search

twitter / util

Wonderful reusable code from Twitter
https://twitter.github.io/util
Apache License 2.0
2.68k stars 580 forks source link

Please reconsider #235 #266

Open sjoerd-vogel opened 4 years ago

sjoerd-vogel commented 4 years ago

Please reconsider #235 as it is required for jdk11 support.

Expected behavior

https://twitter.github.io/finagle/guide/changelog.html mentions initial jdk11 support for 19.11.0 onwards. Hence I expect jdk11 support.

Actual behavior

As far as I can tell issue #235 has not yet been fixed. It results in warnings in jdk11.

Steps to reproduce the behavior

See issue #235 for this.

cacoco commented 4 years ago

@sjoerd-vogel Hi there, thanks for the issue! Reading through the previous ticket, I don't think the intention was that we are done here -- it's just that we don't yet officially support anything other than JDK8 and when we move to full support of JDK11 we will tackle this (it should be within the next few quarters). Since it only results in warnings and not an error we're waiting until the move to address. Hope that helps. Thanks!

ComFreek commented 3 years ago

Are there any news on this issue? JDK 8 has been long superseeded by newer versions.

ComFreek commented 3 years ago

@cacoco Is there any update on this? :smile:

cacoco commented 3 years ago

@ComFreek hey there! We run a very large codebase and are still in the process of certifying the move to JDK11. I don't have a current ETA here but the work is on-going. Hope that helps. Thanks!

sjoerd-vogel commented 3 years ago

I gave this a shot. https://github.com/twitter/util/pull/284

felixbr commented 2 years ago

So, JDK8 is EoL end of March 2022. Also there are several known bugs in JDK8 which would be fixed in JDK11+.

I suspect Twitter still doesn't support anything beyond JDK8 officially for their codebase but maybe we can at least try and support JDK11 for twitter-util, twitter-server, and finagle (= open-source libraries)?

When looking at the CI-config here it looks like the build is already tested for JDK11. In my mind this should be enough to accept PRs which try to improve JDK11 support as long as they don't break JDK8 support.

Am I missing something?