dev-contrib

[static canyon]

Yes

[outer oasis]

Did we ever fix that one?

[static canyon]

Nope

Nothing we can do afaik

sir-lancebot#1087 (for reference)

[dusky shore]

GitHub

[sir-lancebot] #1087 Use buttons for allowing others to receive a bookmark

[outer oasis]

@static canyon , any thoughts?
Poetry 1.2.0 was released 2 days ago, but I'm not seeing anything in https://python-poetry.org/blog/announcing-poetry-1.2.0/ that I would think could cause this

[static canyon]

I can't really look because I'm on mobile, sorry

Wasn't there an issue/PR for pinning the poetry version? Maybe it's related to that?

[outer oasis]

Oh

Yes I just saw this

Hmm
Only pins the version in the Dockerfile

Hi @vale ibex
Please explain bot#2268
Thanks

[dusky shore]

GitHub

[bot] #2268 Use venvs with poetry in Dockerfile

[outer oasis]

Our bird just opened a new PR and and the workflow failed because Poetry said the arguments passed to it were invalid.
But looks like you're now moving Poetry out of the main body and into the Dockerfile and pinning to the previous version?

Do we just need to wait and rerun the CI after your PR is merged?

[karmic silo]

have you guys considered adding some web frameworks to the eval command (just for their utilities and such), i see people asking about them a lot

[outer oasis]

What kind of frameworks are you talking about?

Something like flask?
I don't know if that would be too helpful, because eval can't serve multiple requests

What could you do in a single eval'd code block?

[karmic silo]

again, utilities and such that don’t actually run a web app

im not sure if they work at all without the app though so it might be pointless

[outer oasis]

It does have bs4 already

If that's the kinda thing you're looking for

[cold island]

So I just had a big brain idea

Why not make #voice-verification use buttons

It can then respond with an ephemeral message

No one can mix up the command

And the bot doesn't need to auto-delete messages

[atomic ivy]

nice idea.

[honest lintel]

it's kinda weird to see general have 0 activity. maybe it's because of the time in the most common time zones in this server? But I doubt that.

[atomic ivy]

it does ebb and flow.
https://stats.pydis.com has some stats on that.

[vale ibex]

Too much to type on mobile, but the pr linked in that pr should describe it

Basically Yea, once that has merged, it will fix your ci

[outer oasis]

👍

What're you waiting on?

Just reviews?

[vale ibex]

Reviews

[fossil veldt]

Wait I'm confused, why did we need the user pip installs anyways instead of venv? Was it a caching issue?

[vale ibex]

it was more that we didn't need a venv, since it's a one off image

[fossil veldt]

Also if bot#2268 is switching to venvs shouldn't that fix the issue related to poetry 1.2? But it's also pinning it back to 1.1.15?

[dusky shore]

GitHub

[bot] #2268 Use venvs with poetry in Dockerfile

[vale ibex]

CI is pinning to 1.1.15, so I'm leaving the dockerfile be the same until that is updated

[fossil veldt]

Ah okay, but theoretically it would work for 1.2 now?

[vale ibex]

Yea

[fossil veldt]

So currently lint-test.yml has also been changed to use poetry 1.1.* via HassanAbouelela/actions/setup-python@setup-python_v1.2.1?

[vale ibex]

Yea, that pins it

[fossil veldt]

can we disable the lancebot fuzzy matcher for pygen? Some people were spamming it just now, it's bound to happen again

not that it's always malicious but when one person accidentally invokes it everyone goes 'hey look broken thing' and everyone tries it

related to sir-lancebot#1090

[dusky shore]

GitHub

[sir-lancebot] #1090 Incorrect handling of invalid command names

[vale ibex]

If the fix is too involved, we can disable the misbehaving code path in the meantime, sure

[fossil veldt]

could I look into a fix for 1090? Since it's not currently approved

[gritty wind]

The reason it’s not approved was because it’s a bug without a clear resolution pass, that’s probably going to require discussion to figure something out

Following the philosophy of organization, I thought it shouldn’t be approved till we know what we actually want

No one’s looked into it because people are busy, and it’s not a critical thing

[static canyon]

I mean the solution is pretty easy. Just don't call self.on_command_error.

Imo we should also move the util to botcore since both bot and lance use it, but ig that's not necessary

[olive cloak]

am i understanding correctly that an issue should be opened before PR'ing?

[cold island]

That's the best way to go yes

Gives us a chance to look into the proposed change before you put work into changing the code

[olive cloak]

thanks

[fading galleon]

and also mention the issue in the pr

[gritty wind]

It was proposed before, but it’s not something we want to do in terms of implementation or behavior. It’s hard to implement because the onus is on snekbox to figure out when input is required, communicate back with the bot and the user, and manage the resource and time limits, or parse user code (which is also not easy). In terms of behavior, it’s not something we consider important. Why do you need input instead of just putting in whatever you actually want to run in your program

(If you really need it, you can write directly to stdin)

[karmic silo]

i already suggested this here https://github.com/python-discord/bot/issues/2255 and it seems it’s really just not worth the work

[celest charm]

these didyoumeans are really annoying!

is there a way to disable them if the message starts with ..?

[static canyon]

sir-lancebot#1090

[dusky shore]

GitHub

[sir-lancebot] #1090 Incorrect handling of invalid command names

[static canyon]

It's not gotten approved but I'm planning to fix it as soon as it does

Apparently it's not being considered critical atm though & the core devs are rather busy

[celest charm]

ah

[cold island]

Approval to fix the bug? sure, I think it's indisputable that this is a bug

[static canyon]

@cold island see Scale's messages here

[static canyon]

For me the only question is do we move the util to botcore?

[cold island]

• What util
• Why does it call self.on_command_error in the first place

And let's fix the issue on lance first, we can discuss moving it to botcore separately

[static canyon]

The util that tries to get similar commands

Part of that util triggers on_command_error, but it shouldn't

I described it all in a comment on the issue

[cold island]

How will you check whether it can be run instead?

[static canyon]

The exact same. We just pass instead of handling the error

[cold island]

So before 1064 it would raise an error for not being able to run the command, and then suggest it?

And the only thing that changed now is that it does it multiple times

[static canyon]

It doesn't suggest it

It just raises & sends the error to Discord

It doesn't tell you what the command was

[cold island]

Interesting that it didn't come up before, probably because staff/mods don't get these

[static canyon]

Because when looping the similar commands, there was a return which meant it'd only do it for the first matching command, and ig it seemed intentional at that point

Now there's a continue, so it does it for all matching commands (including subcommands iirc)

The error message also self-deletes after so many (I think 3) seconds so realistically only the invoker would see it

And if as you say staff/mods don't get them, then that's another thing

All contributors to it not being noticed

[cold island]

Anyway that looks like a good solution as far as I can see so I say go for it

[static canyon]

So changing await self.on_command_error(...) to pass?

[cold island]

Do you need to explicitly pass?

[static canyon]

If it'd otherwise be a blank except body then yes

[cold island]

Isn't there a log or something in there

[static canyon]

But just remembered there's a continue there, so no need for pass

[cold island]

Yeah that too

[static canyon]

(also a log, yeah)

So it's just a case of removing that one line, cool

[cold island]

Should be, we should test that doesn't mess things up

[static canyon]

Should I also fix the typehint that 1064 did? I.e. Optional[list] --> list (it's always provided, just is sometimes empty)

def get_command_suggestions(
        all_commands: list[str], query: str, *, cutoff: int = 60, limit: int = 3
) -> Optional[list]:
    """Get similar command names."""
    results = process.extract(query, all_commands, score_cutoff=cutoff, limit=limit)
    return [result[0] for result in results]```the return type for this

[cold island]

Yeah

Shouldn't be a bare list either

[static canyon]

Yep

Cool

Was... this not how you run tests?```
PS C:\Users\tizzy\sir-lancebot> poetry run task test
could not find task "test"
PS C:\Users\tizzy\sir-lancebot> poetry run task tests
could not find task "tests"

[vale ibex]

no tests in lance

[static canyon]

Right

Completely forgot lmao

[vale ibex]

hah no worries

CI does a "test"

it just checks it can import all the exts

[static canyon]

Also forgot I can't access my GH account right now because I don't have my standard phone number since I'm abroad

So I guess the PR will be tomorrow/day after

[vale ibex]

Sure

[static canyon]

Can someone please note on 1090 that it's been approved by @cold island and that I've implemented the fix

(sir-lancebot#1090)

[dusky shore]

GitHub

[sir-lancebot] #1090 Incorrect handling of invalid command names

[fading galleon]

Whats the status of bot#2263 I added a pr for it, when will it be merged?

[dusky shore]

GitHub

[bot] #2263 Remove italics from !help text

[clever wraith]

Hey folks,

Can I get some reviews on bot#2261 and site#770 please ?
I know wookie is supposed to review them, but they appear to be busy with other things so I thought maybe someone else could ?

[dusky shore]

GitHub

[bot] #2261 Add support to fetch rules via keywords
[site] #770 Add a collection of keywords per rule

[stable mountain]

It has arrived!

Here's your reminder: review if not already merged
[Jump back to when you created the reminder](#dev-contrib message)

[molten perch]

Hey! I have a very simple PR, that needs one more review, I'd appreciate if someone took a look at it.
sir-robin#60

[dusky shore]

GitHub

[sir-robin] #60 Bump dpy to v2, and update the project to use BotBase

[static canyon]

RE sir-lancebot#1062

For the part about adding interactivity (react with ❌ to delete etc.) how exactly should this be implemented?

I'm thinking the bot adds the reactions itself, and then a wait_for (potentially in a loop if there's multiple interaction options) will trigger when the user presses one. Maybe even buttons, but I imagine reactions are better here due to the size of buttons.

