entry amongst several hundred that is not working and I have no idea why |
| 15:30 | <annevk> | zcorpan, http://www.w3.org/2008/09/10-xhtml-minutes.html#item06 |
| 15:31 | <annevk> | http://www.windowsfordevices.com/news/NS6599416217.html o_O |
| 15:34 | <hsivonen> | annevk: interesting how the first lines are about 'obligation' and 'simply thanking him' |
| 15:35 | <takkaria> | looks like they didn't really want feedback at all |
| 15:35 | <annevk> | yeah |
| 15:40 | <takkaria> | an interesting link from earlier up in the minutes: |
| 15:40 | <takkaria> | http://www.dev-archive.net/articles/xhtml.html#recommendations |
| 15:40 | <hsivonen> | "even had to disabuse TBL of XHTML as failure canard" |
| 15:40 | <takkaria> | someone in the XHTML WG is working on that, and seems to basically recommend not serving XHTML |
| 15:41 | <annevk> | hehe, that presenter was me I think, though I believe I was clear on the text/html thing |
| 15:42 | <takkaria> | unless you want to use MathML or Ruby, in which case you should use XHTML |
| 15:43 | <annevk> | well, not anymore :) |
| 15:43 | <takkaria> | quite |
| 15:44 | <takkaria> | I wish the XHTML people would stop pretending XHTML served as text/html is XHTML |
| 15:46 | <hsivonen> | takkaria: at least the document has an HTML 4.01 doctype |
| 15:46 | <hsivonen> | and <meta name="keywords" content="$keywords"> |
| 15:47 | <takkaria> | good times |
| 15:49 | <hsivonen> | also, talks about "XHTML WG" |
| 15:52 | <annevk> | I'm somewhat displeased by the HTML WG being referred to as HTML5 WG and the XHTML2 WG as XHTML WG |
| 16:02 | <hsivonen> | otherwise, the content of the article is refreshingly accurate |
| 16:04 | <hsivonen> | http://lists.w3.org/Archives/Public/public-rdf-in-xhtml-tf/2008Aug/0137.html |
| 16:12 | <takkaria> | all the words these RDF people use to talk about things confuse me |
| 16:22 | <hsivonen> | "an obvious next step will be to define HTML+RDFa." -- http://lists.w3.org/Archives/Public/public-rdf-in-xhtml-tf/2008Sep/0028.html |
| 16:25 | <deane> | hsivonen: yeah, I saw that. As if they weren't already stepping over the mark. HTML+RDFa is our domain, surely. Not that anyone's interested in it. |
| 16:26 | <takkaria> | seems a bit lame to say "we're not using CURIEs since the spec isn't ready, we'll just define it ourselves" |
| 16:27 | annevk | -> away |
| 17:56 | <hsivonen> | where's the implementation report for XHTML 1.1? |
| 18:00 | <webben> | hsivonen: http://www.w3.org/MarkUp/2008/xhtml-m12n-11-implementation.html ? |
| 18:03 | <hsivonen> | webben: that's for Modularization--not for the XHTML 1.1 language |
| 18:03 | <hsivonen> | webben: that report doesn't cover implementations of the language in agents |
| 18:03 | <webben> | I see. |
| 18:04 | <hsivonen> | webben: it just says that their schemas worked with three validator engines |
| 18:04 | <webben> | Does someone say there is an implementation report? |
| 18:04 | <hsivonen> | webben: It's a REC and was edited recently |
| 18:05 | <webben> | I'm confused, how can an implementation report be a REC? |
| 18:05 | <hsivonen> | webben: XHTML 1.1 is a REC |
| 18:05 | <webben> | oh right |
| 18:06 | <hsivonen> | (unless new editions of old RECs are exempt of the implementation report requirement) |
| 18:07 | webben | doesn't know what requirement that is. I probably dunno enough of the context of what you're asking for to be helpful, so I'll just shut up it. :) |
| 18:44 | <Hixie> | "XHTML 1.1 doesn't need an implementation report, it's basically just HTML4 with XML, and both of those are in REC" |
| 18:46 | <Dashiva> | What could go wrong? |
| 18:47 | <webben> | heh |
| 18:48 | lbruno | says hi |
| 18:58 | <hsivonen> | Hixie: is that an actual quote? |
| 18:59 | hsivonen | can't detect sarcasm on this topic |
