/irc-logs / w3c / #testing / 2015-10-29 / end

Options:

# Session Start: Thu Oct 29 00:00:00 2015
# Session Ident: #testing
# [00:00] * Joins: zcorpan (~zcorpan@public.cloak)
# [00:15] * Joins: jungkees (~uid24208@public.cloak)
# [00:27] * Joins: anssik (~uid10742@public.cloak)
# [00:46] * Joins: bhill2 (~bhill2@public.cloak)
# [00:51] * Joins: dom (dom@public.cloak)
# [00:53] * Quits: bhill2 (~bhill2@public.cloak) (Client closed connection)
# [00:53] * Joins: bhill2 (~bhill2@public.cloak)
# [00:54] * Parts: bhill2 (~bhill2@public.cloak) (Leaving...)
# [00:54] * Joins: plh (plehegar@public.cloak)
# [01:08] * Joins: jyasskin (~textual@public.cloak)
# [01:09] * terri is now known as terri_offline
# [02:03] * Quits: ArtB (~ArtB@public.cloak) ("Leaving.")
# [02:04] * Quits: zcorpan (~zcorpan@public.cloak) (Client closed connection)
# [02:06] * Joins: zcorpan (~zcorpan@public.cloak)
# [02:06] * Quits: zcorpan (~zcorpan@public.cloak) (Client closed connection)
# [02:09] * Joins: zcorpan (~zcorpan@public.cloak)
# [03:57] * Quits: jyasskin (~textual@public.cloak) ("My computer has gone to sleep. ZZZzzz…")
# [04:06] * Quits: dom (dom@public.cloak) (Ping timeout: 180 seconds)
# [04:12] * Quits: plh (plehegar@public.cloak) ("Leaving")
# [04:19] * Quits: zcorpan (~zcorpan@public.cloak) (Client closed connection)
# [04:20] <gsnedders> jgraham: where be?
# [04:20] <gsnedders> Possibly at lunch, I guess :)
# [04:54] <jgraham> Indeed
# [04:56] <gsnedders> I can't remember what I was going to say now :)
# [04:56] <gsnedders> oh, does wptrunner do anything with the metadata in the CSS tests?
# [04:56] <gsnedders> or just run off the same metadata as in wpt?
# [04:57] <gsnedders> I'm not entirely sure why I'm talking to you on IRC instead of using my vocal calls.
# [04:57] <gsnedders> *chords
# [05:01] <jgraham> wptrunner just uses the [mis]match relationships
# [05:01] <gsnedders> so it runs tests with things like "paged" flags?
# [05:02] <gsnedders> while running on browserswith scrolled media
# [05:04] <jgraham> Yup, but vendors can disable those tests if they want using the metadata files
# [05:05] <jgraham> Or use the mtadata files to set a browser-specific pref that enables that feature
# [05:06] * Joins: jyasskin (~textual@public.cloak)
# [05:11] <gsnedders> "When a test involving a CSS module has been reviewed by one of its CSS spec editors, then such test becomes more trustworthy. Reviewing a test by a competent and independent party is important."
# [05:11] <gsnedders> "There is such a thing as Quality control, Quality assurance, de-subjectivizing the evaluation of your own work. It's even a fundamental principle in justice. I would need more time to explain all this.
# [05:11] <gsnedders> I don't see why.
# [05:11] <gsnedders> It seems hypothetical.
# [05:12] <jgraham> I don't think anyone is saying tests don't need review
# [05:12] <gsnedders> But CSS spec editors don't review them.
# [05:13] <gsnedders> The implication seems to be that only an editor reviewing it makes it trusthworthy.
# [05:13] <jgraham> They are saying that the cost of making the review a very limited group of people is much much higher than the value it offers
