dev-contrib

[patent pivot]

nah that's my one with a new ext

it is a bit temperamental due to the nature of frontend web apps and backend redirects working together

but for the most part for links posted in discord it should work

(that you did not create)

[fallen patrol]

hm

[clever wraith]

What happened here???

[vale ibex]

Alright I've given the help system PR another round of testing and fixed a small bug in our original design, once I fix connection issues I'll push my commits and head off for the night

[sleek steppe]

yes it goes like .github owner/repo

or should it be .github owner repo

[fallen patrol]

i mean

it'd be nice to have the option of both imo

[sleek steppe]

not sure if that could work

I mean it's possible but that would require a converter

with regex

[sleek steppe]

I pushed btw (sir-lancebot#632)

[dusky shore]

GitHub

[sir-lancebot] #632 Make .github a group, and add github repo + user commands

[short snow]

just use ,join() on it, and then allow it to be handled by

if repo.count('/') != 1:

might wanna open a issue on that?

[vocal prairie]

They were discussing it earlier.

Internally, and then they moved it here.

I'll wait to see if joe, ks123, or kutiekaj9 have any more thoughts on it.

[brazen charm]

A discussion here will amount to nothing if it's forgotten; they can continue to discuss on the issue if you open one to make sure it's not lost in conversation

[short snow]

They could comment there thoughts on the issue

so we can have everything together

or it would get lost in the history of the chat

[vocal prairie]

Ok, I'll do it when I wake up tomorrow.

[short snow]

are other users allowed to contribute to dewification project?

[exotic ember]

yes, although there's not much left to do

what do you have in mind?

[short snow]

was willing to contribute if we have any issues open 😄

for site#366 can we use elasticsearch or haystack?

[dusky shore]

GitHub

[site] #366 Dewikification: Search function

[cold moon]

I'm not sure we want to attach external stuff to site after dewikification.

[short snow]

so just a normal search? using fuzzywuzzy?

[cold moon]

We don't have very specific specs yet, but this can come later, after all content apps is done

[short snow]

shouldn't site#449 be part of the meta of a wiki.

Like:

title: How to Contribute a Page
description: Learn how to write and publish a page to this website.
+level: beginner
icon_class: fas
icon: fa-info
relevant_links:
    Contributing to Site: https://pythondiscord.com/pages/contributing/site/
    Using Git: https://pythondiscord.com/pages/contributing/working-with-git/

[dusky shore]

GitHub

[site] #449 Dewikification: Group resources by experience level

[exotic ember]

the experience levels are for resources, not pages

it could fit in site#421, but I'd rather a separate PR after #421 is merged so we don't introduce needless complexity to what is already a big PR

[dusky shore]

GitHub

[site] #421 Dewikification - Display resources based on YAML files

[short snow]

https://pythondiscord.com/pages/resources/ right

[exotic ember]

yeah

[short snow]

in what cases should a dataclass be used, and in what cases should a enum be used?

context: https://github.com/python-discord/bot/blob/8a58d04c277265f2009e46d6fcc9a85f6ddd5896/bot/exts/backend/branding/_cog.py#L23

[exotic ember]

in simple terms, dataclasses are classes that store data, enums are classes that store constants

dataclasses are just 'wrappers' over regular classes, that let you define classes mainly serve to contain data easily without a lot of the boilerplate of classes

enums are a lot more widely known and implemented in programming languages, and are essentially groups of constants that other code can reference to ensure they are well, constant

[brazen charm]

with enums you also usually don't care about what the actual values are

[exotic ember]

elaborating on Num, the values are "banner" and "icon" in this example:

class AssetType(Enum):
    BANNER = "banner"
    ICON = "icon"

in fact in many other languages, enum variants do not need values

enum AssetType {
    Banner,
    Icon,
}

[short snow]

ok, but aren't banner and icon data here?

not constants

[exotic ember]

they are constants, what makes you think they are data?

any function that takes in an AssetType can only take in AssetType.BANNER or AssetType.ICON, anything else can be assumed invalid (I don't say is invalid here because this is Python and you can do whatever you want, a language with stricter type system will error)

[short snow]

haven't gone through the whole code yet, but are they storing the icon/banner for a event

@green oriole is the description better now?

do we follow american or british english?

[exotic ember]

it does not, it is just a constant for other functions to state whether the file at download_url is a banner or an icon when applying it to the guild

[green oriole]

Yep, it is better

And either english is fine honestly

[short snow]

yeah, went through the full code and understood what you are saying now

ok akars

[tough imp]

Is there any desire amongst the maintainers to enforce a more consistent style for log messages & comments, or does it not matter? Some modules prefer to use proper sentences ended with full stops, whereas others use short statements without much punctuation. Is one preferred over the other, or does consistency only matter at the level of individual modules? Logs from the entire project are displayed together after all

The contrib guidelines don't seem to have an opinion

[green oriole]

We haven’t really talked about logging in a long time. I’d say that anything above info should have proper sentences as they will trigger Sentry issue, I don’t mind having shorter statement for others, especially for debug or trace as there may be a lot of them

[tough imp]

Mmm that's a good point with Sentry

I'm planning to do one more documentation & logging revision pass for my PR so I'm asking in case there's a preference to use full sentences I may as well update the code before I submit it for review

I always find it difficult to write good log messages; it's hard to consider the entire amount of contexts in which they may be viewed

[green oriole]

It sure is. Balancing log levels can also be hard, we still only see info and above in production. Is it the PR about the new branding manager?

[tough imp]

Yes, but I'm also wondering in a more general sense

For example, I'm facing a dilemma of how to handle the situation where some inner utility function fails to get some resource (i.e. file from github), and returns None as a result. In my code, it will log a warning, which semantically makes sense to me (it's warning that it failed to acquire the resource and that this may cause trouble for the caller), but then the caller will make a decision based on the return being None and log an error

It is the caller's decision to classify the situation as an error, so it makes sense to me for that log to be there also

However, if I understand Sentry correctly, this will create 2 tickets, which is not ideal

But this is more relating to log levels rather than messages

[download_file]: WARNING: I failed to download the file! This is not good, but I don't know what it means for the caller!
[apply_banner]: ERROR: Download failed, we cannot apply a new banner!

Perhaps reducing the warning to an info log would make sense in this case

[cold moon]

What about logging both of them as warning and send manually alert to #dev-alerts with ping of some role?

[green oriole]

Yeah, it will make two tickets, although I’m not sure if it is an issue

Maybe the first one can be an info indeed

[tough imp]

I don't think that's necessary, it's really just a warning, not a critical error that would require immediate attention

In fact, the reason why it logs a warning in the first place is that it cannot ascertain whether it needs developer attention or not

If GitHub decides to respond with a 500, we probably cannot do much about it anyway

But if it suddenly gives a 403, a dev may need to investigate

[green oriole]

That’s a good point, although having a Sentry issue will still he easier for us to investigate, I think we would rather have too many issues than missing some

[tough imp]

If the first log is info, and the second log is an error, is it trivial for you to see the first log in Sentry? It shows something like breadcrumbs right?

Because the first log would contain the status code, but not trigger a ticket

[green oriole]

Yeah, it does show the log history

[tough imp]

Mm, then I may just info log it yeah, and let the caller decide whether the situation is an error or not

[green oriole]

Like so

I think info would make sense here

[tough imp]

Awesome, I'll update it

[green oriole]

Cool

[brazen charm]

Got a wrong return typehint in the example on https://pythondiscord.com/pages/contributing/style-guide/

[green oriole]

... I didn’t know this page was a thing

Where did you find it?

[brazen charm]

someone mentioned it here iirc

it is linked in the general contributing page

[green oriole]

Hah, that's really nice

The return type hint should be float, shouldn't it?