| 19:11 | <Lachy> | hsivonen, in your html5 live dom viewer, I can't figure out what the last parameter of window.parseHtmlDocument() is supposed to do: |
| 19:11 | <Lachy> | window.parseHtmlDocument(textarea.value, iframe.contentWindow.document, afterParse, null); |
| 19:12 | <Lachy> | it doesn't look like it's actually used for anything in the function instelf, it just looks like an unused parameter |
| 19:12 | <hsivonen> | Lachy: it's an unimplemented error reporting callback |
| 19:13 | <Lachy> | ok |
| 19:13 | <hsivonen> | Lachy: yes, it is unused |
| 19:15 | <Lachy> | does this whole function work by parsing the textarea, creating all the nodes using the DOM APIs, and then inserting them into iframe.contentWindow.document? |
| 19:15 | <Lachy> | effectively replacing the document.write() call in Hixie's dom viewer? |
| 19:17 | <Lachy> | I'm going to try and get both HTML5 and browser parsing integrated into my HTML5 Tools beta, side by side, with a UI to switch between them |
| 19:18 | <hsivonen> | Lachy: yes, and after it si done, it calls afterParse() |
| 20:26 | <gsnedders> | What the derivative of \sin^2 3x? |
| 20:26 | gsnedders | is sure he's doing something stupid |
| 20:28 | <gsnedders> | Hmm. Calculator online says I'm write and answer in textbook is wrong. |
| 20:29 | <Philip`> | I bet it's not 6 sin^2 3x cos 3x |
| 20:30 | <gsnedders> | Philip`: So do I :) |
| 20:30 | <gsnedders> | Philip`: I think 6\sin 3x\cos 3x |
| 20:31 | <gsnedders> | Philip`: Textbook thinks 6\sin 3x |
| 20:31 | <Philip`> | Maxima agrees with you |
| 20:31 | <gsnedders> | I learnt not to trust textbook given answers I while ago :) |
| 20:31 | <gsnedders> | *a |
| 20:31 | <Philip`> | Bah, I've forgotten how the chain rule works :-( |
| 20:31 | <gsnedders> | Philip`: n00b. |
| 20:32 | <lbruno> | I should know this too. |
| 20:32 | <gsnedders> | Times by the power, the bracket with the power reduced by one, and the derivative of the content of the bracket |
| 20:32 | lbruno | is just a random passer by |
| 20:32 | <gsnedders> | s/Times by/Times the/ |
| 20:33 | <gsnedders> | s/the the/the. |
| 20:33 | <gsnedders> | s/the the/the/ |
| 20:33 | <gsnedders> | Hmm, that's not very good typing |
| 20:33 | <lbruno> | gsnedders: that's true! I was deriving to cos, as in (sin 3x)^2. |
| 20:33 | <lbruno> | dumb mistaek |
| 20:34 | <gsnedders> | lbruno: It is (sin 3x)^2 |
| 20:34 | <gsnedders> | lbruno: It's just you need to keep (sin 3x)^1 |
| 20:34 | <Lachy> | I hated finding derivatives, it never made any sense to me |
| 20:35 | gsnedders | concludes this isn't a good place to ask about maths |
| 20:35 | <lbruno> | 2 * sin(3x) * cos(3x)? |
| 20:35 | <gsnedders> | ianloic: No, 3cos3x |
| 20:35 | <gsnedders> | lbruno: sin nx derives to ncos nx |
| 20:36 | Philip` | works out where he got stupidly confused |
| 20:37 | <lbruno> | butbut... I failed Calculus I and II, don't listen to me anyway |
| 20:37 | Philip` | uses the excuse that he's only exposed to discrete maths nowadays |
| 20:37 | <lbruno> | now I see where the 6* came from |
| 20:38 | <Philip`> | (and if I really need to differentiate, I can use Maxima :-) ) |
| 20:38 | <gsnedders> | (This is homework due for last Thursday, BTW) |
| 20:38 | <gsnedders> | :D |
| 20:38 | lbruno | appreciates the refresher |
| 20:38 | <gsnedders> | (That was one of the simple ones) |
| 20:38 | <gsnedders> | Most of the hard things I can do :P |
| 20:38 | <Philip`> | (although that is admittedly slower and more error-prone than calculating it by hand) |
| 20:39 | gsnedders | doesn't claim to know anything about A-Level maths, on grounds he isn't doing it |
| 20:39 | <gsnedders> | so I have no idea what Philip` (once) knew |
| 20:40 | <lbruno> | I can only give my warmest encouragement to those who have to know this |
