make sure Error types from protocols crates implement std::error::Error

[plebhash]

an idea to improve error handling on the roles implementation is to start using the anyhow crate #736 (comment)

as a prerequisite, all custom Error types from the low-level crates under protocols must implement the std::error::Error trait

[plebhash]

once #932 is merged, the proposed changes on this issue can be achieved via:

#[cfg(not(feature = "no_std"))]
impl std::error::error for Error {}

[rockcoolsaint]

This then means that you're handling this issue right?

[plebhash]

This then means that you're handling this issue right?

hey! this is actually on my backlog, if you wanna give it a go, feel free!

my suggestion would be:

• branch off from feat-optional-no-std on my fork, to make sure that we can merge your PR after #932 is merged
• do what I'm describing on the description of this issue, as well as the comment below it
• send a draft PR

once you submit your PR, I'll make sure to add to the description of #932 that we should merge yours afterwards

[rockcoolsaint]

Alright

[rockcoolsaint]

@plebhash should the 'anyhow' crate be applied to every instance of unwrap() or strictly functions that return Result<> type?
Moreover, there are way too many instances where unwrap() is used in roles crate alone. Will refactoring everyone of those instances be appropriate?

[plebhash]

this issue is only related to making sure Error types from protocols crates implement std::error::Error

while the potential introduction of anyhow is the motivation, it is not the focus here... that is future discussion.

[rockcoolsaint]

this issue is only related to making sure Error types from protocols crates implement std::error::Error

while the potential introduction of anyhow is the motivation, it is not the focus here... that is future discussion.

I'm kind of having a hard time identifying the custom Errors, as there seems to be Error scattered everywhere in the protocol crate. Please can you show me a reference custom Error. Seems there are going to be many instances to deal with

[plebhash]

@rockcoolsaint thanks for your effort on this one

we decided to go on a different direction (#932 (review)), apologies if you spent a lot of effort on this one already

[rockcoolsaint]

Alright then, I'll pickup another issue