!e print(type(1/2))

[stable mountain]

@green oriole :white_check_mark: Your eval job has completed with return code 0.

<class 'float'>

[green oriole]

Alright, I am not going crazy

Thanks a lot num

[hardy gorge]

Arguably, the type hint is too specific in this case

It would probably be better to use numbers.Complex for both parameters and the return value

but I'd imagine that would confuse a lot of people

[brazen charm]

I believe there is a number abc that would be a bit clearer, but yeah I don't think that's worth doing for an example like that

[tough imp]

Fyi I decided on a completely different approach - instead of returning None, the download utility can raise on non-200 responses with an informative message, and the caller is responsible for handling that in whatever way appropriate, i.e. error-logging the exc message

This way the status propagates out of the utility and can be logged alongside the actual error

I think this is the best approach

[green oriole]

Sounds like a good idea

[vocal gale]

Thoughts on giving the Python bot a stackoverflow scraper? Either in a command, or in response to a question asked in a help channel ("While you wait, here are some resources which could prove useful: https:.//..."). The API allows 10 000 requests for a given app as far as I know, so if the command has a cooldown it shouldn't be a problem. I'd be glad to pr this, but i'm on the fence about its usefulness. On one hand, many questions here can be answered by a stackoverflow search, but on the other hand it may be a lot of hassle when one could just open ones browser and get the information much easier...

[gritty wind]

I think we already have an issue/PR open for it

But here’s a fair warning

When we last discussed it, we:

1. Didn’t want it in help channels or python-general
2. Didn’t want on python, but sir lance

[vocal gale]

Ah, ok! Took a quick search through the GH issues, didn't find it.

[gritty wind]

No worries, I’m looking through myself to find where that issue went

It may have been decided it isn’t worth the effort

But I’ll look into it more

[vocal gale]

Yeah, understandable. A google search is really not that much effort!

[thorny obsidian]

It was Lancebot: https://github.com/python-discord/sir-lancebot/issues/579

[gritty wind]

Nice, nice

[gritty wind]

It’s rather: if we’re not going to allow it in help channels, and channels where people look for help, that diminishes a lot of its potential for good. That being said, more features are always welcome, and if the author is fine with the feature not getting as much use as it otherwise could, then they are welcome to contribute it

The reason we wanted this away from help channels is explained in the issue, but basically we wanted to avoid people dumping links to stack overflow without knowing the content, or helping the person asking the question

[vocal prairie]

That would also be kind of a thing against the CoC.

Because we try not to say "google it".

It's the same situation in a way.

[vocal gale]

Yeah, "Read the _ manual" behavior is the opposite of useful

On a side note, anyone know is there a specific reason the !user command doesen't have a "trash" reaction? I'm assuming the idea of !user is that non-mods can't get others' information, but leaving the embed open allows others to see ones information. Am I missing something here?

[sleek steppe]

If they're seeing their infractions they probably do it in the staff channels

[vocal gale]

Not the mods, the regular users have !user too, but only for themselves.

[sleek steppe]

Mhm

[vocal gale]

Stuff like date joined, creation date....

[sleek steppe]

There isn't too much info that you can't find yourself

Besides joined at

And infractions

[vocal gale]

True, but some people might want to delete their joindate from the chat anyway, instead of having it there all the time.

[vale ibex]

Join date it public information anyway

For example, my user ID is 126811506632294400

and decoding that snowflake gives the create date

!sf 126811506632294400

[stable mountain]

Snowflake: 126811506632294400

Created at 2015-12-16 22:23:39.225000 (5 years, 3 months and 9 days ago).

[eager fable]

That's a really old user 😄

[gritty wind]

Valid points, all of them, but I wouldn't mind adding the reaction

I think it should be pretty simple with the tools we have available

[vale ibex]

Yea 😄 #bot-commands message

Yea, it's minimal effort to do, and I think it could be good for sure

[short snow]

There is a issue open by vestergukan on this

It is on bot, and is on planning state

[vale ibex]

@gritty wind What do you think about using redis for the .bm issue?

use that to cache the bm messages and have a reaction add listener

[gritty wind]

we can do that, or... just have a regular listener for a few minutes?

Adding redis here feels like adding a lot of unnecessary complexity

[sleek steppe]

Yeah the wait_for is probably better

[gritty wind]

The major downside with a wait_for is you probably shouldn't have it running forever

which means that feature will only work for x period of time after the original command

but I don't mind that tbh

[vale ibex]

We could have it active for 5 min and then clear all reactions

Maybe even delete the embed entirely

[fallen patrol]

ok i understand why pypi info was removed

but it would be nice to add a bit of the info back in the footer

[gritty wind]

cc @vocal wolf with pypi being your baby and all 👀

[vale ibex]

That could be useful, and pretty un-intrusive in screen real estate

[vocal wolf]

oh god I've changed it like 3 times now lol

What kind of info are you suggesting to be placed at the footer?

[fallen patrol]

i was thinking the author

[vocal wolf]

hmmmm

not sure

[vocal prairie]

The Pypi picture takes up more space than the text.

Fitting another line it shouldn't be too intrusive, and wouldn't make the embed longer.

[vocal wolf]

!pypi numpy

[stable mountain]

numpy v1.20.1

NumPy is the fundamental package for array computing with Python.

[vocal prairie]

Actually, maybe not.

[vocal wolf]

for some packages, maybe

[vocal prairie]

Depends on the description length.

[vocal wolf]

ye

[green oriole]

!pypi painful

[stable mountain]

painful v6.6.6

pain pain pain pain pain pain pain pain pain pain pain pain pain

[green oriole]

I thought it was longer

[vocal wolf]

it's a summary so I think it's limited to some length by pypi

it would be problematic if we had the description field instead (from json)

[gritty wind]

I do have some really long ones from when I was doing testing

let me see if I can dig any up

[green oriole]

I don't think it is actually limited

[gritty wind]

!pypi neuraxle-tensorflow

[stable mountain]

neuraxle-tensorflow v0.1.2

TensorFlow steps, savers, and utilities for Neuraxle. Neuraxle is a Machine Learning (ML) library for building neat pipelines, providing the right abstractions to both ease research, development, and deployment of your ML applications.

[gritty wind]

for names there's

!pypi aws-solutions-constructs.aws-dynamodb-stream-lambda-elasticsearch-kibana

[stable mountain]

aws-solutions-constructs.aws-dynamodb-stream-lambda-elasticsearch-kibana v1.93.0

CDK Constructs for Amazon Dynamodb stream to AWS Lambda to AWS Elasticsearch with Kibana integration

[vocal prairie]

Oh, those are long.

I wish you could put something under the thumbnail.

[vocal wolf]

aws likes long names lol

[sleek steppe]

why is that a thing lol

[vocal prairie]

Just name it "All things aws.py"

[gritty wind]

if you visit the user that publishes these packages

there 10 or 20 that look like that

[fallen patrol]

what do you mean?

do you mean the mouse over text?

because that would be nice to be used 👀

[balmy sparrow]

how would you add another library to the !d ?

[gritty wind]

!help docs add

[stable mountain]

Query "docs add" not found.

Did you mean:
docs
docs delete
docs set
docs get

[gritty wind]

!help docs set

[stable mountain]

Command Help

!docs set <package_name> <base_url> <inventory_url>
Can also use: docs s

*Adds a new documentation metadata object to the site's database.

The database will update the object, should an existing item with the specified package_name already exist.

Example:
!docs set python https://docs.python.org/3/ https://docs.python.org/3/objects.inv*

[gritty wind]

We need the url for the actual docs objects

like the one in that embed

[sleek steppe]

sadly if you wanted the docs for set you would have to do !docs get set

[balmy sparrow]

right

[gritty wind]

lol

[balmy sparrow]

when do you decide to add another one?