| 20:40 | Philip` | did that and Further Maths and another three modules just for fun, but has forgotten most of it |
| 20:40 | <lbruno> | I wonder what A-level means |
| 20:40 | lbruno | is from .pt |
| 20:41 | <lbruno> | s,vel ,vel maths, |
| 20:41 | <gsnedders> | .pt? |
| 20:42 | <lbruno> | Portugal |
| 20:42 | <gsnedders> | ah |
| 20:42 | <Philip`> | lbruno: They're the exams (and associated courses) for age ~16-18 (just before university) in sensible parts of the UK |
| 20:42 | <jgraham> | gsnedders: Now integrate sin^2 3x :) |
| 20:42 | <gsnedders> | jgraham: No :P |
| 20:43 | <gsnedders> | jgraham: I always forget how to integrate. |
| 20:43 | <jgraham> | gsnedders: It's much more fun than differentiating it :) |
| 20:43 | <jgraham> | gsnedders: The problem with integration is that in general nothing works. There are a few special cases where you can work things out |
| 20:44 | <gsnedders> | :P |
| 20:44 | <Philip`> | (3x-(sin 6x)/2)/2 - easy! |
| 20:44 | <Philip`> | Um |
| 20:44 | <gsnedders> | Philip`: Nope |
| 20:44 | <Philip`> | (3x-(sin 6x)/2)/6 - easy! |
| 20:44 | <gsnedders> | Philip`: wrong |
| 20:44 | <Philip`> | That's what Maxima says :-) |
| 20:44 | <gsnedders> | Philip`: + c :) |
| 20:44 | <Philip`> | That's just pointless pedantry :-p |
| 20:45 | <lbruno> | Philip`: education here lacks good sense too (generally) |
| 20:45 | <gsnedders> | Philip`: No, it's a requirement in the marking scheme :P |
| 20:45 | <lbruno> | gsnedders: same here. +C is "a very good habit" |
| 20:46 | <lbruno> | but then we all look the other way and proceed to ignore it, as it's too much work |
| 20:46 | <gsnedders> | Philip`: http://xkcd.com/385/ |
| 20:46 | <hsivonen> | Philip`: how does Maxima compare to Mathematica? |
| 20:46 | <Philip`> | (Also, I forgot about the c) |
| 20:46 | <Philip`> | hsivonen: All I know is it's much cheaper |
| 20:47 | <hsivonen> | :-) |
| 20:47 | <gsnedders> | I've only done one more subquestion since this started! |
| 20:47 | jgraham | has a hate/hate relationship with mathematica |
| 20:47 | <gsnedders> | jgraham: So a 1 relationship with mathematica? |
| 20:47 | <hsivonen> | Mathematica has rather dramatic price discrimination going on |
| 20:48 | <jgraham> | Every time I ever tried to use it, it was because the problem was too hard. Firing up Mathematica only gave me two problems |
| 20:48 | <Philip`> | gsnedders: Why do you assume hate is non-zero? |
| 20:48 | <hsivonen> | As a freshman, I was able to afford a student license of mathematica |
| 20:48 | <gsnedders> | Philip`: Because hate is something. |
| 20:48 | <gsnedders> | Philip`: It has to not be zero |
| 20:49 | <Philip`> | gsnedders: Maybe it's just an arbitrarily-ordered enumeration of emotions, and hate happens to be assigned value 0 |
| 20:50 | <Philip`> | Hixie: Please note that I attempt to maintain "never found any XML generating software that can't be tricked into outputting non-wellformed XML" status by ignoring any software that can't be tricked, and claiming that's because it's too simple (e.g. doesn't accept enough user input) or just not saying anything and pretending I never saw it |
| 20:51 | <gsnedders> | Philip`: :D |
| 20:51 | <gsnedders> | I maintain it by not trying, and assuming it is broken |
| 20:54 | <Philip`> | (Also I often cheat and just find general XSS holes, which can be used to make XML ill-formed but are equally problematic in non-XML pages) |
| 20:54 | <Philip`> | ((at least in the context of XHTML-outputting web pages)) |
| 20:54 | <hsivonen> | Philip`: have you tried the new Validator.nu XML serializer yet? |
| 20:56 | <Philip`> | hsivonen: I haven't |
| 20:57 | <gsnedders> | Hmm. I'm stuck. Again. |
| 20:57 | <gsnedders> | I won't try asking here. |
| 20:57 | <gsnedders> | :P |
