| 08:41 | <gsnedders> | So trying to use testharness.js, I want a bunch of tests to run after an iframe has loaded. How do I do this? |
| 08:42 | <annevk> | invoke a function from onload that invokes test.step |
| 08:43 | <annevk> | see the XHR test cases, they deal with asynchronous tests |
| 08:45 | <gsnedders> | But I thought test.step only reported one test back, and if the test object has already failed, it gives up. |
| 08:47 | <annevk> | test.step just makes sure exceptions are caught within the test framework |
| 08:47 | <annevk> | it does not do any reporting |
| 08:49 | <gsnedders> | But Test.prototype.step begins with: if (this.status !== null) return |
| 08:50 | <annevk> | sure, currently stuff stops when something has failed |
| 08:55 | <gsnedders> | Hmm, could call test() from within test.step… |
| 08:56 | <gsnedders> | Or not even that, really… |
| 08:57 | <gsnedders> | http://software.hixie.ch/utilities/js/live-dom-viewer/?%3C!DOCTYPE%20html%3E%0D%0A%3Cdiv%20id%3Dlog%3E%3C%2Fdiv%3E%0D%0A%3Cscript%20src%3Dhttp%3A%2F%2Ftest.w3.org%2Fresources%2Ftestharness.js%3E%3C%2Fscript%3E%0D%0A%3Cscript%3E%0D%0Avar%20t%20%3D%20async_test(%22Simple%20async%20test%22)%3B%0D%0A%20%20%20%20function%20windowLoad()%0D%0A%20%20%20%20%7B%0D%0Atest(function()%7Bassert_true(false)%7D%2C%20%22foobar%22)%3B%0D%0Atest(function()%7Bassert_true |
| 08:57 | <gsnedders> | How evil is that? |
| 08:58 | <gsnedders> | jgraham: for when you return see above |
| 09:00 | <Ms2ger> | gsnedders, incomplete URL |
| 09:01 | <gsnedders> | http://software.hixie.ch/utilities/js/live-dom-viewer/saved/692 |
| 09:04 | <annevk> | http://pivotfinland.com/frozendefence/ is pretty cool |
| 09:05 | <annevk> | gsnedders, http://tc.labs.opera.com/apis/XMLHttpRequest/open-url-javascript-window-2.htm |
| 09:06 | <gsnedders> | annevk: So? I don't want several hundred tests being reported as one. |
| 09:10 | <annevk> | so you start multiple async tests |
| 09:11 | <gsnedders> | annevk: That's kinda cool… |
| 09:24 | <gsnedders> | annevk: Also, a productivity killer :) |
| 13:59 | <karlcow> | http://blogs.adobe.com/designandweb/2010/10/improving-the-web-for-digital-publishing.html |
| 14:03 | <Aleoss> | ReCAPTCHA is broken. http://www.darkreading.com/authentication/security/vulnerabilities/showArticle.jhtml?cid=sem_edit_Security&articleID=226700514&wc=4 |
| 16:02 | <Rik`> | karlcow: where is the syntax/source code ? :( |
| 16:03 | <Rik`> | (for adobe/text wrapping) |
| 16:13 | <karlcow> | Rik`: oui exactement. C'est pour cela que j'ai laissé un commentaire au moins pour une spec |
| 16:13 | <karlcow> | >By karl - 2:10 PM on November 7, 2010 Reply |
| 16:13 | <karlcow> | Did you publish a draft specification document somewhere about it? Could you send it as a W3C Member Submission and/or Editor’s draft to CSS WG. The earlier, the better. |
| 16:14 | <Rik`> | that would be a great effort, but even an example would be enough for now |
| 16:57 | <AryehGregor> | Philip`, tcpcrypt adds no latency if the other side doesn't support it. It adds one round-trip on the first connection to a server if both sides support it, but the resulting secret can be cached, so not every connection. |
| 17:13 | <AryehGregor> | Okay, Firefox 4's "switch to tab" feature is really annoying. Often I have one tab open talking about some site, then I want to open a new tab with that site in it, but the first option is the already-open tab I just left, so I have to go down to the second option . . . |
| 17:26 | <AryehGregor> | Looks like tcpcrypt broke something or other. |
| 17:26 | <AryehGregor> | Cool idea, though. |
| 17:26 | AryehGregor | turns it off |
| 18:50 | <AryehGregor> | Is it just me, or does Google autocorrect searches too aggressively these days? |
