hyper-util: What is the roadmap for the legacy client? #4175
|
Hi I am volunteering for the Tor Project which is making a new implementation in Rust called Arti. Previously we had a crate providing an extra layer on But we noticed that the lack of a custom connector is now missed by our users. In MR!2941 we are planning on resurrecting the But AFAIK this is only a logic choice when the user is able to use the legacy client for I am wondering what the planned roadmap for the legacy client in What exactly is meant by this? Are you planning on supporting this legacy client in the far future too? |
Replies: 2 comments 1 reply
|
I guess a part of your answer is to be found in #3948 . |
|
As long as the replacement has not been completed, we'll fix bugs in the legacy one. But it is not accepting any new features. Those should be layers, such as talked about in https://seanmonstar.com/blog/hyper-util-composable-pools/. Once enough layers exist in hyper-util to re-implement the legacy client, |
As long as the replacement has not been completed, we'll fix bugs in the legacy one. But it is not accepting any new features. Those should be layers, such as talked about in https://seanmonstar.com/blog/hyper-util-composable-pools/.
Once enough layers exist in hyper-util to re-implement the legacy client,
legacy::Clientwill be removed.