| 00:33 | <Hixie> | sweet jesus i hate bidi |
| 00:34 | <Hixie> | what happens if you have "english HEBREW HEBREW HEBREW english" and the HEBREW part is so long it splits into three lines? |
| 00:34 | <Hixie> | in ltr |
| 00:34 | <Hixie> | you get "e1 1h \n 2h \n 3h e2" right? |
| 00:35 | <TabAtkins> | Presumably? |
| 00:35 | <TabAtkins> | Which is weird, of coruse. |
| 00:35 | <TabAtkins> | Or, wait, is that logical or visual string order that you specified? |
| 00:37 | <TabAtkins> | If the logical text was "e1 H1 H2 H3 e2", presumably it would be displayed as "e1 H3 \n H2 \n H1 e2". |
| 00:38 | <TabAtkins> | Since linebreaking should be more-or-less a visual effect where you just break lineboxes, and if the text fit on one line it would display as "e1 H3 H2 H1 e2". |
| 00:41 | <Hixie> | that would be crazy, surely |
| 00:41 | <Hixie> | you never read up |
| 00:41 | <TabAtkins> | Which part is crazy? The one-line layout, or how it's broken into lineboxes? |
| 00:41 | <Hixie> | the part where you have to start reading by skipping a line and then reading up |
| 00:41 | <TabAtkins> | Man, I dunno. You shouldn't mix text directions in a single inline layout anyway. |
| 00:42 | <Hixie> | you don't have much choice when writing hebrew |
| 00:42 | <Hixie> | e.g. THE POPE SAID "whatever the pope said" WHICH IS SILLY |
| 00:42 | <Hixie> | if the quote is long, it'll split across many lines |
| 00:42 | <Hixie> | but you want the start of the quote to be on the first line |
| 00:43 | <TabAtkins> | Solution: stop being rtl, or switch everyone to rtl. |
| 00:43 | <Hixie> | good luck with that |
| 00:53 | <Hixie> | i'm clearly wrong somehow |
| 00:54 | <zcorpan> | hmm. http://html5.org/tools/web-apps-tracker?from=6007&to=6008 - i think several of those are not good indicators |
| 00:54 | <zcorpan> | lots of layout tables with <th> for instance |
| 00:55 | <TabAtkins> | ...why??? |
| 00:55 | <zcorpan> | this is the web? |
| 00:55 | <tw2113> | sparta? |
| 00:55 | <TabAtkins> | I hate people. |
| 00:55 | <zcorpan> | also lots of non-layout tables with cellspacing/cellpadding |
| 00:55 | <Hixie> | zcorpan: file a bug, i'll happily update this accordingly |
| 00:56 | <Hixie> | zcorpan: if we have data in particular that would be fantastic |
| 00:56 | <zcorpan> | it's an area i'd like to research more |
| 00:59 | <Hixie> | cool |
| 00:59 | <Hixie> | in other news: i hate bidi |
| 00:59 | <zcorpan> | still? |
| 00:59 | <TabAtkins> | yeah, I thought you were done with that. |
| 01:00 | <Hixie> | wtf is the difference between "xxx ‮ ab cd ef ‬ yyy" and "xxx HE BR EW yyy"? |
| 01:00 | <TabAtkins> | There... shouldn't be one? |
| 01:00 | <zcorpan> | filed a bug |
| 01:00 | <zcorpan> | nn |
| 01:01 | <Hixie> | nn |
| 01:02 | <Hixie> | http://software.hixie.ch/utilities/js/live-dom-viewer/saved/937 |
| 01:02 | <Hixie> | oh man |
| 01:02 | <Hixie> | i forgot textarea is visual |
| 01:02 | <Hixie> | gah |
| 01:02 | <Hixie> | seriously. hate. bidi. |
| 01:03 | <TabAtkins> | Awesome. |
| 01:03 | <TabAtkins> | Man, I can't even tell which of the three represents the logical ordering. |
| 01:03 | <Hixie> | my brain hurts. |
| 01:28 | <Hixie> | TabAtkins: mail sent to public-css-testsuite |
| 01:28 | <TabAtkins> | kk |
| 01:31 | <Hixie> | christ, faulkner really can't stop himself from filing issues |
| 01:31 | <Hixie> | it's like he has issue turrets |
| 01:32 | <TabAtkins> | tourettes |
| 01:32 | <Hixie> | er, right |
| 01:32 | <TabAtkins> | rather, tourette's |
| 01:32 | <aho> | tourette's |
| 01:32 | <aho> | heh |
| 01:32 | <aho> | :> |
| 01:32 | <TabAtkins> | I like issue turrets, though. |
| 01:32 | <aho> | and torrents |
| 01:40 | <aho> | mh. looks like mozilla still didn't do an update notification for ff4 |
