| 04:09 | <Hixie> | i wouldn't be surprised if acid3 is the main reason IE now does DOM Events |
| 04:09 | <Hixie> | (re the earlier discussion) |
| 04:11 | <jcranmer> | shepazu: fwiw, the only strong use case I've seen for SVG fonts are "it uses SVG" and "ooh, multicolored text!" |
| 04:11 | <shepazu> | jcranmer: have you ever used SVG fonts? |
| 04:13 | <jcranmer> | no |
| 04:13 | <jcranmer> | I will freely admit that |
| 04:13 | <shepazu> | yeah, there you go |
| 04:13 | <jcranmer> | I'm not a font designer |
| 04:13 | <jcranmer> | nor do I try to do anything funky with web pages |
| 04:14 | <shepazu> | if those are the only use cases you've heard of, then you clearly haven't done the research or even thought very hard about it |
| 04:14 | <shepazu> | then maybe you aren't the target audience, hm? |
| 04:14 | <jcranmer> | I intend to mean the use cases that are not served by other font technologies |
| 04:15 | <jcranmer> | I don't suppose you could be more enlightening? |
| 04:15 | <shepazu> | yes, there are other use cases that SVG fonts serve that aren't solved by other font formats |
| 04:15 | <jcranmer> | such as... |
| 04:15 | <shepazu> | multicolored text is useful, even if you want to be dismissive of it |
| 04:16 | <shepazu> | having fills that are not simple colors (or gradients) is another |
| 04:17 | <shepazu> | having overlapping lines, as has been done with hand-drawn and printed glyphs for thousands of years, and is used extensively in print advertising and art, but can't be done with other font formats, is another |
| 04:18 | <shepazu> | being able to customize a font, or draw one from scratch, is another |
| 04:18 | <shepazu> | embedding specific glyphs inside a single document (no external references) is another |
| 04:18 | <jcranmer> | it is possible to handle multicolored text with existing technology (i.e., heavy use of <span>) |
| 04:19 | <jcranmer> | embedding without external references seems like a poor design goal to me |
| 04:19 | <shepazu> | that duplicates the text, and isn't very accessible or maintainable |
| 04:19 | <jcranmer> | color isn't accessible |
| 04:19 | <shepazu> | animated glyphs can also be useful |
| 04:20 | <shepazu> | jcranmer: wtf does that mean? |
| 04:20 | <jcranmer> | if I'm blind, I can't see color |
| 04:20 | <shepazu> | first off, accessibility doesn't mean "blind"… there are lots of different kinds of disabilities |
| 04:21 | <jcranmer> | I'm trying to figure out how manually coloring each glyph would be less accessible than a font which automatically colored some of the glyphs |
| 04:21 | <shepazu> | second, duplicating text to get multicolored glyphs for sighted users means that screen readers will double up the text |
| 04:21 | <jcranmer> | <span> isn't duplicating text |
| 04:21 | <jcranmer> | which is why your response is confusing to me |
| 04:22 | <shepazu> | you'll have to be more explicit in how <span>s make a single glyph multicolored |
| 04:22 | <shepazu> | withut duping it |
| 04:23 | <jcranmer> | I am personally not aware of cases that cause multicoloring of single glyphs |
| 04:23 | <shepazu> | and if you're just talking about changing the colors of individual glyphs in a word, you've just killed search, find, screenreading and all other uses of having a word |
| 04:24 | <shepazu> | jcranmer: then I suggest you look at print advertising, where it's used all the time |
| 04:24 | <jcranmer> | no, a<span>b</span>c is matched when I search for 'abc' |
| 04:24 | <jcranmer> | [I just tested it] |
| 04:25 | <shepazu> | in a screenreader? |
| 04:25 | <shepazu> | in a search engine? |
| 04:25 | <jcranmer> | no, the Ctrl-F |
| 04:25 | <shepazu> | not good enough |
| 04:25 | <jcranmer> | I am also dismissive that print advertising should be written in HTML |
| 04:26 | <shepazu> | make some real, comprehensive test cases, and test them, and show some proof, then I'll credit your research |
| 04:26 | <jcranmer> | for that use case, I would support SVG fonts, but I do not think that warrants them as a necessity for support in a web browser |
| 04:26 | <shepazu> | yeah, I'd say you're pretty dismissive, I agree |
| 04:26 | <shepazu> | you are you to say what HTML should be used for? |
| 04:26 | <jcranmer> | the primary case people are talking about is web browsers |
| 04:27 | <jcranmer> | and whether or not SVG fonts should be a feature implemented by web browsers |
| 04:27 | <shepazu> | you think designers don't want to build on the tradition of print media in browsers? |
| 04:28 | <shepazu> | (and not just print, but TVs, giant screens, etc?) |
| 04:29 | <jcranmer> | why should it be necessary that you send the exact same content for a TV and a large poster advertisement that you do to my desktop computer? |
| 04:29 | <shepazu> | why do you think HTML, SVG, and CSS are just on desktop computers? this isn't the 90s |
