Re: How to take pull -u from a repository instead of doing update only in mercurial using jenkins

[Varghese Renny]

  Yeah, recently one develper has deleted one file in remote repo, so after 
that build is getting failed pointing to the java file he deleted. so it is 
taking update only not pull.. I have given options like 
Repository url   corresponding url...
Branchnil 
Modules   nil
clean build  didn't check this
sudirectory  nil


This is my configuration option
Is there any mistake with tis?


Regards,
varghese

>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>

Antwort: How to take pull -u from a repository instead of doing update only in mercurial using jenkins

[Andreas Schilling]

Hi Varghese,

I'm not sure what problem you mean. The Jenkins hg plugin does a pull and 
an update. If files were deleted within the pulled changesets they also 
become deleted in the workspace.
Are you sure that not something else causes the problems you have?

Kind regards,
 
Andreas Schilling
CAE Processes & Data Management
---
Dipl. Inf. Andreas Schilling
Senior Software Architect

TWT GmbH
Science & Innovation
Bernhäuser Str. 40 - 42
73765 Neuhausen
 
Tel: +49 - 7158 - 17 15 - 673
E-Mail: andreas.schill...@twt-gmbh.de

www.twt-gmbh.de

Geschäftsführung: Dimitrios Vartziotis, Joachim Laicher (stv.)
Registergericht: Amtsgericht Stuttgart, HRB Nr. 212778
Umsatzsteuer: ID-Nr.: DE147841145




Von:Varghese Renny 
An: jenkinsci-users@googlegroups.com
Datum:  07.09.2012 07:27
Betreff:How to take pull -u from  a repository instead of doing 
update only in mercurial using jenkins
Gesendet von:   jenkinsci-users@googlegroups.com




Hi,
 I got an error because some develpers has deleted code from mercurial 
repository, since it is taking update only if something is deleted from 
repository it won't get reflect in the build workspace. What can i do for 
avoiding build getting failed due to this reason?














Regards,
varghese

How to take pull -u from a repository instead of doing update only in mercurial using jenkins

[Varghese Renny]

Hi,
 I got an error because some develpers has deleted code from mercurial 
repository, since it is taking update only if something is deleted from 
repository it won't get reflect in the build workspace. What can i do for 
avoiding build getting failed due to this reason?














Regards,
varghese

Re: how to handle build dependencies?

[Kenneth]

Hi all,
Thanks for so many responses.
I'm currently looking into filefingerprinting an copy artifacts plugin.
It sounds like the right way, and I can almost get it to work. ;-) but...
 
I'm stuck with the problem that the project referes it at ..\..\debug, 
while the jenkins output is different, and i need to make sure the compiled 
dll, is in the path, so it will 'link' correctly in jenkins builds...
 
how should i fix that ?

On Thursday, September 6, 2012 9:15:38 PM UTC+2, Kenneth wrote:

> I am quite new to Jenkins, and am faling in love ;-) (with him sound wrong 
> though)
>  
> Anyway I am trying to figure out how to handle projects that depends on 
> each other. I would really like to only have one job for each project. How 
> should I handle that one project relies on another? a.exe uses b.dll, b is 
> build when its the source changes, and a builds when its source change or b 
> i build...
>  
> I can ofcause place the binaries in a folder, and refer it from a's job, 
> but there must be a smarter way.
> The same problem appears with my unit test projects, that I make as 
> seperate dll's 
>  
> (I am using microsoft VC++ and C# and have subversion as my scm)
>  
> Any tips anyone?
> thanks
> Kenneth
>  
>  
>

Re: [VOTE] Bump Jenkins to Java 6

[Conny Kreyßel]

+1

I know thats not easy to update all kinds of nodes to Java 1.6+. But the 
jenkins core should be cleaned up to depend on Java 1.6.

Conny

Am Mittwoch, 5. September 2012 21:18:41 UTC+2 schrieb domi:
>
> Hi all, 
>
> just after todays meeting on the IRC chat, we startet a discussion about 
> upgrading Jenkins to Java 6. 
> As Java 5 has reached EOL since quite a while, some core developers have 
> asked whether it would be 
> OK to bump Jenkins' runtime dependency from Java5 to Java6. 
> The core is already build on Java6, but until now still backward 
> compatible with Java5. 
> Therefore we would like to know from you (Users) whether you have an issue 
> with this upgrade. 
> This would mean, that in the future you will have to have Java6 installed 
> to run Jenkins (for Master and Slave). 
>
> Here are the current usage numbers (installations we know of): 
> Java 1.5: 655 
> Java 1.6: 29164 
> Java 1.7: 2919 
>
> So please give us some feedback/votes on this. 
> Domi

