#dev-chat
1 messages Ā· Page 22 of 1
Would you recommend one?
And if you have ever used it before, compared to garage how different?
Garage seems to be the more lightweight one, but with less options. Also european backed vs seaweedfs being chinese in case you care about that
But both are seemingly solid options
For new people the ui is prob gona matter more
š āļø
<@&1050493473033289778>
are bots able to view if a user has nitro (and for how long) and also if theyre boosting (and for how long) ? bcs i saw a bot doing it but im not sure if thats something supposed to happen or theyre just scrapping xd
not anymore
the identify.premium is now needed to see nitro tier of a user... but from my understand it does not show how long... just if you are
i would have to double check for boosting
i see, so if i had to guess id say that the bot i saw is prob scrapping the info or something like that, thanks
boosting would be kinda doable with checking e.g. per-server roles
i see
the bot can also get this info from users that dont have it installed in tehir acc so its defo scrap anyways
do you guys know how long intent applications take or if you get notified if it got declined or not? I sent an intent application like a month ago and i was gonna check the ticket status today but its not in the support website even though i got a gmail saying the ticket was recieved and everything... nothing else since
you always get notified after they respond... some users are reporting 6-8 week wait times... but there's no official answer on how long they take
The ticket only becomes visible once they reply
that makes sense!!! well thank you both for the answer! my bot is closing in on 100 servers and it was stressing me out š
i suppose i just have to wait it out
that or temporary pause functionality that uses the intent until your request is approved
obviously not possible for all applications
Is it only verified bots that can set per server pfp, bio etc?
No, all bots can
Is it in the docs?
That's wild
now this is why you make the part with heading and thumbnail one section and the rest its own so its not sooo restricted horizontally
i wasn't showing the embed, but the content in it, anyway
oh who has time to read
Ok so i got garage to work, only now to realize there is no proper user and token policies...
the only time i need ai (other than making first-look website) to check if my codebase has some hidden issues that i couldn't find, it lets me down due to people overloading the servers š„
Its internal only, but either i leave it or replace it with something else that does have policies
Isn't selling permissions against the content guidelines?
I guess... Asian servers are just weird
I guess... Because they're not westren
What do you mean.. 10k is 10,000
35$
Still a lot ngl
What are you even talking about
Also wtf is ban card lmao
Yep, ain't paying a penny to look more cringe in a cringe server
let's stay on topic
Ban people maybe
Hey, staff team I donāt like their username so letās ban them with using Discord API, idk sounds so absurd xd
-# staying on the topic
be aware from AEnterprise
trying to loophole like that is just malicious compliance that will get you removed
I was staying on the topic but alr
Ų§
AE any interesting scams today?
Stay on topic
oh well had a question but i'm unsure if it's on-topic or not š¤ it is a bit bot development related (in a way they do earn a role, so if it's not understood correctly), here's the question:
i know it is, but does this apply to in server currency too? ofc they aren't actual $, but like e.g 100 gems (which you can grind in /daily or /work idk) for sending image permission or something?
-# if it's not on-topic feel free to delete, i do appreciate if it's possible to link me to somewhere i could get an answer though
Iām wondering for a while why Discord doesnāt let people have RPC without local running
not really, messing with garage for s3 storage today
bit more work then anticipated but getting there
Ive had a few spikes on my graph today nothing special really
Pay virtual currency to buy server perks?
I do have similar item related stuff in my economy module, users can buy items ( which can be a role, custom role ) with bot balance
Yeah I added the rizz model cause not many people are good at talking to people so I just did it for fun
Ik ik I was just joking
Ok
yeah I'm aware of items, but what I'm worried about is the "roles" part, but I'm guessing it should be fine, since it's user's guild balance (bot)
š¤·āāļø
Hi Im developer
Hi Developer
sometimes it just ain't active here
Im a dev
timezone
ideally... so is everyone here
lest keep this channel server dev related... not random stuff like this
that is awfully optimistic
i know 
Hm
Could I say one thing off-topic?
No, itās off-topic
There are plenty of other servers where your off-topic banter is on-topic
I canāt go to the official Discord server where I wanted to talkā¦
That doesnāt change anything
this is not an alternative to townhall... if that's what you're saying
Can I say something on topic?
Maybe?
something on topic
Linked roles api is good and so underrated
i agree... it's poorly understood so most just don't know how it works or what it does
To be fair itās above my head
If only we could have separate metadata for guilds separatelyā¦
I also saw some servers got access to add their game as connection in the onboarding
Such a cool feature
agreed. I've never needed it, but I agree
ive not needed it... but could have used it, if i understood it better
do you all remember that dude who was begging for contributors to their project?
Hasn't been active since they were banned
I just happened to still have the repo up on my workstation because I don't close tabs lmao
i hate unicode :) more specifically dealing with it in code
use binary and make it someone elses problem 

Someone know where I can find a well-made doc for new V2 components bot
here's the official docs... unless you are looking for the docs for hour library
or you can use https://discord.builders/ for a web based builder for them
Thx
Hello everyone. I have a question, how can I insert an image that when you click on the image it opens the link? It's that on a server I want to put images that when you click on them they take you to YT, TikTok, and other apps.
have a webserver respond with an image when discord requests it, and an redirect when a browser does
not really easy and generally a bad idea as it's the least obvious and most confusing way to do that for end users
I hate qt so much
but im trying to do it, to agree an image with my Tiktok, Twitch an other apps
and when you click on the image, redirect to the tiktok or the app
again: worst idea i've heard all day
just put links, users know to click those
nobody is gona try to open your images thinking it might go somewhere other then the image
ok, and thanks for the suggestions
Also pretty sure impossible to do on discord i think they mean when they user would click the image which normally full screens it that it would open a link
it works if they also click 'open original'
so again: nobody in their right mind is gona do that and expect it to go anywhere other then the image
ok, then i note it
my main id got banned having 18k members . reason someone drop nsfw video and i ban him made me loss my id under secound
yeah no you're making baseless assumptions here
i made ticket but taking long ig
file an appeal with support and wait for them to reply
but nobody here can make it go any faster
can anyone brute decrypt this algorithm, i just made a new encryption algorithm
DC000A76D4230AC3A77B9261E61E8A8F4D9C4C538E147EFFB5525C85CE50C34D
am owner , why i drop nsfw in my own place .
hi guys, i need advice ASAP: my friend literally just got hacked. he wasnt thinking and gave someone his email and is now logged out of his account. is there any way we can get his account either back, or deleted? because there was sensitive info in his dms with his friends.
ohk
!support
This server is for help with Discord APIs, SDK and other various integrations. For general support, please contact Discord at https://dis.gd/contact, or send us a message on Twitter at @discord_support
problem is, my friend cant log into his discord account to make a support ticket. he was hacked out.
support portal is completely independant
it does not use your discord credentials at all
see the giant banner on the page that links to the article explaining that, and how to submit a ticket
they really can't make it any bigger
alright, thank you. u didnt need to add that last part in, that was lowkey kinda rude
wasn't ment to be rude, just to get you to read the page
cause if that big an banner doesn't do the trick, i'm not sure what else does it
with no context this is just a meaningless byte string
what's the algorithm?
wait
gg i lost my 18k server
please stop and move on. we can't help you
wait for support to reply
I dont think it is feasibly possible to decrypt a message encoded with no information about the algorithm and only having one string of what seems to be only a couple dozen charecters.
