#dev-contrib
1 messages · Page 97 of 1
Do you have an image?
Sure but it might not be the best quality
Emojis are small, so it should be fine.
desaturating the current emoji will have the best result
got this, just getting it put onto the emoji server now


Does that color fit the embed background?
!embed 

thanks
It's contrasting enough imo
how's this looking
oooh, spiffy. What does something like $e^3 - x_2$ look like?
neat
Niiiiice
That sounds like a heavy dependency for just latex rendering
well it was matplotlib or an api
or well sympy which is even heavier
or a local latex installation which is even even heavier
heh
Is a local install really that heavy if it's just pure latex without any extra packages
heavier than matplotlib, surely?
we would also have to worry about untrusted input, etc
Matplotlib requires a latex installation anyway
https://github.com/JelteF/PyLaTeX is an alternative python interface
Oh I see it has its own parser but also support for latex, the latter requires it installed
not necessarily...
lol
The real reason I took lunch here
ah you threw me under the bus
no one said you had to make sir-lancebot#665
wait
.issue
wat that's a command
uhm
that should be whitelisted in this channel, no?
where are the error handlers
oh nvm
should i make a custom exception for latex compilation errors?
normally a try and except should be fine
seems like a bad idea fwiw
at very least
the regex needs to be improved
Why? It's just a link to an issue?
sir-lancsdrbot#1 for example
that regex is bad
Yeah, that could be better. But why is linking bad?
linking in itself is not bad.
I don't think there are many cases where you'd accidentally invoke it, maybe could be whitespace terminate
but making it work in every channel with this regex is a bad idea imo
Oh, got it.
does the in message handling include other orgs?
seems like most of the current issues have been patched
lol sad
We could easily put a \b either side of the regex so it only looks for full words
I think @green oriole is currently fixing/refactoring that cog, what are your thoughts Ak?
he is that's the issue above
Yes I am
I can do that
!remind 13h Add word boundaries around the regex if you haven't already
Your reminder will arrive in 13 hours!
👍
I don't mind PRs to my branch either if you want to take credit for it
wut
I'll push something now
confusidied grammar
lmao
probably just mobile things 😛
Nah, just my brain being smooth, there is a missing word
lol
What are the gh api limits?
no need for an issue, I'll just PR onto Ak's branch
Probably for the esoteric coding challenges we had for a bit
and eval
Can directly push it tbh
Eval is allowed in eso, yeah
The channel's category was special and it doesn't look like the config was updated so it's "special"
good point, but too late I've already got the branch up now lol
We use authenticated requests everywhere, so I think around 50k/h? I don't fully remember
not sure of the exact numbers, but I know its quite forgiving if you supply an api key
where's the code for that?
.src issue
For .issue?
.source should really be allowed in this channel.….
Was thinking of this, where'd it'd either handle <pydis_repo>#<issue_no> or <org>/<repo>#<issue_no>
It doesn't
that would be nice pls add
That was part of the original requirements
The staff decided to not include them
Dunno why, I wasn't there haha
Curious about the reason if someone has it, not really a format that's prone to being used accidentally
Oh, branch protection extends to non-main branches 😦
That's weird, since it doesn't on bot afaik
reasoning?
^
It doesn't though 
You can read the meeting minutes from that staff meeting
ahh, review bot is still there, but I can merge anyway
Well there is not the full reasoning
PRed latex, will respond to any comments tomorrow morning, need sleep 😅
I always read minutes :P
which meeting was it
Long ago
Jan 17 I think, hard to look on my phone
Would the regex work for just repo#num ?
17th january yeah
People just felt odd about it being automatic for all repos, which I still feel like it's weird. It makes sense in our dev categories for our projects
I am guessing it is to avoid silly backtracking if you post chunks of #
Although that can be easily solved
I am not sure why the regex is generated tbh
kutiekat i swear to god the american spelling of organization in your minutes will be the death of me
:3c
Hahaha
fixed
I change it everytime I make the meeting minutes which is always
no you change it AFTER the fact as well
LMAO I do if I go back to review discussion
I think if someone did use it, it'd have to be intentional with the format for different orgs/users, so not exactly surprising
around here #staff-voice message
People were hung up on the automatic part when opened to all repos
yeah
seems that the primary thing was abuse combined with ratelimits
Maybe we just bring it up again on Sunday?
