Buy commercial curl support from
WolfSSL. We help you work out your issues, debug your libcurl
applications, use the API, port to new platforms, add new features and more.
With a team lead by the curl founder himself.
Hyper, is it worth it?
- Contemporary messages sorted: [ by date ] [ by thread ] [ by subject ] [ by author ] [ by messages with attachments ]
From: Daniel Stenberg via curl-library <curl-library_at_lists.haxx.se>
Date: Tue, 16 Apr 2024 09:55:36 +0200 (CEST)
Hello.
In the end of 2020 we merged the initial experimental support for hyper as an
alternative backend for libcurl HTTP functionality. It is a way to increase
the ratio of memory-safe code into a libcurl build.
hyper support remains experimental to this day because we still have fifteen
test disabled for it. Meaning that we still lack the proper adjustments in
curl to make it on par with the native HTTP backend.
We even recently had to back-pedal and remove support for HTTP/2 when hyper is
used because of how we (I) had misunderstood their API and how to properly
integrate HTTP/2 support into curl.
For the last six months, the only changes done in curl's hyper code have been
fixes to make it keep up with refactors and other internal cleanups and
improvements. No one seems to (want to) work on improving curl's hyper backend
and no one seems to use it or care about its lack of HTTP/2 support.
Some 40 months after initial merge, we seem to be at a stand-still and yet
things like [1] happens that require attention and work. Is hyper support now
just a dead weight that adds friction when we want to innovate and improve?
I added "drop Hyper support?" as a conversion topic for curl up 2024.
Thoughts?
[1] = https://github.com/curl/curl/issues/13380
Date: Tue, 16 Apr 2024 09:55:36 +0200 (CEST)
Hello.
In the end of 2020 we merged the initial experimental support for hyper as an
alternative backend for libcurl HTTP functionality. It is a way to increase
the ratio of memory-safe code into a libcurl build.
hyper support remains experimental to this day because we still have fifteen
test disabled for it. Meaning that we still lack the proper adjustments in
curl to make it on par with the native HTTP backend.
We even recently had to back-pedal and remove support for HTTP/2 when hyper is
used because of how we (I) had misunderstood their API and how to properly
integrate HTTP/2 support into curl.
For the last six months, the only changes done in curl's hyper code have been
fixes to make it keep up with refactors and other internal cleanups and
improvements. No one seems to (want to) work on improving curl's hyper backend
and no one seems to use it or care about its lack of HTTP/2 support.
Some 40 months after initial merge, we seem to be at a stand-still and yet
things like [1] happens that require attention and work. Is hyper support now
just a dead weight that adds friction when we want to innovate and improve?
I added "drop Hyper support?" as a conversion topic for curl up 2024.
Thoughts?
[1] = https://github.com/curl/curl/issues/13380
-- / daniel.haxx.se | Commercial curl support up to 24x7 is available! | Private help, bug fixes, support, ports, new features | https://curl.se/support.html -- Unsubscribe: https://lists.haxx.se/mailman/listinfo/curl-library Etiquette: https://curl.se/mail/etiquette.htmlReceived on 2024-04-16