Re: [VOTE] Bump Jenkins to Java 6

[Conny Kreyßel]

Am Mittwoch, 5. September 2012 21:18:41 UTC+2 schrieb domi:
>
> Hi all, 
>
> just after todays meeting on the IRC chat, we startet a discussion about 
> upgrading Jenkins to Java 6. 
> As Java 5 has reached EOL since quite a while, some core developers have 
> asked whether it would be 
> OK to bump Jenkins' runtime dependency from Java5 to Java6. 
> The core is already build on Java6, but until now still backward 
> compatible with Java5. 
> Therefore we would like to know from you (Users) whether you have an issue 
> with this upgrade. 
> This would mean, that in the future you will have to have Java6 installed 
> to run Jenkins (for Master and Slave). 
>
> Here are the current usage numbers (installations we know of): 
> Java 1.5: 655 
> Java 1.6: 29164 
> Java 1.7: 2919 
>
> So please give us some feedback/votes on this. 
> Domi

Re: how to handle build dependencies?

[Les Mikesell]

On Thu, Sep 6, 2012 at 2:15 PM, Kenneth  wrote:
> I am quite new to Jenkins, and am faling in love ;-) (with him sound wrong
> though)
>
> Anyway I am trying to figure out how to handle projects that depends on each
> other. I would really like to only have one job for each project. How should
> I handle that one project relies on another? a.exe uses b.dll, b is build
> when its the source changes, and a builds when its source change or b i
> build...
>
> I can ofcause place the binaries in a folder, and refer it from a's job, but
> there must be a smarter way.
> The same problem appears with my unit test projects, that I make as seperate
> dll's
>
> (I am using microsoft VC++ and C# and have subversion as my scm)
>
> Any tips anyone?

With subversion you can commit the .dll binaries back to the
repository  and use svn: externals to pull them into the workspace of
other projects.  That works the same with/without jenkins.

-- 
  Les Mikesell
 lesmikes...@gmail.com

populating the workspace on a slave

[acozine]

We use Jenkins for our rails projects and are moving to a master-and-slaves 
setup on EC2, where a small master spawns slaves as needed. The Amazon EC2 
plugin works well - it starts new instances and establishes a slave 
connection. However, builds that pass on the master fail on the slave.

On the master each project has a JENKINS_HOME/jobs/project_name/workspace 
directory with a clone of the git project and I can run the tests from the 
command line in the workspace or by triggering the build in Jenkins. I 
thought the master would just copy the workspace onto the slave and execute 
the job there, but it doesn't seem to. At first the slave's project 
workspace directory was empty. I added the Copy to Slave plugin and set it 
to copy *, but the rake task still aborts - and the workspace directory 
only has a few files in it.

The guide to setting up slaves 
(https://wiki.jenkins-ci.org/display/JENKINS/Step+by+step+guide+to+set+up+master+and+slave+machines)
 
seems to focus on getting slaves to run/connect to the master. That part's 
working fine! 

Anybody know where I can find documentation on how the master interacts 
with the slave at the job level? Examples of successful configuration for 
rails projects in a master/slave configuration? Suggestions? I'm new to 
Jenkins, so I may be missing something basic. 

Thanks,
acozine

Re: how to handle build dependencies?

[David Weintraub]

There are a couple of plugins  you might use:

* Copy Artifacts Plugin: This will copy artifacts from Build "A" to Build "B". 
* You can set Jenkins up to trigger one build from another. There are two 
places where this can be done. One is an option to trigger the job when another 
job finishes. That's right above the option to use your SCM to trigger a build. 
The other is a post build action to trigger another build. The end result is 
the same.

Let Build "A" that builds foo.dll that Build "B" needs trigger Build "B". Use 
the Copy Artifacts plugin to copy the foo.dll from Build "A" to Build "B". 
Problem solved.

It is possible to use a Maven repository combined with Ant and Ivy to store 
DLLs in your local Maven repository and fetch them as required. It's a hack, 
but I've seen it work.

