| 00:01 | <MikeSmith> | zcorpan_: so the problems with putting HTML content within annotation-xml is that the user would similarly need to wrap it in something, right? e.g, a <div> |
| 00:02 | <MikeSmith> | zcorpan_: I think I'm starting to agree that dealing with processing of content in annotation-xml might not be worth the trouble.. |
| 00:02 | <TabAtkins> | What's the point of <annotation-xml>? I feel like I've missed something. |
| 00:03 | <MikeSmith> | I guess it's for annotating non-presentational "content" MathML |
| 00:03 | <TabAtkins> | So you use it inside of MathML to represent an annotation? |
| 00:04 | <MikeSmith> | yeah |
| 00:04 | <TabAtkins> | So it's somewhat similar to SVG's <foreignContent> or whatever? |
| 00:04 | <MikeSmith> | yeah |
| 00:04 | <MikeSmith> | sorta |
| 00:05 | <zcorpan_> | http://www.w3.org/TR/MathML2/chapter4.html#contm.annotation-xml |
| 00:05 | <MikeSmith> | but currently different in one really significant way as far as inclusion in HTML5 goes |
| 00:05 | <MikeSmith> | which is that HTML content in <foreignContent> will get processed as expected |
| 00:05 | <MikeSmith> | but HTML content in <annotation-xml> will now |
| 00:05 | <MikeSmith> | *not |
| 00:06 | <MikeSmith> | which I raised a bug for : http://www.w3.org/Bugs/Public/show_bug.cgi?id=9887 |
| 00:06 | <TabAtkins> | Oh. I would naively think it should just be treated the same. |
| 00:07 | <MikeSmith> | yeah, I think that's what others are likely to expect too |
| 00:07 | <zcorpan_> | MikeSmith: from spec example, the content-mathml would go directly in <semantics>, while presentational-mathml would go in <annotation-xml> |
| 00:07 | <MikeSmith> | zcorpan_: yeah, that's how I understand it too |
| 00:07 | <MikeSmith> | (though probably I didn't say that clearly) |
| 00:07 | <zcorpan_> | which makes me wonder why the css profile spec makes it display:none |
| 00:08 | <othermaciej> | AryehGregor: if you think it's a WebKit bug, then please file it in bugs.webkit.org |
| 00:08 | <othermaciej> | AryehGregor: I agree that respecting the constraints without UI for validation failure is a bad state |
| 00:08 | <MikeSmith> | TabAtkins: but I don't think that just because it's what people expect, we necessarily need to make it work the same |
| 00:08 | <TabAtkins> | I think we need a good reason to make it not work the same. |
| 00:09 | <MikeSmith> | TabAtkins: I don't personally feel strongly about how it ends up getting resolved, but I think it does need to get resolved ... and maybe made clear the document-conformance parts of spec, too |
| 00:09 | <MikeSmith> | TabAtkins: "not worth the trouble" isn't a good reason? |
| 00:10 | <MikeSmith> | TabAtkins: users can put HTML content in the MathML <mtext> element instead |
| 00:10 | <TabAtkins> | Not worth the trouble to save implementor effort and make two elements act identically? |
| 00:10 | <MikeSmith> | TabAtkins: the whole <semantics>...<annotation-xml> thing in MathML is a arguably an ugly language hack to begin with |
| 00:11 | <MikeSmith> | well, Jacques Distler is one person that has said that, at least |
| 00:11 | <TabAtkins> | I wouldn't disagree. I'm just wondering what the benefit is of having a special element that is *almost* the same as SVG's foreignContent, but different in subtle ways. |
| 00:12 | <MikeSmith> | true |
| 00:12 | <zcorpan_> | TabAtkins: annotation-xml needs to support embedding mathml elements |
| 00:12 | <zcorpan_> | TabAtkins: that's incompatible with how foreignContent works |
| 00:12 | <MikeSmith> | ah yeah |
| 00:13 | <MikeSmith> | so it's not really "foreign content" ... it's more like "maybe foreign content, or maybe not" |
| 00:14 | <MikeSmith> | btw, Jacques comment: "I really had no idea how to properly include SVG fragments inside a MathML equation. It turns out that the “correct” answer is some horrible kludge, involving embedding the SVG in the MathML <semantics> element. Go figure …" |
| 00:14 | <TabAtkins> | Ah, so it needs to support mathml children that aren't wrapped in <math>? |
| 00:14 | <MikeSmith> | (from http://golem.ph.utexas.edu/~distler/blog/archives/001475.html) |
| 00:14 | <zcorpan_> | TabAtkins: yes |
| 00:14 | <TabAtkins> | That's weird. |
| 00:14 | <bros> | I am at a loss. I can not implement the handshaking element of WebSockets. Would anybody here be willing to review a paste of the code and see if my errors pop out? |
| 00:15 | <MikeSmith> | but fwiw, Minefield doesn't actually seem to render MathML content in <annotation-xml> unless it's wrapped in an <math> element |
| 00:16 | <MikeSmith> | afaict, at least |
| 00:17 | zcorpan_ | might be wrong |
| 00:19 | <Hixie> | bros: ure |
| 00:19 | <Hixie> | sure even |
| 00:19 | <zcorpan_> | <annotation-xml encoding="MathML-Content"> |
| 00:19 | <zcorpan_> | <apply> |
| 00:19 | <zcorpan_> | http://www.w3.org/TR/MathML2/chapter2.html#fund.mix |
| 00:20 | <TabAtkins> | Every single time someone directs a response to bros, I think they're addressing the room like a surfer dude. "Bros, I submit that these waves are totally sweet." |
| 00:20 | <MikeSmith> | zcorpan_: as far as I can tell from the parsing algorithm, it should not be required.. but as far as I can tell from actually testing it in Minefield, Minefield seems to require it |
| 00:20 | <zcorpan_> | <annotation-xml encoding="MathML-Presentation"> |
| 00:20 | <zcorpan_> | <mrow> |
| 00:20 | <MikeSmith> | TabAtkins: :) |
| 00:20 | <zcorpan_> | certainly the spec examples don't use a wrapping <math> |
| 00:21 | <TabAtkins> | So I guess the next thing to do is see if there's much mathml in the wild that uses <annotation-xml> with mathml children that aren't wrapped in <math>. |
| 00:21 | <TabAtkins> | The spec examples can just be fixed if not. |
| 00:21 | <MikeSmith> | zcorpan_: yeah, I know. I copied and pasted some of them.. but can't get them to render in text/html unless I put <math> around them |
| 00:22 | <zcorpan_> | TabAtkins: there isn't much mathml in the wild |
| 00:22 | <bros> | Hixie: Thank you. http://codepad.org/YBQD16oO The problem I face is no matter what, the WebSocket's readyState is always closed, which makes me think I am sending something wrong. |
| 00:22 | <TabAtkins> | zcorpan_: Then let's fix the spec and make the content model of <annotation-xml> more reasonable. |
| 00:23 | <zcorpan_> | TabAtkins: i don't see how it's not reasonable |
| 00:23 | <boblet> | seems that webkit is being smart and running type="email" validation before submit — getting failure for IDN email addys on non-submit ajax form |
| 00:24 | <Hixie> | bros: well to start with, read() might not return the whole handshake at once |
| 00:24 | <Hixie> | bros: read() just returns what the kernel has received so far, which could be on byte, two bytes, the whole handshake plus a few frames, anything really |
| 00:24 | <TabAtkins> | zcorpan_: It seems more reasonable that every direct child of an <annotation-xml> have a root element, rather than just "every child that isn't mathml". |
| 00:24 | <bros> | Hixie: Very true. I can confirm it does in this case. I removed all of my output debugging. It does though, all the way to the end of the 3rd part of the key. |
| 00:26 | <Hixie> | bros: is this code running on a public server anywhere? |
| 00:26 | <zcorpan_> | Hixie: a websocket client wouldn't send frames along with the handshake |
| 00:26 | <Hixie> | zcorpan_: not a browser one, anyway |
| 00:26 | <bros> | Hixie: No. I could make it. |
| 00:27 | <Hixie> | bros: that would be helpful |
| 00:27 | <zcorpan_> | Hixie: iirc, the protocol doesn't allow a conforming websocket client to do it |
| 00:27 | <Hixie> | bros: btw there's a security bug in your code if i'm not mistaken -- if i connect to the server and send only a few bytes, it'll crash with an out-of-bounds buffer read |
| 00:28 | <Hixie> | zcorpan_: ok, not a conforming one then :-) |
| 00:28 | <bros> | Hixie: I wouldn't doubt it. I have been trying very hard to just get this, working. What line specifically? |
| 00:28 | <bros> | Hixie: Permission to private message you? |
| 00:28 | <Hixie> | bros: the "keys[8] = buf[r - 8];" stuff doesn't check that r > 8 |
| 00:28 | <Hixie> | bros: sure |
| 00:29 | <zcorpan_> | bros: why not discuss it here? people can learn from it |
| 00:29 | <Hixie> | i imagine he wants to send me his IP address :-) |
| 00:29 | <bros> | http://98.249.142.158/ |
| 00:29 | <zcorpan_> | oh |
| 00:29 | <Hixie> | bros: which port is the ws server on? |
| 00:31 | <bros> | Hixie: 8080. The link I linked is a page with new WebSocket() |
| 00:33 | <Hixie> | bros: i don't have a good way to sniff the network here. I'll have to go home and do it. Will you still be around in ~ an hour? |
| 00:33 | <bros> | I have Sec-WebSocket-Origin hard-coded. |
| 00:33 | <bros> | Hixie: Most likely. Let me work on making the code "better". |
