dev-contrib

[vale ibex]

https://pythondiscord.com/pages/contributing/setting-test-server-and-bot-account/#setting-up-a-bot-account

I'd highly suggest you read through the contributing guide.

[placid ermine]

also https://pythondiscord.com/pages/contributing/sir-lancebot/

[clever wraith]

i will read it. thanks.

[vale ibex]

Yea, that's the one I linked above 😄

scroll up

[placid ermine]

ah by above you mean above

right

[clever wraith]

Like up

thanks a bunch for help.
i invited bot to server. he is in there. tho im getting some error in docker stuff

pywintypes.error: (2, 'CreateFile', 'The system cannot find the file specified.')

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
  File "docker-compose", line 3, in <module>
  File "compose\cli\main.py", line 67, in main
  File "compose\cli\main.py", line 123, in perform_command
  File "compose\cli\command.py", line 69, in project_from_options
  File "compose\cli\command.py", line 132, in get_project
  File "compose\cli\docker_client.py", line 43, in get_client
  File "compose\cli\docker_client.py", line 170, in docker_client
  File "site-packages\docker\api\client.py", line 188, in __init__
  File "site-packages\docker\api\client.py", line 213, in _retrieve_server_version
docker.errors.DockerException: Error while fetching server API version: (2, 'CreateFile', 'The system cannot find the file specified.')
[15168] Failed to execute script docker-compose

[vale ibex]

Do you have docker running?

[clever wraith]

oh my bad

yay!

[vale ibex]

🎉

[clever wraith]

i checked contri.RD file, there is no stuff about api call errors. am i still missing something?

[vale ibex]

What do you mean by api call errors?

[clever wraith]

i tied to run .connect4

and this happened

its running here fine

[gritty wind]

You don't have the emoji on the server you're running it on

[vale ibex]

Ah, that's likely going to be the emoji

[brazen charm]

You'll need to add the emoji or replace it in constants (or env if that was added)

[vale ibex]

As a note, we should probably make that an env var in lancebot

[gritty wind]

The cross emoji?

[clever wraith]

okay i think its having issue with CROSS_EMOJI

yeah

[gritty wind]

Where else do we use it lol

[clever wraith]

4 files. statically used at 2 of them.

assuming thats the name used everywhere

[vale ibex]

It's going to be the incident_unactioned emoji

[gritty wind]

There are actually quite a few cross emojis

[vale ibex]

Yea, some use unicode emoji

[gritty wind]

\u274C, \u274e, incident_unactioned

[clever wraith]

so should i add it somewhere, or statically add it out there?

wtf

[gritty wind]

!pban 786650529690157066 Self bot

[stable mountain]

:incoming_envelope: :ok_hand: applied ban to @warm musk permanently.

[mellow hare]

Okay, turns out I DO have to use the real Redis in order to properly test this PR. Running into this error still: https://paste.pythondiscord.com/kumaloxepa.py

[vale ibex]

It looks like bot is starting before redis is ready?

unless you're starting that seperatly first?

[gritty wind]

I think the implication here was that he has to switch from fake redis to real redis

in which case

that won't matter

Though admittedly it is trying to connect to redis, so I may have misunderstood

[mellow hare]

I started them separately

[clever wraith]

i have put \u274C, its fine right?
tho what can be permanent solN?

[gritty wind]

Just have it as a constant at the top of the file

[clever wraith]

okay so in pull, im supposed ignore these changes?

[gritty wind]

I'm a little confused, what are you doing?

[clever wraith]

future pull

[gritty wind]

I don't understand still. Are you trying to solve conflicts with the master branch?

[mellow hare]

Here are the steps I did for this:

• Tested using fakeredis, realized I needed regular
• Did docker-compose down to spin the containers down
• Changed the config file to have the redis configs be host: "web" and use_fakeredis: false
• Ran docker-compose up -d web redis and waited until it was running.
• Ran docker-compose up bot and encountered the error almost immediately

[clever wraith]

yes. as i have made this change because i was having this issue, is it acceptable if this change comes on master as well?

[gritty wind]

Its hard to resolve the conflict without knowing the actual content of the files haha. Are you chaning which emoji is used?

[clever wraith]

Hi @gritty wind , would you mind quickly going over my pr on sir-lancebot?

[gritty wind]

Sure, I can give it a look over in a bit

[clever wraith]

Thanks

[clever wraith]

yeah

i have that already, just asking

[mellow hare]

Ah, fair enough

[gritty wind]

gah docker builds take forever

[brazen charm]

I run the bot locally and use localhost, but shouldn't the host be redis?

[mellow hare]

I'm just going by what our contrib guide says

I'll give it a quick shot, though

[gritty wind]

but what are you working on that requires it?

[mellow hare]

The stream permissions thing

[tawdry vapor]

The host for the site should be web and the host for redis should be redis

[mellow hare]

Got it, I'll correct it

[gritty wind]

fair enough

[mellow hare]

I'm testing to see if the PR properly removes the role after if it applied the role then had to restart before it could remove it

And I can't do that with fake

Okay, that fixed it. I'm double checking to make sure I just missed something in the contrib guide or if we need to tweak it

[gritty wind]

I think it's not on there, at least not in the config section

[mellow hare]

https://pythondiscord.com/pages/contributing/bot/#configure-the-bot

Yeah it's not

Okay, I don't feel as dumb now

[gritty wind]

Would you mind adding it to the first section

[mellow hare]

I will in a bit, have to use the little nerd's room

[clever wraith]

idea: can we make @stable mountain use fuzzy matching for the docs command?

[brazen charm]

There is an issue for partial matches

[clever wraith]

hmm?

so one question, do i literally need to restart docker on every change?

[brazen charm]

You can reload the extension

[clever wraith]

im trying to change announcement message and its not giving new

how?

[gritty wind]

!reload ext_name

[clever wraith]

wdym partial matches?

[gritty wind]

Though not on sir-lancebot if that's what you're working on

Or

maybe it is

can't remember

.help reload

[dusky shore]

Query "reload" not found.

[brazen charm]

https://github.com/python-discord/bot/issues/447

Should also be in lancebot

[gritty wind]

ah try .ext reload

maybe spelled right lol

[brazen charm]

I'd guess the src doesn't see the root alais

[clever wraith]

okay that worked

thanks

@brazen charm how i'd imagine i would be useful is, you were talking/helping about something like overwrites, and you don't know the exact name of the class/object. you could just !d overwirtes

[sleek steppe]

that would probably make the docs command much slower than it already is

[brazen charm]

It should be possible, but there are considerations on how to achieve it as there are around 100k symbols which could result in many unwanted results with common keywords

[brazen charm]

The matching shouldn't be very intensive, currently it's slow because of an unnecessarily small cache and making requests to pages for every single item it has to fetch, a more robus solution is implemented in bot#1014

[dusky shore]

GitHub

[bot] #1014 Doc improvements

[clever wraith]

where should I put the link to the discussion for pyfacts?

@gritty wind

[gritty wind]

I don't know what the plan is for gathering facts

[vale ibex]

.topic

[dusky shore]

**What is your favourite color?**

Suggest more topics here!

[gritty wind]

Do you know who originally opened the issue?

[clever wraith]

i did

[vale ibex]

I'd suggest linking the discussion Joe setup in the embed footer

[clever wraith]

i can't set a url in the footer

[gritty wind]

that's a good idea too

I think you can use markdown links, no?

[vale ibex]

Yea, you should be able to

[clever wraith]

no, not in footer

https://discordpy.readthedocs.io/en/latest/api.html#discord.Embed.set_footer

[gritty wind]

set_footer("[link text](https://link.com)")

[clever wraith]

hmm

[clever wraith]

nope not working

[vale ibex]

Ah, maybe just add another line at the end then

[clever wraith]

ok

[vale ibex]

.src topic

[dusky shore]

Command: topic

Responds with a random topic to start a conversation.

