Log file Viewer

[05:21:56] <devesh> Can I know the address of the slack channel, and how can I join it myself

[05:22:30] <devesh> The slack channel for pypa-dev

[13:02:24] <cjc7373> A basic question.. How to debug the tests? When I tried to debug using tox integration in pycharm, it raised some error, see here: https://dpaste.org/NFZL. Environment: Windows 10, python 3.8

[13:03:02] <pradyunsg> cjc7373: I think you're asking about pip?

[13:03:21] <cjc7373> Yeah

[13:05:00] <cjc7373> pradyunsg

[13:06:46] <pradyunsg> cjc7373: I'll respond in a bit, prepping a meal right now.

[13:37:47] <devesh> pradyunsg \o

[15:24:00] <cjc7373> still stuck in that problem.. looks like tests must be run with installed code (e.g. in virtualenv), so I can't simple run pytest

[15:26:31] <travis-ci> pypa/pip#15485 (master - d7bf2cb : Paul Moore): The build passed.

[15:26:31] <travis-ci> Change view : https://github.com/pypa/pip/compare/73bfea6d28b1...d7bf2cb84b22

[15:26:31] <travis-ci> Build details : https://travis-ci.org/pypa/pip/builds/670963562

[15:32:12] <tos9> was the funny :foo: syntax inspired by some other previous tool

[15:34:17] <pradyunsg> cjc7373: I don't think there's any easy way to do that (plug a debugger with pip's tests) since they run in a subprocess, in a carefully crafted environment that's isolated from the system fairly decently.

[15:35:02] <pradyunsg> s/they run/the pip (executed as part of the test) runs/

[15:35:21] <pradyunsg> tos9: happened before I came around.

[15:35:50] <pradyunsg> devesh: o/

[15:38:53] <devesh> Hey pradyunsg, How are you doing?

[15:39:07] <devesh> Relaxing in the weekend?

[15:39:49] <cjc7373> pradyunsg so we can't make use of the debugger vscode/pycharm offers?

[15:40:41] <pradyunsg> devesh: Helping my parents with their volunteer work today. :)

[15:41:01] <devesh> Nice, no pypa/pip today then

[15:41:07] <pradyunsg> cjc7373: I mean, we could, it'd just take someone who wants to do it badly enough to figure out the details.

[15:41:37] <pradyunsg> devesh: not as planned, but I might end up responding to stuff.

[15:42:11] <pradyunsg> i see there's quite a few mentions + notifications (I am back to 200+ after yesterday)

[15:43:02] <devesh> I think there are like 10 PRs open in last two days

[15:43:33] <devesh> half of them by me (sorry)

[15:44:45] <pradyunsg> lol

[15:44:58] <pradyunsg> No need to apologise for doing good work. 🙃

[15:45:06] <pradyunsg> !m devesh

[15:45:06] <pmxbot> you're doing good work, devesh!

[15:45:16] <pradyunsg> !m pmxbot

[15:45:16] <pmxbot> you're doing good work, pmxbot!

[15:45:20] <cjc7373> pradyunsg I see. So what's your workflow? Do you just run tests in terminal and check out the output?

[15:45:48] <pradyunsg> cjc7373: well. I have a couple of approaches.

[15:46:22] <devesh> thanks pradyunsg: Most of them are trivial, apart from one about 'pip show --json'

[15:47:03] <pradyunsg> cjc7373: I have a debug session set up on VSCode, to run src/pip in a specific virtualenv I have for, well, doing experiments.

[15:47:10] <devesh> I will let you be then, I had a question about mypy annotation, can ask you later

[15:47:35] <pradyunsg> cjc7373: and for tests, I'm firmly in the print-based-debugging camp. 🤷🏻‍♂️

[15:47:52] <pradyunsg> devesh: thanks. :)

[15:47:53] <tos9> pradyunsg: /nod

[15:48:14] <devesh> I think you have the whole night ahead of you anyways :)

[15:49:01] <pradyunsg> devesh: maybe. XD

[15:49:33] <devesh> btw, I wanted to ask about the slack channel as well

[15:51:30] <cjc7373> pradyunsg thanks! no more questions then XD

[15:58:10] <pradyunsg> di_codes / @di: ^ devesh is asking about the Slack.

[15:59:04] <devesh> Yup how to join it, I would prefer that instead of slack, I saw there is a bridge between irc and slack as well

[18:48:05] <devesh> hi pradyunsg

[18:49:56] <devesh> Had a question about this PR: https://github.com/pypa/pip/pull/7977

[19:38:52] <pradyunsg> devesh: ask away

[19:39:47] <devesh> No worries, you have already addressed it. Just that2 other reviewers pointed out the same thing as you did (https://github.com/pypa/pip/pull/7977#pullrequestreview-387724438)

[19:40:07] <devesh> So I was wondering if my clarification/communication on the PR was unclear

[19:40:18] <devesh> Something I can improve on in future

[19:45:32] <pradyunsg> devesh: I don't think your clarification was unclear; but it does come across as "this works therefore it's correct" which is dismissive.

