00:34
<Hixie>
ok i'm outta here. bbl.
09:02
<idnc>
hi all
09:19
<idnc>
before I file a feature request(or more of a query)..
09:20
<idnc>
http://pastebin.com/ZmStuMQn
09:21
<idnc>
- the feature part - was this already discussed here? or is this even the right place?
09:23
<idnc>
..I was send here from moz dev nw, although I think its more browser-related then a spec issue
09:46
jgraham
grumbles at the lack of x-archived-at type headers on WHATWG mail
09:56
kennyluck
why need x- ?
09:59
<jgraham>
Oh Maybe no x-
15:31
<wilhelm>
Have any browser vendors written any tests for <details> (and friends) yet?
15:33
<Peter->
There are a few tests on WebKit's SVN, though no formal test-cases
15:33
<jgraham>
wilhelm: Don't think anyone has implemented <details> yet…
15:33
<Peter->
WebKit has
15:33
<Peter->
http://trac.webkit.org/changeset/81035
15:33
<jgraham>
Peter-: With an actual UI?
15:34
<wilhelm>
jgraham: I only care about tests right now. (c:
15:34
<Peter->
Yes, jgraham, albeit simple. It's available in WebKit/Chromium nightlies. Further down that page is a list of changed files which begins with various html files
15:34
<wilhelm>
Peter-: Excellent. Thank you.
15:34
<Peter->
these are tests
15:36
<jgraham>
Peter-: Oh. Indeed. Nice
15:36
<jgraham>
wilhelm: BTW it occurred to me that the open/closed thing is an ideal use for a != ref test
15:40
<wilhelm>
jgraham: Indeed.
19:00
<AryehGregor>
Wow, I finally got IE9 to work with my execCommand() auto-run thing.
19:01
<AryehGregor>
It seems like execCommand() only works . . . well, I have no idea what the actual conditions are, but I eventually fixed the problem by replacing a function call with setting an input's value, focus()ing the input, and click()ing a button that ran the script on the input's value.
19:01
<AryehGregor>
I mean, WTF, but it works.
19:04
<AryehGregor>
So now I can actually research what IE does.
19:34
<Hixie>
AryehGregor: that's messed up
19:34
<Hixie>
AryehGregor: i wonder how devs get it to work on real sites
19:35
<AryehGregor>
Hixie, remember, the usual use-case is interactive stuff.
19:35
<AryehGregor>
It's not going to be common to want to run hundreds of commands on different artificially-constructed selections without user interaction of any kind.
19:35
<AryehGregor>
Also, I'm using the W3C Range API, which is new in IE9. Maybe it would work differently with their proprietary API.
19:35
<AryehGregor>
Anyway, the overall answer to how devs get execCommand() to work on real sites "with much pain".
19:38
<Hixie>
fair enough
20:58
<AryehGregor>
I am seriously ready to murder whoever wrote the default vim indentation script for HTML.
20:59
<AryehGregor>
Most of them are fine, like for PHP or JS. I have quibbles, but they're usable.
20:59
<AryehGregor>
But HTML . . . argh.
21:07
<matijsb1>
AryehGregor: what goes wrong? (thinking of switching to vim from TextMate here)
21:08
<AryehGregor>
matijsb1, lots and lots of stuff. For instance, whenever I try to do "} else {" in an embedded JS block, in K&R style, it tries to indent it an extra line.
21:08
<AryehGregor>
Then when I try to de-indent it, it reindents it when I hit Enter.
21:08
<zewt>
it tends to be nitty about switching to JS mode inside HTML, though increasing the search distance helps a lot
21:08
<AryehGregor>
Then when I de-indent it again, it de-indents all following lines I add by an extra tab.
21:09
<AryehGregor>
zewt, I just use syn sync fromstart for HTML, that solves it.
21:09
<AryehGregor>
But the indentation is infuriating.
21:09
<zewt>
i'd expect it to just use JS's indentation rules for JS nested in HTML, though
21:09
<matijsb1>
AryehGregor: oh god, that sounds bad
21:09
<AryehGregor>
If you try to omit end tags from table cells, it just indents all further lines to a crazy degree.
21:10
<zewt>
i don't have autoindent enabled for JS, though; i probably should
21:10
<matijsb1>
especially the re-indenting after de-indenting would drive me absolutely insane
21:10
<AryehGregor>
It sometimes indents JS that has HTML tags embedded in strings or comments, apparently unable to figure out that it's really a string or comment.
21:10
<AryehGregor>
It also always tries indenting the contents of <script> and <style>, and again, re-indents if you de-indent.
21:10
<zewt>
err, I do in Linux but not in Windows gvim, probably need to sync my configs
21:10
<AryehGregor>
I just do ":set indentexpr=" when I get too fed up, but then it does no real indentation at all.
21:11
<AryehGregor>
Maybe there's an alternative script out there someone's written.