| 01:40 | <aho> | http://gs.statcounter.com/#browser_version-DE-daily-20110320-20110418 |
| 01:40 | <aho> | ff4 got already more than ie8 in germany :> |
| 01:43 | <tw2113> | woo |
| 01:43 | <tw2113> | aho i swear i know you from other freenode rooms |
| 01:44 | <aho> | yes, you do |
| 01:44 | <tw2113> | i just forget which ones |
| 01:44 | <aho> | whois me :P |
| 01:44 | <tw2113> | either LAS or the fedora stuff |
| 01:44 | <tw2113> | [07:48pm] »» [aho] #whatwg |
| 01:44 | <tw2113> | :P |
| 01:45 | <aho> | oh... is that hidden nowadays? :f |
| 01:45 | <aho> | * [aho] #fronteers #impactjs #vp8 #whatwg #commonjs #Node.js #webgl #go-nuts #haxe #fedora-art #inkscape.de #inkscape #lwjgl |
| 01:46 | <tw2113> | aha |
| 01:46 | <tw2113> | small world |
| 01:48 | <aho> | kinda |
| 02:05 | <TabAtkins> | Anyone know a good parser generator? Alternately, anyone want to write a parser for me? |
| 02:07 | TabAtkins | decides to play with PEG.js |
| 02:18 | <doublec> | TabAtkins: I wrote https://github.com/doublec/jsparse a while ago and it still gets some usage |
| 04:55 | <MikeSmith> | can somebody please try http://microformats.org/wiki/existing-rel-values and let me know if you get a response? |
| 04:56 | <MikeSmith> | for me, it's hanging |
| 04:57 | <tw2113> | on it |
| 04:57 | <tw2113> | connecting.... |
| 04:57 | <tw2113> | connecting.... |
| 04:58 | <tw2113> | hanging..... |
| 04:58 | <tw2113> | http://www.downforeveryoneorjustme.com/microformats.org/ |
| 05:14 | <MikeSmith> | hober: can somebody maybe give microformats.org a kick? |
| 05:15 | <MikeSmith> | it seems to not be responding to any requests at all |
| 05:16 | tw2113 | kicks microformats.org's tires |
| 05:30 | <heycam> | do iframes delay the referencing document's load event? |
| 05:31 | <roc> | yes |
| 05:32 | <heycam> | thanks |
| 05:56 | <kennyluck> | That fantasai is the Webmaster of W3C remains on this page :p http://dev.w3.org/cvsweb/~checkout~/csswg/css3-content/Overview.html?rev=1.3;content-type=text%2Fhtml |
| 07:45 | <MikeSmith> | wow |
| 07:45 | <MikeSmith> | http://www.browsium.com/faqs/ |
| 07:45 | <MikeSmith> | "UniBrows is an innovative new product that removes a key impediment to enterprise upgrades by enabling millions of IE6-based line-of-business web applications to run unmodified on Windows 7 or on IE8 on Windows XP (a version designed to work with IE9 is under development). " |
| 07:47 | <MikeSmith> | (by way of hsivonen twitter feed) |
| 08:08 | <MikeSmith> | http://my.opera.com/hallvors/blog/2011/04/20/a-peculiar-cross-browser-onresize-quirk |
| 08:09 | <MikeSmith> | "a cross-browser compatible quirk" |
| 08:16 | <hsivonen> | jgraham: is it intentional that Ragnarök doesn't reset the form pointer when the parent chain of the <input> changes in http://software.hixie.ch/utilities/js/live-dom-viewer/saved/939 ? |
| 08:57 | <jgraham> | hsivonen: Not sure. It could just be a bug. But it is not 100% clear to me without reading the spec closely why 1 is wrong. |
| 09:00 | <jgraham> | Oh, but there is more magic elese where isn't there… |
| 09:03 | <jgraham> | So I guess we don't run the algorithm to reset the form owner of an element when we run the AAA |
| 09:04 | <jgraham> | And it is not really clear to me if we should per spec given the note "The HTML parser overrides this requirement when inserting form controls." |
| 09:05 | <jgraham> | But if it is needed for web-compat I guess we have to |
| 09:05 | <hsivonen> | jgraham: there is evidence suggesting that what Ragnarök does may be more Web-compatible than what Firefox and Chrome do |
| 09:06 | <jgraham> | Unexpected win! |
| 09:08 | <jgraham> | For my next trick I will try to work out what the spec has to say about http://testsuites.opera.com/script-execution/040.html |
| 09:09 | <jgraham> | (chrome, gecko and opera all have different behaviour) |
