dev-contrib

[austere hornet]

sir-lancebot#901 please

[dusky shore]

GitHub

Madlibs

[patent pivot]

yea dm_sent lgtm

[static canyon]

That's a lot of tests for the user command lmao

[patent pivot]

needs a branch update

but otherwise probably can merge

[static canyon]

Thanks joe

I'm rather confused by this. Can it be ignored?

I've got 8 failing tests and no clue why

What's weird is that works fine on the above test

This is the breakdown of the errors:```py
FAILED tests/bot/exts/moderation/test_incidents.py::TestMakeEmbed::test_make_embed_content - AssertionError: 'this is an incident' != 'this is an incident\n\n__*Reported on <t:1:D> at ...'

FAILED tests/bot/exts/moderation/test_incidents.py::TestArchive::test_archive_clyde_username - AttributeError: 'NoneType' object has no attribute 'kwargs'

FAILED tests/bot/exts/info/test_information.py::UserCommandTests::test_regular_user_can_explicitly_target_themselves - AssertionError: expected call not found.

FAILED tests/bot/exts/info/test_information.py::UserCommandTests::test_regular_user_may_use_command_in_bot_commands_channel - AssertionError: expected call not found.

FAILED tests/bot/exts/info/test_information.py::UserCommandTests::test_moderators_can_target_another_member - AssertionError: expected call not found.

FAILED tests/bot/exts/info/test_information.py::UserCommandTests::test_staff_members_can_bypass_channel_restriction - AssertionError: expected call not found.

FAILED tests/bot/exts/moderation/test_incidents.py::TestProcessEvent::test_process_event_confirmation_task_is_awaited - AssertionError: Expected mock to have been awaited.

FAILED tests/bot/exts/moderation/test_incidents.py::TestArchive::test_archive_relays_incident - AssertionError: Expected 'send' to be called once. Called 0 times.```

Hmm I might have an idea

I think I branched from a different PR instead of main

[gritty wind]

That seems to most likely be a bug

The method is already an instance method (it's being called from self.cog), and you don't need to explicitly pass self in those cases

You'd really only pass self explicitly if you're calling it like Class.method() instead of Class().method()

[static canyon]

Changes I wrote on branch x are also showing as unstaged changes on branch y and vice-versa

[static canyon]

So remove the self.cog?

[gritty wind]

I assume so

[static canyon]

Aight

[gritty wind]

Mind you I don't have actual context

[fallen patrol]

oh wait

I'm tired

[gritty wind]

🔨

[fallen patrol]

that never happened

[gritty wind]

I thought you were just making a joke lol

[static canyon]

What confuses me is this

It's fine when another (optional) argument is passed

[fallen patrol]

maybe, idek

[gritty wind]

What's the actual signature for that function

[static canyon]

    @command(name="user", aliases=["user_info", "member", "member_info", "u"])
    async def user_info(self, ctx: Context, user_or_message: Union[MemberOrUser, Message] = None) -> None:

[fallen patrol]

that's tests, isn't it?

[gritty wind]

Strange that it isn't complaining on the first one, really feels like it ought to

[static canyon]

yeah

[fallen patrol]

pretty sure the tests have self.cog

[static canyon]

I mean yeah, that's what's giving the issue

[gritty wind]

Unless that method is mocked in the top example

[fallen patrol]

is it Com and pushed cause I'll pull down and look at the tests

[static canyon]

Nope

[gritty wind]

wery veird

[fallen patrol]

going to my pc now

I told myself I wouldn't develop at all today...

but broke that... 5 minutes after midnight so

[static canyon]

Keep in mind I didn't write these tests, so someone else passed self.cog in and they worked before my PR 🤷

[static canyon]

Nice lol

[gritty wind]

not even new years but we're already breaking resolutions

[fallen patrol]

no, my monitor still works 🤡

[gritty wind]

Do we know for a fact they are collected and ran 🤡

I'd assume so

[fallen patrol]

which branch or pull is this?

[gritty wind]

Anyways, as long as it works, I guess it's fine

From what Tizzy just said, I assume master

[fallen patrol]

master.

checks didn't fail so

seems like incident-archive-msg-improvements @static canyon ?

also is it pushed

[static canyon]

Yeah

Not pushed yet because I'm fixing stuff

[fallen patrol]

ah, could you push what you have so i can debug the tests, or are they broken on what's pushed?

[static canyon]

I removed self.cog from all the function calls and now there's 8 tests failing still but it's a different 8

[gritty wind]

lol nice

whatcha got

[static canyon]

FAILED tests/bot/exts/info/test_information.py::UserCommandTests::test_staff_members_can_bypass_channel_restriction - TypeError: user_info() missing 1 required positional argument: 'ctx'
```niceee

[gritty wind]

Heh

[static canyon]

The call in that one ispy await self.cog.user_info(ctx)

[gritty wind]

Can you show the init where it sets self.cog

[static canyon]

So I guess for some reason self.cog doesn't get passed?

Err there's multiple definitions of self.cog

[gritty wind]

o no

that would help explain the differing behavior

[static canyon]

I think this is the one we want```py
@unittest.mock.patch("bot.exts.info.information.constants")
class UserCommandTests(unittest.IsolatedAsyncioTestCase):
"""Tests for the !user command."""

def setUp(self):
    """Set up steps executed before each test is run."""
    self.bot = helpers.MockBot()
    self.cog = information.Information(self.bot)

[gritty wind]

With that one, it'll definitely not require passing self.cog

Where are the other definitions

[fallen patrol]

lemme take a look

which file?

[static canyon]

class UserInfractionHelperMethodTests(unittest.IsolatedAsyncioTestCase):
    """Tests for the helper methods of the `!user` command."""

    def setUp(self):
        """Common set-up steps done before for each test."""
        self.bot = helpers.MockBot()
        self.bot.api_client.get = unittest.mock.AsyncMock()
        self.cog = information.Information(self.bot)
        self.member = helpers.MockMember(id=1234)```

[fallen patrol]

;-; ~~a screenshot smh I can't copy paste that

tests/bot/exts/info/test_information.py

[gritty wind]

Hm, that's more or less the same

I wonder why it's failing

[static canyon]

class InformationCogTests(unittest.IsolatedAsyncioTestCase):
    """Tests the Information cog."""

    @classmethod
    def setUpClass(cls):
        cls.moderator_role = helpers.MockRole(name="Moderator", id=constants.Roles.moderators)

    def setUp(self):
        """Sets up fresh objects for each test."""
        self.bot = helpers.MockBot()

        self.cog = information.Information(self.bot)
``````py
@unittest.mock.patch("bot.exts.info.information.constants.MODERATION_CHANNELS", new=[50])
class UserEmbedTests(unittest.IsolatedAsyncioTestCase):
    """Tests for the creation of the `!user` embed."""

    def setUp(self):
        """Common set-up steps done before for each test."""
        self.bot = helpers.MockBot()
        self.bot.api_client.get = unittest.mock.AsyncMock()
        self.cog = information.Information(self.bot)
``````py
@unittest.mock.patch("bot.exts.info.information.constants")
class UserCommandTests(unittest.IsolatedAsyncioTestCase):
    """Tests for the `!user` command."""

    def setUp(self):
        """Set up steps executed before each test is run."""
        self.bot = helpers.MockBot()
        self.cog = information.Information(self.bot)
``````py
class UserInfractionHelperMethodTests(unittest.IsolatedAsyncioTestCase):
    """Tests for the helper methods of the `!user` command."""

    def setUp(self):
        """Common set-up steps done before for each test."""
        self.bot = helpers.MockBot()
        self.bot.api_client.get = unittest.mock.AsyncMock()
        self.cog = information.Information(self.bot)
        self.member = helpers.MockMember(id=1234)```

These are the four definitions @gritty wind

[fallen patrol]

!paste

[stable mountain]

Pasting large amounts of code

If your code is too long to fit in a codeblock in discord, you can paste your code here:
https://paste.pythondiscord.com/

After pasting your code, save it by clicking the floppy disk icon in the top right, or by typing ctrl + S. After doing that, the URL should change. Copy the URL and post it here so others can see it.

[gritty wind]

They all just set up self.cog in the same way

[static canyon]

Yeah

So why doesn't it work

[gritty wind]

Which tests fail

[static canyon]

lol

[gritty wind]

Ah I see it up

[static canyon]

Right now it's these that fail:```py
FAILED tests/bot/exts/info/test_information.py::UserCommandTests::test_staff_members_can_bypass_channel_restriction - TypeError: user_info() missing 1 required positional argument: 'ctx'

FAILED tests/bot/exts/info/test_information.py::UserCommandTests::test_regular_user_may_use_command_in_bot_commands_channel - TypeError: user_info() missing 1 required positional argum...

