Options:
- # Session Start: Fri Jul 12 00:00:01 2013
- # Session Ident: #testing
- # [00:01] * Joins: wilhelm (~wilhelm@public.cloak)
- # [00:15] * Quits: AutomatedTester (~AutomatedTester@public.cloak) (Client closed connection)
- # [00:45] * Joins: mdas (~mdas@public.cloak)
- # [00:46] * Joins: mdas_ (~mdas@public.cloak)
- # [00:50] * Quits: mdas__ (~mdas@public.cloak) (Ping timeout: 180 seconds)
- # [00:52] * Quits: mdas (~mdas@public.cloak) (Ping timeout: 180 seconds)
- # [01:04] * Joins: mdas (~mdas@public.cloak)
- # [01:10] * Quits: mdas_ (~mdas@public.cloak) (Ping timeout: 180 seconds)
- # [01:11] * Quits: mdas (~mdas@public.cloak) (Ping timeout: 180 seconds)
- # [01:12] * Joins: mdas (~mdas@public.cloak)
- # [01:28] * Joins: schuki (~quassel@public.cloak)
- # [01:37] * Joins: darobin_ (rberjon@public.cloak)
- # [01:37] * Quits: darobin (rberjon@public.cloak) (Client closed connection)
- # [02:21] * Joins: mdas_ (~mdas@public.cloak)
- # [02:25] * Disconnected
- # [02:26] * Attempting to rejoin channel #testing
- # [02:26] * Rejoined channel #testing
- # [02:26] * Topic is 'http://krijnhoetmer.nl/irc-logs/testing/'
- # [02:26] * Set by Ms2ger on Wed Jun 26 11:22:44
- # [02:26] * Joins: mdas__ (~mdas@public.cloak)
- # [02:27] * Quits: mdas (~mdas@public.cloak) (Ping timeout: 180 seconds)
- # [02:28] * Quits: mdas_ (~mdas@public.cloak) (Ping timeout: 180 seconds)
- # [02:32] * Quits: krijn (~krijnhoetmer@public.cloak) (Ping timeout: 180 seconds)
- # [02:33] * Quits: jhammel (~jhammel@public.cloak) ("leaving")
- # [02:43] * Joins: mdas (~mdas@public.cloak)
- # [02:50] * Quits: mdas__ (~mdas@public.cloak) (Ping timeout: 180 seconds)
- # [02:50] * Quits: mdas (~mdas@public.cloak) (Ping timeout: 180 seconds)
- # [04:03] * Joins: glenn_ (~gadams@public.cloak)
- # [04:07] * Joins: shepazu (schepers@public.cloak)
- # [04:08] * Quits: glenn (~gadams@public.cloak) (Ping timeout: 180 seconds)
- # [04:46] * Quits: darobin_ (rberjon@public.cloak) (Client closed connection)
- # [04:46] * Joins: darobin (rberjon@public.cloak)
- # [05:33] * Joins: mdas (~mdas@public.cloak)
- # [05:38] * Joins: mdas_ (~mdas@public.cloak)
- # [05:45] * Quits: mdas (~mdas@public.cloak) (Ping timeout: 180 seconds)
- # [05:45] * Quits: mdas_ (~mdas@public.cloak) (Ping timeout: 180 seconds)
- # [05:54] * Joins: mdas (~mdas@public.cloak)
- # [05:55] * Joins: mdas_ (~mdas@public.cloak)
- # [06:01] * Quits: mdas (~mdas@public.cloak) (Ping timeout: 180 seconds)
- # [06:31] * Joins: gitbot (~gitbot@public.cloak)
- # [06:31] -gitbot:#testing- [web-platform-tests] sgrekhov synchronize pull request #135: Submission/unipro (master...submission/unipro) https://github.com/w3c/web-platform-tests/pull/135
- # [06:31] * Parts: gitbot (~gitbot@public.cloak) (gitbot)
- # [06:49] * Joins: mdas (~mdas@public.cloak)
- # [06:55] * Quits: mdas_ (~mdas@public.cloak) (Ping timeout: 180 seconds)
- # [06:56] * Quits: mdas (~mdas@public.cloak) (Ping timeout: 180 seconds)
- # [07:47] * heycam|away is now known as heycam
- # [08:06] * Joins: jwang (~jwang@public.cloak)
- # [08:09] * Quits: glenn_ (~gadams@public.cloak) (Client closed connection)
- # [08:59] * Quits: jwang (~jwang@public.cloak) (Ping timeout: 180 seconds)
- # [09:16] * Joins: dom (dom@public.cloak)
- # [09:27] * Quits: Lachy (~Lachy@public.cloak) ("Computer has gone to sleep.")
- # [09:55] * heycam is now known as heycam|away
- # [10:04] * Joins: Lachy (~Lachy@public.cloak)
- # [10:20] * Joins: Ms2ger (~Ms2ger@public.cloak)
- # [10:34] * Joins: tobie (tobie@public.cloak)
- # [11:17] * Joins: AutomatedTester (~AutomatedTester@public.cloak)
- # [11:18] * Quits: AutomatedTester (~AutomatedTester@public.cloak) (Client closed connection)
- # [11:18] * Joins: AutomatedTester (~AutomatedTester@public.cloak)
- # [11:30] <wilhelm> 'Allo. (c:
- # [11:34] <AutomatedTester> allo wilhelm
- # [11:35] * AutomatedTester is doing all his admin now to get it out the way
- # [11:36] <wilhelm> Ditto. Need to get today's share of my day job out of the way; then test automation. (c:
- # [11:39] <andreastt> I have a couple of more work patches to land before I can do anything useful.
- # [11:42] <AutomatedTester> tobie: hey, are you doing TTWF stuff or is it all RHauck?
- # [11:43] <tobie> AutomatedTester: events?
- # [11:43] <AutomatedTester> yea
- # [11:43] <tobie> Depends on the event. I try to stay out of it as much as possible, I have plenty on my plate already. :)
- # [11:44] <tobie> So usually, rhauck and other Adobe folks involved.