| 18:50 | <AryehGregor> | "Did you mean: firefox 2 market share wikipedia" No, I meant "wikimedia", not "wikipedia", that's why I typed "wikimedia". |
| 18:54 | <Philip`> | I find it irritating since I usually don't even notice it's done the auto-correction |
| 18:54 | <Philip`> | so I get confused as to why the search results all seem unrelated to what I wanted |
| 19:02 | <AryehGregor> | I generally notice, but it's an extra click to fix it. |
| 19:02 | <AryehGregor> | But it does that *way* too much. |
| 19:02 | <AryehGregor> | Searching for even a slightly obscure term will often trigger it if it happens to be spelled similarly to a common term. |
| 19:06 | <AryehGregor> | There's something to be said for predictability. |
| 19:06 | <AryehGregor> | Programs that try to be too clever can be annoying. |
| 19:07 | <Ms2ger> | Programs can be annoying. |
| 19:07 | <smaug____> | Google's "autocorrect" usually does something wrong for me. Same with Google Instant (which is why I disabled it pretty fast) |
| 19:07 | <Ms2ger> | Fixed that for you |
| 19:28 | <Lachy> | AryehGregor, re google autocorrect, don't you mean carrots? http://knowyourmeme.com/i/000/081/068/original/5Q9Ok.png?1288919403 |
| 19:41 | <annevk> | idea: move the obsolete IDL interfaces into the other interface blocks (when one exists) shielded with class="impl obsolete" |
| 19:41 | <annevk> | or some such |
| 19:45 | <Ms2ger> | Why? |
| 19:46 | <annevk> | the current setup is more complicated than needed |
| 19:46 | <annevk> | and hence confuses developers |
| 19:48 | <annevk> | it's not that important however |
| 19:57 | <Dashiva> | Roy works for Adobe now? Interesting. |
| 20:00 | Dashiva | ponders the absurdity of requiring minimum contrast levels for greyed-out text |
| 20:07 | <Dashiva> | And IRC is inaccessible to blind people? Really? |
| 20:11 | <annevk> | I've seen blind people use it successfully |
| 20:11 | <annevk> | quite a few actually |
| 20:12 | <Dashiva> | According to Shelley, it isn't |
| 20:13 | <heycam> | yeah, to be honest, i would have thought that irc would be one of the more accessible communication mediums |
| 20:14 | <annevk> | I thought Shelly could see? |
| 20:15 | <othermaciej> | blind people are on W3C IRC all the time |
| 20:15 | <othermaciej> | during meetings and everything |
| 20:18 | <Dashiva> | Maybe she was saying the IRC logs website is inaccessible, but that seems equally odd |
| 20:36 | <AryehGregor> | I've spoken to blind people on IRC. |
| 20:36 | <AryehGregor> | At least one. |
| 20:36 | <AryehGregor> | He did write "smile" instead of ":)", though. Apparently because his screen-reader read ":)" as "colon right parenthesis" or something. |
| 20:36 | <AryehGregor> | Emoticons are kind of fail for blind people, I guess. |
| 20:37 | <Dashiva> | I've been wondering about that |
| 20:37 | <Dashiva> | I would assume the basic smileys are pervasive enough that screen readers would handle them, but at the same time I keep hearing about how screen readers don't support so many apparently obvious things |
| 20:38 | <AryehGregor> | Presumably the screen-reader could translate, just like some IM programs annoyingly translate them to images. |
| 20:38 | <AryehGregor> | But yeah, screen-readers seem to just be garbage. |
| 20:38 | <Dashiva> | (Or not necessary don't support, but hide it behind non-default settings) |
| 20:39 | <AryehGregor> | Maybe there are too few blind people to lend them to much competition, or maybe they have to focus all their resources on heuristics to decide what text to read instead of how to read it, or something. |
| 20:39 | <AryehGregor> | But reading bold and italics the same as regular text? That just seems . . . I dunno. |