FAILED tests/bot/exts/info/test_information.py::UserCommandTests::test_regular_user_can_explicitly_target_themselves - AttributeError: Mock object has no attribute 'author'

FAILED tests/bot/exts/info/test_information.py::UserCommandTests::test_moderators_can_target_another_member - AttributeError: Mock object has no attribute 'author'

FAILED tests/bot/exts/info/test_information.py::InformationCogTests::test_role_info_command - AttributeError: Mock object has no attribute 'send'

FAILED tests/bot/exts/info/test_information.py::InformationCogTests::test_roles_command_command - TypeError: roles_info() missing 1 required positional argument: 'ctx'

FAILED tests/bot/exts/info/test_information.py::UserCommandTests::test_regular_member_cannot_target_another_member - AttributeError: Mock object has no attribute 'author'

FAILED tests/bot/exts/info/test_information.py::UserCommandTests::test_regular_member_cannot_use_command_outside_of_bot_commands - TypeError: user_info() missing 1 required positional ...```

[austere hornet]

Yes you can, right click the image and click "copy image"

[static canyon]

Not what they were after lol

[gritty wind]

big brain

[austere hornet]

Ohh

[static canyon]

They wanted the file path copy-pasteable

[austere hornet]

Ahh

Mb

[fallen patrol]

dw, it was funny

[austere hornet]

Yeah ik

[static canyon]

CC @gritty wind

[fallen patrol]

...

all tests in tests/bot/exts/info/test_information.py pass locally

so uh

have fun 🤡

[static canyon]

lol

[gritty wind]

Well, did you make the changes locally lol

[fallen patrol]

no, because i don't have said changes smh

they weren't pushed or pasted or anything 😔

[static canyon]

The changes I pushed that broke the tests are on the branch you mentioned above

Wait no

user-improvement branch

[fallen patrol]

oh

[static canyon]

The incidents one is for a different pr

[fallen patrol]

oh wait

4 tests fail on the branch i listed above

[static canyon]

Yeah, I'm nowhere near done with that

So just ignore it all

[fallen patrol]

that seems like just not finished rather than bug tho

yeah

although now i'm gonna run the tests on main jussst to have a local baseline

and then check out the actually correct branch

[static canyon]

lol

[fallen patrol]

so disgraceful smh

imagine writing tests lol can't be me

[static canyon]

Should I pass the self.cog back in again? Since that seems to be needed 🤷 @gritty wind

[static canyon]

I had to write my own test suite and tests for my coursework

[gritty wind]

I guess so, but I still want to know what dark fuckery is being done to make that a thing

[fallen patrol]

i copied the pydis mocks and then made some improvements for my own code

bot#1911 to share them upstream

[dusky shore]

GitHub

Upgrade the test suite mocks

[gritty wind]

Git blame says this is KS' doing 👀

[fallen patrol]

git blame always says sometone did it

...i'm leaving that typo

[gritty wind]

Someone mocked the cog, instead of making an actual instance of the cog

Turns out it's not a very good mock /s

At least in the methods where it fails with the changes

[fallen patrol]

wait i actually

huh interesting

[gritty wind]

I honestly have no idea where though lol

[fallen patrol]

what if you use a mock cog and set functions to functions on the mock cog, so you'd be testing just the specific functions

^ that's not relevant but is speculation for later

[gritty wind]

Wait

Hm no

self.cog is a mock when the debugger is inside the actual function

It's because I'm looking at self when it's set to ctx because of the arg fuckery

So ignore that part

[fallen patrol]

oh i get 12 failed tests

[gritty wind]

This might explain it though:

> type(self.cog)
<class 'bot.exts.info.information.Information'>

> type(self.cog.user_info)
<class 'bot.monkey_patches.Command'>

[static canyon]

lol

fgs

[static canyon]

4(?) I fixed because it's just adding False to the end of each create_user_embed

[fallen patrol]

!d unittest.IsolatedAsyncioTestCase

[stable mountain]

unittest.IsolatedAsyncioTestCase

class unittest.IsolatedAsyncioTestCase(methodName='runTest')```
This class provides an API similar to [`TestCase`](https://docs.python.org/3/library/unittest.html#unittest.TestCase "unittest.TestCase") and also accepts coroutines as test functions.

New in version 3.8.

[fallen patrol]

some of these failures are valid....

oh wait

[fallen patrol]

for those questioning my sanity, read what ppl who mentioned me wrote

[static canyon]

I've now got only 5 failing tests

Now clue how but that's good

FAILED tests/bot/exts/info/test_information.py::UserCommandTests::test_moderators_can_target_another_member - AssertionError: expected call not found.

FAILED tests/bot/exts/info/test_information.py::UserCommandTests::test_regular_user_can_explicitly_target_themselves - AssertionError: expected call not found.

FAILED tests/bot/exts/info/test_information.py::InformationCogTests::test_role_info_command - AttributeError: Mock object has no attribute 'send'

FAILED tests/bot/exts/info/test_information.py::UserCommandTests::test_staff_members_can_bypass_channel_restriction - AssertionError: expected call not found.

FAILED tests/bot/exts/info/test_information.py::UserCommandTests::test_regular_user_may_use_command_in_bot_commands_channel - AssertionError: expected call not found.

[fallen patrol]

should just make that have a default in the code lol

!d unittest.mock.patch

[static canyon]

It looks weird with a default though because the command itself (which calls this function) has a default

[stable mountain]

unittest.mock.patch

unittest.mock.patch(target, new=DEFAULT, spec=None, create=False, spec_set=None, autospec=None, new_callable=None, **kwargs)```
[`patch()`](https://docs.python.org/3/library/unittest.mock.html#unittest.mock.patch "unittest.mock.patch") acts as a function decorator, class decorator or a context manager. Inside the body of the function or with statement, the *target* is patched with a *new* object. When the function/with statement exits the patch is undone.

If *new* is omitted, then the target is replaced with an [`AsyncMock`](https://docs.python.org/3/library/unittest.mock.html#unittest.mock.AsyncMock "unittest.mock.AsyncMock") if the patched object is an async function or a [`MagicMock`](https://docs.python.org/3/library/unittest.mock.html#unittest.mock.MagicMock "unittest.mock.MagicMock") otherwise. If [`patch()`](https://docs.python.org/3/library/unittest.mock.html#unittest.mock.patch "unittest.mock.patch") is used as a decorator and *new* is omitted, the created mock is passed in as an extra argument to the decorated function. If [`patch()`](https://docs.python.org/3/library/unittest.mock.html#unittest.mock.patch "unittest.mock.patch") is used as a context manager the created mock is returned by the context manager.

[fallen patrol]

ah

[static canyon]

Now 4 (fixed the attribute error)

[fallen patrol]

i have 4 now too lol

[static canyon]

Okay, cool

[fallen patrol]

8 of them were missing args

[static canyon]

Just to confirm, this is the 4 right?```py
FAILED tests/bot/exts/info/test_information.py::UserCommandTests::test_staff_members_can_bypass_channel_restriction - AssertionError: expected call not found.

FAILED tests/bot/exts/info/test_information.py::UserCommandTests::test_regular_user_may_use_command_in_bot_commands_channel - AssertionError: expected call not found.

FAILED tests/bot/exts/info/test_information.py::UserCommandTests::test_regular_user_can_explicitly_target_themselves - AssertionError: expected call not found.

FAILED tests/bot/exts/info/test_information.py::UserCommandTests::test_moderators_can_target_another_member - AssertionError: expected call not found.

[fallen patrol]

btw you might find this helpful: --sw or --stepwise

also

if you pass tests/bot/exts/info/test_information.py::UserCommandTests::test_moderators_can_target_another_member to pytest

pytest tests/bot/exts/info/test_information.py::UserCommandTests::test_moderators_can_target_another_member
``` you can run just that specific test

[static canyon]

👍

E           AssertionError: expected call not found.
E           Expected: create_user_embed(<MockContext spec_set='Context' id='1788322016032'>, <MockMember spec_set='Member' id='1788321768208'>)
E           Actual: create_user_embed(<MockContext spec_set='Context' id='1788322016032'>, <MockMember spec_set='Member' id='1788321768208'>, False)
```Looks like I just need to update what it expects

Yep, all working now

[fallen patrol]

@static canyon

[static canyon]

create_embed.assert_called_once_with(ctx, self.target)```to```py
create_embed.assert_called_once_with(ctx, self.target, False)```etc.

[fallen patrol]

lol we got it at the exact same time

I have @gritty wind to thank for helping me 10/10 support

[static canyon]

Thanks guys 😄

[fallen patrol]

1 test failed 👀

[static canyon]

?

[fallen patrol]

wait

nvm nothing is wrong

[gritty wind]

Any contribs around to merge bot#2028 please?

