matklad / cargo-xtask

837 stars 23 forks source link

Potential crates to recommend #6

Open epage opened 4 years ago

epage commented 4 years ago

(sorry, needing to rush again or else I'd create a PR)

Immediately useful:

Could be refactored to be useful

epage commented 4 years ago

I'm unsure whether cargo release and clog/jilu (changelog generation) would be useful to refactor into xtasks.

matklad commented 4 years ago

I think in a world where Cargo has per-machine (as opposed to per-project) binary cache, I personally would certainly prefer cargo release to be an xtask, rather than a custom subcommand. Today, the situation is less clear...

epage commented 4 years ago

I personally would certainly prefer cargo release to be an xtask, rather than a custom subcommand

I'm seeing the potential for that now and am really liking it. While there are benefits to pushing people to a standard workflow, there are some things that can't be standardized (like changelog practices) and giving people a toolbox for how to release that they can build up as needed is a really interesting idea.

epage commented 4 years ago

clap-cargo would also be a useful crate for xtasks with custom extensions, like imitating cargo's workspace flags.

matklad commented 4 years ago

I am not actually sure: clap provides high-quality argument parsing, at the cost of high compile times. I think for xtasks, super-fancy arg parsing is not required, so something like pico-args might be a better choice