[gritty wind]

Feel free to suggest one here or in community meta

[balmy sparrow]

having pymongo on there would be useful I think, but i dont know if im the only one who's linking to it

[gritty wind]

I don't think there are any hard guidelines

[sleek steppe]

    @github_group.command(name='repository', aliases=('repo',))
    async def github_repo_info(self, ctx: commands.Context, *repo: str) -> None:
        """Docstring here"""
        repo = '/'.join(repo)

Is this fine?

works fine

and this is the error message

[sleek steppe]

yay now it needs one approval from a core dev

[vale ibex]

@tawdry vapor think I've found the cause of the datetime oddity

>>> datetime.utcnow()
datetime.datetime(2021, 3, 26, 22, 8, 47, 441603)
>>> a = datetime.utcnow().timestamp()
1616821624.207364
>>> a = datetime.utcfromtimestamp(a)
datetime.datetime(2021, 3, 27, 5, 7, 4, 207364)

when my computer clock is utc-7

using just datetime.fromtimestamp(a) works as expected

I'm guessing utcfromtimestamp interprets the timestamp as aware, in local tz, and compensates

Possibly want to check the other places that use utcfromtimestamp() just to be safe

[cold island]

that is..

so silly

Well, I guess there's a warning

[vale ibex]

Hah serves me right for not reading the docs

[short snow]

@sleek steppe Looks good!

[short snow]

is there a way to manage the split area of the while reviewing a PR

like reducing the red area here

[sleek steppe]

For bot#1136, would something like a subclassed Context work?

[dusky shore]

GitHub

[bot] #1136 Add trash option to all bot responses

[short snow]

there is a issue for a embed helper command, if all embeds use that, we don't need to subclass context

but yeah i think there are the only two possibilities

well subclassing seems a better option, since one can miss using the embed helper

on the other hand, am still not convinced with adding trash option to all messages

[fallen patrol]

fwiw it makes sense on a lot of them

because ugh some of them suck and are spammy and never go away

whereas honestly, it'd be nice for some of the error messages to not autodelete

[short snow]

or the embeds sent in help channels

we can add can_be_deleted argument or something like that tho

[sleek steppe]

That makes sense. I'm thinking something like this

embed = GetEmbed(...)
await embed.send(messagable)

[fallen patrol]

like, for the most part, if a package isn't found, I personally don't want my own invocation to be deleted

[short snow]

Enums are like constants, so is it good to keep put capital cased variable names inside enum classes?

[tawdry vapor]

But discord.py uses utcfromtimestamp does that means its datetimes are not in UTC too?

[fallen patrol]

Is the bot emoji server public?

Or are all of the emojis on githib

[gritty wind]

no on the first one, and I'm pretty sure no on the second

What do you need?

[fallen patrol]

None of them, just wanted to see all of them 😓

[gritty wind]

We should probably just put all the emojis somewhere at some point

[fallen patrol]

That'd be a good idea

[gritty wind]

If you have the id

You can just pull them straight from the CDN

https://cdn.discordapp.com/emojis/711475530016489512

https://cdn.discordapp.com/emojis/711475530016489512

[fallen patrol]

bro what

that's very cool

hmmmmm

@gritty wind apparently

webhooks can also use every single emoji on discord

…that's old news

#reddit does that

[gritty wind]

yeah and bots can get pretty much nitro emojis :x

[fallen patrol]

but they must share the server

webhooks and channel topics have totally free range

[fringe river]

why are all these scripts so big

[sleek steppe]

It's a large project

[strong ether]

hi

[sleek steppe]

https://github.com/python-discord/sir-lancebot/pull/642#discussion_r602813409 merge conflicts 😬 😬

[keen valley]

Hey

[short snow]

Is there an example of issue being moved permanently from one PR to another, like I know them, but I dunno the original issue number.

[exotic ember]

looking through concrete examples, most of the issues moved are from public to private pydis repositories.

I will amend the issue to treat 301s as 404s.

[patent pivot]

!remind 10M fire pr from f1re

[stable mountain]

You're the boss!

Your reminder will arrive in 10 minutes!

ID: 1813, Due: 2021-03-28T10:24:52

[short snow]

I wanted to see what is the json got on 301s

[patent pivot]

I can test shortly

!remind 10M this

[stable mountain]

You got it!

Your reminder will arrive in 10 minutes!

ID: 1814, Due: 2021-03-28T10:28:23

[exotic ember]

{
  "message": "Moved Permanently",
  "url": "",
  "documentation_url": "https://docs.github.com/v3/#http-redirects"
}

[patent pivot]

aha

[stable mountain]

@patent pivot

It has arrived!

Here's your reminder: fire pr from f1re.
[Jump back to when you created the reminder](#dev-contrib message)

[tough imp]

Does anyone know if this also allows maintainers to push new branches in my fork? The tooltip seems to imply that not

But I'd like that maybe I'd just have to add them as contributors

Although it warns that they 'potentially' could gain access to other branches

what does that even mean

[stable mountain]

@patent pivot

It has arrived!

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

[patent pivot]

I think it's only the PR branch

[tough imp]

okay, I'll just manually give reviewers access if they want

or I guess they could fork my fork too

[short snow]

Right, so how to get the url of the moved issue. Like if it is moved from bot to Lance, how to get the issue number on Lance

[exotic ember]

@short snow that is no longer required, see this message I sent earlier

[vocal prairie]

Question. Was it intended that https://git.pydis.com/blackbox still redirects to https://github.com/lemonsaurus/blackbox even though that was moved to Lemon's personal GitHub? Is it because it still relates to PyDis operations?

[gritty wind]

This isn't something we configured

It's something github does by default

if a resource is moved, it's old URL will auto redirect to the new URL until a new resource is added to the old url

[patent pivot]

@sullen phoenix review left 👍 https://github.com/python-discord/bot/pull/1486#pullrequestreview-622741233

[safe gorge]

is there any reason this message was not pinned in the help channel
<#help-burrito message>

[vale ibex]

Looks like it lines up with when @stable mountain was restarting, see #dev-log

[safe gorge]

Matches up

[fallen patrol]

Bro

https://github.com/lemonsaurus/blackbox#quick-start

Blacbox

[sleek steppe]

blacbox --init

[fallen patrol]

Not blackbox

[fallen patrol]

Shouldn't the bot redeploy with a reconnect rather than a reinit? This would mean that when updated it could get its previous history

[tawdry vapor]

The ci environment contains a gh api token. If you execute code checked out from a branch to which others have write permissions, the token could be leaked. If someone has the api token, then they can use the api to write to other branches I guess.

The leak happens via malicious code that would read the token and send it to the attacker's server, for example

[tough imp]

Oh right, so it explains the causality there. Maintainers could edit the workflow and through that access other branches, makes sense.

I was struggling to understand exactly what it's trying to tell me

[tawdry vapor]

Yeah it's actually really easy I think, compared to the other way around (fork getting access to main)

[tough imp]

I find it really difficult to think of everything that could go wrong with the workflows and secret access

I was talking to Sebastiaan about it earlier because I had some doubts, but he explained that the token's scope actually changes depending on where the workflow is being triggered from

[tawdry vapor]

I don't know that much about it

[sullen phoenix]

cheers, will address them soon

[patent pivot]

contrib/core dev review on https://github.com/python-discord/bot/pull/1486 would be neato 🙂

[balmy sparrow]

Guessing the lack of the space is a mobile bug?

[vale ibex]

!src server

[stable mountain]

Command: server

Returns an embed full of server information.

Source Code

Go to GitHub

bot/exts/info/information.py:152

[vale ibex]

Yea, it looks like a mobile-only issue

seems like the \ and the lack of {features} being populated is making it miss the newline char?