[dusky shore]

GitHub

Add missing infraction id parameter & change ctx.bot to self.bot

[brazen charm]

why the change from ctx.bot to self.bot?

[static canyon]

Scale asked for it

[brazen charm]

and why is it being passed around when it's causing issues and is available from the global instance

[gritty wind]

To clarify, the typehint isn't wrong

What we're passing is wrong compared to the type hint

afaik we don't patch the internal ctx.bot object

just our global bot instance

Unless those are linked somehow?

[brazen charm]

pretty sure that should be our subclasses bot

[hoary haven]

ot names didn't change today

[gritty wind]

Ah, that makes sense

[fallen patrol]

bot redeployed at exactly the change time #dev-log message

[hoary haven]

oh lol i was like ok yeah makes sense bot wants a day off for christmas

[fallen patrol]

which actually means... all tasks that were supposed to run probably

hmm did anything important not run

nah seems like they all ran or aren't important

[gritty wind]

!int e ```py
from bot.constants import Channels, MODERATION_ROLES
CHANNELS = (Channels.off_topic_0, Channels.off_topic_1, Channels.off_topic_2)

channel_0_name, channel_1_name, channel_2_name = await bot.api_client.get(
'bot/off-topic-channel-names', params={'random_items': 3}
)
channel_0, channel_1, channel_2 = (bot.get_channel(channel_id) for channel_id in CHANNELS)

await channel_0.edit(name=f'ot0-{channel_0_name}')
await channel_1.edit(name=f'ot1-{channel_1_name}')
await channel_2.edit(name=f'ot2-{channel_2_name}')
log.debug(
"Updated off-topic channel names to"
f" {channel_0_name}, {channel_1_name} and {channel_2_name}"
)

[stable mountain]

