Windows builds/tests for core disabled

25 views
Skip to first unread message

R. Tyler Croy

unread,
Feb 4, 2017, 11:11:32 AM2/4/17
to jenkin...@googlegroups.com
I just merged this[0] change to the Jenkinsfile which removes the Windows
platform from testing.

Due to an issue in the JUnit plugin (which has since been fixed) builds were
appearing to pass despite numerous platform specific test failures.

Unfortunately this means we have no useful testing on Windows for Jenkins core,
again. Those interested in making Jenkins and Windows play nicely together may
wish to coordinate their efforts, perhaps at the Jenkins hackathon on
Monday[1]?


[0] https://github.com/jenkinsci/jenkins/pull/2733
[1] https://www.meetup.com/jenkinsmeetup/events/236370750/

- R. Tyler Croy

------------------------------------------------------
Code: <https://github.com/rtyler>
Chatter: <https://twitter.com/agentdero>
xmpp: rty...@jabber.org

% gpg --keyserver keys.gnupg.net --recv-key 1426C7DC3F51E16F
------------------------------------------------------
signature.asc

Slide

unread,
Feb 4, 2017, 8:12:52 PM2/4/17
to jenkin...@googlegroups.com

I won't be at the hackathon, but I'll take a look. Some of the tests were written without any windows considerations taken into account, so the sooner we can get this fixed (so people who write new tests get feedback on windows being broken by them) the better.


--
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/20170204161125.GS26439%40blackberry.coupleofllamas.com.
For more options, visit https://groups.google.com/d/optout.

R. Tyler Croy

unread,
Feb 6, 2017, 12:11:30 PM2/6/17
to jenkin...@googlegroups.com
(replies inline)

On Sun, 05 Feb 2017, Slide wrote:

> I won't be at the hackathon, but I'll take a look. Some of the tests were
> written without any windows considerations taken into account, so the
> sooner we can get this fixed (so people who write new tests get feedback on
> windows being broken by them) the better.

Is there a compelling reason not to just mark all these failing tests as
skipped on the Windows platform and move on?
signature.asc

Slide

unread,
Feb 6, 2017, 12:24:09 PM2/6/17
to jenkin...@googlegroups.com
(replies inline)

On Mon, Feb 6, 2017 at 10:11 AM R. Tyler Croy <ty...@monkeypox.org> wrote:
(replies inline)

On Sun, 05 Feb 2017, Slide wrote:

> I won't be at the hackathon, but I'll take a look. Some of the tests were
> written without any windows considerations taken into account, so the
> sooner we can get this fixed (so people who write new tests get feedback on
> windows being broken by them) the better.

Is there a compelling reason not to just mark all these failing tests as
skipped on the Windows platform and move on?

Well, from what I can see, several of the tests are just newline issues, people expect \n only, when on Windows it should be \r\n. I don't think we should skip tests for just that reason, we should fix as many of the tests as possible so that we have as good of coverage on Windows as on Linux. Just skipping them seems too hacky to me. 

 

- R. Tyler Croy

------------------------------------------------------
     Code: <https://github.com/rtyler>
  Chatter: <https://twitter.com/agentdero>
     xmpp: rty...@jabber.org

  % gpg --keyserver keys.gnupg.net --recv-key 1426C7DC3F51E16F
------------------------------------------------------

--
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.

Daniel Beck

unread,
Feb 7, 2017, 2:06:20 AM2/7/17
to jenkin...@googlegroups.com

> On 06.02.2017, at 18:23, Slide <slide...@gmail.com> wrote:
>
> Just skipping them seems too hacky to me.

IMO we should get Windows testing back ASAP -- so if skip on Windows + TODO comment gets us there more quickly, we should do so.

While it doesn't improve the Windows coverage, it will at least prevent regressions.

Baptiste Mathus

unread,
Feb 7, 2017, 3:29:40 AM2/7/17
to Jenkins Developers
+1. Better reenable Windows tests ASAP. Agree to fix low hanging fruits, and just @ignore the harder ones to get back to a green build.

--
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-dev+unsubscribe@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/jenkinsci-dev/44B2E1D6-B1D7-4938-BCE3-0983EDEE4909%40beckweb.net.

Jesse Glick

unread,
Feb 7, 2017, 11:17:17 AM2/7/17
to Jenkins Dev
On Tue, Feb 7, 2017 at 3:29 AM, Baptiste Mathus <m...@batmat.net> wrote:
> @ignore the harder ones

Or rather, `Assume` we are running on Unix unless and until a more
portable test can be devised.

(Making serious changes to existing tests is a bit risky, because you
might inadvertently cause it to no longer prove the effectiveness of
the original bug fix. To be on the safe side, you need to comment out
the fix and verify that the test still fails in the expected way.)

Slide

unread,
Feb 7, 2017, 3:06:48 PM2/7/17
to Jenkins Dev

Most of the failures are due to tests assuming \n as the line ending if I remember. I was wondering if there was a log or CI available for someone to look at. I know it was up on ci.j.io, but doubt that many bills are kept that it would still be there.


--
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/CANfRfr3EB5gy%2BVPCKMxpj%2But5RjX_vOmDMMcZ_fnNQCKg-YB1A%40mail.gmail.com.

Jesse Glick

unread,
Feb 7, 2017, 3:42:24 PM2/7/17
to Jenkins Dev
On Tue, Feb 7, 2017 at 3:06 PM, Slide <slide...@gmail.com> wrote:
> I was wondering if there was a log or CI available for someone to
> look at.