[sullen phoenix]

thanks @patent pivot, @vale ibex, and @cold moon for the quick reviews

[sleek steppe]

oof that review came right after the merge

[brazen charm]

was a bit too late but submitted my review although not likely, handling for the None paste_link should be added or it'll result in an exception if it does fail

[sullen phoenix]

riiight, i can open another PR later today to address those reviews

better late than never!

[sleek steppe]

bot#1471 needs approval from kosayoda, and sir-lancebot#632 needs an approval from a core developer 🙂

[dusky shore]

GitHub

[bot] #1471 Remove reactions from everyone when paginating and waiting for trashcan reaction.
[sir-lancebot] #632 Make .github a group, and add github repo + user commands

[vocal wolf]

@clever wraith Are you planning on finishing this PR off?

https://github.com/python-discord/bot/pull/1410

[clever wraith]

sorry, i was afk, i'll work on it a bit

ok, it's waaay behind

[vocal wolf]

ty on the update

[clever wraith]

how do I pull from the repo and still have my changes

[sleek steppe]

git pull upstream main

You'll still have your changes. If there are merge conflicts then you won't lose your changes, you'll just need to resolve them.

[tawdry vapor]

Don't do that

You just create unnecessary merge commits

There are currently no conflicts with master. Your PR is pure text content so there's no benefit for testing to have the latest changes

[sleek steppe]

For bot#1485, should I just add allowed_mentions, or should I filter codeblock escapes as well?

[dusky shore]

GitHub

[bot] #1485 Filter bypass in !raw command, can make the bot @mention

[clever wraith]

is there an other character looks like `?

maybe we could use that?

[fallen patrol]

so to be clear,

[sleek steppe]

I'm using an invisible character.

[fallen patrol]

@stable mountain no longer relays dms to staff.

[vale ibex]

Not automatically, no

[clever wraith]

what does it look like?

[vale ibex]

but we are able to get it on request if needed.

[fallen patrol]

okay that's kind of being a dick, python

[sleek steppe]

​

[clever wraith]

oh

[patent pivot]

yeah, we changed it so if required we can pull DMs into a text file, but otherwise don't relay live

[fallen patrol]

well, it relays them to the pastebin, no?

[patent pivot]

not live

no

but yes pastebin/text file

[fallen patrol]

pls add that to the privacy policy now that dms are able to be pulled off of the platform

👀

as that was the reason it didn't need to be in the privacy policy last time i was concerned about this

but now that it may be pulled to a pastebin/stored off of discord, that needs to be properly disclosed, no?

[vale ibex]

it's already there.

[patent pivot]

https://www.notion.so/pythondiscord/Privacy-Policy-4ceb50b84d164541bd5b0b7046aa7e6a#ea6b615c140144e9bfe84c7ac779d092

[subtle crater]

can someone ping me im testing smth

[fallen patrol]

👀

why does it refer to a pastebin

[sleek steppe]

These are the changes I've made:

# bot\exts\info\information.py
# line 450-452
            # replace backticks as it breaks out of code blocks. An invisble character seemed to be the most reasonable solution.
            # we hope it's not close to 2000
            paginator.add_line(content.replace('`', '`\u200b'))

# line 471
            await ctx.send(page, allowed_mentions=AllowedMentions.none())

[fallen patrol]

yes im giving you a hard time

[patent pivot]

autocorrect, amended.

[fallen patrol]

every where else in the privacy policy it refers to it as hastebin

cool

damn its an instant update

[patent pivot]

lol

yea

notion is markdown as a service

[sleek steppe]

are these changes fine?

[sharp timber]

lol

[patent pivot]

I think so, yeah

[vale ibex]

Would appreciate people's opinions on https://github.com/python-discord/sir-lancebot/issues/341#issuecomment-808939964

[sleek steppe]

Looks good 👍

[vocal prairie]

Lgtm

[vocal wolf]

@sleek steppe https://github.com/python-discord/bot/pull/1488#commitcomment-48815459

You're able to edit comments with the three little dots to the right of the post.

[sleek steppe]

I made a mistake on the commit message

[vocal wolf]

ah

[clever wraith]

.bm 825809428870856746 bm issue

[vocal wolf]

https://github.com/python-discord/sir-lancebot/pull/644

thing

[vale ibex]

Is this blank line supposed to be here? https://github.com/python-discord/sir-lancebot/pull/644/files#diff-95c311f77020e1a02cceeb127aaf97a7ee6d297bce44caa245f94c0dab4db894R26

[vocal wolf]

n o p e

just noticed that lmfao

give me a second

[vale ibex]

smh

[patent pivot]

yeah you can answer one of the designated topics

or

the void

[vale ibex]

[vocal wolf]

aaaaaaaand it's gone

you saw nothing

[vale ibex]

i am welcoming bikeshedding on this message

[cold island]

@vale ibex can you re-approve?

[vale ibex]

@clever wraith Glad you agree, already PR'd it 😄 sir-lancebot#645

[dusky shore]

GitHub

[sir-lancebot] #645 Bookmark command allows other users to bookmark the same message

[clever wraith]

@vale ibex I'll leave a review in a bit!

[vale ibex]

No rush 🙂

[fallen patrol]

hey if you're adding that what about reply support?

aka message.reference support

[cold island]

hmmmm.. do we want a separate message? you could potentially keep the reaction on the original message, and then DM whoever presses it

And auto-remove the user's reaction

[fallen patrol]

👀 why tho?

imho i hate bots that do that~~

[patent pivot]

why not?

[cold island]

we do that with our paginators.. keeps things neat

[patent pivot]

I'd prefer it remove the reactions, similar to how our paginator does things

yeah

[fallen patrol]

what about removing all of the user reactions when it closes?

[cold island]

It's the same amount of API requests

[patent pivot]

there is a clear all route

[vale ibex]

The reason for the embed was mostly for discoverability

[cold island]

yeah but in this case you avoid the bot's reaction

[patent pivot]

right, but when it ends then all reactions are removed

[fallen patrol]

but after 5 minutes if its inactive it is nice to remove the dead reaction

[patent pivot]

so whatever we do we'll trigger a remove all at the end of 5 minutes

[cold island]

I guess that's true

[patent pivot]

but I also am fairly partial to removing user reactions

it is just neater

[vale ibex]

This embed is deleted entirely after the timeout too

[patent pivot]

noice

[fallen patrol]

if you do that but keep them there before they're all done, it would be nice since this is something that multiple would like to get, so keeping the other reactions kinda lets people know that a message was bookmarked

[cold island]

hmm.. it still seems quite bulky to me, but I'm not incredibly passionate about not having it

[patent pivot]

just feels like an inconsistency to me

[fallen patrol]

because of the nature of what this is, sending a person's msg to dms...
keeping the user reactions there to see who reactions for just a few minutes helps some people be less paranoid

[patent pivot]

we keep reactions where the counts are important, like in polls, and remove when not, like in paginators

this isn't important imo

[patent pivot]

eh, i mean, you remember all the message IDs you send? lol

bear in mind .bm works in DMs lol

[cold island]

I'm not sure I get the argument, you see the result of the reaction in your DM's

[fallen patrol]

yes of course

[fallen patrol]

👀

[vale ibex]

I'm heading to bed anyways, can you add your thoughts to the PR after this discussion is seen through?

[cold island]

yeah

[fallen patrol]

oh dang

[patent pivot]

lol

[fallen patrol]

okay anyways imma get back to figuring out how to get rid of these gitshit errors srsly if someone can help i'll be in #tools-and-devops

[patent pivot]

uh

that's a p simple one iirc

[fallen patrol]

I cannot connect the repo to github

[patent pivot]

you need to do uhhh

[fallen patrol]

go to #tools-and-devops

[clever wraith]

does sir lancebot has a math quiz command

idk whag to contribute so i wanna make like a command that gives you math problems to solve

[short snow]

bump

[sleek steppe]

bump

[exotic ember]

!remind 4h bot#1471

[dusky shore]

GitHub

[bot] #1471 Remove reactions from everyone when paginating and waiting for trashcan reaction.

