-
Notifications
You must be signed in to change notification settings - Fork 631
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Add AsyncRead::poll_read_buf and AsyncWrite::poll_write_buf #1826
Conversation
df29530
to
9432c94
Compare
I have no strong opinion about whether to block this with bytes 0.5-alpha. |
I don't think this is a right addition for Direction for
|
@yoshuawuyts thank you for spending the time to write this up. I think there are some really good ideas here!
I think this is very reasonable and makes sense. 👍
Yes, this is something we really do want to achieve because I think this is and will become a large pain point throughout the ecosystem. The tough part is that we have certain requirements that we want to maintain and this has become hard once the io traits got moved out of tokio. Which is why we still have our own versions since we have not had the full opportunity to explore them yet in relation to new futures.
I think this approach would work fine for our use case and maybe the correct approach here. Though I'd like to hear what @taiki-e has to say.
Now circling back to this section. What are the plans around this? I don't think the ecosystem has had enough time to really explore all the traits besides Again, thanks for the write-up and I do hope soon we can all live on the same traits 😄 |
So another thing related to the extension trait that I totally forgot was that we can no longer specialize implementations with that method. Meaning we can't override the poll fn. This is something we would like to do but I'm not sure how we can support that without this. 😞 |
@LucioFranco can you give more information about the kinds of overriding implementations you'd like to provide that the current system doesn't support? Maybe we can find another way to make those work nicely. |
@cramertj one use case is that the caller can avoid going through the Edit: and besides just a binary "supported" or not, it also allows the |
We discussed this before, and I still don't think i ever quite understood why this is something you'd want-- |
I think that's the basic crux of my argument here-- I'm not opposed to adding this as an unstable optional feature, but I'm opposed to ever having functions on |
I think this is a perfectly valid thing and I actually agree! But this just leads me to think that these traits may not be ready for So from that, I think it may make sense to close this or throw it behind unstable. |
I'm not sure how that conclusion relates-- these traits seem fine to me, and correspond pretty precisely to |
Sure, then I say we close this and we can explore, for now, in our own traits. |
Closes #1818