| 00:34 | <Hixie> | k. bbiab. |
| 00:34 | <zcorpan_> | bros: WebSocket handshake failure, expected field 'sec-websocket-location' to have value 'ws://98.249.142.158:8080/', found 'ws://98.249.142.158'. |
| 00:34 | <bros> | That is something. Thank you. How did you find that error? I was using Chrome developer tools and couldn't find anything. |
| 00:35 | <zcorpan_> | i used an internal build of opera :) |
| 00:36 | <bros> | very nice. :) |
| 00:37 | <MikeSmith> | zcorpan_: um, please ignore what I said about needing to wrap MathML content in <math> to get it work in annotation-xml in Minefield... you don't need to .. not sure what I was doing wrong before |
| 00:37 | <MikeSmith> | clearly a sign that I need a morning coffee break |
| 00:38 | MikeSmith | goes to make coffee |
| 00:44 | <AryehGregor> | othermaciej, I don't know if it's a WebKit bug, since I normally only use Linux, and there don't seem to be non-Chromium WebKit builds readily available for Linux. |
| 00:44 | <AryehGregor> | I could try using a Windows machine, but I bet there are lots of people here with copies of Safari or WebKit around. |
| 00:44 | <AryehGregor> | To those people: in the data URL data:text/html,<!doctype html><form><input required><input><button>Submit</button></form> |
| 00:45 | <AryehGregor> | Enter text into the second field, leaving the first blank, and try to submit. Does it work? |
| 00:45 | <MikeSmith> | AryehGregor: you know about Arora? (QtWebkit-based Linux browser) |
| 00:45 | <zcorpan_> | AryehGregor: it focuses the first field |
| 00:45 | <AryehGregor> | Nope. Should I try in that |
| 00:45 | <AryehGregor> | ? |
| 00:45 | <AryehGregor> | zcorpan_, right, so that's the bug. |
| 00:45 | <AryehGregor> | I'll file at webkit.org, then. |
| 00:46 | <MikeSmith> | AryehGregor: yeah -- Arora is actually a pretty decent browser |
| 00:47 | <MikeSmith> | AryehGregor: and there's an #arora channel on freenode where you can bug the developers when needed |
| 00:48 | <MikeSmith> | http://code.google.com/p/arora/ |
| 00:49 | <GPHemsley> | Am I right in interpreting that <input> and other form elements do not actually require a <form> parent? |
| 00:50 | <MikeSmith> | GPHemsley: depends on if you want them to actually do anything |
| 00:50 | <MikeSmith> | they do render without a <form> ancestor |
| 00:50 | <GPHemsley> | if I don't need them to submit with a form, then it's valid? |
| 00:51 | <MikeSmith> | yes |
| 00:51 | <MikeSmith> | as far as document conformance goes |
| 00:51 | <GPHemsley> | ok, cool |
| 00:55 | <AryehGregor> | othermaciej, https://bugs.webkit.org/show_bug.cgi?id=40747 |
| 00:59 | <othermaciej> | AryehGregor: ty |
| 01:19 | <Hixie> | bros: did you get it working? |
| 01:23 | karlcow | wonders if there is a use case where an input element has to be in the document without a form and then if not why is it conformant. |
| 01:23 | <karlcow> | not that important maybe. |
| 01:24 | <Dashiva> | Anything that doesn't submit to the server is justified outside a form |
| 01:24 | <Dashiva> | E.g. a javascript calculator |
| 01:25 | <karlcow> | ah I see the use case, UI widget for client side application |
| 01:28 | <AryehGregor> | "FWIW I think Ian’s, er, disregard for social-niceties is sub-optimal. But at this point, I’m so frustrated with the glacial progress of web markup since 1997, I don’t care if he tells the w3c to go fuck itself and publishes HTML5 in limited editions by personally carving it into the skins of dead baby seals. Or ex-XHTML2 WG members. Using Princess Diana’s hipbone. While stamping on the Dali Lama’s spectacles. On stage. On fire. In |
| 01:28 | <AryehGregor> | front of the Queen. Just get it out." |
| 01:30 | <Dashiva> | Well, that's certain an opinion! |
| 01:30 | <Dashiva> | *certainly |
| 01:30 | <boblet> | I MIS |
| 01:31 | <boblet> | I misread it as “just get it out … (in) front of the Queen” to start with (reading from the bottom up |
| 01:31 | <AryehGregor> | Writing HTML5 on the skin of dead XHTML2 WG members would certainly provide a strong incentive for working groups to pay attention to user needs. |
| 01:33 | <Dashiva> | I liked this one too: http://twitter.com/SickingJ/status/16284113128 |
| 02:01 | <MikeSmith> | AryehGregor: I sense a subtle hint of frustration in whoever wrote that |
| 02:01 | <MikeSmith> | but maybe that's just me |
| 02:01 | <MikeSmith> | I'm not a psychologist |
| 02:01 | <AryehGregor> | MikeSmith, that's one possible interpretation of that comment. |
| 02:01 | <MikeSmith> | heh |
| 02:01 | <AryehGregor> | Perhaps he was just obliquely expressing his antipathy toward baby seals. |
| 02:04 | <MikeSmith> | AryehGregor: did the CSS Squirrel write that? Or somebody from the Last Week in HTML5 collective? I ask because it seems to have the same kind of subtle hint of frustration that characterizes those works |
| 02:04 | <Dashiva> | I don't agree |
| 02:05 | <Dashiva> | It seems very different from last week, at least |
| 02:05 | <AryehGregor> | MikeSmith, both of those parties would tend to express the opposite viewpoint. It was a commenter on CSS Squirrel's latest post. |
| 02:05 | <AryehGregor> | "Mattur". |
| 02:05 | <AryehGregor> | http://www.cssquirrel.com/2010/06/15/693comic-update-html5-unicorn-heuristics/#comment-31994 |
| 02:09 | <Dashiva> | http://www.cssquirrel.com/comic/?comic=62 |
| 02:09 | <Dashiva> | I'm pretty sure this isn't a real language |
| 02:09 | <othermaciej> | that is an amusing comment |
| 02:14 | <Dashiva> | Wow, the blog post itself is pretty ignorant |
| 02:15 | <AryehGregor> | Yes, most of Kyle's posts on HTML5 process are pretty ignorant. |
| 02:15 | <AryehGregor> | Kind of a shame. |
| 02:17 | <AryehGregor> | Good thing we don't have to give those ignorant non-implementers any kind of say in anything important, am I right? |
| 02:19 | <MikeSmith> | AryehGregor, Dashiva - I see. I guess it was the subtlety of his wording that threw me off |
| 02:19 | <MikeSmith> | the felicitous language |
| 02:20 | <MikeSmith> | He should state himself more strongly |
| 02:20 | <MikeSmith> | you know, tell us what he really thinks |
| 02:20 | <MikeSmith> | instead of beating around the bush |
| 02:22 | <AryehGregor> | Maybe you should comment on the blog to advise him. |
| 02:22 | <Dashiva> | If it wasn't for the point espoused, I could have been convinced it was squirrel writing the comment |
| 02:22 | <AryehGregor> | Yes, the tone is very much like him. |
| 02:22 | <Dashiva> | But I don't think he would sock puppet himself that way... |
| 02:22 | <AryehGregor> | People tend to pick up the tone of the sites they post on, I notice. |
| 02:23 | <AryehGregor> | If you go to the discussion forums for Dinosaur Comics, for example, you'll find everyone writing in run-on sentences and spouting non sequiturs. |
| 02:23 | <MikeSmith> | Mattur is definitely no sock puppet |
| 02:24 | <MikeSmith> | I think he does some self-sock-puppetry occasionally |
| 02:24 | <MikeSmith> | But I do that too |
| 02:24 | <MikeSmith> | in the privacy of my own home, of course |
| 02:24 | <MikeSmith> | but hey, we all do that, right? |
| 02:24 | <Dashiva> | I may have inadvertently sock puppeted myself with my second IRC client |
| 02:24 | <AryehGregor> | . . . what? |
| 02:25 | <Dashiva> | I don't think anyone in here know who it is |
| 02:25 | <MikeSmith> | I like the Squirrel too |
| 02:25 | <MikeSmith> | that Beavers Taste Great with Mustard stuff is solid gold |
| 02:25 | <karlcow> | Dali Lama is the extravaganza of buddhism |
| 02:25 | <MikeSmith> | he needs to give us more stuff like that |
| 02:26 | <MikeSmith> | and more violence, and more nudity |
| 02:26 | <MikeSmith> | if not outright intercourse |
| 02:26 | <Dashiva> | He makes good comics, and he serves a purpose, I just wish his blog posts would stick to facts and rational criticism |
| 02:27 | <MikeSmith> | as Karl Marx once said, there is nothing funnier in the entire world than depictions of cartoon characters have sex with each other |
| 02:27 | <Dashiva> | I don't think he said that |
| 02:27 | <MikeSmith> | oh, I am pretty sure he did |
| 02:27 | <MikeSmith> | it was either him or Albert Einstein |
| 02:27 | <Dashiva> | Unless by Karl Marx you mean Mark Twain and Oscar Wilde (co-authored) |
| 02:27 | <MikeSmith> | I think the Dalai Lama even quoted that one |
| 02:28 | <MikeSmith> | Karl Marx said Huckleberry Finn is his favorite book of all time |
| 02:28 | <MikeSmith> | and he also said Oscar Wilde is crap |
| 02:28 | <Dashiva> | Is the Dalai Lama allowed to find things funny, or does that ruin his chances for nirvana? |
| 02:29 | <MikeSmith> | and Marx said that Nabokov is "even more crap" |
| 02:29 | <karlcow> | Dashiva: rationalism… comics… hmmm I'm worried about mental health of Dashiva now. do you have Specification orgasm? :p |