Source Code

Go to GitHub

bot/exts/evergreen/conversationstarters.py:40

[clever wraith]

how do I reload en extension?

[vale ibex]

.ext r <cog>

or just .reload <cog>

[clever wraith]

ok

huh

AttributeError: 'NoneType' object has no attribute 'cancel'

pretty vague

[green oriole]

That's the aoc

[clever wraith]

aoc

[green oriole]

It is pretty broken

[clever wraith]

oh

[green oriole]

I believe @hardy gorge was looking into this

[vale ibex]

oh wait, I don't think sir lancebot support root aliases, until my PR at least, so it looks like its a copy-paste oversight 😛

or a hidden feature

[clever wraith]

ok

[vale ibex]

This is nothing to do with your issue by the way, just something I noticed with extensions reload

[clever wraith]

what's this mean?

bot/exts/evergreen/pythonfacts.py:24:121: E501 line too long (144 > 120 characters)
bot/exts/evergreen/pythonfacts.py:25:9: E128 continuation line under-indented for visual indent

[vale ibex]

first one is your line is too long. Second is your continuation line isn't indented enough

[clever wraith]

ok

how can I fix it?

[vale ibex]

Separate the long line into different lines

And indent the line to match up with the other indents

[hardy gorge]

Into which part?

[clever wraith]

Markdown doesn't work for footers

[green oriole]

In making sure that the aoc cog was properly unloading and could only get accessed during December

[gritty wind]

yeah apparently not haha

[hardy gorge]

Right

I forgot about that

The changes currently required are fairly minimal

There's one command that requires an event to be ongoing that doesn't check it

And a task has changed names which wasn't properly applied to all references

[clever wraith]

I'd have to suggest making an addition field

riight

why didn't I think of adding a field to the embed?

[green oriole]

@hardy gorge alright, do you still want to take care of it? I can do that this weekend otherwise

[clever wraith]

why can't i get under 120 characters

[vale ibex]

What does your line look like?

[clever wraith]

value='Want to suggest more facts? Suggest more facts here!')

[vale ibex]

.topic

[dusky shore]

**Do you speak a language other than English?**

Suggest more topics here!

[vale ibex]

You could remove the first sentance

It'll be consistent with this then.

Otherwise I'd suggest soemthing like ```py
value=("Want to suggest more facts? "
"Suggest more facts here!"
)

[gritty wind]

@clever wraith honestly don't blame you at all for all the format changes haha

the docs are a pain

[clever wraith]

ok

[clever wraith]

understatement

why is linter so picky

[green oriole]

Because having a consistent and healthy code style is important

[clever wraith]

ok

[hardy gorge]

@green oriole Which command was the broken one again in AoC?

[green oriole]

Hmmm

[hardy gorge]

Not the None error

[green oriole]

.aoc join

[dusky shore]

The Python Discord leaderboard for 2021 is not yet available!

[hardy gorge]

refresh

maybe?

[green oriole]

I can't refresh actually

[hardy gorge]

.aoc leaderboard

that one

[green oriole]

Only admins can

Oh, the leaderboard

[dusky shore]

:x: Unable to fetch leaderboard!

[hardy gorge]

alright

thanks

!

[gritty wind]

I think it's actually on startup

It logs a missing token error

But I may be misremembering

[clever wraith]

@green oriole can you review my PR?

sir-lancebot#596

[dusky shore]

GitHub

[sir-lancebot] #596 Add .pyfacts command

[green oriole]

I can't, I'm not on my computer, sorry

[clever wraith]

id review if I was a contributor 🙃

ok

[gritty wind]

Actually you are still welcome to review

[clever wraith]

oh?

[gritty wind]

it helps other reviewers, even if doesn't count

[clever wraith]

can I review

?

[gritty wind]

You may have seen @short snow's reviews around

[green oriole]

You can, but not your own PR haha

[gritty wind]

It tells other reviewers that you think the PR is good

which is a helpful sign when looking over something

[clever wraith]

ok

@gritty wind can you review? :rooThink:

[green oriole]

We sometimes even merge using a contributor review

[clever wraith]

ok

@gritty wind i made the requested change

it seems the linter did not like

[green oriole]

You have a failed lint

Yep

[gritty wind]

while you're at it

th :earth_americas: fro

extra space there 👀

line 43

[clever wraith]

oop

[gritty wind]

If you're failing on code length

you could extract ?utm_source=sir_lancebot&utm_medium=referral into a variable

[clever wraith]

I failed on whitespace, escaped with slash

that failed lint, oops

[green oriole]

Sounds like something you can store as a constant and use an urllib function to create

I don't mind storing it as a constant though

[clever wraith]

@gritty wind the lint passed

[gritty wind]

Hmm I'm not sure it complies with the style guide

[clever wraith]

the \?

[gritty wind]

nah

it would be name on a new line, and the closing parenthesis on a new line too:

add_field(
  name="Author",
  value=f"Photo by [{username}]({profile}" + rf + ") \
  on [Unsplash](https://unsplash.com?" + rf + ")."
)

[clever wraith]

oh

alright, ill fix

[gritty wind]

Also you're already using f-strings, I think you don't need to do string concat with plus

[clever wraith]

oh im stupid

ill get that fixed!

[gritty wind]

👍

I'm not sure if this is in the style guide or not, so I'm not going to enforce it, but I'd prefer if you avoid the break

you can built the value on another line, it doesn't have to be in the function

author = f"Photo by [{username}]({profile}{rf}) on [Unsplash](https://unsplash.com?{rf})."
add_field(name="Author", value=author)

that took too many edits haha

Alrighty, that works for me

[clever wraith]

awesome

[gritty wind]

I will ask you to fix the extra space in the embed description though because that is user facing

[clever wraith]

I thought I got that!

@gritty wind i fixed the space there

[gritty wind]

thank you

[clever wraith]

no problem, i thought i got that there

i cant believe how much things i messed up doing that!

[gritty wind]

Ok everything looks fine to me

Lol my first contrib was flaming too haha

[clever wraith]

good to know

my first also holds second place for most comments on a pr

[gritty wind]

Gonna spin it up locally to test, but other than that I think it's just getting the API approval now

lol I'll take partial blame for that

[clever wraith]

lmao its fine

!topics

[stable mountain]

Off-topic channels

There are three off-topic channels:
• #ot0-psvm’s-eternal-disapproval
• #ot1-perplexing-regexing
• #ot2-never-nester’s-nightmare

Their names change randomly every 24 hours, but you can always find them under the OFF-TOPIC/GENERAL category in the channel list.

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

[clever wraith]

!topic

[stable mountain]

Off-topic channels

There are three off-topic channels:
• #ot0-psvm’s-eternal-disapproval
• #ot1-perplexing-regexing
• #ot2-never-nester’s-nightmare

Their names change randomly every 24 hours, but you can always find them under the OFF-TOPIC/GENERAL category in the channel list.

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

[clever wraith]

.topic

there

@clever wraith i don't understand your review

its for the styling of embeds

oh, 1, these are facts, not topics

f***

hang on

there

the add_field, not the entire embed

[vale ibex]

@clever wraith congrats! 🎉

[clever wraith]

Thanks!

excited for when its available!

[vale ibex]

Once @dusky shore says its rejoined in #dev-log you should be good to go

should be less than 1 minute

[gritty wind]

👀

[clever wraith]

exciting!!!

[gritty wind]

it's pretty much just waiting for the actions to run

hmm if things break

blame Joe

[clever wraith]

lmao

[green oriole]

Even if it is your fault

blame joe

[clever wraith]

@gritty wind so how's my PR?