| 04:30 | <jcranmer> | I do not think that |
| 04:30 | <jcranmer> | let me rephrase it slightly |
| 04:30 | <jcranmer> | should my web browser support the same APIs as, say, a Tivo? |
| 04:31 | <jcranmer> | I am arguing that the domain for which svg fonts is suitable is not the same domain for which a web browser is intended to be used |
| 04:32 | <shepazu> | and I would need to see much more evidence that that is true, because all the evidence I've seen is that the convergence of TV and the Web is already here |
| 04:32 | <shepazu> | TV is now a service, as much as a device |
| 04:33 | <shepazu> | one of my TV providers, Netflix, uses a Web browser on my PS3 |
| 04:33 | <shepazu> | it's written in HTML |
| 04:34 | <shepazu> | (at least on some devices, if not on PS3, I'll have to check that) |
| 04:35 | <shepazu> | HTML, CSS, JS, SVG, APIs, etc are now a primary programming platform, not just a "Web browser" ghetto |
| 04:36 | <shepazu> | I'm surprised I have to make that point |
| 04:37 | <shepazu> | font designers have approached the SVG WG about improving SVG fonts because the existing font formats don't meet the use cases they have |
| 04:37 | <shepazu> | they know their business, and what they (and their clients) want to do |
| 04:38 | <shepazu> | maybe we should listen to them? |
| 04:38 | <shepazu> | anyway, I'm off to bed |
| 07:41 | <zcorpan> | anyone know a url that webkit will fail to resolve? |
| 08:29 | <asmodai> | morning |
| 08:37 | <zcorpan> | Living Acid! awesome |
| 08:46 | <annevk> | on Acid? Sounds like fun :) |
| 08:48 | <annevk> | shepazu, I think the debate was more about whether those additional features were supported |
| 08:48 | <annevk> | shepazu, at least that's why I said as long as it doesn't do per-character-... |
| 09:05 | <asmodai> | Love the discussions that spring up now that Acid 3 was 'eased' down a bit. |
| 09:12 | <robman> | evening all - wondering if anyone knows of a design pattern seen in the wild where the content of an entity can be serialised inline but the canonical id is provided in as an href or similar (e.g. perhaps an IRI/URI)? |
| 09:14 | annevk | is not sure what that means |
| 09:14 | <robman> | 8) fair enough |
| 09:14 | <robman> | atom:link could perhaps be used like this |
| 09:14 | <robman> | but in html5 link is defined as not allowed to have content |
| 09:15 | <robman> | but sometimes it's useful to serialise content (e.g. cached/latest known) but still provide a reference to a source of truth |
| 09:15 | <robman> | just wondering if there's a good reason this design pattern should NOT be used - it seems useful |
| 09:18 | <annevk> | so you have a copy of a resource and link to the original using <link>? |
| 09:21 | <robman> | well in the link example that's how it could work |
| 09:22 | <robman> | we're exploring this design pattern for the POI WG and danbri suggested I ask whatwg people if there was a good reason this was excluded with <link> in html5 |
| 09:22 | <annevk> | why does it have to be inside the <link>? |
| 09:23 | <robman> | for example in a feed...often you want to serialise the latest known copies of things |
| 09:23 | <annevk> | (the reason <link> cannot contain anything is because of HTML parsers) |
| 09:23 | <robman> | yet still allow them to be externalised - it's just a convenience for serialised delivery |
| 09:23 | <robman> | ok...so just a logical convention? |
| 09:24 | <annevk> | a syntax constraint |
| 09:24 | <robman> | k |
| 09:27 | <robman> | my goal was to make the POI model more linked data like and use the existing IANA link relation types where possible rather than create a wave of new entity types - but having too much as links just leads to many network requests so this seemed like a good compromise |
| 09:27 | <robman> | but thanks for your feedback annevk |
| 10:04 | <Ms2ger> | if (node.isSameNode ? node.isSameNode(otherNode) : node == otherNode) |
| 10:05 | <leaverou> | AryehGregor: hi! are you there? |
| 10:06 | <matjas> | zcorpan: other browsers throw a TypeError: DOM object constructor cannot be called as a function when you use `Option` without `new` — do you know if that’s a bug or a feature? |
| 10:11 | <annevk> | Web IDL should be able to tell you that |
| 10:11 | <annevk> | serializing HTML in that matter seems somewhat icky btw |
| 10:15 | <zcorpan> | matjas: that's a bug per webidl i think |
| 10:19 | <zcorpan> | it's not serializing, it's escaping |
| 10:19 | <matjas> | annevk: no argument there :) |
| 10:20 | <annevk> | new Text(html) |
| 10:20 | <annevk> | at some point |
| 10:21 | <annevk> | or just appendChild(htmlString) |
| 10:21 | <annevk> | at some point |
| 10:31 | <erlehmann> | Ms2ger, what is this i don't even. |
| 11:25 | <annevk> | Ms2ger, and xmlVersion too! |
| 11:33 | <Ms2ger> | Is Opera removing those, btw? |