[dusky shore]

GitHub

[sir-lancebot] #1062 improving the .latex command

[cold island]

Isn't it just a regular ?

[static canyon]

.help

[static canyon]

Right, the delete specifically is

[cold island]

yeah, it uses wait_for iirc

[static canyon]

Cool

with contextlib.suppress(NotFound):
    try:
        _, new_message = await self.bot.wait_for(
            'message_edit',
            check=_message_edit_predicate,
            timeout=REDO_TIMEOUT
        )
        await ctx.message.add_reaction(REDO_EMOJI)

        await self.bot.wait_for(
            'reaction_add',
            check=_reaction_add_predicate,
            timeout=10
        )

        await ctx.message.clear_reaction(REDO_EMOJI)
        with contextlib.suppress(HTTPException):
            await response.delete()

    except asyncio.TimeoutError:
        # One of the `wait_for` timed out, so abort 
        await ctx.message.clear_reaction(REDO_EMOJI)
        return None

    else:
        # Both `wait_for` triggered, so return the new content to be re-run
        return new_message.content

return None```Where does the bottom `return None` ever get reached? I'm thinking it's only when `await ctx.message.clear_reaction(REDO_EMOJI)` fails due to the emoji not having been added (i.e. when the first `wait_for` fails)

[vale ibex]

it'll be whenever NotFound is raised

since it will exit the suppress block

[static canyon]

Right

[vale ibex]

so it could be for that clear reaction, or even for the add reaction

[static canyon]

So also add_reaction

Yeah

async def continue_job(ctx: Context, response: Message) -> Optional[str]:
    """
    Check if the job should be rerun.

    For a job to be rerun, the user must edit their message within `REDO_TIMEOUT` seconds,
    and then react with the `REDO_EMOJI` within 10 seconds.
    """
    # Correct message and content did actually change (e.g. wasn't a pin status udpate)
    _message_edit_predicate = lambda old, new: new.id == ctx.message.id and new.content != old.content

    _reaction_add_predicate = lambda reaction, user: all((
        user.id == ctx.author.id,  # correct user
        str(reaction) == REDO_EMOJI,  # correct emoji
        reaction.message.id == ctx.message.id  # correct message
    ))

    with contextlib.suppress(NotFound):
        try:
            _, new_message = await self.bot.wait_for(
                'message_edit',
                check=_message_edit_predicate,
                timeout=REDO_TIMEOUT
            )
            await ctx.message.add_reaction(REDO_EMOJI)

            await self.bot.wait_for(
                'reaction_add',
                check=_reaction_add_predicate,
                timeout=10
            )

            await ctx.message.clear_reaction(REDO_EMOJI)
            with contextlib.suppress(HTTPException):
                await response.delete()

        except asyncio.TimeoutError:
            # One of the `wait_for` timed out, so abort
            await ctx.message.clear_reaction(REDO_EMOJI)
            return None

        else:
            # Both `wait_for` triggered, so return the new content to be re-run
            return new_message.content