^
GF(2^8)
prime based brutal bit rotation
circulant 8Ć8 MDS Matrix Diffusion
feistel network structure
xor round subkey Injection
keccak based key expansion
nlfsr entropy folding
Arnold's cat map torus permutation
constant time cubic hardening
quantum save feistel
memory hard temporal fortress
post quantum lattice binding concept
blake3 integrity seal concept
multi round non linear diffusion layer
GF(2^8) finite field arithmetic
avalabche oriented mixing layer
dynamic prime rotation schulde
hyburd symmetric chiper architecture
block based cryptographic transformation
entropy amplification pipeline
:v
You claim to be using a Post Quantum Feistel Network in your design - I'm not aware of any in existence, certainly none that have been standardized
honestly it sounds like you just looked up "terms related to symmetric encryption" and put this list together
hmm
I mean these are just terms. you haven't described an actual algorithm
Hello iām brazilian developer and iām not speak English so mutch
This server is for help with Discord APIs, SDK and other various integrations. For general support, please contact Discord at https://dis.gd/contact, or send us a message on Twitter at @discord_support
š
Is there a way to keep containers from "Shrinking" eg always make them full-width without hacky solutions? Eg I could add a bunch of empty characters, or an image to force it, but curious if there are any other known solutions?
Probably adding select
hey i create my discord bot how i get a badge?
don't advertise
The active developer badge was removed a long time ago.
u cant dev badge got deleted
Nah adding a select doesn't do the trick either. Plus that's hackier than the other solutions xD
What are the other solutions
The only ones I know are arbitrary invisible characters or a custom image divider
You know it's rendering differently from mobile to web, invis characters would break mobile view hard
mobile usually always stretches the container to max
on tablet, containers have variable size but separator stretches them a considerable amount
and web is just fully variable
hi, long time no see, guys
Yeah the problem is if you insert invis characters on mobile there will be awkward blank lines

Would seem the image is the only way to go then..
I can do a 1px X 500px image that will force it, but I have to make sure it loads FAST or it'll end up looking jenky as all getout.
iceman
What is the objective of trying to make the container wider.
yes, I am still making the full documentation in a whitepaper in docx format.
also, it's not 100% pure yet because it's still in rust, C, PHP and others are still having errors, so it will still be a long time before it's published.
how can i make my first project
An algorithm publication does not typically contain code bound to any particular language... pseudocode, sometimes, but primarily formal proofs. This is a symmetric scheme, correct...? I'm wondering why you're using lattice module cryptography, which is almost exclusively used in asymmetric schemes for key encapsulation/decapsulation
First project for what? Just in general or related to Discord?
first project in c++
i dont really know where to start though
Have you learned the basics of c++?
most of them
Then you should start simple. Something like a guessing game. I will send a website with a list of starter projects
First project in C++: decide on which build system to use
New programmers often need small projects to work on as they hone their skills. This is a list of project ideas that beginners can tackle.
fair point. lox-617 ( my own algorithm) is a symmetric block cipher. Iām using lattice-based logic specifically to harden the Key Schedule, not for the data encryption itself. The goal is Key Entanglementāmaking the relationship between subkeys as mathematically complex as a "hard problem" to prevent algebraic key-recovery.
As for the 'Post-Quantum Feistel' claim, it refers to the integration of these lattice-inspired layers and the 16,384-byte memory-hard Temporal Fortress to resist future hardware-accelerated attacks. I agree that formal proofs are the standard; this document is currently a technical specification of a research prototype intended for benchmarking and peer-review
Is that AI LLM slop
AI slop
I can't speak English so I asked AI to help translate
Could you put the original message through Google translate and paste the result here
"Temporal Fortress" is not a term I have seen in any cryptography literature, so I'm not sure exactly what you mean there.
Going for the "memory-hard" angle is a peculiar choice for an encryption algorithm - you generally want it to be fast, but you're deliberately slowing it down.
You still haven't said anything about the encryption algorithm itself - with the Feistel network.
temporal fortress is my internal term for the memory hard key derivation phase. iām aware encryption is usually optimized for speed, but the "memory-hard" choice is a deliberate trade-off to increase the cost of ASIC/FPGA based brute-force attacks during the key setup. once the subkeys are expanded, the actual encryption remains fast.
As for the Feistel Network itself:
itās a 16-round balanced structure. the f-function integrates a GF(2^8) S-Box, an 8x8 Circulant MDS Matrix, and Arnoldās Cat Map for bit-permutation. this ensures that even with a feistel structure, we achieve full diffusion much faster than traditional designs. essentially, I'm using SPN-style diffusion layers inside a feistel round to maximize the avalanche effect.
good lord
I made a random security bot just for fun and I'm finally done
š¤
exactly. youāve summarized the architecture perfectly.
The core philosophy of lox-617 is to use the Feistel Network as the secure "container" while utilizing SPN-style primitives (MDS and S-Box) inside the F-function to solve the slow diffusion problem typically found in classic Feistel ciphers.
By integrating Arnoldās Cat Map specifically for bit-shuffling, we ensure that the avalanche effect isn't just limited to byte-level changes but reaches bit-level chaos within the first 2 rounds. The Memory-Hard setup is simply a modern "gatekeeper" to ensure that the speed of the cipher serves the legitimate user, not an attacker with an ASIC farm.
I'm currently benchmarking the exact Avalanche Coefficientāearly tests show it hits ~50% bit-flip probability very rapidly

i got pinged here?
Hi sora
So - is this one algorithm, or multiple? Because you're describing a KDF and an AEAD here. And AES256GCM is already Post Quantum, so there isn't a lot of motivation in the crypto community right now to expend any energy on symmetric ciphers.
Ping
Where's your peer review? Where are your sources? Where is the cryptanalysis?
63 hours..
thatās exactly why Iām sharing it here. as an independent research project, Iām currently in the pre-publication phase.
The cryptanalysis and formal peer review are what Iām looking for from the community. Iāve documented the internal logic and mathematical primitives (MDS, Arnold's Cat Map, GF arithmetic) in my technical spec, and now I need eyes on it to find the weaknesses I might have missed.
As for sources, the design draws inspiration from established concepts like Rijndael's MDS diffusion and Argon2's memory-hardening, but the hybrid integration is my own. I'm not claiming itās a "standard"āIām presenting it as a candidate for review.
Presenting it to whom? Also it's strange of you to refer to this as a Feistel cipher and at the same time make multiple claims of drawing inspiration from Rijndael - which uses a completely different network and is not a Feistel cipher at all
If you want peer review, show me the spec.