| 11:35 | <annevk> | I've been filing bugs |
| 11:39 | <annevk> | Ms2ger, we are a little behind still, having not yet removed e.g. attachEvent, but we'll get there |
| 11:40 | <Ms2ger> | Great |
| 11:43 | <roc> | I find it harder to yell at Alex Danilo since I found out he went to my wedding |
| 11:44 | <jgraham> | roc: I am the opposite of an expert in this field, but doesn't one usually invite people to one's wedding? |
| 11:45 | <jgraham> | So it seems surprising to find out about it years later |
| 11:46 | <roc> | at the SVG WG meeting, I found out he used to work at CISRA (Canon Research Australia) |
| 11:46 | <roc> | I asked him if he'd met my wife, who was working there when we got married |
| 11:46 | <roc> | 12 years ago |
| 11:46 | <roc> | he said "yeah, I went to her wedding along with her other workmates" |
| 11:49 | <roc> | this was of course before I'd even contemplated working on Web browsers ... it's hard to remember there being such a time |
| 11:51 | <Ms2ger> | Starting to feel old? :) |
| 11:51 | <jgraham> | So was this some part of the wedding where any interested party could attend, or had you actually invited him, but forgotten about it in the intervening time? |
| 11:51 | <roc> | jgraham: Janet just invited her workmates, and I'd forgotten their names |
| 11:51 | <jgraham> | Fair enough |
| 11:52 | <roc> | I probably didn't even remember their names on the day; I was nervous |
| 11:52 | jgraham | is no longer confused |
| 11:52 | <roc> | Ms2ger: not "starting", no :-) |
| 11:54 | <annevk> | heh |
| 12:04 | <annevk> | jgraham, why are assert_exists and assert_not_exists deprecated? |
| 12:06 | <jgraham> | annevk: Because they confused people |
| 12:07 | <jgraham> | assert_exists is basically assert_own_property |
| 12:07 | <jgraham> | s/basically/exactly/ |
| 12:09 | <annevk> | and not? |
| 12:10 | <jgraham> | I don't think there is a replacement at the moment |
| 12:10 | <jgraham> | It would be assert_not_own_property |
| 12:11 | <jgraham> | I wish I had made things more composable |
| 12:11 | <Ms2ger> | Sounds like a weird thing to test |
| 12:11 | <jgraham> | Or, at least had and, or, not functions |
| 12:11 | <Ms2ger> | Also, assert_truthy would be nice |
| 12:12 | <jgraham> | Ms2ger: I don't like the sound of that so I am quite happy for people to use assert_true(truthy == true) |
| 12:12 | <Ms2ger> | assert_true(!!truthy), more likely |
| 12:13 | <jgraham> | Well whatever |
| 12:13 | <Ms2ger> | :) |
| 12:13 | <jgraham> | But doing type-converting assertions seems bad |
| 12:13 | <annevk> | Ms2ger, I was thinking of testing the historical section of DOM4 |
| 12:14 | <Ms2ger> | Oh? |
| 12:14 | <annevk> | Ms2ger, to make it easier to find out what bugs need to be filed |
| 12:14 | <Ms2ger> | Not stopping you :) |
| 12:14 | <annevk> | I'm now using stuff like assert_equals(document[name], undefined) |
| 12:15 | <annevk> | which works |
| 12:15 | <Ms2ger> | assert_false(name in document)? |
| 12:18 | <annevk> | could do that too |
| 12:18 | <annevk> | it's shorter... |
| 12:19 | <annevk> | but maybe the error messages are less nice? |
| 12:19 | <jgraham> | Well if it exists you won't get to find out its value |
| 12:20 | <Ms2ger> | Mm, true |
| 12:21 | <Ms2ger> | dev.w3.org down? |
| 12:24 | <annevk> | Ms2ger, it was for me some time ago |
| 12:25 | <zcorpan> | annevk: getComputedStyle's second argument should not stringify null to 'null' |
| 12:25 | <zcorpan> | annevk: http://www.google.com/codesearch#search/&q=getcomputedstyle%5C(%5B%5E,%5D%2B,%5Cs*null%5Cs*%5C)%20lang:js&type=cs |
| 12:25 | <Ms2ger> | Makes sense |
| 12:26 | <Ms2ger> | annevk, happen to have the latest Progress Events draft lying around, then? |
| 12:27 | <annevk> | Ms2ger, for this one time, http://www.w3.org/TR/progress-events/#interface-progressevent should prolly suffice |
| 12:29 | <Ms2ger> | The horror :) |
| 12:30 | <annevk> | Ms2ger, hey, Opera has only 13 fail, Mozilla 15! http://w3c-test.org/webapps/DOMCore/tests/submissions/Ms2ger/historical.html |
| 12:30 | <annevk> | and I'm comparing a Mozilla nightly to a shipping Opera (prolly doesn't matter) |
| 12:30 | Ms2ger | applies his CDATASection patches |
| 12:31 | <annevk> | and we can still do easy stuff like removing Entity and such |
| 12:33 | <foolip> | is it Hixie that administers wiki.whatwg.org? |
| 12:34 | <Ms2ger> | foolip, Lachy and AryehGregor, IIRC |
| 12:35 | <foolip> | Lachy, AryehGregor, logging in doesn't currently work, gives "There seems to be a problem with your login session" and the following printed at the bottom: |
| 12:35 | <foolip> | Warning: Unknown: write failed: No space left on device (28) in Unknown on line 0 |
| 12:35 | <foolip> | Warning: Unknown: Failed to write session data (files). Please verify that the current setting of session.save_path is correct (/tmp) in Unknown on line 0 |
