Using #Discord as a forum is a weird experience. Even trying to view this feels overly .. crowded? And search seems overloaded with searching in channels, people - like a "omni-search" that's more likely to give something _I don't need_ versus actionable info.
@jalcine I feel the same way, I find Discord to be so clunky. I feel like I’m missing something because so many others seem real stoked about it
That said, I'll repost my question from there here for any #Rust eagle-eyed folks.
I'm getting a case of "not general enough" when attempting to use their macro to define a command with Tauri. I've pasted the output and code in question here: https://paste.ubuntu.com/p/4KyvJ2W58r/
@jalcine Discord actively fucks my head up whenever I try to use it. Absolute clusterfuck of an interface. Fortunately, their signing in process is almost equally cuntastic, so I'm rarely able to enter.
@jalcine discord forums are also completely unusable on narrow browser windows and on mobile web
The Discord UX challenges my perception of my own age, so as a neophile, I have a love-hate relationship with it.
@jalcine it's weird that discord is designed for the gaming community and inadvertently found itself also being one of the gotos for all kinds of other communities. I think it's bad at both.
@jalcine Oof, yeah. That's so dense and crowded, and gives you so little room for the actual post.
(Totally aside from the thing where they're shut inside an app where you need an account to even *read* them…)
@DavidDarnes mobile is where it shines - I much rather use Discord on that (and I think it's because I have more time spent on it, esp for non-tech spaces) @mayank
@cferdinandi I try to meet people where they are, tbh. I know that's more of a non-answer but it feels like it's hard to find a one-size-fits-all thing.
That said, I try to nudge towards either Signal or e-mail if I can.
@jalcine @heydon Sadly, since Musk bought twitter, "where people are" is like a dozen places for me now.
Some here, some on Bluesky, some on Slack, some on Discord. Twitter is still THE place to go for Black culture stuff. This place was _crickets_ when the whole Kendrick/Drake beef happened, for example.
I miss having many weird communities in one place.
@jalcine I don't know enough to answer this question, but I'd start by first changing the code so the tauri::generate_handler! result is stored in a let var before being passed into the function, then doing type inference (IDE or let () = varname), just to see what it is you've really got.
I did some google poking and found other examples where people fixed an error like this by putting the value in a Box<>. And another example where an error message like this turned out to be a compiler bug D:
So after doing a quick refresher on `git bisect`, I found the (potential) culprit. This was this change. https://git.sr.ht/~jacky/shock/commit/1eb4731f769c484d80116bf5ac4277a99f09d4bc
So examining the docs for `chunks` at https://docs.rs/futures/latest/futures/prelude/stream/trait.StreamExt.html#method.chunks and source gives me no indicator as to _why_ removing this would have my code compile.
This does produce another whack-a-mole moment but that's fixable.
@jalcine I mention the first thing because I found the generate_handler! doc in question didn't seem to explicitly explain its return value.
Also, there is a second "Rust Programming Language Community" Discord I would go to if Tauri discord is confused by the question, as this other discord very good at spotting weird issues that can be rephrased to be about general Rust primitive issues rather than specific libraries. Can pass that link on if it helps.
@mcc Ah, it's asking for an inference on a type and looking even into where it's passed in, it'd ask for a function definition (so it's like building a function to call another [set of?] functions)
This is more tangled than I expected
@jalcine If it turns out the problem is something very different and the compiler error was misleading, it might actually be worth filing a compiler bug on Rust's github. They're actually really intent on improving error message when it turns out a misleading element to a message is mechanically detectable (which isn't always ofc…)
@mcc I'll keep this in mind. I'm going to run a bisect to see if I can get to the point of no return.
That said, I _do_ remember upgrading Rust - if that's affecting this, that feels like compiler bug space, no?
@jalcine The code worked before and doesn't now? That's interesting if so, I don't know but I *assume* that violates the forward compat promises the Rust team makes. I think (?) backcompat breaks are only supposed to happen with new Editions.
You may already know this but I think Rustup will let you if you want to double check whether that's the case
@mcc well, I found the issue but have no idea as to _why_ it is. Going to add to main thread in a bit
@jalcine Try this URL: https://discord.com/channels/273534239310479360/1120124565591425034
If that doesn't work, try: https://discord.gg/rust-lang-community
I've got the most help in the rust-question-and-answers general channel. If you don't get an immediate answer the forum might be good for a more involved question like this. I've occasionally had luck w/ the subject channels but they're much slower moving. They react well to questions that can be phrased in Godbolt form & have some bot that lets you run a command and execute a ``` block in Godbolt