IRC log of htmlt on 2013-04-09

Timestamps are in UTC.

15:57:53 [RRSAgent]
RRSAgent has joined #htmlt
15:57:53 [RRSAgent]
logging to http://www.w3.org/2013/04/09-htmlt-irc
15:59:04 [Bin_]
Bin_ has joined #HTMLT
16:00:00 [krisk]
Here is the agenda I posted http://lists.w3.org/Archives/Public/public-html-testsuite/2013Apr/0004.html
16:00:47 [krisk]
We can wait a few more minutes to see if anyone else attends (tobie?)
16:01:52 [mdyck]
mdyck has joined #HTMLT
16:02:31 [Bin_]
present+ Bin_Hu
16:03:40 [krisk]
OK let's get rolling!
16:04:02 [krisk]
* maybe mdyck will come back?
16:04:23 [mdyck]
mdyck has joined #HTMLT
16:04:38 [krisk]
* Indeed mdyck is back
16:05:17 [krisk]
The first agenda item I had was the April F2F at eBay
16:05:28 [krisk]
Will others be attending?
16:06:05 [Bin_]
I will be attending
16:06:26 [krisk]
One item t discuss (part webapps) would be organization of the github repository
16:06:53 [krisk]
For example I noticed some stuff that had approved/submitted folders
16:07:12 [krisk]
Which I didn't think was the intent
16:08:04 [krisk]
bin: good to hear
16:08:14 [Bin_]
Yes, I am listening
16:08:26 [Bin_]
sorry
16:08:49 [Bin_]
so what is your proposal of organizing the github repository?
16:08:53 [krisk]
darobin/jgraham are you following?
16:10:48 [Bin_]
Kris: only you and me here?
16:11:16 [krisk]
I am not proposing and thing per se - just wondering what the 'intent' is since it looks like some stuff is not following the practice that was intially setup
16:12:24 [Bin_]
I guess when some group migrates to github, they may have their own naming convention or convention of folder structure
16:14:53 [Bin_]
Kris: when you mentioned "following the practice", is the practice documented somewhere?
16:15:12 [krisk]
hang tight and I'll add a link
16:17:07 [mdyck]
mdyck has joined #HTMLT
16:21:35 [Bin_]
Kris: are we done here?
16:21:47 [krisk]
Here bin - https://github.com/w3c/html
16:22:17 [krisk]
Now this is for the HTML spec
16:23:15 [Bin_]
I see. Thanks. So the diagram specifies the convention for HTML spec. But other WG may not follow it, right?
16:23:28 [mdyck]
mdyck has joined #HTMLT
16:24:52 [krisk]
Now look at this example pull request https://github.com/w3c/web-platform-tests/pull/69
16:26:04 [krisk]
Then look at https://github.com/w3c/web-platform-tests/tree/CR
16:26:41 [mdyck]
mdyck has joined #HTMLT
16:27:48 [Bin_]
OK, the first pull request added to directory "IndexedDB/submissions/TestTWF_Paris/rhuet/"
16:29:33 [krisk]
Now every so often a script is ran that takes stuff in master and the cr branches and moves them to http://w3c-test.org/web-platform-tests/CR
16:29:38 [krisk]
and http://w3c-test.org/web-platform-tests/master
16:30:15 [krisk]
If you look at http://w3c-test.org/web-platform-tests/master/ you'll see all kinds of stuff that is not specific to the HTML WG
16:31:39 [krisk]
Now back in December it was discussed and CR would map to the HTML5 spec and HTML5.1 would be in 'master'
16:31:58 [Bin_]
I hear you. So irrelevant stuff were put there
16:32:21 [krisk]
Maybe irrelevant is too strong of a word :)
16:32:46 [Bin_]
sorry :)
16:32:53 [krisk]
I think it's valueable to have a known good set of tests, thus the 'CR' branch
16:33:29 [Bin_]
I think so too. We need a known, stable, and goos set of test suites
16:33:32 [mdyck]
mdyck has joined #HTMLT
16:33:51 [mdyck2]
mdyck2 has joined #htmlt
16:33:58 [krisk]
Though if stuff just gets merged right into master then master starts to lose value since it's not clear that the tests are correct
16:34:17 [Bin_]
Do you want to discuss it in F2F in San Jose? Request an agenda item there?
16:34:29 [krisk]
yes
16:35:16 [krisk]
For example we have some test suites in the 'master' branch and the spec is at REC
16:35:18 [krisk]
e.g. http://w3c-test.org/web-platform-tests/master/SelectorsAPI/Status.html
16:35:27 [Bin_]
I support to discuss it, because a stable, known good test suite is essential to vendors and service providers to have more confidence of the test result
16:35:47 [krisk]
Which I'd expect to have this set of tests in the 'CR' branch
16:36:12 [krisk]
but it's only in the 'master' branch
16:36:41 [krisk]
OK, thanks
16:36:42 [Bin_]
That's right. Once spec is REC, and test suite is stable, "master" branch should be "locked"
16:37:43 [krisk]
The next agenda item is that their is a test the web forward event this Friday and Saturday in Seattle
16:38:03 [Bin_]
"CR" branch should be open to those CR spec, and the new tests that have been reviewed and known good
16:38:41 [Bin_]
And the tests for others should be reviewed and ratified first before puttint it to CR
16:39:08 [Bin_]
Yes, from our AT&T side, Bryan, Dan and Jen will be there
16:39:14 [krisk]
yes
16:39:24 [krisk]
at the F2F?
