Changes to 8u and 11u process

Andrew John Hughes gnu.andrew at
Thu Aug 8 21:44:00 UTC 2019

On 08/08/2019 11:50, Andrew Haley wrote:
> On 8/7/19 4:36 AM, Andrew John Hughes wrote:
>> I would point out that the decisions made by Oracle for their
>> proprietary fork are opaque to us, and so to do as proposed would be
>> to essentially hand control of the majority of our work to a black
>> box. I am vehemently of the opinion that the OpenJDK projects should
>> remain independent of such proprietary forks.
> Everybody's vehement today. :-)
> It's not in the best interests of Free Software that we diverge
> significantly from Oracle because we must maintain an easy path from
> closed to open. Sure, we can add features like, say, Shenandoah, but
> we shouldn't miss any. So, decisions about whether to accept a
> backport that Oracle has already accepted are going to be pretty much
> automatic.
> While it'd be nice if we were the masters of all we survey, it's not
> going to happen: better that we accept or place in all of this, which
> is to do our part to help the masses reach the sunlit uplands of freedom.

Indeed. I'm familiar with these realities from our days on GNU Classpath
and GCJ.

However, it is worth acknowledging another reality; we only have a
window into their JDK, not the whole picture. While the shared bug
database means we can see some of the changes being made to the
proprietary forks, there are enough private bugs to mean that we can
never claim to not miss any. For example, in the last CPU, Oracle added
the option to toggle the ECC implementation used, which I haven't seen
referenced anywhere in the bug database.

Although it may be used sparingly, if at all, I think we should retain
the right to approve all bugs, especially as these legacy versions
continue to diverge further from the main development tree.
Andrew :)

Senior Free Java Software Engineer
Red Hat, Inc. (

PGP Key: ed25519/0xCFDA0F9B35964222 (hkp://
Fingerprint = 5132 579D D154 0ED2 3E04  C5A0 CFDA 0F9B 3596 4222

More information about the jdk8u-dev mailing list