Commit Graph
2 Commits
Author SHA1 Message Date
Ran Benita e3b4c99a33 codec: a few suggestions (#1418)
How the buffer is managed is often critical for performance. Not
taking care of it will be catastrophic for performance beyond the
initial buffer size with the current implementation (a loop of
`reserve(1)`).
2019-08-09 12:20:31 -07:00
jesskfullwood 6b9e7bdace codec: update to use std-future (#1214)
Strategy was to

- copy the old codec code that was temporarily being stashed in `tokio-io`
- modify all the type signatures to use Pin, as literal a translation as possible
- fix up the tests likewise

This is intended just to get things compiling and passing tests. Beyond that there is surely
lots of refactoring that can be done to make things more idiomatic. The docs are unchanged.

Closes #1189
2019-06-27 10:10:29 -07:00