In [1]: from bot.constants import Channels, MODERATION_ROLES
   ...: CHANNELS = (Channels.off_topic_0, Channels.off_topic_1, Channels.off_topic_2)
   ...: channel_0_name, channel_1_name, channel_2_name = await bot.api_client.get(
   ...:     'bot/off-topic-channel-names', params={'random_items': 3}
   ...: )
   ...: channel_0, channel_1, channel_2 = (bot.get_channel(channel_id) for channel_id in CHANNELS)
   ...: await channel_0.edit(name=f'ot0-{channel_0_name}')
   ...: await channel_1.edit(name=f'ot1-{channel_1_name}')
   ...: await channel_2.edit(name=f'ot2-{channel_2_name}')
   ...: log.debug(
   ...:     "Updated off-topic channel names to"
   ...:     f" {channel_0_name}, {channel_1_name} and {channel_2_name}"
   ...: )
   ...: 
Out[1]: 
```... response truncated; full contents at https://paste.pythondiscord.com/daqipamozu.py

[gritty wind]

First time getting to see the changes roll out

[hoary haven]

internals-and-pepsi yum

lol go to sleep scale

[gritty wind]

Maybe >.>

Have to do midnight patches

[fallen patrol]

oh?

what tz?

[patent pivot]

coordinated scaleios clock

[fallen patrol]

.time CSC

TI(re)L .time is a command

[austere hornet]

Wait is it? .timed is though

I just tried in #sir-lancebot-playground and it's not a thing

[fallen patrol]

its an alias

[austere hornet]

Oh

But you gave me a great idea - we should have a .timezone-convert <time> <from which timezone> <to what timezone> that gives you the difference in # of hours between 2 timezones. What do people think about that?

(If we don't already have that, that is)

[tawdry vapor]

That's probably easy enough to do with an eval already

[fallen patrol]

well yes but also... its implemented in tzdata which is imported with dateutil.tz and then you pass that to a method and then use datetime.datetime to get the current time and then shift it... etc

given we're adding .epoch, i don't see how a timezone command shouldn't be added because it can be done in eval

[tawdry vapor]

I misunderstood it as converting a time to a different timezone, which would be simpler. Anyway, that bot is just meant to be for fun and learning so I don't mind what gets added.

[fallen patrol]

ah gotcha 👍

[austere hornet]

I mean, that's basically what I was proposing. I might open an issue about it soon

[tawdry vapor]

Ah i missed the <time> argument you had

So it would essentially be like

!e import arrow; print(arrow.utcnow().to("Europe/Berlin"))

[stable mountain]

@tawdry vapor :white_check_mark: Your eval job has completed with return code 0.

2021-12-26T05:02:19.194466+01:00

[fallen patrol]

say....

do the bots share the same redis instance?

[tawdry vapor]

Yes, I think so.

They connect to the same host and I only see one redis container in prod

[fallen patrol]

neat

oh come on!

github!!!

[short snow]

hmm?

[fallen patrol]

my actually important review comment wasn't saved

[short snow]

oh lol

[fallen patrol]

was about api requests

[clever wraith]

@thorny obsidian I believe the second PR you meant to link in your form issue is https://github.com/python-discord/forms-backend/pull/92, you linked our PR twice

also fwiw

returned to the frontend is documented on

I actually wonder in which cases 500 is returned

I guess a general 500 handler should be added

oi, I found a bug when reading the unittesting code (misuing strip)

and also it doesn't catch base exception from the user code meaning you could trick it into generating a 500

[clever wraith]

how do you make a token

I guess you must run the frontend for that?

it would be nice to have a utility script to generate one

or even have just the backend drop one in the console if debug mode is enabled (cc @gritty wind)

[gritty wind]

Nah, it raises a user code load error, which is caught, and grouped with other bad user code

And reported as "failed" for the suite

[gritty wind]

Not possible, can't do the discord oauth

But we've added a special form for debug that will be sent if no other form exists, which does have oauth

so yes, you do need to run the frontend, but then you'll get a token

This is all documented in the getting started guide

The backend also has /admin_dev route which allows you to add any admin user without requiring a pre-existing admin account, which resolves that cyclical dependency

Because I've just done it right now, the steps to getting a full fledged admin token are:

1. Run backend and frontend
2. Go to frontend
3. Auth with the only form there
4. Copy the cookie
5. Make a request to /admin_dev with your ID
6. You're done

[clever wraith]

No, because the code catches Exception and not BaseException, and SystemExit inherits from the latter

[gritty wind]

Ah, I see what you mean

[clever wraith]

also could you assign me to the issue?

[gritty wind]

Sure, but don't expect a review from anyone :P

Assigned

[clever wraith]

haha, thanks

[gritty wind]

Actually, if you get it in now, I can review

[clever wraith]

yeah

just gimme some time to fork and all

it is a quite small PR

[gritty wind]

Right

[clever wraith]

no form currently in prod uses uppercase letters in its ID?

[gritty wind]

No

[clever wraith]

https://github.com/python-discord/forms-backend/pull/126

Sending DMs also sound interesting

[gritty wind]

It's more or less done iirc

[clever wraith]

oh wait you are working on it

bad squishie not being assigned to the issue

[gritty wind]

Is it possible to have the form ID be automatically lowercased when fetching/writing from the DB? I have a feeling it'll get forgotten eventually

It might especially be lost in testing because people won't think to

[clever wraith]

that's what the constraint is doing

[gritty wind]

I mean for the other places where you're manually calling lower

[clever wraith]

I submitted a form called siGNUps and it saved it to the db as signups

ah, the problem is the frontend here

hmm

maybe a validator could do the trick

nope

I don't think that's a thing you can do

[gritty wind]

Alright that’s fine

Thanks Chris, but I’ve hooked up my brain to the forms database

I know everything

[vale ibex]

lol

[clever wraith]

plot twist, chris hooked his brain onto yours

[vale ibex]

I imagine you could do that is we used request models

then the pydantic convertor would work

[gritty wind]

Not possible, he hasn’t gone mad yet

[gritty wind]

Replying to this

[vale ibex]

lol yea

[clever wraith]

I thought that's how the admin team worked

disappointed

[gritty wind]

I think we do that with the validation already, just in some places we only need ID instead of an entire form object

[clever wraith]

would it actually run validation before fetching the DB?

[gritty wind]

This just means we don’t need to make all props optional, or require all props on all requests

Hm we could create a base form model that’s just ID

It would run it on the arg passed to the requestitsef

[clever wraith]

also I'm not sure why you have pre=True on some models with a single validator haha

[vale ibex]

I more mean ```diff

• async def get(self, request: Request, form_id: some_con_str) -> JSONResponse:

• async def get(self, request: Request) -> JSONResponse:
  """Returns single form information by ID."""
  admin = request.user.admin if request.user.is_authenticated else False

•   form_id = request.path_params["form_id"].lower()
  

then you wouldn't need to lower() all the path params

[gritty wind]

How would the request actually look

[vale ibex]

since it's done by the type hint

[clever wraith]

I guess it is the same problem as you have to remember to use the constraint

[vale ibex]

true

[vale ibex]

the same

I usually do this with the decorator, so it might be slightly different syntax like this

IE ```py
@app.get("/foo/{bar}")
async def read_bar(bar: ham):

[clever wraith]

well you'd have to swap _id for form_id wouldn't you

[vale ibex]

depends what the actual request looks like

I'm not familiar with forms directly, but that's a format that would work at least

I had assumed since it was path_params["form_id"] the path param was form_id

[clever wraith]

it is a bit of a mixed bag, going by ak's memory haha

yeah in that case it is form_id

cool

[vale ibex]

cool

[clever wraith]

should be giveOwOy but oh well can't have everything in life

oh also fyi @vale ibex

Getting into a bad data situation like this is directly blocked in the command anyway, so shouldn't be possible.
Our github account is linked to our three discord accounts haha

if it doesn't break that's fine, we can just poke y'all :3

[vale ibex]

Ah right

well it won't break yea, dict generators just work

just the same as if you overwrite a dict key normally

[clever wraith]

yeah yeah, not sure why she wrote that

[vale ibex]

I wouldn't mind a subsequent PR to support multiple links, but don't have the time to do it myself atm

[clever wraith]

I would say it is fine

[vale ibex]

at least not until the new year

[cold island]

@vale ibex whenever you've got a moment, your requested changes on bot#1889 were addressed

[dusky shore]

GitHub

feat: added url parsing to the filter

[clever wraith]

now I wish we wouldn't have deleted our old test server

[vale ibex]

ah, not quite, I thought they were going to be implementing this pseudo code that we talked about #dev-contrib message

if not, it's something we can do in a following pr anyway

.

[cold island]

The main difference is that it still relies on there being a schema

[vale ibex]

Yea

[cold island]

So I guess I'm fine with either

(now or in a future PR)

[vale ibex]

I mean if this is a clear improvement, then we can merge

[cold island]

There's no particular rush, I'm just trying to clear filter-related PRs

[vale ibex]

yea makes sense 😛

Ah, of course, there's lock conflicts 😅

[cold island]

yeah I can relock in the evening if no one gets to it

[vale ibex]

We'd lose policy bot then tho

[cold island]

oh hmm

[vale ibex]

I can do it right now, lol

@gritty wind are you here to give a proxy approval?

[clever wraith]

I can give a contrib approval, that should work?

[vale ibex]

yea should do

jsut checking out the pr now

poetry lock takinga while on this laptop

might do a @hoary haven and buy a new macbook

[clever wraith]

I get the error that blue mentioned on the aoc role PR

it will error if someone is on the leaderboard and doesn't have any linked account, from what I can see

[vale ibex]

Ahhhhh, I thought he was referring to the member_aoc_info["name"] slice

but now I see it was the slice that uses that

[clever wraith]

I don't know how you managed that one chris lol

there are multiple programming error

diff --git a/bot/exts/events/advent_of_code/_cog.py b/bot/exts/events/advent_of_code/_cog.py
index 7315d166..90c42783 100644
--- a/bot/exts/events/advent_of_code/_cog.py