| 02:29 | <MikeSmith> | Dalai Lama is a punk, seriously |
| 02:29 | <MikeSmith> | or a chump |
| 02:29 | <MikeSmith> | or a jackass |
| 02:30 | <Dashiva> | I think MikeSmith had a few too many beer breaks today |
| 02:30 | <othermaciej> | most exciting thing about those comments: reading Shelley's confident predictions of what Apple would do in particular situations |
| 02:30 | <Dashiva> | othermaciej: Or the comment about all the browser companies leaving WHATWG |
| 02:31 | MikeSmith | struggles to find a good other term that accurately captures the nature of the petty fraud that is the Dalai Lama's entire life |
| 02:31 | <karlcow> | hehe |
| 02:31 | <Dashiva> | MikeSmith: Sort of like every other holy man and woman after a little close examination? |
| 02:31 | <MikeSmith> | no, no |
| 02:32 | <karlcow> | MikeSmith: what about Dalai Lamarketing |
| 02:32 | <MikeSmith> | some holy men are genuine |
| 02:32 | <MikeSmith> | like Pope Ratzinger |
| 02:32 | <MikeSmith> | he will be a Saint one day |
| 02:32 | <Dashiva> | His excuses are full of holes, yes |
| 02:32 | <MikeSmith> | Saint Ratzinger |
| 02:32 | <MikeSmith> | Dashiva: beer bong |
| 02:32 | <MikeSmith> | not beer |
| 02:32 | <MikeSmith> | today |
| 02:32 | <MikeSmith> | http://www.hecklerspray.com/owen-wilson-vs-the-dalai-lama/2005686.php |
| 02:33 | <MikeSmith> | "corny purveyor of goofball ideas |
| 02:33 | <Dashiva> | karlcow: Also, I'm just acting as MikeSmith's foil, I don't think it's fair to make judgements about my own state |
| 02:33 | <MikeSmith> | that's another site filled with great subtlety of expression, btw |
| 02:33 | <karlcow> | Ratzinger could fit for a beer name |
| 02:34 | <MikeSmith> | Pope Ratzinger Beer |
| 02:34 | <MikeSmith> | I would buy it |
| 02:34 | <Dashiva> | Is there a beer you wouldn't buy? |
| 02:34 | <Dashiva> | Except, like, Asahi |
| 02:35 | <MikeSmith> | Ringnes |
| 02:36 | <Dashiva> | Nice google-fu |
| 02:36 | <karlcow> | There is a beer I would not buy. Budweiser, though it doesn't really qualify as a beer |
| 02:36 | <Dashiva> | Yeah, it was a trick question |
| 02:37 | <MikeSmith> | othermaciej: are you suggesting that Apple doesn't make critical technology decisions based on confidently stated third-party assertions found in blog comments? |
| 02:38 | <Dashiva> | Apple doesn't make forward-looking statements, so he can't answer your question |
| 02:38 | <othermaciej> | ^ this |
| 02:39 | <karlcow> | Apple doesn't decide. Steve does. |
| 02:39 | karlcow | notes he has not drank, just ate chocolate :p |
| 02:40 | <MikeSmith> | Dashiva: no google-fu involved.. I have experienced the wonder that is Ringnes firsthand... had my life's fill of Ringnes, and then some |
| 02:41 | <MikeSmith> | Norway needs to do something about its affordable-beer situation |
| 02:41 | <bros> | Should /query Hixie Hello |
| 02:42 | <karlcow> | Ringnes NOR way |
| 02:46 | <myakura> | uh oh http://fantasai.inkedblade.net/style/specs/css2.1/px-unit |
| 02:48 | <myakura> | it doesn't look right in html5-enabled-Gecko, </ins> doesn't close the <p>. |
| 02:49 | <MikeSmith> | myakura: which part is rendering wrong? |
| 02:49 | <MikeSmith> | it seems to look correct in my Minefield |
| 02:50 | <myakura> | Chrome 6 (dev) doesn't close either, while Opera 10.60 beta and IE8 does. |
| 02:50 | <kbrosnan> | should have a lot less green |
| 02:50 | <MikeSmith> | or at least it looks the same as it does in other browsers |
| 02:50 | <MikeSmith> | oh |
| 02:50 | <myakura> | MikeSmith: lines 45-47. <ins><p>In cases where the <a href="cascade.html#usedValue">used</a> length cannot be supported, user agents must approximate it in the <a href="cascade.html#actual-value">actual value.</a></ins> |
| 02:50 | <myakura> | oops |
| 02:51 | <MikeSmith> | yeah, I see now |
| 02:56 | AryehGregor | needs to block tvtropes.org at the firewall or something to prevent himself from gratuitously staying up late and wasting huge amounts of time. |
| 02:56 | <AryehGregor> | (although the "Jews in Space" skit it linked to was possibly awesome enough to be worth it this time) |
| 02:56 | <kbrosnan> | livedom.validator.nu gives the same dom tree as Firefox trunk |
| 03:52 | <MikeSmith> | man, I really am slow on the uptake |
| 03:52 | <MikeSmith> | I am just now realizing that annotation-xml content is, well, annotations |
| 03:53 | <MikeSmith> | with the implication that it's not displayed in the actual body text of a document |
| 03:53 | <MikeSmith> | but made available in some other way |
| 03:53 | <MikeSmith> | I guess |
| 03:53 | <MikeSmith> | a pop-up |
| 03:53 | <MikeSmith> | or whatever |
| 03:54 | <MikeSmith> | at least I think that's the actual intent |
| 03:54 | <MikeSmith> | based on seeing the rule that Simon cited in http://www.w3.org/TR/2009/CR-mathml-for-css-20091215/#d1e2465 |
| 03:55 | <MikeSmith> | default UA stylesheet |
| 03:55 | <MikeSmith> | annotation, annotation-xml |
| 03:55 | <MikeSmith> | {display:none;} |
| 03:55 | <MikeSmith> | which Opera apparently conforms to |
| 03:55 | <MikeSmith> | but Firefox does not seem to |
| 04:06 | <MikeSmith> | ah geez |
| 04:07 | <MikeSmith> | Opera Dragonfly apparently does not let you edit CSS from "local user stylesheets" |
| 04:07 | <MikeSmith> | of which I guess the MathML stylesheet it uses is one |
| 04:09 | MikeSmith | finds ./Contents/Resources/styles/mathml.css |
| 04:13 | <boblet> | re: earlier conversation about IDNs silently failing <input type="email"> validation (blocking form submission with zero user feedback) http://pukupi.com/post/2070/ Ouchies |
| 04:13 | <boblet> | Opera bug filed |
| 04:19 | <MikeSmith> | after reading more, I find the whole <semantics><annotation-xml> markup in MathML to be seriously overloaded |
| 04:20 | <MikeSmith> | annotations can either be "content" annotations (=non-presentational) or presentational annotations |
| 04:20 | <MikeSmith> | and it is not clear at all from the spec under what circumstances (if any) there are meant to be presented to the user by default |
| 04:21 | <MikeSmith> | and it is not clear at all from the spec under what circumstances (if any) there are meant to be presented to the user by default." |
| 04:22 | <MikeSmith> | oops |
| 04:22 | <MikeSmith> | "A particularly important case is the use of a presentation MathML expression to indicate a preferred rendering for a content MathML expression." |
| 04:22 | <MikeSmith> | "Note that when a presentation MathML annotation is present in a semantics element, it may be used as the default rendering of the semantics element, instead of the default rendering of the first child." |
| 04:23 | <MikeSmith> | ..."may be used as the default rendering" |
| 04:24 | <karlcow> | my house is a computer code. No tag soup there. - Sim City in Ixtapaluca, Mexico - http://hydeparkblvd.com/?p=1494 |
| 04:26 | <MikeSmithX> | "may be used as the default rendering" = "might also not be used as the default rendering, it's up to you!" |
| 06:13 | <kennyluck> | Hmm... MikeSmith told me that the browser vendors are not that interested in W3C Device API (http://www.w3.org/2009/dap/ ) because of the policy framework stuff (XACML?). But are there any plans for the browser vendors on implementation things like the Contact API or Calendar API ? |
| 06:55 | <boblet> | is there anything in the spec on what UAs should do when an input fails type validation? |
| 06:56 | <Hixie> | yes |
| 06:57 | <boblet> | hey Hixie — following checkedness link now, I may be getting close :) |
| 07:03 | <MikeSmith> | kennyluck: Mozilla has some interest in the Contacts API |
| 07:04 | <MikeSmith> | iirc |
| 07:07 | <MikeSmith> | kennyluck: http://lists.w3.org/Archives/Public/public-device-apis/2010Jun/0147.html |
| 07:07 | <boblet> | Hixie: atm IDN values in type="email" cause silent submission failure in all supporting browsers. any suggestions apart from using type="text"? i.e. should I file a bug to request spec say if input fails validation this should be communicated to the user? |
| 07:07 | <MikeSmith> | [[ |
| 07:07 | <MikeSmith> | Where |
| 07:07 | <MikeSmith> | specifications worked on in the DAP WG lend themselves to implementation |
| 07:07 | <MikeSmith> | plans, I think Mozilla participants interested in these can comment on |
| 07:07 | <MikeSmith> | them (e.g. Contacts API, at least for now). |
| 07:07 | <MikeSmith> | ]] |
| 07:08 | <boblet> | or is there a way to override per-input validation (or browser default chrome for eg date picker)? |
| 07:19 | <MikeSmith> | boblet: is there not an attribute for that? |
| 07:19 | <MikeSmith> | "novalidate" or whatever |
| 07:19 | <boblet> | MikeSmith: seems to be for form |
| 07:19 | <boblet> | let me doublecheck |
| 07:20 | <boblet> | oh nice, yeah it’s also an attribute on input |
| 07:20 | <boblet> | doh |
| 07:20 | boblet | must skim less |
