Originally posted by ehansin
View Post
These HTTP/3 features have been researched, discussed and tested for a very long time and yet there are still some unknowns around the implementations and practical benifit. It could either be that some of the projects are waiting for this new OpenSSL 3.2 release (it has been a long time coming) or that the project simply doesn't see HTTP/3 or QUIC as a priority for them. I would also suggest if you don't need any HTTP/3 features then stick with HTTP/2 for a few years.
For example: I have been using QUIC standalone (without HTTP/3) for redundant voice communication over multiple WAN connections. I have not seen the need to update my web server to support QUIC because the performance improvements doesn't outweigh the management costs in my niche use case.
If you want to get your hands dirty I would suggest to take a look how Quiche (cloudflare), Nginx, Caddy, A̶p̶a̶c̶h̶e̶, Jetty and others have decided to make use of HTTP/3. You could also read some of the docs or specs: https://github.com/bagder/http3-expl...ter/SUMMARY.md and https://datatracker.ietf.org/doc/html/rfc9000
PS: It will take many years for me to finish learning and understanding the entire spec in my own time.
Comment