    return None  # triggered whenever a `NotFound` was raised

```does this look good?

I do think the function name should be changed, but not sure what to

Maybe rerun_job

Or check_rerun_job/maybe_rerun_job since it's not guaranteed to rerun and the function doesn't do the rerunning itself

[cold island]

Maybe you can take inspiration from the structure in eval

[static canyon]

Yeah, I have

This is copy-paste from eval, then adjusted

[cold island]

!src eval

[stable mountain]

Command: eval

Run Python code and get the results.

Source Code

Go to GitHub

bot/exts/utils/snekbox.py:483

[static canyon]

I've also added the comments for clarity

[cold island]

out of the options above, adding maybe is my preference

[static canyon]

I'm thinking check_rerun_job since the docstring is literally "Check if the job should be rerun."

And I feel like maybe_rerun_job sounds like the function will do the rerunning (which it doesn't)

[cold island]

ah

ok

[static canyon]

I think I'm just about ready to PR

Is this good for a bot-core docstring?```py
async def check_rerun_job(ctx: Context, response: Message) -> Optional[str]:
"""
Check if the job should be rerun.

For a job to be rerun, the user must edit their message within `REDO_TIMEOUT` seconds,
and then react with the `REDO_EMOJI` within 10 seconds.

Args:
    ctx: The command's context
    response: The job's response message

Returns:
     The content to be rerun, or `None`.
"""```

[cold island]

If this is bot-core then it should be as generic as possible

variable duration for editing, maybe also take a callable to process the response

Also you should try implementing it in eval to see if what you did covers this use case, since that's why you're writing it in bot-core

is this function the only thing you're adding to bot-core, or are there other parts as well?

[gritty wind]

Use double backticks

[cold island]

is that a sphinx thing?

[gritty wind]

Yes, single backticks do different formatting stuff than md

Double backticks are the equivalent

[static canyon]

Just this util

[fossil veldt]

just a typing suggestion but otherwise looks good

[molten perch]

Much appreciated! 😄

[fallen patrol]

additionally it shouldn't use global variables, unless botcore is basically on pypi for pydis use only

[gritty wind]

It’s not on pypi

[timid sentinel]

Would be good to get another opinion on https://github.com/python-discord/bot/pull/2272

[hoary haven]

something happened earlier and this help channel did not close
#help-dumpling message

for a second i even saw @stable mountain reply with an embed to the OP saying "did you mean !close?" or something like that

[fading galleon]

I think the previous use of the lock from the channel's id was still being holded by some await statement, possibly in claim_channel(/bot/bot/exts/help_channels/_cog.py) could be something else too like the lock from user's id being holded somewhere(hard to traceback) becuase when you tried to close it, it closed

[kindred sun]

Apologies if this is the wrong place to ask this. I've been referencing the Python bot's testing code in order to learn to setup testing in my own bot, and I noticed a tiny issue. In bot/tests/helpers.py line 320, additional_spec_asyncs features an attribute, redis_ready, that (seems to) no longer exist in Bot or BotBase. I believe it was removed in fc05849. I'm looking to make small contributions since I've never contributed to an open source project before, would this be something worth making an issue for?

[vocal prairie]

edit: nvm, i was confused what you were asking

[kindred sun]

😄

i'll try to be more clear

[vocal prairie]

nah you were fine, i just didn't read properly

[static canyon]

@timid sentinel re bot#2233 do you mean the command docstring?

[dusky shore]

GitHub

[bot] #2233 Allow referencing message as argument to !remind edit content

[timid sentinel]

yeah

[static canyon]

Cool, will do that soon™️

[static canyon]

!remind 28h if not done already

[stable mountain]

ROGER THAT

Your reminder will arrive on <t:1663001978:F>!

ID: 5153

[timid sentinel]

Good spot. Feel free to create an issue, it's also fine if you'd like to create a PR without an issue since it's a small change.

[kindred sun]

Thanks, will do

[clever wraith]

@timid sentinel would you mind having a look on bot#2261 again please?
Mark has suggested that we'd allow the rule numbers and keywords to be sent in any orders but also lookup based on multiple keywords and just one.
I thought that it wouldn't really matter after all so I changed it.
Thank you!

[dusky shore]

GitHub

[bot] #2261 Add support to fetch rules via keywords

[kindred sun]

Since I'll be poking around in the testing code for my own purposes, is there anything else there that I could help with? Didn't see anything in issues under the testing label

[gritty wind]

Tbh, the testing setup for bot is not great

Testing discord.py is already a terrible process, so most contributors struggle with the unittests

We don’t pay too much attention tests, we just keep them passing

What is there is fine, nothing wrong with it, but it probably shouldn’t be your reference

[kindred sun]

I see. I've been trying to find a way to add tests for my own bot projects and Mocks seem to be the most sane option. Never used them before so I was looking to see how other discord bot projects used them.

Discord bots make up the bulk of my "portfolio" projects I show to employers, so it feels wrong to not have any sort of testing.

[timid sentinel]

!charinfo !closе

[stable mountain]

Character Info

\u0021 : EXCLAMATION MARK - !
\u0063 : LATIN SMALL LETTER C - c
\u006c : LATIN SMALL LETTER L - l
\u006f : LATIN SMALL LETTER O - o
\u0073 : LATIN SMALL LETTER S - s
\u0435 : CYRILLIC SMALL LETTER IE - е

Full Raw Text

\u0021\u0063\u006c\u006f\u0073\u0435

[timid sentinel]

Wasted a bunch of time looking into that 😑

[timid sentinel]

Ah sorry I forgot about that, I'll try and take a look tomorrow

[kindred sun]

hilarious

[timid sentinel]

Was literally about to send a message here about how I had traced through the logs and code and was convinced what was happening was impossible and made no sense at all, then figured I should just double check...

Made the rookie mistake of entertaining the idea that there could be a bug in our code

[hoary haven]

lmaooo

very obvious now that i look at it now

surwei

so what was that log about notify_session_participants and something being locked?

[austere hornet]

it's not a bug, it's a feature

[fading galleon]

I should have also checked that earlier, spent a lot of time on my test server to reproduce it and understand the lock code

[cursive relic]

Where do the results of .http_status come from?

[atomic ivy]

.src http_status

[dusky shore]

Command: http_status

Choose a cat or dog randomly for the given status code.

Source Code

Go to GitHub

bot/exts/fun/status_codes.py:29

[atomic ivy]

https://github.com/python-discord/sir-lancebot/blob/main/bot/exts/fun/status_codes.py#L8-L9

[stable mountain]

bot/exts/fun/status_codes.py lines 8 to 9

HTTP_DOG_URL = "https://httpstatusdogs.com/img/{code}.jpg"
HTTP_CAT_URL = "https://http.cat/{code}.jpg"```

[cursive relic]

Oh I forgot the .src command.. thank you!

[stable mountain]

It has arrived!

Here's your reminder: if not done already
[Jump back to when you created the reminder](#dev-contrib message)

[wind ruin]

not sure if there was any discussion about it before but
now that python 3.11 rc2 is out, should snekbox be updated to use it or just ignore that version?

[atomic ivy]

I thought about it, but seeing as how there are not going to be any changes included other than bug fixes which have open PRs now till release, I thought there's no point

[gritty wind]

Bumping doesn’t cost us anything, feel free to PR it (to the correct branch) if you like

It just might take a few days to build and deploy since we didn’t finish the CI for it

[static canyon]

!remind 10h30M

[stable mountain]

Aye aye, cap'n!

Your reminder will arrive on <t:1663146007:F>!

ID: 5163

[stable mountain]

It has arrived!

Here's your reminder: See referenced message.
[Jump back to when you created the reminder](#dev-contrib message)

[static canyon]

Is there a reason that the docstring of the remind edit command includes that of remind edit duration? Surely the help for a subcommand shouldn't be included in the parent? If you want to know how the subcommand works then you can invoke help on the subcommand -- or even press the button for the subcommand since we have that

!help remind edit

[stable mountain]

Command Help

!remind edit
Can also use: remind change, remind modify

*Commands for modifying your current reminders.

The expiration duration supports the following symbols for each unit of time: - years: Y, y, year, years - months: m, month, months - weeks: w, W, week, weeks - days: d, D, day, days - hours: H, h, hour, hours - minutes: M, minute, minutes - seconds: S, s, second, seconds

For example, to edit a reminder to expire in 3 days and 1 minute, you can do !remind edit duration 1234 3d1M.*

Subcommands:

!remind edit content <id_> <content>
Edit one of your reminder's content.
!remind edit duration <id_> <expiration>
Edit one of your reminder's expiration.
!remind edit mentions <id_> [mentions]...
Edit one of your reminder's mentions.

[static canyon]

I'm thinking of just removing it as part of my current PR, since it's already in the duration subcommand

[mortal whale]

1

[outer oasis]

So... I tried to rip off @dusky shore, but I broke it.
I stole this:

for ext in walk_extensions():
  bot.load_extension(ext)

But it gives me this:
RuntimeWarning: coroutine 'BotBase.load_extension' was never awaited bot.load_extension(ext)
How/when/where do these get awaited?

[cold island]

I think lancebot is still using an alpha version of dpy2

Where these are synchronous methods

[outer oasis]

Oh
Welp
Who can I steal from then?

[cold island]

The Python bot

[static canyon]

Bump

[cold island]

Sure

[glossy berry]

Could someone look at my PR and give me an idea of why linting & tests is failing. I only changed one line of code and I'm a bit confused now.

[gritty wind]

Due to a recent change in our dependency manager, our CI is broken temporarily

[static canyon]

Fyi just pushed this 👍

What changelog version should a PR to botcore have currently? The next one is 8.3.0, but there's already a PR with that version so should I go latest that isn't taken by a pre-exisiting PR?

[gritty wind]

If you’re coordinating with someone to merge PRs in a certain order, sure

Otherwise no just bump as you need from main

We follow semver strictly

[static canyon]

I'm not coordinating with anyone, so I just go for 8.3.0 and whichever gets merged after will have to be changed?

[gritty wind]

Yeah

[static canyon]

Aight

[gritty wind]

I don’t think you’ll pass linting

Also please use two backticks for the docstring

Single backtick in Sphinx is a different thing than markdown’s

[static canyon]

Ah yeah

I forgot about that

Should be good now

[gritty wind]

I don’t understand if you want an actual review or not

[static canyon]

I mean you can, but I've not tested it yet so things might end up changing slightly to fix bugs with it

[kindred sun]

@timid sentinel hey do I need to do anything else to get bot#2275 reviewed?

[dusky shore]

GitHub

[bot] #2275 Removed "redis_ready" from additional_spec_asyncs in MockBot

[grand holly]

!contribuute

[stable mountain]

Contribute to Python Discord's Open Source Projects
