Preconditions with missing handlers! What do?

Preconditions: What should happen with preconditions (global or not) that need to run on a message command but do not have a handler for them? Skip? Throw error?

UndiedGFX•953d ago

Maybe skip them till you add preconditions for slashies?

vladdy•953d ago

Not the question at hand

24•953d ago

can you define an example?

vladdy•953d ago

Preconditions will happen across the board, the question is what should happen if a command that can run both via / and message receives a precondition that only handles /

24•953d ago

skip means you could define everything in 1 array buuut

vladdy•953d ago

but should it be skipped Or should it throw an error

UndiedGFX•953d ago

Maybe throw an ephermal message? As error

vladdy•953d ago

Not doing user responses for them

UndiedGFX•953d ago

vladdy•953d ago

We don't even have an in-chat error message by default

24•953d ago

IMO skipping them means you have to deal with mixed preconditions (slashes and messages) which can result in either nasty nested arrays, or w/e

UndiedGFX•953d ago

I mean we do handle it On commandDenied

vladdy•953d ago

There is no listener for commandDenied included by default

UndiedGFX•953d ago

So just like how normal preconditions work would be OK ish Erhm

vladdy•953d ago

I'm also not talking about the preconditions array But running the preconditions specified

UndiedGFX•953d ago

Maybe throwing an error could be helpful

24•953d ago

can you define an example? I still don't quite follow exactly

vladdy•953d ago

Pretty sure the question speaks for itself.. Take this precondition

import type { Message } from 'discord.js';
import { Identifiers } from '../lib/errors/Identifiers';
import { Precondition, PreconditionResult } from '../lib/structures/Precondition';

export class CorePrecondition extends Precondition {
    public chatInputRun(message: Message): PreconditionResult {
        return message.guild === null
            ? this.ok()
            : this.error({ identifier: Identifiers.PreconditionDMOnly, message: 'You cannot run this command outside DMs.' });
    }
}

import type { Message } from 'discord.js';
import { Identifiers } from '../lib/errors/Identifiers';
import { Precondition, PreconditionResult } from '../lib/structures/Precondition';

export class CorePrecondition extends Precondition {
    public chatInputRun(message: Message): PreconditionResult {
        return message.guild === null
            ? this.ok()
            : this.error({ identifier: Identifiers.PreconditionDMOnly, message: 'You cannot run this command outside DMs.' });
    }
}

UndiedGFX•953d ago

What vlad says it that it isn't handler for interactions What should he do at that point

vladdy•953d ago

Now let's assume your command that says this precondition is needed handles message based commands AND slash commands

UndiedGFX•953d ago

But it isn't handled for slash commands

vladdy•953d ago

Or messages or contextMenus

UndiedGFX•953d ago

Because maybe you forgot to make a handler I mean aren't they handled by sapphire for messages?

vladdy•953d ago

message handlers will be optional

UndiedGFX•953d ago

vladdy•953d ago

and well chatInputRuns are considered the new abstract run you must always provide

24•953d ago

that makes more sense, I couldn't quite think the way through the question. I think in these cases, if we error them, we would then have more runtime errors (unless they can be typed) If we skip them, we could run into some conditions that would appear to be running, but don't, thus causing frustration for an end user I would accept having them error due to having a much better understanding of why a precondition didn't run (If it errored, I didn't provide a valid handler / I used the wrong precondition- that should error, makes it much easier to debug and better (D)UX)

vladdy•953d ago

I'm leaning towards hard erroring because it's safer for the user

24•953d ago

I'd agree

UndiedGFX•953d ago

I'd agree too

24•953d ago

I can see skipping becoming a hassle for people in #old-sapphire-support i i.e. "Why did my precondition not run for interactions but do for messages"

vladdy•953d ago

or permission escalations

24•953d ago

yup I don't see any valid use cases for skipping it just seems like it would cause more headaches for everyone (devs, you)

vladdy•953d ago

But then SHOULD the chatInput ones be mandatory or not.. I want to make chatInputCommands the pushed way bc well, they're nice

24•953d ago

good question in doing so you would force people to think about using slash commands over message commands win-win right?

vladdy•953d ago

which is what Discord (and ngl I) want

24•953d ago

I can see that being annoying for some users but really, really important down the road

vladdy•953d ago

I mean the bitching will only bitch for TS users

24•953d ago

:) yeah I can see this being slightly annoying at first, but nothing a find and replace command can solve

𝖄𝖚𝖌𝖊𝖓•953d ago

If no handler is defined. Dont run it. I think a same approach like the commands would be beneficial

vladdy•953d ago

Yeah BUT it can mean permissions escalation Or accidentally forgetting to handle it

𝖄𝖚𝖌𝖊𝖓•953d ago

I mean. Skipping error'ing is just as bad And since commands already have a seperate function to handle slashcommand functionality. It wouldnt be weird to have the same logic apply to other pieces imo

vladdy•953d ago

..well, yes

Sapphire - Imagine a framework Join

Welcome to the Sapphire Discord server! The next-gen object-oriented Discord.js bot framework can be found here.

2.2KMembers

View on Discord

Want results from more Discord servers?

Add your server

Context Menus vs Application Commands - separate or not?**Question time** ❓ Context menus.. they exist. They're like used once a blue century. How should Interaction Handlers after usageWant to give interaction handlers a shot **now**? Well, you can! Just run the command below and giv Interaction Handlers!!Hello peeps! I know, I know, allllll of you (myself included) are just basically begging me for app Initial Commit ThoughtsI've pushed the initial commit for interaction handlers: heavily wip, totally untested and raw (just The statefulness or statelessness of Interactions1. Most likely, but it would be a nice addition if I was able to have non-state "callbacks" as menti Interaction handlers in the command instead of globallyHonestly. I would much rather add interactions from my command and from there tell to continue in (f Reply sess 2**1.** Yes. This also follows the design pattern that we more-or-less employ already for Sapphire. W dont feel like creating a review on GitHub...PR is up: https://github.com/sapphiredev/utilities/pull/152 CC @kyra 🩵🩷🤍🩷🩵 @vladdy

Gaming

Programming

Preconditions with missing handlers! What do?