- # [11:45] <tobie> but some events organized by Google, MSFT, etc.
- # [11:45] <tobie> AutomatedTester: how can I help.
- # [11:45] <tobie> s/./?/
- # [11:45] <AutomatedTester> tobie: I wanted to hijack one of their events :D
- # [11:45] <AutomatedTester> I will email rhauck
- # [11:46] <tobie> AutomatedTester: which event? what kind of hi-jacking?
- # [11:46] <AutomatedTester> tobie: the one before TPAC
- # [11:46] <AutomatedTester> and to get WebDriver conformance tests
- # [11:46] <tobie> AutomatedTester: we probably need a mailing list to coordinate.
- # [11:46] <tobie> AutomatedTester: sounds feasible.
- # [11:47] <AutomatedTester> tobie: well I will email her about it and then if she doesnt have a plan then we can bring to the mailing list
- # [11:47] <AutomatedTester> sound good?
- # [11:47] <tobie> sure. /cc me.
- # [11:47] <AutomatedTester> ok
- # [11:48] <tobie> Also, we need a mailing list. :)
- # [11:48] <AutomatedTester> we have browser testing
- # [11:48] <AutomatedTester> and TTWF has a mailing list
- # [11:48] <AutomatedTester> there are 2
- # [11:48] <AutomatedTester> but no point putting the horse before the cart if we arent ready
- # [11:48] <tobie> yeah, none of which fit the bill of having a dedicated mailing list for event coordination.
- # [11:49] <tobie> (I get about 2 emails a week on the subject).
- # [11:49] <AutomatedTester> i'll add a p.s to it ;)
- # [11:49] <tobie> oh, please do!
- # [11:49] <tobie> thanks
- # [12:06] * Joins: abarsto (~abarsto@public.cloak)
- # [12:06] * abarsto is now known as ArtB
- # [12:07] <AutomatedTester> is it ok to sign up for TPAC now?
- # [12:12] <wilhelm> Yes. I should send a ping to the list encouraging people to do so.
- # [12:12] <wilhelm> I signed up already.
- # [12:14] <AutomatedTester> ok cool
- # [12:14] <AutomatedTester> I just noticed its all set
- # [12:14] <AutomatedTester> setup*
- # [12:14] <wilhelm> We should grab some observer seats in semi-random other groups to keep an eye on what they're up to, and see if their work has any implications on ours.
- # [12:15] <AutomatedTester> definitely
- # [12:15] <wilhelm> If you can stick around for more than just our meeting, that's encouraged. (c:
- # [12:15] <AutomatedTester> it should be doable
- # [12:37] * Joins: sstewart6 (~simons@public.cloak)
- # [12:37] <sstewart6> Greetings everyone
- # [12:37] <AutomatedTester> greetings
- # [12:38] <wilhelm> 'Ello!
- # [12:38] <sstewart6> Wotcha
- # [12:38] <sstewart6> I'm in the middle of perf season so may disappear occasionally to write up stuff about colleagues, but I'll be here all day
- # [12:39] <AutomatedTester> ahh I start that next week apparently
- # [12:39] <sstewart6> *shudder*
- # [12:40] <sstewart6> I hatses the perf
- # [12:40] <AutomatedTester> I am only a tech lead with no underlings so it should be simples for me
- # [12:41] <sstewart6> ahahahahahahahahahahahahaha
- # [12:42] <andreastt> sstewart6: \o
- # [12:43] <sstewart6> yo
- # [12:43] <andreastt> So.
- # [12:43] <andreastt> AutomatedTester: Did you check in your runner yet?
- # [12:44] <AutomatedTester> andreastt: not yet sorry
- # [12:44] <andreastt> It's quite painful to run the tests as it is now, so we should consider preserving the WD session.
- # [12:45] <AutomatedTester> andreastt: we will
- # [12:46] <andreastt> This thing does the trick for unittest: https://gist.github.com/andreastt/5875264
- # [12:46] <andreastt> But I don't like unittest.
- # [12:46] <wilhelm> My day job merely involves performance testing of software today. That's a bit less painful
- # [12:48] <andreastt> What's the link to the repo BTW?
- # [12:48] <andreastt> s/the repo/the new repo/
- # [12:48] <andreastt> Also, what's the process for landing patches now? We send it to critic?
- # [12:49] <wilhelm> andreastt: https://github.com/w3c/web-platform-tests
- # [12:50] <wilhelm> It pulls in the kitchen sink. Which, for a change, is a good thing. Since _we're_ the kitchen sink.
- # [12:50] <sstewart6> Feel free to add me PRs
- # [12:50] <sstewart6> Will happily review stuff and things :)
- # [12:51] <andreastt> Ugh, we have to go through Github's PR system?
- # [12:51] <andreastt> )-:
- # [12:51] <wilhelm> We should probably get the usual suspects here direct access.
- # [12:51] <sstewart6> I have the commit bit
- # [12:52] <Ms2ger> We tend to prefer PRs to allow review
- # [12:52] * andreastt would massively prefer jgraham's critic instance
- # [12:53] <Ms2ger> andreastt, then you need PRs
- # [12:53] <Ms2ger> Every PR gets a critic review
- # [12:53] <wilhelm> Yeah, there's a critic bot involved. (c:
- # [12:53] <andreastt> I see.
- # [12:54] <wilhelm> andreastt: Here's an example: https://github.com/w3c/web-platform-tests/pull/246
- # [12:55] <Ms2ger> If there are people whose tests don't need to be reviewed, I think I can hand out commit bits
- # [12:56] <andreastt> We haven't really done any review on the tests that are there now.
- # [12:56] <andreastt> They're primarily from me, but I do think we need to review both them and any new tests.
- # [12:57] <wilhelm> Ms2ger: For the people in this conversation, for the subset of tests in the /webdriver directory, reviews should not be necessary at this point in time.
- # [12:57] <Ms2ger> I'm getting conflicting messages here :)
- # [12:58] <sstewart6> I'm happy to loop through reviews
- # [12:58] <andreastt> Me too.