| 12:35 | <foolip> | looks like someone filled up the disk space... |
| 12:37 | <Ms2ger> | Hmm, we seem to have a working implementation of NameList |
| 12:37 | <Ms2ger> | But no way to get your hands on one |
| 12:39 | <annevk> | Ms2ger, see, this test was a great idea :) |
| 12:40 | <annevk> | WebKit has 17 fail |
| 12:54 | <Lachy> | foolip, if it's a disk space issue, then Hixie would have to adjust the quota. |
| 12:55 | <Lachy> | But there doesn't seem to be any such problem, since I can log in with SSH and write files without a problem. |
| 12:55 | <foolip> | Lachy, can you log into the wiki? |
| 12:55 | <Lachy> | unless it's a problem with the database quota, which is more likely to fill up. |
| 12:56 | <Lachy> | foolip, no. I get the same error. I can only ssh into the server |
| 12:56 | <foolip> | ok, so wait until Hixie wakes up? |
| 12:56 | <Lachy> | yeah, either he or AryehGregor will have a better chance of fixing than myself. |
| 13:04 | <hsivonen> | why does Web Socket have ping messages? (as opposed to letting the app define its own pinging) |
| 13:06 | <hsivonen> | allowing ping frames contain application data seems like a great way to invite authors to transfer significant traffic in a place that wasn't meant for it |
| 13:07 | <zcorpan> | hsivonen: a ping frame can only have 125 bytes payload |
| 13:08 | <Philip`> | You can't even piggyback Twitter messages on it? |
| 13:09 | <hsivonen> | zcorpan: why does have payload at all? |
| 13:09 | <zcorpan> | hsivonen: i don't know |
| 13:09 | <hsivonen> | s//it / |
| 13:09 | <hsivonen> | zcorpan: ok |
| 13:09 | <zcorpan> | design by hybi |
| 13:17 | <Philip`> | Sounds a bit like "another of those things which it was easier to allow to be put in there than it was to argue against it" (from an unrelated standardisation context) - someone says they might possibly want it, so might as well add it to the spec |
| 13:18 | <annevk> | that's how the CSS WG operates |
| 13:25 | <annevk> | hmm I thought Gecko had "DOMUserData" |
| 13:26 | <annevk> | guess I need to add tests for the methods that return that too |
| 13:26 | <annevk> | ooh, it has "UserDataHandler" |
| 13:30 | <Ms2ger> | 17 |
| 13:35 | <annevk> | I'll do some shopping and then up that number some more :) |
| 14:38 | <hsivonen> | I'm a bit surprised at the successful 386 resistence of the usual suspects regarding the "Benefits of polyglot XHTML5" thread on the help list |
| 14:39 | <annevk> | Have you considered they might not be subscribed? |
| 14:39 | <Ms2ger> | annevk, so... |
| 14:39 | annevk | isn't for instance |
| 14:39 | <Ms2ger> | Given two XML documents |
| 14:39 | <Ms2ger> | <!DOCTYPE foo [ <!ELEMENT foo (#PCDATA)> ]><foo/> |
| 14:39 | <Ms2ger> | and |
| 14:39 | <Ms2ger> | <!DOCTYPE foo [ <!ELEMENT foo EMPTY> ]><foo/> |
| 14:39 | <Ms2ger> | What should doctype1.isEqualNode(doctype2) return? |
| 14:40 | <annevk> | throw a COME_ON_ERR |
| 14:40 | <annevk> | but euh, true |
| 14:41 | <Ms2ger> | OK |
| 14:41 | <hsivonen> | is the internal subset retrievable via the DOM? |
| 14:42 | <annevk> | that is, I still think we should nuke DocumentType.internalSubset, and I think we should not care about whether or not DocumentType nodes make a whole lot of sense, because they should not have been there in the first place |
| 14:42 | <Ms2ger> | hsivonen, not anymore per spec |
| 14:46 | <hsivonen> | Ms2ger: nice |
| 14:50 | <Ms2ger> | Hmm |
| 14:50 | <Ms2ger> | http://software.hixie.ch/utilities/js/live-dom-viewer/saved/1142 |
| 14:50 | <Ms2ger> | Three browsers, three results |
| 14:51 | <jgraham> | Try more versions of the browsers |
| 14:51 | <Ms2ger> | I'm lazy |
| 14:51 | <Ms2ger> | Anyone want to check what IE does? |
| 14:54 | <annevk> | Ms2ger, I think WebKit wins |
| 14:54 | <hsivonen> | I just got to the email where Hallvord described the brokenness of scripts when a browser fires both readystatechange and load on scripts. I'm not really all that surprised, but it's still so sad. |
| 14:54 | <Ms2ger> | Mm |
| 14:54 | <annevk> | Ms2ger, it's the intersection of Opera and Gecko and the minimum we can get away with :) |
| 14:55 | <Ms2ger> | Not changing that just yet, though |
| 14:58 | <annevk> | changing? |
| 14:59 | <annevk> | Ms2ger, http://html5.org/specs/dom-parsing.html#concept-serialize-xml is wrong for Text because e.g. & would not be escaped |
| 14:59 | <Ms2ger> | Almost everything in that algorithm is wrong :) |
| 15:20 | <annevk> | Cunning how http://www.glazman.org/weblog/dotclear/index.php?post/2011/09/19/HTML-UI references the battery API and not one of the many text layout thingiebingies |
| 15:48 | <annevk> | Ms2ger, even within DOM3Core NameList is not returned by anything; curious |
| 16:02 | <Ms2ger> | annevk, Validation |
| 16:04 | <Ms2ger> | And down to 12 |
| 16:04 | <annevk> | you win |