| 09:10 | <jgraham> | But IE9 matches Gecko |
| 09:26 | <jgraham> | So, I think the spec makes this case explicitly racy |
| 09:26 | <jgraham> | hsivonen: Do you have any opinions here? |
| 09:27 | <hsivonen> | jgraham: both Gecko's behavior and the behavior marked as expected on the test case could be explained to make sense. |
| 09:28 | <hsivonen> | jgraham: changing this is Gecko would be annoying and I wouldn't want to be the person changing things |
| 09:28 | <hsivonen> | jgraham: bz might have opinions |
| 09:28 | <hsivonen> | jgraham: also, if Gecko matches IE, it's a good reason not to change Gecko |
| 09:28 | <jgraham> | hsivonen: The other behaviour I think makes sense (moreso than the expected one) is internal script #1, end script #1, JS URL, frame script, inline script #2 |
| 09:29 | <hsivonen> | jgraham: we're not gonna do that |
| 09:29 | <jgraham> | hsivonen: Why? |
| 09:29 | <hsivonen> | jgraham: since that behavior would require synchronous parsing of the document it the iframe |
| 09:30 | <jgraham> | Rwally? |
| 09:30 | <jgraham> | *Really |
| 09:30 | hsivonen | opens the test case again |
| 09:31 | <jgraham> | Isn't it just the Gecko behaviour with the javascript: URI dereferenced before the second script is run rather than after> |
| 09:31 | <hsivonen> | jgraham: correction: it would require blocking scripts on the parent until scripts in the child frame have executed |
| 09:31 | <hsivonen> | hmm. |
| 09:32 | <hsivonen> | which I admit is kinda similar to blocking scripts on style sheet loads on another documnet |
| 09:32 | <jgraham> | Getting it consistently like that might make it blocking |
| 09:33 | <jgraham> | I guess I should test what happens if you have a slow non-javascript-uri resource in the iframe |
| 09:33 | <hsivonen> | blocking scripts on the parent until the child has run all its scripts would effectively mean blocking scripts on the parent until the whole iframed doc has loaded |
| 09:33 | <hsivonen> | which would be perf poison |
| 09:33 | <jgraham> | yes |
| 09:34 | <jgraham> | I assumed this case was inherently racy with javascript: URIs racing rather fast |
| 09:34 | <hsivonen> | yeah, seems so |
| 10:15 | <jgraham> | hsivonen: Does Gecko always wait to execute scripts in subframes until all the scripts on the main page have run? |
| 10:17 | jgraham | is sure he is doing something wrong |
| 10:18 | <hsivonen> | jgraham: I would imagine not, but I haven't tested |
| 13:43 | <hsivonen> | has anyone done the research on whether non-initial navigations to about:blank need to be synchronous? |
| 15:23 | <MikeSmith> | yee-hah http://microformats.org/wiki/existing-rel-values is back |
| 15:24 | MikeSmith | now starts to think about how to parse the needed info out of that page |
| 15:24 | hsivonen | battles with about:blank |
| 15:25 | jgraham | feels more sorry for hsivonen than MikeSmith, on the whole |
| 15:25 | <MikeSmith> | heh |
| 15:25 | <MikeSmith> | me too |
| 15:25 | <MikeSmith> | :) |
| 15:26 | <MikeSmith> | anyway, the taxonomy of maturity states on that wiki page is quite a bit more complicated than what the spec has |
| 15:26 | <jgraham> | Although unless they added lots of hidden metadata, that page doesn't look optimised for machine readability |
| 15:26 | <jgraham> | Which is ironic |
| 15:27 | <MikeSmith> | yeah |
| 15:27 | <MikeSmith> | is this a mediawiki? |
| 15:28 | <MikeSmith> | yeah, it is |
| 15:28 | <jgraham> | Indeed. It is surprisingly non-ugly for a media wiki install |
| 15:28 | <jgraham> | s/ // |
| 15:29 | <MikeSmith> | anyway, the spec implies the only statuses we need concern ourselves with are "proposed" or "ratified" and "discontinued" |
| 15:29 | <MikeSmith> | but here I find "brainstorming", "POSH", more |