| 07:21 | <MikeSmith> | only on input/type=submit, right? |
| 07:23 | <boblet> | aah yeah, checking the table we can only use formnovalidate on submit (and novalidate on form) |
| 07:23 | <boblet> | so all or nothing |
| 07:29 | <kennyluck> | Thanks, MikeSmith. I was trying to find a relevant bug in bugzilla@Mozilla but found nothing related. |
| 07:30 | <MikeSmith> | kennyluck: see also http://lists.w3.org/Archives/Public/public-webapps/2010AprJun/0909.html |
| 07:30 | <MikeSmith> | though that was qualified a bit further in http://lists.w3.org/Archives/Public/public-webapps/2010AprJun/0911.html |
| 07:31 | <MikeSmith> | not that I necessarily agree with any of those assessments |
| 07:31 | <MikeSmith> | but just to give you an sense of some of the discussion that's been taking place |
| 07:32 | kennyluck | is reading |
| 07:33 | <MikeSmith> | kennyluck: reading through the whole thread at http://lists.w3.org/Archives/Public/public-device-apis/2010Jun/thread.html#msg128 would not be a waste of time |
| 07:33 | <MikeSmith> | in particular the messages from Arun and Brian |
| 07:34 | <kennyluck> | Yeah, I am actually following those "in-reply-to" because I am a bit confused. |
| 07:56 | <Hixie> | boblet: do you mean non-ASCII text input or non-ASCII value=""s? and why is it silent? |
| 07:58 | <boblet> | non-ASCII text input (eg example@日本語.jp), and apparently there’s no indication to the user that this fails validation, and that in doing so the form can’t be submitted. |
| 07:59 | <boblet> | this has been a problem for a long time due to \w in regex treating double-byte characters as if they had a space on each side (again apparently) |
| 08:02 | <Hixie> | sounds like a lot of bugs |
| 08:02 | <Hixie> | file them :-) |
| 08:04 | <Hixie> | looks like the spec doesn't mention it, but really if the user types in a unicode e-mail address it should get converted to punycode |
| 08:06 | <boblet> | Hixie: heh http://code.google.com/p/chromium/issues/detail?id=45804 by AryehGregor and https://bugs.webkit.org/show_bug.cgi?id=40761 by me |
| 08:08 | <boblet> | the word is using formnovalidate gets around this, so I’m planning to use that with type="email" for iPhone keyboard (mobile-targetted site) |
| 08:09 | <boblet> | I just wonder given how broken this is if adding a note about error handling to the spec would be a good idea, ref: http://krijnhoetmer.nl/irc-logs/whatwg/20100617#l-61 |
| 08:10 | <boblet> | focussing the form field of the first validation error is not equal to providing a user UI to this |
| 08:10 | <Hixie> | agreed |
| 08:11 | <Hixie> | feel free to file a bug on the spec or send mail to the mailing list if you think the spec needs updating a bit |
| 08:11 | <Hixie> | though i'm way behind on dealing with feedback |
| 08:12 | <boblet> | rgr, will do |
| 08:27 | <MikeSmith> | Hixie: would it be possible to have the parsing algorithm just read the contents of annotation-xml elements into the DOM but do any processing on them at all? |
| 08:27 | <MikeSmith> | that is, even if those contents are SVG |
| 08:27 | <MikeSmith> | or even if the contents are MathML |
| 08:28 | <MikeSmith> | *into the DOM but _not_ do any processing on them at all |
| 08:29 | <MikeSmith> | for one thing, those contents are, after all, annotations that are not intended to be rendered in the main text flow |
| 08:30 | <MikeSmith> | both Gecko and Presto have default UA stylesheets with "annotation-xml { display: none }" |
| 08:31 | <othermaciej> | what is annotation-xml? |
| 08:31 | <MikeSmith> | and the MathML spec says: "The default rendering of a semantics element is the default rendering of its first child" |
| 08:31 | <MikeSmith> | where the first child is <mn> or whatever |
| 08:32 | <MikeSmith> | and the second or subsequent children are annotation-xml elements |
| 08:32 | <MikeSmith> | othermaciej: it's a mechanism in MathML for providing an "alternate representation" of MathML content |
| 08:33 | <Hixie> | MikeSmith: "processing"? |
| 08:33 | <MikeSmith> | http://www.w3.org/TR/MathML3/chapter5.html#mixing.alternate.representations |
| 08:33 | <MikeSmith> | Hixie: not render it as whatever it would otherwise render it as |
| 08:33 | <MikeSmith> | even if it's SVG, don't render it as such |
| 08:33 | <Hixie> | aren't they display:none? |
| 08:33 | <MikeSmith> | yeah |
| 08:33 | <Hixie> | by default? |
| 08:33 | <MikeSmith> | by default |
| 08:33 | <MikeSmith> | yeah |
| 08:34 | <Hixie> | so...? |
| 08:34 | <othermaciej> | MikeSmith: does MathML call for it to be display: none, or is that convergent evolution among Gecko and Presto? |
| 08:34 | <MikeSmith> | so, that's overrideable |
| 08:34 | <Hixie> | isn't that enough? |
| 08:34 | <othermaciej> | if annotation-xml is itself display: none, nothing inside it will render by default |
| 08:35 | <MikeSmith> | Hixie: ewhat if a site stylesheet sets it to actually display? |
| 08:36 | <Hixie> | then we'd want it to display, no? |
| 08:36 | <Hixie> | why else would it be set to display? |
| 08:38 | <MikeSmith> | OK, the problem currently is that the parsing algorithm doesn't allow you to put HTML content into <annotation-xml> elements |
| 08:38 | <MikeSmith> | but I think users will expect to be able to put HTML content there |
| 08:39 | <MikeSmith> | just as they can put SVG and MathML |
| 08:39 | <Hixie> | not sure how to allow it without hard-coding a list of elements |
| 08:39 | <MikeSmith> | yeah |
| 08:40 | <MikeSmith> | it could be allowed by allowing it |
| 08:41 | <Hixie> | annotation-xml didn't seem important enough to hard-code a lot of elements there |
| 08:41 | <MikeSmith> | and instead treating SVG and MathML elements in <annotation-xml> as arbitrary unknown elements |
| 08:41 | <Hixie> | it'd be a lot of complexity in every parser for a very obscure case |
| 08:41 | <MikeSmith> | Hixie: I fully agree it's not important |
| 08:42 | <Hixie> | i supported svg since that seemed useful and easy to support |
| 08:42 | <Hixie> | some mathml examples do it |
| 08:42 | <MikeSmith> | yeah, I know the examples do |
| 08:43 | <MikeSmith> | but I don't know if there are any real-world use cases for it in Web content |
| 08:43 | <MikeSmith> | other than those examples |
| 09:10 | <MikeSmith> | Hixie: another (hopefully easy to answer question) - |
| 09:10 | <MikeSmith> | the MathML spec current says this: "The draft HTML5 Recommendation, [HTML5], contains details for how MathML is included in HTML5 (and XHTML5)." |
| 09:11 | <Hixie> | not fully true yet but it will be |
| 09:11 | <MikeSmith> | well, I think it could be refined |
| 09:12 | <MikeSmith> | to say this: |
| 09:12 | <MikeSmith> | > > The draft HTML5 Recommendation, [HTML5], contains details for |
| 09:12 | <MikeSmith> | > > how MathML is included in HTML5 (and XHTML5), and, in turn, |
| 09:12 | <MikeSmith> | > > details on where HTML content is allowed to be included in |
| 09:12 | <MikeSmith> | > > MathML content within HTML documents. |
| 09:12 | <MikeSmith> | to make it more clear that doesn't just say how MathML can be included in HTML5 |
| 09:13 | <Hixie> | wfm |
| 09:13 | <MikeSmith> | OK |
| 09:13 | <MikeSmith> | no suggestions for tweaking it any further? |
| 09:13 | <Hixie> | looks good to me |
| 09:13 | <MikeSmith> | OK, thanks |
| 09:14 | <othermaciej> | MikeSmith: it seems inaccurate to refer to the HTML5 Working Draft as "The draft HTML5 Recommendation" |
| 09:14 | <othermaciej> | or at least, imprecise |
| 09:14 | <Hixie> | draft HTML5 standard? :-) |
| 09:15 | <othermaciej> | I forget, does the W3C make standards this week? |
| 09:15 | <MikeSmith> | the W3C makes technologies |
| 09:15 | <MikeSmith> | normal people don't care about standards |
| 09:16 | <MikeSmith> | they do care about new technologies |
| 09:16 | <MikeSmith> | shiny stuff |
| 09:16 | <asmodai> | One thing raised on the Unicode list is that right now it is hard for i18n work to know which character ranges a particular UA supports. That way if someone lacks a particular font you could use some way (JavaScript?) to supply missing support via font-face. There's no plans for any API or detection way for supported scripts I guess? |
| 09:16 | <Hixie> | http://www.whatwg.org/specs/web-apps/current-work/ says "Draft Standard" |
| 09:17 | <MikeSmith> | othermaciej, Hixie - I don't see why they could not just refer to there simply as "HTML5" or as "the HTML5 specification" |
| 09:17 | <MikeSmith> | right? |
| 09:17 | <MikeSmith> | I can ask for a change to that as well |
| 09:17 | <MikeSmith> | they are in LC with the spec now |
| 09:18 | <othermaciej> | I think both of those would be accurate ways to refer to HTML5, and less likely to be invalidated over time |
| 09:18 | <MikeSmith> | ok |
| 09:18 | <Hixie> | yeah |
| 10:07 | <zcorpan_> | http://www.w3.org/mid/6797BD1D-BA7A-440D-B6D6-4D0162E66883⊙gc here, roy is making a whole lot of sense |