16:39:27 [mdyck]
mdyck has joined #HTMLT
16:39:37 [krisk]
Here is a link to the event http://testthewebforward.org/
16:39:45 [Bin_]
I meant "Test the Web Forward" event in Seattle
16:39:52 [krisk]
Oh nice!
16:40:16 [krisk]
I'll be asking for some people to focus on writting HTML5 tests
16:40:18 [Bin_]
I won't be there, because I am located in San Jose area.
16:40:30 [krisk]
no problem
16:41:02 [Bin_]
So I will definitely be attending everything in San Jose area, like this F2F and any future events
16:41:09 [krisk]
cool
16:41:32 [krisk]
The last agena item was to do some more spec classification of the HTML5 spec
16:41:49 [Bin_]
This time, we may not be able to contribute HTML5 tests to this Test the Web ofrward event in Seattle
16:41:54 [krisk]
starting with http://www.w3.org/TR/html5/infrastructure.html#conformance-requirements
16:42:02 [krisk]
any reason?
16:42:34 [krisk]
CSS WG also needs some tests to move a few spec forward (background, border radius)
16:43:44 [Bin_]
we are trying, but cannot promise anything
16:43:52 [Bin_]
resources and time constraint
16:44:05 [krisk]
IMHO though 2.2 Conformance requirements has some 'MUST's it's just definitions and no actual tests are needed
16:44:27 [krisk]
Which is not clear on the coverage report http://w3c-test.org/html-testsuite/master/tools/coverage/
16:44:36 [mdyck]
mdyck has joined #HTMLT
16:45:29 [krisk]
does anyone disagree?
16:45:32 [Bin_]
That's right
16:45:37 [Bin_]
agre
16:45:42 [krisk]
mdyck?
16:45:50 [mdyck2]
yes?
16:46:09 [krisk]
OK lets move on to http://www.w3.org/TR/html5/infrastructure.html#conformance-classes
16:47:24 [krisk]
This is a bigger section and the coverage report lists that it has no tests and normative requirements exist
16:48:07 [mdyck]
mdyck has joined #HTMLT
16:48:37 [krisk]
Indeed this part of the spec needs some tests
16:48:58 [krisk]
For example
16:48:59 [krisk]
A conforming XHTML processor would, upon finding an XHTML script element in an XML document, execute the script contained in that element. However, if the element is found within a transformation expressed in XSLT (assuming the user agent also supports XSLT), then the processor would instead treat the script element as an opaque element that forms part of the transform.
16:49:12 [krisk]
and...
16:49:13 [krisk]
Web browsers that support the XHTML syntax must process elements and attributes from the HTML namespace found in XML documents as described in this specification
16:50:23 [Bin_]
Good catch
16:51:17 [krisk]
So specifically in this section "Web browsers and other interactive user agents" need a few tests
16:51:31 [Bin_]
Agree
16:51:39 [krisk]
Conforming HTML5 documents just looks like a definition
16:52:06 [Bin_]
That's right
16:52:09 [krisk]
Same with 'Non-interactive presentation user agents'
16:52:18 [mdyck]
mdyck has joined #HTMLT
16:53:13 [krisk]
Visual user agents that support the suggested default rendering is also a definition with the expectation that tests will come from the 'rendering section' of the spec
16:53:22 [krisk]
e.g. http://www.w3.org/TR/html5/rendering.html#rendering
16:53:32 [Bin_]
Correct
16:53:59 [krisk]
Looks like User agents with no scripting support is also a definition
16:54:11 [mdyck2]
is the test suite just for testing user agents, or also for testing things like conformance-checkers?
16:54:31 [krisk]
That is a good question
16:54:58 [krisk]
The role of the test suite to make sure that two or more implementations can be created that are interoperable
16:55:27 [krisk]
So I would presume that we would want/expect a conformance checker to be able to be created using the spec
16:56:01 [krisk]
So you would need a test for a conformance checker - maybe a page that then is ran by the checker
16:56:55 [krisk]
...itself and would report conformance issues
16:57:26 [Bin_]
Is the Conformance Checker itself considered a sort of "test suite"?
16:58:07 [krisk]
Let's talk about this at the next meeting, it's alomst 10am (pacific time) and I need to head to another meeting
16:58:17 [mdyck2]
ok
16:58:24 [Bin_]
Me too, another meeting in 2 minutes
16:58:36 [krisk]
Since I expect this will not quick converstation
16:58:38 [Bin_]
Thank you Kirs, and see you in San Jose
16:58:44 [krisk]
OK let's adjourn
16:59:00 [krisk]
sounds good
16:59:15 [krisk]
RRAAgent, make logs public
16:59:21 [krisk]
RRSAgent, make logs public
16:59:46 [krisk]
rrsagent, generate minutes
16:59:46 [RRSAgent]
I have made the request to generate http://www.w3.org/2013/04/09-htmlt-minutes.html krisk
17:03:00 [mdyck]
mdyck has left #HTMLT
18:17:17 [Ms2ger]
Ms2ger has joined #HTMLT