| 16:05 | <Ms2ger> | Now I need to get all these reviewed... |
| 16:59 | <annevk> | Ms2ger, where else is TextIsOnlyWhitespace() used? |
| 17:00 | <Ms2ger> | CSS optimizations |
| 17:06 | <annevk> | After all these years, we still have DOM Load & Save in Opera |
| 17:06 | <annevk> | That's why we have DOMError apparently |
| 20:03 | <AryehGregor> | jgraham, is it safe to nest test()? |
| 20:04 | AryehGregor | doesn't see why it wouldn't be |
| 20:07 | AryehGregor | realizes he's not actually nesting tests anyway here |
| 20:34 | <theoros> | is there a reason the css3 text spec doesn't seem to have a way to inherit the color for text-shadow? |
| 20:34 | <theoros> | i can see some mention of this at http://lists.w3.org/Archives/Public/www-style/2011Mar/0126.html but can't find much else |
| 20:34 | <divya> | no reason |
| 20:35 | <divya> | just like same for box-shadow |
| 20:42 | <Hixie> | theoros: there's a color called "currentcolor" that is the value of the 'color' property |
| 20:49 | <theoros> | can that be used in the value of the text-shadow property? |
| 20:59 | <theoros> | Hixie: ^ |
| 20:59 | <theoros> | sorry, i should just.. try it. :) |
| 21:02 | <theoros> | true enough. http://cssdesk.com/AxcyD thanks for pointing that out. |
| 21:11 | <TabAtkins> | Are js strings ucs-2 to utf-16? |
| 21:12 | <zewt> | (huh?) |
| 21:12 | <TabAtkins> | s/to/or/ |
| 21:13 | <annevk> | they're 16-bit code units and don't deal with surrogates |
| 21:13 | <TabAtkins> | So, ucs-2. |
| 21:14 | <zewt> | do any implementations actually vomit utf-16 on you if high codepoints come in from outside? |
| 21:14 | <zewt> | ("here you go, good luck!") |
| 21:14 | <zewt> | (to be fair, that's probably better than whatever other hack might be devised in that situation) |
| 21:15 | <annevk> | TabAtkins, I think it's UTF-16, but there's not a native concept of character |
| 21:15 | <matjas> | “Serious about XHTML” — http://mathiasbynens.be/notes/xhtml5#comment-46 |
| 21:15 | <annevk> | TabAtkins, just 16-bit code units |
| 21:15 | <matjas> | must… resist… |
| 21:15 | <TabAtkins> | annevk: Ah, okay, I get the distinction you're drawing. |
| 21:15 | <zewt> | annacc: aka "it'll break your stuff but it's the best we've got" |
| 21:16 | <zewt> | nothing ucs-16 (or utf-8) exposes data per-codepoint anyway |
| 21:19 | <annevk> | matjas, haha |
| 21:20 | <annevk> | matjas, nothing changed since 2002 or so it seems |
| 21:20 | <matjas> | TIL browser vendors are lazy |
| 21:20 | Philip` | wonders why laziness is seen as a bad argument |
| 21:21 | <Philip`> | Surely both facilitating laziness and exploiting laziness are good approaches to take in any design |
| 21:21 | <matjas> | Philip`: because XHTML! |
| 21:21 | <annevk> | matjas, oh I didn't read the specific comment; I started at the top... |
| 21:21 | <zewt> | heh ... even xhtml has its place :P |
| 21:22 | <annevk> | matjas, seemed like a lot of omg have to use XML it will improve the web kind of comments |
| 21:22 | <zewt> | (why's it so quiet in here all of a sudden) |
| 21:22 | <Philip`> | (It's quiet because HTML is dying) |
| 21:22 | <matjas> | annevk: that pretty much sums it up. i gave up on replying |
| 21:22 | <AryehGregor> | jgraham, I just tweaked testharness.js so that it will load testharness.css correctly even if the <script> was removed by something in the interim. |
| 21:23 | <zewt> | the only case i've found xhtml useful is on things that ... aren't the web |
| 21:23 | <zewt> | (eg. isolated protocols where you want to define simple markup, where very limited xhtml profiles can be handy) |
| 21:24 | <annevk> | matjas, but yeah whatever, no need to 386 TechZilla |
| 21:26 | <AryehGregor> | jgraham, I needed this for my editing tests to work right in IE (although I could have worked around it on my side, of course). |
| 21:26 | <AryehGregor> | jgraham, feel free to critique/revert/refactor/whatever. |
| 21:35 | <timeless> | AryehGregor: obviously you need to use cmd.exe w/ prefs per app to have different default background colors per session. -- oh wait, does the ubuntu terminal support that? :) |
| 21:35 | <timeless> | zewt: fwiw, on my w7 i can fairly easily distinguish windows |
| 21:35 | <timeless> | but i rely on two or three things: 1. window shapes (because not all windows are maximized), 2. window titles (because microsoft doesn't hide them and they aren't all useless in my case). 3. pictures (which aren't totally useless for my windows) |
| 21:35 | <AryehGregor> | timeless, . . . what? |
| 21:35 | <timeless> | AryehGregor: very high latency reply |
| 21:35 | <timeless> | you were talking about how Unity was unhelpful |
| 21:35 | AryehGregor | doesn't get the context |
| 21:35 | <AryehGregor> | Oh. |
| 21:35 | <AryehGregor> | I worked around that already. |
| 21:36 | <timeless> | good; what did you do? |