On Sep 6, 2012, at 3:15 PM, Kenneth  wrote:

> I am quite new to Jenkins, and am faling in love ;-) (with him sound wrong 
> though)
>  
> Anyway I am trying to figure out how to handle projects that depends on each 
> other. I would really like to only have one job for each project. How should 
> I handle that one project relies on another? a.exe uses b.dll, b is build 
> when its the source changes, and a builds when its source change or b i 
> build...
>  
> I can ofcause place the binaries in a folder, and refer it from a's job, but 
> there must be a smarter way.
> The same problem appears with my unit test projects, that I make as seperate 
> dll's
>  
> (I am using microsoft VC++ and C# and have subversion as my scm)
>  
> Any tips anyone?
> thanks
> Kenneth
>  
>  

Re: how to handle build dependencies?

[Loren]

Sorry again, I thought you were using a repository. Wow, Maven really has 
me hard wired. I don't have an easy answer for you other than possibly 
executing a "Pre step" that sets up your downstream project with the 
upstream artifact, and linking the jobs together using a post-build action 
in your upstream job config.

Re: Subversion plugin failing with bad_record_mac

[John R. Cary]

Answering my own question:

I could not get jenkins to use sslv3, but by modifying my
server to also allow tlsv1 on the SSLProtocol definition, I
got the Jenkins subversion plugin to work.  The magic line
appears to be

  SSLProtocol -all +SSLv3 +TLSv1

John Cary

On 9/6/12 11:57 AM, John R. Cary wrote:

I have been trying to get jenkins (v1.478) to svn checkout
my repos.  I keep getting the error

org.tmatesoft.svn.core.SVNErrorMessage: svn: OPTIONS ...
...
bad_record_mac

Intensive searching has not found the answer to this
problem.  From http://issues.tmatesoft.com/issue/SVNKIT-176
I tried
JENKINS_JAVA_OPTIONS="-Djava.awt.headless=true -Dsvnkit.http.sslProtocols=SSLv3"
to no avail.

If anyone has any hints on how to fix this problem, I would
be most grateful, but in the absence of that...

The last time I had this working was with the svn plugin
1.34, so I would like to try downgrading to that, but
the jenkins GUI does not appear to give me that option.
The plugin manager gives me the update to 1.42 or the
downgrade to 1.39.

Is there a way to downgrade the subversion server
to version 1.42 in jenkins 1.478?

Thx.John Cary

Re: how to handle build dependencies?

[Loren]

Sorry, that was a Maven answer. Add a post-build action and you can specify 
what is to be built after a project builds.

>  
>  
>

Re: how to handle build dependencies?

[Loren]

Check the build trigger checkbox that says "Build when a snapshot 
dependency is built". 
So if A depends upon B and you check the box in A, then A will build after 
B builds.

Submit config.xml for Maven job but Maven release is not enabled

[Loren]

I have been using Hudson/Jenkins for a few years now. I manage roughly 350 
Maven artifacts and probably double that number of jobs.

To help with Maven releases I wish to automate the sequence of jobs that 
are to be executed and also test those jobs in a sandboxed env prior to 
release (so as not to disturb regular development). This requires 
programmatically creating and executing jobs.

My problem is that when I create a job by submitting a config.xml, Jenkins 
ignores the  section of the config.xml. This results in the 
"Maven release build" checkbox being unchecked in the interface, no 
"Perform Maven Release" button, and of course no ability to 
programmatically kick off a release. I can take the XML from an existing 
job that has this checkbox checked and submit it and the result is the same 
- no Maven release build. Jenkins just ignores an entire block of XML.

I examined the post that the configuration interface performs when a user 
checks this box and it appears to send every configuration value to the 
server in key/value pairs plus sending duplicate information as a JSON 
string (very odd but whatever). The parameter that sets the checkbox is 
org-jvnet-hudson-plugins-m2release-M2ReleaseBuildWrapper=on. I have tried 
sending subsets of values that I wish to edit but get an error telling me 
that it expects a post request (yes, I am posting). I assume that is just a 
catch-all error message that has nothing to do with get/post, but again, 
whatever.

I really don't want to have to reverse engineer this huge request (and 
build out the entire JSON string) just to be able to send a single boolean 
value to enable the release build. Is there a better way to create a Maven 
job with the "Maven release build" checkbox enabled?

Any suggestions would be appreciated.

how to handle build dependencies?

[Kenneth]

I am quite new to Jenkins, and am faling in love ;-) (with him sound wrong 
though)
 
Anyway I am trying to figure out how to handle projects that depends on 
each other. I would really like to only have one job for each project. How 
should I handle that one project relies on another? a.exe uses b.dll, b is 
build when its the source changes, and a builds when its source change or b 
i build...
 