[stable mountain]

I got you.

Your reminder will arrive in 4 hours!

ID: 1816, Due: 2021-03-29T06:47:45

[short snow]

how we want to handle bookmarks which are embed?
cc @vale ibex

currently the content is none, so it is blank

[short snow]

site#465 is basically part of the staff pywek project (aaaaAAAA)

[dusky shore]

GitHub

[site] #465 Build-Your-Own Ducky!

[short snow]

https://github.com/python-discord/aaaaAAAA/tree/main/img/duck-builder

[gritty wind]

That was created before we had a team lol

And it probably won't be done before the competition tbh

[stable mountain]

@exotic ember

It has arrived!

Here's your reminder: bot#1471.
[Jump back to when you created the reminder](#dev-contrib message)

[exotic ember]

wtf it's been 4 hours already?

[short snow]

@tough imp What do you think about using python3.9 in the branding repo?

[green oriole]

Well we can reuse the assets but not the code

[tough imp]

It should use the same Python that bot uses

[short snow]

why?

[green oriole]

3.8 then because the library devs sucks 😔

[short snow]

yeah ik.

[gritty wind]

I have news 👀

not now tho

[tough imp]

Because it verifies that bot will be able to load the meta.md files, so it should use the same Python and the same dependency versions

[green oriole]

That’s nice

[short snow]

right, ok.

In order to/for -> to/for, what do you think about this, since in order to/for becomes too wordy.
You seem to use this alot in your doc-strings and explainations

and about using colorma or rich to give colors to the [PASS] and [FAIL] 😛

[tough imp]

sorry, I need to attend a few calls now

I'll respond laater

[short snow]

no worries!

[tough imp]

I don't feel about this in any way, we can use the shorter expression

[tough imp]

It would be nice, but I also think that we should keep a very minimal set of dependencies there

unless the contribution makes drastic changes, it will probably only fail 1 event, so the output shouldn't be difficult to read

currently the reqs.txt only declare dependencies that are absolutely necessary for it to function, I see some value in that

[short snow]

i think there is a way to do it without those, using some special characters string

[green oriole]

You can but it won’t be portable

On some terminal emulators on Windows (such as cmd.exe) it will look like garbage

[short snow]

hmm

[vale ibex]

We don't currently do anything with it, so my PR doesn't change that. Maybe we should raise an issue to discuss it

[green oriole]

@short snow the image is broken lol

[short snow]

fixed it

tenor gifs go to the page not gif 😦

[green oriole]

Noice

[short snow]

maybe if we just add rich to dependencies, won't cuz much harm 🙃

[gritty wind]

Really doesn’t sound like a very useful feature

[storm breach]

Hi guys! Tell me where I can find a link to snakebox on github? (with commits for 2020-2021)

[storm breach]

https://github.com/python-discord/snekbox

[clever wraith]

@vale ibex for the .bm, could you make it so it accepts a reference too?

[vale ibex]

wydm by reference?

[clever wraith]

like this

[vale ibex]

Ahh right, yea that could probably be done. I'd rather keep it for a separate issue though

Do you feel like raising it? 😄

[cold island]

That seems pretty spammy tbh

why not just react?

[balmy sparrow]

The reaction is to bookmark a message that someone already bookmarked right?

[balmy sparrow]

I think they mean to use this as the..first bookmark if I'm explaining that right

[cold island]

So any reply to a message will cause the bot to DM it to you? not sure I follow

[green oriole]

The suggestion is that for example if I do

[green oriole]

.bm

It will bookmark your last message

[cold island]

ah

I guess that's not bad

although it would be annoying to get pinged for it

[green oriole]

I was about to say that

[fallen patrol]

.bm just. Turn. Off. The. Mention.

[fallen patrol]

Mentions can be turned off

Although I uh may have forgotten to turn off that mention intentionally

i obviously

forgot

[sleek steppe]

still didn't review

Also would it be cool for !raw to take a message reference to?

[patent pivot]

I don't think we need reply support for bookmarks or raw

as mentioned it relies on people to turn off pings

which is not the default

[green oriole]

Well, the default is on and I doubt most people disable them

[fallen patrol]

fwiw almost always its a bookmark within the conversation

[patent pivot]

Hm

[fallen patrol]

so like, one person replies to it making it easier to everyone see what the bookmark is referencing

[patent pivot]

what's that based on

[fallen patrol]

.bm

[patent pivot]

you don't see every bookmark in the server 😄

[fallen patrol]

i'll put it this way

replies only work in the same channel
if its not the current conversation, its much less likely to reply to an old message with .bm rather than using the link in #bot-commands like the bot makes you.

i rarely see people reply to messages from old conversations with a message that doesn't relate, and i feel that the ping wouldn't be too annoying

if it annoys people enough it could be removed, but i doubt it will

[patent pivot]

I just think that this isn't a great use case for replies, especially if we migrate to slash commands

[fallen patrol]

you intend to migrate?

ngl slash commands do not have the amount of greatness that the current bot system has

arguably i wouldn't migrate

[green oriole]

Maybe one day if d.py has a solid integration

Right now it isn't ready at all

[fallen patrol]

and the api is better

[thorny obsidian]

I think it'd just be nice that if I emoji-react to an existing .bm it'll also send it to me. That way we don't have half a dozen people calling the same command on the same message

[fallen patrol]

yknow

[green oriole]

It is part of Chris' PR

[fallen patrol]

hm

[green oriole]

at least I think?

[patent pivot]

I think for some of our simple commands slash commands work very well

[fallen patrol]

instead of the replies then, add a link to the message in the response to people can see what they're gonna bookmark since an ID doesn't give much
or quote the message

[fallen patrol]

some, perhaps. Def not all

[patent pivot]

any examples?

I'm trying to think of bot command functionality which wouldn't work out of the box

[fallen patrol]

fwiw, the current system is all or none

[patent pivot]

it isn't, we can migrate components as necessary

[fallen patrol]

so all moderation commands would be nada

no i mean

the slash api is users can see every registered command or none

[patent pivot]

that's being added

[fallen patrol]

as of now, they haven't changed it yet

[patent pivot]

they have said they are adding it

[sleek steppe]

coming (discord.py 2.0)

[green oriole]

I'd totally be for a /note command

Like, that's the only command that I would really like to have

Because you wouldn't need to switch to a mod channel to note

[patent pivot]

https://github.com/discord/discord-api-docs/issues/2315

I've had a peek through and can't find much that isn't supported with slash commands

So I'm fairly confident that as the feature matures and some QoL stuff comes such as permission locking then we should migrate

[green oriole]

I agree

I hope that the UI will be better though

Because we have a good 30 or 40 commands on sir lance

[patent pivot]

yeah

[green oriole]

How can we count how many top level commands we have

Nevermind, there is no int e in lance

[fallen patrol]

limit is 100 fwiw

[fallen patrol]

the biggest concern is

every response is webhook api

so allowed mentions is free range

a large bot didn't fuck that up or anything resulting in thousands of pings for people

[green oriole]

It was a bug on discord's end iirc

[fallen patrol]

nope

its allowed mentions

[patent pivot]

no

that was a bug

[fallen patrol]

no

[fallen patrol]

its allowed mentions

[patent pivot]

right, why is that an issue

we just disable it

[green oriole]

You are too fast Joe

[patent pivot]

lol

[green oriole]

But allowed mentions didn't apply

[patent pivot]

but yeah I mean... @stable mountain is currently free range with mentions hahaha

we just... disable them

[green oriole]

It can ping mods, admins and owners, right?