| 21:36 | <zewt> | timeless: i have no idea what you're responding to :) |
| 21:37 | <timeless> | you were complaining about the w7 taskbar's unusable previews |
| 21:37 | <AryehGregor> | timeless, Alt-F2, gnome-panel |
| 21:37 | <timeless> | (part of aero peak) |
| 21:38 | <timeless> | AryehGregor: alt-f2 sounds like win-r :), glad to hear that the unity update gave you your feature back :) |
| 21:40 | <AryehGregor> | Okay, so in IE, getSelection().addRange() decides to throw "Unspecified error." |
| 21:40 | <AryehGregor> | For no obvious reason. |
| 21:40 | <timeless> | heh |
| 21:40 | <AryehGregor> | I think I'm just not going to debug this. I've put more than enough time into it. |
| 21:40 | AryehGregor | tries in IE10PP2 to give it one last chance |
| 21:41 | <AryehGregor> | Oh, nice, it doesn't fail as horribly. |
| 21:43 | <zewt> | gushing review |
| 21:43 | <ChrisWilson> | That reminds me of an IE poster I have somewhere from around IE5 - the tagline is "You were expecting a new version, but you weren't expecting this." Doesn't have a version number. Could apply to any version, really. :) |
| 21:44 | <timeless> | heh |
| 21:44 | timeless | sighs |
| 21:44 | <timeless> | the Mozilla Nightly uninstaller isn't friendly to Windows [8] App Certification Kit -- it doesn't support unattended uninstall |
| 21:45 | <AryehGregor> | . . . does anyone else have an issue where IE sometimes displays backslashes as struck-through W's? |
| 21:45 | AryehGregor | wonders if it's a font problem |
| 21:45 | <timeless> | There's a blog explaining it |
| 21:45 | <timeless> | you're in Korea/Japan iirc |
| 21:45 | <timeless> | http://blogs.msdn.com/b/michkap/archive/2005/09/17/469941.aspx |
| 21:45 | <timeless> | When is a backslash not a backslash? |
| 21:46 | <timeless> | The character in question is U+005c, the REVERSE SOLIDUS, also known as the backslash or '\'. It is the path separator for Windows, which is encoded at 0x5c across all of the ANSI code pages. |
| 21:46 | <timeless> | Well, on Japanese code page 932, 0x5c is the YEN SIGN, and on Korean code page 949, 0x5c is the WON SIGN. |
| 21:46 | <ChrisWilson> | When it's someone saying "http backslash backslash"? |
| 21:46 | <timeless> | -- you should read the whole article |
| 21:46 | <timeless> | in fact, it should be required reading |
| 21:47 | <divya> | ChrisWilson: has your nocompete expired :)) |
| 21:47 | <hober> | divya: on the 21st, iirc |
| 21:47 | timeless | wonders where ChrisWilson went |
| 21:47 | <divya> | o i c |
| 21:47 | divya | wonders the same |
| 21:48 | <AryehGregor> | timeless, that is terrifying. |
| 21:48 | <timeless> | AryehGregor: i'm moderately horrified to discover you haven't learned about that already |
| 21:49 | <AryehGregor> | timeless, why would I? The page is UTF-8, why the heck should it interpret anything as Korean? |
| 21:49 | <zewt> | ... is a yen sign a struck-through W? :) |
| 21:50 | <timeless> | heh |
| 21:50 | <AryehGregor> | No, the won sign is. |
| 21:50 | <zewt> | the yen thing is one of the most continuously obnoxious long-term cluster^Wscrewups MS has given us, heh |
| 21:50 | <zewt> | for people who have to see it, anyway |
| 21:50 | <AryehGregor> | Why am I seeing it, though? |
| 21:51 | <zewt> | what's your codepage? |
| 21:51 | <timeless> | zewt: totally unfair |
| 21:51 | <zewt> | japanese systems always display yen for backslash |
| 21:51 | <timeless> | it's not microsoft's fault |
| 21:51 | <zewt> | also, what's with unicode's obsession for calling things by obscure names? it's a damn backslash guys |
| 21:51 | <timeless> | it's more or less the fault of the japanese/koreans |
| 21:52 | <zewt> | timeless: it's their OS that's shipped it that way for a couple decades, so I blame them |
| 21:52 | <timeless> | zewt: i'm not a fan of REVERSE SOLIDUS either.. |
| 21:53 | <ChrisWilson> | whoops, sorry Divya |
| 21:53 | <Philip`> | It's Microsoft's fault for not using Unicode before it was invented |
| 21:53 | <ChrisWilson> | I have 2 days, 3 hours and 3 minutes left. |
| 21:53 | <ChrisWilson> | :) |
| 21:53 | <divya> | hehehehe |
| 21:54 | <divya> | are you throwing a party when it ends ChrisWilson? |
| 21:54 | <timeless> | ChrisWilson: not counting the seconds? :) |
| 21:54 | <AryehGregor> | zewt, my OS language is English. The page's encoding (as determined from the View menu) is UTF-8. |
| 21:54 | <aho> | solidus ⁄ slash / |
| 21:54 | <aho> | there is a difference |
| 21:54 | <ChrisWilson> | I'm thinking it's going to be a rolling party. :) But yes, I think I'll host a Seattle party on Oct 1. |
| 21:55 | <divya> | o noes i wont be in town :( |
| 21:55 | <divya> | (not that I will be invited or anything) |
| 21:55 | <divya> | :P :P |
| 21:55 | <ChrisWilson> | timeless: well, yes, I'M counting - but people think I'm weird when I say how many seconds. And then keep saying how many seconds. |