- # [12:58] <AutomatedTester> sstewart6: you get emails for webdriver PRs
- # [12:58] <AutomatedTester> sstewart6: and so do I
- # [12:58] <sstewart6> Right
- # [12:59] <wilhelm> Alright. Ignore my previous comment. (c:
- # [12:59] <andreastt> I've set myself up as a reviewer for /webdriver in critic.
- # [12:59] <andreastt> So presumably I'll also get emails now.
- # [12:59] <AutomatedTester> should do
- # [12:59] <AutomatedTester> I only care about critic emails now
- # [12:59] <AutomatedTester> too much noise on web-platform-tests
- # [13:01] <AutomatedTester> AFAIK only sstewart6 and I can push the webdriver tests
- # [13:01] <AutomatedTester> but anyone can review
- # [13:06] <andreastt> sstewart6: So I don't get the parentNodeVisibleWhenAllChildrenAreAbsolutelyPositionedAndOverflowIsHidden test in Selenium.
- # [13:06] <sstewart6> Hang on a tick. Lemme fire up an IDE and have a look at it
- # [13:06] <sstewart6> Did I write that?
- # [13:06] <andreastt> We're not checking the parent node at all in that test AFAICT.
- # [13:07] <sstewart6> IDE almost started
- # [13:07] <andreastt> Yes you did.
- # [13:07] <andreastt> sstewart6: My emacs-init-time says 0.7 seconds (-:
- # [13:07] <AutomatedTester> please leave your flame wars at the door
- # [13:07] <AutomatedTester> ;)
- # [13:08] <sstewart6> Hmmm…we do appear to be missing a node there
- # [13:08] <andreastt> Either that or the title of the test is wrong.
- # [13:08] <andreastt> Not really sure what it's intending to test.
- # [13:11] <sstewart6> Given the length of the title, I'd suggest that the test is wrong
- # [13:11] <sstewart6> Or there's a missing test
- # [13:11] <andreastt> Do you remember what it was meant to cover?
- # [13:12] <sstewart6> I know my name is on there, but I don't think I actually wrote that test
- # [13:13] <andreastt> Looks like something that came in via the issue tracker (-:
- # [13:13] <sstewart6> Ah.
- # [13:13] <AutomatedTester> I have a sinking feeling thats my fault
- # [13:13] <sstewart6> I guess we wanted to make sure that the parent node was visible even when all children were positioned absolutely
- # [13:14] <sstewart6> And when the parent of the parent node had overflow "hidden" set
- # [13:14] <sstewart6> I guess that means that we'd like to be sure that the node can actually be seen
- # [13:15] <sstewart6> Presumably at some point we were claiming that "Hello world, I like cheese" wasn't actually visible
- # [13:16] <andreastt> Okay. I can write a test for that.
- # [13:18] * Joins: glenn (~gadams@public.cloak)
- # [13:24] <AutomatedTester> andreastt: if you find issue with the spec please feel free to raise bugs against it https://www.w3.org/Bugs/Public/enter_bug.cgi?product=Browser%20Test%2FTools%20WG&content=webdriver
- # [13:24] <AutomatedTester> that way we can come back later and fix them
- # [13:26] <andreastt> So far I haven't really been concentrating on the spec as much as porting over Selenium tests.
- # [13:27] <andreastt> Hence the need to review element_state/visibility_test.py against the spec.
- # [13:27] <AutomatedTester> awesome
- # [13:27] <AutomatedTester> ahh that reminds me
- # [13:28] <AutomatedTester> I need to add that hidden thing we discussed yesterday
- # [13:28] <AutomatedTester> i will raise a bug now
- # [13:28] * Joins: Lachy_ (~Lachy@public.cloak)
- # [13:28] * Quits: Lachy (~Lachy@public.cloak) (Ping timeout: 180 seconds)
- # [13:36] * Quits: Lachy_ (~Lachy@public.cloak) ("Computer has gone to sleep.")
- # [13:36] <AutomatedTester> wilhelm: can you add to your todo about TPAC
- # [13:37] <AutomatedTester> I think it would be best if the email comes from you
- # [13:38] <wilhelm> Yup.
- # [13:39] <AutomatedTester> thanks!
- # [13:42] <AutomatedTester> ok, just having a quick break and then will do some code
- # [13:49] <andreastt> sstewart6: This test is kind of similar to the MAP outside of viewport scenario that gdennis gave.
- # [13:50] <andreastt> I'm not sure we should consider the parent visible…
- # [13:50] <andreastt> https://gist.github.com/andreastt/5983873
- # [14:05] <sstewart6> The test seems to make sense
- # [14:06] <sstewart6> And it's suitably cheesy
- # [14:06] <sstewart6> Question: when you render that page in the browser, are the children actually visible?
- # [14:07] <andreastt> Yes.
- # [14:07] <sstewart6> Then the test LGTM
- # [14:07] <sstewart6> :)
- # [14:07] <andreastt> The parent is not, for example if you give it a background colour.
- # [14:07] <sstewart6> Interesting
- # [14:07] <andreastt> But webdriver should think it's visible, right?
- # [14:07] <sstewart6> Ugh. Brain hurts
- # [14:07] <andreastt> (=
- # [14:08] <andreastt> I think this is the same issue we discussed with gdennis.
- # [14:08] <sstewart6> I think you may be right
- # [14:08] <andreastt> It also becomes a cascading problem I think.
- # [14:08] <sstewart6> s/may/are/
- # [14:08] <andreastt> Should the parent's parent be visible? If the parent is considered visible by webdriver, we then have to use a different metric for visibility for deciding the visibility of the parent's parent's visibility.