Could you at least make a thread if you are going to post LLM slop
i'm aware that Rijndael is an SPN and not a Feistel network. when I say I draw inspiration from it, I'm referring specifically to the mathematical primitives, not the network topology.
LOX-617 uses a Feistel structure as the outer framework, but inside the F-function, Iāve integrated SPN-style components like the GF(2^8) S-Box and the 8x8 Circulant MDS Matrix (similar to the MixColumns transformation in Rijndael).
The goal is to bring the high-diffusion efficiency of SPN into the reversible structure of a Feistel network. As for "presenting it"āIām presenting it here to the developer community for initial peer-review and technical scrutiny before I finalize the formal paper.
lol what
Pong
š¤Ø

š
Okay so let's see it then?
Where .env
The goal is to make the containers, across multiple messages, with different/variadic text, a consistent width.
Eg this, to me, looks like hot garbage all jumbled up together.
I've said it, sir, I said earlier that it's still not 100% and the documentation is still lacking, now only rust C is working yet
I'll let you know later when it's done
a lot of the info there is unnecessary, the real issue is clutter imo
I didnt use a .env its in the config file
Why not .env
Dotenv files are cool
Unnecessary how? You don't even know what it does or it's purpose so your statement is entirely unfounded....
cause i just made a config file and desided to add it there
Does your config file contain stuff other than constants?
e.g. you dont need all of this, its all redundant
I'm not a "sir", and no you haven't said anything of substance - you're just tossing around LLM-generated crypto jargon without any real understanding. If you had a spec, you could produce it in 10 seconds. You don't.
yeah it stores everything that can be configured
No that's not redundant, sometimes it shows @unknown-user so the other fields make it possible to track it down. The display name can change, and the username is the source of truth
If you know how to make it always show the user with the mention, I'm all ears, but I've never been able to get it redundant, and this is an updating application to data truthiness is very important.
The discord ID is important in case someone wants to create a report/warning for a user that may have left the discord through another system, a record that can be trusted on member leaves.
But it's clickable though? The mention, that is
plus yea the mention is clickable
The mention will not expose the underlying information if it's in "Unknown" status
^ the link would work
iām not trying to hide behind jargon. here is the F-function logic directly from my source to clarify the design. Itās a work in progress, but the implementation of the MDS matrix and Arnoldās Cat Map is already functional
// Simplified look at the F-function logic
void f_function(uint8_t *half_block, uint8_t *subkey) {
// 1. Non-linear layer
sub_bytes(half_block);
// 2. Linear Diffusion
mix_columns_mds(half_block);
// 3. Bit-level Permutation
apply_arnolds_cat_map(half_block);
// 4. Key Addition
xor_with_subkey(half_block, subkey);
}
iām currently finalizing the LaTeX documentation for the full spec. I'll share the PDF/docx as soon as the Test Vectors are verified. I'm 15 and still learning, but the math behind the MDS and Feistel structure is what I'm focused on, not just "jargon".
im not sure if it works on the pc app though
I know it does work on the mobile
there are tons of viable workarounds too. for the purpose of showing the outcome of a form (?) that sort of info is not important, and i assume you're storing the data elsewhere so you can retrieve later if needed. a view should contain the relevant information for its purpose, not everything for the sake of it
It opens Chrome lol
The link is a thought, I'm going to see if that's viable. On PC it's rendering as the full link, so from a display perspective, it'd be a terrible user experience.
š only a mobile feature then?
link buttons/hyperlink
yeah it prolly doesn't work on pc
ok mobile it opens up the user profile with no lag
if you're genuinely interested in learning about cryptography, that's an endeavor that I can support, but please don't send LLM-generated content in this server and expect that people won't see right through it
i get why it looks that way, especially since I'm using english which isn't my first language to explain complex stuff. but the logic is mine.
for example, implementing the 8x8 MDS Matrix in C was a headache because i had to make sure the branch number was optimal for diffusion. or how i had to adapt Arnold's Cat Map (which is usually for image pixels) to work on 128-bit half-blocks for bit-shuffling. thatās not something i just "generated" i spent nights debugging why the avalanche effect wasn't hitting 50% in the first 2 rounds.
iām genuinely trying to learn. if you want to support that, I'd appreciate a critique on the MDS multiplication logic or how i handle the Key entanglement once i post the snippet. no LLM, just me trying to build something cool
That's totally fair, but most of the information on those forms is hidden behind that "Info" button, this is just the quick access information to help staff onboard users safely. It's an onboarding system with manual approval processes. It also has information for when users are inactive, have been banned, and various other functions/features to help discord servers keep abreast of guilds, communities, etc.. Users have expressly asked for that information to always be front and center, and every attempt I've made to either reduce or move has resulted in backlash. So even if you're right, the user wins
is anyone here able to help me with official discord support
Sure, I'd be happy to look it over
!support
This server is for help with Discord APIs, SDK and other various integrations. For general support, please contact Discord at https://dis.gd/contact, or send us a message on Twitter at @discord_support
appreciate it. Hereās a snippet of the F-function where I integrate the MDS matrix and the bit-shuffling.
i'm specifically curious about the MDS transformation I usedāI implemented it using a circulant matrix in GF(2^8). Do you think an 8x8 matrix is overkill for a 128-bit half-block, or does it provide a significantly better diffusion margin compared to the 4x4 used in AES?
pub fn lox_gf256_mul(
mut a: u8,
mut b: u8
) -> u8 {
let mut p = 0u8;
for _ in 0..8 {
if (b & 1) != 0 {
p ^= a;
}
let hi =
a & 0x80;
a <<= 1;
if hi != 0 {
a ^= 0x1B;
}
b >>= 1;
}
p
}
pub fn lox617_mds_apply(
block: &mut [u8; 8]
) {
let mds: [u8; 8] = [
0x17, 0x31,
0x59, 0x83,
0xA7, 0xC1,
0xD3, 0xEF
];
let mut temp =
[0u8; 8];
for i in 0..8 {
for j in 0..8 {
temp[i] ^=
lox_gf256_mul(
block[j],
mds[
(i + j) % 8
]
);
}
}
*block = temp;
}
Also, Iām still fine-tuning the Arnold's Cat Map implementation to ensure it doesn't introduce any unintended periodicities in the bit-shuffling. Any thoughts on that?
code snippet from rust
How much storage capacity does a Discord server have?
it does yield more diffusion, but the compute is higher - real tests would show if it's worth the tradeoff. how did you derive the 8 constants in mds?
