00:38
<kennyluck>
Firefox does have this triple store implementation. Just wonder how old it is.
00:39
<kennyluck>
I missed the discussion I should have joined :(
01:53
<JonathanNeal>
Woot
01:53
<JonathanNeal>
http://sandbox.thewikies.com/createElement/
01:53
<JonathanNeal>
vs http://sandbox.thewikies.com/createElement/before.html
01:54
<TabAtkins>
Should I be looking at that in IE?
01:56
<TabAtkins>
Also: your loop syntax is so horrible. ;_; I really don't understand why you value a microsecond or two over the clarity of a for loop.
01:58
<TabAtkins>
Otherwise, though, seems very clever. I approve.
02:03
<TabAtkins>
Argh, I'm introducing too much abstraction here, but I'm not sure of the best way to collapse one of the abstraction levels. >_<
02:03
<TabAtkins>
Anytime you have to introduce *two* new levels of abstract directional terminology, you're doing something wrong.
03:23
<jlebar>
Hixie, Is there a reason that offline manifests don't treat form feed characters (and other unicode space characters) as whitespace?
03:24
<jlebar>
e.g. step 17 in http://www.whatwg.org/specs/web-apps/current-work/multipage/offline.html#writing-cache-manifests
03:24
<annevk>
consistency
03:24
<jlebar>
annevk, with what?
03:24
<annevk>
text/plain, HTML, CSS, etc.?
03:25
<jlebar>
the split on spaces algorithm does consider ff to be whitespace.
03:26
<jlebar>
http://www.whatwg.org/specs/web-apps/current-work/multipage/common-microsyntaxes.html#skip-whitespace
03:26
<annevk>
sure, but newlines are always just CR and LF or CRLF
03:27
<jlebar>
Right. But the step in the algorithm I first mentioned is stripping whitespace off the end of a line.
03:29
<annevk>
aah I see now
03:29
<annevk>
sorry
03:30
<jlebar>
np
03:30
<annevk>
seems like a bug of some sorts
03:30
<jlebar>
Perhaps. But Hixie may have some grand plan. He often does. :)
03:32
<annevk>
maybe he tries to keep FF out of cache manifests but the inconsistency seems confusing
07:32
<Hixie>
anyone remember what the term is for a block's overall bidi direction?
07:35
<annevk>
base writing direction of blocks?
07:35
<annevk>
(comes from a google search)
07:35
<Hixie>
doesn't sound right
07:36
<Hixie>
i mean the post-bidi-algorithm direction
07:36
<Hixie>
the one that decides whether 'text-align:start' is left or right
07:37
<Hixie>
wait, text-align is dependent on 'start', not on the actual direction?
07:37
<Hixie>
er, on 'direction', i mean, not 'start'
07:37
<annevk>
dunno, sorry
07:38
<annevk>
haven't followed that stuff recently since in discussions vertical text is added on top of it which makes my brain explode (or at least wander off)
07:38
<Hixie>
i guess i mean the value that decides whether a "." at the end of the string goes to the left or the right in the rendering
07:41
<Hixie>
looks like i mean the direction of the character found in the bidi algorithm step P2
07:42
<Hixie>
which seems to be called the "paragraph direction"
07:43
<Hixie>
ah, the paragraph embedding level
07:43
<Hixie>
there we go
07:45
<Hixie>
man it's been so long since i've worked on bidi stuff
07:49
<Hixie>
annevk: yeah, bidi spec says it's sometimes called the "base direction", thanks
07:53
<annevk>
what are you working on btw? there's no open bidi issues afaict (apart from the unfinished i18n document I guess)
07:55
<tagawa>
hello!
07:55
<annevk>
hey tagawa :)
07:55
<tagawa>
(sorry, just testing)
07:56
<Hixie>
annevk: positioning of subtitles
07:58
<annevk>
ah right, forgot about captioning
07:58
<annevk>
o_O
07:58
<Hixie>
heh
09:14
<nessy>
good that somebody with knowledge looks at i18n issues for subtitles!
09:42
<JonathanNeal>
Updated http://sandbox.thewikies.com/createElement/before.html and http://sandbox.thewikies.com/createElement/ to include multiple libraries working with html5 elements successfully.
09:44
<jgraham>
JonathanNeal: Which code is the code that you actually wrote?
09:44
<jgraham>
i.e. what should I be looking at?
09:47
<JonathanNeal>
I wrote the tests, and I wrote the "shimproof" script.