# [05:15] <gsnedders> No, they aren't, as far as I can tell?
# [05:15] <gsnedders> They're saying it's more worthwhile when CSS spec editors do it.
# [05:18] * Joins: zcorpan (~zcorpan@public.cloak)
# [05:22] * Joins: plh (plehegar@public.cloak)
# [05:31] * Joins: gitbot (~gitbot@public.cloak)
# [05:31] -gitbot:#testing- [web-platform-tests] hayatoito closed pull request #2284: Updates frameElement.sub.html: frameElement should return null if cross origin. (master...frameElement-test) https://github.com/w3c/web-platform-tests/pull/2284
# [05:31] * Parts: gitbot (~gitbot@public.cloak)
# [05:40] <jgraham> gsnedders: i think you misunderstood the antecendent of "they" in that sentence. I meant "people who don't think the group of reviewers should be strictly controlled"
# [05:41] * Joins: dom (dom@public.cloak)
# [05:53] * Joins: fwtnb__ (~uid123457@public.cloak)
# [06:07] <gsnedders> oh right
# [07:32] * Disconnected
# [07:34] * Attempting to rejoin channel #testing
# [07:34] * Rejoined channel #testing
# [07:35] * Quits: hober (~ted@public.cloak) (Client closed connection)
# [07:38] * Joins: fantasai (~fantasai@public.cloak)
# [07:43] <MikeSmith> regarding https://lists.w3.org/Archives/Public/public-webapps/2015OctDec/0116.html
# [07:43] <MikeSmith> > Testharness tests are not currently able to test the core Pointer Lock
# [07:43] <MikeSmith> features, which would require user gesures (mouse clicks) and synthesizing
# [07:43] <MikeSmith> mouse movement. No progress is seen here for this specification, but the
# [07:43] <MikeSmith> challenge is present for several. E.g. Web Bluetooth brought up this
# [07:43] <MikeSmith> concern recently [Platform Testing] on public-test-infra, but with no
# [07:43] <MikeSmith> response. I have requested Jeffrey Yasskin and Sam Uong from Chrome to
# [07:44] <MikeSmith> discuss this topic at TPAC.
# [07:44] <MikeSmith> ...
# [07:45] <jyasskin> MikeSmith: We brought this up in the Web Driver session on Tuesday, and in the "getting people to write/run tests" session Wednesday. For user gestures, it seems like there's consensus that Web Driver is a decent way to accomplish it. Sam's interested in getting that working in ChromeDriver, but I don't know the state of running tests using ChromeDriver in our CI system.
# [07:45] <MikeSmith> Maybe Sam mentioned this in the WebDriver f2f meeting?
# [07:45] <MikeSmith> jyasskin: OK thanks
# [07:47] <jyasskin> There's another half for Bluetooth and Geolocation, in which we need to tell the UA to return certain results. Web Driver might also be a good place for this part, but it's farther afield from what Web Driver's currently doing. We didn't talk about this half enough to find consensus.
# [07:47] <MikeSmith> jyasskin: btw last time I was in a conversatio with Sam about the state of running tests using ChromeDriver in y'alls CI system, he was mainly concerned of about cost to everybody in additional build time
# [07:47] <MikeSmith> ok
# [07:47] <jyasskin> MikeSmith: Yep, that's the sort of problem I'm worried about.
# [07:48] <MikeSmith> so that's seems measurable and an evaluation could be made on measuring it
# [07:50] <MikeSmith> I know Joshua Bell is a strong advocate for trying to get people on the blink/chrome team to be writing more testharness.js-based tests, and upstreaming them. I think he might be willing to help try to to measure the costs of the CI integration
# [07:52] <MikeSmith> and of course Mozilla has been running all of the web-platform-tests suite in their CI for some time now and could give guidance based on what they've learned over time from that experience so far
# [07:52] <MikeSmith> though of course all the browser build systems are very different
# [07:53] <MikeSmith> jyasskin: anyway, as far as the WebDriver integration with testharness.js I imagine you heard this at the WebDriver f2f, but the main thing that's been blocking that for a year or more now is just the lack of a person committed to do the work, with the time to do it
# [07:54] <MikeSmith> if we could get someone to work on it, I think we've estimated they could get it done within 3 months
# [07:54] <MikeSmith> part of that needs to be looking at all the requirements
# [07:55] <MikeSmith> but the basic requirement we're starting from is just to be able to automate as many of our existing manual tests as we can
# [07:55] <jyasskin> MikeSmith: Right. I think Sam might be interested in doing this. I'm trying to avoid it, especially if it's going to take 3 months away from implementing Web Bluetooth, but we may have to take it on if nobody's done it by the time we want to go to CR.
