You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
cap-async-std has been removed without a replacement (see #399), because async-std is no longer maintained.
Now there is no way of using async code with this crate.
It would make sense to add either a:
cap-tokio (I would prefer this one)
or a
cap-smol
crate to this repository to retain the async functionality of this project.
Personally, I think it's simpler and actually more efficient in some ways to just encapsulate high level cap-std workflows (i.e. more than one operation) into a tokio spawn_blocking thread.
To be honest, I'm preparing to either borrow code from #414 for my project or vendor a copy of tokio::fs::read_dir where I swap out the wrapped std::fs::read_dir operations for cap_std::fs::Dir::read_dir ones.
If for no other reason, because I tend to work with folders large enough that being able to cleanly abort retrieving a list of contents is valuable and Tokio already figured out the DOs and DON'Ts of incorporating read_dir into it.
I hope this won't be another case of "The image crate closed my request, so now I have to vendor a copy of their code because what I want is literally their glue code that can't be meaningfully written any other way, but with a means to use the APIs provided by the underlying png crate to inject some XDG Thumbnailing Spec metadata in the process".
Activity
sunfishcode commentedon Feb 15, 2026
Either of those would indeed be nice to do. I myself don't have the bandwidth to do them, however.
Ericson2314 commentedon May 11, 2026
I just started trying to slop
cap-tokiointo being.cgwalters commentedon Jun 8, 2026
Personally, I think it's simpler and actually more efficient in some ways to just encapsulate high level cap-std workflows (i.e. more than one operation) into a tokio
spawn_blockingthread.ssokolow commentedon Jul 15, 2026
To be honest, I'm preparing to either borrow code from #414 for my project or vendor a copy of
tokio::fs::read_dirwhere I swap out the wrappedstd::fs::read_diroperations forcap_std::fs::Dir::read_dirones.If for no other reason, because I tend to work with folders large enough that being able to cleanly abort retrieving a list of contents is valuable and Tokio already figured out the DOs and DON'Ts of incorporating read_dir into it.
I hope this won't be another case of "The
imagecrate closed my request, so now I have to vendor a copy of their code because what I want is literally their glue code that can't be meaningfully written any other way, but with a means to use the APIs provided by the underlyingpngcrate to inject some XDG Thumbnailing Spec metadata in the process".