| 15:29 | <MikeSmith> | and "dropped" rather than "discontinued" |
| 15:30 | <MikeSmith> | I guess those status levels were for the whatwg rel-values wiki page |
| 15:35 | <hsivonen> | Whoa! document.title on about:blank is non-blank in Opera! |
| 15:40 | <zcorpan> | about:blank isn't actually blank in opera |
| 15:41 | <hsivonen> | zcorpan: spec violation! |
| 15:41 | <jgraham> | hsivonen: We know :p |
| 15:41 | <zcorpan> | yes |
| 15:42 | <jgraham> | I think fixing it is non trivial for some reason) |
| 15:42 | <jgraham> | +( |
| 15:42 | <hsivonen> | hooray. non-Opera browsers don't perform non-initial navigation to about:blank synchronously |
| 15:42 | <jgraham> | (but I don't remember what that reason is so maybe I imagined it) |
| 15:43 | <zcorpan> | we also use standards mode for about:blank |
| 15:43 | <MikeSmith> | jgraham: sounds like a dispassionate violation |
| 15:44 | <MikeSmith> | if you're going to do some violation, make it passionate violation |
| 15:44 | <zcorpan> | you're right, we should be more passionate about our spec violations |
| 15:44 | <MikeSmith> | willful, with a flourish |
| 15:45 | <zcorpan> | get ready for some willful and passionate <time> violations |
| 15:45 | <jgraham> | zcorpan: Sshh, the time travel feature was supposed to be secret |
| 15:45 | <zcorpan> | oops |
| 15:47 | <hsivonen> | I'm failing at basic testing |
| 15:47 | <hsivonen> | iframe.contentWindow.addEventListener("load", iframeLoad, false); should work, right? |
| 15:47 | <jgraham> | hsivonen: If Opera is doing screy things with about:blank navigation, or navigation in general, it would be nice to know what we are doing wrong |
| 15:47 | <hsivonen> | to listen to load events inside the iframe? |
| 15:48 | <zcorpan> | yeah, should work i think |
| 15:48 | <hsivonen> | found the problem |
| 15:53 | jgraham | accidentially runs across the brilliant term "Eigenplot" to describe a store in which each of the character's unique traits, however normal, is required to overcome some obstacle that would otherwise be insurmountable |
| 16:06 | <jgraham> | So, how do people on touch devices browse xkcd? |
| 16:08 | <Rik`> | jgraham: m.xkcd.com |
| 16:09 | <hsivonen> | it's still more useful to have a user stylesheet that render the title attribute right away |
| 16:11 | <jgraham> | Rik`: Oh, interesting. s/xkcd/Girls with Slingshots/ or Abtruse Goose or some other comic that has started doing the same thing |
| 16:12 | <jgraham> | hsivonen: In this case the reason the title is used is explicitly because the UI makes it hidden until user interaction occurs |
| 16:12 | <MikeSmith_> | it's arguably even more useful to not use the title attribute for that to begin with |
| 16:12 | <hsivonen> | jgraham: I know. I fight the design with my user style sheet. |
| 16:13 | <jgraham> | MikeSmith: Sure, but it is easy so it's not surprising that people do it |
| 16:13 | <jgraham> | hsivonen: Why? Doesn't that rather weaken the payoff? |
| 16:14 | <MikeSmith_> | jgraham: yeah, but for a site like this that's used by so many people, you'd think doing something more than the easy thing would be worthwhile |
| 16:15 | <jgraham> | For xkcd maybe. Most of the people doing this are hardly experts though |
| 16:15 | <MikeSmith> | sure |
| 16:15 | <jgraham> | It seems unreasonable to expect webcomics authors to also master HTML and javascript and whatever |
| 16:16 | <jgraham> | Just to have a different-to-everyone-else-slightly-hidden-but-still-discoverable extra joke in the comics |
| 16:16 | <MikeSmith> | true, but I'm sure in this case he could get lots of help |
| 16:16 | <MikeSmith> | well, it's kind of a key feature of that particular comic |
| 16:17 | <jgraham> | I think Randall would likely be fine. But he's not the only person doing this |