+++ b/bot/exts/events/advent_of_code/_cog.py
@@ -93,25 +93,25 @@ class AdventOfCode(commands.Cog):
                 # Only give the role to people who have completed all 50 stars
                 continue
 
-            member_id = aoc_name_to_member_id[member_aoc_info["name"]]
+            member_id = aoc_name_to_member_id.get(member_aoc_info["name"], None)
             if not member_id:
                 continue
 
-            member = members.get_or_fetch_member(guild, member_id)
+            member = await members.get_or_fetch_member(guild, member_id)
             if member is None:
                 continue
 
             if member in completionist_role.members:
                 if await self.completionist_block_list.contains(member_id):
-                    await members.handle_role_change(member, member.remove_roles)
+                    await members.handle_role_change(member, member.remove_roles, completionist_role)
                 continue
 
-            await members.handle_role_change(member, member.add_roles)
+            await members.handle_role_change(member, member.add_roles, completionist_role)
             members_who_deserve_role.add(member)
 
         for member in set(completionist_role.members) - members_who_deserve_role:
             # Remove the role from members who have it but don't deserve it (been removed from the leaderboard)
-            await members.handle_role_change(member, member.remove_roles)
+            await members.handle_role_change(member, member.remove_roles, completionist_role)
 
     @commands.group(name="adventofcode", aliases=("aoc",))
     @whitelist_override(channels=AOC_WHITELIST)```

if you want to quickly fix them, this code should work

[gritty wind]

Sorry I caught a glimpse of myself in the mirror and have been distracted staring

Do you still need me for anything

[clever wraith]

Fame and money

Oh and a proxy approval

[vale ibex]

kk bot#1889 is pushed up

[dusky shore]

GitHub

feat: added url parsing to the filter

[vale ibex]

I blame baileys

but it's pushed now

will squash before merge

[clever wraith]

Lmao

[static canyon]

RE your comment on bot#2008: https://github.com/python-discord/bot/pull/2008#discussion_r771915170@vale ibex

[dusky shore]

GitHub

fix: check for the modteam role before reassigning the user's moderation role

[static canyon]

(I asked for it to be a log.warning since it means we'd need to remove them via int e)

I suppose sending to mod-meta would work the same though (and arguably makes more sense)

[vale ibex]

imo, it shouldn't be a log, they should just be removed from the cache entirely

a message to mod-meta would make more sense over sentry, since this is mod domain, not core-dev

[static canyon]

I wasn't sure whether that's something we wanted to do

[static canyon]

Yeah, I just realised that

[static canyon]

If this is an option then +1

[vale ibex]

Yea, I can't think of a reason why not to

[static canyon]

Maybe just an info log saying they were removed and why

I think that could be helpful

[vale ibex]

👍

Added a comment

[austere hornet]

Yeah that's basically what I was proposing, a simple time converter from one timezone to another

[hoary haven]

[clever wraith]

dammit i missed my merge

anyways https://learn-python.com is safe to use now

[fallen patrol]

uhhh is lance broken

playing status, now I wonder if it was just a fluke

"playing since 60 minutes ago"

could someone on devops redeploy lance

[gritty wind]

If probably doesn’t have a reset after the last day of aoc

.int e 0;exit()

[fallen patrol]

smh you could have used king arthur

[gritty wind]

I don’t know the commands 😭

[fallen patrol]

arthur deployments redeploy sir-lancebot

[vocal prairie]

restart, not redeploy iirc

[gritty wind]

I think it’s aliased to both

[fallen patrol]

I've seen Chris use redeploy for restart

[vocal prairie]

oh interesting

[fallen patrol]

I don't think there's any difference really

[vocal prairie]

arthur help deploy restart

sad

[fallen patrol]

you have to have the devops role

[vocal prairie]

yeah

[fallen patrol]

speaking of which @gritty wind congrats on devops!
I don't remember seeing it last I stalked viewed your profile

[gritty wind]

Slash commands when

Ah ty ty

[fallen patrol]

when you guys either move to a fork or implement them yourself. there's actually methods on the http object in dpy for slash commands, soooooo you don't actually need a fork.

[gritty wind]

Slash commands are actually really easy to write without a library

[patent pivot]

I can't run Arthur commands from my Apple Pencil

[gritty wind]

If we’re not doing gateway stuff

[fallen patrol]

now I want to take the challenge

[patent pivot]

We'll do slash commands once they have good permissions

[fallen patrol]

everything exists in dpy 2.0 to have slash commands

[gritty wind]

It’s not a challenge, it’s doable in 200 lines with a decent http library

[patent pivot]

KA will get slash commands fairly early

[gritty wind]

Dude

Voice activated Arthur commands when

AI too so I don’t have to remember the exact thing

When can I just shout restart the mainframe and have it do something

[patent pivot]

TRUE

[fallen patrol]

last you said it was when you moved to a fork I thought

[patent pivot]

No

It's once we move to a fork and can hide commands

[fallen patrol]

that doesn't make the most sense to me

you can make some commands slash commands that never are intended to be hidden

before there's discord support to hide them

like /docs /pypi /tags /topic etc

[patent pivot]

All king Arthur commands are hidden

I'm talking specifically KA here

[fallen patrol]

I wasn't talking about ka specifically, but all the bots

ah

[patent pivot]

Lance might move before that

¯_(ツ)_/¯

[fallen patrol]

yeah I haven't added any commands that require permissions to slash commands because discord

[patent pivot]

Apple Pencil is great I can write ¯_(ツ)/¯ and it writes ¯_(ツ)/¯

[dry folio]

joe uses 3 shrugs in a row-

[patent pivot]

¯_(ツ)_/¯

[austere hornet]

Lmao

[fallen patrol]

imagine not shrugging

[fallen patrol]

!e 5

[stable mountain]

@fallen patrol :warning: Your eval job has completed with return code 0.

[No output]

[fallen patrol]

@patent pivot is this about to switch from @gritty wind's server back to pydis?

[patent pivot]

yes

[fallen patrol]

noice

[patent pivot]

once i like

switch it

lol

it hasn't deployed yet

[fallen patrol]

ah

wait it failed? #dev-log message

oh

lol

I see you already fixed that: https://github.com/python-discord/snekbox/commit/94627e22c118060bdec1676f9dca16691617fa7d

is the self hosted runner ran by pydis?

[patent pivot]

yeah

[fallen patrol]

ah

[patent pivot]

it's on a deb 11 box we span up just for this

it means that all external contribs will need approval to run snekbox CI for all commits, but that isn't a major problem

[fallen patrol]

yeah

what happened here? https://github.com/python-discord/snekbox/runs/4637369027?check_suite_focus=true#step:7:1125

something went wrong somewhere: shouldn't really affect the outcome but

[patent pivot]

uhh I don't see it

[fallen patrol]

oh, wait until the run finishes then 💀

github streaming logs doesn't show them retroactively until the run is done

[patent pivot]

yeah makes sense

[fallen patrol]

#19 3.419 WARNING: Ignoring invalid distribution -istlib (/root/.local/lib/python3.10/site-packages)

sometimes happens from a pip install

obviously meant distlib

[patent pivot]

oh yeah this is fine

it's just a warning

[fallen patrol]

but how'd it do that in the first place

[patent pivot]

lol

¯_(ツ)_/¯

arthur deploy restart bot

[radiant merlin]

:white_check_mark: Restarted deployment bot in namespace default.

[patent pivot]

!eval import platform; print(platform.uname())

[stable mountain]

@patent pivot :white_check_mark: Your eval job has completed with return code 0.

uname_result(system='Linux', node='snekbox', release='5.10.0-9-cloud-amd64', version='#1 SMP Debian 5.10.70-1 (2021-09-30)', machine='x86_64')

[patent pivot]

okay we're back on pydis infra

[vale ibex]

In cluster?

Just a plain debian box?

[patent pivot]

in cluster

we figured that the stuff with cgroups wasn't exactly a bug

[vale ibex]

Nice nice

Nah, just a happy accident

[patent pivot]

it was just a thing we had to handle

[vale ibex]

I'm guessing we'll go back in April?

[patent pivot]

what happens in april

[fallen patrol]

I mean

April is after March and before may soooo

you kinda have to be prepared 🤡

[vale ibex]

I remember something happening in april

[tawdry vapor]

Only thing that comes to mind is new Ubuntu LTS coming out

[vale ibex]

Ahhh yea

[tawdry vapor]

But that's only relevant for CI

[fallen patrol]

thinking of discord message content intent?

[vale ibex]

No, what mark was saying

[fallen patrol]

ah

[tawdry vapor]

We might go back to a debian box if we can get gvisor working on it

[gritty wind]

I wonder if I can take down my domain from mobile

Bleh I’ll just leave it up

[fallen patrol]

👀

[patent pivot]

kubernetes 1.24 is released in april

[fallen patrol]

pay no attention to my IP ty

[tawdry vapor]

What's that update going to bring?

[gritty wind]

I can still IP ban you yk

[patent pivot]

hmmmm, full deprecation of dockershim is targeting 1.24 I think

[fallen patrol]

yes, yes you can

so please don't ty

[patent pivot]

https://github.com/kubernetes/enhancements/tree/master/keps/sig-node/2221-remove-dockershim

KEP 2221 schedules removal of Docker support in Kubernetes for 1.24 now

it was gonna be 1.23

but got pushed forward, so April 2022

[fallen patrol]

how does that affect snekbox and or pydis?

[patent pivot]

not much

scenario 1 is that Linode continue using an externally managed version of Dockershim and LKE clusters continue using Docker

[fallen patrol]

how does pydis use dockershim*

[patent pivot]

scenario 2 is that Linode pick a better container runtime like containerd or CRI-O, then I have to go through our manifests and change alpine to docker.io/alpine, lol, and that's about it

[patent pivot]

container runtime

kubernetes doesn't work 100% with docker

so the kubernetes team wrote a tool called "Dockershim" which makes Kubernetes compatible with Docker runtime

but it's a big maintenance burden

and there are better runtimes approaching like containerd and CRI-O

that are built specifically for use in Kubernetes

whether we are on or off Kubernetes, we are looking at migrating away from Docker

there are much nicer container runtimes nowadays, if we're off Kubernetes there is https://podman.io/ which is absolutely fantastic

If we're on Kubernetes, probably containerd or CRI-O, which are good runtimes for kube