- # [14:08] <sstewart6> We check for visibility for two reasons: a) we want to interact with it, b) we want to tell whether it would be considered "displayed"
- # [14:09] <AutomatedTester> we had this discuss in the wrt interactable vs visible
- # [14:09] * sstewart6 hides under desk, rocking backwards and forwards
- # [14:09] <AutomatedTester> discussion*
- # [14:09] <sstewart6> Right
- # [14:11] <AutomatedTester> we do a number of traversals up the tree atm
- # [14:12] <AutomatedTester> and doing visible going all the way back up for different reasons makes things take longer and longer
- # [14:12] <AutomatedTester> I think if we make certain assumptions (yes they could be wrong) and stick to them we should cover a large percentage of use cases
- # [14:12] <sstewart6> We should probably cache results as we figure out visibility
- # [14:12] <AutomatedTester> sstewart6: yup
- # [14:13] <andreastt> Applying some memoization there makes sense, yes.
- # [14:13] <AutomatedTester> we should memoise the path
- # [14:13] * AutomatedTester has said we should do this for a very long time
- # [14:13] <sstewart6> It's pretty obvious
- # [14:13] <AutomatedTester> but I have failed at making a patch :/
- # [14:13] <sstewart6> We've just never done it
- # [14:13] <sstewart6> snap
- # [14:13] <andreastt> It's just a scary thing to code, I tink.
- # [14:13] <andreastt> think.
- # [14:13] <AutomatedTester> not really
- # [14:14] <andreastt> But with a sufficiently advanced test suite we should be covered.
- # [14:14] <andreastt> That's what we're writing now (-:
- # [14:14] <AutomatedTester> for memoisation if our current tests failed we have a problem
- # [14:14] <AutomatedTester> I dont think we need added tests just for that
- # [14:14] <AutomatedTester> sinces its a perf thing
- # [14:17] <sstewart6> We should make sure that our visibility calculations aren't cached too long
- # [14:18] <AutomatedTester> sstewart6: for the life of the method only
- # [14:18] <sstewart6> right
- # [14:20] * Joins: gitbot (~gitbot@public.cloak)
- # [14:20] -gitbot:#testing- [web-platform-tests] chrislo pushed 4 new commits to master: https://github.com/w3c/web-platform-tests/compare/0aad38267289...7f5c1d1d529d
- # [14:20] -gitbot:#testing- web-platform-tests/master d5b8864 Chris Lowis: remove interface test helper. We'll use WebIDL now.
- # [14:20] -gitbot:#testing- web-platform-tests/master 973f9ae Chris Lowis: refactor bufferLoader to use decodeAudioData
- # [14:20] -gitbot:#testing- web-platform-tests/master 4954657 Chris Lowis: remove unused helpers and add assert_array_approx_equals
- # [14:20] * Parts: gitbot (~gitbot@public.cloak) (gitbot)
- # [14:20] * Joins: gitbot (~gitbot@public.cloak)
- # [14:20] -gitbot:#testing- [web-platform-tests] andreastt opened pull request #247: Add visibility test for hidden parent with visible children (master...andreastt/overflow_parent_visible_with_absolute_children) https://github.com/w3c/web-platform-tests/pull/247
- # [14:20] * Parts: gitbot (~gitbot@public.cloak) (gitbot)
- # [14:21] <andreastt> Ms2ger: When does the critic bot pick up on it?
- # [14:21] <AutomatedTester> srsly... http://24.media.tumblr.com/tumblr_m7qmdyN8a01qf5ifco1_500.gif
- # [14:21] <sstewart6> andreastt: do we have that case covered in the spec?
- # [14:21] <andreastt> Unsure.
- # [14:22] <sstewart6> Cos if it's not in the spec, we may have to update that too :)
- # [14:22] <sstewart6> (We can handle it in a follow up diff if necessary)
- # [14:22] <sstewart6> (but we should make sure that it doesn't drop off the map)
- # [14:22] <AutomatedTester> raise a bug
- # [14:22] <AutomatedTester> if in doubt... raise a bug
- # [14:22] <andreastt> We discussed it at the meeting, but I don't know if either of you have added it yet.
- # [14:23] <sstewart6> I've not added it
- # [14:23] <andreastt> Where's the bug tracker BTW?
- # [14:23] <AutomatedTester> i linked it earlier
- # [14:23] <andreastt> Oh, sorry.
- # [14:23] * andreastt goes to read the backlog
- # [14:23] <AutomatedTester> https://www.w3.org/Bugs/Public/enter_bug.cgi?product=Browser%20Test%2FTools%20WG&content=webdriver
- # [14:23] <andreastt> What's the login there? The W3C one?
- # [14:24] <AutomatedTester> I think you may need to create one
- # [14:24] <AutomatedTester> since non-w3c wg members can create bugs
- # [14:25] <Ms2ger> Yeah, it's a separate one
- # [14:25] <andreastt> Thanks (-:
- # [14:25] * Joins: gitbot (~gitbot@public.cloak)
- # [14:25] -gitbot:#testing- [web-platform-tests] Ms2ger closed pull request #247: Add visibility test for hidden parent with visible children (master...andreastt/overflow_parent_visible_with_absolute_children) https://github.com/w3c/web-platform-tests/pull/247
- # [14:25] * Parts: gitbot (~gitbot@public.cloak) (gitbot)
- # [14:25] * Joins: gitbot (~gitbot@public.cloak)
- # [14:25] -gitbot:#testing- [web-platform-tests] Ms2ger reopened pull request #247: Add visibility test for hidden parent with visible children (master...andreastt/overflow_parent_visible_with_absolute_children) https://github.com/w3c/web-platform-tests/pull/247
- # [14:25] * Parts: gitbot (~gitbot@public.cloak) (gitbot)
- # [14:26] <AutomatedTester> Ms2ger: what are you doing!?!?
- # [14:26] <AutomatedTester> :P
- # [14:27] * Ms2ger pokes jgraham
- # [14:28] <AutomatedTester> did we break it?
- # [14:28] <Ms2ger> Seems like something broke it
- # [14:29] <AutomatedTester> wasnt me
- # [14:29] <AutomatedTester> I think
- # [14:29] <AutomatedTester> I blame the Opera people
- # [14:29] <andreastt> Isn't jgraham officially Mozilla people now? d-:
- # [14:30] <AutomatedTester> no
- # [14:30] <AutomatedTester> is he in no-man's land?