# [07:55] <MikeSmith> OK
# [07:56] <wilhelm> Yes, we have a proof of concept. But no engineer to do the actual work. If Google or someone else can free up resources, that would be excellent.
# [07:56] <jyasskin> I'll also make sure Josh and Ojan are aware of the need for a person.
# [07:56] <MikeSmith> well I have a open mandate to help make it happen, though I also can't personally set aside the weeks needed to do the real work
# [07:58] <MikeSmith> at least intially I could help with things like going through the existing manual tests to identify what WebDriver features we'd need to hook into to automate them
# [07:58] <MikeSmith> but I'm reluctant to spend much time on that until we have somebody committed to spending time on the implementation
# [07:59] <MikeSmith> anyway, maybe we are finally reaching critical mass on enough people needing this to happen sooner rather than later
# [08:11] * Quits: fwtnb__ (~uid123457@public.cloak) ("Updating details, brb")
# [08:11] * Joins: fwtnb__ (~uid123457@public.cloak)
# [08:18] * Quits: zcorpan (~zcorpan@public.cloak) (Client closed connection)
# [08:22] <gsnedders> I know jgraham wanted to avoid separate files and people having to know Python, but I wonder if we should just accept nobody is going to work on the metacircular stuff.
# [08:22] * Joins: zcorpan (~zcorpan@public.cloak)
# [08:23] <Mek> not knowing anything about webdriver/chromedriver, I have at least seen code in chromedriver somewhere that seems to do mock geolocations, but I have no idea how that would be used in actual tests
# [08:25] <ato> gsnedders: I didn’t read the full backlog here, but Google has expressed interest in working on it, and Mozilla is doing a “quarter of contribution” PR thing to attract interns to work on it.
# [08:27] * hober2 is now known as hober
# [08:29] <ato> Mek: Yes, there are experimental implementations of it.
# [08:29] <ato> Mek: Geolocation is quite easy to do, and will probably be one of the first extension we try to standardise for WebDriver I think.
# [08:29] <gsnedders> Also as jyasskin pointed out yesterday it's likely authors will want to mock geolocations and similar for testing their apps
# [08:29] <gsnedders> so it's not purely browser testing APIs
# [08:30] <gsnedders> OTOH, there are some of the existing testing APIs which go into implementation detail and that we shouln't expose in any general way
# [08:30] <gsnedders> probably provide some extension mechanism so they can be pathed over WEbDriver
# [08:31] <Mek> ato: cool, where would I go to read more about webdriver/follow progress on standardizing things like geolocation?
# [08:32] <wilhelm> Mek: WebDriver has an extension interface that should be the basis for new endpoints like that: http://www.w3.org/TR/webdriver/#protocol-extensions
# [08:32] <ato> gsnedders: Yes, we need to talk to some spec people on how it would be best to spec.
# [08:32] <gsnedders> I know I've said I'll help before, but I have too much to do. :)
# [08:32] <wilhelm> Those endpoints should ideally be defined by the WG working on the feature, but with help from the WD subject-matter experts.
# [08:32] <ato> Yep, like wilhelm points out that explains how they are done in practice.
# [08:33] <ato> gsnedders: I think we can do the work but we could do with advice from more experienced spec people.
# [08:33] <ato> Yes, what wilhelm says is what we discussed as a possible approach.
# [08:33] <ato> Like, putting the actual prose in GEOLOCATION or whatever.
# [08:34] <wilhelm> Mek: We briefly discussed a general method for accepting permissions (like for geolocation, camera, whatever) in a way a little similar to how we handle prompts: http://www.w3.org/TR/webdriver/#user-prompts
# [08:34] <wilhelm> There seem to be patterns emerging on how those are presented to users.
# [08:34] <gsnedders> ato: perhaps. there's a fair bit of process nonsense to creating a revision of a published W3C spec
# [08:34] <ato> Yes, that’s my worry.
# [08:35] <wilhelm> It may be fine to add the extensions to our spec, but we do not have the expertise or bandwidth to do the actual work on each particular feature.
