Backporting features [was Re: RFR (11u, XXL): Upstream/backport Shenandoah to JDK11u]

[Andrew Haley]

On 2/17/20 7:13 PM, Claes Redestad wrote:

> I think our best bet at both retaining existing users and growing
> the ecosystem is by putting our collective resources on delivering
> new features of high quality into the mainline. Backports are a
> necessary evil, though, and JFR is an obvious "gap". I'm just not
> convinced it's obviously the right thing to do, especially
> considering what we could have done instead with that time and
> effort.

I'm not either. I wrestle with thoughts like this every day.

Some random thoughts: a significant difference between OpenJDK and
closed projects is that we get to have these kinds of conversations
out in the open. Not only that, but our deliberations are preserved
forever for everyone to see. All we can say in the end is that while
the decisions we make may not be the ideal ones from everyone's point
of view, or indeed match those of any single individual, I can
guarantee we've thought about them very hard. This conversation is
part of that process.

Also, having several significant backports coming at the same time is
unusual and entirely accidental. I believe that the volume of
backports will slacken considerably from this point onwards.

-- 
Andrew Haley  (he/him)
Java Platform Lead Engineer
Red Hat UK Ltd. <https://www.redhat.com>
https://keybase.io/andrewhaley
EAC8 43EB D3EF DB98 CC77 2FAD A5CD 6035 332F A671