| 10:12 | <zcorpan_> | "and that an <img> without a @role attribute is assumed to be the equivalent of <IMG @role="img"> (and would follow the rules in #1 above)" - http://www.w3.org/2009/06/Text-Alternatives-in-HTML5 |
| 10:12 | <zcorpan_> | this doesn't make any sense given the definition of role="img" (which is closer to <figure>) |
| 10:13 | <Hixie> | holy cow, roy agreed with something html5 says? |
| 10:14 | <zcorpan_> | annevk: http://daringfireball.net/2007/12/fastcompany has just one sentence in the heading afaict |
| 10:14 | <annevk> | lol |
| 10:14 | <annevk> | fail |
| 10:15 | <annevk> | fortunately james example had plenty |
| 10:16 | <zcorpan_> | TR/ versions should have a floating message about the snapshotness, not just a paragraph in the SotD which engineers don't read |
| 10:17 | <annevk> | yeah, but that would need more than HTML WG buy in |
| 10:18 | <zcorpan_> | or the image in the margin should say "THIS DOCUMENT IS OBSOLETE DON'T READ IT" |
| 10:18 | <Hixie> | "Is there some kind of multiple-sentence-header abuse you are trying to combat?" lol |
| 10:19 | <Hixie> | zcorpan_: the paragraph right now isn't even in the SotD, it's in the History section |
| 10:19 | <zcorpan_> | Hixie: engineers don't read that either |
| 10:19 | <Hixie> | i guess we could make it more obvious |
| 10:20 | <zcorpan_> | Hixie: the way they end up in TR is they google on what they're looking for |
| 10:20 | <Hixie> | true |
| 10:20 | <zcorpan_> | maybe the solution is to change google's algorithms |
| 10:21 | <zcorpan_> | or change the links in wikipedia from TR/ to dev |
| 10:27 | <hsivonen> | this bug is awesome: https://bugzilla.mozilla.org/show_bug.cgi?id=515171 |
| 10:27 | <hsivonen> | see comment #6 |
| 10:33 | zcorpan_ | changes his language to 'DROP TABLE *;' |
| 10:33 | <Hixie> | zcorpan_: well i looked into what it would take to add a floating warning on TR/ drafts, and it's doable pretty easily, but if i do it just like that i'll have half the wg on my ass |
| 10:33 | <Hixie> | zcorpan_: so if you want it please try to shore up some support in the wg first |
| 10:38 | <annevk> | should I file a bug on adding a magic square example to the spec? |
| 10:38 | <annevk> | not sure whether it's worth it, but they're kind of cool |
| 10:38 | <Hixie> | sure |
| 10:39 | <Hixie> | worse that can happen is i reject it and someone raises an issue for it |
| 10:40 | <jgraham> | Or a sudoku puzzle |
| 10:40 | <jgraham> | I hate them but it would be kinda cool to have a puzzle in the spec :) |
| 10:41 | <Hixie> | better idea than a magic square, file a bug :-) |
| 10:41 | <annevk> | nice one |
| 10:41 | annevk | files |
| 10:43 | <annevk> | http://www.w3.org/Bugs/Public/show_bug.cgi?id=9943 |
| 10:45 | <zcorpan_> | the puzzle should be interactive |
| 10:45 | <zcorpan_> | let's also add pacman to the canvas section |
| 10:46 | <Hixie> | so long as you get permission from the pacman rights owners... |
| 10:46 | <annevk> | and was Google pacman even done using <canvas>? |
| 10:46 | <annevk> | it worked in IE, no? |
| 10:47 | <Hixie> | that was a <div> disaster |
| 10:47 | <Hixie> | :-) |
| 10:47 | <Hixie> | http://www.google.com/pacman/ btw |
| 10:48 | <annevk> | yeah I know, it's fun |
| 10:53 | <Hixie> | so did anyone ever get around to giving some URLs of example pages that are actually helped by this role=button stuff? |
| 10:55 | <MikeSmith> | hsivonen: "For modsecurity 2.1.7 with core rules ver 1.5.1, workaround is to remove string “|id” (yes, remove that pipe character also) on the regular expression of rule 950006 in file modsecurity_crs_40_generic_attacks.conf" |
| 10:55 | <MikeSmith> | I guess they got a little carried away there |
| 10:56 | <zcorpan_> | annevk: a script isn't run if it's inserted but doesn't have any childnodes and doesn't have src set, iirc |
| 10:57 | <Dashiva> | http://www.cssquirrel.com/2010/06/15/693comic-update-html5-unicorn-heuristics/#comment-31995 |
| 10:58 | <zcorpan_> | annevk: step 5 |
| 10:58 | <annevk> | ah |
| 10:59 | <annevk> | resolved |
| 11:02 | <hsivonen> | MikeSmith: at leaast it's not a default feature of apache |
| 11:03 | <zcorpan_> | http://twitter.com/opera/statuses/16372312957 lol |
| 11:06 | <annevk> | haha |
| 11:06 | <Lachy> | haha, isn't it great when April Fools jokes just keep fooling people long after April Fools? :-) |
| 11:07 | <Dashiva> | Unless they know something you don't! |
| 11:08 | <Dashiva> | I've been wondering why the latest Opera builds seem to activate my webcam at random times... |
| 11:09 | <annevk> | that's part of our world dom^H^H^H^H... |
| 11:31 | <hsivonen> | hmm. Opera still sends the U; token in the UA string. I wonder if that's really needed for Web compat |
| 11:31 | <hsivonen> | IE doesn't send it |
| 11:38 | <karlcow> | opera doesn't support face gestures… no. shame on you ;) |
| 11:39 | <annevk> | jaja |
| 11:46 | <Lachy> | what was the U; token for originally? |
| 12:01 | <hsivonen> | Lachy: distinguishing between U.S., French and other versions of Netscape back when the U.S. and France had worse crypto regulations than they have today |
| 12:02 | <hsivonen> | U is U.S. ("strong" crypto), N is France (no crypto) and I is international (weak crypto) |
| 12:08 | <hsivonen> | Would anyone care to load http://www.delorie.com:81/some/url.txt in a Japanese version of IE8 and pastebin the result? |
| 12:09 | <hsivonen> | any non-English version of IE would be interesting, too, but not quite as interesting as Japanese |
| 12:09 | hsivonen | only has English IEs around |
| 12:09 | <hsivonen> | and non-Ultimate Windows, so I can't switch languages |
| 12:10 | <daedb> | hsivonen: http://pastebin.com/YUtFfz2r <-- swedish IE8 |
| 12:11 | <hsivonen> | daedb: thanks! |
| 12:12 | <hsivonen> | confirms my hypothesis that IE doesn't put the UI language in the UA string |
| 12:27 | <Lachy> | why doesn't windows let you switch languages if you don't have Windows Ultimate?! |
| 12:28 | <Lachy> | that seems like such a basic feature, and so unnecessary to force people to purchase the Ultimate package for. |
| 12:30 | <gsnedders> | It's of use to a small number of people, and you get more profit if you force people to do so |
| 12:36 | <Lachy> | But it would be cheaper for them if they only had to produce 1 master copy of the discs to distribute globally, instead of having to then produce individual masters for each language in which Windows is distributed. |
| 12:37 | <Lachy> | and it would make it easier for foreigners to buy a copy that supports their own native language while in a foreign country |
| 12:37 | hsivonen | expects MS is competent at calculating the cost of tiered versions against the profit extracted by exploiting different points on the demand curve |
| 12:38 | <gsnedders> | Lachy: AFAIK they produce a single master disk, it's just it's locked by license key at install time |
| 12:39 | <Lachy> | so then each disc actually has all of the langauges on there anyway, and it's just their greed that prevents people from taking advantage? |
| 12:39 | <hsivonen> | I specifically ordered my Windows 7 Professional from Amazon.co.uk to get more googlable error messages |
| 12:39 | <hsivonen> | for some reason, the English version was priced unusually high compared to Finnish and Swedish versions in the local stores I usually buy from |
| 12:40 | <Hixie> | i downloaded ubuntu to get fewer error messages overall |
| 12:40 | <hsivonen> | the UK package allowed me to install and activate en-US, though |
| 12:40 | <Lachy> | fair enough. I might have to do the same, since I definitely don't want norwegian if and when I get a copy. I had previously assumed I would just be able to select English as install time, like I can with OSX |
| 12:40 | <hsivonen> | Hixie: I run it on top of Ubuntu |
| 12:40 | <Hixie> | :-P |
| 12:42 | <hsivonen> | however, the only times Windows 7 has done an ungraceful shutdown so far have been when X on Ubuntu has died due to USB devices getting plugged in... |
| 12:53 | <Lachy> | hsivonen, other than the language issue, does windows 7 professional generally have all the necessary features? Is there anything in Ultimate you wish you had? |
| 12:56 | <Lachy> | the compaison only shows bitlocker as the other difference, and I probably don't need that. |
| 13:02 | <hsivonen> | Lachy: The only other difference I am aware of is Bitlocker, which I don't need, since I only use Windows 7 for browser testing |
| 13:02 | <hsivonen> | and I assume Windows is untrusted, so I don't put my site login credentials there |
| 13:10 | <Lachy> | ok, in that case, I might got for Professional instead of Ultimate too, when I decide to get a copy soon. |
| 13:33 | <hsivonen> | Lachy: note: I've had Windows 7 for a day now |
| 13:43 | <Lachy> | ah, ok. |