| 21:55 | <zewt> | aho: \ is a backslash; unicode can shake their fist angrily all they want, it's still a backslash |
| 21:55 | <ChrisWilson> | divya: of course you would! |
| 21:56 | <jgraham> | AryehGregor: that ptch looks OK. I would have just written a function that found the prefix and then used that to set a variable rather than having the function set a variable in the enclosing scope |
| 21:56 | <divya> | hahaha thnx ChrisWilson you saved me from an AWK moment. |
| 21:56 | <AryehGregor> | jgraham, you mean just have the function return the value? That makes sense. |
| 21:56 | <AryehGregor> | Didn't think of that. |
| 21:56 | <aho> | my favorite is still octothorpe for # |
| 21:57 | <jgraham> | AryehGregor: Yeah |
| 21:57 | <Philip`> | aho: At least everyone can agree that's the wrong name, unlike e.g. "pound" |
| 21:57 | <aho> | i like hash most |
| 21:57 | <zewt> | gar someone sending mail as "whatwg@hostname" with no name |
| 21:57 | <timeless> | heh |
| 21:58 | <zewt> | i reply to him directly (asking him to fix it) and it shows up in gmail as "Glenn Maynard to whatwg" as if i sent it to the list, heh |
| 21:58 | <jgraham> | AryehGregor: If we were being pickly, I would mildly prefer the comment inside the function |
| 21:58 | <zewt> | hello mr. whatwg |
| 21:58 | <zewt> | aho: heh yeah i was trying to remember that one |
| 21:58 | <zewt> | and yet they're so honest with PILE OF POO |
| 21:58 | <jgraham> | *picky |
| 21:58 | <aho> | speaking of octo*, how is github's mascot called? everyone seems to call it octocat :3 |
| 21:58 | <jgraham> | I assume pickly is when one is covered in vinegar |
| 21:59 | <aho> | haha |
| 21:59 | <aho> | i wonder if they will also add steaming pile of poo and pile of poo with some flies |
| 21:59 | jgraham | -> sleep |
| 21:59 | <divya> | aho: http://octodex.github.com/ |
| 22:00 | <aho> | nya :3 |
| 22:05 | <AryehGregor> | Okay, so do IE and Opera decide whether to add closing tags for things like <p> based on astrology or something? |
| 22:06 | <AryehGregor> | In innerHTML, I mean. |
| 22:06 | <AryehGregor> | When serializing. |
| 22:06 | <timeless> | heh |
| 22:06 | <AryehGregor> | Because they seem really inconsistent. |
| 22:06 | <timeless> | AryehGregor: have you looked at the source material? |
| 22:06 | <AryehGregor> | Which causes them to fail tests in unrelated ways, but if I can't work around it, I guess that's too bad . . . |
| 22:06 | timeless | wonders if they try to avoid adding </p> when the original content didn't have </p> |
| 22:08 | <AryehGregor> | I imagine it's something comparably insane to that. |
| 22:09 | <AryehGregor> | Or maybe they have a special case where setting innerHTML and then getting it will return the same value. |
| 22:09 | <AryehGregor> | That would explain what I'm seeing. |
| 22:09 | <timeless> | oh, yeah, that makes sense |
| 22:10 | <AryehGregor> | IE10PP2 seems not to quite do that. |
| 22:10 | <timeless> | heh |
| 22:10 | <AryehGregor> | IE's parsing is completely wacko, by the way. It parses attribute names to non-lowercase, normalizes attribute values . . . |
| 22:11 | <timeless> | how different is uppercase from non-lowercase? |
| 22:11 | <AryehGregor> | It does things like parsing contenteditable="foo" to contentEditable="foo". |
| 22:11 | <AryehGregor> | And style="font-weight: bold" to style="font-weight: bold;". |
| 22:12 | <timeless> | oh |
| 22:12 | <timeless> | canonicalcase :) |
| 22:12 | <Philip`> | How about <span style="font-weight:bold" style="color:red"> ? |
| 22:12 | <AryehGregor> | Of course, Firefox changes <video> to <video tabindex="0">, so . . . |
| 22:12 | <timeless> | heh |
| 22:28 | <annevk> | I always knew ChrisWilson wanted to hang out here but couldn't because of Microsoft :p |
| 22:28 | <annevk> | Kind of funny considering what http://en.wikipedia.org/wiki/Web_Hypertext_Application_Technology_Working_Group has to say about Microsoft |
| 22:37 | <ChrisWilson> | annevk: I dropped in a few times even when I worked for MS. |
| 22:37 | <timeless> | ChrisWilson: so when did you leave? |
| 22:37 | <ChrisWilson> | Actually, wikipedia's article was substantially correct - although I did list a couple of other reasons. |
| 22:37 | timeless | is clearly out of the loop |
| 22:38 | <ChrisWilson> | Microsoft? |
| 22:38 | <timeless> | yeah |
| 22:38 | <ChrisWilson> | I left one year minus two days, two hours nd 19 minutes ago. |
| 22:38 | <paul_irish> | lol |
| 22:38 | <timeless> | wow |
| 22:38 | <gsnedders> | Well, that's certainly accurate. |
| 22:38 | <timeless> | does ms give you sufficient severance to afford to be without an employer for that long? |
| 22:39 | <ChrisWilson> | timeless: I've been working for Google since last November 1st. |
| 22:39 | timeless | isn't sure if severance is the right word, perhaps compensation? |
