Re: Changing the upcoming LTS baseline?

76 views
Skip to first unread message

oliver gondža

unread,
Dec 30, 2015, 1:26:22 PM12/30/15
to Jenkins Developers, Daniel Beck
That sounds fine with me. I will prepare branch for both 1.636 and 1.642
so we can publish rc as soon as we decide what the baseline will be.

--
oliver

Daniel Beck

unread,
Dec 30, 2015, 5:46:18 PM12/30/15
to Jenkins Developers
We decided to base the upcoming LTS line on 1.636 in the last meeting on Dec 9. 1.637 introduced two regressions (JENKINS-31458 and JENKINS-31518), that were only resolved in 1.639 and 1.640, so we didn't consider those to be great choices. 1.640 in turn introduced another bug, JENKINS-31954, which was only fixed in the then-unreleased 1.642.

It has been three weeks since the meeting, and we've had very positive feedback for the last two releases, Jenkins 1.642 and 1.643.[1] There have been no notable further regressions reported in JIRA either.

Since we've skipped the Dec 23 meeting, the LTS schedule has been moved two weeks: The next LTS release is planned for Jan 20, with backporting finished Jan 6 (so far, it hasn't really started[2]).

Given all of the above, it may be feasible to advance the version chosen for the LTS baseline to 1.642 or 1.643, rather than stick to the previous choice made necessary by the regressions in later releases, while keeping the expected schedule.

WDYT?

1: http://jenkins-ci.org/changelog (click 'Community ratings')
2: https://github.com/jenkinsci/jenkins/commits/stable-1.636

Daniel Beck

unread,
Dec 30, 2015, 5:46:18 PM12/30/15
to Jenkins Developers, oliver gondža

Daniel Beck

unread,
Dec 30, 2015, 5:46:18 PM12/30/15
to Jenkins Developers, oliver gondža
For some reason my original email didn't make it to the list (twice), so reposting as an answer to Oliver's response, hoping this will work.

---
---
> --
> You received this message because you are subscribed to the Google Groups "Jenkins Developers" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to jenkinsci-de...@googlegroups.com.
> To view this discussion on the web visit https://groups.google.com/d/msgid/jenkinsci-dev/op.yagytzwxsbfict%40localhost.localdomain.
> For more options, visit https://groups.google.com/d/optout.
>

Kanstantsin Shautsou

unread,
Dec 30, 2015, 6:03:41 PM12/30/15
to Jenkins Developers, ogo...@gmail.com, m...@beckweb.net
Imho, better wait for some time for 1.642 user testing, then schedule meeting and pick new base (or backport fixes?).
Almost all latest LTS's had regressions, rush is not acceptable for "LTS" imho (too much fun with weekly dev builds).

Daniel Beck

unread,
Dec 30, 2015, 6:13:53 PM12/30/15
to Jenkins Developers

> On 31.12.2015, at 00:03, Kanstantsin Shautsou <kanstan...@gmail.com> wrote:
>
> wait for some time for 1.642 user testing

I understand the concern, but I doubt we'll get significantly more testing of a specific version that we have with these two releases already. And we got quite strong positive feedback on the changelog already. There were only 2 cloudy and 2 storm votes for 1.642, but no issue IDs provided. No negative feedback for 1.643 at all.

Or do you think the feedback from e.g. companies is missing due to the December time frame? Aren't they on LTS only anyway?

Kanstantsin Shautsou

unread,
Dec 30, 2015, 6:17:23 PM12/30/15
to jenkin...@googlegroups.com
Even IRC become quiet. I think because of holidays better wait for feedbacks bit longer time.
>
> --
> You received this message because you are subscribed to a topic in the Google Groups "Jenkins Developers" group.
> To unsubscribe from this topic, visit https://groups.google.com/d/topic/jenkinsci-dev/K06ny0sozWM/unsubscribe.
> To unsubscribe from this group and all its topics, send an email to jenkinsci-de...@googlegroups.com.
> To view this discussion on the web visit https://groups.google.com/d/msgid/jenkinsci-dev/6F5DC217-43B6-433B-B04B-9626AB19D510%40beckweb.net.
signature.asc

Jesse Glick

unread,
Jan 5, 2016, 11:31:50 AM1/5/16
to Jenkins Dev
On Wed, Dec 30, 2015 at 6:13 PM, Daniel Beck <m...@beckweb.net> wrote:
> do you think the feedback from e.g. companies is missing due to the December time frame?

Definitely a concern. On the other hand, we actually have _more_
ratings on both 1.642 and 1.643 than we do on 1.636—and they are
better. Not sure what that means. Maybe admins took advantage of the
quiet time to upgrade their systems, and did not see any obvious
problems, but users have not been doing much after the upgrade either?

As the author of both the regressions and their fixes my opinion
perhaps counts for little. :-)

Has anyone checked the status of ATH? There were some test failures
which just corresponded to changes in HTML in 1.637, which I think I
fixed.

When do we need to choose? Looks like there is a project meeting
tomorrow. KK is on vacation AFAIK.

ogondza

unread,
Jan 6, 2016, 11:19:34 AM1/6/16
to Jenkins Developers
I have prepared stable branch for both 1.636 and 1.642[1]. Running ATH agains both:

https://jenkins.ci.cloudbees.com/job/core/job/acceptance-test-harness-stable/549/
https://jenkins.ci.cloudbees.com/job/core/job/acceptance-test-harness-stable-1.642/1/

[1] https://github.com/olivergondza/jenkins/commits/stable-1.642 (yes, only one commit to backport if we pick such a new baseline)

--
oliver

ogondza

unread,
Jan 7, 2016, 2:57:43 AM1/7/16
to Jenkins Developers
On the meeting we agreed that the baseline will change to 1.642.

--
oliver
Reply all
Reply to author
Forward
0 new messages