Fixes of copyright headers: Should we downport them?

[Gil Tene]

Sent from my iPad

On Mar 25, 2020, at 11:26 PM, Aleksey Shipilev <shade at redhat.com> wrote:

On 3/19/20 4:39 PM, Lindenmaier, Goetz wrote:
Hi Andrew,

There are a row of changes fixing copyrights downported by Oracle to 11.0.8,
similar to this one:
https://bugs.openjdk.java.net/browse/JDK-8220414

See also
https://bugs.openjdk.java.net/issues/?filter=38639

I'd consider them pretty pointless given the rules you set.
So should we skip them?
If so, I would mark them all as openjdk-na.
(In my opinion we should downport them, but I'll follow
your advice.)

Not Andrew, but I believe copyright headers are important enough to consider for backports. If the
reverse was true, we would not care about rejected hunks in copyright lines during the normal backports.

Umm...

It literally says ”DO NOT ALTER OR REMOVE COPYRIGHT NOTICES OR THIS FILE HEADER.”

So unless we are copying the entire file from upstream
intact (replacing the entirety of the old copyrighted file
in 11u with a new copyrighted file copied from upstream),
I don’t think anyone other than Oracle should be messing
with the existing Oracle copyright notice text in 11u code.

And if we do replace the file in its entirety (and e.g. the only
difference between the two is the copyright text) we’d likely
want the change records to indicate a removal of the old file
and an addition of a new replacement file, and not just a
change of the one or two lines of copyright text that differed
between the two.

If you can get Oracle to do the backport of the copyright
changes to 11u, that’s fine. But otherwise, I’d leave them
as they are.

I know this may seem pedantic, but copyrights and notices
matter, and they carry a lot of implications. We all depend
on the code in the updates projects not messing with them.


--
Thanks,
-Aleksey