- # [14:30] * Joins: gitbot (~gitbot@public.cloak)
- # [14:30] -gitbot:#testing- [web-platform-tests] Ms2ger opened pull request #248: Add a test for empty names for Element.children. (master...HTMLCollection-empty-children) https://github.com/w3c/web-platform-tests/pull/248
- # [14:30] * Parts: gitbot (~gitbot@public.cloak) (gitbot)
- # [14:30] * Joins: Lachy (~Lachy@public.cloak)
- # [14:30] <AutomatedTester> he doesnt start for a few weeks
- # [14:45] <AutomatedTester> there we go
- # [14:45] <AutomatedTester> got a critic email
- # [14:46] <AutomatedTester> but not for the PR
- # [14:46] <AutomatedTester> ok...
- # [15:05] * Joins: gitbot (~gitbot@public.cloak)
- # [15:05] -gitbot:#testing- [web-platform-tests] chrislo pushed 1 new commit to master: https://github.com/w3c/web-platform-tests/commit/cabf1bdae1cd90a0ab126749011a24045cb8c2f8
- # [15:05] -gitbot:#testing- web-platform-tests/master cabf1bd Chris Lowis: add a longer timeout to the test
- # [15:05] * Parts: gitbot (~gitbot@public.cloak) (gitbot)
- # [15:16] * jgraham discovers he should read the backscroll
- # [15:20] * Joins: gitbot (~gitbot@public.cloak)
- # [15:20] -gitbot:#testing- [web-platform-tests] jgraham closed pull request #247: Add visibility test for hidden parent with visible children (master...andreastt/overflow_parent_visible_with_absolute_children) https://github.com/w3c/web-platform-tests/pull/247
- # [15:20] * Parts: gitbot (~gitbot@public.cloak) (gitbot)
- # [15:20] * Joins: gitbot (~gitbot@public.cloak)
- # [15:20] -gitbot:#testing- [web-platform-tests] jgraham reopened pull request #247: Add visibility test for hidden parent with visible children (master...andreastt/overflow_parent_visible_with_absolute_children) https://github.com/w3c/web-platform-tests/pull/247
- # [15:20] * Parts: gitbot (~gitbot@public.cloak) (gitbot)
- # [15:21] <jgraham> So the critic repo got a little behind somehow
- # [15:21] <jgraham> I need to fix it so that if that happens it tries to fetch rather than erroring
- # [15:22] <jgraham> But I kicked it now
- # [15:23] <andreastt> sstewart6: https://critic.hoppipolla.co.uk/r/217
- # [15:24] <sstewart6> Reviewed. Will critic push into the repo automagically?
- # [15:24] <AutomatedTester> no
- # [15:24] <sstewart6> Bah!
- # [15:24] <sstewart6> One moment, then, caller.
- # [15:24] <sstewart6> Did you file a bug to make sure we don't lose _why_ this test was written?
- # [15:25] <andreastt> AutomatedTester: We do that at Opera, but there seems to be very little security here.
- # [15:25] <andreastt> sstewart6: Working on it, but got derailed by some intern follow-up.
- # [15:25] <sstewart6> OK. NP
- # [15:26] * Joins: gitbot (~gitbot@public.cloak)
- # [15:26] -gitbot:#testing- [web-platform-tests] chrislo pushed 1 new commit to master: https://github.com/w3c/web-platform-tests/commit/d9277da9b5847724b45d1979f4dafa0c5c1dac0f
- # [15:26] -gitbot:#testing- web-platform-tests/master d9277da Chris Lowis: remove old assert helpers, they're in testharness.js now.
- # [15:26] * Parts: gitbot (~gitbot@public.cloak) (gitbot)
- # [15:27] <sstewart6> What's the workflow for merging?
- # [15:27] <sstewart6> Shall I just follow the seleium approach?
- # [15:27] <AutomatedTester> sstewart6: we dont require rebasing
- # [15:28] <AutomatedTester> merge commits are ok
- # [15:28] <sstewart6> Do I hit that nasty green button on github then?
- # [15:28] <tobie> sstewart6: yes please.
- # [15:28] <AutomatedTester> sstewart6: you can if you want
- # [15:28] * Joins: mdas (~mdas@public.cloak)
- # [15:28] * Joins: gitbot (~gitbot@public.cloak)
- # [15:28] -gitbot:#testing- [web-platform-tests] shs96c closed pull request #247: Add visibility test for hidden parent with visible children (master...andreastt/overflow_parent_visible_with_absolute_children) https://github.com/w3c/web-platform-tests/pull/247
- # [15:28] * Parts: gitbot (~gitbot@public.cloak) (gitbot)
- # [15:28] * Joins: gitbot (~gitbot@public.cloak)
- # [15:28] -gitbot:#testing- [web-platform-tests] shs96c pushed 2 new commits to master: https://github.com/w3c/web-platform-tests/compare/d9277da9b584...1c76adbcb5b7
- # [15:28] -gitbot:#testing- web-platform-tests/master 757bce5 Andreas Tolf Tolfsen: Add visibility test for hidden parent with visible children...
- # [15:28] -gitbot:#testing- web-platform-tests/master 1c76adb Simon Stewart: Merge pull request #247 from andreastt/andreastt/overflow_parent_visible_with_absolute_children...
- # [15:28] * Parts: gitbot (~gitbot@public.cloak) (gitbot)
- # [15:29] <sstewart6> andreastt: there you go
- # [15:29] <andreastt> Well. Something appear to have happened.
- # [15:29] * Joins: mdas_ (~mdas@public.cloak)
- # [15:29] * AutomatedTester runs off to twitter to say that sstewart6 used the github greeen button...
- # [15:29] <AutomatedTester> :P
- # [15:29] <sstewart6> Forget you, AutomatedTester. Forget you and the forgetting horse you forgetting rode in on
- # [15:29] <sstewart6> :P
- # [15:29] * AutomatedTester takes his pony away in a sulk
- # [15:30] <AutomatedTester> screw you guys, I'm going home
- # [15:30] <sstewart6> ha
- # [15:30] <sstewart6> You are home
- # [15:30] <AutomatedTester> fine... I'll stay
- # [15:30] <AutomatedTester> :D
- # [15:30] <sstewart6> :)
- # [15:30] <andreastt> That button is very green though. I think that's probably the nicest thing I can say about it.