I really wonder why it can ping owners haha

[patent pivot]

we figured if people found a way then we'd like to know lol

[green oriole]

Hah

[clever wraith]

@patent pivot will @stable mountain or @dusky shore ever have slash commands?

[patent pivot]

Yeah, was just saying above how we'd likely eventually migrate

There are some features that need to be added still, but yes that is probably a direction we'll go down

[fallen patrol]

ohhhhh cool https://docs.github.com/en/developers/overview/secret-scanning

^ if someone remembers how I used to talk about the regex for bot tokens, this is how discord bot tokens are revoked when they're uploaded.

[tawdry vapor]

That is a general description of the process GitHub uses. It doesn't go into details of Discord's token format.

[fallen patrol]

no ik

but it does mean that discord has an endpoint that could be used to deactivate posted tokens

perhaps something on this page https://discord.com/developers/docs/topics/oauth2#shared-resources-oauth2-urls

[tawdry vapor]

Sorry, I misunderstood your original point.

[fallen patrol]

basically if one was to upload a token to github it gets invalidated by discord. But rather than use github, I wonder if the https://discord.com/api/oauth2/token/revoke url would actually work for this

[gritty wind]

Oauth isn’t related to the bots afaik, but if you have the endpoint, you can test

[clever wraith]

hello!

[gritty wind]

Hi

So, all our projects are available at github.com/python-discord

But the most frequently contributed to are @dusky shore, @stable mountain, and our website. We have a page describing how to get started with each of the projects here: https://pythondiscord.com/pages/contributing/

Sir-lancebot is specifically targeted at complete beginners for the most part. It's a good starting off point for open source contribution.

The main selling point of sir-lancebot is you can contribute a feature, become familiar with an open source process, then see your work here on the server

All of the features available in sir-lancebot were contributed by someone from the community or staff team

If you're interested in any of these projects, have a look through the issues tab on the github repo (https://github.com/python-discord/sir-lancebot/issues for example), and leave a comment

You may also want to keep this somewhere handy, it's a long list of open source projects organized by language
https://github.com/MunGell/awesome-for-beginners#python

[clever wraith]

Awesome!

Looks like i have some reading to do 🙂

[gritty wind]

Yeah I ended up monologuing a bit, but feel free to ask here if you're unsure about anything

[clever wraith]

will definitely do, and thanks for the intro 🙂

[fallen patrol]

mmm, pygit2

with a healthy dose of pain

[gritty wind]

Some of these are pretty good

I've been meaning to check out pytest for a while now

they recently got a discord server, which is great

[fallen patrol]

fwiw using pygit2 isn't pain

installing it is

[fallen patrol]

pygit2 isn't bad, just to install it i had to go through 3 different webpages for install instructions for a dependency

lets just say it was recursive hell

pygit2 needs libgit2 which needs cmake and each one had more exposition than explantion

[gritty wind]

I really hate how dependent some things are on other libraries. I'm by no means innocent, but I've been in dependency hell for a while trying to upgrade a package to 3.9

There simply are waaaaaay too many abandoned packages that probably won't be upgraded in our lifetimes

[fallen patrol]

dependency hell is fun!

out of curiosity why won't some packages just run on 3.9?

[gritty wind]

All of the ones I have trouble with do run fine on 3.9

and I got the project running locally

thing is, they have no published wheels for 3.9

so I need buildtools

[fallen patrol]

ok, why won't tensorflow then

[gritty wind]

Tensforflow is huge

[fallen patrol]

pfft, you're talking to a guy who built 3.9 from scratch and then realized that pyenv would just do it nicely

[gritty wind]

they probably do some super crazy stuff that's directly dependent on python specifications, which can change each version

or maybe they're also in dep hell

[fallen patrol]

ah

[gritty wind]

Ah yeah, it's the latter lol

https://github.com/tensorflow/tensorflow/issues/44485

4K issues

ouch

[fallen patrol]

smh

[gritty wind]

Nov 17, 2020

I feel that so hard lol

[fallen patrol]

anyways, imma head to sleep

[gritty wind]

Catch you later

[fallen patrol]

I made progress today, I implemented the logic of a .src command 😄

[gritty wind]

👀 👀 👀 👀 nice

[fallen patrol]

but as you might see, a bit of a twist because its a dev bot, so the branch might switch

using pygit2 for that which i THANKFULLY had experience installing when i was self hosting a different bot

so it was actually already installed

[fallen patrol]

fwiw i actually only looked at the src of @stable mountain to see the Path usage, I actually wrote everything myself

I wouldn't steal code

[gritty wind]

Lol it's MIT licensed

[fallen patrol]

without including a license obviously

[gritty wind]

You can use it

but there is a lot to learn by doing things

[fallen patrol]

hehe if you check my admin cog i stole that verbatim from a different bot

granted the purpose for that was, I don't wanna spend time writing a load cog command

or an eval command

not when learning to write those without them already would be kinda hard

so I just copied it so I would have a framework to use

[fallen patrol]

The cog I borrowed from a different bot is mozilla 2.0 licensed so I have no idea what that really means

[keen valley]

Hey

[short snow]

Hello!

[keen valley]

There is a issue in Sirlance

Its bit harder to understand

[gritty wind]

Can you describe it or demonstrate it in some way?

Do you want to work on that issue?

[keen valley]

yes

[gritty wind]

It's already assigned and has a PR open

[keen valley]

But for that what should i do

[gritty wind]

There isn't much left to do for that issue. It has already been reviewed, and now it's waiting for the author to address the review

If you want to review it, feel free to

But the code is already there

[keen valley]

so did Any issue that I can work on it

[gritty wind]

If you're looking for an issue to work on, this will filter the search to approved issues that aren't assigned
https://github.com/python-discord/sir-lancebot/issues?q=is%3Aopen+is%3Aissue+label%3A"status%3A+approved"+no%3Aassignee

The ones labeled good first issue are usually the best starting point

[keen valley]

https://github.com/python-discord/sir-lancebot/issues/643
This one or what

[gritty wind]

If you'd like to work on that one, it is available

just leave a comment on the issue

[keen valley]

Okay

[short snow]

what is "that" here?

[gritty wind]

It was deleted lol

@keen valley if you want to work on that issue, you have to leave a comment first

Otherwise we end up with miscommunications

It doesn't have to be anything crazy, just: "I want to work on this"

[keen valley]

Defiantly I will

If I take some time to solve that issue then it will not a big issue I guess @gritty wind

[gritty wind]

I think it should be fine

[short snow]

#dev-log message discord just removed it xd

[vocal wolf]

@turbid cedar Recruitment isn't allowed on this server.

[short snow]

looks like xith is back on duty, poking ppl 🙃

[vocal wolf]

I've been back for a while

[patent pivot]

do we have a policy on full stops in logs

[short snow]

#dev-contrib message 🤷

[sullen phoenix]

not afaik

[patent pivot]

full sentences isn't full stops

I don't think we do it, and if we do we probably shouldn't

commas, however, is of course fine

[short snow]

uhhh it is not needed imo, but commas can be be good to break sentences while reading the logs

[patent pivot]

commas are necessary in some places, full stops aren't necessary in most places

[short snow]

ok well, that upon part was from grammarly, and I wasn't sure about it, so i added probably @green oriole

[green oriole]

Sure, no problem

[tough imp]

I added full stops to log messages in that module, so for the sake of local consistency that line should have it too

but I'm unsure because overall the project isn't consistent

[patent pivot]

hmm

[tough imp]

some modules have them, some dont, some mix

[patent pivot]

lol

I think we should have a logging revamp at some point

[short snow]

there is a issue open on that, on sir lance

[patent pivot]

we're having discussions internally on stuff like trace logging, maybe setting out some grammar stuff could go alongside that

[tough imp]

there's a similar inconsistency for comments

[patent pivot]

yeah

[tough imp]

I used to never end comments (proper comments, not docstrings) in full stops

but sometimes you want to write something longer

and then you either have to write an awkwardly long sentence, or introduce inconsistency

so for this PR I tried something new with the full stops

but of course if maintainers lean one way or the other, I'll comply

last time I asked I had the feeling that there's no consensus

[patent pivot]

yeah, we don't have consensus

if we do then we'll make that clear in guidelines somewhere

[tough imp]

the way I see it is that no full stops may even encourage writing better, more expressive logs and comments

and that was my preferred style for a long while

but I also come across situations where you just cannot push everything into a single sentence

and I'd rather have stops everywhere than inconsistency

[short snow]

Well, in my opinion I don't think full stops are necessary in logs.

but yes for comments

[short snow]

@tough imp #dev-log message I mean noun here.