| 20:39 | <Dashiva> | Especially since replacing <i> and <b> was phrased by many (not all, of course) as an accessibility thing |
| 20:40 | <annevk> | gregory told me the other day he uses some kind of weird config file to make a mapping from :-) to "smile" |
| 20:40 | <annevk> | which seems annoying |
| 20:40 | <annevk> | until then he did not know what o_O meant either |
| 20:40 | <Dashiva> | "emoticon: left-facing noseless smiling face" |
| 20:41 | <Dashiva> | I bet unicode has good descriptions we could reuse |
| 20:41 | <AryehGregor> | Emoticons must be pretty opaque if you've been blind your whole life. |
| 20:41 | <AryehGregor> | (presumably not if you lost your sight later in life) |
| 20:41 | <AryehGregor> | Although I guess you can figure out from context eventually. |
| 20:41 | <Dashiva> | I actually have that problem as a non-disabled user |
| 20:42 | <Dashiva> | When I communicate with heavy MSN users, they will use complex smileys and my client doesn't translate |
| 20:42 | <Dashiva> | I ran into ({) earlier this week |
| 20:43 | <AryehGregor> | Oh, well, that's because they're really writing in some markup language that just masquerades as plaintext. |
| 20:44 | <Dashiva> | Unicode will save us |
| 20:44 | <AryehGregor> | Actually, it includes explicit markup for things like bold, so why doesn't it mark up smilies explicitly? |
| 20:44 | <AryehGregor> | Will Unicode be able to keep up with lists like this? http://www.astro.umd.edu/~marshall/smileys.html |
| 20:44 | <Dashiva> | I think w3c worked on some kind of emoticon markup language |
| 20:44 | <AryehGregor> | ******:-) Marge Simpson |
| 20:45 | <AryehGregor> | :-F Bucktoothed vampire with one tooth missing |
| 20:45 | <Dashiva> | Unicode is outpacing us by far |
| 20:45 | <mgdm> | MSN messenger can be set up to intercept more or less any bit of text and replace it with an emoticon of some sort |
| 20:45 | <Dashiva> | POUTING CAT FACE |
| 20:46 | AryehGregor | places an order that "is 100% secure using SSL technology." |
| 20:46 | <Dashiva> | #unicodebandnames is my favorite twitter hashtag |
| 20:47 | <annevk> | except that without an iPhone with emoji enabled Unicode emoticons are hard |
| 20:47 | Ms2ger | is reminded of http://ln.hixie.ch/?start=1031829070&count=1 |
| 20:47 | <Dashiva> | Doesn't the iPhone have an emoticon shortcut panel? |
| 20:48 | <AryehGregor> | !⃤ |
| 20:48 | <AryehGregor> | Doesn't render well in my font. |
| 20:48 | <AryehGregor> | !⃤ |
| 20:48 | <Dashiva> | Narrow exclamation point and narrow box for me |
| 20:49 | <annevk> | Dashiva, yes, if you download some app to enable it |
| 20:49 | <annevk> | Dashiva, I did this last weekend, I'm still discovering new icons |
| 20:49 | <annevk> | euh, last week |
| 20:49 | <AryehGregor> | Sadly, the more intricate Unicode icons really need a large font size to display properly. |
| 20:50 | <hendry> | Is SVG in HTML5 just done with an img tag? http://svg.dabase.com/ -- or can someone point me out to a better example (that works in FF)? |
| 20:50 | <AryehGregor> | hendry, you can do SVG with an img tag in all reasonably recent browsers that support SVG, except Firefox 3.x. Firefox supports it in 4.0. |
| 20:50 | <AryehGregor> | For earlier Firefox versions, you have to use <object> or <iframe>, IIRC. |
| 20:50 | <AryehGregor> | (which both stink) |
| 20:51 | <Hixie> | <!DOCTYPE html><title>SVG in HTML</title><svg><circle cx=10 cy=10 r=10 /></svg> |
| 20:51 | Ms2ger | gives Hixie a badge saying "I live on the edge" |
| 20:51 | <AryehGregor> | By "that works in FF" I assume he means "that works in final released versions of Firefox without turning on the HTML5 parser in about:config", but maybe I'm wrong. |
| 20:52 | <hendry> | Hixie: inline doesn't work for me :) http://svg.dabase.com/inline.html |