- # [15:35] * Quits: mdas (~mdas@public.cloak) (Ping timeout: 180 seconds)
- # [15:36] * Quits: mdas_ (~mdas@public.cloak) (Ping timeout: 180 seconds)
- # [15:42] <jgraham> The button is very nice. It just does totally the wrong thing
- # [15:43] * Joins: mdas (~mdas@public.cloak)
- # [15:43] * Quits: Ms2ger (~Ms2ger@public.cloak) ("bbl")
- # [15:44] * Joins: mdas_ (~mdas@public.cloak)
- # [15:44] <andreastt> jgraham: Yep (-:
- # [15:45] <jgraham> sstewart6: You might want to add your email address in critic
- # [15:46] <sstewart6> ok
- # [15:46] <sstewart6> Where?
- # [15:46] <sstewart6> On it
- # [15:46] <jgraham> https://critic.hoppipolla.co.uk/home
- # [15:46] <sstewart6> Yeah :)
- # [15:46] <sstewart6> Just found it :)
- # [15:46] <sstewart6> Thank you
- # [15:47] <sstewart6> Done
- # [15:47] * jgraham wonders why tobie keeps making PRs and merging them himself
- # [15:47] <sstewart6> Because you're meant to make a PR and then do a pull?
- # [15:47] <sstewart6> s/pull/merge/
- # [15:47] <sstewart6> He's following process
- # [15:48] <jgraham> sstewart6: Well usually the process has review in the middle
- # [15:48] <jgraham> That clearly isn't happening here
- # [15:48] <sstewart6> We could just ask him?
- # [15:48] <jgraham> (this is testtwf-website or whatever that repo is called)
- # [15:48] <tobie> I'm slightly schizophrenic.
- # [15:49] * sstewart6 waits for the obvious follow up comment by tobie
- # [15:49] <jgraham> I don't object to you just merging stuff, but getting two emails per change gets old fast
- # [15:49] <tobie> I actually like the paper trail PR + merge gives.
- # [15:50] * jgraham considers "git log" to be equivalent
- # [15:50] * Quits: mdas (~mdas@public.cloak) (Ping timeout: 180 seconds)
- # [15:50] <tobie> That's arguable.
- # [15:52] <AutomatedTester> if we have merge commits then its not really arguable...
- # [15:53] <AutomatedTester> zomg I found tea on my desk that is still warm
- # [15:53] <AutomatedTester> totally forgot about that
- # [15:53] <AutomatedTester> its like finding £5 in a pocket of jeans
- # [15:54] <gsnedders> Thermal jugs are the best for that.
- # [15:54] <AutomatedTester> gsnedders: that requires forethought
- # [15:54] <AutomatedTester> and I lack that mostly
- # [15:55] <gsnedders> (Opera's xmas present last year was… a thermal jug. It's wonderful. I'd never have bought one myself, as why would I when I have a thermos?)
- # [15:55] <AutomatedTester> see example of me making tea and forgetting to drink it
- # [15:55] <gsnedders> Yeah, I do that too. :)
- # [15:55] * jgraham just had to desugar his kitchen wall
- # [15:56] <AutomatedTester> desugar?
- # [15:56] <gsnedders> jgraham: Have you actually moved yet, BTW?
- # [15:56] <jgraham> I felt like slightlyoff
- # [15:56] <jgraham> gsnedders: No
- # [15:56] * andreastt just discovered that was a word in the English language
- # [15:56] <AutomatedTester> why was there sugar on the wall?
- # [15:56] <jgraham> AutomatedTester: Over entusiastic whisking
- # [15:57] <AutomatedTester> ok...
- # [15:57] <AutomatedTester> interesting
- # [15:57] <AutomatedTester> :)
- # [15:57] <jgraham> I wouldn't have mentioned it
- # [15:57] <jgraham> Except for the pun
- # [15:57] <gsnedders> AutomatedTester: Damn it now I want tea.
- # [15:58] <AutomatedTester> gsnedders: you're welcome
- # [15:58] <gsnedders> (Never mind the fact that Glasgow is "warm" and "sunny", words I struggle to use living here. I forget their meaning.)
- # [15:59] <AutomatedTester> gsnedders: you live in Glasgow?
- # [16:05] <sstewart6> oh. tea
- # [16:10] * AutomatedTester twiddles his thumbs waiting for the repo to clone
- # [16:15] <gsnedders> AutomatedTester: Yes.
- # [16:24] * AutomatedTester pokes andreastt for that bug before MFBT
- # [17:01] <stearns> AutomatedTester: will you be physically present at the TPAC TestTWF?
- # [17:03] <AutomatedTester> stearns: I was hoping to
- # [17:03] <AutomatedTester> stearns: I need to be at TPAC for the Monday for WG
- # [17:04] <stearns> AutomatedTester: doing WebDriver tests would be good - it's best when one or more topic experts are there to guide folks
- # [17:04] <AutomatedTester> stearns: I will be there and I am sure we can coax sstewart6 to be there too
- # [17:04] <stearns> AutomatedTester: also, it works better when there's a test plan (here are the tests we need, here's how to write them…)
- # [17:05] <stearns> AutomatedTester: the more that gets planned in advance, the better the results
- # [17:05] <sstewart6> It'd be good to be there
- # [17:05] <AutomatedTester> stearns: we have started working on that already
- # [17:05] <tobie> sstewart6: yeah.
- # [17:05] <stearns> AutomatedTester: great!
- # [17:06] <AutomatedTester> stearns: I want to make sure that I have everything in place to hit CR the week after TPAC
- # [17:06] <AutomatedTester> stearns: so will make sure all the minutae are in order
- # [17:06] <tobie> I'm looking forward to the discussion around settling for which language to use to write WebDriver tests.
- # [17:06] <AutomatedTester> tobie: python
- # [17:06] <sstewart6> python 2.7
- # [17:07] <tobie> AutomatedTester: there are a number of folks disagreeing with this.