[gritty wind]

lol haven't had a chance to look at it, I was doing the earth one

I'll try and get to it before I head to bed

[clever wraith]

oh yeah, you're located in saudi arabia

ok

woohoo!

its on

lol sometimes it gives an image of an egg #sir-lancebot-playground message

[gritty wind]

I got a globe haha

[clever wraith]

something with people categorizing their images on unsplash

[gritty wind]

it is accurate lol

anyways, it turned out quite well

[clever wraith]

im happy with it

[gritty wind]

For future reference

Your PR stated closes issue 147 which didn't actually close the issue

It'd be something like closes 147 or resolves 147

[clever wraith]

@green oriole are you free/available to review my PR?

[green oriole]

I am not until this weekend

[clever wraith]

oh ok

[green oriole]

@clever wraith awesome job, thank you for the PR!

[clever wraith]

no worries

Thanks!

@clever wraith good job!

It was pretty fun!

thnks for the review too

no pron

prob*

[gritty wind]

@clever wraith merging the earth PR caused conflicts on the contants haha

[clever wraith]

haha

[gritty wind]

could you resolve them?

[clever wraith]

pardon?

how?

you mean the constants?

yea

[gritty wind]

You can try and merge master into your branch, which would make git prompt you to resolve

Or you can do it online

https://github.com/python-discord/sir-lancebot/pull/596/conflicts

[clever wraith]

so what do I click?

[gritty wind]

You need to go down to the area highlighted in red

and essentially remvove the >>>>>

Make sure the content of the file is correct, which in this case means adding the extra constant from master

[clever wraith]

sorry that it messed something up!

[gritty wind]

It's not your fault, it shouldn't have

git is just being

[clever wraith]

not your problem

[gritty wind]

well, git

[clever wraith]

phew

git is intersting

[gritty wind]

Oh you commited the arrows themselves

Ok

everything else looks fine besides that

[clever wraith]

whhops

[gritty wind]

haha the github conflicts menu can be confusing

I don't think I've ever actually used it, just used my local setup

[clever wraith]

@clever wraith congrats on the merge!

thanks!

[gritty wind]

that's been changeloged

[clever wraith]

awesome!

congrats

[gritty wind]

hopefully we'll get a full unspash token soon so we can changelog earth

[clever wraith]

it takes 5-10 days is what ive seen

just a reminder that .earth is an alias, i just use it cause its shorter

[gritty wind]

The full command is .earth_image, yeah?

[clever wraith]

i believe so

[patent pivot]

.pyfact

[dusky shore]

Python Facts

If you type import antigravity in the Python REPL, you'll be directed to an xkcd comic about how easy Python is.

Suggestions

Suggest more facts here!

[patent pivot]

nice

is there a link to all the pyfacts

i'm going to clear the comments of the discussion and edit it in

[clever wraith]

.earth_photos @gritty wind

[gritty wind]

hmm it's in a text doc right now

[clever wraith]

oops

[vale ibex]

https://github.com/python-discord/sir-lancebot/blob/master/bot/resources/evergreen/python_facts.txt

[patent pivot]

cheers

discussion updated & comments cleared

[gritty wind]

Alrighty, that's all I got in me for tonight. Grats to both of you, and keep up the good work 🥳

[clever wraith]

Night @gritty wind , and thanks!!!

[short snow]

how much time does pipenv run test take? Earlier when u used error out at RedisCache it took 30 seconds

but now when i fixed it

it is running from 20 minutes

[brazen charm]

Maybe a couple of minutes at worst

[short snow]

