feat: print sources on alternate display

[Frando]

Description

This prints sources when using the alternate display ({:#}).

Example:

        let ep = Endpoint::builder()
            .bind_addr_v4("123.0.0.3:3".parse().unwrap())
            .bind()
            .await;
        if let Err(err) = ep {
            let err = n0_snafu::Error::from(err);
            println!("err display:");
            println!("{err}");
            println!("err alternate display:");
            println!("{err:#}");
        }

prints

err display:
Failed to create bind sockets
err alternate display:
Failed to create bind sockets: Cannot assign requested address (os error 99)

Before, the alternate display would be the same as the regular display (no sources).

Breaking Changes

Notes & open questions

Change checklist

• Self-review.
• Documentation updates following the style guide, if relevant.
• Tests if relevant.
• All breaking changes documented.

[github-actions]

Documentation for this PR has been generated and is available at: https://n0-computer.github.io/n0-snafu/pr/4/docs/n0_snafu/

Last updated: 2025-08-29T07:36:20Z

[matheus23]

Before we merge, we had a discussion about this on discord (and I agree):

I am not a fan of this
because this assumes a single source
so I think it should always be a list
my failure: because a: because b: because c
will now be the result for multiple sources
which is really not great
I think it should be

my failure:
 0: because a
 1: because b
 2: because c

or sth like that

(source)

[matheus23]

Happy with this in the sense that we really need this 😅 (I keep hitting this)

But also: Given we use the same format as color-eyre for spantraces and backtraces, perhaps it'd make sense to use the same format for errors, too? https://lib.rs/crates/color-eyre

[dignifiedquire]

it's an improvement and we have a test to track the behaviour, we can always iterate on this

[dignifiedquire]

are you sure this prints the same stack now? especially it doesn't skip the root anymore, so you will end up printing that twice now I think

[Frando]

I was not getting the root at all before, in the new test I added. Maybe needs more testing, also with all the different enum types of Error (tested only the whatever variant so far, I think).

[Frando]

btw: what is the difference between the Source and Message variants, couldn't they be merged?