| 20:57 | <jgraham> | Hey |
| 20:57 | <jgraham> | I remember A-level-level maths |
| 20:59 | <lbruno> | gsnedders: do try :) |
| 20:59 | <Philip`> | hsivonen: (Is it meant to be unbreakable by an API user, rather than just an indirect web user?) |
| 21:00 | <gsnedders> | It's a question of how you prove the derivative of e^{3x} |
| 21:00 | <gsnedders> | (actually, it isn't, but I know what the derivative is, and I want to understand it :P) |
| 21:00 | <GregHouston> | Philip: Thanks for pointing out the namespace change with VML last night. It doesn't see to work if the VML is appended to the page though, and thus not with Excanvas or Moocanvas. |
| 21:01 | <hsivonen> | Philip`: it's not meant to be unbreakable by the seralizer Java API user |
| 21:01 | <hsivonen> | Philip`: it is meant to be unbreakable by the Validator.nu Web service API user |
| 21:02 | <hsivonen> | Philip`: the Java API leaves NCNameness and correct nesting to the application for perf reasons |
| 21:02 | <hsivonen> | but it does sanitize character data and attribute values |
| 21:03 | <hsivonen> | and it generates ns declaration for the crazy cases |
| 21:03 | <lbruno> | gsnedders: How (or why) d(e^{ex})/dx == 3 * e^{3x}, IIRC? |
| 21:04 | <gsnedders> | jgraham: d(e^{3x})/dx == 3 * e^{3x} |
| 21:04 | <gsnedders> | (not e^{ex}) |
| 21:05 | <jgraham> | gsnedders: Yeah. What's the question? How do you prove it? |
| 21:05 | <gsnedders> | jgraham: Yeah (it's my question, the question in the exercise is just what it is) |
| 21:05 | <hsivonen> | Philip`: and cutting the output stream by provoking Out Of Memory in a stream write operation does not count |
| 21:06 | <jgraham> | df/dx = lim dx->0 (f(x+dx) - f(x))/dx |
| 21:07 | <jgraham> | (exp(a(x+dx)) - exp(ax))/dx = exp(ax) * (exp(a dx) - 1) / dx |
| 21:07 | <gsnedders> | Ya |
| 21:08 | <jgraham> | = exp(ax)((1+a*dx) - 1)/dx in lim dx->0 |
| 21:08 | <jgraham> | = a*exp(ax) |
| 21:08 | <jgraham> | But that's kind of cheating because you need to use the taylor series approximation for exp which is based on the derivative |
| 21:09 | <Philip`> | GregHouston: Have you tested whether it works if you use document.namespaces.add(prefix, uri, behaviour) to bind the behaviour? |
| 21:09 | <gsnedders> | jgraham: I can't really read maths written in text. Meh. :P |
| 21:10 | <Philip`> | GregHouston: Actually, you probably mean it doesn't work if the VML content itself is written dynamically, in which case that probably won't affect anything |
| 21:10 | <GregHouston> | Yeah, that is what I meant. |
| 21:11 | jgraham | guesses you could define exp by its taylor series approximation and so circumvent the problem |
| 21:12 | <gsnedders> | jgraham: I don't quite see where the final = a*exp(ax) comes from |
| 21:12 | gsnedders | is probably being silly |
| 21:12 | lbruno | wrote e^{ex} by mistaek. sorry |
| 21:13 | <GregHouston> | Philip: in the ticket it says that tagUrn is not set when adding the VML dynamically. I don't know what tagUrn is. Do you know if it is something that can be manually set? |
| 21:13 | <jgraham> | gsnedders: (1+a dx)-1 = a dx ; a dx/dx = a |
| 21:13 | <jgraham> | gsnedders: It turns out that mathematicians are quite anal http://planetmath.org/encyclopedia/DerivativeOfExponentialFunction.html |
| 21:14 | <gsnedders> | jgraham: Is that unexpected? |
| 21:14 | <gsnedders> | :) |
| 21:14 | <jgraham> | gsnedders: Not at all :) |
| 21:15 | jgraham | should go home |
| 21:15 | <lbruno> | I think I prefer the Wikipedia version. You are not asking why d(e^x)/dx == e^x |
| 21:15 | <jgraham> | Wikipedia version? |
| 21:16 | <lbruno> | the chain rule |
| 21:17 | <lbruno> | i'm not making any sense |
| 21:17 | <gsnedders> | lbruno: Link? |
| 21:17 | <lbruno> | just before this section: http://en.wikipedia.org/wiki/Exponential_function#Formal_definition |