Hmmm, even if I re-run it, it takes the same amount of time

this only happened after i added id attribute to all MockMessage instances

......................................................................................................................./usr/lib/python3.8/unittest/mock.py:2030: RuntimeWarning: coroutine 'Infractions.apply_voice_ban.<locals>.action' was never awaited
Coroutine created at (most recent call last)
  File "/usr/lib/python3.8/unittest/async_case.py", line 88, in _callMaybeAsync
    return self._asyncioTestLoop.run_until_complete(fut)
  File "/usr/lib/python3.8/asyncio/base_events.py", line 603, in run_until_complete
    self.run_forever()
  File "/usr/lib/python3.8/asyncio/base_events.py", line 570, in run_forever
    self._run_once()
  File "/usr/lib/python3.8/asyncio/base_events.py", line 1851, in _run_once
    handle._run()
  File "/usr/lib/python3.8/asyncio/events.py", line 81, in _run
    self._context.run(self._callback, *self._args)
  File "/usr/lib/python3.8/unittest/async_case.py", line 102, in _asyncioLoopRunner
    ret = await awaitable
  File "/usr/lib/python3.8/unittest/mock.py", line 1342, in patched
    return await func(*newargs, **newkeywargs)
  File "/home/user/PycharmProjects/bot/tests/bot/exts/moderation/infraction/test_infractions.py", line 134, in test_voice_ban_mod_log_ignore
    self.assertIsNone(await self.cog.apply_voice_ban(self.ctx, self.user, "foobar"))
  File "/home/user/PycharmProjects/bot/bot/decorators.py", line 153, in wrapper
    await func(*args, **kwargs)
  File "/home/user/PycharmProjects/bot/bot/exts/moderation/infraction/infractions.py", line 366, in apply_voice_ban
    await self.apply_infraction(ctx, infraction, user, action())
  setattr(_type, entry, MagicProxy(entry, self))
RuntimeWarning: Enable tracemalloc to get the object allocation traceback
....................................................
``` it displays this and then stops

If add id=123 to line 553 and 533 of test_incidents.py it stops

otherwise it doesn't

[short snow]

give me a ping if anyone knows

[short snow]

https://github.com/python-discord/bot/pull/1446 we need to stop this, the check is running from quite some time now

it is gonna eat up your github action minutes

waitt, nvm, we have unlimittted minutes

but why is this happening

[hardy gorge]

Not sure

looks like it's hanging on the test suite

[short snow]

Yeah, samething happened to me locally, look just above the link of the pr

[short snow]

^

[tawdry vapor]

Find out which tests are the culprits

[short snow]

here are the culprits Mark

[tawdry vapor]

Ok I don't see anything obvious but I can try to look closer tomorrow

[short snow]

Alright thanks

[balmy sparrow]

Right now the docs command group when invoked without subcommand just calls the docs get subcommand right?

[cold moon]

Yes

[hardy gorge]

@short snow It's related to events loops and the RedisSession object

I'm not exactly sure what's going on yet, though

But, I have a work meeting starting in a few minutes

[short snow]

Yeah no hurries, just ping me when u get back so I can be here

[balmy sparrow]

Basically, just not having to type out the entire name

[short snow]

I think you can suggest this idea on the on-going doc improvents

something like !doc search

[balmy sparrow]

was hoping to catch numerlor here. i'd suggested something similar before

[short snow]

bot#1014

[dusky shore]

GitHub

[bot] #1014 Doc improvements

[short snow]

They are here 🙃

[brazen charm]

It has already been suggested, was asked yesterday #dev-contrib message. I'll probably take a look at it after 1014 is merged but there are more things to consider with the amount of inventories present in the bot

[balmy sparrow]

alright 👍

[short snow]

scal is faster than xith

👀

[short snow]

@hardy gorge free now?

[hardy gorge]

No, still working.

The issue is that the setup is creating another event loop while the RedisSession was originally initialised in a different loop.

It may seem like you're creating a new RedisSession here, but a RedisSession is a singleton to make sharing sessions within an application more easy

Another way of handling this would by using the asyncSetUp method of the test class

but it means that it would run before each test method

However, I've tested it, and it doesn't really come with all that much of a performance hit.

[short snow]

Which different loop?

[hardy gorge]

There are two different testing modules that set up RedisSession in this way, right?

[short snow]

Yeah

Oh yes

[clever wraith]

@clever wraith first thing before developing, please do git pull 😄

[short snow]

Just do a reset kn the main branch

and then checkout from there

[clever wraith]

wat

[short snow]

If that’s what u mean

[clever wraith]

ok

[short snow]

So should use asyncSetUp of the test class? and how should i use it. Any parts of the doc i should refer to

[clever wraith]

Aight

[vale ibex]

Do you know how to fetch updates from an upstream origin?

[short snow]

@hardy gorge

[vale ibex]

https://docs.github.com/en/github/collaborating-with-issues-and-pull-requests/configuring-a-remote-for-a-fork
https://docs.github.com/en/github/collaborating-with-issues-and-pull-requests/syncing-a-fork
Could be good if you don't yet 😄

[clever wraith]

Yep

I pulled last night, but there were 3 commits this morning

[short snow]

@vale ibex can u fix the merge conflicts on sir-lancebot#597

[dusky shore]

GitHub

[sir-lancebot] #597 Move pfp commands into executors

[vale ibex]

That should do it

[vale ibex]

@sleek steppe re you comment on the above PR, I've tried adding @commands.max_concurrency(1, wait=False) to the group command pfp_modify, but spamming .pfp_modify prideavatar 5 times still outputs all five

I send all 5 commands before the bot even responds to the first

[clever wraith]

@clever wraith the campaign's should be in yml, just a note

[clever wraith]

hmm?

oh right

ok

we should do reactions

That's what I was thinking, yeah

b/c sending a message every time for the path would be bad

@hardy gorge does the bot have perms to manage reactions?

Yee

.help

The way sir lancebot does the .help command

Campaign's will go in bot/resources/evergreen/adventure_campaigns if that's ok

ok

so, first things first, what's the story line?

My idea was heading into the temple of(insert name here) to steal the holy python idol(just a statue of the python logo) kind of like an Indiana Jones thing

But you can write a different one if you want

I made support for multiple campaigns

@vale ibex what's with the huge number as your status?

[vale ibex]

I'll give you a hint, 0x at the start of a number denotes a hex value

[clever wraith]

That's one long hex value

What kinda color is that

[vale ibex]

Hex isn't just for colours, it's just another way to represent a number

Try and convert it to text

[clever wraith]

oh

@clever wraith if its ok we will change from yaml to json yaml is acting up

[clever wraith]

LMAO

https://youtu.be/dQw4w9WgXcQ

[vale ibex]

😄

[stable mountain]

Hey @clever wraith!

Uh-oh! It looks like your message got zapped by our spam filter. We currently don't allow .txt attachments, so here are some tips to help you travel safely:

• If you attempted to send a message longer than 2000 characters, try shortening your message to fit within the character limit or use a pasting service (see below)

• If you tried to show someone your code, you can use codeblocks
(run !code-blocks in #bot-commands for more information) or use a pasting service like:

https://paste.pythondiscord.com

[clever wraith]

oh

[gritty wind]

Usually the pastebin is the best choice

[clever wraith]

its too long for the pastebin somehow

[gritty wind]

huh, what are you uploading lol

[clever wraith]

https://pastebin.com/K06xzejg

[gritty wind]

I'm going to open that

but if its a rick roll on the other side

aww deleted

[vale ibex]

It was a link to this https://dis.gd/threads /s

[clever wraith]

ok

[clever wraith]

nice idea!

let's get started

can I have write access?

I sent a contribution invite

[gritty wind]

Did you guys get assigned to the issue yet btw?

Ah yes

good luck then :)

[clever wraith]

@clever wraith not yet walking home I'll let you know

[clever wraith]

Sent contrib invite

[clever wraith]

It was a hexadecimal containing the entire bee movie script

[gritty wind]

lmao

[clever wraith]

ok

Pastebin didn't like

@clever wraith i didn't get it, or I don't know how to check

[sleek steppe]

Interesting, it worked differently for me.

[vale ibex]

it might be due to the monkey patching we're doing for root aliases

I'll give it a go on a fresh bot when I get a chance

[sleek steppe]

I'll try testing again

Tried again and it still works the exact same.

[clever wraith]

Check email

[vale ibex]

Odd, I just removed the monkey-patch and it still isn't limiting concurrency

So I'm guessing i'm doing something wrong somewhere

[clever wraith]

go it

Awesome

[vale ibex]

out of interest, are you running it on my PR, or on your own bot?

[sleek steppe]

oh I think I can see the problem because I didn't have invoke_without_command=True

Now when I set it to True the checks don't actually work

[vale ibex]

hmm

I can work with that

So the interesting thing with this, is that there's a top level group, with some commands, but it also contains a subgroup, with some commands itself

the concurrency at the top level group seems to impact the commands against it, but not the sub group

[sleek steppe]

Yeah but when it isn't True it actually runs the code inside the group plus the subcommand invoked.

I think, but let me try it out some more.

Yeah I think a custom check would be more suitable for this

[cold island]

How can I get add an emoji to the code like that? 🤔
https://github.com/python-discord/bot/blob/b34da23b925bcee132b98d5aa5e953ab6b4fe94e/bot/exts/fun/duck_pond.py#L211

[obsidian patio]

If you’re on Mac, there’s a “Text” menu where you can select Unicode characters

On other OSes, I don’t know about the specific tools available. However, you can always copy paste emojis into the code. They behave just like any regular Unicode characters

[cold island]

If I copy an emoji from Discord it will display it as the name between colons

[obsidian patio]

Yeah, discord is weird in that sense

Just put a \ before sending it on discord, and it should show up as the “actual” emoji

😄
\😄

[cold island]

ah, that worked, nice

[obsidian patio]

Perfect 😄

[cold island]

thanks

[obsidian patio]

Happy to help

[fallen patrol]

but only if on pc it gets tough on mobile

[obsidian patio]

Yeah, it’s not always consistent

[short snow]

That was a long PR.

[hardy gorge]

Hey @short snow, how's it going?

I have some time now

I had to finish some work yesterday and It was finished a little bit too late

[short snow]

Gimme sometime, lets say 10 minutes and i will get back to you. Kinda busy with some work

[hardy gorge]

No problem.

[short snow]

Ok am back.

[hardy gorge]

Alright

Instead of what you're currently doing, we can also solve it in a different way

In your TestCase class, you add this:

class ...:
    async def flush(self):
        # Flush everything from the database to prevent carry-overs between tests
        with await self.session.pool as connection:
            await connection.flushall()

    async def asyncSetUp(self):
        self.session = RedisSession(use_fakeredis=True)
        await self.session.connect()
        await self.flush()

    async def asyncTearDown(self):
        await self.session.close()

[balmy sparrow]

you've named both the methods the same?

[short snow]

in TestIncidents?

shouldn't this be added to the setup? loop.run_until_complete(redis_session.connect())

[tough imp]

How will it behave with fakeredis in this case, do you get a fresh / clean cache for each test?

[short snow]

i think so, since when there was no redis session in test_incidents.py it was getting redis intialize something something errors

looks like i need to update the discord message link on my repo

(#community-meta message) is also there

[hardy gorge]

Thanks, oversight

[hardy gorge]

No, you won't get that. I'm thinking of adding a subclass of IsolatedAsyncioTestCase to the package to simplify testing

[tough imp]

That sounds like a good idea

oh to the package itself, even better

so this way the store would persist for the whole testing "session"?

[hardy gorge]

There's already one in there, but it's not packaged with the package

it's only in the test suite that the package itself uses

[short snow]

what if we could add a main class to the tests which would contain all these sessions like Bot class, won't that also simplify stuff?

[short snow]

and these two questions about your code

shouldn't there be a if statement on the tear down function if the session is not initialized

[hardy gorge]

That would be an odd thing to happen, because the set up method should take care of that

[short snow]

and about this and the question below it?

[hardy gorge]

No, we don't have to create and run a loop because we can await connect here

this is already a coroutine that gets run before each test method

[short snow]

oh ok. And this should be added to TestIncidents right?

[hardy gorge]

This is currently a solution that needs to be added to all test classes that rely on an actual (fake)redis backing

[short snow]

oh ok.

need some help with regex here

message_links = DISCORD_MESSAGE_LINK_RE.findall(str(message.content))
``` how i get it

@hardy gorge No class TestCase

u meant commandtestcase right?

