02:22
<JonathanNeal>
Hello!
02:31
<Hixie>
MikeSmith: yeah. though really people _should_ real the rationale before they comment.
02:31
<Hixie>
JonathanNeal: hello
02:31
<Hixie>
JonathanNeal: btw if you have any questions about the sort stuff don't hesitate to ping me
02:31
<JonathanNeal>
Hixie: I wrote up a little something, it doesn’t address all the use cases, but it works.
02:32
<Hixie>
the spec doesn't, or your impl doesn't?
02:33
<JonathanNeal>
As of right this second, my implementation is limited to ascending alphabetical sorting.
02:34
<Hixie>
ah
02:34
<Hixie>
the autosorting algorithm is the best part of the proposal :-)
02:39
<JonathanNeal>
Hixie: it’s kind of the point, yes?
02:40
<caitp>
i didn't read all of the details of the proposal, does it offer different sorting algorithms or even custom ones for people who want those?
02:41
<caitp>
well not so much algorithms
02:41
<caitp>
but comparison routines
02:42
<Hixie>
JonathanNeal: well, the point is sorting. it could do the job with a dumb algorithm, or with a selector for picking a simple algorithm, or with a callback or something
02:43
<JonathanNeal>
caitp: if it does, i missed that part, and was wondering the same thing. It has a lot of built in sorting rules. The thing that was most vague was how table[sortable] might be exposed to the ui.
02:43
<Hixie>
JonathanNeal: but instead it has the algorithm it has, which imho is pretty cool.
02:43
<Hixie>
caitp: i encourage you to read the spec
02:43
<Hixie>
JonathanNeal: re the UI, the idea was just that you'd click the <th> to sort by it.
02:43
<caitp>
i'll read it if i have time to hack together an implementation for blink :p
02:44
<caitp>
but my schedule is pretty tied up the next few months :(
02:44
<Hixie>
that is true of all of us
02:46
<JonathanNeal>
Let me share what I have. It’s the first draft, but it’s fun to see, maybe just for me though.
02:47
<JonathanNeal>
http://sandbox.thewikies.com/table-sorting/
02:48
<caitp>
cool stuff
02:49
<JonathanNeal>
I figured the standard sorting is based on the visible content of the cells, eg innerText and not textContent.
02:49
<Hixie>
JonathanNeal: if you're planning on implementing the spec's algorithm, i have a test dataset somewhere i can mail you if you want to test the spec against what i think the spec says
02:49
<JonathanNeal>
Yes, that would be wonderful.
02:50
<Hixie>
JonathanNeal: send me an e-mail, i'll find it and mail it to you (maybe monday when i get to work)
02:51
<JonathanNeal>
Sure thing.
02:52
<caitp>
it doesn't look like there's very much in the wpt about that feature yet, apart from idl assertions
02:52
<caitp>
you could probably contribute some while working on the polyfill
02:55
<JonathanNeal>
There are many steps. So far this: finds eligible tables, fires and checks the status of a sort event, gets the order of the heading cells, orders the body cells by that heading cell order. Thanks to Mr. DOMSubtreeModified, we can respond to changes instantly.
02:58
<JonathanNeal>
Oh and there’s a good catch, the event can’t remove the sortable attribute from the table, so it needs to be checked again. That’s wisely in the spec.
05:48
<Hixie>
JonathanNeal: sent
09:08
<smaug____>
annevk: why was get() added to Notification?
09:10
<smaug____>
looks like because of https://bugzilla.mozilla.org/show_bug.cgi?id=899574#c0
09:10
<smaug____>
rather error prone API
09:12
<annevk>
hmm, no better suggestions so far
09:28
<smaug____>
thinking..
09:40
<smaug____>
well, making click cancelable doesn't really make the API any worse
09:41
<smaug____>
but would be really really nice if one user input wasn't represented as several click events
09:44
<jgraham>
gsnedders: Yeah, we need a list of tests that are expected to fail. It's the only sane way to deal with the situation where you have tests from "upstream"
12:54
<gsnedders>
jgraham: not entirely clear how to implement that, though :(
13:22
<jgraham>
gsnedders: It's not impossible though. You just need to make up an id for the tests
13:23
<gsnedders>
jgraham: indeed
14:23
<smaug____>
annevk: have you seen this http://mxr.mozilla.org/mozilla-central/source/dom/webidl/XMLHttpRequest.webidl?mark=53-57#53
14:24
<smaug____>
apparently it was added in https://bugzilla.mozilla.org/show_bug.cgi?id=788369
14:24
<smaug____>
(just noticed while looking at some random .webidl files)
14:29
<caitp>
hah
15:42
<annevk>
hmm, have not, but that seems obsolete given that the constructor no longer takes any argument
15:47
<annevk>
caitp: the way <base> works is perfectly normal btw per URL parsing rules
15:47
<caitp>
it's not that it's anomalous or anything, it's just weird
15:48
<caitp>
"oh you want to go to /home/caitp/home/caitp? okay I'll send you to /home/caitp instead because I'm confused"
15:49
<caitp>
second guessing what the programmer wants, that's the weird part
15:53
<gsnedders>
"If all the bytes up to the last byte converted by the current decoder have the same Unicode interpretations in both the current encoding and the new encoding" - how do you check what characters have been seen?
15:53
<gsnedders>
do people just store whether non-ASCII chars have been seen and have a table?
16:11
<gsnedders>
html5lib is all over the place with spec complaince :(
16:11
<gsnedders>
up-to-date in places and nowhere elsewhere
16:26
<gsnedders>
jgraham: the big problem with the up to date tests is that we implement the scripting enabled case but don't support scripting, so the #script tests fail
20:18
<erlehmann>
JonathanNeal have you written a polyfill for table sortable?
21:01
<JonathanNeal>
erlehmann: yes, but it's still in development.
21:02
<erlehmann>
JonathanNeal show and tell! :)
21:08
<JonathanNeal>
Ok
21:09
<JonathanNeal>
http://sandbox.thewikies.com/table-sorting/ erlehmann
21:13
<erlehmann>
thx JonathanNeal
21:22
<JonathanNeal>
erlehmann: thanks. what would you like to see?
21:24
<paulohp>
hello guys! do you have more content about streams? eg.: https://whatwg.github.io/streams/
21:25
<erlehmann>
JonathanNeal ▼ U+25BC BLACK DOWN-POINTING TRIANGLE in header of primary sorting column. ▽ U+25BD WHITE DOWN-POINTING TRIANGLE in header of other columns.
21:25
<erlehmann>
using ▲ U+25B2 BLACK UP-POINTING TRIANGLE for reverse sort.
21:25
<erlehmann>
like wikipedia does