Looking to contribute to Open Source Projects for the first time? Want to add a feature or fix a bug on the bots on this server? We have on-going projects that people can contribute to, even if you've never contributed to open source before!

Projects to Contribute to
• Sir Lancebot - our fun, beginner-friendly bot
• Python - our utility & moderation bot
• Site - resources, guides, and more

Where to start

1. Read our contribution guide
2. Chat with us in #dev-contrib if you're ready to jump in or have any questions
3. Open an issue or ask to be assigned to an issue to work on

[grand holly]

Hey anyone here

I would like to be assigned to this issue
Add number guessing game #1082

[lancebot]#1082

it's for the lancelotbot

Or I can add a cellular automaton generator to the bot

[outer oasis]

sir-lancebot#1082

[dusky shore]

GitHub

[sir-lancebot] #1082 Add number guessing game

[grand holly]

yeah that's the one

is there anything to do before I clone it and pull request?

[outer oasis]

It's still not approved - so - get a core dev to approve it
(of which I am not)

[grand holly]

oh

[grand holly]

ok what about this idea

I already have the code

and it's functional on a another bot I made

https://github.com/Sas2k/PokeStrike-Utility-Bot/blob/main/Cogs/Fun.py#L19-L72

[outer oasis]

Go ahead and make an issue for it
You can technically go ahead and make PRs too - there's just no guarantees if they're not approved

[grand holly]

done
https://github.com/python-discord/sir-lancebot/issues/1098

[clever wraith]

Hello folks,
I'm having a look at our _create_redis_session function in bot.__main__.py and i saw that the type hints are off.

async def _create_redis_session() -> RedisSession:
    """Create and connect to a redis session."""
    redis_session = RedisSession(
        host=constants.Redis.host,
        port=constants.Redis.port,
        password=constants.Redis.password,
        max_connections=20,
        use_fakeredis=constants.Redis.use_fakeredis,
        global_namespace="bot",
        decode_responses=True,
    )
    try:
        return await redis_session.connect()
    except RedisError as e:
        raise StartupError(e)

I'm saying they're off because the connect call returns None.
I know that this isn't that critical, but I'm also guessing it isn't intentional and consistent, right ?

Shouldn't we instead star the connection then return the redis_session ?

[atomic ivy]

the RedisSession.connect method does return self

[clever wraith]

Well, I don't see that.
Maybe my packages are outdated

[atomic ivy]

probably

[clever wraith]

I'll run the command to update them & check again

[timid sentinel]

had a busy week 😓. Am taking a look now but it seems like there are a bunch of commits on the PR that shouldn't be there, i'm not sure how that happened?

[clever wraith]

Yeah the history got messy when I merged from main again to get my branch up to date since the linting Github action was failing :/

Did I do it wrong ?

[fossil veldt]

@static canyon I'm looking at your pr for lance command suggests, not sure if it's fully within scope but would you want to add an exclusion for commands beginning with multiple (.) to not report a command error at all? It's a pretty common issue and was part of the original issue post

[clever wraith]

Yep, that was it.

[timid sentinel]

The other commits shouldn't be there so I think so. I'm not sure how you'd fix it though, maybe someone else here knows?

[clever wraith]

Maybe by rebasing ? It shouldn't conflict with was is currently there

And in the worst case, I can make a new PR & cherry-pick my commits since there aren't so many, what do you think ?

[static canyon]

Addressing that wasn't really part of the issue, and so I'd argue hasn't exactly been approved

I don't see why we need to specifically address that anyway

It just means if you do ...help it'll recommend help since ..help and help are seen as close enough

Is that so bad?

(The thing of having multiple responses is nothing to do with the multi-prefix, if that's what you were referring to by "it's a pretty common issue")

[timid sentinel]

I'll quickly try the solution mentioned here, if it doesn't work you could give rebasing a go, idk https://stackoverflow.com/questions/16306012/github-pull-request-showing-commits-that-are-already-in-target-branch

[static canyon]

@stable mountain help

Pretty sure the bot is ded

Ah, it's restarting nvm

[outer oasis]

Do all the bots respond to mentions?

[timid sentinel]

seemd to work

[gritty wind]

Most do

[clever wraith]

Yep !
I'll leave you to have a look then.
And maybe we should discuss how we should display the keywords for each rule, and maybe how many keywords per rule are we having now and how many could we have in the future ?

[fossil veldt]

hmm yeah that's true

is our CI broken again?

why is the check failing on that

[static canyon]

Yeah, there's a PR for it iirc

Hmm maybe not actually

This is what's failing:

[stable mountain]

.github/workflows/lint.yaml lines 73 to 78

# Install our dependencies if we did not restore a dependency cache
- name: Install dependencies using poetry
  if: steps.python_cache.outputs.cache-hit != 'true'
  run: |
    pip install poetry
    poetry install --no-interaction --no-ansi```

[static canyon]

Not sure why

@gritty wind @vale ibex hello, CI seems to be broken on lancebot. Is this a known thing? (sir-lancebot#1096 and sir-lancebot#1097)

[dusky shore]

GitHub

[sir-lancebot] #1096 Fix command suggestions.
[sir-lancebot] #1097 Fix issue #1050

[gritty wind]

Yes it’s a regression in poetry 1.2 that’s affected all our projects

We’re rolling out fixes as quickly as we can, but we’re limited by a lack of reviewers

So far bot is fixed

We’ll get them all eventually

[static canyon]

Cool 👍

[static canyon]

CC @fossil veldt (why CI is broken on lance)

And thanks for the review 👍

[cold island]

I don't mind being a trigger finger, but I really don't feel like I know enough about these things to review it

Considering the importance, if both you and Chris approve it it's enough for me

[static canyon]

@timid sentinel Fyi I've addressed your review on bot#2232, thanks!

[dusky shore]

GitHub

[bot] #2232 Add ability to delete multiple reminders at once & display mentions in !remind list.

[timid sentinel]

Nice, looks good

[clever wraith]

@timid sentinel All the changes should be in place for bot#2261.
I've tested it against the cases you've mentioned in your last comment & it works accordingly.
However, what do you think about making the command return the matched rules and we add a test for it ?

Have a look whenever you can, thanks !

[dusky shore]

GitHub

[bot] #2261 Add support to fetch rules via keywords

[fossil veldt]

bot#2278 is a pretty simple bug and fix if someone wants to approve it

[dusky shore]

GitHub

[bot] #2278 !pypi command TypeError on packages without description

[timid sentinel]

done 👍

[static canyon]

As the author, were you wanting to implement the fix yourself?

[fossil veldt]

yeah I'll probably get to it today, but if anyone else really wants to I don't mind

[static canyon]

I'll leave it to you 👍

[vale ibex]

@timid sentinel I don't mind removing that fwiw

it was just copy + paste

we might use it in future

[timid sentinel]

Nah might as well leave, does no harm

[vale ibex]

I've been adding it to all the bots 😅

[timid sentinel]

Thanks, have submitted a review. Tests aren't required, although it could be a nice if you'd like to, up to you!

[clever wraith]

Awesome, thank you.
I've resolved your comments.

This should be the last review for that PR.

[brazen charm]

does the doc build here https://github.com/python-discord/bot-core/pull/88 fail for anyone else? I can't get it locally with the poetry docs task https://github.com/python-discord/bot-core/actions/runs/3077983162/jobs/4973229590

[static canyon]

If you're talking about netlify I believe it's a known issue

[brazen charm]

I don't think netlify is involved there

[vale ibex]

I've ran into similar issues

if you delete docs/build and docs/output you should get it locally

not sure how it deals with invalidating the build cache

[brazen charm]

didn't get anything when I cleared the build and output

[vale ibex]

I got ```
Warning, treated as error:
C:\Users\chris\src\bot-core\botcore\utils\cooldown.py:docstring of botcore.utils.cooldown.block_duplicate_invocations:8:py:class reference target not found: botcore.utils.cooldown.ParamSpec

[brazen charm]

got it with a recreated environment, not sure what was going on

[vale ibex]

maybe needs a poetry install --sync?

to remove dangling deps

[clever wraith]

Does any one know why my lint-test action is failing here ?
https://github.com/python-discord/site/actions/runs/3078077698/jobs/4973387884

[timid sentinel]

It's currently failing on all PRs, https://github.com/python-discord/site/pull/772 should fix it when merged

So don't worry about it

[clever wraith]

Alright, all good then

I'm curious @timid sentinel, how would people know that they can now use keywords to search for rules ?

Aside from watching people who know about it using it

Idk 🤷‍♂️

[timid sentinel]

We'll post about it in #changelog when it's merged

That will let people know they can use keywords, although doesn't help so much with knowing what keywords they can use :P

[clever wraith]

Ahahahah true, I'll open up an issue for the new command that lists the available keywords

[timid sentinel]

Yeah, probably best for a follow up PR

[timid sentinel]

Thanks @vale ibex

[outer oasis]

Thank you for writing all of them

[vale ibex]

lol

[vale ibex]

@timid sentinel can you sanity check me?

Reviewing sir-lancebot#1097

[dusky shore]

GitHub

[sir-lancebot] #1097 Fix issue #1050

[vale ibex]

The url key on the original object doesn't seem to be documented on giphy's api

https://developers.giphy.com/docs/api/schema#image-object

I've tested locally and it's returned by the API

So I'm not sure if this is a mistake on giphy's part, or if we're using an undocumented key that's likely to cause issues again in future

I might suggest changing to data["data"]["images"]["downsized"]["url"]

as that's documented

and also guarantees the image is <2MB

[timid sentinel]

Ah yeah, your suggestion sounds good, similar to what I suggested a while ago when I looked into it (https://github.com/python-discord/sir-lancebot/issues/1050#issuecomment-1116368854).

I do wonder how they came up with what's in the PR though, maybe just from looking at the API response?

[vale ibex]

possibly

or giphy changes it's api (and docs) frequently

[tawdry vapor]

@vale ibex @gritty wind What's the reason for using the Poetry installer in the Dockerfile? I fail to see what benefits isolating Poetry itself has when it's going to be the only package installed outside of a venv.

[gritty wind]

Yeah I don’t mind either way, it’s just what was in the source we copied from

Since we’re using the base image, it doesn’t really matter

[tawdry vapor]

Hmm okay. I will PR a change then. Would make the Dockerfile simpler and avoid installing curl

[gritty wind]

👍

[tawdry vapor]

On second though, I think I'll leave it alone.

Because it's in a separate image now, the manner of installation is a bit opaque

And we can't guarantee consumers will not install other packages globally

@gritty wind I'm not following how your action configures the cache directories. Can you explain?

You're caching ~/.cache/py-user-base, but where is this actually being set as the user base dir?

[gritty wind]

It’s just the default location described in the docs

[tawdry vapor]

For the setup-python action?

[gritty wind]

Yeah

[tawdry vapor]

I did not consider that it would change the default user site.

[gritty wind]