[hardy gorge]

No, just the testcase in which you need it at the moment

I don't think we need to do this for every test we have that doesn't rely on Redis

Just in your class where your tests are currently having issues with Redis

[tough imp]

hmm, I think the way those tests are designed is that TestIncidents is a superclass that already provides some setUp methodology to the children

so you can put it there, but then you may be getting a redundant session that you won't use for each test

if you put it in the subclass, make sure that the superclass doesn't also have an asyncSetUp that you may have to call manually (because you'll be overriding it)

[short snow]

test incidents is child class of unittest.IsolatedAsyncioTestCase

oh now i get what u are doing

[tough imp]

yes, but it doesn't have any tests of its own, it's only to provide the setUp to the children

so that they don't have to duplicate

but looking at it, it doesn't have an async setup, so it should be safe to just add it where necessary to the children

in this specific case maybe we could make the redis async setup a mixin, I don't know

[short snow]

well mock_session is a unresolved attribute, not there in the parent class

[hardy gorge]

naming mistake

should be fixed now

[short snow]

time to run the test...

@hardy gorge 🎉 Works like a charm!

welp one problem @hardy gorge

ah thought so, we needed thatn if statement

Uhhh

https://github.com/python-discord/bot/actions/runs/627160537 shouldn't this be ignored?

[cold island]

why should the lint test be ignored?

[short snow]

not the lint test, it should ignore test/* dir ig (it is there in the tox.ini) and asyncSetUp iis the parent class function which we need to over-write so can't really fix it

[cold island]

Looks like the other instance of this function in the codebase has a # noqa: N802

But you should lint before you push smh

[short snow]

i thought it wouldn't be required since it is a change in test/ dir

[cold island]

you should still write good code in tests 🙂

run pipenv run precommit and it just won't let you commit if it fails to lint

[short snow]

Ok PR is up for review! And i have commented on 1/2 of your comments i don't were needed zig

how do u add commit body with git (CLI tool, not gh)

[patent pivot]

git commit -m "title here" -m "body here"

[cold island]

Or just don't write a commit message and it'll prompt you in a text editor

[patent pivot]

very small PR that won't affect many people but will make my day https://github.com/python-discord/bot/pull/1449

[hardy gorge]

Just don't use -m at all, like @cold island mentioned. It will open a text editor for you to write your commit message in.

If you want, you can even configure git to use the text editor you want to use

[short snow]

or best, pycharm git commit tool 😄

[hardy gorge]

That;s what I use

It has a summary line length checker (50 chars) and hard warp commit message body (72 chars) options built in that you can enable

[short snow]

where can i enable them from?

[obsidian patio]

That's built in for my Vim installation as well, which is pretty nice

[hardy gorge]

File -> Settings
Version Control -> Commit -> Commit Message Inspection
Limit Body Line -> 72 chars + wrap when...
Limit Subject Line -> 50 chars

[short snow]

Do u need to approve once again if someone made changes?

[gritty wind]

not really, it still would've counted my approval

But its probably good practice if there have been major changes

Or if you don't approve, you could dismiss or rerequest review

because you technically signed off on the old code, not the new

[short snow]

So in joe’s case there is no need. Since it is not a big change

[gritty wind]

I did it anyways, it doesn't make much of a difference

[short snow]

No need now lol

[gritty wind]

Just more of a: I continue to approve of this

[short snow]

okie.

@vocal wolf sir lancebot repo needs label updates

[vocal wolf]

on the PRs?

[brazen charm]

Depending on the project, I usually go with 72 on the subject too, as 50 is quite limiting and github will display 72 without wrapping

[patent pivot]

yeah I go 72 for the title as well

[patent pivot]

that's a configurable setting

[vocal wolf]

Or do you just mean the labels in general?

[patent pivot]

on some of our repos new changes will dismiss reviews

[vocal wolf]

This spacebar is going to be the end of me lol

gets jammed every sentence

[short snow]

Yes, like on pr changing status needs review to wip, etc.

[gritty wind]

where are you hitting it 👀

[vocal wolf]

on the right

[gritty wind]

do we still use those?

[vocal wolf]

yes

[vocal wolf]

Haven't been active much this week, I now have time to do stuff

ty on update tho

[short snow]

Np!

[vocal wolf]

@cosmic moat please stop playing factorio for a couple seconds and finish up this PR lol

https://github.com/python-discord/sir-lancebot/pull/586

@gritty wind poke https://github.com/python-discord/sir-lancebot/pull/595

[gritty wind]

sir-lancebot#595

[dusky shore]

GitHub

[sir-lancebot] #595 Update Tic Tac Toe Command

[gritty wind]

ah

we never figured out what to do with that PR

it's a style change

I don't know if it works or not

[vocal wolf]

into needs review it goes

[gritty wind]

Also tic tac toe is disabled now, so ¯_(ツ)_/¯

[vocal wolf]

why's that?

[gritty wind]

Blocking things and stuffs

[vocal wolf]

I see

@short snow everything has been updated. Bask in the glory that is, updated labels.

[cosmic moat]

later today or tomorrow™️ i promise

[vocal wolf]

ty

[cosmic moat]

i've been without my dev environment for a couple weeks, which helps me procrastinate

[short snow]

awesome job

@vocal wolf i will tak over sir-lancebot#485

[dusky shore]

GitHub

[sir-lancebot] #485 Pride Leader(pl)

[short snow]

i will take over the rest too but in a while

got one week holiday so free

[exotic ember]

@short snow if you comment on that PR I'll assign you to it

[vocal wolf]

@exotic ember u should assign yourself to the emoji cog pr

(probably, that is)

[short snow]

done

[short snow]

If any coredev is here, need assignment here https://github.com/python-discord/sir-lancebot/issues/446#issuecomment-791545350

(Hello scal)

[gritty wind]

done

[short snow]

Cool thanks

https://github.com/python-discord/bot/issues/1059 Has good first issue and staus:planning, both are contradicting labels ig

[brazen charm]

Why do you think they're contradicting?

[short snow]

Well good first issue is something like a new contributor can start wokring on

but status planning is like it still planning and u can’t work on it

[vale ibex]

status:approved means you can start working on it

good first issue is just a difficulty gauge

[brazen charm]

It can still be a good first issue during planning, as new contribs can contribute to the discussion or push it forward

[short snow]

ah right

[lofty parrot]

hey
so basically the python help channels get their channels name changed
i wanted to know
whats the rate limit
with the discord.py api
who can I talk to eg staff

[gritty wind]

The API isn’t from discord.py, but the point is

It’s not very high

We hit when we were smaller

It’s not viable at all right now

[lofty parrot]

kk

[short snow]

@exotic ember Reviewed!

[short snow]

@vocal wolf reset your branch to the main, and then push your commits to it

will fix the commits from different branch

[vocal wolf]

the thing is there was only one commit, I just needed to remove the unnecessary file. Also I thought the force pushes would combine into one.

I thought I saw that before, apparently I was wrong.

Yeah if I had multiple commits having each file this would be fine, but since there was only one commit, I could not do your suggestion.

I would probably just reset to a hash

and then do whatever

[short snow]

oh ok.