09:47
<JonathanNeal>
It is an evil script that proves a point, but beyond that, it should not be used. Rather, it should be implemented into js libraries.
09:47
<jgraham>
You know it would be friendly to include a non-minimised version if you want people to read it :)
09:48
<JonathanNeal>
I thought shimproof.min.js made it obvious, but sure http://sandbox.thewikies.com/createElement/shimproof.js
09:48
<JonathanNeal>
It's just an experiment :D
09:49
jgraham
wonders if he should propose a way of accessing the AST from javascript to TC39 so that it is easy to implemnt pretty-printers
09:49
<jgraham>
(as well as all sorts of other exciting things)
09:49
<jgraham>
(I guess they would hate it though)
09:52
<Philip`>
Why do you need a standardised API for implementing pretty-printers?
09:52
<Philip`>
Isn't it fine to just have a standalone pretty-printing tool that uses private APIs of a specific implementation?
09:55
<jgraham>
Philip`: I don't really. But it's not like there are lots of existing implementations that expose private APIs to do what I want either
09:56
<Philip`>
Getting an implementation to expose a private API sounds like far less effort than getting every implementation to expose a public API
09:57
<jgraham>
It's really unclear that that is true :)
09:58
<jgraham>
Also I guess it would be nice if you don't need java in order to use most existing minifiers
09:58
<jgraham>
*didn't
09:59
<Philip`>
How can it not be true? The public API is as much effort as doing the private API multiplied by the number of implementations, plus the cost of coordination
10:00
<jgraham>
Because the cost to me isn't the implementation cost it is the cost of convincing people
10:00
<Philip`>
I was assuming you weren't being so selfish as to only consider the cost to yourself ;-)
10:01
<Philip`>
(Given the apparent pain and suffering involved with SpiderMonkey's decompiler, I doubt implementors want to do much more of that kind of thing - computing and storing perfect lossless ASTs seems to conflict with optimisation goals, and deriving an AST from more optimised data structures is hard)
10:03
<jgraham>
JonathanNeal: Why the use of the foo[constant] syntax everywhere rather than foo.constant? The latter is more readable and, at least in some modern engines, more performant (IE might be different of course)
10:04
<JonathanNeal>
Just to make the script under 1k minified
10:04
<JonathanNeal>
It doesn't really matter.
10:06
<Dashiva>
http://www.youtube.com/watch?v=MHrTE-PLAsc&feature=related#t=6m20s
10:07
<Dashiva>
I wonder if you can justify subtitling something people wouldn't normally be able to hear properly anyway
10:23
<webben>
Dashiva: Sure. TV often does that when voices are hard to hear for some reason (e.g. bad recording).
10:41
<Philip`>
(It's interesting to see when TV shows consider English-speaking people to have a sufficiently peculiar accent that they have to be subtitled)
10:43
<Philip`>
(e.g. it seems Americans sometimes find Scottish accents tricky)
12:12
<Dashiva>
Philip`: There's a funny video like that
12:12
<Dashiva>
http://www.dailymotion.com/video/x3084g_newsreportfromiraq_fun
12:26
<hsivonen>
http://en.wikipedia.org/wiki/HTML5_video doesn't seem to conform to Wikipedia policies
12:27
<hsivonen>
lots of [citation needed] and [weasel words] missing there
12:27
<Philip`>
Dashiva: I like how the subtitles get completely obscured by the popup advert
12:33
<mr_daniel>
I wonder where the 'dblclick' event is specified? I just read the DOM Level 2 Events recommendation, specifically the mouse events. The 'dblclick' event is not described: http://www.w3.org/TR/DOM-Level-2-Events/events.html#Events-MouseEvent
12:34
<mr_daniel>
but on the other hand the wikipedia article about DOM Level 2 Events states that there is a 'dblclick' event http://en.wikipedia.org/wiki/DOM_events#HTML_events
12:35
<mr_daniel>
so where does the 'dblclick' event come from? Is it supported by all browsers, which support DOM Level 2 Events? Or should I avoid the use of the 'dblclick' event?
12:48
<nessy>
Dashiva - hehe - I really enjoyed that video ;)
13:31
<Henrik`G>
hey peoples i have a box with links of latest posts in my sidebar
13:31
<Henrik`G>
and i am wondering if it is a section or a div would be the best use
13:31
<Henrik`G>
my sense says section because it starts with a header and then th ul
13:31
<Henrik`G>
the
15:22
<ray>
hehe, [[deep pockets]]
17:40
<Dashiva>
I think Dean needs a timeout...