| 14:16 | <kamathln> | Is there anything in HTML5 that gives AJAX like behaviour without script |
| 14:16 | <kamathln> | scripting* |
| 14:17 | <kamathln> | Something that allows the server side to tell the otherside "Replace only X part of the page, with Y item" |
| 14:19 | <zcorpan_> | kamathln: i guess that would be <iframe> and <a target> |
| 14:19 | <kamathln> | :D |
| 14:20 | <kamathln> | I wanted the server side to decide on the target |
| 14:20 | <kamathln> | I think thats not possible with iframe and target . |
| 14:21 | <zcorpan_> | it can if you're ok with it deciding up front when the first page loads :) |
| 14:22 | <kamathln> | zcorpan_: didnt catch you |
| 14:22 | <annevk> | there was a feature like that but it was dropped |
| 14:22 | <annevk> | though it did not quite do what you want |
| 14:23 | <kamathln> | annevk: Oh! any hint where I can look at the discussion, and any particular term they used that I can search for? |
| 14:23 | <kamathln> | Ok |
| 14:23 | <zcorpan_> | why do you want to do it without js? |
| 14:24 | <kamathln> | zcorpan_: I want it to work on devices/browsers that dont depend on js |
| 14:25 | <zcorpan_> | what is it you're trying to do? |
| 14:25 | <kamathln> | Some people also have Javascript disabled |
| 14:26 | <kamathln> | zcorpan_: Nothing at the moment .. But I have always found this as a major hinderence. That just disabling scriipting Simply makes the whole Server side programs a lot heavier |
| 14:29 | <kamathln> | Disabling javascript makes browsing a tad bit secure. |
| 14:29 | <kamathln> | Well, I think I will just keep quiet for a while till I gather the right words |
| 14:30 | <annevk> | there has been discussion on a feature that's closer to what you want on the WHATWG list |
| 14:31 | <kamathln> | again, any term I can search for |
| 14:31 | <kamathln> | or a date range |
| 14:31 | <kamathln> | ? |
| 14:31 | <annevk> | I forgot the terminology :/ |
| 14:31 | zcorpan_ | also wonders what annevk is talking about |
| 14:31 | <daedb> | onlyreplace? |
| 14:31 | <zcorpan_> | oh |
| 14:32 | zcorpan_ | didn't read that thread so no wonder he doesn't recall it |
| 14:32 | <kamathln> | http://lists.whatwg.org/htdig.cgi/whatwg-whatwg.org/2009-October/023661.html |
| 14:33 | <annevk> | http://lists.whatwg.org/pipermail/whatwg-whatwg.org/2007-April/010801.html |
| 14:33 | <kamathln> | whoa! back in 2007! |
| 14:34 | <annevk> | nothing is novel |
| 14:36 | <annevk> | the proposal that was in a spec at some point was form pre-filling |
| 14:36 | <annevk> | material from 2004 or so |
| 14:36 | <kamathln> | All you guys were right. This comes close enough. But my point makes one but big change .. In my idea target(s) are to be decided by the server |
| 14:36 | <annevk> | which in turn is based on some XForms stuff |
| 14:38 | <hsivonen> | whoa. scary. I got an automated prompt to install Flash when going to msn.com in IE8. |
| 14:39 | <hsivonen> | I wonder how many people click Yes to random prompts like that. |
| 14:39 | <zcorpan_> | hsivonen: 50% |
| 14:40 | <hsivonen> | seriously, it would be better for the security behavior of users to bundle Flash than to make the default home page autoprompt the install |
| 14:42 | <hsivonen> | wow. IE is hard to use. |
| 14:45 | <hsivonen> | oh great. after a day of using Windows, Minefield already has NPAPI plug-ins installed into it by 3 other browser vendors without explicitly telling me |
| 14:46 | <hsivonen> | kudos to Opera for not injecting native code into my Gecko-based browser |
| 14:49 | <Lachy> | oh, I hate when other programs try to install firefox extensions without asking. |
| 14:50 | <Lachy> | I uninstall them immediately. May require registry hacking. |
| 14:50 | <Lachy> | but what extensions did Chrome and Safari install? |
| 14:51 | <hsivonen> | (for those wondering what I'm talking about: Installing Chrome also installs a Google Update plug-in for Firefox/Minefield, installing a fully functional version of Safari also installs QuickTime Plug-in for Firefox/Minefield and to install legacy IE as part of XP mode, you have to go through Genuine Validation which apparently sticks a Windows Activation Technologies plug-in into Firefox/Minefield) |
| 14:51 | <Lachy> | is that true if you get the standalone Safari installer, instead of the Safari+QuickTime installer? |
| 14:52 | <hsivonen> | Lachy: I assume not, but I said "fully functional" |
| 14:52 | <hsivonen> | Lachy: I gather Safari 4 didn't support <video> without QuickTime |
| 14:53 | <Lachy> | if I install quicktime, then I expect it to install the NPAPI compatible plugin in the appropriate location for use by other browsers |
| 14:53 | <hsivonen> | I didn't test if Safari 5 can fall back to MS-provided decoders |
| 14:54 | <hsivonen> | Lachy: well, the other plug-ins can sorta be explained to be legitimate, but what is Google Update doing in Firefox after I install Chrome? |
| 14:54 | <Lachy> | no idea. |
| 14:55 | <daedb> | Actually, <video> does work without installing quicktime |
| 14:55 | <Dashiva> | Oh look, another YSOD, good thing XHTML is so easy |
| 14:55 | <hsivonen> | daedb: ooh. interesting! |
| 14:56 | <daedb> | I think, anyway |
| 14:56 | <daedb> | wait, no, I was wrong... quicktime is required :-/ |
| 14:57 | daedb | should remember to actually test things before speaking... |
| 15:02 | <zcorpan_> | hsivonen: it's there so google can investigate why people switch back to firefox |
| 15:04 | <hsivonen> | awesome. now Windows Activation Technologies crashed |
| 15:04 | <hsivonen> | thanks to out-of-process plug-ins, Minefield didn't crash |
| 15:04 | <fagan> | hsivonen: trying to use IE? |
| 15:05 | <hsivonen> | fagan: trying to install XP Mode via Minefield after failing to install it via IE8... |
| 15:05 | <fagan> | sounds hard |
| 15:05 | fagan | is a happy linux user |
| 15:06 | <hsivonen> | I'm a Linux user, too, though an unhappy one |
| 15:06 | <fagan> | hsivonen: why unhappy? |
| 15:06 | <fagan> | its getting better year on year |
| 15:06 | hsivonen | filed two customer dissatisfaction surveys on Canonical support case handling today |
| 15:07 | <hsivonen> | fagan: 1) I have situations that need support. 2) Support hasn't fully solved a single problem yet. |
| 15:07 | <fagan> | hsivonen: wow paid for support ill never do that :) |
| 15:07 | <fagan> | hsivonen: the best way I found is to ask on the forums sometimes its better than canonical support |
| 15:08 | <hsivonen> | support has partially solved one out of six so far. |
| 15:08 | <fagan> | im an Ubuntu community developer |
| 15:08 | <hsivonen> | fagan: before I purchased support, I tried Launchpad Questions |
| 15:08 | <fagan> | hsivonen: the forums are the place or IRC |
| 15:09 | <hsivonen> | fagan: I also tried the Finnish Ubuntu IRC channel first |
| 15:09 | <fagan> | launchpad answers is kinda dead |
| 15:09 | <hsivonen> | I guess I need to try the forums, then |
| 15:09 | <fagan> | hsivonen: must be a complicated problem |
| 15:10 | <fagan> | I wonder why the whatwg site is so basic |
| 15:11 | <fagan> | shouldnt you guys have an insane video/audio experience to show off what your doing? :D |
| 15:13 | <hsivonen> | I want to do radical stuff with Ubuntu, such as configure a button on my Wacom stylus (has GUI on Windows and Mac), make a bootable clone of the system (productized for Mac) or attach USB devices while the system is running |
| 15:13 | <hsivonen> | and distribution upgrade to completion |
| 15:14 | <Lachy> | fagan, because the WHATWG is more focussed on developing the spec, than on creating and publishing fancy demos. There are plenty of other sites doing that already |
| 15:15 | <Dashiva> | It's definitely a task worth doing, though |
| 15:15 | <Lachy> | sure, if there are volunteers |
| 15:19 | <hsivonen> | oh great. Windows 7 crashes when trying to setup the XP Mode |
| 15:19 | <hsivonen> | bah |
| 15:21 | <Lachy> | that's not a bug, that means XP mode is functioning normally ;-) |
| 15:21 | <hsivonen> | I wonder if VirtualBox kills the guest if the guest tries to run an VT-x instruction |
| 15:22 | <hsivonen> | so much broken stuff |
| 15:22 | hsivonen | heads to the gym instead of fighting with operating systems |
| 15:23 | <fagan> | Lachy: very true I suppose it takes some effort to do good tutorials and thats what w3schools is for |
| 15:24 | <Lachy> | fagan, try http://html5doctor.com/ |
| 15:25 | <Lachy> | I wouldn't trust W3 schools. They're tutorials are notoriously bad quality |
| 15:25 | <fagan> | Lachy: I actually have been reading the spec so I dont really need to look at tutorials :D |
| 15:25 | <fagan> | but I only subbed the mailing list today |
| 15:26 | <fagan> | I just popped in and had a look at it as it was being built to see what was going on |
| 15:26 | zcorpan_ | wonders if there's a way to make the material on w3schools be good |