[vocal wolf]

The first force push was me combining all commits into one, and then I noticed the other file lol

was an "oh shit" moment

[short snow]

lol

[tough imp]

I think the contrib guide for bot may be missing a bit on the Redis host

https://github.com/python-discord/bot/blob/8f1ef21bc2e140d39dabb16de8fbbd6077805a25/config-default.yml#L15

I believe this should be set to redis when running with compose locally

[gritty wind]

I think hemlock was going to add that a couple days ago

If he hasn’t added it yet, I can add it in a little bit

[hardy gorge]

We could also make it an environment variable and give it the correct value in the compose

that means people won't have to configure anything

[gritty wind]

Works for me

[left flume]

I have a suggestion for @stable mountain

[cold island]

do tell

[left flume]

Sometimes people test out their code on the server using the eval command. This is sometimes very helpful and some people wish to save the result.
The bot already reacts with a 🗑to delete the message if the user doesn't want it anymore. My proposal is to make the bot react with '💾' as well. If users react with this emoji, the bot 'saves' the message aka DMs them the same message. This might allow users to save the results for the codes that they test out on this server.

[vale ibex]

We have something quite similar within @dusky shore with

.help bm

[dusky shore]

Command Help

.bookmark <target_message> [title=Bookmark]
Can also use: bm, pin

Send the author a link to target_message via DMs.

[vale ibex]

There has been talks in the past about extending this feature though

[left flume]

Not many know about this tho

It's literally on another bot

It isn't very easy to save the message in this way

If the option is provided with delete option users are more likely to use that

That just requires a reaction directly on the message. It doesn't even require a command so users can instantly and easily save their msgs.

This one is for the eval command

Since many users would like to save those msgs

[cold island]

What if the code snippet was written by someone else?

Which is likely to be the case

[left flume]

I'm talking only abt the eval command so for example, lets say I did !e print("HelloWorld") and it sends out a message. I will have an option to delete that message as well as easily react with 💾 to send it directly to my dms.

This is obviously done for the commands that you type

but if someone else uses !eval and you want to save their message we can innovate on this by allowing others to react with 💾 and then that particular message reaches their DMs too. just to make it better, we can also have the bot to react with ✅ and ❎ after another user reacts with 💾 so that the user can confirm if they want to save it or not...

[gritty wind]

I’m alright with the feature if it only allows the original author to get the DM

But I’m very, very, very wary of allowing anyone to do it

DM rate limits are

Not great

And we need DMs for more important things

[left flume]

that was my idea at first as well

it should only be used for your codes

its a much easier way to save your message instantly and its only supposed to be used for the eval command

and after the user reacts to the message with 💾 both the bots and the users reactions are removed so that the user cant misuse the feature

[gritty wind]

At the same time, why not just use sir-lancebot’s book mark commands, and not risk getting rate limited on the more important bot

Right now, I can’t think of anything on @stable mountain that would easily allow a user to rate limit it

[left flume]

how will it become rate limited?

[gritty wind]

You spam a bunch of evals in a bot command channel

Then go back and click on all of them really quickly

You cause the bot to effectively spam you, which discord won’t like

And suddenly, we are unable to do some moderation actions

[left flume]

You know how lancebot deletes all of its reactions after the inactivity timeout?

Python could do the same if the user plans to use the command again

[gritty wind]

We can definitely add safeguards

But

I don’t see why when we can shove it somewhere lower risk

[left flume]

its easily accessible if we include it there

its like running a code and then going and taking a screenshot of it and then saving it instead of just directly clicking on the save button in your IDE

[gritty wind]

Except

We already have another form of save button

So that’s not a fair comparison

I think the book mark command allows you to do it for anyone’s code, and can keep the main bot safer

[left flume]

alright

[short snow]

with dewification the site is basically gonna be static? With no login/bot api,since API is a separate project now?

[obsidian patio]

We’ll still have a django backend, but the site will be static from what I’m aware. The idea is to simplify things a lot

[short snow]

by things what do you mean?

[obsidian patio]

The way it works. Like, we don’t need authentication and a database, considering what we do on the site. New pages be added with PRs instead

[short snow]

is the api going to be connected to the site atleast to manage interactions between the bot and the api?

[gritty wind]

Bot -> API (something.pythondiscord.com) -> Database
User -> Site (pythondiscord.com)

something like that

the important part is that the site and API will be different projects

[patent pivot]

yeah

django site will just be a glorified static site generator

[cold island]

@short snow do you still want bot#1423 ?

[dusky shore]

GitHub

[bot] #1423 Fix tp list command

[short snow]

i can do it, but after 1 week or so, since am assigned to ~3 more issues

[short snow]

why not just use something like hugo then?

[patent pivot]

Ultimately the tech stack used will fall to the media team. Using Django still makes some sense since we can tightly control how content is displayed, for example loading in YAML and rendering the resources page from that.

We are using Hugo elsewhere in our stack so if it goes well it's something to think about, but I'm not sure Hugo is powerful enough for pythondiscord.com

We might at some point look into a Python-based solution that outputs static HTML because then we can do things like deploy previews which is a real nice way of working.

However, we've got several large rewrites on our hands at the moment, so I don't think we should complicate it when it can be avoided.

[green oriole]

Hot swapping statics

[cold island]

That's no problem. I was asking because I'm working on a related issue and can do it simultaneously

[short snow]

oh alright

[short snow]

If u wanna take it, go ahead!

[cold island]

👀 not done yet

I might need to split the talentpool into two files

[short snow]

You can split talentpool into helper files

like the helpchannel

[green oriole]

oh god

Plz split it

[timid sentinel]

when you're reviewing some code on github does it not let you add a comment to parts of code that haven't been changed? It doesn't seem to be letting me

[short snow]

Nope

[timid sentinel]

that's seems pretty stupid ngl

[short snow]

But you can get the permalink of that line and then make a comment

That’s what i do

[timid sentinel]

cool, thanks

[green oriole]

I am not 100% sure why, but my client let me do that

But I think I'm the exception

For example

[timid sentinel]

It seems to let me go three lines up or three lines down from an edit, but I want to make a comment 4 lines down 😔

[gritty wind]

You'd need to use an external editor

[green oriole]

oh interesting

[short snow]

Uhh it doesn’t let me

[short snow]

Hmm, pycharm doesn’t give me the + sign on non changed lines

[gritty wind]

Don't know what to tell you

Are you on pro?

[short snow]

Yeah

[gritty wind]

It also won't let you do expanded sections like github, but it's still a little larger

[short snow]

when u review a pr, u get the + sign and then u can comment

but in non changed sections it doesn”t

[timid sentinel]

Quick question, I've tested the talentpool bot PR with the site one, and my only comments were really going to be about the style of the !tp search command results, i.e. it currently looks like this which I feel is a bit messy. Do you think changes to this would be in the scope of the PR, or would you want to do it along with fixing the tp list command?

[green oriole]

What does the PR currently change?

[timid sentinel]

it's this, https://github.com/python-discord/bot/pull/1445, so basically allowing multiple nominations at once

I guess it is fairly relevant, as the the changes from the PR are what has made it not that neat

my thoughts were a few small things like

• The "Nomination ID" for inactive nominations should be after status and date like with active infractions.
• "Actor", "Created", "Reason" would be a better order than "Actor", "Reason", "Created"
  but I don't know if it might be worth messing around with some bigger changes to the formatting to neaten it up

[obsidian patio]

I personally think getting the functionality should be the highest priority. If it’s easy to fix the styling, that would be great. Otherwise, I’d rather see it in a follow up PR

[green oriole]

Those are really small changes, I don't mind either way or even a master commit

[timid sentinel]

Makes sense, thanks. I'll approve the PR, mention them as optional changes, and I can always open a PR later if I want to change anything else.

