Respinning RC3 seems a better approach if you are willing to take the effort.
I could see 475853 being done in RC4 if necessary;
But 476147 would have global impact I guess which I personally would not be comfortable with in RC4.
Martin Oberhuber, SMTS / Product Owner – Development Tools,
direct +43.662.457915.85 fax +43.662.457915.6
From: eclipse-pmc-bounces@xxxxxxxxxxx [mailto:eclipse-pmc-bounces@xxxxxxxxxxx]
On Behalf Of David M Williams
Sent: Thursday, September 03, 2015 10:50 PM
To: eclipse-pmc@xxxxxxxxxxx; platform-releng-dev-bounces@xxxxxxxxxxx
Subject: [eclipse-pmc] Request to respin RC3 ... or early request for RC4 change
There are two bugs that I think must be addressed for Mars.1. Both small in their effect, but large in being a professional release.
Bug 476147- Update JDT compiler in maintenance stream, so final product is build with the released compiler
I'm assuming JDT is "really done" :) and we should be sure our final release is built with the compiler that will be released (which, was just changed, for RC3).
I have already checked in this change, assuming it's relatively routine to do ... only question is to do it now? Or next week?
Bug 475853- Reminder to increase Mars service fields, if changed
Two features still needed "service increments". One is fixed in Gerrit, and the other is attached as patches to the bug (if someone from PDE is around to commit it).
I'm fine with either RC3, or to wait for RC4, but advocate the former, just so that our RC3 really is as "final" as we can make it, and, partially so there is less of a psychological effect of
putting in other fixes merely because "we are rebuilding RC4 anyway", and partially since Monday is a holiday for some ... so, would be good to be "final" as early as possible.
I'll look to PMC guidance if we should respin RC3, or declare it done, and do a "quick build" for RC4?