| 15:27 | <fagan> | zcorpan_: well I found the best way to develop websites was to just look at other sites you like and try to make it look the way you want it |
| 15:28 | <fagan> | then use tidy to fix the mistakes |
| 15:29 | <fagan> | so for the most recent site I made I copied the look of apple.com ish and then changed the color scheme and images..etc so then it looked good |
| 15:29 | <fagan> | if w3schools broke down a few good websites to show people how to do it like a pro it would be a lot more helpful I think |
| 16:00 | <zcorpan_> | clearly shelley doesn't know safari 5 supports <aside> |
| 16:12 | <karlushi> | zcorpan_, since march 2010? https://bugs.webkit.org/show_bug.cgi?id=32943 |
| 16:12 | <karlushi> | http://en.wikipedia.org/wiki/Comparison_of_layout_engines_%28HTML5%29#Elements |
| 16:14 | <fagan> | karlushi: thats out of date I think for opera's engine |
| 16:14 | <fagan> | that has most of them done in the most recent version |
| 16:14 | <fagan> | or it should from what I remember |
| 16:14 | <gsnedders> | No, it doesn't |
| 16:15 | <gsnedders> | I think the only thing wrong in that section is canvas should be 2.4 |
| 16:16 | <karlushi> | hmmm |
| 16:16 | <karlushi> | a common format for publishing release notes would be cool. That would help to dynamically create implementation status report across browsers engines |
| 16:20 | <karlushi> | that would require a feature id (maybe the uri of the feature), version number of the engine, status: [bug|partial|ok|no], comment (optional), uri of bug report (optional) |
| 16:20 | <karlushi> | then people could publish their own report if they disagree with the report published by the company. |
| 16:22 | <Rik`> | would be cool to have release notes for Safari :( |
| 16:23 | <gsnedders> | When is something buggy and when is something not? |
| 16:25 | <karlushi> | gsnedders, to be defined :) maybe when there is an open bug about this feature. Though people would open open bug report just to change the status :p |
| 16:25 | <karlushi> | looking at the changelog of Opera 10.6 beta http://www.opera.com/docs/changelogs/mac/1060b1/ |
| 16:26 | <karlushi> | the level of granularity is a big issue too |
| 16:27 | <karlushi> | I see for example in opera 10.6 beta 1 |
| 16:27 | <karlushi> | "# Removing support for div nowrap" |
| 16:28 | <karlushi> | so something which was implemented in the past is not anymore. But in an implementation report it doesn't say, if because it was too buggy, or if the feature was dropped from the technology. |
| 17:35 | <jgraham> | I wonder what "officially objecting" is |
| 17:36 | <zcorpan_> | i wonder if daniel objects to all non-machine-checkable requirements |
| 17:36 | <jgraham> | Logically he has to |
| 17:37 | <jgraham> | Given his second sentence |
| 17:38 | <jgraham> | In other news, I loath the python unittest module |
| 17:38 | <jgraham> | *loathe |
| 17:42 | <hober> | zcorpan_: yeah, just asked him that on the list |
| 17:43 | <Dashiva> | Removing non-machine-checkable requirements, hwh |
| 19:28 | <gsnedders> | http://gsnedders.html5.org/outliner/process.py?url=http%3A%2F%2Fgsnedders.com |
| 19:28 | <gsnedders> | Looks like it's trying to use the wrong Python version's folder |
| 20:51 | <jgraham> | gsnedders: Seems to have charcter encoding issues |
| 20:52 | <gsnedders> | jgraham: I see |
| 21:36 | <jgraham> | Can a <select> token ever be processed as if In Body when the current node is a <colgroup> element? |
| 21:37 | <jgraham> | Or rather when in the In Column Group insertion mode |
| 21:41 | <zcorpan_> | jgraham: afaict, <select> in column group just implies </colgroup> (which switches to in table) and gets reprocessed |
| 21:41 | <zcorpan_> | jgraham: or ignored in the fragment case |
| 21:42 | <jgraham> | zcorpan_: That's what I think |
| 21:43 | <jgraham> | So the text about using In Select In Table if the insertion mode is In Column Group is redundant? |
| 21:44 | <zcorpan_> | which text? |
| 21:45 | <zcorpan_> | If the insertion mode is one of in table", "in caption", "in column group", "in table body", "in row", or "in cell", then switch the insertion mode to "in select in table". Otherwise, switch the insertion mode to "in select". |
| 21:46 | <zcorpan_> | yeah |
| 21:46 | <zcorpan_> | while you whine about that, there's a missing quote also :) |
| 21:52 | <jgraham> | Whined |
| 22:07 | <AryehGregor> | Now, how long until the YouTube HTML5 beta works on embedded videos? |
| 22:07 | <TabAtkins> | I'm counting the days. :\ |
| 22:07 | <AryehGregor> | Does that imply you know how many days are left? |
| 22:07 | <TabAtkins> | I wish it did. |
| 22:08 | <AryehGregor> | They'd have to change the embed code they provide, of course. Probably just make it an iframe or script tag. |
| 22:08 | <jgraham> | How high is your count? |
| 22:08 | <TabAtkins> | jgraham: Q. |
| 22:08 | <zcorpan_> | AryehGregor: would be cool if it could be made to work retroactively i.e. without changing the embed code |
| 22:09 | <Lachy> | there are some issues that I think need to be solved before before YouTube will start allowing 3rd party sites to embed HTML5 video |
| 22:09 | <AryehGregor> | The embed code is <object type="application/x-shockwave-flash">, so I don't see that working. |
| 22:09 | <AryehGregor> | Lachy, really? Like what? |
| 22:09 | <Lachy> | in particular, full screen support, more widespread support in browsers |
| 22:09 | <Lachy> | for webm |
| 22:09 | <zcorpan_> | AryehGregor: hmm, yeah |
| 22:09 | <Lachy> | in released products, not just betas |
| 22:10 | <AryehGregor> | I'm saying they could support the beta for embeds. |
| 22:10 | <AryehGregor> | I.e., use HTML5 if you're logged into YouTube and have opted in. |
| 22:10 | <AryehGregor> | Not for general deployment, of course, except as a fallback if Flash isn't installed. |
| 22:13 | <Lachy> | they could make it easier to deploy HTML5 later, if they changed the embed code from <embed> to a <script> that then dynamically generates the markup. |
| 22:14 | <Lachy> | that way, even if such a script simply embedded the current flash player, they can update the script later to generate code for HTML5 when that's supported |
| 22:14 | <othermaciej> | or they could use an iframe embed |
| 22:14 | <othermaciej> | that would let them conditionally serve flash or html |
| 22:14 | <Lachy> | yeah, that too. |
| 22:15 | <othermaciej> | without having to convince sites to let them inject a <script? |
| 22:15 | <othermaciej> | er, <script> |
| 22:15 | <othermaciej> | we at Apple have been recommending <iframe> embedding to video content providers who ask us |
| 22:20 | <AryehGregor> | Why would sites be upset by <script> but not mind <object>? |
| 22:20 | <TabAtkins> | Because people are dumb? |
| 22:20 | <AryehGregor> | <iframe> is certainly more secure than either, if you don't trust Google's JS. |
| 22:20 | <othermaciej> | I don't know of sites would mind <script> |
| 22:20 | <othermaciej> | apparently the way Web video embedding works now is... |
| 22:20 | <AryehGregor> | They use <script> for ads and analytics. |
| 22:21 | <othermaciej> | many of the popular providers have their player SWFs on a whitelist for embedding into posts / comments / user generated content / etc |
| 22:21 | <othermaciej> | like, popular blogging platforms, things like facebook, etc |
| 22:21 | <othermaciej> | if they come up with a new mechanism, it will take some time for them to get those whitelists updated |
| 22:22 | <othermaciej> | of course, an iframe is safer than <object> and maybe does not even need whitelisting but that might be a hard sell |
| 22:23 | <Lachy> | it would be nice if they also provided the <video> markup directly for authors to copy and paste, so that sites can apply their own custom controls for videos. |
| 22:24 | <AryehGregor> | I doubt they'll want to do that. It gives them much less control. |
| 22:24 | <jgraham> | So what am I missing: |
| 22:24 | <jgraham> | <!doctype html><body><rp><rp> |
| 22:24 | <jgraham> | er |
| 22:24 | <AryehGregor> | Actually, maybe part of the reason that they haven't done this is that the video URLs are obfuscated and they don't want to publicize the URL generation algorithm. |
| 22:24 | <jgraham> | <!doctype html><body><rp></rp> |
| 22:24 | <TabAtkins> | jgraham: <ruby>? |
| 22:24 | <jgraham> | stack of open elements is |
| 22:24 | <jgraham> | [html, body, rp] |
| 22:24 | <AryehGregor> | They probably don't want people hotlinking their videos without their scripts and so on being there as well. |
| 22:25 | <jgraham> | </rp> has no special processing |
| 22:25 | <AryehGregor> | Technically it's always been possible, but I guess <video> makes it a lot easier. |
| 22:25 | <jgraham> | So we do any other end tag |