Yeah, sounds like a plan
I'm not heavily against the idea, but we should lookup rate limits
Oh that was the meeting for the help channel names
Sounds good
5k/hr
Don't mind me
but we use them elsewhere for other projects as well
don't think the feature is worth it for staff only
I think providing we lock it to a tight regex it should be fairly difficult to accidentally invoke
I really doubt that amount of requests would be reached even if the whole server was aware of the functionality (w/o malicious uses of it)
Something like [a-zA-Z0-9_-]+\/[a-zA-Z0-9_-]+#\d+ should be fine-ish
I think a simple regex that looks for sirl---bot#--for example works for accidental lancebots.
Probably a shorthand for this, but heh
-- being anything
the boundaries and lancebot being matched first should take care of it
duck
lol
Yes, [\w-] would be shorthand
Oh nice
Also a bonus that it works with Unicode rather than ASCII only
Actually it doesn't seem like GitHub would allow Unicode in the URLs anyway
i just realized, does this mean that sir-lancebot is authenticated into the python discord private repos
Nope
In that case \w could be set to ASCII with a regex flag
The token doesn't have any scope AFAIK
so
lancebot can read private repos
but the code used to fetch all pydis repos is unprivileged
so it doesn't actually see repos like organisation, admin-tasks, etc.
and therefore they don't go in the regex
Oh really
makes sense
My new code make it be authenticated
@patent pivot can we drop the perms for the token?
I don't think we can, no
hmmmmmm
which token do we use on lancebot
hold on let me delve into k8s
Is k8s short for kubernetes?
just make a request to https://api.github.com/user
oh wait no int e yet
yep
ah
looks like that token can't read private repos
x-oauth-scopes: gist, public_repo, read:discussion, read:enterprise, read:org, read:packages
Cool
we should get someone on that eval implementation
fwiw
@green oriole if you're into it, i'd add implementation to check if its public or not
do we need that
Is it just ripping it from python/black knight?
Well if @hardy gorge has a minute that'd be perfect
that way its proofed against security
the current implementation works since we just fetch it unprivileged
we can check when we add the org/repo syntax that fetched private repos can only be used in staff channels
but there are use cases for this feature for using it to fetch organisation issues
why did that bot never become public 
security
security by obscurity
Cool I'll see if I can rip it over since I wanted to dive into eval internald anyway
if we make our anti-abuse features public, people anti the anti-abuse
That's... Not security by obscurity
but aren't all of those features on @stable mountain with anti abuse anyhow
It's security by security not obscurity.
no, they're not
unless there's some cogs in @stable mountain that are .gitignored
then where are they?
certain components that we had in mind for BK such as improved defcon were ported over to defcon on @stable mountain
can't remember off the top of my head, but not sure we need to detail too much on that
Discord has gotten better with raid stuff so we didn't feel the need for a whole separate bot. We're making sure the anti pattern knowledge stays with us so if we need to protect against the raid we have it
And worst case scenario, have an admin turn off send messages for the everyone role. Simple.
i would be surprised if there wasn't a !hush entire server
I think the server silence feature got ported over? That or it's on zig's to do
i also wouldn't be surprised if defcon was actually enabled today
Nice, the defcon upgrades are so nice
….but that description is technically wrong :/
huh?
What's steakbot? Is it a new project?
Shut down the server by setting send permissions of everyone to False.
Steakbot is joe's glorious slashbot child
me playing around with slash commands a few months back
Testing slash commands? Nice
when the support is there we can migrate core projects to use them
but until then it'll remain playful things like steakbot lol
huh?
if the server was being raided wouldn't you want to shutdown voice as well?
in previous raids
the approach there has just been uhh
delete the voice channels
lol
lock_vc()
for u in vc:
u.disconnect()
Literally just turn off view channels.
no, voice channels are special
And set Rules to override.
feel free to PR
done
why
i can still talk
yeah, we handle that in voice ban
But how do they get in in the first place?
just #voice-verification lol
That's not an issue.
it's if role is removed during a voice session
which we account for in voice ban by removing from voice session
i opted out of zoot
Raid accounts generally won't be voice verified, so /shrug
i'd opt in to zoot then in that case

Oops.
I'm so confused, since Dave only has a couple files, yet it generates so much data. It's cool.