# [08:35] <Mek> for new features I agree that it makes sense to spec how it can be tested in the same spec that specs the feature
# [08:35] <jyasskin> wilhelm: camera is a little risky: we're talking about moving that to a chooser model, where you'd select the camera rather than just saying "yes".
# [08:35] <gsnedders> I think what you may well have to do is add the extensions in your own spec if there's no active work for the other spec.
# [08:36] <ato> wilhelm: I think the feelings of the WG was that the user prompt handler approach didn’t make sense because permissions dialogues typically don’t block the JS runtime; but yes, there are converging behaviours for permissions for different features.
# [08:36] <wilhelm> jyasskin: Yes, but that fits into the same theme as your choose-which-BT-device need, or the choose-which-second-screen need.
# [08:36] <jyasskin> We can write small parallel specs like GEOLOCATION-TESTING that define the necessary extensions without revving GEOLOCATION. I'm starting that sort of thing in https://webbluetoothcg.github.io/web-bluetooth/tests/, and will update it to WebDriver once we know how to add that support.
# [08:36] <jyasskin> Yep
# [08:37] <wilhelm> This is yet uncharted territory, and we'll need your help to find the patterns. :)
# [08:37] <ato> jyasskin: What we discussed was a generic approach where you could query what permissions were currently being requested, which would return an (arbitrarily typed?) enumeration of options, and a primitive to select one or dismiss.
# [08:37] <ato> jyasskin: Plus an option to reset to default.
# [08:37] <wilhelm> ato: Yes. Related to, but not equal to.
# [08:37] <Mek> wilhelm: well, I'd love to help from the geolocation side
# [08:37] <ato> jyasskin: The GEOLOCATION-TESTING idea is interesting!
# [08:38] <wilhelm> Mek: That would be excellent! Could you perhaps read through how we do the other endpoints, and see if you can napkin-sketch what you'd need for geolocation?
# [08:39] <wilhelm> Where to put the mocking is one of the challenges, I guess.
# [08:39] <Mek> wilhelm: yeah, I'll try to read up on all this testing/webdriver stuff
# [08:39] <ato> I think the current Chrome implementation is very simplistic. It allows you to spoof the current position and reset it to default.
# [08:39] <wilhelm> Mek: Wonderful! Just let us know if you need input, or reviews.
# [08:40] <Mek> the current chrome implementation for geolocation mocking I think also has issues (and might not work for the geofencing API I'm working on)
# [08:40] <gsnedders> what jyasskin is true, but you can, within the W3C, get into debates as to whether that's within the remit of your charter.
# [08:40] <wilhelm> gsnedders: Worst case, then we'll just stuff things in our level 2 spec. We just can't do all the work. ;)
# [08:42] <jyasskin> Meh :) The important thing is to write the spec and get it implemented. The WICG could be a fine place. It can land in any WG when it's done.