- # [17:07] <tobie> :)
- # [17:07] <AutomatedTester> tobie: well they havent come to us
- # [17:07] <sstewart6> What do they suggest as an alternative?
- # [17:07] <tobie> It'll be fun.
- # [17:07] <AutomatedTester> tobie: all the companies working on webdriver have agreed on python
- # [17:07] <andreastt> The alternatives all have worse issues than Python IMO.
- # [17:07] <stearns> I expect that will limit the pool of available TestTWF peons (we've never asked for python experience before) :)
- # [17:07] <sstewart6> tobie: even MS
- # [17:08] <sstewart6> stearns: we're out of process, so js isn't a great fit
- # [17:08] <AutomatedTester> tobie: so they can disagree amongst themselves but until they come to us I will treat it as nothing more than rumour
- # [17:08] <stearns> I don't have a preference - just setting TestTWF expectations
- # [17:08] <andreastt> OOP JS essentially means using V8, which understandably doesn't go down well with all vendors.
- # [17:09] <sstewart6> right, and that's why we discounted it
- # [17:09] <sstewart6> :)
- # [17:09] <AutomatedTester> andreastt: its not V8, its node which isnt as up to date as V8 all the time
- # [17:09] <andreastt> True.
- # [17:09] <sstewart6> For reference, the options we seriously considered were: JS, Java, C#, Python and Ruby
- # [17:09] <sstewart6> JS is out for reasons already discussed here
- # [17:09] <sstewart6> Java won't fly with some companies
- # [17:09] <AutomatedTester> c# isnt brilliant cross platform
- # [17:10] <sstewart6> Right
- # [17:10] <tobie> AutomatedTester: this isn't about what you guys are going to use to run WD tests as much as it is what the rest of the specs requiring server-side code will use a s a programming lang
- # [17:10] <sstewart6> Which leaves ruby and python
- # [17:10] <sstewart6> The same logic should apply, though
- # [17:10] <AutomatedTester> tobie: can you tell me who is complaining so I can speak to them
- # [17:10] <tobie> Well, so far, everyone had agreed on using WPHP
- # [17:10] <sstewart6> Frankly, the rails fanboys did a lot of damage to the credibility of ruby in some eyes
- # [17:10] <sstewart6> WPHP?
- # [17:10] <tobie> s/WPHP/PHP/
- # [17:11] <sstewart6> jesus
- # [17:11] <sstewart6> sorry
- # [17:11] <AutomatedTester> I highly doubt Mozilla agreed on PHP
- # [17:11] <tobie> (woops)PHP
- # [17:11] <AutomatedTester> and seriously... they want to use PHP
- # [17:11] <sstewart6> "everyone"?
- # [17:11] <AutomatedTester> and I highly doubt Google agreed on PHP
- # [17:12] <sstewart6> biab
- # [17:12] <tobie> I'll ping the mailing list with the current situation shortly.
- # [17:12] <AutomatedTester> tobie: Opera, Mozilla and Google use python in their builds/tests already
- # [17:12] <wilhelm> There are two distinct discussions here. They shold be kept separate.
- # [17:12] <tobie> No point in spoiling the fun on irc.
- # [17:13] <tobie> wilhelm: agree to a certain extent.
- # [17:13] <AutomatedTester> ok... then I will use this opportunity to drive home
- # [17:13] <wilhelm> One question is how to do <?php echo "this"; ?> with the least friction possible.
- # [17:13] <wilhelm> Another is how to automate a browser with the least friction possible.
- # [17:14] <AutomatedTester> bbl
- # [17:14] <tobie> wilhelm: and another how much legacy content we have relying on a given lang
- # [17:14] <tobie> wilhelm: and yet another: how much contributions will we get if we choose a given lang
- # [17:15] <tobie> tobie: what are the cost and benefits of using multiple languages, etc.
- # [17:17] * Quits: Lachy (~Lachy@public.cloak) ("Computer has gone to sleep.")
- # [17:17] <wilhelm> Yup. For the record, I think PHP is perfectly fine for the simple "return my passed POST variables" case. I encouraged that exact approach when at Opera, since mod_php is _everywhere_.
- # [17:18] <wilhelm> For the WebDriver case, I agree with the gentlemen above. (c:
- # [17:18] <tobie> The problem arises when you start having both in the same repo.
- # [17:19] <tobie> Which is exactly what where we're at right now.
- # [17:19] <tobie> s/what//
- # [17:19] <wilhelm> Are we talking only about web servers now, or in general?
- # [17:20] <andreastt> (For the sake of the argument, all of Opera's new remote server tests are written using local Python servlets.)
- # [17:20] <tobie> Soonish, folks will start authoring tests which were previously manual tests using WebDriver
- # [17:20] <wilhelm> Yes.
- # [17:20] <wilhelm> That's the goal. (c:
- # [17:22] * Quits: AutomatedTester (~AutomatedTester@public.cloak) (Ping timeout: 180 seconds)
- # [17:22] <wilhelm> But those tests have three components: the HTML+CSS+JS run in the browser, the server or file system serving those files (with or without server-side magic) and the human or robot pushing buttons.
- # [17:22] * gsnedders votes we use wilhelm as "the human"
- # [17:23] <tobie> Absolutely.
- # [17:23] * wilhelm votes against.
- # [17:26] <tobie> We could very well agree on using JS/PHP/Python (in that order) or JS all the way, or lots of other combinations.
- # [17:26] <wilhelm> Yes, in theory.
- # [17:27] <tobie> Not sure how "in practice" is going to change my very open ended comment. :)
- # [17:27] <wilhelm> It's tricky. (c:
- # [17:28] <gsnedders> BLOODY PEOPLE AND THEIR OPINIONS.
- # [17:28] <wilhelm> As of 2011, my position was JS-all-the-way-please. But it turned out to be more complicated than I had hoped.
- # [17:29] <tobie> For technical or political reasons?