I guess I did remove it, so I'll grant it back lol
if not, expect a dm in about 30 minutes
*Dave
I should add some notes to the privacy policy saying there are downsides of opting out
i mean
we've considered message gating functionality before
if you do opt out
after getting voice verifed
then you got the role
and should be fine
yeah, but if a) you then request clearing or b) we change thresholds then you may be ineligible
but yeah, opting out doesn't reset the countner
uhhhh hmmmmm yeeeeah. That sounds like a thing
so if you have 50 messages and opt out then you are still eligible
since we still have that data
then c) opt back in and requalify and then opt out again and dm joe to clear it
fwiw
joe is the one who will be more inconvenienced than me
Because you'll spam his dms? /j
i have 30 days to handle a deletion request lol
not entirely /j, he has a legal obligation to clear it.
but 30 days to do so
opting out is less legal compliance and more a generosity move
we don't store any information that requires opt out functionality to be present
user data is still in metricity
if you didn't have the opt-out*
it isn't actually handled in the code. all voice channels can still be joined
a portion of the solution would just be to remove connect perms from everyone as well
like
one line code
I can make a pr
yeah, but at that point we may consider metricity data essential to the operation of the community
How does metricity get so much data? I cannot find a place in the repo where it gets so many different types?
which data?
then me and the other 12 who have opted out will leave the server.
isn't that just an on_message event?
Isn't that what gets the data for https://stats.pythondiscord.com?
Public statistics for the Python Discord server
the data there comes from bot
hmm
is it worth leaking something that provides info on this
Ohhhh. So then what kind of data does metricity get?
rich details on message counts per channel
Yes. PyDis operations are amazing.
I see.
medium smh
it's not medium lol
Wait can you see my post through that?
lol i unpublished your thing kutie don't worry
Oh thank god
hahahahahah
@cold island can you clarify which role has the connect permissions
i assume its the default_role
aka the (dangerous) everyone role
everyone has connect, voice verified has speak
interesting
entirely irrelevant to any pr, why does the Muted role then block from connecting to all vcs?
seems like it should just block from speaking?
you can still be a nuisance if you have connect privileges
with it blocking connect, people cant spam connect and leave
ah
voice banned users generally don't fuck around after being voice banned
and if they do, it's nice to completely bar entry
No, voice ban just prevents you from verifying
this has become #community-meta
lol
its 0.001% relevant to a pr
it's semi related to implementing this extension to defcon so we're good
/bot yeah
yeah, don't think so
what about /sirbot
there is on lancebot
how can i view that 
uhh
best to look at the raw version, it has comments https://raw.githubusercontent.com/python-discord/sir-lancebot/main/.github/pull_request_template.md
it.… redirects.
so i saw the raw version and only the raw version
wait
huh
why
how
it somehow redirected but normally doesn't
ok
pr created
removes the connect permission when the server goes sour
?
nvm
@green oriole What do you need?
For bot#1511 , I think we should set speak=False
Int e command from black knight was the conversation context
It's already false for the everyone role
It's granted with the voice verify role
And changing the permission while in a vc affects nothing because of when discord checks permissions
So?
Read the next part
Changing permissions while in a vc affects nothing because of when discord checks permissions
Permissions are reflected when joining the vc and never again.
Until you leave and rejoin the vc
So if one was to turn that off, it wouldn't affect anything whatsoever
Only if for some reason there is a channel which has the everyone role with connect overrides to True
@sleek steppe ^ that make sense?
Why not hide channels?
that would not be a simple single role change
that would affect text channels as well
No, I mean why not hide the vc channels?
because it would need to be an edit on every single vc channel
True.
or it would affect text channel as well if turning on view channels
this fix disables voice for everyone as well if you aren't already in a channel
that was me, yeah
I restarted it to reset some mitigations
what mitigations
anti-abuse things
like this thingy?
nope, that was just a token in our mod alerts
so anti-abuse is closed source?