I can ofcause place the binaries in a folder, and refer it from a's job, 
but there must be a smarter way.
The same problem appears with my unit test projects, that I make as 
seperate dll's 
 
(I am using microsoft VC++ and C# and have subversion as my scm)
 
Any tips anyone?
thanks
Kenneth
 
 

RE: MSTest Plugin doesn't fail the job on failing test.

[William Soula]

Is there anything in the log to denote a failure?  You could use the log parser 
to fail if it finds test failures.  But I agree with the guy below they should 
be marked unstable, which can also be done with the log parser plugin.

Will

From: jenkinsci-users@googlegroups.com [jenkinsci-users@googlegroups.com] on 
behalf of Mark Waite [markwa...@yahoo.com]
Sent: Thursday, September 06, 2012 12:50 PM
To: jenkinsci-users@googlegroups.com
Subject: Re: MSTest Plugin doesn't fail the job on failing test.

You might consider the comments in this thread before you mark a job as failed 
due to test failures:

http://jenkins.361315.n4.nabble.com/plugin-A-build-with-failed-tests-should-be-marked-as-Unstable-or-Failure-td3585663.html

If that guidance doesn't meet your needs and you decide that you still want to 
fail the build on test failures rather than mark the build "Unstable" on test 
failures, then I think the easiest way is to invoke mstest.exe directly 
(instead of using the MSTest plugin to launch it) and then have the MSTest 
batch file exit with a non zero exit code if any test fails.

Mark Waite




From: bearrito 
To: jenkinsci-users@googlegroups.com
Sent: Thursday, September 6, 2012 10:04 AM
Subject: MSTest Plugin doesn't fail the job on failing test.

I would like the MSTest plugin to fail the job when a test fails. Is this 
feasible ?

Re: problem installing plug-ins on windows pc

[Kenneth]

Found a fix.
The problem isn't with the browser, but with the fact that jenkins was 
installed under ..Program Files...
I unistalled and moved to C:\Jenkins\.. and erverything works thus far.
and the potential limit on path and filename lengths of 256 is a bit 
further off.
cheers
 
kfn
On Saturday, September 1, 2012 11:57:27 PM UTC+2, Kenneth wrote:

> Hi all,
> I am trying to get Jenkins to work, and I keep getting blocked by not 
> being able to download and install plugin files.
> Has anyone a solution for this?
> it might be a setting in internet explorer or something.
>  
> kfn
>

iPhone App To Help You Learn Chinese (Mandarin) Faster By Using Flashcards With Pictures

[Chinese Flashcards with Pictures]

http://goo.gl/VFim3 - "iChineseFlashcards" is an iPhone app that will help you 
learn Chinese (Mandarin) faster by using flashcards with pictures (learn over 
300 most commonly used words in the English / Chinese language from A to Z), 
thanks.

Subversion plugin failing with bad_record_mac

[John R. Cary]

I have been trying to get jenkins (v1.478) to svn checkout
my repos.  I keep getting the error

org.tmatesoft.svn.core.SVNErrorMessage: svn: OPTIONS ...
...
bad_record_mac

Intensive searching has not found the answer to this
problem.  From http://issues.tmatesoft.com/issue/SVNKIT-176
I tried
JENKINS_JAVA_OPTIONS="-Djava.awt.headless=true -Dsvnkit.http.sslProtocols=SSLv3"
to no avail.

If anyone has any hints on how to fix this problem, I would
be most grateful, but in the absence of that...

The last time I had this working was with the svn plugin
1.34, so I would like to try downgrading to that, but
the jenkins GUI does not appear to give me that option.
The plugin manager gives me the update to 1.42 or the
downgrade to 1.39.

Is there a way to downgrade the subversion server
to version 1.42 in jenkins 1.478?

Thx.John Cary

Re: MSTest Plugin doesn't fail the job on failing test.

[Mark Waite]

You might consider the comments in this thread before you mark a job as failed 
due to test failures:

http://jenkins.361315.n4.nabble.com/plugin-A-build-with-failed-tests-should-be-marked-as-Unstable-or-Failure-td3585663.html 


If that guidance doesn't meet your needs and you decide that you still want to 
fail the build on test failures rather than mark the build "Unstable" on test 
failures, then I think the easiest way is to invoke mstest.exe directly 
(instead of using the MSTest plugin to launch it) and then have the MSTest 
batch file exit with a non zero exit code if any test fails.