| 21:18 | <lbruno> | I think this is what you're asking about; if not... flames gladly accepted :) |
| 21:19 | <lbruno> | s/just before/just above/ |
| 21:19 | <gsnedders> | lbruno: just before is fine |
| 21:19 | <gsnedders> | (so is above) |
| 21:20 | <lbruno> | thanks |
| 21:20 | <Philip`> | GregHouston: I don't know, but I would assume it can't be set, in the same way that tagName can't be set |
| 21:21 | lbruno | is always doubting if the idioms he uses are correct |
| 21:21 | lbruno | finds writing about himself in the 3rd person quite peculiar |
| 21:22 | <jgraham> | lbruno: Using the chain rule is cheating unless you prove the chain rule :) |
| 21:22 | jgraham | has no idea how to do that |
| 21:23 | <lbruno> | jgraham: true, that was on my mind as well! |
| 21:23 | <lbruno> | but you've got to believe in something |
| 21:24 | <lbruno> | I mean, in some axiom |
| 21:24 | <lbruno> | this one looks useful |
| 21:25 | <gsnedders> | I am now sure I'm really being stupid |
| 21:25 | Philip` | would assume you prove the chain rule by calculating limits, but you probably don't want to prove that calculating limits is valid |
| 21:26 | <gsnedders> | Can \frac{2x^2+6x}{4x^2+12x-9} be simplified? |
| 21:33 | <Philip`> | Probably not, but you should factorise it first |
| 21:34 | <gsnedders> | How? |
| 21:34 | <lbruno> | by finding zeros, I think, and expressing it as (x-<zero1>)(x-<zero2>). |
| 21:35 | <lbruno> | .oO(is that what factorise meant?) |
| 21:37 | <Philip`> | lbruno: That's what I mean |
| 21:38 | <lbruno> | I'm having some seriously silly difficulty to use the Magic Formula on the bottom part of the fraction. |
| 21:49 | <lbruno> | this is silly, I be able to do this from muscle memory by now. I sure flunked calculus enough times |
| 21:49 | <lbruno> | I cheated |
| 21:50 | <lbruno> | I found an online graphing calculator, plotted both parts of the fraction, and saw they didn't share common zeros. |
| 21:51 | <lbruno> | so I think the fraction can't be simplified that way. |
| 21:51 | <hsivonen> | https://bugzilla.mozilla.org/show_bug.cgi?id=436083 |
| 21:55 | lbruno | guesses it's enough OT for now |
| 21:59 | <gsnedders> | lbruno: you want to be on-topic? why? |
| 22:00 | <lbruno> | just that I find a bit peculiar thinking about calculus on #whatwg |
| 22:00 | <lbruno> | not complaining, though |
| 22:00 | <lbruno> | I sure need the exercise |
| 22:00 | <gsnedders> | We discuss everything here :P |
| 22:00 | <lbruno> | mathematical and physical :-) |
| 22:00 | <gsnedders> | Oh, and other things |
| 22:01 | <lbruno> | oy, it's IRC |
| 22:01 | gsnedders | yawns |
| 22:01 | gsnedders | is too tired to do this |
| 22:01 | <lbruno> | OT is bound to happen anyway |
| 22:01 | <lbruno> | it /is/ getting rather late to do math calisthenics |
| 22:02 | <lbruno> | gsnedders: why the interest in math? |
| 22:02 | <gsnedders> | lbruno: I have homework to do :P |
| 22:03 | <gsnedders> | lbruno: For last Thursday, no less |
| 22:04 | <lbruno> | can't say I envy you. homework always seemed to deflate my interest in math |
| 22:05 | <gsnedders> | lbruno: Love does not envy. |
| 22:05 | <lbruno> | ? |
| 22:05 | gsnedders | shrugs |
| 22:05 | lbruno | felt a bit of a draft overhead |
| 22:05 | gsnedders | is feeling vaguely Biblical |
| 22:06 | <gsnedders> | (1 Corinthians 13) |
| 22:06 | <lbruno> | ah. |
| 22:06 | <lbruno> | never did get to that part of the bible |
| 22:06 | <gsnedders> | Somewhere between 4 and 7, IIRC |
| 22:06 | gsnedders | has just read random bits |
| 22:07 | <gsnedders> | Quite a lot of random bits, but compared with the size of the Bible, next to nothing |
| 22:07 | <lbruno> | true, you can spend a lot of time around that book and read just an insignificant part |