it's not closed source, it's just that we program certain snippets as and when
like there is no source code
o
like add a whole cog lol
we have a policy for not adding commands in production, so not really
ah
smh you and your intelligent policies that stop stuff breaking
lul
its more fun when you get called at 4am because everything broke /s
ikr my bot i actually changed the bot to a duck instance
just for the lulz yesterday
one such mitigation from today did have adverse side effects lol
lol
oh just 2004 events match this spam regex
nothing important
we just banned spaces its fine
not_much_happened_dont_worry_everyone
my craving for ci/cd hasnt gone away im scared lol
lol
I am working on this today https://grafana.com/blog/2019/08/21/how-grafana-labs-effectively-pairs-loki-and-kubernetes-events/
CI/CD makes me go 🥴
@patent pivot I got conflicting info on this
^ no token used
should i make an issue before working on a pr for that?
hmm
fwiw im not so inclined to make the pr for this one
yeah if you don't want to make the PR then issue 100%
but the ratelimits go boom
i mean
just copy the code from the other cog lol
or
perhaps make it request from github in a util function
¯_(ツ)_/¯
okay definte issue either way since there's multiple ways to approach this
and i might try and make the fix myself but that would require me setting up the bot on my servers
…
is this technically a bug
also uh
I mean, it doesn't matter that much. The template.
that's why i was a bit curious on the checking if its a public repo thing because if we use the token on this cog and then sometime give it public support this is a major security hole for being able to see code of closed source repos in the python-discord organisation on github @patent pivot
for the .github command we can definitely sanity check the repos are public
I will type organisation so sue me kutiekat
when we rework the org/repo#123 syntax then we'll handle public and private safely
im thinking like
the request function for both of these could possibly be moved into a util file
and do all the checks in one place
wait we can do any org not just pydis? or am i reading that totally wrong lol
yes
sir-lancebot#642 sir-lancebot#630 there's an issue and pr on that exact thing
but like
No it isn't
there's three seperate things here
That's completely different, isn't it?
we will be able to eventually
ah nice
well, we'll probably discuss it on sunday
its in the pr
i'll try be awake 
Oh, I see.
no, it is just specific to python discord
it uses tasks, to get the list of repos in every 30 minutes
right now it is
and then makes a regex out of that list
That would need a very tight regEx, or it'd get ratelimited, right?
i think you overestimate how often people actually type that and underestimate the size of github's ratelimits lol
@patent pivot it'd be kinda cool if everyone could view the meeting minutes but that's just me
There are things we discuss and put up there that we don't want shared with the community early.
Ahh
True, I guess people don't type #s much.
A lot of our internal updates happen there including the events we plan
even if people type a # it still wouldnt be in the right format most of the time
Makes sense.
Fair
(\w+/)?\w+#\d+ with a bit of refinement should be fine
and the debates over organisation vs organization.
Yeah they wanna keep that under wraps
Lolol that's not even a debate. That's me waging a one woman war against the brits
im american and
im not happy that my precious quackstack has been invaded by 'color'
Organisation on notion yells at me, so I change it to organization :3c
@patent pivot put a constraint on the meeting notes that doesn't allow organization
lol
just like the db doesn't allow you to be infracted
brb getting job
But ignoring ratelimits, what if multple things are named the same. eg) ondiscord/bot#1 and pythondiscord/bot#1. That might be an hard if we want it to be really accurate.
That wasn't supposed to happen.
We should only count it if it has a space before.
we should go bot#1
fine if we're gonne be pedantic about including the whole thing, \b(\w+/)?\w+#\d+\b
notlikebot#1hahahaha
Hahabot#22
sir-lanceldjejsjejribot#69
sir-lancebot#670 can someone change it to enhancement and not feature
it gives me literally no option when creating so please don't tell me i could.
done
danke
Why'd you check both?
because i'd perfer to implement it but it's going to be like two weeks before i bother to set up a sir lancebot clone
so if anyone wants to do it before then they can go for it and i won't mind
lol i just realized i created my first prs in the pydis org today
Just join the test server Jason made.
He has premade configs
What's the problem with that?
i don't feel like cloning the repo rn
not until im a little more competent with git
¯_(ツ)_/¯
This is my config.yml's urls section, which worked when you did:
$ docker-compose up -d
$ docker-compose up bot
But it gives couldn't connect to api when you do:
$ docker-compose up -d
$ pipenv run start
And, all the images are up
I can also go to 127.0.0.1:8000 and open the site
random unrelated q, are you running a snekbox
yes
you're running everything lolz
@tawdry vapor had a missing message caused a 400 for you in the past? If so, do you remember what specifically created that error?