| 16:17 | <MikeSmith> | yeah, I know |
| 17:05 | <jgraham> | gsnedders: Happy birthday BTW. Are you celebrating it on by being on IRC whilst playing computer games? |
| 17:34 | <MikeSmith> | I don't understand what the "IE10" column in http://crockford.com/javascript/performance.html represents |
| 17:34 | <MikeSmith> | oh |
| 17:34 | <MikeSmith> | now I do |
| 17:34 | <MikeSmith> | "relative to IE10"? |
| 17:35 | <Rik`> | that's my understanding too |
| 17:36 | <MikeSmith> | funny that Crockford has such deep insight into what JS engine developers are doing |
| 17:36 | <MikeSmith> | "lacking credible benchmarks, engine developers are tuning to what they have" |
| 17:36 | <MikeSmith> | if he says so, I guess it must be true |
| 17:36 | <MikeSmith> | because he is a true sage |
| 17:36 | <jgraham> | Lacking credible benchmarks, I decided that my program was obviously a credible benchmark |
| 17:37 | <MikeSmith> | heh |
| 17:37 | <jgraham> | Because I entirely missed the point of that Microsoft paper |
| 17:37 | <Rik`> | I'm surprised about the difference between IE9 and IE10 |
| 17:38 | <Rik`> | I don't remember any announcement around JS perf in IE10 |
| 17:38 | <Rik`> | maybe it's "native ES5" ? |
| 17:38 | <MikeSmith> | JS engine developers are clearly just a bunch of asses who have no clue what they are doing |
| 17:38 | <jgraham> | Since Crockford helpfully made it easy to rerun his test you cantry and reproduce his results |
| 17:38 | <MikeSmith> | they should all just be replaced by instances of Crockford |
| 17:39 | <MikeSmith> | browser projects should all hire Crockford to come in and educate their dev teams about how to develop JS engines propertly |
| 17:39 | <MikeSmith> | that would be fun |
| 17:39 | <MikeSmith> | I would pay real money personally to attend such sessions |
| 17:40 | <MikeSmith> | Crockford is like Eric Clapton during the Cream days |
| 17:41 | <MikeSmith> | except he doesn't even need a Jack Bruce and a Ginger Baker |
| 17:41 | <MikeSmith> | he is a world-class power trio solely on his own |
| 17:41 | <MikeSmith> | like the holy trinity |
| 17:41 | <miketaylr> | he's even on tour |
| 17:41 | <MikeSmith> | sweet |
| 17:42 | <miketaylr> | coming to brooklyn to speak at etsy next week |
| 17:42 | <MikeSmith> | well, count yourself blessed |
| 17:43 | <MikeSmith> | it will be the greatest thing that Brooklyn has seen since Walt Whitman |
| 17:43 | <miketaylr> | been a long time coming |
| 17:44 | <MikeSmith> | or at least since RZA |
| 17:45 | <MikeSmith> | Jim Jarmusch should do a "Coffee and Cigarettes" reprise with Crockford and RZA and DZA |
| 17:45 | <miketaylr> | nah, i think wu tang is all from staten island |
| 17:46 | <MikeSmith> | RZA and DZA is from Brooklyn |
| 17:46 | <MikeSmith> | I thikn at least |
| 17:46 | <miketaylr> | ah, wikipedia confirms |
| 17:46 | miketaylr | stands corrected |
| 17:47 | <MikeSmith> | not to knock the various islands and what not |
| 17:47 | <MikeSmith> | the Ramones were for Queens… |
| 17:47 | <MikeSmith> | *from |
| 17:48 | <akahn> | RZA is from staten island |
| 17:48 | <akahn> | oh i should read |
| 17:48 | <akahn> | :-$ |
| 17:48 | <akahn> | i don't know who DZA is though ;) |
| 18:13 | <MikeSmith> | goddammit |
| 18:13 | <MikeSmith> | I really hate browser caching behavior |
| 18:14 | <karlcow> | :) |
| 18:14 | <MikeSmith> | why the fuck browsers can't all make it easy for me to bypass their cache is beyond me |
| 18:14 | <MikeSmith> | jesus |
| 18:14 | <MikeSmith> | the hours of wasted time… |
| 18:17 | <MikeSmith> | the only thing that I have found that does it right is the Develop > Disable Caches option in Webkit/Safari |
| 18:17 | <MikeSmith> | everything else is fuckt |
| 18:17 | <MikeSmith> | sorry to have to say |
| 18:17 | <MikeSmith> | but it's the truth |