there is a nice FAQ about Dockershim removal at https://kubernetes.io/blog/2020/12/02/dockershim-faq/ and a good blog post at https://kubernetes.io/blog/2020/12/02/dont-panic-kubernetes-and-docker/

me and Volcyy are writing a paper on pydis devops

mayyyyyybe public we'll see

[fallen patrol]

so docker-compose for running the projects wouldn't exist anymore?

[patent pivot]

It would

[fallen patrol]

so just what you use internally then

[patent pivot]

We're talking production here

[fallen patrol]

gotcha

[patent pivot]

Podman supports compose so we might advise you use that

[clever wraith]

My IP is 2a0f:85c1:0029:4269

Actually it isn't

[patent pivot]

But podman has full Docker compatibility

[clever wraith]

But it will be!

Hasn't k8s already dropped Docker runtime support?

[patent pivot]

You can alias Docker to it and it works

[patent pivot]

April 2022

Kubernetes 1.24

[clever wraith]

Huh, I thought linode already switched to another runtime

[patent pivot]

However, Docker will be maintained by a third Party

So Linode may stay with Docker

[gritty wind]

Come to do, we’re already running containered 😔

[patent pivot]

They have not confirmed anything yet

[gritty wind]

Do you actually need to change to docker repo in the image?

I think I’m running fine from just the pydis deployments, at least for snekbox lol

[patent pivot]

Podman advises you do, containerd might figure it out

Podman can support it

[gritty wind]

Good

To know

[patent pivot]

Podman can also run kubernetes pod specifications which is so cool

https://docs.podman.io/en/latest/markdown/podman-play-kube.1.html

it also runs as non-root

which is cool

[vocal prairie]

Do you like that more than Docker?

[patent pivot]

https://www.redhat.com/sysadmin/podman-docker-compose

[patent pivot]

personally yeah

it's a drop in replacement that's getting a lot of community traction

[vocal prairie]

I've been hearing a lot about it, maybe I'll try it out

[patent pivot]

this is what I wrote about it

daemonless is pretty nice

[fallen patrol]

if it's drop in, why use it in place of docker?
why does it exist if docker exists?
if its drop in, what doesn't actually work that does work with docker?

[clever wraith]

So would you yeet k8s for podman?

Or k8s will rely on podman as it's engine

[patent pivot]

why use it in place of docker? docker makes a lot of decisions which are... decisions. it's just an alternate container runtime option, more details at https://podman.io/whatis.html. some of the great things about podman include that you can run things rootless containers and so on.

why does it exist? alternatives have to exist, it's what makes OSS software great. it exists because people didn't like how docker do specific things.

what doesn't work that works with docker? it's mostly 1-to-1, both docker and docker-compose will work with Podman socket. there are probably intricacies that are different but vast majority is 1-to-1.

it's important to remember that docker is a fairly thin layer, the majority of docker is in the build process and Dockerfiles and so on, the runtime just uses a lot of built-in Linux features

[patent pivot]

it's more if we removed k8s, we'd use podman, it's not the reason we'd drop k8s

[clever wraith]

I see

[fallen patrol]

so

[patent pivot]

they explicitly don't support this, they recommend CRI-O for Kubernetes, but Podman will generate YAML for Kubernetes and can intepret it, it's just not meant to be a CRI tool

[fallen patrol]

if people didn't like how docker does specific things but podman is 1:1, isn't that counteractive?

[clever wraith]

Interesting

[patent pivot]

no, it's about how docker does things, not how docker is used

it's under the hood changes

the API is the same because Docker is so widely used that it's almost standard

Dockerfiles, all in all, are pretty great

but for runtime, Docker uses a daemon, Podman is daemonless, which is a huge difference

[fallen patrol]

er, what does that mean? 😅

[clever wraith]

Daemonless sounds very interesting

[clever wraith]

You don't need to have a master process running all the time, dockerd in the case of docker

Which allows you to more easily deploy on demand and stuff like that

[fallen patrol]

but how would it work without one?

[patent pivot]

yeah, pretty much that

[fallen patrol]

more importantly, how does it work with one?

also should we go to #tools-and-devops at this point lol

[patent pivot]

well, whenever you run a docker command, it is basically just an API call to the docker daemon

then that daemon builds/starts whatever you are working with

[clever wraith]

Through /var/docker.socket if I remember the path correctly

[patent pivot]

yeah something like that

podman still supports that, which is how docker-compose support works

but when you do podman run or whatever to start a container, instead of calling out to some API it interacts directly with runc

which is the tool for spawning OCI compliant containers https://github.com/opencontainers/runc

[clever wraith]

How is bookkeeping done?

Like what container maps to what PID

[patent pivot]

hahaha a very good question

[clever wraith]

I'm assuming some form of file storage?

[patent pivot]

mayyyybe, or maybe it can pull it back out of runc

[clever wraith]

Wouldn't that mean runc would run on a dameon then?

I guess it could list namespaces and all from the kernel

But I doubt every information can be taken back from the OS

[patent pivot]

don't think runc has a daemon nope

[clever wraith]

Yeah I'm pretty sure it doesn't

~/.local/share/containers/storage it seems like?

Or is that just the images

[patent pivot]

think that's just the storage

[clever wraith]

Someone just do strace -eopen podman run ... haha

[fallen patrol]

maybe when you start a container it starts a daemon or something?

[clever wraith]

Strace best friend for life

[fallen patrol]

or writes it to a file

[patent pivot]

it doesn't do daemons no, it just uses the containerisation tools that are built into the linux kernel

you'd be surprised how much of docker and containers are just linux features

[fallen patrol]

lol ;-;

[patent pivot]

https://jvns.ca/blog/2016/10/10/what-even-is-a-container/ a great blog post

Basically there are a few new Linux kernel features (“namespaces” and “cgroups”) that let you isolate processes from each other. When you use those features, you call it “containers”.

[fallen patrol]

enter docker desktop for windows

[patent pivot]

which spawns a linux vm, lol

[clever wraith]

Is that the "containers aren't a thing" blog post

[patent pivot]

uhhh maybe

but yeah more or less it just breaks containers into "containers are fancy processes"

which actually cleared things up a lot for me

[clever wraith]

Namespaces, cgroups and overlayfs basically makes up Docker

Or runc?

[fallen patrol]

!remind 40M

[stable mountain]

Aye aye, cap'n!

Your reminder will arrive on <t:1640552840:F>!

ID: 3823

[patent pivot]

uhhh i thought moby was the bit that docker used

unsure

oh maybe it is runc actually yeah

[clever wraith]

https://github.com/p8952/bocker/

That one is a great read

[clever wraith]

To be fair, it is a mess haha

[patent pivot]

hahahhaa yeah

[clever wraith]

You have CRIO, Moby, runc, docker, docker-compose

[patent pivot]

it is all part of why I like podman though, it makes things a bit clearer

[clever wraith]

It is such a confusing stack

[patent pivot]

podman interfaces directly with runc, and that's it

it uses a cool tool called Buildah when you build Dockerfiles (well, you rename them to Containerfiles, lol) https://buildah.io/

[gritty wind]

You know docker runs non-root :x

It’s in the troubleshooting guide of all places

[clever wraith]

Does it now

[patent pivot]

wtf fun ruined

[clever wraith]

Pretty sure it needs sysadmin capabilities for all the cgroup actions it has to do

[gritty wind]

Wait what do you mean by root

[patent pivot]

root user

[gritty wind]

Like, root user to start the daemon or run commands

[clever wraith]

Well a sysadmin user

Who is usually root

[gritty wind]

Because if it’s the latter, that’s what I’m referring to

[patent pivot]

run the daemon

[clever wraith]

To start the daemon

Yeah of course if you set your socket perms properly you can use the client without root

[gritty wind]

Right, nvm then

[clever wraith]

But I heard from security consultants that you shouldn't

[patent pivot]

oh cool, my podman isn't using runc

it uses crun, hahahaha, what great naming

crun works with cgroups v2

[tawdry vapor]

wtf https://github.com/python-discord/snekbox/runs/4637679833?check_suite_focus=true

[clever wraith]

Hold on lmao

Was it not confusing enough for ya lmao

[patent pivot]

i had this earlier when deploying that fix where self-hosted was deploying

the hosted runner took like 4 mins to boot because it kept failing to pull the action

good to see our self-hosted runner going strong though lol

[tawdry vapor]

boom another reason to drop k8s 😎

[patent pivot]

lmfao

hmmm no status reports on github

[tawdry vapor]

I tried re-running and it's still failing on it

I guess I'll just try again in an hour or something

[clever wraith]

Do we self host now?

[patent pivot]

lol https://twitter.com/jesseproudman/status/1475170998198554632

Jesse Proudman (@jesseproudman)

Any one else seeing extraordinarily slow checkout speeds on @github today? This is from a Github Actions worker...

Twitter · Dec 26, 2021, 6:25 PM

[gritty wind]

Nice

[patent pivot]

looks like there might be some generic networking problems on the hosted runners today, 13 KiB/s pull speed lmfao

[patent pivot]

for snekbox we do, so that we get a cgroups v1 and v2 install to test on

[gritty wind]

Genuinely faster than my cellular rn 🥲

[clever wraith]

cool, that's smart

[fallen patrol]

yeah, I had a 5+ minute checkout earlier

[clever wraith]

@patent pivot is the runner set up on k8s and if yes are the manifest available somewhere?

[patent pivot]

it isn't nah

it's just on a debian 11 nanode we've spun up

install was fairly easy, just set up the runner & docker on that box

and it's good to go

[clever wraith]

Cool

Sorry, saying we instead of you out of habit

Would be nice to get more reviews on sir-lancebot#991

[dusky shore]

GitHub

Give AoC completer role to members with 50 stars

[fallen patrol]

I don't reaaaally think that's a problem, given that you are still part of pydis (being a member) and a contributor

[obsidian patio]

@static canyon heyo, just so you know, I’ve addressed your requested changes

[fallen patrol]

this about the twemoji command?

[static canyon]

Yep, I saw. I'm busy until new yearish probably; will try to review when I'm free

[fallen patrol]

what issue number was that again (I can't remember if my review comment got saved or not)