Context: https://github.com/python-discord/bot/pull/1414#discussion_r594641273
Not necessarily
But there are some endpoints that do that
The API is inconsistent like that
ah
I don't remember what it was for
because it pulls from the invite, not the actual server
I'll be leaving now sorry for the ping
we don't want presences intent enabled, but we want presences
I thought it was some kind of error of sorts
Hard coded in what sense? Our Python bot is intended to only really work with our server

just checked, it isn't hard coded
it's a config option
here is the code in !server that fetches the counts from that configured invite https://github.com/python-discord/bot/blob/main/bot/exts/info/information.py#L169-L176
Well yeah, but if something happens to the pydis invite of python it'd be nice to be able to not have it hard coded, which it isn't, so the point is moot
@fallen patrol Partnered servers are warned 4 weeks in advance if they're on the way to losing partnership.
That definitely long enough to change things.
!remind 10h metricity#3
Your reminder will arrive in 10 hours!

why
why are you doing this to the 10 of us who opted out
because i put my privacy hat on
and then i put my data hat on
and came to the conclusion that the opt out functionality is more harmful than it is good
😔
It's not any personal data
it's data that one can already accumulate through discord search or our existing message logs
no
that's stated in the privacy policy
metricity messages have an is_deleted property
But mods can see that too.
And they're the only ones who see metricity metrics.
So it's not really that much of a change.
so metricity logs more than python in a sense
Not really.
not really, metricity never has contents, ever
api_messagedeletioncontext & api_deletedmessage
There were 285 message deletion contexts in the database, within these there were 1,300 stored deleted messages.
These have been deleted since they are stored only for review and are deleted after no longer than 30 days.
that's Python, yeah
just dont rename anything, joe ;) lolll
those are the anti-spam held messages including context
so metricity has more info on when i sent a message than python does.
if it was spam then yes, otherwise prob not
which means you could figure out every time i've deleted or edited a message, no?
deleted, yes, edited, no
they log deleted and edited stuff with @stable mountain i think
you don't store edits?
not with metricity
python handles edits
and yeah, to an extent python can see deletions and edits
You're not losing any privacy then.
The logging of the typical deletion and edits is kept within discord though
Crap, then you see all my embarrassing spelling mistakes.
yeah, that's true, messages get an edit timestamp at the discord api side
hahahahaha
if it makes you feel any better, you're not alone there lol
but yeah, the use of metricity data is pretty vast, and this opt out isn't worth the loss of clarity
It's okay, I flood that log with my terrible spelling when I use my phons
are the uses detailed in depth anywhere?
I detailed some of them on the issue
metricity#3
nominations use metricity data now, voice verification, channel usage metrics
a few uses inside moderation as well
what cogs in python
like?
metabase, grafana, the postgres shell
yeah, exactly that
sources?
I'm not talking like systems here, this is just ways that we query the data, we have uses inside moderation
all three are open source
The grafana stuff is obnoxiously useful for talentpool stuff and general quick data trends
I'm curious, how many folks have actually opted out?
Can we see a limited version of the grafana?
Or is that exactly what the stats page is?
only ever a handful
for sure!
oneeee second
let's see
in that i see the reason to remove it but also to not remove it since its only a few people
lol trying to find a dash without internal info
You could possibly reach out to those few if it's not that many to give a heads up
It'll be changelogged
that data comes from the metricity table
metabase is where we tend to crunch message data more often
What are those categories at the top? I can't read it on my phone
the categories on the top graph? not sure what you are referring to
Yeah
red = not passed native gate, green = passed
The title
Oh ok. Ty
Users pending native gate
That's pretty cool
I mean like a warning ahead of time of some kind, never know if people will want to abandon the server or something haha
Probably not a super big concern since its not that many though
its like 6 of us ffs
Will one day of metrics hurt them?
"you opted out! We just opted you back in!" is a pretty bad announcement
True
it's anonymised, yeah
i think they'd have to give a date to remove it by and changelog it before its actually removed. I think.
I still want to make a gif of the peak active times changing from Oct -> Feb
Ooh, those are some nice insights.
The AoC activity shift is pretty amusing
for this PR?
it's not a compliance thing, this isn't personal data
AoC?