| 20:53 | <AryehGregor> | hendry, it does in Firefox 4, and IE9, and some version of WebKit somewhere. |
| 20:53 | <Hixie> | the above works for me in chrome, didn't test ff |
| 20:53 | <AryehGregor> | Oh, it works in Chrome too now? |
| 20:53 | <AryehGregor> | Seems to. |
| 20:53 | <AryehGregor> | Just dev channel, or has that been pushed out to beta/stable yet? |
| 20:53 | <hendry> | Chrome 9.0.570.1 renders half my face with http://svg.dabase.com/inline.html |
| 20:53 | <Hixie> | what is this "stable" you speak of? :-) |
| 20:53 | <AryehGregor> | inline.html looks fairly broken in HTML. |
| 20:53 | <AryehGregor> | Er. |
| 20:53 | <annevk> | AryehGregor, the iPhone just renders the glyphs using an image |
| 20:53 | <AryehGregor> | In Chrome. |
| 20:54 | <annevk> | taking fonts to a whole new level |
| 20:54 | <AryehGregor> | annevk, like a colored bitmap image? A PNG or something? |
| 20:54 | <annevk> | something like that |
| 20:54 | <AryehGregor> | http://svg.dabase.com/ works in Chrome, but not Firefox 4 for some reason. |
| 20:55 | hendry | sighs |
| 20:55 | <AryehGregor> | Oh, I guess SVG-in-img hasn't been turned on yet in the Firefox 4 betas. |
| 20:55 | <AryehGregor> | Man, I wish they'd get with the program and adopt a Chrome model for releases, at least something like the dev channel. |
| 20:55 | <Hixie> | firefox has a dev channel |
| 20:56 | <Hixie> | it's called "nightly builds" |
| 20:56 | <Hixie> | and they autoupdate just like chrome |
| 20:56 | <Hixie> | it's what i use |
| 20:56 | <AryehGregor> | They auto-update? I didn't know that. |
| 20:56 | <AryehGregor> | My betas don't even seem to auto-update. |
| 20:56 | <annevk> | mine do |
| 20:56 | <annevk> | well, I don't use beta |
| 20:56 | <Hixie> | well it's not as silent as the chrome updates, but yes, they have (chatty) autoupdate |
| 20:56 | <hendry> | most importantly it doesn't work on a mobile OS like Android, so I guess I need to experiment with <object> |
| 20:56 | <AryehGregor> | What platform? |
| 20:56 | annevk | fetched something wrong http://ftp.mozilla.org/pub/mozilla.org/firefox/nightly/latest-trunk/ |
| 20:57 | <Hixie> | mac |
| 20:57 | <hendry> | Hixie: i doubt http://nightly.mozilla.org/ updates on Linux |
| 20:57 | <Hixie> | but they do it on windows and linux |
| 20:57 | <annevk> | from* |
| 20:57 | <AryehGregor> | I've used nightlies on Linux and they don't seem to auto-update. |
| 20:57 | <AryehGregor> | Chrome gives me a .deb that installs a software source that shows up with all the other updates. |
| 20:57 | <AryehGregor> | That's how it should be done. |
| 20:58 | <Hixie> | oh you have to install it locally, it won't do it using the package thingy, yeah |
| 20:58 | <Ms2ger> | AryehGregor++ |
| 20:58 | <AryehGregor> | Firefox nightlies are a .tar.bz. |
| 20:58 | <AryehGregor> | Doesn't seem likely to auto-update. |
| 20:59 | <AryehGregor> | (although in principle it could) |
| 20:59 | <Hixie> | it does |
| 20:59 | <AryehGregor> | Well, it doesn't for me when I've used it. |
| 20:59 | <Hixie> | at least it did for me when i used linux desktop years ago |
| 20:59 | <AryehGregor> | Maybe their Linux support has regressed. |
| 20:59 | <hendry> | AryehGregor: not getting much luck with http://svg.dabase.com/object.html either. doh. |
| 21:00 | <AryehGregor> | hendry, works for me in Firefox 4, although the picture is cut off. |
| 21:01 | <AryehGregor> | I'd test in Firefox 3.6, but of course, Firefox doesn't let me run two different versions at once unless I jump through hoops like make new profiles or something, which I'm not interested enough to do. |
| 21:01 | <AryehGregor> | (Chrome doesn't let you run two different versions at once, but its versions are only ever separated by like a month or two of development time, so it's not really necessary.) |