Wrong server; no recruitment
This seever have somw good staffs i guss
no hiring/recruitment here
Not the server for that
this is a dev community not a staff community + recruitment not allowed here
Ok sorry
For everyone?
for the constants, i aimed for a circulant MDS Matrix to keep the implementation efficient while maximizing the branch number. i selected the coefficients in GF(2^8) to ensure the matrix is non-singular and provides an optimal differential and linear branch number.
to be honest, I'm still refining the search for the "perfect" constants using a script to verify the MDS property (ensuring every sub-square matrix is non-singular). currently, I'm testing with a set of constants that minimize the number of XORs/shifts in the multiplication to keep the performance hit from the 8x8 size manageable.
do you have a recommended method for verifying the branch number more formally, or perhaps a specific generator polynomial you'd suggest for an 8x8 over GF(2^8)?
bluishpurple gradient moment
Blurplish*
Remove the site
Faah
The screenshot tells nothing what's its about
seems like a discord copy
Ahhh the confidence
will do this after eating :3
:horror:
what's up with the horror š
i was just getting annoyed with writing if else everywhere there was an optional enum
What if you have taste of blue color
You can just say to bypass that words used on you
š
Guys, also are you still counted as dev/coder when you paste codes as from snippets or tracing an AI art to draw it, is it making you artist too?
Holy @daring wasp
I wish Sora was not deprecated
š¤¦
čæęÆååäŗåļ¼ęéå°čæēåļ¼
post https://discord.com/api/v9/applications
{"message": "ē¼ŗå°ęé","code": 50001}
čæęÆååäŗåļ¼ęéå°čæēåļ¼
try a different name maybe? and this server is English only
This has to be the absolute cleanest onboarding experience ever xD
smooth
it is, wow.
I would not read all that
Only took me 2yr of coding to get it to this point lol.
You wouldnāt read 2 sentences, 3 at most? Thatās crazy work
(also most of that is literally just replicating onboarding but worse)
Youāre a troll clearly. Onboarding is ass, and any serious discord knows it.
Letās auto-hide channels when no one asks for it. Letās surface channels when staff accidentally moves a channel. Letās hide all channels when a user has it enabled because a channel was created or deleted.
would you read all this
no
Also, this function has manual approval processes and moderation you donāt see going on behind the scenes.
not really, its good for a non programmer and offers enough customisability for an average discord server
bots are more customizable
^your hide channel point is very understandable i dont like it either
it's like windows babysitting you
outside of the third container?
The application flow is the only reason I built it.
the third container does seem like it's not just onboarding, everything else seems like it very much is
i just looked at message object so scary š i have to implement it (partially)
Holy text
One word⦠āUsersā
clearly havent seen this
all or that is duplicate
š
well true people tend to forget that discord ui exist when they read dev created content
like the "hello" automod for example
I think most people understand ephemeral content just fine, esp. qhwn the message already says it
also the text literally just lists the buttons right below it for some reason
Hopefully that makes my point however. It puts discords onboarding to shame š
it's all duplicate text
it really doesn't
Ragebait much? Muted
theyre both equally viable
most of it is actively worse and you have a single feature onboarding doesn't support in there to justify replacing an entire flow
that's what we all do
not me
create a whole new thing to add 1 single feature to the existing one
I agree it has its place.
like this totally isn't from risky's honeypot
(it isn't, it has a different filtering)
most of what was created here could just be done using the native feature, and it'd be easier to just have this one bit as separate than reimplement the native feature
because native re-implement what was missing, but only when we start doing it ourselves.
Don't you love reinventing the wheel?
to be fair, "honeypot" is not exactly an novel concept
i mean it should be called flycatcher or something
you realize risky didn't invent that design or the term right
honeypot is a technical term :p
i know...
though tbh I'm not quite sure these things qualify 
I first seen this term when visiting RIT and looked it up.
I've seen servers employ similar things with bots long before risky's honeypot was a thing
they're not exactly trying to appear legitimate, at least not to normal users
and yeah it's been a thing for a while, inc. with onboarding roles
Hi guys I'm back
You are so ignorant itās wild. Here is everything it does.
- Auto role assignments before and after application
- Manual application process with form collection.
- Application status DMs
- Inactivity Monitoring
- Reporting Monitoring
- Per app notes
- Application history and frequency
- Analytics tracking
- Data & Reporting
- Cross-server watchlist Notifications
Basically a whole bunch of stuff that onboarding doesnāt dream to do and certainly more than āOne thingā.
Itās built with the express purpose of preventing people hoping discord to get a guild invite in a game or similar, and leaving making staff wonder where someone is or if they should or should not have access.
whoa whoa whoa
I don't think he will read all of that 
I actually did 
morning chat!
morning
Morning
but I'm grouping all of that as a single feature in my description nonetheless
morning guys
Okay
that's a little uncalled for I'd say, also I disagree
and my comment was specifically referring to the third container in the non-native onboarding flow as the only unique feature
Heās attempting to rage bait and assuming with absolutely no context. Itās deserved
I don't think it's a bad feature (though not like I've used it), my issue is with the rest of the containers, which I do think should be replaced by the native onboarding flow
Thatās a much better way to state it. Be clear in your intention or other people will find you to be a problem.
Bro can we make bot from vibe coading
I wouldnāt disagree with that statement, if youāre okay with onboarding. Personally I am not
Sure, until it breaks š
specifically the role selects here could easily be done natively, and the feature onboarding has for showcasing channels works great also
No it not brake it work
AI is bad at it
Why not learn how to code without using AI
In my opinion, just use a public bot
Or actually pay someone to make an actual good bot without a thousand vulnerabilities
So vibe coding is not good
AI has like months old api reference unless you explicitly point it to the docs
no, its not for anything beyond prototyping
If you know what youāre doing AI can help, but if you expect to get away with no coding, youāre gonna burn some serious tokens over one little thing not working. It wonāt happen early, only once youāre invested, and then you can be left with slop and nothing done at the end of the day.
Criticism of someone's design, approach, or opinion is valid as long as it's delivered in a respectful way, but personal attacks are not welcome
Itās gonna be a ā;ā out of place and youāll spend $200 and 4 days finding it xD
^ yep it can only help you if you already know what you're doing
Or itāll DROP your database, if you can manage to connect it lol
Actually things like missing semi colon would be solved by it reading the error from linter/lsp. Agents nowadays don't reread thru all the code to spot the error first
> "Fix this code"
> Comments all your code
I was making a funny
Pure. No side-effects. Immutable. I approve.
# :rocket: Super Fast ā Production Ready Version
š¤®
In fact I think all code should just be commented out
Yes. This is the way.
Seriously, AI is really good at simple, repetitive things that are very well defined. Eg interfaces and typing support.
Itās also really good if you have a strong architectural pattern and coding standards.
If you have neither, your only hope is to tell it āOne shot and make no mistakes
I one shot a whole python library for Discord with Claude
Claude, gimme a million dollar app
and then 800 errors
šš
How do you even get Claude to run that long without absolutely nuking your tokens? Iāve heard of people having long running sessions but no clue how they do that crap
madeline is secretly anthropic ceo
it failed a few times and I just clicked the button "keep trying"
by the end of the day it was done
That was peak
Yeh, you know it doesnāt work right?
but it does
Tests not required lol
the JavaOrJavaScript language by claude was peak
the point wasn't really to produce something useable
that was brilliant
Matt did something similar too, I think
I give myself a lot of credit though for the design decisions on JSoJ
one time I timed my usage so perfectly it reset at the moment I hit full, retried once and I was back to 0%
by pure coincidence
what timing
I had claude and codex do some flag hunting in a competition while I find another flag alongside.