github was bugging out and didn't save a few comments I made on different issues so I'm not sure

[static canyon]

It's easy enough to find if you just go to the repo

Github.com/python-discord/sir-lancebot/pulls

It'll be basically at the top

[fallen patrol]

I'm on mobile smh

sir-lancebot#988 def didn't scroll back to find it

[dusky shore]

GitHub

Add Twemoji utility command

[fallen patrol]

ah yeah & I couldn't check out the branch so I need to debug that

I really ended up with a detached head and a load of other goodies

[static canyon]

It's still right on top though lol

[fallen patrol]

¯\_(ツ)_/¯

[fallen patrol]

!remind 90m

[stable mountain]

Yep.

Your reminder will arrive on <t:1877201974:F>!

ID: 3825

[elder granite]

Hey thanks for sharing that. I've never worked with containers and this article shed some light on what are they at their core

[static canyon]

Edit duration

[fallen patrol]

!remind edit duration 3825 90M

[stable mountain]

Of course!

That reminder has been edited successfully!

ID: 3825

[patent pivot]

💯! It's a great article, containers are nifty tech and understanding them for real makes them even more magic.

[obsidian patio]

Yep, no stress 😄

[fallen patrol]

(also if someone has some ideas as to why I kept getting invalid reference from the gh cli client, I'd love that)

[austere hornet]

I'm sorry if I'm annoying anyone by this, but sir-lancebot#901 has been up for a long time now and needs one more core dev review to get merged. Thanks!

[dusky shore]

GitHub

Madlibs

[stable mountain]

It has arrived!

Here's your reminder: https://jvns.ca/blog/2016/10/10/what-even-is-a-container/ a great blog post
[Jump back to when you created the reminder](#dev-contrib message)

[fallen patrol]

that was a fast 40 minutes.

[elder granite]

The article is small and reads in less than 5 minutes hah

[austere hornet]

Haha

[elder granite]

And Joe how are you a first year student ? Did you start to code at 2 yo ?

[patent pivot]

8 years old, on a raspberry pi :)

So 10 years now

[elder granite]

impressive 🎩

[stable mountain]

It has arrived!

Here's your reminder: I really ended up with a detached head and a load of other goodies
[Jump back to when you created the reminder](#dev-contrib message)

[clever wraith]

I really ended up with a detached head

take that out of context

easy win

[austere hornet]

Lmao

[trail pilot]

i can review rn

Ahh I have some very small grammar nitpicks in the docstring for the AOC completer role but I don't know if I should post them given that they're the smallest nitpicks, and are probably unnecessary

[clever wraith]

haha, you can always approve and comment with a suggestion

as chris will squash the PR down anyway

[trail pilot]

Ahh

[tawdry vapor]

Why isn't my workflow executing in this PR https://github.com/python-discord/snekbox/pull/133

Do PRs only execute the workflows in the target branch, even if they've been changed in the source branch?

Oh never mind, it just errored out but the UI wasn't making it clear.

[fallen patrol]

github failing to load for anyone else?

[fiery kernel]

Hello

[elder granite]

works for me

[austere hornet]

nope it's fine for me, maybe it's your wifi or something

[clever wraith]

Whoa writing

[tawdry vapor]

Can anyone think of a workaround for this? https://github.com/actions/runner/discussions/1419

Having to create a new commit to update the refs sucks.

[clever wraith]

That's an interesting problem. If no interpolation is done I can hardly think of a workaround.

pushing to a specific branch could be a quite janky solution

probably not worth doing at that point

[fallen patrol]

workaround for which thing?

the whole issue?

[tawdry vapor]

Workaround for lack of interpolation support for the ref. I was looking at the JS actions to see if there's any way to execute a workflow from a YAML or something of that nature.

The only workaround for the whole issue that I know of is to use composite actions, but those suck because 1. it requires each action to be in a separate directory 2. actions cannot contain multiple jobs

[gritty wind]

There’s something planned that’s somewhere between composite actions and normal actions (aka composite actions with jobs and what not), but that won’t happen for a while. Even then, it’s only announced for enterprise, but I imagine it’ll happen eventually

For now we just live with the duplication and dedup where we can (HassanAbouelela/actions was a start to that)

[tawdry vapor]

Makes me miss azure pipelines

[fallen patrol]

what's HassanAbouelela? (joke, obviously)

[gritty wind]

My GH

[fallen patrol]

and I wonder why I overuse the clown emoji 😩

[fallen patrol]

help.

it works for every pull

except for this one

sir-lancebot#988

[dusky shore]

GitHub

Add Twemoji utility command

[fallen patrol]

wait it happens on other pulls now

something with git config perhaps?

yeah idfk what i did here

[cold island]

Can you checkout the usual git way?

[fallen patrol]

what is the usual git way?

[vale ibex]

git switch -c twemoji upstream/twemoji

Might need to git fetch upstream first

[cold island]

upstream is usually inferred, so just git pull && git checkout twemoji should work

[fallen patrol]

!remind 7H review twemoji pr plz thx. and to actually stay focused, have you thought about a todo list?

[stable mountain]

Okay.

Your reminder will arrive on <t:1640716396:F>!

ID: 3832

[fallen patrol]

what do you use for managing your dev todo? 🤡

[green oriole]

it is inferred if the branch name is only present on one remote

[cold island]

Aren't you supposed to not be awake rn? 🤔

[fallen patrol]

also fwiw I did pull from upstreamz even went as far to nuke the folder and reclone

[fallen patrol]

.. maybe

what do you all use for managing what you need to do or something? like, what app or tool?

specifically talking like: fix up dev commit here, review that pull, finish an issue, etc

[gritty wind]

Jetbrains VCS integration >.>

[green oriole]

did lance just die lol

[gritty wind]

It probably oomed on something, could look into it later

[green oriole]

I was about to say, a devops should be able to pull that haha

then checked roles

[gritty wind]

I don’t really have PC access this month :D

[green oriole]

Kubenav is a great app haha

[austere hornet]

I don't really use any, but I know a few. One is Workflowy. Another is Google Tasks. Another is Outlook reminders maybe if you have that. Those are the only ones I can think of at the moment but give them a try.

[vale ibex]

I just use github notifications

[dim pelican]

OneNote for work, and for developing I just work slow enough to have 1 thing at a time

[austere hornet]

Good idea haha

And yeah OneNote also works

[vocal wolf]

github notifications

@brisk brook @last patio https://github.com/python-discord/site/pull/508 merged, thanks for the reviews!

@eternal owl https://github.com/python-discord/bot/pull/1568 unstalled.

[brisk brook]

Awesome!

[gritty wind]

Lol just downloaded it a couple days ago

But internet is way too slow to reliably get anything

[eternal owl]

👍

[vocal wolf]

[green oriole]

lol

[vale ibex]

Forcepush to the AoC branch was just a squash

[green oriole]

We all saw the token grabber you tried to push to that branch

[vale ibex]

It just emails me your dotenv on boot

[green oriole]

I mean if you want it I can mail it to you

Like literally mail it

But you pay the postage

Hey @vocal wolf, feels like reviewing? https://github.com/python-discord/sir-lancebot/pull/991

Poking the poking lead, we've come full circle

[vocal wolf]

jokes on you I already poked myself heheheheheh

[green oriole]

nooooooo

[fallen patrol]

how is the message info calculated

its off by.... 20k messages

unless it doesn't include commands?

bot#2016

[dusky shore]

GitHub

Include message counts in all channels

[stable mountain]

It has arrived!

Here's your reminder: review twemoji pr plz thx. and to actually stay focused, have you thought about a todo list?
[Jump back to when you created the reminder](#dev-contrib message)

[fallen patrol]

i timed that perfectly then

[fallen patrol]

what the fuck

[summer garden]

Do you guys fix style errors and write docstrings as you go, or do you do it all at the end?

[fallen patrol]

all commands:

[fallen patrol]

both, personally

[fallen patrol]

like, I tend to do docstrings at the end, right before committing, but linting all the way through with a formatter

[summer garden]

I despise writing docstrings and commit messages. It's almost more work than the coding itself

[dim pelican]

Conversely; it is the most important stuff so that people like me can understand what a function or a commit does

[fallen patrol]

so that people can understand
FTFY

basically its not just you, its helpful for everyone

[summer garden]

HAHA you know what? It's because I have to decipher what my code does and then explain it in a concise and eloquent way so nobody else gets confused

[vale ibex]

This

Message count is used for voice verification, so we ignore both bot command channels

[fallen patrol]

so a bot command in a different channel is counted then

[vale ibex]

Yes

[fallen patrol]

so I have 20k messages in #bot-commands and #sir-lancebot-playground

[cold island]

Counting starts August 25th 2020

[fallen patrol]

I was here after then

and I only have about 2-3k messages in bot command channels so....

[dim pelican]

I find the same thing happens with emails / presentations at work. Trying to summarize a complex thing in a few bullet points that gets the key info right is really tough to do

[fallen patrol]

oH yeah

is this Stat metricity

[cold island]

yes

[fallen patrol]

that's why it's wrong then

[cold island]

?

[fallen patrol]

I previously asked Joe to delete my metricity data 🤡

[cold island]

ah

[fallen patrol]

any thoughts on why this isn't working

[cold island]

Did the other approach work?

[gritty wind]

Git fetch?

[fallen patrol]

oof I missed it, will try shortly

[fallen patrol]

@gritty wind i ran that

[gritty wind]

Can you list remotes

[fallen patrol]

.

oh.

[fallen patrol]

yes, yes i can

[gritty wind]

Idk try checkout instead of switch /shrug

[fallen patrol]

this could be why....

the remote references aren't being saved, so its something with fetching

oh.

git config

remote.upstream.fetch=+refs/heads/main:refs/remotes/upstream/main

there. we. go.

[fallen patrol]

for reference change this to remote.upstream.fetch=+refs/heads/*:refs/remotes/upstream/*

apparently github does this when you use gh repo clone

[gritty wind]

Why are you even using it?

[fallen patrol]

using what

[gritty wind]

GH

[green oriole]

That activity block count seems off