| 22:39 | <ChrisWilson> | (and no - severance =0. :) |
| 22:40 | <ChrisWilson> | (or more accurately - severance = whatever office supplies you can steal on your last day) |
| 22:40 | <timeless> | how much impact has the non-compete had on your everyday life / what you worked on? |
| 22:40 | <Philip`> | timeless: Hush money? :-) |
| 22:40 | <timeless> | ChrisWilson: in my case, i ordered a phone from my previous employer a month or two before i left |
| 22:40 | <timeless> | i was overdue by over a year for one :( |
| 22:40 | <timeless> | and the one i got is kinda lame |
| 22:40 | <timeless> | otoh, someone's using it |
| 22:40 | <timeless> | which means someone appreciates it... |
| 22:41 | <ChrisWilson> | timeless: well, I couldn't work on the core web platform team. :/ |
| 22:41 | <timeless> | 12mp cameras aren't incredibly common today.. |
| 22:42 | <ChrisWilson> | timeless: I presume you mean 12mp cameras aren't incredibly common on cell phones today? |
| 22:42 | <gsnedders> | AryehGregor: innerHTML serialization and optional tags is based upon the input in Opera, IIRC |
| 22:42 | timeless | doesn't know much about cameras |
| 22:43 | <timeless> | but sure, if it's necessary to specify cell phones, then make it so :) |
| 22:43 | <gsnedders> | ChrisWilson: You and your 5D MkII… |
| 22:43 | <timeless> | the last camera i bought was >5 years ago (and it was 8mp) |
| 22:44 | <ChrisWilson> | ah. Current crop of point-and-shoot cameras average 10-12mp. |
| 22:44 | <annevk> | ChrisWilson, fair enough; another party at TPAC? ;) |
| 22:45 | <ChrisWilson> | 12mp seems to be the point where people finally stop caring about "give me MOAR pixels!" and start realizing they've well overshot the resolving power of the crappy lens in their point and shoot. :) |
| 22:46 | <timeless> | heh |
| 22:46 | <ChrisWilson> | gsnedders: yeah, although I gotta say - the massive RAW files off the 5Dmkii actually do create storage problems. I've had to move photos off to another disk. |
| 22:46 | <timeless> | heh |
| 22:47 | <gsnedders> | ChrisWilson: Heh. I had to do that while my laptop was stil my main computer (with 500GB HDD). |
| 22:47 | <ChrisWilson> | annevk: I need to figure out if I'm going to TPAC. (I know, it's in Santa Clara, not much of an excuse not to.) |
| 22:49 | <ChrisWilson> | gsnedders: I'm lazy, and don't want to figure out how to seamlessly replace my 1TB disk with a 3TB in my iMac. |
| 22:50 | <gsnedders> | ChrisWilson: I now have a 3TB disk, but 1.5TB Linux and 1.5TB Windows, so all my photos have to fit into the latter (admiteddly most stuff in Windows in on SSD, so the HDD is pretty much just photos) |
| 22:50 | <timeless> | heh |
| 22:54 | <zewt> | dual-booting in 2011? D: |
| 22:54 | <ChrisWilson> | heh |
| 22:54 | <ChrisWilson> | dual web browsers is the new dual booting |
| 22:54 | <gsnedders> | zewt: Games! |
| 22:54 | <gsnedders> | zewt: And still have a sane POSIX terminal :P |
| 22:54 | <zewt> | if i need two OS's, i use two computers :P |
| 22:54 | <zewt> | (or a VM) |
| 22:54 | <zewt> | (though that doesn't help with games) |
| 23:01 | <timeless> | zewt: can i show you something which will probably make you cringe? :) |
| 23:02 | <gsnedders> | zewt: Yeah, I'd need to use Windows as the host OS, which means the majority of the time I'd be in the virturalized env, which isn't great for perf. |
| 23:05 | <timeless> | gsnedders: how *bad* for perf is it? |
| 23:06 | timeless | thought vm's had gotten to the point of mostly not sucking |
| 23:06 | <gsnedders> | timeless: A few %, AIUI |
| 23:07 | <timeless> | does that really matter? |
| 23:07 | <timeless> | if a build takes an hour, a percent is less than a minute |
| 23:07 | <timeless> | a few percent is maybe 2 minutes |
| 23:08 | timeless | tries using FishIE w/ Firefox (6?) on Metro @1600x1200 running in Scaled mode inside Windows 7 |
| 23:08 | <timeless> | it's getting 10-12 fps |
| 23:10 | <timeless> | oh gosh, these guys cross posted to both dap and dom? |
| 23:11 | <timeless> | http://lists.w3.org/Archives/Public/public-device-apis/2011Sep/att-0086/OMM_deviceAPI.txt |
| 23:12 | <timeless> | oh no, dom is realy dom the person, not the group :) |
| 23:12 | timeless | sighs |
| 23:20 | <Hixie> | TabAtkins: what do you recommend on the @global thing? (i got a request to finish speccing this out asap since people are implementing <style scoped>) |
| 23:23 | <TabAtkins> | I hadn't suggested @global on www-style yet, though that's on my list for today. |
| 23:23 | <TabAtkins> | Chrome people doing the implementing are aware of the proposal, so no worries about immediate speccing. |
| 23:24 | <Hixie> | TabAtkins: k |
| 23:24 | TabAtkins | goes ahead and writes the email now. |
| 23:25 | <gsnedders> | WRITE ALL THE EMAILS. |
| 23:56 | <TabAtkins> | Hixie: Wrote the email. |