At mobile so can’t comment

[tough imp]

it doesn't matter, 'to ensure that we provide embed' is wrong, the article should be there

were you using some kind of software for the grammar?

the sentence can definitely be improved, but not by removing the article

[short snow]

Grammarly

[tawdry vapor]

@vale ibex I think the timezones are still wrong in the PR. For example, there's a case where utcnow(), a naïve UTC datetime is being added with fromtimestamp(), a naïve local datetime (cause the function converts to local time). utcfromtimestamp is technically fine. The problem is calling datetime.timestamp() on naïve datetimes; it will always assume they are in local time. Since discord.py returns naïve datetimes in UTC, created_at.timestamp() results in a timestamp that's relative to local time despite the original time being relative to UTC.

I think I'd feel safest by making everything tz-aware, as ugly as that may make the code.

[vale ibex]

I think the code was previously tz-aware, before this PR anyway

I'll give it a look this evening

[tawdry vapor]

I have it open now and I have some time

I could take care of it now

I also wonder if using arrow will result in nicer code here. It's already a dependency for something else apparently.

[vale ibex]

I think it's currently only used in the init for logging timestamps

but yea, I imaging it'll be much nicer

[tawdry vapor]

Some parts were tz aware others weren't it seems

[vale ibex]

feel free to work on it 😄

[tawdry vapor]

But the calc used to be much simpler so tz-aware didn't play a role

idle_time = (datetime.utcnow() - msg.created_at).seconds

[vale ibex]

yea, that's true

[tawdry vapor]

Now it does since we need to be mindful of converting timestamps back and forth

Such a headache 😔

[vale ibex]

I try to avoid timezone work as best I can at work 😄

[tawdry vapor]

Apparently arrow was only used for arrow.utcnow() once in the branding cog

Can't say I agree with keeping the dependency when porting it over

But that's history now

[whole forge]

well, maybe it can become useful if we already have it

[tawdry vapor]

Yes I am using it now

[whole forge]

👍

[vale ibex]

Ahhh, it's in lancebot where it's used in the init

[tawdry vapor]

Also, correction: it relied on humanisation from arrow too

Though IIRC we already have a util for humanisation

[green oriole]

We do have one

But I guess we don't need it

[vale ibex]

yea, bot.utils.time.humanize_delta

[green oriole]

Interesting dotted path haha

[vale ibex]

you saw nothing

especially nothing cpp releated

[green oriole]

Haha

[vale ibex]

When linting takes longer than getting code reviews smh

[patent pivot]

reviewing #1470 now

uhhh

ah

what are the possible values of closed_on?

I see "cleanup", "command", "claimant_timeout"

what is the first one?

cc @vale ibex

[vale ibex]

# If for some reason we have more than `max_available` channels available,
# we should move the superfluous ones over to dormant.

It's done on init, ig if the max available is lowered

[patent pivot]

ah, right

and that goes to the help session stats reporter?

[vale ibex]

ye

they're all auto.{reason}

except command

[patent pivot]

right

so the reporter doesn't differentiate between closed after claimant and closed after non-claimant?

[vale ibex]

There's also others_timeout

for non-claimants

[patent pivot]

ah right

is that documented somewhere that I am missing

either way I have a pending comment for a docstring update so I will continue having a gander

[vale ibex]

from what I can see, it's not documented

I didn't write that part, so I'll have a look around

So the docstring on _cog.unclaim_channel has some examples, but not an exhaustive list

It looks like it's only missing deleted and others_timeout

deleted being for when the claimant deleted their message, so we use the shorter timeout

[patent pivot]

review left

[vale ibex]

actioned your comment about the docstring

[patent pivot]

noice

resolved my naïve comments on the spelling of naïve

approved

[vale ibex]

I prefer naïve when reading docs or w/e, but usually write naive just because I'm lazy

[tawdry vapor]

That wasn't an approval. It was a comment.

[patent pivot]

got too used to cmd + ennter

done

[vale ibex]

huh wait, I updated the docstring in unclaim_channel() since that's where it was previously

didn't notice that you had tagged your comment against the _stats func

should we add some info there too?

[tawdry vapor]

I think it's fine

It's more visible on unclaim_channel anyway

Oh I see what you mean

The docstring for the stat function still describes it as a boolean

Yeah it'd be good to fix that

By the way, I was considering making it an enum but decided against it cause it felt like it was over complicating it.

[vale ibex]

yea, the docstring is just wrong since it's not is_auto anymore

[tawdry vapor]

I just feel iffy about passing around strings, especially since we compare against one.

[vale ibex]

I'll update it now

an Enum would be good here

[tawdry vapor]

If you think so too then go for it

Strings are kind of evil since they're more prone to programmer error

[vale ibex]

hmmm, would _stats be the best place for the enum?

[tawdry vapor]

Yeah

[vale ibex]

👍

[patent pivot]

uhh

we compare against it in the main file as well though right?

so it should go in _channel

imo

[tawdry vapor]

Either one works for me 🤷

[vale ibex]

Hah lucky we did, found a reason that I didn't mention

[patent pivot]

will have a re-review of that and then we need one more staff/contrib review and we can merge

[vale ibex]

👍

[tawdry vapor]

Well, I reviewed it

It probably doesn't count cause I made commits but we can overlook that

[vale ibex]

Just glance over this pls

[patent pivot]

ah, that's why it won't count, yeah

oooooone second

[gritty wind]

if you'all are good with it, just force

[patent pivot]

yeah, need to alter a branch protection rule to have the option to force

[gritty wind]

ok

I can just approve then

[patent pivot]

i altered it it's all good

it's just an "Include Administrators" toggle

which we can flip off for single PRs like this where there is a strong review consensus anyway

[sullen phoenix]

addressed some of your comments in bot#1491 @brazen charm

[dusky shore]

GitHub

[bot] #1491 Small fixes/improvements for !dmrelay

[fallen patrol]

https://github.blog/2021-03-30-github-advanced-security-security-overview-beta-secret-scanning-private-repos/

[patent pivot]

I don't think we get access to either of those lol

[tawdry vapor]

We should once they're out of beta then

[patent pivot]

they're part of the paid plan

GitHub Advanced Security

https://docs.github.com/en/github/getting-started-with-github/about-github-advanced-security

[tawdry vapor]

All public repos on GH have it

[fallen patrol]

And this just means it expanded to private repos

Like python-discord/secret-project

[vale ibex]

🎉 It didn't crash and burn on upgrade

[tawdry vapor]

The docs you linked say that at the end too.