Ive never really run out of Claude Usage even for the free plan even since i dont rely on it too much
codex usage hit 100%, retried once, and it went back to 0%
for the python lib I actually told it not to do any testing cuz I knew it would have no chance of completing if it tried to test
nowadays claude consumes more token if your prompt is big
i wanna use codex but id need to use vscode instead of code oss for that
that is literally how tokens works
you dont have to
you can modify the marketplace url at a risk of microslop possibly suing you
I should've frsmed it this way
codex doesnt work on native termux (where i use code oss)
id have to use proot to simulate glibc to run vscode and codex
although i havent tried the extension version, the cli just wouldn't load up
makes sense, most things wont work without glibc except those that use musl
android uses Bionic not musl
so basically i can't run anything closed source without proot sim
yeah, that's what I said, no?
Did anyone of you sell a bot and how much money you take
I did not realize the existence of bionic so I only knew musl and glibc. My bad then.
you should try ubuntu touch
its not common to know about it
Bionic is a specific lightweight one only used in android
So it was not used in other
open source stuff usually you can get to work
because you can compile/tweak it for android
https://discord.com/channels/613425648685547541/1505074485177614436
sorry, i have trouble
Don't crosspost it here
hi inbestigator
you can use the push api to implement server events without a continuous connection
sorry, i japanese , i don`t know
You are japines
š
Did you live in japan
watashi ha nihon ni sunde imasu
Bro what are you typing
Japanese. This is an English only server though @dense bone
#rules Only english is allowed here, you can use a translator to talk in english
Can you use translator to under stand@daring wasp
They said "I live in Japan" that's it
That's offtopic for this server
for general talks you should go in https://discord.gg/discord-townhall
@daring wasp what app did you use to code and make a bot and like use it to make a working bot
I use VSCode (GUI) for long sessions / fresh-editor (Terminal Based) for quick edits
So which is batter for big project and is it free
Ok thanks
šš
Did you know which is best coding language to make a bot and big project python or js
Whatever you're more comfortable with
There is no "best"
same.
do you know how to get autocomplete as i type rather than ctrl + space?
Like I'm learning so which I learn
that's a big reason I'm still using vscode
oh wait, as you type
Take a look at both
Python is more english-y like you can literally read it
JS is used in a lot of places like webdev
Ohh I hadn't seen this
So like js Is more good for development so I will learn J's and it is also used to make bot righy
Soory right not righy
Python is also the same. It's really up to you which language you're gonna learn and use
Python is also used a lot
You're also not limited to just JavaScript and Python
there's other languages too but I feel like a high level language like python, js is better to start
Yeah that's right
So I will learn J's than python because J's is like use more in web dev
js isn't better for development
neither
there is no "better"
Bro like which is batter for web dev
Hello I want to ask for an anti nuke bot
no
both can do bots, both can do web, both can do backend
js is used a lot in webdev for running client-side js, and it also has some popular backend frameworks, but so does e.g. python or c#
Uhh i mean I want an antinuke bot to protect my server
we can't help with that
Why
"anti-nuke bots" are inherently dangerous; you're giving some random person admin privileges in order to provide the same "protection" for your server that you can have by simply not handing out high level permissions to people who aren't trusted or don't need it. if an anti-nuke bot actually did what it claimed to do, the first thing it would do would be to nuke itself.
but isn't that what it's supposed to stop? 
using "nuke" loosely here
it's true that you've uncovered a bit of a paradox though @neat olive. if an anti-nuke bot is also a nuke bot, can the bot faithfully execute either of its roles without compromising the integrity of its identity?
<@&1050493473033289778>
They can see logs bro
š
Deleting message doesn't make you safe
soviet air defense
š?
Hu
hi samtino
Thx mods!
-# I didn't see what it was
server adv
Ahh
Same Timor
https://en.wikipedia.org/wiki/A-35_anti-ballistic_missile_system for context (they really wanted to nuke the nukes mid-air)
(to be fair it's not an exact analogy since the idea was to nuke other nukes instead of yourself, but still)
This adds complexity though, because in our scenario we are the soviets, a nuke is heading our way which was fired by the soviets, we have a nuke which we can use to nuke the nuke midair, and we are the soviets
I am using railway to keep my bot online but it says that i have 17 days left(Until my free subscription expires). Do you guys know any free bot running website?

how should we act, if we are to act in the interests of the soviets?
some hyperscalers have free tiers
follow a guide to setup a free tier vm on any of the big cloud providers e.g. aws, azure, gcp, they each give you a year for free
I think by historical standards... we break
maybe the paradox solves itself
perhaps you're right. we CANNOT exist in both states simultaneously, therefore we MUST not
I was going to check the website but i saw the comment
I'm not even surprised at this point
@misty cipher
seems like they arent really in a position to be making educational content yet š
Failed to send DM to user: DiscordAPIError[40003]: You are opening direct messages too fast.
Failed to send DM to user: SyntaxError: Unexpected end of JSON input
@modest geyser it must have been a spammy night to get these errors
ive gotten it way too many times i can't tell whats going on
like am i having too many paralell connections to discord and multipliexing gets confused?
or does it somehow end before actually receiving the last few chars

Please open a github issue about that. The error may be intentional but it shouldn't give you invalid json at least
oh they are 2 diff events
don't you have a 100ms queue
let me check if i have it uploaded on github sec
but im not sure if its due to me aborting it after 2.5s and maybe Bun just cahceling the reading of body or whats wrong
there are alot of possible things and i don't have enough of a repo or often enough to be able to definitely blame discord
something like this https://pastes.dev/lJ46KPPLlY
should do the trick most of the time
so my bot works indeed in a queue from redis list that WS pushes to, but not sure what has to do with this
the http sending is done by @discordjs/rest
queue the HTTP sending to prevent excessive dms in a short span
oh also docker for all we know could be funny - there are many layers of proxy to discord themself
oh, uh i only try for 2.5sec coz i dont want to delay the dm for too long
like once they are removed from server, there is a low chance i can dm again
i use the custom made queue for most stuff (i have it set as 100, not 1000)
i only try for 2.5sec
š¤ so if fail you try again and again? or am i misunderstanding something
i do cache the route for getting channel id incase i ban people in 2+ servers quickly
like i have an AbortSignal.timeout
which means that the fetch query will abort after that time if its not completed
aha, not what i mean
the design choice of mine is as following:
- user does something, needs to be kicked (softbanned + message)
- bot adds to queue that this exact user needs to be kicked
- the queue pretty much runs one by one, prevents all of them to be ran at once
- if message fails i abort, but still kick
does it make more sense now
i mean isnt that similar to what i do?