Advent of Code
Got it
joe 
I'll just amend all relevant documentation at once when I merge the PR
I agree, and that's why we don't word it like that 😛
Alexandria O Cortez
AOC != AoC
really hope i spelled that correctly
*''
I think the body of the PR explains it fairly well, it's a fairly misleading opt out policy since it only actually opts you out from the part that is fully anonymous
….
so you mean to tell me i've opted out of being anonymous this entire time
brb dming the data controller
no, that's not what I'm saying
Would it hurt to share that like a day or two early though? Still seems like a weird thing to do without public notice imo.
I'm saying that the component that you opted out of, the message counters, is the part that is anonymous
guys who's the data controller again
oh yeah for sure I'll give it a couple of days
i have nothing but bad vibes if its not early, its legit shady business practice to change stuff with no time before going into effect
Yeah that's all I'm saying, to just make sure the folks it affects are aware if possible ahead of time
But then again I have no say in this so... carry on 😂
I wouldn't be surprised if you at least have some. They consider a lot of community ideas. 🤷♂️
this is the current structure of the held data
so the former two are handled as part of the data review and removals, amendments, etc. are carried out there
and metricity message data is handled in the opt out functionality of metricity
Nice diagram
Makes sense
So you're not opting out of anything specific to you?
but the misleading factor there is that opting out of messages is in no way an opt out of other data collection, and I think that the best way to make that clear is not to have misleading stuff there and instead to integrate that into the data review process
and as I mentioned, the message metrics harvested by metricity are proving very useful in many areas of the organisation, so it makes sense for them to be treated as such
True. When I first read the opt out thing, I assumed you opted out of everything. It's much clearer when you remove that whole idea
Metricity#3
but I do agree that forewarning on things like this is important, so making the privacy policy amendments beforehand and changelogging it is a good move to make, so I'll give it a good week between publishing that and these changes going live
metricity#3
Nice talking with you all, but now I must sleep. Good night!
https://github.com/python-discord/sir-lancebot/pull/666 needs one last review
ty @cold moon
I have idea for @dusky shore : command that fetches information about compound from PubChem API.
Here is API docs https://pubchemdocs.ncbi.nlm.nih.gov/pug-rest-tutorial
Would that be useful here? I feel like lancebot has a few too many api wrappers which don't really improve the project for beginners, and I believe WolframAlpha can get the information too
@green oriole
Here's your reminder: Add word boundaries around the regex if you haven't already.
[Jump back to when you created the reminder](#dev-contrib message)
Chris got it
would love a review on lancebot#668
!remind 5H review sir lancebot-668
Your reminder will arrive in 5 hours!
How long does this take the render?
EG if I gave it some long, complex, input, could it take a few seconds?
I'd say if its >1 second, with any input, then we should run it in an executor
@placid ermine Do you have experience with running code in executors?
btw, I am talking about the max case, otherwise it is instant rendering.
yeah, somewhat
just run it in a ProcessPool right
yeah would probably be worth it, timing the _render it seems mostly fine but takes a bit longer on very complex inputs lol
👀 that input
oh no no i just added the timing to test it
lol I just did r"\frac{0}{0}".replace('0', r'\frac{0+0}{0}').replace('0', r'\frac{0+0}{0+0}').replace('0', r'\frac{0+0}{0+0}').replace('0', r'\frac{0+0}{0+0}').replace('0', '1'), that's as much as discord's text limit allows 😔
✅ Your eval eval output is posted here: https://paste.pythondiscord.com/qibocukoyu.txt?noredirect
hmm
that is pretty much what i had done in https://github.com/gurkult/gurkbot/blob/b28d4349986e6a078f4306eccf0790388a90a7fc/bot/exts/fun/bonker.py#L108-L109
I have a question regarding the test server, since everyone is a super user or a admin on the test server, anyone can run fork bombs on the user's bot, so do you guys prevent that in some way, or just trust everyone 🕊️
trust, that's why it's staff only
ah, are contributors also in it?
no, they're not staff
i mean
you could make an issue to possibly make it a little more friendly for building
yeah it's helpers+
huh?
as in setting it up?
that is a project of ours
You should make a bot that has all the emojis
cc @green oriole
And it can put them all in a server
shouldn't smartconfig resolve some pain here Ak?
smart config
an additional config in @stable mountain to like disable most of the stuff except for from bot.owner
Henlo
Hi
Why?
i could possibly make a script for server setup
That isn't functionality we need
Yeah, Smartconfig will come with an helper command to generate the whole server and config with one command
We need multi-user access during our testing
Ooh, that's cool.
no but
you could do it in like
4 lines of code for the benefit of most contributors
Well, not really
I don't think it is for the benefit of most contributors
It's for the benefit of server staff maybe
in the test server
but even then like this isn't a protection we need
because a) people don't run dev bots 24/7 and b) if we have trust issues with staff then that's a wider problem, this shouldn't be a problem that needs a solution
as for contributors, just run it on a bot account that is in a server with only yourself
To make sure I'm understanding correctly, smartconfig will allow you to set up test servers with one command?