- # [17:29] <sstewart6> It's not tricky in the slightest
- # [17:29] <sstewart6> browser = js, client side = python, hosted in a server = php (apparently)
- # [17:29] <sstewart6> *gag*
- # [17:29] <andreastt> There are many more factors to consider. The politics of the technology, its platform availability.
- # [17:30] <sstewart6> joy
- # [17:30] <tobie> andreastt: agreed.
- # [17:30] <andreastt> I think out of process JS is difficult for both reasons.
- # [17:30] <andreastt> For the context of controlling a browser remotely.
- # [17:31] <andreastt> I'm not saying it's technically not possible.
- # [17:31] <sstewart6> It's kind of funky.
- # [17:32] <tobie> andreastt: I'm not sure I understand the issue here. But I think we don't have the same use case in mind.
- # [17:32] <andreastt> Which side of it can I explain? I guess you're refering to JS?
- # [17:32] <tobie> Also, everytime I feel like I've finally grasped WebDriver, someone (usually sstewart6) proves me wrong.
- # [17:32] <sstewart6> ?
- # [17:32] <sstewart6> What've we done this time?
- # [17:33] <andreastt> (He usually has that effect.)
- # [17:33] <tobie> sstewart6: see ^
- # [17:33] * Joins: mdas (~mdas@public.cloak)
- # [17:33] <wilhelm> sstewart6: Solved a difficult problem. (c;
- # [17:33] <sstewart6> It's a well known phenomenon
- # [17:33] <sstewart6> :)
- # [17:34] <tobie> andreastt: yeah. I'm unsure why the out of process JS is an issue given the JSON protocol in the middle.
- # [17:35] <tobie> But that's probably because I'm missing something.
- # [17:35] <sstewart6> Because once you start writing JS people start trying to run it in browsers
- # [17:35] <sstewart6> And then you head down a path of fugly nesting
- # [17:35] <tobie> Oh.
- # [17:35] <sstewart6> or impossible to reason about futures
- # [17:35] <tobie> That.
- # [17:35] <sstewart6> Both of which are suboptimally entertaining
- # [17:35] <andreastt> On the technical side we'd have to bring in some sort of promise library to make the blocking nature of the WebDriver API nice to deal with, which means introducing a third party dependency.
- # [17:36] <sstewart6> webdriverjs would actually be okay for that
- # [17:36] <sstewart6> It has its own promises library baked in
- # [17:36] <tobie> andreastt: that's arguably not an issue.
- # [17:36] * Joins: mdas__ (~mdas@public.cloak)
- # [17:36] <andreastt> As for politics, it's difficult for browser vendors (Moz in particular) to use another company's JS engine.
- # [17:36] <sstewart6> (cos we started using promises before they were trendy)
- # [17:37] <sstewart6> got to go
- # [17:37] <sstewart6> biab
- # [17:37] <andreastt> tobie: Third party dependencies isn't an issue?
- # [17:37] <andreastt> I actually also have to go fairly soon.
- # [17:37] <andreastt> The interns here are throwing a party.
- # [17:37] <wilhelm> Aw, my interns are just programming.
- # [17:37] <tobie> In the case of node.js based-program where there's a solid packaging system
- # [17:38] <tobie> I haven't seen deps be a serious issue.
- # [17:39] * Quits: mdas_ (~mdas@public.cloak) (Ping timeout: 180 seconds)
- # [17:39] <wilhelm> It has been in the past. Fewer constraints would be good news here.
- # [17:40] * Quits: mdas (~mdas@public.cloak) (Ping timeout: 180 seconds)
- # [17:57] * Joins: mdas (~mdas@public.cloak)
- # [17:57] * Joins: kkania (~kkania@public.cloak)
- # [17:58] <tobie> jgraham: care to press the green button, or should I? https://github.com/w3c/testtwf-website/pull/56 :P
- # [18:03] * Quits: mdas__ (~mdas@public.cloak) (Ping timeout: 180 seconds)
- # [18:16] * Joins: jimevans (~jimevans@public.cloak)
- # [18:16] <jimevans> g'day ladies and gentlemen
- # [18:16] <jimevans> my apologies for being late.
- # [18:16] <jimevans> i'm fighting an illness.
- # [18:19] <sstewart6> g'day jimevans
- # [18:19] <sstewart6> Hope you're feeling better soon
- # [18:19] <jimevans> me too. :)
- # [18:44] * Quits: darobin (rberjon@public.cloak) (Client closed connection)
- # [19:03] * Joins: chrisgao (~chrisgao@public.cloak)
- # [20:08] * Parts: sstewart6 (~simons@public.cloak) (sstewart6)
- # [20:22] * Quits: dom (dom@public.cloak) ("")
- # [20:23] * Quits: mdas (~mdas@public.cloak) (Client closed connection)
- # [20:23] * Joins: mdas (~mdas@public.cloak)
- # [20:30] * Quits: mdas (~mdas@public.cloak) (Ping timeout: 180 seconds)
- # [20:59] * Joins: AutomatedTester (~AutomatedTester@public.cloak)
- # [21:18] * Quits: ArtB (~abarsto@public.cloak) (Ping timeout: 180 seconds)
- # [21:19] * Joins: abarsto (~abarsto@public.cloak)
- # [21:19] * abarsto is now known as ArtB
- # [22:00] * Quits: ArtB (~abarsto@public.cloak) ("Leaving.")
- # [22:15] <AutomatedTester> stearns: if I ever want to organise something for TTWF, are you the best person to email
- # [22:16] <AutomatedTester> stearns: since I originally emailed rhauck
- # [22:22] * Quits: chrisgao (~chrisgao@public.cloak) ("Page closed")
- # [22:40] * Quits: AutomatedTester (~AutomatedTester@public.cloak) (Client closed connection)
- # [22:40] <kkania> are we supposed to fail a test if the impl doesn't follow a SHOULD from the spec?
- # [23:06] * Joins: jhammel (~jhammel@public.cloak)
- # [23:19] * Quits: glenn (~gadams@public.cloak) (Client closed connection)
- # Session Close: Sat Jul 13 00:00:00 2013
The end :)