# [08:43] <gsnedders> BTW, do any of you guys have any idea what you're doing for dinner?
# [08:43] <gsnedders> am debating going back to hotel for a bit
# [08:43] <ato> jyasskin: Good idea.
# [08:44] <ato> gsnedders: I’ll be visiting one of the seven bars my hotel has. (I’m only staying five days )-: )
# [08:45] <ato> jyasskin: Even unofficial drafts might help here.
# [08:45] <jyasskin> +1
# [08:45] <ato> What is important from WebDriver’s side, is getting the base specification to a REC level.
# [08:45] <ato> (For various political and legal reasons.)
# [08:45] <ato> What we pile on top of that will be easy.
# [08:45] <gsnedders> I expect much of these things might have to be done in the Web Platform WG, but I also expect one just defining a testing API wouldn't be that hard
# [08:46] <jyasskin> (However, every WG's charter should include "and testing APIs for all our things.")
# [08:47] <gsnedders> (Which won't happen for forever because rechartering is too much work.)
# [08:47] <jyasskin> forever == 2 years, right?
# [08:47] <wilhelm> gsnedders: No plans.
# [08:47] <gsnedders> (Realistically things will get rechartered several times before rechartering.)
# [08:49] <ato> I should point out that the vendors represented in the WebDriver WG (who are doing the actual work of implementing these extensions) are a friendly bunch. I don’t think it will matter much where the specification prose lives to them.
# [08:49] <wilhelm> Indeed.
# [08:50] <wilhelm> There's the just-for-lawyers-spec, and then the pile of real stuff we need.
# [08:51] <ato> wilhelm: ++
# [08:51] <gsnedders> ato: yeah, this is AB level politics
# [08:54] <gsnedders> wilhelm: I was thinking I haven't had any Hokkaido ramenm which perhaps I should
# [08:56] <gsnedders> but maybe that's best left for Saturday lunchtime.
# [08:57] <wilhelm> gsnedders: After the third izakaya couldn't fit a football team-sized group yesterday, we went for that. It's good.
# [09:01] <gsnedders> Anyhow, -> hotel
# [09:07] * Quits: rego (~smuxi@public.cloak) (Client closed connection)
# [09:09] * Quits: zcorpan (~zcorpan@public.cloak) (Client closed connection)
# [09:12] * Quits: jyasskin (~textual@public.cloak) ("My computer has gone to sleep. ZZZzzz…")
# [09:14] * Joins: rego (~smuxi@public.cloak)
# [09:14] * Quits: rego (~smuxi@public.cloak) (Client closed connection)
# [09:23] * Joins: rego (~smuxi@public.cloak)
# [09:29] * Quits: plh (plehegar@public.cloak) ("Leaving")
# [09:34] * Joins: Ms2ger (~Ms2ger@public.cloak)
# [09:41] <sangwhan> jgraham: managed to dig out the iframe ticket/PR?
# [09:46] <gsnedders> wilhelm: let me know if you obtain any plan
# [09:48] <wilhelm> gsnedders: Obtaining a plan is a good idea.
# [09:48] <wilhelm> Anyone up for food soon-ish?
# [09:50] * gsnedders suggests somewhere townish, but that's rather implied by food
# [09:51] * Ms2ger plans his kitchen table
# [09:51] * Joins: jjj (~jjj@public.cloak)
# [09:52] * Quits: jjj (~jjj@public.cloak) ("Page closed")
# [09:56] * gsnedders has heard rumours of what Ms2ger looks like
# [09:56] <gsnedders> it's terrifying
# [09:56] <gsnedders> it's almost like Ms2ger might actually exist
# [09:56] <Ms2ger> Hah
# [09:56] <Ms2ger> So, you coming to Orlando?
# [09:59] <gsnedders> Nah.
# [10:05] * Quits: dom (dom@public.cloak) ("")
# [10:06] <AutomatedTester> wilhelm: the webdriver folk are going to Sapporo beer garden (We havent been yet)
# [10:06] <AutomatedTester> gsnedders: ^
# [10:06] <AutomatedTester> we are leaving the Century Royal Hotel at ~6:30
# [10:10] <wilhelm> AutomatedTester: Ah, fun. I'm still in the AC meeting, may ping you later. :)
# [10:10] * fwtnb__ is now known as fwtnb
# [10:10] <gsnedders> Hmm. Maybe I should.
# [10:11] <gsnedders> AutomatedTester: how are you getting tere? taxi?
# [10:11] <AutomatedTester> gsnedders: not too sure, not sure if Taxi or subway is the best way (or walking)
# [10:11] <AutomatedTester> I have no idea where it is
# [10:12] <gsnedders> not that close to any subway, ~40 minutes walk
# [10:13] <AutomatedTester> gsnedders: we are meeting up in the lobby at 6:30, when I know more I can send you a message via twitter
