cmsis_update::update()

[Yatekii]

Hi!

Is there a reason that cmsis_update::install() for example requires some type with a trait implemented to be passed into it just to configure it and get some progress update?
Why can't this be fixed? Is it because of the python interface? Even then I think this can be implemented with a less complicated interface which I'd highly prefer.

Also, is there a reason you force users to use slog or did you just not care enough? Because I don't want to use slog but I now have to include the dependency anyways. And exactly for this scenario there is https://docs.rs/log/0.4.8/log/ which provides a nice and unified frontend for maaaany different backends (slogbeing one of them).

I know this crate was maybe not written with intent to be reused as a lib, so don't take my criticism too harsh, I just wanted to ask and would make those improvements in a PR if that's ok =)

Best,
Yatekii

[theotherjimmy]

some type with a trait implemented to be passed into it just to configure it and get some progress update?

There is probably a better way to do this. I'm open to suggestions and PRs for simpler interfaces. Originally it was to allow the python interface to handle the messages differently than the CLI.

Is there a reason you force users to use slog or did you just not care enough?

I don't think slog is needed at this point. I must have had a reason for adding it, but that reason is probably no longer applicable. I think it would be some work to pull out the usage of slog.

I just wanted to ask and would make those improvements in a PR if that's ok =)

I would love to see these improvements. Feel free to submit PRs.