| 21:01 | <AryehGregor> | I guess I'll still with trying out Firefox 4 until its final release before I decide whether to switch back to Chrome. |
| 21:04 | AryehGregor | thinks he meant "I'll stick with" rather than "I'll still with" |
| 21:10 | <AryehGregor> | The kind of psychological warfare I'm waging against myself is sort of fascinating. To prevent wasting time, I banned myself from doing my routine morning web tasks (checking non-important e-mail, feeds, common sites, etc.) more than once per day, and wrote some filters to eliminate a lot of the useless mail I get, and cut back on posting comments on the feeds I read. Now, rather than doing any real work, I still manage to spend my time sta |
| 21:10 | <AryehGregor> | ring at IRC chats. |
| 21:10 | AryehGregor | wonders what strategy he'll adopt to combat this. Possibly "stop procrastinating and do the work already". |
| 21:31 | <Chilaquiles> | anybody here helps with html5? |
| 21:34 | <boogyman> | Chilaquiles: this isn't an HTML5 support channel, please visit #html for those discussions |
| 22:38 | <Chilaquiles> | somebody told me that people help in #html but nobody knows there |
| 22:38 | <Chilaquiles> | they just are a bunch of noobs like me |
| 22:43 | <AryehGregor> | boogyman, people can ask their HTML5 support questions here, doesn't hurt. |
| 22:43 | <AryehGregor> | It's one of the only places they're likely to get an informed answer to certain types of questions. |
| 22:44 | <boogyman> | AryehGregor: in that case, please define, and I quote "stylistically offset from the normal prose without conveying any extra importance" |
| 22:45 | <boogyman> | (sorry for "enter" punctuation), and how that's not presentational html |
| 22:45 | <AryehGregor> | The translation of that is "do whatever you want, nobody actually cares about this stuff but we have to appease the semantics mafia". Next question. |
| 22:46 | <Dashiva> | <em> conveys emphasis, not importance, so that seems to apply :) |
| 22:48 | <AryehGregor> | Or at more length, "we obviously need a <b> element because Unicode doesn't let us do bold, but bold is used for a huge variety of basically unrelated semantics, so we're stuck with having a <b> tag, and since we want to maintain a veneer of 'HTML doesn't do presentation, only semantics', and we don't want to not say what <b> is used for, we add a lame attempt at summarizing what bold fonts are generally used for in typography." |
| 22:49 | <boogyman> | Dashiva: semantics != importance |
| 22:54 | <AryehGregor> | boogyman, I filed a bug for you: http://www.w3.org/Bugs/Public/show_bug.cgi?id=11254 |
| 22:54 | <AryehGregor> | Alas, I expect WONTFIX, because Hixie is a member of the semantics mafia. |
| 22:55 | <AryehGregor> | Or at least a fellow traveler. |
| 22:55 | <AryehGregor> | (not to be confused with the Semantic Web mafia) |
| 22:57 | <boogyman> | I'm surprised you actually included all non-semantic elements :P even though <b> is equivalent to font-weight:bold, therefore is applicable with CSS |
| 23:01 | <AryehGregor> | You need some element to apply bold to stuff, though. |
| 23:01 | <AryehGregor> | CSS can't apply to unadorned runs of text. |
| 23:02 | <AryehGregor> | So you need some presentational stuff at a level below CSS, like HTML or Unicode. |
| 23:02 | <AryehGregor> | Unicode is totally presentational, but it has gaps, and CSS can't fill all of them, so HTML has to fill some. |
| 23:02 | <AryehGregor> | Pretending that <br> isn't presentational is just silly. |
| 23:03 | <boogyman> | unadorned runs for text? |
| 23:03 | <boogyman> | of* |
| 23:03 | <aho> | horray for white-space:pre-line :> |
| 23:07 | <AryehGregor> | Runs of text that have no element enclosing them. |