[cold island]

I think we can make something work. @glass pecan has been working on polishing the infraction embeds, so maybe we can draw inspiration from that

[glass pecan]

yeah that embed is pree ugly, but your primary changes sound important too, so I'd focus on getting your primary stuff pushed first and have the formatting done later

it's defs a thing to do though, so might be worth adding as an issue ticket at the least so it's not forgotten

[cold island]

Yeah I'll keep the changes lean to not block the pr, and then we can bikeshed an exact ui

[clever wraith]

@patent pivot how do you feel about a tag for the humanize library?

[patent pivot]

not sure what it is

[vocal wolf]

!pypi humanize

[stable mountain]

humanize v3.2.0

Author

Jason Moiron

Requires Python

=3.6

Summary

Python humanize utilities

License

MIT

[vocal wolf]

@patent pivot it's like arrow.humanize() but for more stuff

[patent pivot]

ah

[clever wraith]

a library that's pretty useful for converting stuff into human redable text

[celest charm]

Hey, I found this thing https://www.gitpod.io/
Maybe we could set it up for lancebot or something like that?

seems like there's a free option for public repos

TL;DR it's a remote editor that sets up environment for you

[gritty wind]

There was an issue opened yesterday for providing support for it

I'm just unsure of adding editor specific support because

it's opening the floodgates slightly

[cold moon]

I use GitPod for my own repos. Can't use for PyDis stuff, because not enough permissions.

[cold island]

What's the max characters in a message a bot can send, 2000?

[obsidian patio]

Yeah, I think that’s it for regular messages

Embeds have some special limits as well

These are the exact limits for embeds https://discord.com/developers/docs/resources/channel#embed-limits

[clever wraith]

@patent pivot what do u think of the tag idea?

[patent pivot]

fine by me

[clever wraith]

ok

[cold island]

Thanks

Do we have a function or something for splitting text across multiple messages and posting it?

[obsidian patio]

We have a paginator, if that’s what you’re looking for, like in the help command

[cold island]

Hmm I'm looking for something without an embed, but I'll check how the split works there

[obsidian patio]

Alright. I don’t know if we have something for splitting it up, that doesn’t use an embed. Sounds like a good idea to check the source code

[cold island]

yeah, thanks for the suggestion

textwrap.wrap() seems promising

[clever wraith]

I think that was ne

Me*

[obsidian patio]

Yeah, and I’m getting more and more convinced that we should have a hidden gems video on Textwrap

[vocal prairie]

That's not a bad idea.

[vale ibex]

Yea it's so useful 😀

Learned about it myself via a lancebot pr I reviewed recently 😅

[short snow]

How does reddit bot automatically post the embeds to other channels, i added announcement option to my dev-log channel on a test server and it didn't happen. Like i get share with followers option everytime i need to do it manually

[subtle kraken]

https://discordpy.readthedocs.io/en/latest/api.html#discord.Message.publish you probably are looking for this

[short snow]

hmm, what if i want to do it the github webhook?

like the #dev-log channel

[subtle kraken]

webhook you don't send to?

still same method

just look for the message in on_message and figure out if it is in the channel, is the message you want to publish etc @short snow

[short snow]

oh alright. thanks

[tawdry vapor]

I need some help investigating some relatively-advanced Docker stuff

I ran into a roadblock and spent hours investigating it but couldn't find an explanation to the behaviour I was observing.

[vocal wolf]

Is it possible to store objects within the Redis cache?

Such as a message or channel object

[tawdry vapor]

No. You have to serialise them.

[short snow]

they are only locked to certain data types

like the id of them the message in your case can be stored

[vocal wolf]

right

[tawdry vapor]

Here's the problem: I have two separate run commands, but for some reason, the second command prevents the cached layer for the first command from being used.

# Pipenv installs to the default user site since PIP_USER is set.
RUN pipenv install --deploy --system

# Install numpy when in dev mode; one of the unit tests needs it.
RUN if [ -n "${DEV}" ]; \
    then \
        pipenv install --deploy --system --dev \
        && PYTHONUSERBASE=/snekbox/user_base pip install numpy~=1.19; \
    fi

Where DEV is a build arg passed to Docker. If I build with it and then without it (or vice versa), it will always run the first command again instead of using the layer from the previous build.

There isn't enough documentation on how exactly buildx determines whether a layer is valid or not. I thought it just checked if the command was the same.

I tried using a tool to inspect the layers and they came up as 0 bytes on this specific step for some reason??? I don't really remember. If anyone has any leads on this, let me know.

[hardy gorge]

I have no idea, but I can discuss it with someone from work who knows way more about docker than I do. I'm just not sure how soon I'll be able to talk to them.

[tawdry vapor]

Thanks, I appreciate that. It's not urgent, considering I've sat on this for like a month already.

[vocal wolf]

What would be the best way of serializing a discord.Message object for storage within a Redis cache?

pickle?

[tawdry vapor]

They aren't pickleable IIRC

But the question is, why must you do this?

[vocal wolf]

Let me get the context for this

[tawdry vapor]

Speaking of context, for my Docker issue, here is the build output https://paste.pythondiscord.com/raw/eyajuvoton and here are the commands I'm running https://paste.pythondiscord.com/raw/dehulolito with this dockerfile https://paste.pythondiscord.com/raw/tujalawawo

[hardy gorge]

.bm #dev-contrib message

[tawdry vapor]

Notice how some parts of venv are cached, but as soon as it reaches step 3, it refuses to keep using it despite the previous build creating a layer for it.

[hardy gorge]

@tawdry vapor Did you try moving the ARG declaration to between the two RUN commands?

https://docs.docker.com/engine/reference/builder/#impact-on-build-caching

ARG variables are not persisted into the built image as ENV variables are. However, ARG variables do impact the build cache in similar ways. If a Dockerfile defines an ARG variable whose value is different from a previous build, then a “cache miss” occurs upon its first usage, not its definition.** In particular, all RUN instructions following an ARG instruction use the ARG variable implicitly (as an environment variable),** thus can cause a cache miss. All predefined ARG variables are exempt from caching unless there is a matching ARG statement in the Dockerfile.

[tawdry vapor]

oh shit

Man that's such a huge gotcha

fuck that

I'm gonna try it

[hardy gorge]

I was looking at your file and that was my guess, because Docker can't know if a command run using RUN uses an environment variable

this seems to back it up

but I can't test it here on my work machine

[tawdry vapor]

Yeah, I forgot they are implicitly environment variables too.

Thanks, I knew I needed a 2nd pair of eyes on this.

Waiting for the build...

Yes, that fixed it.

Good catch

[vocal wolf]

Ignore what's going on in #dev-log, my branch is fine lol

[tawdry vapor]

Hey are you still gonna provide context or did you figure it out?

[vocal wolf]

I will be providing context, I just got sidetracked with squashing

@tawdry vapor the current PR that I'm working on is https://github.com/python-discord/bot/pull/1414. It adds a message at the bottom of #❓｜how-to-get-help, showing the currently available help channels. kwzrd suggested that I add Redis caching so I don't have to set the variables every time on startup, since they'd instead be ready for me in the cache.

The current problem I'm facing is here:
https://github.com/python-discord/bot/blob/71ea5a6c9e6f3f4c09ee1b4b26fafb8e31e9ffbf/bot/exts/help_channels/_cog.py#L506-L511

Since self.how_to_get_help will still be an ID after the first block, the .history method will throw AttributeError. I thought it may just be better to instead replacing some API calls with putting the objects directly into the cache instead.

[hardy gorge]

Nice, glad I could be that second pair of eyes.

