All patches are pushed now.

Aside from the below problems, I have found the following problems:

ttp: bad javadocs
sxp: bad javadocs

The following two projects require attention from their committers:

snmp4sdn: needs cleanup, as it seems to be pulling very old yangtools versions and master does not seem to build at all
tsdr: H2 plugin tests are failing due to HashMap ordering change -- not sure if tests should be updated or implementation needs to be fixed

I would say that overall status is good for enabling JDK8 for all projects and I think we can get to Beryllium release being built using JDK8 -- we'll know more as the fixes trickle in and autorelease has a chance of finishing.

Bye,
Robert

I am already pushing per-project patches for projects participating in Be autorelease. The list can be found here: https://git.opendaylight.org/gerrit/#/q/topic:jdk8, which includes any fixes needed as I encounter problems in local build.

So far the only problems encountered are:

aaa: two wrong @links
dlux: overriding jacoco to an ancient version
gbp: fails in unit tests (have not looked further)
lacp: fails in unit tests (but also fails its regular verify)

I expect to have the patches pushed out today. I would very much prefer those jobs be voting, as otherwise breakage will creep back in (as it did in AAA some six weeks ago).

Bye,
Robert

On 2015-10-31 15:00, Keith Burns wrote:

Maybe we could start with Jenkins verify jobs. Is it possible to kick off a java7 (voting) and java8 (non voting) verify for each opted in project or is that adding more burden to our Jenkins system?

On Oct 31, 2015 4:16 AM, "Robert Varga" <nite@...> wrote:

On 2015-10-30 11:09, Robert Varga wrote:

So this is a three-step process:

1) Release with JDK7, support build with JDK8
2) Release with JDK8, support build with JDK7
3) Require JDK8

We want to get to 3) due to language features.

We were at 1) at the end of Lithium, if memory serves me, but have regressed. The original intent was for Beryllium release to be at step 2), which I think is still attainable, but we have to get to 1) first and make sure we don't ever regress.

I have re-opened BUG-2625, moved it to releng/autorelease and targeted it for Beryllium -- thus we have an overall tracker issue.

Bye,
Robert

_______________________________________________
release mailing list
release@...
https://lists.opendaylight.org/mailman/listinfo/release