# [10:13] <gsnedders> k.
# [10:13] <AutomatedTester> or you could meet us there at 7
# [10:13] <gsnedders> I may well subway up and walk down to it
# [10:14] <AutomatedTester> I am guessing we wont be too far away by that time
# [10:16] <gsnedders> you'll be there if you taxi
# [10:16] <gsnedders> even if it takes forever for you to leave :)
# [10:17] <AutomatedTester> :)
# [10:17] <AutomatedTester> jgraham: you're welcome to join to (if you aren't asleep already)
# [10:19] <AutomatedTester> too*
# [10:47] <gsnedders> I so can't be bothered drinking beer.
# [10:53] * Joins: ArtB (~ArtB@public.cloak)
# [10:59] * Joins: jyasskin (~textual@public.cloak)
# [11:06] * Joins: bhill2 (~bhill2@public.cloak)
# [11:19] <ato> gsnedders: Especially not Japanese beer.
# [11:20] <ato> Ms2ger: Will we have the pleasure of seeing you in Orlando?
# [11:20] <gsnedders> ato: you not out with them? :)
# [11:21] <ato> gsnedders: I’m in Tokyo.
# [11:21] * Quits: jyasskin (~textual@public.cloak) (Ping timeout: 180 seconds)
# [11:21] <Ms2ger> 8ball says "Most Likely"
# [11:22] <ato> https://irccloud.mozilla.com/file/4ItSDRiP/IMG_20151029_191547.jpg
# [11:22] <ato> Two girls dancing and singing false on stage outside my hotel.
# [11:23] <ato> Ms2ger: Excellent. I mean, it’s excellent seeing _you_. Not so much the Orlando bit.
# [11:24] <Ms2ger> Not a fan?
# [11:24] <ato> Not to sound negative, I’m sure I’ll find a big rollercoaster I’ll enjoy (-:
# [11:25] <gsnedders> ato: oh. I thought you were leaving /tomorrow/ morning
# [11:25] <ato> I left this morning.
# [11:26] <gsnedders> And leaving for home tomorrow morning?
# [11:26] <ato> Tuesday next week, spending a few days in Tokyo.
# [11:26] <gsnedders> ah
# [11:26] <Ms2ger> hsivonen> Could someone with Japanese-localized Gmail & browser and a Japanese IP addr use Gmail to send me an email with "日本語" as subject and body?