like whats a few ms differance of sending the event going to do
its prob a hourly ratelimit or smth
prevent spamming discord's api most likely
š¤ interesting
or at very least one that isn't counting paralell connections
it's a token bucket algorithm, the frequency of the requests is irrelevant, it's about the number of requests made during a sliding time window
ya
i mean the way i handle it is, in a way, so that the number of requests made during a sliding time window decreases
e.g send message at 0, kick at 100, inform in specific channel at 200
I really doubt this would have any impact
i don't see any other solution for this to prevent exhausting the requests in that number of sliding time window
Btw how many events are you sending?
if you receive a rate limit warning you should just throw the request into your queue and have it retry the next req in the queue (or batch if API supports for that type of action), if successful, continue to next in queue, otherwise try again when the rate limit advises you to
Ya d.js already does this for me
the queue should be allowed to accumulate as many backlogged requests as you can handle, and process them when possible, even if it's minutes or hours later
im also on djs with the queue included to prevent heavy amount of logs being sent at once/prevent any unfunny/unsafe non-queued (bypassed?) processes if there is one by mistake
hmm
(in this specific scenario btw; there are other scenarios, not really bot related, where you want to avoid queueing requests)
Realistically in my case (the reason for max time spend for dm msg), its indeed only ban that i want to get this level of queueing if needed.... but i dont know if ive hit much ratelimits there
i don't think i see any scenarios where you'd want to avoid queuing requests to prevent 3rd party/your own servers from getting exhausted
DM is the only one you're gona have trouble with
isn't DM considered as 2 requests sent in same time:
- opening dm
- sending msg
wdym. there is no "open dm". it's just sending the message
huh
I could be mistaken but afaik it's just a normal message send?
you need to get dm channel first
wouldn't you have that cached
if its not cached
if it's first time, it's not cached
oh ig not for this scenario
ie youve never dmd them before
i dont like how that event doesn't check if you can even dm them to begin with - like you can get a channel with anyone... but its only the send msg that will error if you dont have mutual
yeah pretty much each time he messages someone it's 2 requests being done
anyways probably irrelevant as i doubt those are in the same token bucket to begin with
i mean everything is part of the global bucket, but you're not gonna hit that limit
i'm pretty much preventing myself from reaching both the global bucket limit and any weird limits that may suddenly show up but i'll consider what was previously mentioned
Failed to softban user: DiscordAPIError[10026]: Unknown Ban
im impressed by some servers tho... as either this is due to ratelimiting or a bot insta unbans before i can (?)
the delayed requests strategy here doesn't really do anything though - because in the mean time you could just be receiving more events that require requests
i've honestly yet to have any issues with it, rather, it's been much more helpful, though my bot is mostly logging (messages, guild members, user profile, etc), it does have softban ability (message, ban, unban)
i'm not saying it would cause issues, just that it has zero effect
so guys if i have a seperate script for proxying discord reqs for ratelimit (ie so i can have many procs of event handler)... should i transfer an abort or still send the request? like for context any update to code and such would restart the bot and thus nullifying in process stuff potentionally
yeah i agree

mk, oh well, i mean i can still keep it for the log system i've done, but seems like i need to find solution for others (a bit of an example, for context, is that it sends attachments or long messages in separate messages, for easy reading + follow discord limitations)
thanks for letting me know of the issue(s)
the lifetime of an abort signal should only be as long as the request takes to resolve (or timeout)... if you get a 429, the abort signal should be canceled just as you would with a 200 response. then you create a brand new copy of the request and pass it off to your queue, along with some kind of indicator to the queue about which bucket it's in, when the request should be retried, etc. then whatever part of your system is managing the queue makes sure queued requests get sent out in the order they came in as rate limit timeouts expire
why not upload the attachment in the cv2 msg?
oh ofc the proxy would handle retries, but i meant if the internal connection dies should i cancel the actual discord one?
as some stuff may have later dependents for all i know
the best idea is to try and let docker keep it for as alive as needed (not accepting any new events) and idealy by then not have any stuff left
wdym by internal connection? between your main program and the proxy?
ya
does the proxy ever receive responses, or are those directed back to the main program
ig my question is: the proxy receives data from your main process, but does it ever send data to your main process
yeah it would retry as much as it needs (idealy 0 coz good ratelimiting data) and then send data back
i mean logically i dont actualy often read the response, but sometimes i do
<@&1050493473033289778>
hmm... that design feels shaky to me. if the proxy is a separate service, it should be capable of handling the request and response all the way through - no backwards communication
wdym? you are saying it should never get the actual response?
and for those i send directly?
if it does, it should be able to handle it
i.e. sending a request to your proxy should mean "this is no longer my concern"
like the whole point is to avoid saving thate ratelimit data twice and potentionally having outdated info
i could try to make my own cursed setup and store that data in redis i guess, but i dont wana fork any ratelimiters
honestly rarely the case as i do at least check the status (ie if error code then i can send msg that i failed to do action)
i would question the two-system setup first. it's unlikely that you actually need it
up to a point, be careful you don't get into a situation where the proxy infinitely waits on timeouts of 24h+, keeping them queued up even after the requester gave up on a timeout. or it does retries, keeps getting 429s and then the downstream app their own retry logic kicks in to amplify the mess once the proxy finally gives up
i havent needed it yet, but im just thinking about it in future you know (like how do people handle clusters with ratelimits - or when restart pesist data to avoid making discord mad)
i do get for clusters its not needed as generally (not my setup tho) you'd be using sharding properly where each handler is for a specific range of servers
oh for sure. there must always be an upper limit on the number of retries, and it's context-based too - for some cases, you'd wanna keep it very short. if a request is going to sit in the queue for awhile it should be side-effect free in relation to your whole system, basically
my setup is diff as i condence all shards into one event handler queue as i dont have really any shared state/cache i need
any wone plss help me
retries and awaiting the rate limit sounds convenient in theory but i dunno if id want to be dealing with the potential for requests to take unexpectedly long vs basic error handling
With what
anyone got any idea if ip bans are a concern when only dealing with websocket connections, where no api requests are being sent from the same
e.g. if you have a bunch of bots all connecting off the same ip, is there a limit where discord decides nope
probably not a major concern at least, I don't know if it's possible to get ip banned like that 
How to get the š± badge
You can't
Not anymore?
