| 08:50 | <annevk> | MikeSmith: DOM build script makes the validator throw: https://travis-ci.org/whatwg/dom/builds/364015253? |
| 08:56 | <MikeSmith> | annevk: looking right now |
| 08:56 | <MikeSmith> | ah yeah |
| 08:56 | <MikeSmith> | same issue as https://github.com/validator/validator/issues/634 |
| 08:57 | <MikeSmith> | will fix it right now |
| 08:59 | <annevk> | hsivonen: miketaylr: it seems there's no Firefox bug for Event's returnValue? |
| 09:00 | annevk | can only find one for window.event and Event.prototype.srcElement |
| 09:20 | <nox> | annevk: Care to review https://github.com/w3c/web-platform-tests/pull/10356 and https://github.com/w3c/web-platform-tests/pull/10357 if you have some time, please? |
| 09:23 | <annevk> | nox: does idlharness even care about dictionaries? |
| 09:23 | <nox> | annevk: It ERROR'd without that. |
| 09:24 | <nox> | annevk: And other tests have untested dictionary IDL definitions. |
| 09:25 | <annevk> | nox: okay, r+ |
| 09:27 | <nox> | annevk: So am I supposed to hit the big green button myself? |
| 09:28 | <annevk> | nox: please do "rebase and merge" or "squash" (to clean up commit message if you want to) |
| 09:28 | <annevk> | nox: unless you really want me to do it |
| 09:29 | <nox> | Ok. |
| 09:36 | <hsivonen> | annevk: so it seems |
| 09:36 | <hsivonen> | no bug for returnVaule that is |
| 09:36 | <hsivonen> | Value |
| 09:36 | <annevk> | hsivonen: I'll file one then |
| 10:24 | <MikeSmith> | annevk: OK build now un-broke https://travis-ci.org/whatwg/dom/builds/364015253 |
| 10:25 | <JakeA> | In WPT's python file handlers, is there anything special I need to do to import a module in the same folder? |
| 10:27 | <JakeA> | `from utils import create_wav_header` fails with "No module named utils", but `utils.py` exists in the same folder |
| 10:27 | <jgraham> | JakeA: I think the pwd is the server root rather than the directory with the file in it |
| 10:27 | <gsnedders> | JakeA: I am heavily unsurprised that doesn't work |
| 10:27 | <jgraham> | JakeA: sys.path.insert(0, os.path.basename(__file__)) |
| 10:28 | <jgraham> | and then import? |
| 10:28 | <jgraham> | JakeA: Not sure how much effort it would be to fix this, but you could file an issue? |
| 10:29 | <JakeA> | jgraham: shall do. The sys.path.insert thing doesn't seem to change anything in this case |
| 10:32 | <annevk> | MikeSmith: thanks |
| 10:34 | <annevk> | mkwst: reportedly there's other PRs timing out too |
| 10:35 | <annevk> | mkwst: the suggestion from GitHub support is indeed to open a new PR if you want to continue working on it now... |
| 10:36 | <annevk> | mkwst: given that you have write access to whatwg/fetch you should be able to pull up the PR through GitHub Desktop and such at least |
| 10:37 | <JakeA> | jgraham: gsnedders: https://github.com/w3c/web-platform-tests/issues/10378 |
| 10:51 | <jgraham> | JakeA: Takk |
| 12:21 | <nox> | TabAtkins: I'm pretty curious about the discussion in https://github.com/w3c/css-houdini-drafts/issues/729#issuecomment-379731917 |
| 12:21 | <nox> | AFAICT people were mildly in favour of it, and no one was against it, so… it was rejected? |
| 12:22 | <nox> | Ah never mind, skipped a log line, disregard that. |
| 12:23 | <gsnedders> | nox accepts me as people! success! |
| 12:24 | <nox> | What. :D |
| 12:43 | <TabAtkins> | JakeA: When importing from your own directly, use absolute relative imports - `from .utils import create_wav_header`. |
| 13:16 | <JakeA> | TabAtkins: that didn't work either in this case |
| 13:17 | <TabAtkins> | Weird. |
| 13:18 | <gsnedders> | JakeA, TabAtkins: you'd need a __init__.py in the directory for that |
| 13:18 | <JakeA> | gsnedders: I have that too |
| 13:18 | <gsnedders> | oh actually you need to import it as a package so yeah that probably doesn't work |
| 13:18 | <TabAtkins> | gsnedders: Oh, is it a directory? I thought he said it was just an update.py file. |
| 13:18 | <TabAtkins> | In which case it's a module already, nothing else needs to be done |
| 13:19 | <gsnedders> | TabAtkins: no, for "from ." anything you have to already be in a package (with an __init__.py) |
| 13:19 | <TabAtkins> | Oh, you have to be *in* a package, yeah. wpt not being in a package has annoyed me in the past. |
| 13:20 | <gsnedders> | and being in a package is partly down to how you import that Python script first |
| 13:21 | <gsnedders> | I mean we could try and import stuff as a package, but we can't really expect __init__.py in every file up to the root |
| 13:21 | <TabAtkins> | Sure you can? I mean, like, Bikeshed is a package and contains packages. It's easy and convenient. |
| 13:22 | <gsnedders> | well I mean I think people would object if we added an __init__.py to every directory in WPT |
| 13:23 | <TabAtkins> | Oh, all the *data* folders. They don't contain Python files, do they? |
| 13:23 | <gsnedders> | They can, you can have Python files that get loaded by the server. |
| 13:24 | <TabAtkins> | Ah. So why are you expecting pushback about that? |
| 13:24 | <gsnedders> | Because there's relatively few of them. |
| 13:25 | <TabAtkins> | So you can add them to just the folders with Python in them, and their parents. |