For pip and poetry specifically, the GitHub python action actually mentions those by name

Under caching

[tawdry vapor]

I'm not seeing that in its docs

But I guess the poetry and pre-commit directories used are the defaults as well?

[gritty wind]

You’re right actually, I can’t find where py-user-base is listed

But the other two are listed in their docs for sure

[tawdry vapor]

Are you sure it's actually caching the user base then :D

[gritty wind]

It would emit a warning if a path is missing haha

[tawdry vapor]

Though is there any point in caching the user base if we're no longer doing user installs in CI?

Wouldn't you want to cache the poetry venv instead?

Or the site-packages inside of that. IDK... venvs are more finicky since they aren't portable

[gritty wind]

I’m not sure if it does, but I figured it shouldn’t hurt

But yes I hate the new setup

Installing to the system is no longer a supported use case however

[tawdry vapor]

In general I think caching user base was a "weird" thing to do - I don't see anyone else doing that, probably cause it's fragile to some degree. But the reasoning behind that is even if the package is downloaded, there is extra time required to install it. not having to do that can save time.

[gritty wind]

“Not using venvs in an anti-pattern”

[tawdry vapor]

I think the non portability of venvs is mostly with the activation scripts. CI is a controlled environment though. As long as we don't move the venv and cache based on Python version then I think we should be okay?

[tawdry vapor]

I think this justification is bogus but the other reason of not wanting to rely on pip anymore was perfectly reasonable. Though I'm not sure if I'd characterise user installs as too niche to replicate as a Poetry feature.

https://tenor.com/view/christian-bale-american-psycho-walk-jam-gif-16732505

me tuning out packaging bullshit and using requirements.txt

[gritty wind]

I’ve come to understand you ways, science man. And I must say, you’re a genius

[timid sentinel]

If I commit a suggestion from someone does their approval still count (as they'd be co-author, no?)?

[vale ibex]

Yea, coauthors are fine

It's only when some commits to the branch they can't approve

Whether that's an entirely new commit or a rebase

[gritty wind]

Did you review your own PR

Power move

[grand holly]

Any updates on this?

sir-lancebot#1098

[dusky shore]

GitHub

[sir-lancebot] #1098 1D Cellular Automaton Generator

[brazen charm]

I concur with mbrauh there, apart from it making a grid of cell based on some rules, I have no idea what it's doing with the arguments

[grand holly]

the arguments there make the generation happen

the rule-number says which type of automaton happens

the position tell where to place the starting cell

[brazen charm]

Explain how it works in the issue then, e.g. how the number relates to what's going to happen

[grand holly]

I did

i think

done

[static canyon]

The bot repo has a CustomLogger class that adds a trace method to prevent linting highlights. I've noticed that sir-lancebot has logging, uses .trace, but gives linting warning because it doesn't have the custom class.

Can we do a PR moving the class to bot-core instead, so that it's accessible by both bots?

[vale ibex]

It's already in botcore

Lance just isn't using it yet, until my pr

[static canyon]

Ah right

That makes sense

[vale ibex]

It's being bot core that is, my pr didn't change logging

Didn't want to double the size of the pr lol

[static canyon]

So bot and lancebot (currently) don't use the bot-core class?

[vale ibex]

Not sure

[static canyon]

It doesn't seem so

[vale ibex]

Should be possibly to check

[static canyon]

Yeah, it's not

Just checked

Can I open PRs for that?

[grand holly]

here is an example of a generation gif

[vale ibex]

Sure

You can open a pr on lance too if you want, just base it off my branch

[grand holly]

this is rule 60, with a single cell in the center

[static canyon]

I'll start with bot, and then check out lancebot

Can you link your PR please?

[vale ibex]

Not easily, I'm on mobile

[static canyon]

Ah no worries

[vale ibex]

There's only like 4 prs on lance atm

[static canyon]

Yeah, I'll look

[grand holly]

the bot will send a picture of the output not gifs since a gif takes time

[vale ibex]

That comment was a response to tizzy, not you

[grand holly]

I know I mistaking replied to it

[vale ibex]

Ok

[static canyon]

Am I right in saying that this still needs to stay in bot? https://github.com/python-discord/bot/blob/main/bot/log.py#L45-L129

It's the setup(), setup_sentry() and _set_trace_loggers()

[clever wraith]

I thought that adding tests never hurts, so take a look at the new changes if you'd like.

[brazen charm]

It's currently waiting on some other things, but could use a review or two on https://github.com/python-discord/bot-core/pull/88I

[vocal prairie]

Can the "the code jam is under way" banner be removed from this page? https://www.pythondiscord.com/events/

[tawdry vapor]

Yes. Want to create the PR? :)

[vocal prairie]

can't do it now, i'm afraid. i might be able to do it tomorrow evening if someone else doesn't, i don't particularly care.

[sacred fossil]

@static canyon sir-lancebot#1096 fixes the on command error but doesn't address the ...... part as in ......what is still triggering the command suggestion while it shouldn't be since to me it seems like Lance should only trigger if the message matches ^\.[^\.\s]+. Should I open a new issue and reference sir-lancebot#1900?

[dusky shore]

GitHub

[sir-lancebot] #1096 Fix command suggestions.

[static canyon]

It was never supposed to address that

It wasn't in the issue, and so didn't get approved, and so I didn't implement it

You can make another issue if you'd like to, and if it's approved I'm happy to implement

[sacred fossil]

Lovely

[cold island]

Do we have some way to keep up with new posts in Discourse? #mailing-lists message

[wind ruin]

they seem to support RSS feeds for most things, but we also have to consider how much of it (more specifically, which categories) we would want to keep up with

[gritty wind]

No longer reaches the proper audience
Why? Because they all retired of old age?

Mailing lists man

[cold island]

Yeah I think we'll want to keep track of specific categories like the PEPs

[static canyon]

@timid sentinel RE bot#2232

Would something like Mentions: <@user> (user#discrim) be good?

[dusky shore]

GitHub

[bot] #2232 Add ability to delete multiple reminders at once & display mentions in !remind list.

[timid sentinel]

yeah that sounds fine

[cold island]

iirc there should be a method in member utils for that

[static canyon]

Doesn't seem to be

There's get_or_fetch and handle_role_change

[cold island]

Hmmmmm

[vale ibex]

its in the messages utils iirc

[cold island]

Ah yeah

Why is it there lol

[vale ibex]

Hey @brazen charm your change seems to be merged into sphinx autodoc typehints. I just tried it locally and it seems to work.

If you want to pin the version to https://github.com/tox-dev/sphinx-autodoc-typehints/archive/d8a9b9c19450268f46fc89b048f8f1efb9bc248c.zip and relock I'd be happy to merge once CI passes

obviously when a full release is made we can pin to that version instead

re bot-core#88 for context

[dusky shore]

GitHub

[bot-core] #88 Add a decorator that stops successive duplicate invocations of commands

[brazen charm]

I think a proper release will go out today or tomorrow, if my last contribution there is anything to go by

[vale ibex]

ah alright fair enough, no need to rush then

[static canyon]

The util returns the id alongside the mention, not the name#discrim

Which imo is pointless here because the whole point is we want to display extra information for when the ping doesn't render

If the ping doesn't render then people will already see the id

[static canyon]

Fyi just pushed @timid sentinel

[static canyon]

I'm writing a PR where I move the re-run logic of !e into bot-core, and I'm trying to also fix a bug with the logic.

After an edit, the bot just assumes everything after !eval is the code, even though it could be a different command argument (in this case, specifying the python version to use)

!eval 3.10 print("hello world")

[stable mountain]

@static canyon :x: Your 3.10 eval job has completed with return code 1.

001 |   File "<string>", line 1
002 |     3.10 print("hello world")
003 |          ^^^^^
004 | SyntaxError: invalid syntax

[static canyon]

I.e. this is a valid invocation, but doesn't work when you do it through an edit

What's the best way to fix this?

My current solution is to have the user specify what args can be changed, and then see what actually changed, and go from there (and return what can be changed if everything went correctly)

[tawdry vapor]

Is it possible to re-parse the message as a command and get it arguments, without actually executing the command again?

On second thought, that command may rely on additional parsing within rather than just on d.py

So if that's the case, maybe refactor the parsing into a separate function?

[static canyon]

Yeah, it is

!d discord.ext.commands.Bot.get_context

[stable mountain]

discord.ext.commands.Bot.get\_context

await get_context(origin, /, *, cls=...)```
This function is a [*coroutine*](https://docs.python.org/3/library/asyncio-task.html#coroutine).

Returns the invocation context from the message or interaction.

This is a more low-level counter-part for [`process_commands()`](https://discordpy.readthedocs.io/en/latest/ext/commands/api.html#discord.ext.commands.Bot.process_commands "discord.ext.commands.Bot.process_commands") to allow users more fine grained control over the processing.

The returned context is not guaranteed to be a valid invocation context, [`Context.valid`](https://discordpy.readthedocs.io/en/latest/ext/commands/api.html#discord.ext.commands.Context.valid "discord.ext.commands.Context.valid") must be checked to make sure it is. If the context is not valid then it is not a valid candidate to be invoked under [`invoke()`](https://discordpy.readthedocs.io/en/latest/ext/commands/api.html#discord.ext.commands.Bot.invoke "discord.ext.commands.Bot.invoke").

Note

In order for the custom context to be used inside an interaction-based context (such as [`HybridCommand`](https://discordpy.readthedocs.io/en/latest/ext/commands/api.html#discord.ext.commands.HybridCommand "discord.ext.commands.HybridCommand")) then this method must be overridden to return that class...

[static canyon]

Via that ^

[static canyon]

I mean I got close, but then dpy didn't have a feature to return {command_arg: value_used, ...} from a Context object (or at least, I can't find it)

Which I needed

If I had that I'd be set

The closest I can get is using ctx.args and pairing to ctx.command.clean_args, but there's no fool-proof matching since first provided arg (given by .args) isn't necessarily the first possible arg (given by .clean_args)

As an example of my end goal:```py
@bot.command()
async def greet(ctx, name: str = "World"):
"""Will keep saying hello until the user doesn't edit name anymore."""
print("running")
while True:
response = await ctx.send(f"Hello, {name}!")

    changed_args = await check_rerun_job(ctx, response, args_user_can_change=('name',))
    if not changed_args or "name" not in changed_args:
        break  # user didn't change the name, so stop saying hello```

[tawdry vapor]

I'll look into this later this evening and let you know if I think of a good solution.

[static canyon]

Thanks, appreciate it 😄

[cold island]

Honestly if it's resending the result anyway, we could make it re-run the command on the message