Nope, you had to play april fools game on discord
i dont know if it being a thing, but i feel like it wouldnt be surprising if there was some sort of limit before things get flagged
Ic ic
e.g. if 1000 bots are all connecting from the same ip, surely discord would think that looks a bit sus
I don't know if any of the events you need to send for websocket connections count against any relevant limits but if not there wouldn't be any limit except how many connections you can actually open at once from a single ip (presumably you need a port per connection) 
i havent heard of any shared hosts having issues there, but they'd first get issues because of the get bot info would error
(ie to see how many shards and concurrent)
even that, you'd be doing rarely, definitely not to the frequency that would cause too many requests from the single IP
The only one that would is /gateway/bot
protocol translating proxy
like the dev tos is pretty clear against the idea of hosting other peoples bots thing (if thats what you are doing)
Mmm the wording is intentionally vague
is there precedence for that being the stance?
considering mee6 is alive and still offers it
but i heard they gave us per server bot profiles to minimise the reason to want
mee6 being able to do something doesnt hold much weight tbf
yeah
tbh i have a few times wonderd if i can make a service which translates gateway events into http events so people can do serverless to their hearts content
exactly the thing im wondering. that or reversing the client/server ownership to properly support ws hibernation on cf workers DOs
ya
issue here is the events can get very spammy so like even if you can do serverless its prob a bad idea
the thing with the IP bans was that it could support a free shared IP for some number of bots, and if there is any risk there, offer paid dedis to get around that risk
that wasn't a thing before, and, I kinda like it this way too
ws hibernation
you'd not last long because like every x infinitival you have to send heartbeat
my intention is to do all the lifecycle stuff, just forward the meaningful events
and also you get alot more events than you care about so just waisting cpu time
ah
at very least you need to get them to select the events they want as otherwise itll get messy
would abstract away sharding entirely, you'd just get events through your single point you can do all the easy load balancing on to your hearts content
I need someone who knows how to code a WhatsApp bot im having some problems
tbh any bot using this (and not their own system) prob doesn't need to deal with sharding
2.5k servers is alot
i dont see any reason that'd necessarily be the case
well when you get to a state of geting heeps of events, the serverless cost would get high to make you question why
I'd honestly do anything to make it so I don't have to shard my bots ngl, but to be honest, that's the easiest part of all
if it were trivial to make a (for all intents and purposes) infinitely scalable bot, that sounds like a pretty good value proposition, like any other serverless platform
my bot is effectively serverless idea (ws -> redis queue -> event handler)
clustering is what I'd struggle (multiple nodes) as I do not have a proper way to handle clustering
sooo basically you dont want a big bot
imagine being serverless I could never
-# except if you look at me pushing serverless and writing an entire guide on it
you could then switch over to containers and continue operating the exact same way
yes/no
there's a lot of pain in this
i use docker
the main benifit of my strat is i can restart my event handler often and not have to restart/resume the actual ws itself
as that stays steady, and then i use docker swarm like one could with http server and always having at least one alive at the moment
yup, that's the beauty of decoupling
honestly i think only few bots can do that because most actually rely on caching
and when you cache, may as well just keep everything on same cluster
oh also whats a normal events per sec for a 9k server bot would you say
sharding between 2 nodes for me rn is a bit of a impossible to do, I don't know when the bot is supposed to start/what happened to it/etc
and having to handle cluster rn would mean I need to think of an effective system to easily broadcast data if necessary to another cluster (keyword: if; it doesn't mean I will, I don't think there is any day I will ever do that too)
sharding is part of what makes clustering not so trivial for a bot, but if you only needed to load balance across webhooks, a message queue, etc, then you can do LB like any other service would

basically the same idea, just that i think itd be cool to make it a managed multi-tenant service
above all else, its an excuse for me to have fun with durable functions and getting in the weeds of some cloud architecture stuff š
-# arent you meant to be a aws fan
i use aws stuff at work but azure stuff at home
is it me or are you using redis as an alternative to "websocket", in a way you're doing as if:
discord -> main service -> redis -> external service (handles events only) OR external service (handles commands only)
this does seem a bit like it (though, not fully correct)
More like a message broker
if just commands only, just set http interaction url at that point
aren't websockets more efficient at being a message broker than redis or is that not true
I mean yes, fair enough, but I couldn't give another example currently lol
would it be possible to know WHEN to and WHEN NOT to?
message brokers can provide reliability that websockets can't without extra work
weirdly, it's one of those things that you'll know when you need it
there's definitely a "checklist" you could use...but it's just more instinct too...and architecture
ok so ill explain my goals
- every time i update code, i don't want to have to deal with pain of downtime - whether that is directly needing to re-identify or still connect you still have some time where nothing can be processed
- means i can filter my events before hitting event handler
- they can scale interdependently depending on what uses more resources (tbh i thought WS was cheaper than event handler)
I mean I usually only rely on either socket or websocket
also it means that if event handler crashes, then i still have a queue of events to replay through on restart
(this was before i knew you could just save url and reconnect)
honestly i prob wouldnt have bothered as much if i knew this single fact
wdym by filter your events before hitting event handler
also somehow there is a substantial tls or something overhead that i dont know where - as its not fully json parse as when i proxied everything the cpu usage on WS was way higher
I see
like i dont care about MESSAGE_UPDATE or messages from a diff channel (to what i subscribed to) for instance
so why bother sending smth that'll just get dropped
yeah websocket is a bit of ? when it comes to CPU usage, though redis could end up worst if not handled properly
redis is just a lpush and blpop (resolve when there is smth to pop - and run this in a loop)
uhh, doesn't that also work the same way in websocket
Am not from china
im just saying how im doing it (i store the channel ids i care about in redis and check before adding the event to the queue)
the important distinction is that by abstracting away the websocket, you're free to use whatever tool best suits your priorities
rather than being locked into one architecture you can do things however makes most sense. and compute efficiency is only one metric, that often is not actually that important, especially when weighing up other priorities
H
robo's back online
if that was the only thing that mattered we'd all be writing bots in C
I mean yeah but it's nice to know the advantages & disadvantages of using websocket vs redis as a message broker
you say this, but im literally making my ws thing in rust for better cpu usage (3% -> 0.5% and doing more)
3% local is 10% on vps
the main benifits are due to not fully parsing the json which is faster and has less GC cleanup requirements
why am i always that one unnoticed guy
i havent personally looked at using redis in that way but there are countless other approaches
one downside is you've still got a bottleneck, you've just shifted it off to your redis instance š
haha true
i could make the redis myself tbh
just put direct socket connections and then choose randomly which to send events to
to me this seems like a lot of effort to replace something that ought to just work fine already
idk this is fun to me to make it peak perf
idk about leymbda, but its intresting to work out what parts can be done to improve effeciency and what gains/downsides it yeilds
in terms of performance this seems like it'd always just be worse than just sticking with websockets, no?