[19:46:26] <devesh> Sure, more research is needed from my end to prove my point I guess, including looking at typeshed like you did

[19:47:49] <pradyunsg> devesh: I'm basically sure that it was not the intent you had there, but it is often difficult to convey intent in pure-text mediums where we don't have the ability to, for example, modulate our pitch to convey intent.

[19:48:18] <devesh> Yes, but I agree on doing more research for your own learning purpose too

[19:49:23] <devesh> BTW, can I now either revert the file and the affected with the incorrect annotation, or add an Any annotation?

[19:49:41] <devesh> Because the other file I have added is fine with the added annotations

[19:50:58] <devesh> So instead of putting in the wrong annotation, we either use any, or visit that file __init__.py later

[19:51:53] <pradyunsg> devesh: I think we should: (1) file an issue with typeshed that "hey, your annotation of difflib.get_closest_matches` is wrong (2) change the PR have the correct annotating and (3) use `pip._internal.utils.typing.cast` (I think that's where it lives now?) to cast the value to the correct type, adding a comment that describes why we're doing a cast there, linking to the issue that was filed with typeshed.

[19:52:10] <pradyunsg> oops, I messed up the backticks, but whatever.

[19:52:39] <pradyunsg> I mean, or we can skip (2) and (3), and come back later. :P

[19:52:44] <pradyunsg> (like you suggested)

[19:52:54] <pradyunsg> Actually, yea, I think I like that better.

[19:53:44] <pradyunsg> Skipping (2) and (3) is basically "do less work now, have a clear indicator that work is needed (the comments at top of file) and don't create more work for later".

[19:53:46] <pradyunsg> devesh: ^

[19:54:12] <devesh> Sure, I can do 1, 2, and 3, or just do 1, I guess as a maintainer whatever is best for you

[19:55:36] <pradyunsg> Just 1 would be fine for pip, yea. Oh, and you can also file a PR to typeshed, where it would a 1 line change. :)

[19:56:08] <devesh> I will revert the __init__.py, and just addi the comment to the PR I am going to file and the old PR

[19:56:15] <pradyunsg> You can also revert that file, as you suggested, so that we can merge in the rest of the PR. That makes sense!

[19:56:23] <devesh> and just leave the other file I added (check.py)

[19:56:30] <pradyunsg> Makes sense, yea.

[19:56:43] <devesh> Let me do that, yes I don't want the PR reviews and work of maintainers go to waste

[19:56:52] <devesh> Have seen so many PR's and good work abandoned

[19:57:57] <devesh> And did you look at the comment here: https://github.com/python/typeshed/issues/2067#issuecomment-382412265 , Is that a valid concern?

[20:18:30] <devesh> Hi pradyunsg: I have filed a bug with typeshed and have removed the file in question as well, please take a look whenever you can :)

[20:24:00] <pradyunsg> devesh: I did look at the typeshed issue you linked to -- I think that's a separate issue.

[20:24:57] <pradyunsg> devesh: thanks for filing the typeshed issue! ^>^

[20:27:42] <devesh> No worries pradyunsg, I didn't know of typeshed and it's relation to type annotations before this

[20:28:13] <pradyunsg> devesh: now you know that! yay!

[20:28:14] <devesh> But feel free to add more details if needed in the ticket at https://github.com/python/typeshed/issues/3906

[20:29:25] <devesh> Hehe, please also relook at the PR pradyunsg :), especially if the comment in the __init__.py correctly notes down the issue :)

[20:29:41] <pradyunsg> devesh: sure, I'll do that in a bit. :)

[20:29:57] <pradyunsg> need to go do some chores now.

[20:30:18] <devesh> chores at 2 am, woah

[20:34:03] <pradyunsg> those dishes don't wash themselves. :P

[20:42:36] <devesh> yep

[20:43:02] <devesh> Anyways, I'll take off, seems like you have a few more pr's to review :)

[20:44:01] <devesh> I've responded to your questions anyways, and I see uranusjr is awake too, is he not on IRC?

[20:47:42] <pradyunsg> devesh: Nope.

[22:29:50] <travis-ci> pypa/twine#1353 (master - f7402e0 : Brian Rutledge): The build was fixed.

[22:29:50] <travis-ci> Change view : https://github.com/pypa/twine/compare/5482d6877df0cd90545c4708e5abe13b359f8a08...f7402e0d2e1c1e6ecde61dc8bcba515b807a48a0

[22:29:50] <travis-ci> Build details : https://travis-ci.org/pypa/twine/builds/671082691