On Sat, 21 Oct 2023, Stephen Farrell via curl-library wrote:

I'm not clear how you like to handle experimental things like our ECH PR. [1] Is that something that'd be suited for proposing for this release cycle or not?

I would like to land this as soon as we think it is usable, but labeled EXPERIMENTAL and only built if opted in at build-time.

This way, we can fine-tune it while people get a chance to try it out and work on it in the master branch. It should then only possibly hurt users who opt in to participate in the experiment.

We can do this during this feature window if we think it is good enough. I feel I've neglected your work there recently, but I want to get back and give it a good review again and help merging it.

Since the function is to be merged as experimental at first anyway, it is not terribly important to stick to the merge window, even if I think it is a generally good idea to do that.

I'd probably recommend waiting 'till the next cycle maybe, on the assumption that the wolfSSL+HRR bug may be fixed then, and the ECH spec may have gotten to working group last call in the IETF. But I'm happy to try for an earlier merge if that makes more sense here.

I'm happy to follow your recommendation. Then we can also make sure that we have it reviewed and in perfect shape by then, merge and then convince people to try it out for real!

--

 / 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.html

Reply via email to