Pretty sure there's something in dpy for that

The code change will be mostly not having a loop anymore, since if the rerun reaction is pressed, it creates a task to rerun the command and then just ends

[tawdry vapor]

Yeah, I think you're correct.

One difference would be that the stats would count every re-eval but that's fine, and we probably should've been doing that already.

[static canyon]

I was trying to avoid that, but I don't think I can really

So that's most likely what I'll end up doing

Which is annoying considering how close I was to making my way work

I'll need to see whether error handling works properly when reinvoking, but I don't see why it wouldn't

[cold island]

Yeah it can happen, some case that's only discovered in testing. But it seems like a good way to solve it, I use re-invocations in the filtering rewrite and it seems to work well

[static canyon]

Yeah, that's fair. I'll test reinvoking whenever I get time

Good to hear it works for filters

[static canyon]

So I thought this was going well, until I tried subcommands...

Apparently bot.get_context will get the parent command, not the subcommand

And so when I check to see if the command changed, it says it did (even though it didn't)

Now I can just not check the command is the same, but that means it allows you to invoke other commands when editing, which I imagine we don't want

Tl;Dr; when using subcommands ctx.command != (await ctx.bot.get_context(ctx.message)).command even though it's the same message being parsed

From how I see it, we either let the user edit the message into a different command (meaning e.g. an !eval invocation could become a !ping invocation), or don't let users re-run through edit

I don't see a big issue with letting users edit into other commands, other than it's not intended behaviour and can't be done unless the command has the util for re-running

[midnight jasper]

@clever wraith I have a project you can contribute to.

[clever wraith]

ill see if i can be of any use

[midnight jasper]

Do you know what an api is

[clever wraith]

yes

[midnight jasper]

Do you know what an api wrapper is?

[clever wraith]

*not new to python or coding in general

[clever wraith]

yes

[midnight jasper]

Alr. What’s ur most advanced/complex project

[clever wraith]

worked on https://localmonero.co (EDIT, i worked with the OWNERS regarding escrowing funds)

created https://bitdaily.xyz

just bored tbh

no longer work on any of them, for clarification

[midnight jasper]

Wait u made ur own cryptocurrency and platform for it?

[clever wraith]

no, i made my own platform for cryptocurrency transactions

the simpliest way to explain what i made:
a framework for verifying, sending, and notifying if a transaction has been complete

[austere hornet]

@clever wraith @midnight jasper this channel is not for discussing projects other than the ones here on PyDis. You should move your conversation to one of the off-topic channels

!ot

[stable mountain]

Off-topic channel: #ot2-never-nester’s-nightmare

Please read our off-topic etiquette before participating in conversations.

[clever wraith]

my bad

want to go dms?

[midnight jasper]

The dog guy told us to come here bc he was looking for a project to xontrivute to

Maybe it was jsut to contribute to pydis projects?

Idk

[clever wraith]

ye

[austere hornet]

Yes, that's probably what they meant.

[cold island]

Why do you need to know if the command changed? Also if it's not in the command maybe it's ok the args?

If you run the command from the ctx it should just work

[static canyon]

Why do you need to know if the command changed?
To stop them editing it into a different command 🤷

We don't have to do that, but it seems odd to allow behaviour that isn't a feature (i.e. you can't edit any command into a different command)

[cold island]

Hmmm right

[static canyon]

maybe it's ok the args?
That... might work

[cold island]

In the args*

[static canyon]

Yeah

It doesn't work unfortunately

Because ctx.command is a Group, which doesn't have .args

It presumably has something to realise there's a subcommand though, so maybe I can find that

[cold island]

Ctx.args

[static canyon]

An empty list

[cold island]

Weird

[static canyon]

Gonna delve into the source code and see if I can find anything

This seems to be the code that detects a subcommand https://github.com/Rapptz/discord.py/blob/master/discord/ext/commands/core.py#L1619-L1632

[stable mountain]

discord/ext/commands/core.py lines 1619 to 1632

ctx.invoked_subcommand = None
ctx.subcommand_passed = None
early_invoke = not self.invoke_without_command
if early_invoke:
    await self.prepare(ctx)

view = ctx.view
previous = view.index
view.skip_ws()
trigger = view.get_word()

if trigger:
    ctx.subcommand_passed = trigger
    ctx.invoked_subcommand = self.all_commands.get(trigger, None)```

[cold island]

What is that get_word

How does it get the word

[static canyon]

I think I've got it

I just need to a recursion support so sub-subcommands, sub-sub-subcommands, etc. work

It's looking good

[vale ibex]

oh hey, sir-lancebot#1092 is looking mighty fine to review

[dusky shore]

GitHub

[sir-lancebot] #1092 Python 3.10 & bot-core migration

[static canyon]

Maybe not. Don't think I can make it recursive, since I need a ctx.view again which won't exist for the command object

I can probably figure out an iterative approach (maybe)

Work all of the commands, and see if any match

new_ctx = await ctx.bot.get_context(ctx.message)
if new_ctx.command is ctx.command:
    await new_ctx.command.invoke(new_ctx)
    return

else:
    print("not same command, so trying to get subcommand")
    if not isinstance(new_ctx.command, Group):
        print("not a group so can't have subcommand")
        return
    group: Group = new_ctx.command

    for command in group.walk_commands():
        if command is ctx.command and new_ctx.message.content.startswith(ctx.prefix+command.qualified_name):
            print("FOUND SUBCOMMAND!")
            await new_ctx.command.invoke(new_ctx)
            return

await ctx.send(":warning: Aborted since you changed the command.")```this *seems* to be working perfectly

Nvm

If you invoke a parent, you can then edit into a subcommand

[summer garden]

What do we think about adding a !http command to look up what all of the codes mean? I know we have one that is kind of memish with the cat images, but it doesn't offer an explanation

The one we have is !http_status

[atomic ivy]

we have those in lancebot, yeah

might be nice to have

[vale ibex]

Currently .http_status picks a between cat and dog randomly if not specified

We can change that, so that if you don't specify it gives you the "formal" description

Adding the .http alias if it isn't there

[wild prism]

ok but can it tell me the correct code to use

[vale ibex]

There's a flowchart for that

https://www.codetinkerer.com/2015/12/04/choosing-an-http-status-code.html

Quite old, so doesn't have the new "hip" codes twitter is trying to make happen, but gets the job done for everything else

Oh, I lied, it does

[cold island]

"are you twitter" hahaha

[vale ibex]

@summer garden did you want to do this yourself?

[summer garden]

yeah I would

[vale ibex]

[summer garden]

I'll probably do this over the weekend

I envision an embed for it with a link in the title to some more concrete info

can we just do a big enum for all of the codes?

[vale ibex]

there's one in the http built-in

it'll resolve codes to their short meaning

and as you say, we can link out for more info

[summer garden]

The python http library?

[vale ibex]

Yea

[summer garden]

I'm looking and there seems to be some missing. I was looking for 111

[vale ibex]

http.HTTPStatus

[summer garden]

Are there a bunch of different standards?

[vale ibex]

111?

don't think that's in the standard

[cold island]

.http 111

.http_status 111

[dusky shore]

Unable to find status floof for 111.

[summer garden]

code 111 came up earlier today so i had it memorized

https://webconcepts.info/concepts/http-warn-code/111

[vale ibex]

I thought warn codes were deprecated?

https://developer.mozilla.org/en-US/docs/Web/HTTP/Headers/Warning

Yea

Was deprecated in rfc9111

[summer garden]

Should they be included for the sites that may still be using them?

[vale ibex]

rfc7234, the one that webconcepts links to, has at the top that it was Obsoleted by: 9111

https://datatracker.ietf.org/doc/html/rfc7234 https://datatracker.ietf.org/doc/html/rfc9111

[vale ibex]

I'd rather not, since it would mean maintaining our own list, and potentially signalling that they're still ok to use

warn codes are in addition to http response codes

they're a separate header

[static canyon]

After way, way too many hours, I've finally got a solution for a re-run util. I'll upload it to GitHub tomorrow, but here's the code on a paste: https://paste.pythondiscord.com/jazuvifome
CC @cold island @tawdry vapor

[gritty wind]

Wikipedia has some good reference articles for all the codes that goes beyond just their usage (history and stuff). You could use that for the external link

Don’t need to hardcode it either, just ask the api for the url, or use a string template

[vale ibex]

https://developer.mozilla.org/en-US/docs/Web/HTTP/Status/ {code}

mdn good

history and stuff could be cool though

Maybe both? Not sure what the embed would look like

[summer garden]

The specific code link can be in the title and the "find out more" wikipedia link in the footer?

[vale ibex]

Yea, that could be good

pushing people to mdn in general is good, it's an amazing resource

[static canyon]

I've now pushed to GitHub, and it's ready for review 👍

bot-core#137

[dusky shore]

GitHub

[bot-core] #137 Add check_rerun_command util to commands.py

[sweet glacier]

Hello guys

long time no see

[brisk brook]

Hi 🙂

[static canyon]

  EnvCommandError

  Command C:\Users\tizzy\AppData\Local\pypoetry\Cache\virtualenvs\sir-lancebot-INSnasTY-py3.10\Scripts\pip.exe install --no-deps C:\Users\tizzy\AppData\Local\pypoetry\Cache\artifacts\21\17\d4\ad0f3a8b93e95e364982481a40ecad444c6a57d5fad7c43f74c4283a04\emoji-2.1.0.tar.gz errored with the following return code 1, and output: 
  Processing c:\users\tizzy\appdata\local\pypoetry\cache\artifacts\21\17\d4\ad0f3a8b93e95e364982481a40ecad444c6a57d5fad7c43f74c4283a04\emoji-2.1.0.tar.gz
    Installing build dependencies: started
    Installing build dependencies: finished with status 'error'
    error: subprocess-exited-with-error
    
    pip subprocess to install build dependencies did not run successfully.
    exit code: 1
    
    [3 lines of output]
    WARNING: Skipping page https://pypi.org/simple/setuptools/ because the GET request got Content-Type: Unknown. The only supported Content-Types are application/vnd.pypi.simple.v1+json, application/vnd.pypi.simple.v1+html, and text/html
    ERROR: Could not find a version that satisfies the requirement setuptools>=40.8.0 (from versions: none)
    ERROR: No matching distribution found for setuptools>=40.8.0
    [end of output]
    
    note: This error originates from a subprocess, and is likely not a problem with pip.
  error: subprocess-exited-with-error
  
  pip subprocess to install build dependencies did not run successfully.
  exit code: 1
  
  See above for output.
  
  note: This error originates from a subprocess, and is likely not a problem with pip.
  

  at ~\AppData\Roaming\pypoetry\venv\lib\site-packages\poetry\utils\env.py:1195 in _run
```Trying to redo the env for my local lancebot, and getting this error (failing to get setuptools>=40.8.0). Anyone have any ideas?