# [11:26] <ato> The temperature is better than Sapporo (-;
# [11:26] <ato> Ms2ger: I think MikeSmith helped.
# [11:26] <Ms2ger> So it seems
# [11:27] <ato> Helpful chap.
# [11:27] <Ms2ger> And it's all utf8 \o/
# [11:27] <ato> gsnedders: Unfortunately my hotel has ten bars, and I only have five days to spend here.
# [11:28] <gsnedders> ato: two a night? you can totally do that.
# [11:29] <ato> I can tell you’re a problem-solver.
# [11:36] * Joins: jyasskin (~textual@public.cloak)
# [11:49] * Quits: jungkees (~uid24208@public.cloak) ("Connection closed for inactivity")
# [12:12] * Quits: jyasskin (~textual@public.cloak) ("My computer has gone to sleep. ZZZzzz…")
# [12:12] * Quits: anssik (~uid10742@public.cloak) ("Connection closed for inactivity")
# [12:31] * Quits: bhill2 (~bhill2@public.cloak) (Client closed connection)
# [13:41] * Joins: zcorpan (~zcorpan@public.cloak)
# [14:42] * Joins: bhill2 (~bhill2@public.cloak)
# [15:03] * Joins: jyasskin (~textual@public.cloak)
# [15:03] * Parts: ArtB (~ArtB@public.cloak)
# [15:04] * Joins: ArtB (~ArtB@public.cloak)
# [15:19] * Joins: pwz3n0 (~pwz3n0@public.cloak)
# [15:27] * Quits: pwz3n0 (~pwz3n0@public.cloak) (Ping timeout: 180 seconds)
# [15:45] * Quits: jyasskin (~textual@public.cloak) ("My computer has gone to sleep. ZZZzzz…")
# [15:59] * Joins: ArtB1 (~ArtB@public.cloak)
# [16:03] * Quits: ArtB (~ArtB@public.cloak) (Ping timeout: 180 seconds)
# [16:18] * terri_offline is now known as terri
# [17:03] * Quits: rego (~smuxi@public.cloak) (Client closed connection)
# [17:11] * Joins: rego (~smuxi@public.cloak)
# [17:30] * Quits: zcorpan (~zcorpan@public.cloak) (Client closed connection)
# [18:02] * Quits: Ms2ger (~Ms2ger@public.cloak) ("nn")
# [18:04] <M-hrjet> Hi! Are the CSS21 tests supposed to be run with a display having a specific DPI?
# [18:05] <M-hrjet> For example, this reference: https://test.csswg.org/suites/css21_dev/nightly-unstable/xhtml1/reference/border-bottom-width-006-ref.xht
# [18:06] <M-hrjet> .. specifies a height / width of 96 pixels. While the tests use inches / points.
# [18:06] <M-hrjet> These tests pass on screens with 96 DPI, but fail on other displays.
# [18:09] <gsnedders> M-hrjet: so, AFAIK, nowadays they should all pass everywhere
# [18:09] <gsnedders> M-hrjet: also my laptop is about to die :)
# [18:10] <gsnedders> M-hrjet: tl;dr: 96 CSS pixels equal 1 CSS inch
# [18:10] <gsnedders> M-hrjet: but a CSS pixel isn't necessarily a device pixel
# [18:11] <M-hrjet> oh, I didn't know about the distinction. Thanks!
# [18:13] <gsnedders> M-hrjet: https://drafts.csswg.org/css2/syndata.html#length-units
# [18:15] <gsnedders> M-hrjet: what are you working on, BTW?
# [18:18] <M-hrjet> I am developing this https://github.com/UprootLabs/gngr
# [18:18] <M-hrjet> And, as can be seen, I am not entirely familiar with the spec :)
# [18:20] <M-hrjet> The tests are helping a lot though! We have just started running the tests in CI, and have already pinned down a lot of layout / rendering issues.
# [18:27] <gsnedders> You may well be interested in recent discussions on public-css-testsuite, about reducing the amount of metadata in tests, and about what's needed to get more tests running in CI and more tests into the testsuite
# [18:28] <gsnedders> (I'd also warn that the 2.1 test suite really is quite shallow.)
# [18:28] <gsnedders> (It has depth in places, but in general it's relatively shallow.)
# [18:31] <gsnedders> Anyhwo, I need to sleep :)
# [18:31] <M-hrjet> will have a look, thanks! Though, to be honest, we have only just wet our toes. The shallowness, if any, isn't an immediate concern.
# [18:32] <M-hrjet> gsnedders: bye and thanks.
# [20:26] * terri is now known as terri_offline
# [21:20] * terri_offline is now known as terri
# [21:35] * Quits: rego (~smuxi@public.cloak) (Client closed connection)
# [21:42] * Joins: rego (~smuxi@public.cloak)
# [22:04] * Joins: gitbot (~gitbot@public.cloak)
# [22:04] -gitbot:#testing- [web-platform-tests] metromoxie opened pull request #2287: Update scripthash normalization test to *fail* on normalization (master...csp-scripthash-normalization-change) https://github.com/w3c/web-platform-tests/pull/2287
# [22:04] * Parts: gitbot (~gitbot@public.cloak)
# [22:14] * Quits: ArtB1 (~ArtB@public.cloak) ("Leaving.")
# [23:21] * Quits: bhill2 (~bhill2@public.cloak) (Client closed connection)
# Session Close: Fri Oct 30 00:00:00 2015

Think these logs are useful? Then please donate to show your gratitude (and keep them up, of course). Thanks! — Krijn