[fallen patrol]

Is that a clone or actually this bot

[patent pivot]

actual bot

mods + core devs can access prod logs

[fallen patrol]

Ngl I'm really glad that wasn't smth else

[vale ibex]

yea, now we just gotta wait until the first close for a new reason

[fallen patrol]

For what reason

[patent pivot]

my favourite logs are metricity

[vale ibex]

We changed how help channels work slightly, expect a changelog soon ™️

[fallen patrol]

Personal thing about this but should it let the user know when a mod views their dm logs?

[vale ibex]

haha yea, the constant syncing spam

[fallen patrol]

Damn

[patent pivot]

I don't think so? We only look into DMs for moderation purposes + we do state in our privacy policy that we have that capability.

[fallen patrol]

is this the 10 minute and 30 minute thing?

[vale ibex]

yea

[fallen patrol]

Ah

[vale ibex]

Although we may play with those numbers

[fallen patrol]

The autohotkey discord runs a variation on the help channel system here.

[tawdry vapor]

We've heard

[fallen patrol]

Yes you've heard from me

[patent pivot]

We inspired the one on the TypeScript server as well lul

[tawdry vapor]

I don't think it's too helpful to compare since we have a much larger community and much higher volume of questions

[gritty wind]

It's a good system provided you don't have as many members as we do 💪.

[tawdry vapor]

How big is TS?

discord.gg/typescript

[patent pivot]

they have uh

[fallen patrol]

18k lmao

[tawdry vapor]

Is that the one?

[patent pivot]

18 hour timeouts

yeah

[gritty wind]

hahahahaha

[fallen patrol]

Seriously?!

[patent pivot]

yes, it's lower volume

[fallen patrol]

Holy

[sleek steppe]

and it's verified

[patent pivot]

When we were that size we could probably have had similar

[fallen patrol]

So do they have problems that take longer to fix or what

[patent pivot]

they work with msft

[patent pivot]

less questions

[sleek steppe]

oh I see

[vale ibex]

We can't be verified as we aren't officially python, since the PSF doesn't do that kind of thing rn

[fallen patrol]

Ngl this server is like one of two partnered servers in my entire server list and I'm on 100 servers

Granted, the verified icon takes presidence over partnered, so I'd have to double check

[vale ibex]

bot#1492 just a small PR to fix incorrect reporting of close reason

[dusky shore]

GitHub

[bot] #1492 Use deleted reason if help channel is closed due to being empty

[fallen patrol]

Does someone have all of the github emotes?

Like the pr above and whatnot

[patent pivot]

based off https://github.com/primer/octicons

[fallen patrol]

O dope

[gritty wind]

You can also use any that you have on your keyboard as far as I could tell

If on windows, try windows + .

[brazen charm]

the links in codeblocks aren't displayed properly https://pythondiscord.com/pages/contributing/contributing-guidelines/supplemental-information/#logging

[patent pivot]

cc @green oriole

[green oriole]

Thanko

!remind 20M fix it

[stable mountain]

Yep.

Your reminder will arrive in 20 minutes!

ID: 1831, Due: 2021-03-31T11:55:22

[green oriole]

Notion exported the links as

`[text](link)`
```That's silly

[exotic ember]

we might have a feature freeze in the wiki soon while I migrate content to dewikification

[stable mountain]

@green oriole

It has arrived!

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

[cold moon]

Also can we get freeze for resources?

[short snow]

freeze meaning?

[patent pivot]

we haven't made many amendments to resources anyway as of recent, but yeah

@cold moon @exotic ember how close are we in terms of getting those migrations done? I'm fine with calling a freeze now unless critical

[cold moon]

Basically content app itself is ready, there is PR open for improvements. My resources PR is waiting for review.

[patent pivot]

so that leaves uhh

redirects & events?

[cold moon]

Event pages are already migrated

[patent pivot]

aha

so just getting those open PRs done and then redirectss

[cold moon]

I will do redirects after resources PR get merged

[exotic ember]

@cold moon if you want you can start work on it, there shouldnt be any major changes left for the resources PR

[cold moon]

Okay

[green oriole]

So you have any estimated time for the freeze?

[cold moon]

There isn't much to redirect, I think. Only resources and events.

[exotic ember]

@green oriole should be a week at most, then changes should target dewikification

Is anything urgent?

[green oriole]

That's pretty fine then

[patent pivot]

Okay, I've made necessary amendments so we have no private pages in the wiki anymore

[exotic ember]

@cold moon guides content will need redirects too, I'll get you the changes when I work on the migration

[patent pivot]

71 articles total

only private page is the index lol

we never really did do that

[exotic ember]

hahaha

[short snow]

what is freeze, again?

[patent pivot]

stopping all new content

[green oriole]

No changes during this period

[short snow]

oh ok

[green oriole]

So we should enter the freeze now?

[short snow]

like no commits until dewifikation happens

[green oriole]

No wiki changes in this case, if I understand correctly

[exotic ember]

no wiki/resource changes, site wide feature freeze won't happen until a quite a bit later

[patent pivot]

https://paste.pythondiscord.com/muzidatido.sql?noredirect

all articles on site

to access them, pythondiscord.com/pages/{id} can be used as a shortcut

[exotic ember]

danke

[short snow]

https://github.com/python-discord/site/issues/459 i have left a comment here, would be good if someone can reply 😄

[cold moon]

@exotic ember I think I need resources routes for redirects.

[exotic ember]

@cold moon alright then, I'll try to get that merged as soon as I can

[gritty bolt]

What's Xithrius on github's discord username here?

[patent pivot]

lol

@vocal wolf

sticking to the theme

[gritty bolt]

@vocal wolf , yo I got your message on github. My laptop broke recently and so I can't make the changes until it is repaired, I am really sorry for this

[short snow]

allow someone else to take over or wait until lappy gets fixed?

[green oriole]

Hey @gritty bolt, do you know how long it will take before you can continue the PR?

[gritty bolt]

Depends on how quickly my laptop can get fixed, but if I were to give an estimate, I would say a week from now

[green oriole]

Alright, I think it is okay for us if you want to pursue the PR, it isn't really time sensitive

[gritty bolt]

In that case, then I can work on it when I get my laptop back

[green oriole]

Cool, sounds good

Are you able to send a message on github?

[sleek steppe]

Can sir-lancebot#632 get an approval from a core dev?

[dusky shore]

GitHub

[sir-lancebot] #632 Make .github a group, and add github repo + user commands

[patent pivot]

I'm going to resolve some of the addressed comments on bot#1463, getting hard to read things now

[dusky shore]

GitHub

[bot] #1463 Rewrite branding manager

[vocal wolf]

No problem, I hope everything works out for you. Thanks for the update.

[short snow]

cool go ahead, I don't think i can do that

[green oriole]

If you created the discussion you should be able to

[cold moon]

I think this is one of weird things of GH

[patent pivot]

pending comment for you here: https://github.com/python-discord/bot/pull/1463#discussion_r604729943

[short snow]

@patent pivot https://github.com/python-discord/bot/pull/1463#discussion_r604270842 for the text wrap thing you mentioned

right, i will have a look at it tomorrow, too sleepply rn

https://github.com/python-discord/bot/pull/1463#discussion_r602640328 this is not resolved as of now

[patent pivot]

It appears to be on an outdated diff anyway, let me double check

[patent pivot]

I think the response from kwzrd clears it up, the code is fine functionally

Actually, the logging comment probably isn't worth it since we're considering removing trace logging soon, it shouldn't be in production

[short snow]

uhh it is basically you need to log fetch succesfull messages, and before raising runtime error, just log a message, so it closes the log.debug("Fetching..., i will comment on the pr, once i get back to my pc

[patent pivot]

right, I can see to that, it'd be nice to get it merged by today which is why I'm having a look now

[short snow]

ok cool, anyway i will give it a tomorrow, tested it locally and worked like a charm

tomorrow meaning, 6-7 horus

[patent pivot]

the runtime error will propagate through anyway won't it