A manual pip install emoji==2.1.0 seems to have worked 🤷

[vale ibex]

Try poetry cache clear --all

WARNING: Skipping page https://pypi.org/simple/setuptools/ because the GET request got Content-Type: Unknown. this seems like the important line

Also, what version of poetry are you on?

updating might help if you're not on latest

[static canyon]

It should be the latest

Or near enough

1.1.14 apparently

RE sir-lancebot#1079

The code snippets feature uses wait_for_deletion, which currently isn't in lance or bot-core. In fact, bot-core is missing all the message utils, so should I create a PR on bot-core adding these before doing the lance PR?

[dusky shore]

GitHub

[sir-lancebot] #1079 Port GitHub code display embed from Python

[vale ibex]

Try updating to 1.2

1.1.16 apparently fixed a caching issue

[static canyon]

What's the best way to update it?

[vale ibex]

Poetry self update is a thing in a recent version, not sure if you have it

If not, do the pipe to python thing shown on the website

[static canyon]

Yeah, it seems to be running 👍

[static canyon]

The command worked, but now poetry is broken```py
Traceback (most recent call last):
File "C:\Users\tizzy\AppData\Local\Programs\Python\Python310\lib\runpy.py", line 196, in _run_module_as_main
return run_code(code, main_globals, None,
File "C:\Users\tizzy\AppData\Local\Programs\Python\Python310\lib\runpy.py", line 86, in run_code
exec(code, run_globals)
File "C:\Users\tizzy\AppData\Roaming\Python\Scripts\poetry.exe_main.py", line 4, in <module>
ImportError: cannot import name 'main' from 'poetry.console' (C:\Users\tizzy\AppData\Roaming\pypoetry\venv\lib\site-packages\poetry\console_init.py)

[vale ibex]

Lol love to see it

Run the command on the website

Should fix it

[static canyon]

Aight

[static canyon]

I... don't think it did?```py
PS C:\Users\tizzy\sir-lancebot> poetry --version
PS C:\Users\tizzy\sir-lancebot>

I mean the error is gone, but no output

Same for things like poetry --help

Re-invoking the command on website just says it's already installed

Huh

[vale ibex]

That's likely the problem 😅

[static canyon]

bump btw

[timid sentinel]

if anyone has a chance, a second review on bot#2245 would be good

[dusky shore]

GitHub

[bot] #2245 Update autoreview system

[rapid swallow]

I'll try to review it tomorrow

[vale ibex]

yea, having that in bot-core seems reasonable

[vale ibex]

reviewed

you're welcome

[timid sentinel]

Thanks!

[vale ibex]

lol

[static canyon]

Should I move all the message utils? (from both bot and lance)

[vale ibex]

If they're easy enough sure

Not sure how much they rely on constants

[static canyon]

Don't think they rely on any :)

Ah, bot has Emojis, MODERATION_ROLES, NEGATIVE_REPLIES

[static canyon]

So how do we approach that?