Yep
The end goal is that you create a server, invite the bot, run a command grab a coffee because of ratelimits, restart the bot and you are good to go
It should also be able to sync with the configuration when changes are made
Ratelimits smh
so the bot will be a hosted bot from pydis?
No, it's a package
Nah, it is your development bot
Oh, that's useful
then how tf can it sync to this server if its not in it??
No need to copy and paste
It will be inside the source code of @stable mountain, only loaded when debug mode is enabled
the server configuration is defined in the repo yeah
The sync command will also be inside the repo
because we don't want to exact mirror of this server
it's bloated for development server setups
users don't need to synchronise modmail threads, archived channels and other internal things with their dev servers
till then, i willl just come up with a setup script for this server
(Hence why the template we provided doesn't have all the channels)
(We will have to think on a way to merge some channels in one inside the configuration but it should be fairly simple)
@crude gyro In regards to the second part of site#453, I was wondering if I should place anything where the timeline used to be on the home page?
Also, is there a specific place in the "more" section of the navbar you want the link? I have it under Contributing at the moment.
uhh. I wasn't aware we removed the timeline.
to me it doesn't really make sense to remove it until we find something to replace it with.
@green oriole any idea when this component disappeared?
oh it didn't, it's still there?
hmm I still see that component.
it is just not on the top row anymore
it was earlier nah my bad, the 100k banner was there
I don't think we should remove the interactive timeline undernearth
Didn't it mentioned the timeline?
Apparently not
Oh, I though the issue mentioned it.
Oh, I misread, my bad.
But I'll still open the pr to move the timeline to the navbar.
Any thoughts about placement for that?
Yep, it would be nice to have it up there
.
Oh, that image is my copy
ah ok
Above FAQ is probably good
I'll go ahead an open a pr.
PR has been made. Also, is there a plan to add anything where the 100k banner was? I
move seems to indicate you're removing the widget from the landing page
Oh, I mean "add". My bad.
Do we want to remove the mobile banner?
?
On mobile, the 100k banner still appears.
It's this portion of the code.
It’s not wrong, but it definitely conveys the message that we just hit 100k
When we’re almost at 200
In my opinion, we shouldn't remove it, but rephrase it for the timeline.
@patent pivot
Here's your reminder: metricity#3.
[Jump back to when you created the reminder](#dev-contrib message)
@short snow
Here's your reminder: review sir lancebot-668.
[Jump back to when you created the reminder](#dev-contrib message)
sirlancebot#668
o
Just need @exotic ember's review on sir-lancebot#642 to merge it 🥺
I guess if kosa isn't available someone can review for them and dismiss
Looks like few good ppl already reviewed it and I have nothing to say, but well I am gonna do some bikeshedding on the output message 
and to be clear, this change won't make it output properly, it just won't output anything 😄
Yeah
I'd have a couple of nitpicking to do but I'll refrain myself haha
For lance, why are some hardcoded and some not?
Lol, I would like to add timing to it
Who knows
Some are not hard coded, as they are not needed for normal testing ig
This will be obliterated by smartconfig which should come further down the road
If you need any for testing feel free to introduce an env var
How long have we been talking about smartconfig coming soon™️
But I don't think it is worth switching everywhere
You really don't need most of the roles in development
I said "further down the road" haha
the only real use is for bypassing checks
Lol, I was gonna just ask that scal, but AK’s pr needs review
and admins=mods in sir-lance for the most part
Hopefully we should merge the main PR soon-ish
and admin=owner in both
hows the progress on quackstack?
We don’t really need postgres on quackstack
progress not postgres :P
I wouldn't care if it isn't completed this year, compared to other projects
Autocorrect, iPad isn’t that smart