at my current rate of event spammsing, ill eventually need to do multicore or try and reduce the cpu buttlenecks earlier on
all the possible benefits would be at the cost of perf in my eyes 
i cant remember whats being argued here
but my method of ws and event handler being seperate has worked very well for me
for me its about the convenience and flexibility of being able to host in ways not otherwise possible
ya
if you wanna chuck all your bot code in a lambda and call it a day then you can
the idea behind https://pylon.bot is nice, so if you can easily just make a simple bot without needing an actual server its nice
pylon...has its own issues
including the fact that you needed to use their libraries
from what I recall
i said the idea behind it š
looks abandoned too, which would be the biggest flaw of any sort of service like this
@solid wind's solution is just proxying the events and letting you handle however you see fit
but it does miss out on cache
you'd still need an actual server proxying everything 
unless you'll let you subscribe to added data with cache
cache can also be in redis 
thats what they are considering making š
you'd be implementing caching on the bot's end not the proxy
it just seems like there's just more and more complexity rather than an actually simplified workflow
but one can't simply cache well in serverless
but
maybe I'm just not seeing the vision
redis š
like unless you want to get a bunch of waisted member update events to handle and send only to cache
honestly the serverless pricing on some things are insaine compared to what you can get on basic vps
guys init_arena or arena_init
(if using serverless we'd also now be paying for both regular hosting and all the serverless use)

funciton name
my idea is to make it a managed service rather than a self hosted one, which has its own pros and cons too, but at least means you arent managing your own VMs
new Arena
huhhh
ya, i gathered this :)
and honestly its a way better idea than actually running peoplz code
all you'd be doing is setting up an extra third party service then pointing it to your lambdas or whatever, so all you're managing in terms of infra is your lambda
id honestly ask discord support on if they'd condone this as the dev tos does kinda allude to it being a bad idea
they dont answer policy questions like that
like if you store encrypted and handle ratelimits... you in theory aren't actually consuming any of their payloads yourself
wait really? i that everyone always says ask support for those cases here
my naming convention is so inconsistent I just realised
Arena *
init_arena()
void
*arena_malloc(Arena *arena, const size_t size)
arena_init
oke then
i did some number crunching a while ago and figured that it'd be feasible to do a model like $1 per shard, and while thats certainly more at greater scale, it is predictive billing, and ends up cheaper for most when compared to non-free-tiers
thats alot of dolla just for being a proxy
like in theory an actual shard itself shouldnt really cost you that much
now if you can cache and send other data with it, that changes things (but ram can get really high so prob not a good idea)
way i see it is thats $1 and effectively free serverless vs a few bucks on a vm
at least have $1 give many shards
and have free teir with max events proxied per mo as trial
not to get too into the weeds of businessy things as opposed to the actual technical viability, but that would also afford a single shard free tier
i'm not using the ternary operator enough man
dont go crazy with them, you're better off underusing than overusing
what are the downsides?
can be messy and unreadable
fair
interesting
Y.
in python we do something similar i think
Looks pretty clear
How should this situation be resolved?
!support
This server is for help with Discord APIs, SDK and other various integrations. For general support, please contact Discord at https://dis.gd/contact, or send us a message on Twitter at @discord_support
I want to try making my own Discord API wrapper in Python ._.
How should this situation be resolved?
Read what the bot said above
OK
Does anyone know how to unblock this?
by reading what the bot says and follow
this server cannot help you.
I couldn't figure it out on the Discord website either; here's what you need to do
For some reason, a few accounts are experiencing this issue.
but you can't do anything at all here. Are you using translator by any chance?
we're just people
not developers that work on discord, just normal people
I applied for two new accounts, and both have this issue. I'm not sure what's going on.
morning guys
please.. just stop.. we can't help you even if you continue to ask here.
Does anyone know the staff's contact information?
!support
This server is for help with Discord APIs, SDK and other various integrations. For general support, please contact Discord at https://dis.gd/contact, or send us a message on Twitter at @discord_support
this ^^^^^
Arena *
arena_init()
{
Arena *arena = malloc(sizeof(Arena));
if (arena == NULL) {
return NULL;
}
arena->size = 0;
for (int i = 0; i < DEFAULT_ARENA_SIZE; i++) {
if (pool_init(arena) < 0) {
arena_destroy(arena);
}
}
return NULL;
}
``` so dumb
Why is there a return type if the only thing that can be returned is NULL
exactly
i was wondering why it wasn't working
until I checked that function... so yeah
an iq
if it gonna return null everytime why all of this
fr lmao
Is this a serious question or is this a funsy statement
Is it clearing a space in memory? Not sure what the point of that would be tho
I just created my first python discord bot using discord.py coming from c# this wasn't to difficult
Python is cool
I have it running on my Ubuntu laptop server
I'm still newbie at it
Do you think you can teach me Python?
That sucks
no body likes js
Do you know how to exit the python repl
Anyways, iconic how these lines of codes worked on first try
import os
import discord
from dotenv import load_dotenv
load_dotenv()
intents = discord.Intents.default()
intents.message_content = True
class GamedayClient(discord.Client):
async def on_ready(self):
print(f'Logged in as {self.user}')
async def on_message(self, message):
if message.author == self.user:
return
if message.content.startswith('!hello'):
await message.channel.send('Hello!')
client = GamedayClient(intents=intents)
client.run(os.getenv('TOKEN'))```
exit (or exit() on 3.12 below)
is my bot
Why not using the commands extension
šÆ
very hd
because I havent gotten that far yet
Why using prefix command :/
I wanted the blurry ultra hd but it lasted like 2 milliseconds
It's the file/code preview, not the actual codebase
I know that :p
@olive minnow
this was to get something running to make sure I could connect to the discord gateway
ext.commands is basically needed if you're using prefix :p
Nuh-uh >:(
@olive minnow šššš
(or if you wanna use cogs)
which are really nice
Cogs are a pain
Cogs are wheels
true
I've never had problems with cogs
hey i like cogs
cogs...i still don't know that...
they're nice
Extensions are nice. But I have only used a cogs feature like twice
eh?
in dpy aren't they the same..?
no
Extensions are cool too, 'cause they extend
what's the diff
Greetings
oh right like load_extension can be used for other stuff too right?
Extensions are modules
Extension is the file with the setup function. It is what is loaded and reloaded. Cogs are the class that just has extra stuff like an error handler for all commands in a cog
i believe so
Because it will organize your help commands (prefix jumpscare)
In my newer bots I just don't use cogs.
https://github.com/Icebluewolf/Meowtivation/blob/main/src%2Fcommands%2Fview_goals.py#L112-L115
Ooo meowtivation
Cogs also auto add all commands and listeners within the cog. But I actually prefer doing it manually
pycord:3
i go eat now my hands tired from codin
Same concepts just different paths for where stuff is at
I have a question. My server was breached for two days and they said it would be back online on the 16th, but it's still not online. Will it be online tonight?
reply to your support ticket instead, discord does not have or do live chat support
I just asked a question, I didn't say you solved my problem.
it's an off topic question nobody here can even try to answer
Will it only open tonight?
nobody has any way of knowing
The only question is yes or no?
thank you again
Hello, could you tell me how to add a photo to the voice channel like in the screenshot?
!support
This server is for help with Discord APIs, SDK and other various integrations. For general support, please contact Discord at https://dis.gd/contact, or send us a message on Twitter at @discord_support
AE have you seen the crypto scams with the msg content as OMGG
not something i track
Oh My, Good Game
huh.. what am i supposed to do now
ignore it, its a scam
does anyone know how to block some bots from interacting with my msgs because some bots do not have a user settings interface to do it
My lib be just: ```py
from disthon import Client
Client().start()
what do you mean interfacing? you can just not use it
Hi, AE O/ How ya doin'
:awesome:
Hello Guys Help me?
It requires the dev to configure their bot token as an environment variable, typically in a dotenv :D
With what?
Bot
Specifically
The discord
hey everyone, is there a discord server for report any dc server?