[vale ibex]

depednds on the nature of how they're used

they could be hard coded into bot-core if appropriate

alternatively, we could do something similar to how the roles get passed into the generic view

[static canyon]

I'll take a look and see

[rapid swallow]

lmao

[static canyon]

The only issue I'm having atm is figuring out the botcore equivalent of bot's bot.instance

right_reaction = (
        user != bot.instance.user
        and reaction.message.id == message_id
        and str(reaction.emoji) in allowed_emoji
    )```

[vale ibex]

you could refactor that to be not user.bot

slightly different behaviour, but still fine

[static canyon]

Very true

Yeah

    try:
        try:
            await bot.instance.wait_for('reaction_add', check=check, timeout=timeout)
        except asyncio.TimeoutError:
            await message.clear_reactions()
        else:
            await message.delete()
    except discord.NotFound:
        log.trace(f"wait_for_deletion: message {message.id} deleted prematurely.")
```in `wait_for_deletion`

I guess make the util take bot as an arg, with a typehint of commands.Bot?

[static canyon]

This is what I've gone with for now

[static canyon]

A PR has been opened! bot-core#141

[dusky shore]

GitHub

[bot-core] #141 Migrate all message utils from bot and sir-lancebot.

[cold island]

hmmmm is there a way to do something when a modal is canceled?

I guess there's no response so the modal object wouldn't know...

[cold island]

My code is such a hot pile of garbage, I'm surprised it actually does what I want

[clever wraith]

Hello folks !
I was going through the different issues that we have in our bot project, and I've seen a lot of ones related to filters so it got me intrigued to how it works.

Is there a particular way you'd recommend to have a high level overview of our filtering system & then deep dive ? Or shall I just browse the code ?

[cold island]

If you mean message filtering, I'm probably handling most of them atm. Maybe I should assign myself properly

[clever wraith]

I know you are, I saw all the issues & I was wondering if I contribute, but then I was like "I have no clue of how it works"

But yeah, maybe they should be assigned to you

[cold island]

I'll try to document it properly when I open the PR, but atm it's mostly work that can't really be split

[clever wraith]

I was mostly looking at bot#1379 now, is that something you're working on ?

[dusky shore]

GitHub

[bot] #1379 Adjust rich_embeds filter so it doesn't catch edge case

[cold island]

Ah no, like the issue says it's not blocked by my work

[clever wraith]

Yep, I saw that, but then I when you said "I'm probably handling most of them atm", I thought this could be another thing you're working on & unassigned to you

Alright !

[cold island]

yeah there are some things there which aren't blocked, just wasn't sure which you were referring to

Anything that's related to how a message is processed is probably not blocked

if you need help to find the relevant code that needs changing lmk

[clever wraith]

I'm actually looking at it atm.
I just need to reproduce the issue with a mobile twitter link.
I'll make an account & see what I can make out of it, thanks !

[cold island]

Should I assign you then?

[clever wraith]

Gladly !

[cold island]

ok, can you comment on the issue?

[clever wraith]

Sure thing

[cold island]

man, it's too easy to click the "assign yourself" button. Assigned you now

[clever wraith]

Ahahahha I saw that, thanks !

[cold island]

Is there an extra step when running migrations on site?

found it

Is it possible to do it in the container? it says the fs is read-only and I'd rather not set up everything locally

[tawdry vapor]

I believe what you're seeking was discussed here https://github.com/python-discord/site/pull/622#discussion_r757771435 and we never figured out a solution

[cold island]

Ah there's poetry run task dockermigrations, nice

[vale ibex]

yea, my intention was to add that in that PR

but there was some odd file permission issues on linux that I couldn't fix as I didn't have a linux machine to reproduce the error and investigate

[cold island]

wait, but the PR description says it adds that

oh it's still open

ok

[vale ibex]

Yea :D

[cold island]

What about Mark's suggestion in https://github.com/python-discord/site/pull/622#discussion_r776490897 ?

[vale ibex]

Yea I think that's what we'd want to change the task to

it doesn't solve the permission issue though

[cold island]

then.. it doesn't work?

[vale ibex]

The PR is in a broken state on linux, yea

due to the volume mounts

[cold island]

why specifically linux?

[vale ibex]

File permission differences

[cold island]

hmmm

[vale ibex]

as in, the PR works fine on my windows machine

but obviously we want it to work on linux before merging too

[cold island]

alright.. I'll just steal the command from the PR in the meanwhile

[vale ibex]

you'll need to add the compose volume(s) too

[cold island]

https://tenor.com/view/thanos-a-small-price-to-pay-for-salvation-a-small-price-to-pay-price-to-pay-avengers-gif-22079658

[vale ibex]

lol

[clever wraith]

How do I setup the webhook in these cases ?
Am I supposed to create a webhook to the same TextChannel and put its id in the config.yml ?

Is that documented somewhere ?

Ok found it

Found out what to do I mean

[clever wraith]

But still curious about this

[tawdry vapor]

I think the setup guide for the bot has a page that explains how to create a webhook and get its ID.

Does anyone have comments on this? https://github.com/python-discord/bot-core/pull/88#discussion_r897390979

[clever wraith]

I don't remember seeing it, maybe I didn't pay much attention

Also, on this line https://github.com/python-discord/bot/blob/71fbd5aa7ef8fdf3547dbae845cf7e6b9bc08277/bot/exts/moderation/watchchannels/_watchchannel.py#L131
remove_cog isn't awaited, isn't that a bug ?

[stable mountain]

bot/exts/moderation/watchchannels/_watchchannel.py line 131

self.bot.remove_cog(self.__class__.__name__)```

[tawdry vapor]

If it's a coroutine function and it's not awaited then yes it's a bug

[clever wraith]

It's an async function, yes

[clever wraith]

I've just opened the issue, and it's bot#2281

I'll include the fix in another PR that I'm working on, so it'll be great if you can assign it to me.
Thanks!

[dusky shore]

GitHub

[bot] #2281 Async function remove_cog not awaited

[clever wraith]

Hello folks,

I need to make a function that does some string operations on a URL.
This will be used in the watch_rich_embeds filter to fix a bug mentioned in bot#1379

Does it need to live in bot only ? ( If yes, maybe in bot.utils.helpers ? ) or in bot-core ?

[dusky shore]

GitHub

[bot] #1379 Adjust rich_embeds filter so it doesn't catch edge case

[cold island]

bot is fine

Can also just be a helper function in the same module

[clever wraith]

Alright, thanks !
If someone wants to have a look, it should be rather quick in bot#2283

[dusky shore]

GitHub

[bot] #2283 1379 tranform urls upon rich embed filter application

[grim prairie]

Wait: where did the announcement about the shortcut tags go for rules?

[wild prism]

#changelog

[grim prairie]

Sheesh. Okay thanks. I thought I was going crazy

... or at least more crazy than normal

[cold island]

fwiw I just tried it on Windows and it gave me the read-only error

[summer stone]

I need help with solving task based on Lagrange theorem

Does somebody here know math good enough?

[vale ibex]

Interesting, I wonder if it's a docker version change, I haven't tried it myself recently

The thing is we're mounting .:/app as read only, and then a more specific path as writable

So there may be odd interactions there

[vale ibex]

Which issue are you working on?

[summer stone]

Express number as a sum of 4 squares

[vale ibex]

No, i mean which issue on our github repo?

[summer stone]

What's github repo?

[vale ibex]

This channel is for discussing Python Discord projects, found here https://git.pydis.com/

Check the channel description

If you're not here for help related to one of our projects, your best bet it to read #❓｜how-to-get-help

[vale ibex]

anyway, for now it might be easier to run the site locally, connecting to the other services using compose

DATABASE_URL=postgres://pysite:pysite@localhost:7777/pysite
METRICITY_DB_URL=postgres://pysite:pysite@localhost:7777/metricity
SECRET_KEY=suitable-for-development-only
STATIC_ROOT=/var/www/static
DEBUG=1

[cold island]

Is it possible to mount it as writable?

[vale ibex]

we could mount app as writable, the only reason it isn't is that if the code gets altered by the images somehow, it wouldd be reflected on the host too

but I guess that's no different to running on host

[vale ibex]

we could make a new docker compos spec, with the app volume mounted as writable

then if you want to run migrations in a fully docker setup, you can use that compose file instead.

[cold island]

Update the compose file and dockerapp file.

The development app from the compose file will now apply migrations on boot and additionally, it will use Django's development server along with the current directory mounted read-only into the container in order to support code reloading, helpful while development.

@last patio a commit you made in 2018. Are you saying that without read-only code reloading won't work?

Or were you saying it's mounted to support reloading

[vale ibex]

If it was the case back then, it isn't anymore

[last patio]

this

mounted read only to prevent permission messes

[vale ibex]

just get a better os

[last patio]

thanks

[vale ibex]

I heard rocky was good

[last patio]

docker issues

A banned

[grand holly]

bot#1

[dusky shore]

GitHub

[bot] #1 Add "Now playing" status

[grand holly]

sir-lancebot#1

[dusky shore]

GitHub

[sir-lancebot] #1 Set up some simple Discord bot boilerplate.

[grand holly]

bot[1]

bot.1

hmm....

[gritty wind]

Could I get some help with pre-commit

[vale ibex]

No

go away

thanks

[gritty wind]

Flake8...................................................................Failed
- hook id: flake8
- duration: 0.92s
- exit code: 1

C:\Python\Python310\lib\site-packages\setuptools\command\install.py:34: SetuptoolsDeprecationWarning: setup.py install is deprecated. Use build and pip and other standards-based tools.
  warnings.warn(
The command line is too long.

You can probably ignore the poetry warning, I'm just not sure if the The command line is too long. is part of it or not

Thanks Chris

Flake passes on it's own

(This is the most verbose output I could generate)

It's the flake and pre-commit versions from this PR https://github.com/python-discord/site/pull/769

[vale ibex]

can we bump versions in .pre-commit-config?

[gritty wind]

flake8 = "5.0.4" and pre-commit = "2.20.0"

Sure, but pre-commit uses system local flake

So it doesn't matter

[vale ibex]

Ah

[gritty wind]

Using local flake is a source of so much pain

But specifying all linting dependencies twice is not great either

Here's the flake config (unchanged from master):

  - repo: local
    hooks:
      - id: flake8
        name: Flake8
        description: This hook runs flake8 within our project's environment.
        entry: poetry run flake8
        language: system
        types: [python]
        require_serial: true

[vale ibex]

Do we need to bump the version of setuptools?

Try setuptools==58.2.0

[gritty wind]

That would be my system setuptools, and I'm pretty sure it's producing a warning because it's too up to date

I'm already on Version: 65.4.1

Poetry is doing something in a fashion that has been deprecated

So I'd need to downgrade to remove that warning

Or wait for poetry to fix it

[vale ibex]

Fun

[gritty wind]

I technically could just ignore this since CI does not run flake through pre-commit

But then no one would be able to commit to site

Which does seem like an oversight

[gritty wind]

Yeah I fixed that locally

[vale ibex]

Ahh right, the error above from locally?

[gritty wind]

site-py3.10C:\Github\PyDis\site>flake8
.\pydis_site\apps\api\tests\test_github_utils.py:31:68: B026 Star-arg unpacking after a keyword argument is strongly discouraged, because it only works when the keyword parameter is declared after all parameters supplied by the unpacked sequence, and this change of ordering can surprise and mislead readers.


site-py3.10C:\Github\PyDis\site>flake8

site-py3.10C:\Github\PyDis\site>

Yeah so flake passes now

Just pre-commit is broken big-time

[vale ibex]

try remaking the precommit env?

[gritty wind]

Would pre-commit clean do that?

Because I'm not sure where it's installed

[vale ibex]

pre-commit uninstall iirc

[gritty wind]

Nah that would just get the hook

[vale ibex]

clean is just for cache

[gritty wind]

Clean seems to have done it

As well as deleted all the deps for all the projects on my system

Use with caution lol

[vale ibex]

nice

nah dude, that's optimisation

[gritty wind]

Anyway, flake is still failing in pre-commit

I've pushed all my local changes now, you should be able to see the same things I'm seeing

I imagine CI is going to pass, so I've locked down the PR

[vale ibex]

running locally now

[gritty wind]

Uhhh

CI failed lol

Interestingly, I did get that warning locally, but only on one file

[vale ibex]

I removed require_serial: true from the pre commit config and it ran fine locally

[gritty wind]

I can confirm the same result

I also got like 400 deprecation notices, which is interesting

[vale ibex]

I didn't

oh

in tests

yea didn't run those

[gritty wind]

From pre-commit

What version of setuptools are you on

[vale ibex]

65.4.1

[gritty wind]

Ah, they only show up if you run with -v

verbose

[vale ibex]

🕙 16:05:38 ❯ poetry run task lint -v
Check for merge conflicts................................................Passed
- hook id: check-merge-conflict
- duration: 0.24s
Check Toml...............................................................Passed
- hook id: check-toml
- duration: 0.21s
Check Yaml...............................................................Passed
- hook id: check-yaml
- duration: 0.35s
Fix End of Files.........................................................Passed
- hook id: end-of-file-fixer
- duration: 0.32s
Mixed line ending........................................................Passed
- hook id: mixed-line-ending
- duration: 0.28s
Trim Trailing Whitespace.................................................Passed
- hook id: trailing-whitespace
- duration: 0.3s
check blanket noqa.......................................................Passed
- hook id: python-check-blanket-noqa
- duration: 0.23s
Flake8...................................................................Passed
- hook id: flake8
- duration: 4.84s

Unable to find qualified name for module: manage.py

[gritty wind]

Lol what

If you still have it up, could you run tests?

These are straight up exceptions not just deprecations

And I can't repro

[vale ibex]

yea will do now

just need to rebuild everything

I cleared my docker system not long ago

[gritty wind]

Ok I found where the deprecations are coming from

Weird it didn't pop up when running the project itself, seems django's deprecation warnings only come up when the code is activated

I guess that works out for us since we have a 100% coverage requirement on site

from datetime import datetime this is going to be the bane of my existance today

[vale ibex]

tests are frozen half way through

😅

[gritty wind]

Oh nah they are just kinda slow

Kill it if it takes >30 seconds

[vale ibex]

it completed 248 tests without any warnings

out of 285

rerunning a few times it never got passed that

[gritty wind]

You have to explicitly enable warnings as an env var

set PYTHONWARNINGS=error

Deprecation warnings are disabled by default

It's fine, I've gone ahead and fixed everything

[vale ibex]

cool cool

[gritty wind]

Tests seem to be taking significantly longer now

Fuck

2 3 4m and going compared to the 18s on main

I wonder if it's broken

[vale ibex]

My test suite is consistently getting stuck

[gritty wind]

Ok, perhaps one of the tests is now broken on linux machines, any idea which one?

[vale ibex]

I'm on windows

running test suite locally

[gritty wind]

Hm, I'm not getting that

Oh tbf I'm not running through the manage.py

[vale ibex]

when I ctrl+c it's always on File "C:\Users\chris\src\site\pydis_site\apps\redirect\tests.py", line 32, in test_redirects

[gritty wind]

Since I'm running through pycharm

[vale ibex]

I'm running poetry run task test

[gritty wind]

As I've said before, we should delete the redirects app

[vale ibex]

test = "coverage run manage.py test"

[gritty wind]

Ok, this is a good lead, thanks chris

[vale ibex]

https://paste.pythondiscord.com/izujoqiqom

The only other line I can see in our code is File "C:\Users\chris\src\site\pydis_site\apps\resources\views\resources.py", line 117, in get

[gritty wind]

How are you still getting the whitenoise error

That should be fully suppressed now

Ok repro confirmed

And now I've confirmed it's only if you run through manage.py

It seems to be a django test runner issue, not something in our config/usage specifically. If I bypass all the test stuff in our config and just have it run through manage.py normally, it still freezes

Seems to be an issue with django 4.1.*

Running it with a debugger, there doesn't seem to be any place where it freezes or takes an unexpected amount of time

Although I've only tested one redirect

Ok, seems it isn't all redirects, just a specific app most likely

[gritty wind]

Ok, seems that all 10 resource redirects take >12 seconds per resource

These are all dynamic

Seems like the django render function is responsible

Maybe the template is pulling in network resources?

Well it would've failed my original test then hmm

Ok found it, there does seem to be something here: ```html
<div class="content is-flex is-justify-content-center">
<div class="container is-fullwidth">
{% for resource in resources.values %}
{% include "resources/resource_box.html" %}
{% endfor %}
</div>
</div>

This block alone takes 15 seconds per resource

That includes ~72 resources in total, the question is why does it start taking so long after 4.1

Huh, it takes just as long on django 3

Hmm, running now with the same version as CI, it's still really slow

Does include make a subrequest of some sort

[gritty wind]

Well actually, if downgrading did not resolve the issue, perhaps that's a red herring

I did actually make another change in this PR to the redirects app itself, but I don't see why that would be the cause when it's resources that's failing, not all redirects

It was not the cause

I reverted all package changes to what's on master, and reverted to python 3.9 which fixed the issue. Now i have the incredibly fun task of again bumping each dep one by one till the issues start again

It was sentry

[vale ibex]

sentry is pretty recent in the call chain for my error

[gritty wind]

Well sentry would be present in literally anything that uses the client since it wraps the requests

[vale ibex]

true

[gritty wind]

I have no idea why it slows things down, why specifically for this page, and why on includes

[vale ibex]

So does it all work well fi you bump everything but leave sentry?

[gritty wind]

I'm working on that now

Downgrading sentry does indeed resolve the issue

Do we just disable sentry in tests

Changelog btw https://github.com/getsentry/sentry-python/blob/master/CHANGELOG.md

[vale ibex]

currently writing a script that checks every version between 0.20.3 and latest to see where it fails

[gritty wind]

I'd start with 1.9.8 since 1.9.9 does not install in python 3.9

Smells of broken release

[vale ibex]

I'm on 3.10

everything on latest version

[gritty wind]

Yeah it's definitly 1.9.9

1.9.8 works correctly

[vale ibex]

So IG we just pin to 1.9.8 for now

Their django update is marked as ongoing

so maybe it'll be fixed

[gritty wind]

Why does pip just now show anything if I run pip show sentry-sdk

Mildly/highly concerning

Ah wasn't in venv

Ok agreed

[vale ibex]

poetry run pip

ez

[gritty wind]

Literally

Ok PR ready

Ran 285 tests in 18.672s
Cool

[vale ibex]

Ran 285 tests in 10.943s

suck it

[gritty wind]

Homeslice

how do you do that

[vale ibex]

still getting whitenoise warning btw

lol

[gritty wind]

Stickers don't work for me, so I just assume that's the skill issue sticker

Would be a real shame if it loaded and it wasn't that

[vale ibex]

I won't spoil the surprise for you

[gritty wind]

So much so that you might just have to edit your message now before it loads

[gritty wind]

Now this is the real skill issue

Are you running through manage.py?

Or the poetry task

[vale ibex]

poetry task uses manage.py

[gritty wind]

Oh static preview failed

[vale ibex]

so both?

[gritty wind]

Yeah

Ohh yeah the site 500x lol

See #dev-alerts

[vale ibex]

nnice