Pretty sure I had >30k blocks

You can look at my messages in mod-spam back when we were testing the lag

[fallen patrol]

you use github smh

[vocal prairie]

I can't find anything like that

Seems like you're at 16k

[green oriole]

Really

Well, thanks for checking

[gritty wind]

The GH client not just GitHub lol

[fallen patrol]

why not use the gh cli client?

[gritty wind]

I’m genuinely asking why you use it

Like what benefits does it give you over regular git

[fallen patrol]

ah

fwiw I use it in combination with regular git

but it has api, browse, actions, gist, issue, pull, etc commands

so gh api /repos/python-discord/bot to make an api request to their,
gh browse to open the repo in the browser
gh gist to make gists
gh issue view num to view an issue
gh pr checkout num to checkout a pull, etc etc

[vocal prairie]

I use the gh client since I'm too lazy to set up ssh and the auth is easy

[fallen patrol]

yeah pretty much

[gritty wind]

Most of these things are pre-integrated into my IDE, so I’ve gotten kinda complacent about the whole thing

My IDE has the benefit of doing most of the heavy lifting while I give it vague instructions

[green oriole]

I use it to check out PRs because I'm a lazy bastard

That is aaaallll

This isn't a self attributed lazy bastard title btw

[fallen patrol]

lmfao @vocal wolf

[vocal wolf]

Yes, I joined a server

[fallen patrol]

its more like

all of you at once

but also

I didn't know that bots could join servers with invites 🤡

[gritty wind]

Our raid channel says otherwise

[fallen patrol]

yeah, saw it yesterday on stats.pydis

[tawdry vapor]

Looking for reviews on snekbox#132 and snekbox#136 - they're not too scary, I hope

[dusky shore]

GitHub

Store relative paths in coverage data
Fix passing arguments to python -m snekbox and support Python args

[fallen patrol]

💀 omg those are scary

[austere hornet]

Also sir-lancebot#901 needs a core dev review! Again, sorry if I'm being annoying about this but it's been up for a long time now. Thanks so much!

[dusky shore]

GitHub

Madlibs

[static canyon]

Is anyone able to give some insight as to what these test errors mean?```py
FAILED tests/bot/exts/moderation/test_incidents.py::TestArchive::test_archive_clyde_username - AttributeError: 'NoneType' object has no attribute 'kwargs'

FAILED tests/bot/exts/moderation/test_incidents.py::TestArchive::test_archive_relays_incident - AssertionError: Expected 'send' to be called once. Called 0 times.

FAILED tests/bot/exts/moderation/test_incidents.py::TestProcessEvent::test_process_event_confirmation_task_is_awaited - AssertionError: Expected mock to have been awaited.

[green oriole]

No line number?

Because without one it would be hard to exactly know what happened, at least with the first one

The first one is the code didn't call send, and the second one à certain task never got awaited

[static canyon]

It says the test

Gimme a few mins, uploading what I've got at the moment to GitHub

Right, okay

@green oriole ```py
______________________________________________________________________________________ TestArchive.test_archive_clyde_username ______________________________________________________________________________________

self = <tests.bot.exts.moderation.test_incidents.TestArchive testMethod=test_archive_clyde_username>

async def test_archive_clyde_username(self):
    """
    The archive webhook username is cleansed using `sub_clyde`.

    Discord will reject any webhook with "clyde" in the username field, as it impersonates
    the official Clyde bot. Since we do not control what the username will be (the incident
    author name is used), we must ensure the name is cleansed, otherwise the relay may fail.

    This test assumes the username is passed as a kwarg. If this test fails, please review
    whether the passed argument is being retrieved correctly.
    """
    webhook = MockAsyncWebhook()
    self.cog_instance.bot.fetch_webhook = AsyncMock(return_value=webhook)

    message_from_clyde = MockMessage(author=MockUser(name="clyde the great"))
    await self.cog_instance.archive(message_from_clyde, MagicMock(incidents.Signal), MockMember())

  self.assertNotIn("clyde", webhook.send.call_args.kwargs["username"])

E AttributeError: 'NoneType' object has no attribute 'kwargs'

tests\bot\exts\moderation\test_incidents.py:439: AttributeError
============================================================================================== short test summary info ==============================================================================================
FAILED tests/bot/exts/moderation/test_incidents.py::TestArchive::test_archive_clyde_username - AttributeError: 'NoneType' object has no attribute 'kwargs'

It's the webhook.send.call_args.kwargs["username"] giving the attribute error

But I didn't change that and it worked before my commits

[green oriole]

Hmm

Try a poetry install?

[static canyon]

hmm

Let's see

[static canyon]

It updated two libraries:```c
(bot-6mLjCNeK-py3.9) C:\Users\tizzy\bot>poetry install
Installing dependencies from lock file

Package operations: 0 installs, 2 updates, 0 removals

• Updating pycares (4.0.0 -> 4.1.2)
• Updating rapidfuzz (1.7.1 -> 1.8.0)

Installing the current project: bot (1.0.0)

[green oriole]

Interesting

Is your branch up to date with main?

[static canyon]

Not the current main

It's like a monthish behind

[green oriole]

Hmm, could you try to merge? (fetch and merge origin/main)

[static canyon]

That worked, no conflicts

[green oriole]

What about the tests

[static canyon]

But test still fails

Exact same error

[green oriole]

How weird

What files did you change

[static canyon]

Only incidents.py

bot/exts/moderation/incidents.py iirc

[green oriole]

Wait, so your code probably broke it then haha

[static canyon]

But like, I don't get how

I've not changed any of the core functionality

[green oriole]

Can I get a diff?

[static canyon]

It's the exact same just the archive message has more info

[static canyon]

https://github.com/python-discord/bot/compare/incident-archive-msg-improvements

If you run tests you'll also get once saying the embed.description doesn't match but that's just because I forgot to remove the on in the test so that test can be ignored (have resolved locally)

[green oriole]

Okay

One of the problem is the mock user doesn't have display_name working properly, I'm pretty sure

[vale ibex]

- message_from_clyde = MockMessage(author=MockUser(name="clyde the great"))
+ message_from_clyde = MockMessage(author=MockUser(display_name="clyde the great"))

[green oriole]

Basically this

[static canyon]

Right, lemme try

Yep, that fixed it lol

Thanks

Now the other 2 tests

[green oriole]

All three?

Sadge

Thanks for the role Chris

If that was you

Oh BTW you shouldn't be using now in your test

[static canyon]

2nd test I'm looking at might be the same thing

[green oriole]

They should be deterministic

[static canyon]

Yep, it was

[static canyon]

Yeah, I'll fix that later

[austere hornet]

Yeah it was him sir-lancebot#991

[dusky shore]

GitHub

Give AoC completer role to members with 50 stars

[static canyon]

_________________________________________________________________________ TestProcessEvent.test_process_event_confirmation_task_is_awaited __________________________________________________________________________

self = <tests.bot.exts.moderation.test_incidents.TestProcessEvent testMethod=test_process_event_confirmation_task_is_awaited>

    async def test_process_event_confirmation_task_is_awaited(self):
        """Task given by `Incidents.make_confirmation_task` is awaited before method exits."""
        mock_task = AsyncMock()

        with patch("bot.exts.moderation.incidents.Incidents.make_confirmation_task", mock_task):
            await self.cog_instance.process_event(
                reaction=incidents.Signal.ACTIONED.value,
                incident=MockMessage(id=123),
                member=MockMember(roles=[MockRole(id=1)])
            )

>       mock_task.assert_awaited()

tests\bot\exts\moderation\test_incidents.py:543:
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _

self = <AsyncMock id='1744593019040'>

    def assert_awaited(self):
        """
        Assert that the mock was awaited at least once.
        """
        if self.await_count == 0:
            msg = f"Expected {self._mock_name or 'mock'} to have been awaited."
>           raise AssertionError(msg)
E           AssertionError: Expected mock to have been awaited.

..\AppData\Local\Programs\Python\Python39\lib\unittest\mock.py:2188: AssertionError
========================================================================================= short test summary info =========================================================================================
FAILED tests/bot/exts/moderation/test_incidents.py::TestProcessEvent::test_process_event_confirmation_task_is_awaited - AssertionError: Expected mock to have been awaited.

Last one @green oriole

[red grotto]

that looks weird

[static canyon]

name isn't passed here in the MockMember but do I need to add display_name still?

[static canyon]

Doesn't seem to fix it so guessing not

[static canyon]

Yeah, I don't really know how to fix this one

[green oriole]

I'm guessing on a traceback

Currently playing domino with my grandma so gl

[static canyon]

What do you mean?

[vale ibex]

Is there a traceback further up in the output?

[static canyon]

Nope

[static canyon]

This is the enire traceback @vale ibex

[vale ibex]

can you do poetry run task retest and put the full output into a paste

[static canyon]

yep

[static canyon]

https://paste.pythondiscord.com/obaviciwoc.py

It's the same thing as what I sent above

[vale ibex]

alright, lemme pull you changes and take a look

[static canyon]

There's no PR at the moment but this is the diff

[vale ibex]

Yea, I'm just gonna pull the branch

[static canyon]

👍

[vale ibex]

What was your diff to fix test_make_embed_content

[static canyon]

Just remove the on

[vale ibex]

👍

[static canyon]

        self.assertEqual(
            f"{incident.content}\n\n__*Reported {day_timestamp} at {time_timestamp} ({relative_timestamp}).*__",
            embed.description
        )
```works

[vale ibex]

Alright, the problem is the same

it's because display_name isn't set

MockMember by default has it's name attribute set, but not display_name

see tests.helpers.MockMember

[static canyon]

I tried specifying display_name but it didn't fix it

[vale ibex]

it's because Incidents.archive uses incident.author.display_name

which comes from the message that get's passed

not the member

[static canyon]

Ah right

[vale ibex]

MockMessage uses MockMember for the author

which only has member