Create a branch/PR reënabling Windows-based tests and follow the logs.

There are a bunch that use, say, `hudson.tasks.Shell` rather than some
portable test builder or a switch to `BatchFile`.

Slide

unread,
Feb 7, 2017, 3:52:04 PM2/7/17
to Jenkins Dev

I thought the Jenkinsfile wasn't picked up from anything but master. If I am mistaken, I'll gladly do that.


--
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.

R. Tyler Croy

unread,
Feb 8, 2017, 6:18:04 AM2/8/17
to jenkin...@googlegroups.com
(replies inline)

On Tue, 07 Feb 2017, Slide wrote:

> I thought the Jenkinsfile wasn't picked up from anything but master. If I
> am mistaken, I'll gladly do that.


That depends on whether you are able to commit to the repository or not. As a
committer, which IIRC you are, the GitHub Branch Source plugin (IIRC) will use
the pull request's Jenkinsfile.
signature.asc

Slide

unread,
Feb 8, 2017, 3:58:39 PM2/8/17
to jenkin...@googlegroups.com
It doesn't look like it did, I only see a Linux build happening see [1].  I don't think I have direct commit rights on jenkinsci/jenkins.



--
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.

Daniel Beck

unread,
Feb 8, 2017, 4:13:06 PM2/8/17
to jenkin...@googlegroups.com

> On 08.02.2017, at 21:58, Slide <slide...@gmail.com> wrote:
>
> It doesn't look like it did, I only see a Linux build happening see [1]. I don't think I have direct commit rights on jenkinsci/jenkins.

I created a branch and added the Windows build back on it.

https://github.com/jenkinsci/jenkins/tree/windows-ci

PRs against that branch should do what you need them to do.

Slide

unread,
Feb 12, 2017, 11:30:54 AM2/12/17
to jenkin...@googlegroups.com

I've made some progress on this. We started at about 60 something tests failing and we're now down to about 15 tests failing. Some of them need to be skipped on windows at this point. I don't remember how to do that, so if someone knows off the top of their head, that would be great.


--
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.

Baptiste Mathus

unread,
Feb 12, 2017, 7:27:35 PM2/12/17
to Jenkins Developers
Not tested, but as advised by Jesse, something like:

Assume.assumeFalse(System.getProperty("os.name").toLowerCase().startsWith("win"));
To simply disable those tests on Windows as a quick fix for current builds?
I agree it's even better than disabling them through @Ignore, to get us back to green, then take more time to possibly fix them more thoroughly/durably.

2017-02-12 17:30 GMT+01:00 Slide <slide...@gmail.com>:

I've made some progress on this. We started at about 60 something tests failing and we're now down to about 15 tests failing. Some of them need to be skipped on windows at this point. I don't remember how to do that, so if someone knows off the top of their head, that would be great.

On Wed, Feb 8, 2017, 14:13 Daniel Beck <m...@beckweb.net> wrote:

> On 08.02.2017, at 21:58, Slide <slide...@gmail.com> wrote:
>
> It doesn't look like it did, I only see a Linux build happening see [1].  I don't think I have direct commit rights on jenkinsci/jenkins.

I created a branch and added the Windows build back on it.

https://github.com/jenkinsci/jenkins/tree/windows-ci

PRs against that branch should do what you need them to do.

--
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-dev+unsubscribe@googlegroups.com.

--
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-dev+unsubscribe@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/jenkinsci-dev/CAPiUgVeN1xBtKM6RH1DcqBEtcLbffbsgukDRMHt2R8pnh6XeLg%40mail.gmail.com.

Slide

unread,
Feb 12, 2017, 8:23:01 PM2/12/17
to Jenkins Developers
We have Functions.isWindows() that I can use as well. I found out how to do it and never replied back to my original email. I am applying the last fixes and should have Windows green soon.

On Sun, Feb 12, 2017 at 5:27 PM Baptiste Mathus <m...@batmat.net> wrote:
Not tested, but as advised by Jesse, something like:

Assume.assumeFalse(System.getProperty("os.name").toLowerCase().startsWith("win"));
To simply disable those tests on Windows as a quick fix for current builds?
I agree it's even better than disabling them through @Ignore, to get us back to green, then take more time to possibly fix them more thoroughly/durably.

2017-02-12 17:30 GMT+01:00 Slide <slide...@gmail.com>:

I've made some progress on this. We started at about 60 something tests failing and we're now down to about 15 tests failing. Some of them need to be skipped on windows at this point. I don't remember how to do that, so if someone knows off the top of their head, that would be great.

On Wed, Feb 8, 2017, 14:13 Daniel Beck <m...@beckweb.net> wrote:

> On 08.02.2017, at 21:58, Slide <slide...@gmail.com> wrote:
>
> It doesn't look like it did, I only see a Linux build happening see [1].  I don't think I have direct commit rights on jenkinsci/jenkins.

I created a branch and added the Windows build back on it.

https://github.com/jenkinsci/jenkins/tree/windows-ci

PRs against that branch should do what you need them to do.

--
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.

--
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.
--
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/CANWgJS63nwb4ydjh_8ubJK62XMh0xMTmpvFMp%2Bg9OgdSOVFX0A%40mail.gmail.com.
Reply all
Reply to author
Forward
0 new messages