| 22:25 | <jgraham> | The first step is to generate implied end tags, which removes the rp element from the stack of open elements |
| 22:26 | <jgraham> | this is bad |
| 22:26 | <jgraham> | Should it generate implied end tags excluding rp? |
| 22:26 | <jgraham> | (or whatever the current token name is) |
| 22:28 | <Lachy> | in theory, the optimal markup to offer to authors would be <video style="binding:url(http://example.com/player.xml)"><source ...>...</video>, where player.xml is an XBL template that dynamically creates the custom UI. |
| 22:28 | <Lachy> | that would also allow 3rd parties to create and distribute video player UIs that sites can easily apply to their own videos |
| 22:30 | <zcorpan_> | jgraham: isn't <!doctype html><body><option></option> the same issue? |
| 22:32 | <jgraham> | zcorpan_: Yeah, but <rp> is shorter to type |
| 22:34 | <zcorpan_> | jgraham: yeah. it should say "except for elements with the same tag name as the token." in any other end tag |
| 22:34 | <jgraham> | In step 2.1? |
| 22:35 | <zcorpan_> | yes |
| 22:35 | <jgraham> | Good, because that's the bug I field |
| 22:35 | <jgraham> | *filed |
| 22:36 | <zcorpan_> | i thought bugs like these had been ironed out by now :) |
| 22:40 | <zcorpan_> | hmm, why does <rt> outside <ruby> not generate implied end tags? |
| 22:41 | <jgraham> | zcorpan_: Me too :) |
| 22:41 | jgraham | just managed to crash emacs |
| 22:44 | <zcorpan_> | wonder how ie handles rt and rp outside ruby |
| 22:47 | <zcorpan_> | seems ie implicitly closes rt and rp for <rt> or <rp> even outside ruby |
| 22:47 | <zcorpan_> | and doesn't implicitly close <p> even inside ruby |
| 22:48 | <zcorpan_> | while html5 treats rt and rp as unknown elements outside ruby, and forcefully pops anything inside ruby |
| 22:48 | <Hixie> | IE's behaviour is nutto |
| 22:48 | <Hixie> | i based the parser on actual pages |
| 22:49 | <Hixie> | i literally scanned every page in the google index that used <ruby> and then tried to come up with something that worked with all of them |
| 22:49 | <zcorpan_> | Hixie: did you look at pages that use <rt> or <rp> outside <ruby>? |
| 22:51 | <Hixie> | i don't recall there being a significant number |
| 22:58 | <jgraham> | <b><button>foo</b> should reopen the <b> inside the button these days, right? |
| 23:00 | <fagan> | jgraham: well the button itself isnt going to be bold the text is so it should be inside the button tag |
| 23:00 | <jgraham> | fagan: Indeed, but not quite the point |
| 23:00 | <jgraham> | The point is how the error recovery is supposed to work |
| 23:01 | <jgraham> | button used to insert a marker on the list of active formatting elements so that the <b> would never apply inside the <button>. That seems to have changed |
| 23:01 | <fagan> | jgraham: just tested it in chrome and <b> outside the button tag doesnt work |
| 23:02 | <jgraham> | fagan: That could be why it changed :) |
| 23:02 | <fagan> | true |
| 23:02 | <jgraham> | Although I guess it is more likely to be IE compat |
| 23:02 | <jgraham> | Er, wait |
| 23:03 | <jgraham> | The chrome behaviour is the old behaviour, obviously |
| 23:03 | <fagan> | lol |
| 23:06 | <jgraham> | Hmm, I am clearly too tired for this |
| 23:06 | <jgraham> | Hixie: If you fancy looking at parser bugs relatively soon it would be appreciated |
| 23:08 | <fagan> | jgraham: which browser ? |
| 23:09 | <zcorpan_> | fagan: jgraham is tearing apart the html5 parsing algorithm |
| 23:09 | <fagan> | ah ok |
| 23:10 | fagan | knows the engine side of things but not the algorithm stuff |
| 23:13 | <fagan> | I wonder how up to date the html checker is on the w3 site for html5 |
| 23:14 | <fagan> | Im doing a few websites in it at the moment and they are passing fine but id hate for them to not be valid later on |
| 23:14 | <fagan> | no matter anyway |
| 23:14 | <Hixie> | use validator.nu, that's probably the most up to date |
| 23:15 | <Hixie> | validators will catch more and more errors as they get better though |
| 23:15 | <Hixie> | so it's normal to see new errors be caught over time :-) |
| 23:15 | <zcorpan_> | not to mention that html5 can still change out from under you |
| 23:16 | <fagan> | isnt it last call for changes now? |
| 23:16 | <zcorpan_> | sure |
| 23:16 | <Hixie> | last call doesn't mean no changes :-) |
| 23:16 | <zcorpan_> | it can still change |
| 23:16 | <Hixie> | the spec will continue changing its whole life, as people find new problems |
| 23:16 | <Hixie> | when it stops changing is when it's dead |
| 23:17 | <fagan> | true but still I would have thought that the biggest changes have already happened |
| 23:18 | <Hixie> | the biggest ones probably have, yeah |
| 23:20 | <fagan> | I just love the video and audio if nothing else was done but that I would have been very very very happy |
| 23:20 | <fagan> | but thats not the only thing I know |
| 23:33 | <AryehGregor> | Where's the big combined Web Apps 1.0 spec? |
| 23:33 | AryehGregor | is trying to determine how the size of Hixie-edited specs compares to the total output of the CSSWG across its entire existence |
| 23:34 | <AryehGregor> | Oh, drat, not all the W3C specs are single-page. |
| 23:34 | <zcorpan_> | http://www.whatwg.org/specs/web-apps/current-work/complete.html |
| 23:34 | <AryehGregor> | Thanks. |
| 23:35 | <AryehGregor> | I didn't see a link anywhere on the website. |
| 23:35 | <Hixie> | it's there, but i try not to scare people off too much |
| 23:35 | <fagan> | I would be nice if it was split up a little |
| 23:35 | <Hixie> | fagan: http://www.whatwg.org/specs/web-apps/current-work/complete/ |
| 23:35 | <AryehGregor> | Web Apps 1.0: ~5.3 MB |
| 23:35 | <AryehGregor> | All of the stuff I could easily get from the CSS WG without looking at every spec: <3 MB |
| 23:36 | <Hixie> | AryehGregor: you really want to compare the source documents, the post-processor adds a ton of cross-reference stuff |
| 23:36 | <AryehGregor> | Are the source documents for all the CSS specs readily available? |
| 23:36 | <Hixie> | i use way more hyperlinks in my specs than most people |
| 23:36 | <Hixie> | AryehGregor: no idea |
| 23:36 | <fagan> | Hixie: that is a huge page :) |
| 23:36 | <AryehGregor> | The CSS specs have a lot of repeated boilerplate, though, from the headers on every one. |
| 23:37 | <AryehGregor> | Source is almost as large. |
| 23:37 | <AryehGregor> | >4 MB. |
| 23:37 | <Hixie> | fagan: complete/ is a split up version of complete.html |
| 23:42 | <fagan> | Hixie: but still its about a 2 day read start to finish |
| 23:42 | <fagan> | Oh and I love the joke in the how to read this document part :D |
| 23:42 | <Hixie> | splitting it up doesn't make it smaller, no :-) |
| 23:42 | <Hixie> | just makes it have multiple pages |
| 23:43 | <fagan> | true but most people would be interested in specific parts |
| 23:43 | <fagan> | but I suppose they could just look at the index and jump around |
| 23:44 | fagan | understands |
| 23:46 | <zcorpan_> | fagan: what joke? |
| 23:46 | <fagan> | zcorpan_: read it the entire way through, read it again, read it backwards then jump around to the bits you are interested in |
| 23:47 | <zcorpan_> | fagan: i think that's not a joke |
| 23:47 | <fagan> | has to be joking |
| 23:47 | <Hixie> | it's one of those "it's funny because it's true" things |
| 23:47 | <fagan> | reading stuff backwards makes my head hurt |
| 23:48 | <fagan> | words stop making sense |
| 23:48 | <fagan> | :) |
| 23:48 | <annevk> | it just means you have to read it a lot |
| 23:48 | <annevk> | which not many people do, so there may be some issue |
| 23:49 | <fagan> | annevk: I just like to keep up with it :) |
| 23:50 | <annevk> | for that all you need is http://html5.org/tools/web-apps-tracker |
| 23:50 | <annevk> | :) |
| 23:50 | <zcorpan_> | in my experience just reading a spec doesn't give you a useful understanding of it, no matter how many times you read it |
| 23:50 | <fagan> | annevk: true but I hate that page |
| 23:51 | <zcorpan_> | you have to do something actively with it to properly understand it, like implement it or write a testsuite for it |
| 23:51 | <fagan> | dont ask me why |
| 23:51 | <zcorpan_> | and you're better off focusing on one part at a time instead of reading cover to cover |
| 23:52 | <fagan> | zcorpan_: I focus on the bits im interested in and the bits I would use most |
| 23:52 | <annevk> | zcorpan_, I can answer quite a lot of questions within seconds/minutes by having read most bits of the specification a few times |
| 23:52 | <fagan> | (and just skip the rest) |
| 23:52 | <fagan> | I read all the common stuff too though |
| 23:52 | <zcorpan_> | annevk: sure, but you can't find fundamental spec bugs by just reading it, or can you? |
| 23:52 | <zcorpan_> | i can't |
| 23:54 | <zcorpan_> | bedtime, nn |
| 23:54 | <annevk> | oh yeah, finding the real big issues requires implementation experience |