@vocal wolf If you know the content, the channel ID, and the message ID, you should be able to edit the message directly

The API call only requires this information: https://discord.com/developers/docs/resources/channel#edit-message

[vocal wolf]

That's what's happening here:

https://github.com/python-discord/bot/blob/71ea5a6c9e6f3f4c09ee1b4b26fafb8e31e9ffbf/bot/exts/help_channels/_cog.py#L517

oh wait, channel ID

[hardy gorge]

No, you're using a message object to edit the message there by locating it

If you're going to keep track of the message ID of the last message posted by the bot (but think about what you're going to do if this cache is somehow stale or incorrect) and the channel ID (which is basically what you're doing if you were to cache an entire message object), you can just store the channel ID + message ID.

And then use the underlying API call to edit the message.

[short snow]

you can set the channel id, and then get the channel id while editing the message

[vocal wolf]

self.bot.http.edit_message(channel_id, message_id, content=fresh_content)

[hardy gorge]

You don't need an entire message object as a wrapper

[vocal wolf]

I completely looked over this in kwzrd's review, sorry about that

I see what I must do, thank you @hardy gorge!

Wow this simplifies everything by a great amount

Redis caching is something else

[short snow]

yeah, it is really good

how about adding this to .pl

[short snow]

and this

[vocal wolf]

Thank you, @hardy gorge and @tough imp for helping me with this PR, especially with the implementation of the Redis cache.

'Tis no longer a draft, and can be seen here: https://github.com/python-discord/bot/pull/1414

[short snow]

bot_1       | ModuleNotFoundError: No module named 'pythonjsonlogger'

uhh even when i did pipenv install

[vocal wolf]

Happens to me as well. pipenv --rm then pipenv install only sometimes resolves it.

[short snow]

Courtesy Notice: Pipenv found itself running within a virtual environment, so it will automatically use that environment, instead of creating its own for any project. You can set PIPENV_IGNORE_VIRTUALENVS=1 to force pipenv to ignore that environment and create its own instead. You can set PIPENV_VERBOSITY=-1 to suppress this warning.

You are attempting to remove a virtualenv that Pipenv did not create. Aborting.
Aborted!

did pipenv install pythonjsonlogger as a temporary solution

uhhh

Error:  An error occurred while installing pythonjsonlogger!
Error text: Looking in indexes: https://pypi.python.org/simple

ERROR: Could not find a version that satisfies the requirement pythonjsonlogger
ERROR: No matching distribution found for pythonjsonlogger
✘ Installation Failed 

[exotic ember]

!pypi python-json-logger

[stable mountain]

python-json-logger v2.0.1

Author

Zakaria Zajac

Requires Python

=3.4

Summary

A python library adding a json log formatter

License

BSD

[exotic ember]

you shouldn't have to install manually though

[short snow]

doesn't work with install

[exotic ember]

what virtualenv are you using? pipenv is saying you're in a virtual environment it did not create

[short snow]

am using pipenv, venv made by pycharm when u slect that, and i am inside (bot)

[exotic ember]

are you sure Pycharm is using pipenv when creating the virtual environment? you should be able to check it in the project settings

[short snow]

it doesn't seem to work with docker, even after installing manually

i can screen share in dms to show you if u want

[exotic ember]

sure

[short snow]

yes, it shows on bottom right corner Pipenv (bot)

[vale ibex]

try docker-compose up --build

[short snow]

worked!

kosa helped m out in dms

[vocal wolf]

noice!

[vale ibex]

🎉

[short snow]

(did what you had suggested)

@vocal wolf looks like there is one problem

[brazen charm]

It shouldn't be in the same category

[short snow]

oh yes, my bad

whoops, kosa already mentioned one change

[glass pecan]

@short snow hey, you mentioned you tested emoji cog, correct?

[short snow]

yes

[glass pecan]

would you be able to have a quick squiz via mobile client on the output of a certain command for me?

[short snow]

ok

one minute

[glass pecan]

cheers. it's when you do !emoji count asd

and if it outputs the invalid embed

it should show a list of valid categories

[short snow]

Oh, i have alread tested that with jasonnn instead of asd

Yeah it works

it shows the list of categories

[glass pecan]

yep. does it render correct on the mobile client?

[short snow]

not sure about that, do u have a mobile rn? i can quickly send u a invite

[glass pecan]

i do

and sure

feel free to dm me

confirmed a bug but i'm thinking it shouldn't block the PR tbh, especially as the bug is hardly noticeable + it's applicable across many of our command outputs

i'll make an issue ticket specifically for the bug to be addressed as a separate thing

[short snow]

lol @obsidian patio

#dev-log message

[obsidian patio]

What's up?

[short snow]

it is dismissed nvm

i thought u commented

and then it got merged the same moment

[obsidian patio]

Ah, no

[vocal wolf]

I have 3 of my own PRs that need to be reviewed, all relatively small except the last one:
https://github.com/python-discord/bot/pull/1453 - Purge ban message update
https://github.com/python-discord/bot/pull/1452 - PyPi message deletion on error
https://github.com/python-discord/bot/pull/1414 - Dynamic #❓｜how-to-get-help message w/ Redis caching.

[short snow]

Exams are ending tomorow, would try to do them.

did u mention the changes in 1414?

[vocal wolf]

Things were changed, yes.

[short snow]

Am using https://en.wikipedia.org/api/rest_v1/page/random/summary to get random wikipedia articles, how should i give the question. Currently i use description but it is not sufficient to get the correct. and extract contains the answers in a lot of places

my idea is to use extract.replace(title, ""), is that fine?

[cosmic moat]

@vocal wolf my computer should be good again today, if i have time, i'll finish it tonight

i left a message for fix last saturday

i don't know if they read it yet

i'll check that when i work on this

[left flume]

I have a suggestion

You can add an embed visualizer command

so that people can see how embeds look like

Title, description, colour etc. can all be separated using a |

[green oriole]

We do have an !embed command, but we limit it to mods to avoid users faking to be the bot

[left flume]

oh...

okay

[subtle kraken]

yeah I imagine main use people would have from it is to try and break our rules with it by things such as advertising their servers and what not

there already are way better UI/UX wise tools online for that either way

[left flume]

I see

in that case, why do you guys have the eval command?

[subtle kraken]

for evaluation of python code

[left flume]

There are many IDE's that users can use instead, even on web

[subtle kraken]

not for just making bot send plain text/embed

it is useful for help sessions

to show expected/result of a code snippet

[left flume]

Ye makes sense

[green oriole]

It is easier in this case to have the output directly in Discord

Also we do run the filters on the snekbox output

[left flume]

I see

[subtle kraken]

theres also is a big output difference between eval and embed

!embed most people might not even notice who invoked the command, especially if it was deleted

[stable mountain]

most people might not even notice who invoked the command, especially if it was deleted

[subtle kraken]

it can easily be taken out of context

[left flume]

cool

hmm

Maybe you can add a text

Add a text just like how you do for the eval command

@User Here is your embed

anything that goes against the rules would be moderated I guess

[subtle kraken]

generally no bots have an open-to-everyone command that allows people to send text/embeds via command

other thing being, it isn't really a thing that'd be much of a use outside of projects such as discord.py bot

[left flume]

oh right

I understand

[gritty wind]

Honestly there are ways to get around most of these issues, but whatever we make will just be a discounted version of the embed visualizer sites

We probably won’t give it full customization like color, title, fields etc, because it’ll be clunky to use over IRC

[left flume]

why dont you add an embed colour to all of @stable mountain's embeds?

[short snow]

cc @obsidian patio as the orginal author

[obsidian patio]

Adding some (maybe blurple) default color would definitely be nice, but I think that would be best implemented using something like bot#1348, so we keep things DRY

[dusky shore]

GitHub

[bot] #1348 Embed helper

GitHub

[bot] #1348 Embed helper