| 18:18 | <wilhelm> | Oh, that's a nice feature. |
| 18:19 | <miketaylr> | Preferences > Advanced > History and disabling disk & memory cache is a bit less friendly |
| 18:19 | <Rik`> | MikeSmith: I've heard that the Disable Caches option in Safari is not that reliable |
| 18:19 | <miketaylr> | (in opera) |
| 18:19 | <Rik`> | (but I have no source that I remember to support that) |
| 18:22 | <MikeSmith> | Rik`: you have heard wrong |
| 18:22 | <MikeSmith> | miketaylr: yeah, just a wee bit |
| 18:26 | <TabAtkins> | jgraham: On the subject of hidden jokes in comics, were you aware the Dinosaur Comics embeds 3 separate jokes? |
| 18:26 | <TabAtkins> | (I have a Chrome extension solely concerned with showing all three of them automatically. |
| 18:27 | <jgraham> | TabAtkins: I have to say I don't read Dinosaur Comics regularly |
| 18:27 | <jgraham> | so no |
| 18:27 | <jgraham> | This could be a failing on my part |
| 18:28 | <TabAtkins> | It is. |
| 18:28 | karlcow | doesn't even know what Dinosaur Comics |
| 18:28 | <MikeSmith> | I have a Chrome extension that tells me when the latest/next earthquake has happened |
| 18:28 | <TabAtkins> | qwantz.com |
| 18:28 | <karlcow> | MikeSmith already made my day today. |
| 18:29 | <MikeSmith> | yay for that |
| 18:30 | <MikeSmith> | though I have no idea what I did |
| 18:30 | <MikeSmith> | if I got something productive done today, it's a surprise to me |
| 18:30 | <karlcow> | handwritten manuscript |
| 18:31 | <karlcow> | Walt Whitman |
| 18:35 | <MikeSmith> | ah |
| 20:14 | <gavin> | is window.getSelection specced somewhere? |
| 20:15 | <gavin> | oh, http://www.w3.org/Bugs/Public/show_bug.cgi?id=10798 |
| 20:15 | gavin | reads http://html5.org/specs/dom-range.html |
| 23:03 | <yuhong_> | Here is a big site using real XHTML, BTW: http://us.battle.net/wow/en/ |
| 23:04 | <gsnedders> | http://www.tesco.com/ *used* to. |
| 23:04 | <gsnedders> | (It was fairly trivial to get it to output something non well-formed, though) |
| 23:07 | <Philip`> | http://us.battle.net/wow/en/search?q=potato%ef%bf%bf - that does look like XML |
| 23:14 | <yuhong_> | Old twitter used to generate well-formed XHTML, but then they added <meta http-equiv="X-UA-Compatible" without the slash. |
| 23:14 | <yuhong_> | I reported this to @twitter/team and they fixed it. |
| 23:17 | <espadrine> | Philip`: syntax error. ugly. |
| 23:18 | <yuhong_> | Here is an embarrassing one, BTW: |
| 23:18 | <yuhong_> | http://twitter.com/#!/yuhong2/status/60591084190969856 |
| 23:29 | <TabAtkins> | Yay for parser generators! In not too much time (most was just learning my way around), I built a parser for the stylesheet in Lists, so I can systematically rewrite it. |
| 23:42 | <yuhong_> | Sorry, this was a mistake, they are using HTML5 now. |
| 23:48 | <yuhong_> | Twitter's XHTML is broken again: They put "<meta http-equiv="X-UA-Compatible" content="IE=edge">" without the slash. |
| 23:50 | <yuhong_> | I mentioned this about how use of XHTML in gitweb prevented XSS: |
| 23:50 | <yuhong_> | http://www.no-ack.org/2010/12/cross-side-scripting-vulnerability-in.html |
| 23:58 | <TabAtkins> | Uh, no, that's not true at all, yuhong. By accident, this particular XSS hole is slightly harder to exploit, because the page spams the filename into the code twice. If anything, you should be saying that poor accessibility prevented an XSS, because they're using the filename as the @alt and @title value of an image. |
| 23:58 | <aho> | he left the building |
| 23:59 | <TabAtkins> | Ah well. I never understand what he's trying to say anyway. He comes in, spams some link that's vaguely about xhtml, then doesn't say anything else. |
| 23:59 | <aho> | using the file name for title and alt is kinda... uhm... pointless :> |
| 23:59 | <aho> | heh |