feat(0.3): rework stream error handling/remove wasi:io usage#143
feat(0.3): rework stream error handling/remove wasi:io usage#143lukewagner merged 1 commit intoWebAssembly:mainfrom
wasi:io usage#143Conversation
wasi:io usage
lukewagner
left a comment
There was a problem hiding this comment.
The change to the constructor/new is actually pretty interesting, because it demonstrates how the "return a future<result<_, error-code>>" trick is necessary not just for functions returning a stream, but also functions taking a stream, and thus even the stream<T,SuccessT,ErrorT> extension (presented here) wouldn't eliminate it. This raises some interesting questions on whether the ErrorT should be able to flow both ways... but that's all for later, and I think what you're doing here is about the only possible thing we can do in a 0.3.0 timeframe to not regress from 0.2, which provided error-codes upon write failures. Nice work! (cc @dicej )
Signed-off-by: Roman Volosatovs <rvolosatovs@riseup.net>
60c4d31 to
9286513
Compare
|
@lukewagner I've rebased on latest |
lukewagner
left a comment
There was a problem hiding this comment.
Given all the things in flight, I wasn't sure when's a good point to merge (and iterate in later PRs), but seems fine to merge now?
I think it's fine to merge now indeed |
Remove
wasi:iousage inwit-0.3.0-draftFollowed the same stream error handling convention as currently in
wasi-0.3.0draft wasi-filesystem#164wasi-0.3.0draft wasi-sockets#111