21:11
<zewt>
don't see problems with "} else {" with a .js file, though it's pretty dumb about indentation when statement-ending semicolons are missing
21:11
<AryehGregor>
For other languages it's fine, although occasionally annoying (like the PHP formatter always tries to leave whitespace at the end of lines in comments when it auto-breaks lines, and gets confused if you remove it).
21:11
<zewt>
eg. if you say foo() instead of foo(); it indents the next line
21:11
<AryehGregor>
No, .js is okay.
21:12
<AryehGregor>
Well, I don't do that, so I'm fine. :)
21:12
<zewt>
(i'm from a C background so I always stick a semicolon there anyway, so I don't hit that much)
21:12
<zewt>
that's what I mean--if it's sane, javascript inside html should use the same indentation rules as a .js file--not, of course, claiming that it's *actually* sane
21:12
<AryehGregor>
I did most of my early JavaScript for MediaWiki, whose style conventions say to use semicolons in JS, presumably so it looks more PHP-like.
21:12
<zewt>
yeah I see what you're saying for embedded JS
21:13
<zewt>
why would anyone want to make things look more like PHP? :)
21:13
<AryehGregor>
Because MediaWiki is written in PHP.
21:13
<matijsb1>
lol
21:15
<zewt>
ew:
21:15
<zewt>
let js = '<script.*type\s*=\s*.*java'
21:16
<zewt>
not sure what affect that has, but it looks like it doesn't know that <script> = <script type="text/javascript"> now
21:16
<AryehGregor>
Anyone want to write an HTML parser that does incremental updates as the user types? :)
21:16
<AryehGregor>
<script> without a type works for me in vim . . .
21:16
<zewt>
it does seem to work--but something in the HTML indentor, at least, is matching that
21:18
<zewt>
bleh: vim's indentation/highlighting is great when it works, but it makes me want to cut myself whenever I have to look at its configuration
21:18
<AryehGregor>
Yes, the configuration is horrifying.
21:19
<gsnedders>
emacs ftw!
21:19
gsnedders
hids
21:19
<gsnedders>
*hides
21:19
<zewt>
run, coward, run
21:19
<AryehGregor>
The highlighting is almost always superb, much better than I've seen in any other editor (I haven't tried emacs, but assume it's comparably good).
21:19
<AryehGregor>
Although it has a few bugs, like parsing ?> as a PHP end tag for PHP even if it's inside a string literal.
21:19
<AryehGregor>
(comes up fairly often with regex)
21:30
<zewt>
AryehGregor: the } else { problem does go away if I say <script type="text/javascript">
21:30
<AryehGregor>
Oh, really.
21:30
<AryehGregor>
I'll have to try that.
21:30
<espadrine>
zewt: some odd stuff then!
21:30
<zewt>
i'm guessing that can be fixed by fiddling with that regex
21:30
<AryehGregor>
If it works, I'm definitely going to be writing up a patch.
21:31
<zewt>
(the syntax coloring is still totally wrong--that is, not the same as it is in a .JS file)
21:31
<AryehGregor>
It's syntax-colored as JS, although it uses different actual colors.
21:31
<zewt>
(which I find more annoying than the indentation, I should probably look into that)
21:32
<zewt>
yeah, it should be using the exact same ruleset
21:32
<AryehGregor>
FWIW, in PHP, it switches into another language's colors if you use a heredoc. Like: $foo = <<HTML
21:32
<AryehGregor>
<b>Hello!</b>
21:32
<AryehGregor>
HTML;
21:32
<AryehGregor>
The <b>Hello!</b> gets highlighted as HTML.
21:32
<AryehGregor>
So that suggests maybe it's possible to embed another language's ruleset, but maybe it just reimplements everything too.
21:33
<zewt>
http://www.vim.org/scripts/script.php?script_id=3081 this seems to have some improvements, too, though I'm not crazy about forking Vim's configuration stuff--wonder if he knows to send patches
21:33
<zewt>
well ... it's on vim.org, so heh I don't know
21:36
<zewt>
comments in the HTML highlighting script like '" JAVA SCRIPT' don't inspire a whole lot of confidence
21:55
<jgraham>
AryehGregor: Well hober is writing a HTML parser for emacs whcih I assume will provide kickass highlighting and indentation :)
21:55
<jgraham>
Probably not js-in-HTML though
21:55
<AryehGregor>
Not quite enough to get me to consider switching from vim to emacs at this point, I'm afraid.
21:56
<jgraham>
Oh just admit it, you're only using vim because you accidentially started it once and still haven't figured out how to quit it
21:56
<zewt>
bleh guess I should try vim 7.3 before sending any mails about the highlighting stuff
21:58
<AryehGregor>
jgraham, no, I'm using it because the server I originally used didn't have emacs installed, so when I graduated from nano I naturally switched to vim.
21:59
<zewt>
heh
21:59
<zewt>
:help!
21:59
<AryehGregor>
I think I might have even been root, but figured it wasn't worth an apt-get to try out emacs when vim was already installed.
22:04
<zewt>
ever loaded an XPM in vim? heh