Mark Waite





>
> From: bearrito 
>To: jenkinsci-users@googlegroups.com 
>Sent: Thursday, September 6, 2012 10:04 AM
>Subject: MSTest Plugin doesn't fail the job on failing test.
> 
>
>I would like the MSTest plugin to fail the job when a test fails. Is this 
>feasible ?
>
>
>
>
>
>
>
>

Problem with MAVEN_HOME in a build

[mpapo - Michaël Pailloncy]

Hi,

We faced in a problem with Maven during a release build. The job is 
configured with the Release Plugin.
MAVEN_HOME variable is correctly set during the first build (and any 
'normals' builds), but when we launch a release and the build arrive to the 
command 'mvn .. release:clean release:prepare release:perform' to launch 
the Maven3 release, it seems that the MAVEN_HOME variable is not set 
correctly and the build crashs with this error : 

class not found:  org.codehaus.classworlds.Launcher  -  
org.codehaus.classworlds.Launcher 


We have also Maven2.2.1 and Maven3.0.3 installed on the master (AIX) which 
build this job.
In .profile file, M2_HOME is by default set to Maven2.2.1.

The problem is appeared since we have updated Jenkins with version 1.478.
The same job worked correctly before the update (last release has been 
built just before our update of Jenkins and the mvn command worked 
correctly).

Is this a bug ? Does I create a ticket in Jira ?

Thanks in advance from any help !
Michaël

Another Maven Release plugin error

[gsilverman]

I'm getting the following output when I build a project with the Maven
Release Build checked in the Maven Build Environment section of the project
configuration (My artifact pom.xml file has the maven release plugin,
version 2.3.2 property configured):

[INFO]

[INFO] BUILD FAILURE
[INFO]

[INFO] Total time: 48.671s
[INFO] Finished at: Thu Sep 06 12:24:19 EDT 2012
[INFO] Final Memory: 5M/15M
[INFO]

mavenExecutionResult exceptions not empty
message : Failed to execute goal
org.apache.maven.plugins:maven-release-plugin:2.3.2:prepare (default-cli) on
project edispense.csrfiles: Unable to commit files
Provider message:
The svn command failed.
Command output:
Fatal: no entropy gathering module detected
/bin/sh: line 1: 29846 Aborted svn --non-interactive commit
--file /tmp/maven-scm-1499199491.commit --targets
/tmp/maven-scm-8319584718810155363-targets

cause : Unable to commit files
Provider message:
The svn command failed.
Command output:
Fatal: no entropy gathering module detected
/bin/sh: line 1: 29846 Aborted svn --non-interactive commit
--file /tmp/maven-scm-1499199491.commit --targets
/tmp/maven-scm-8319584718810155363-targets

Stack trace : 
org.apache.maven.lifecycle.LifecycleExecutionException: Failed to execute
goal org.apache.maven.plugins:maven-release-plugin:2.3.2:prepare
(default-cli) on project edispense.csrfiles: Unable to commit files
Provider message:
The svn command failed.
Command output:
Fatal: no entropy gathering module detected
/bin/sh: line 1: 29846 Aborted svn --non-interactive commit
--file /tmp/maven-scm-1499199491.commit --targets
/tmp/maven-scm-8319584718810155363-targets

at
org.apache.maven.lifecycle.internal.MojoExecutor.execute(MojoExecutor.java:213)
at
org.apache.maven.lifecycle.internal.MojoExecutor.execute(MojoExecutor.java:153)
at
org.apache.maven.lifecycle.internal.MojoExecutor.execute(MojoExecutor.java:145)
at
org.apache.maven.lifecycle.internal.LifecycleModuleBuilder.buildProject(LifecycleModuleBuilder.java:84)
at
org.apache.maven.lifecycle.internal.LifecycleModuleBuilder.buildProject(LifecycleModuleBuilder.java:59)
at
org.apache.maven.lifecycle.internal.LifecycleStarter.singleThreadedBuild(LifecycleStarter.java:183)
at
org.apache.maven.lifecycle.internal.LifecycleStarter.execute(LifecycleStarter.java:161)
at org.apache.maven.DefaultMaven.doExecute(DefaultMaven.java:319)
at org.apache.maven.DefaultMaven.execute(DefaultMaven.java:156)
at
org.jvnet.hudson.maven3.launcher.Maven3Launcher.main(Maven3Launcher.java:79)
at sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method)
at
sun.reflect.NativeMethodAccessorImpl.invoke(NativeMethodAccessorImpl.java:39)
at
sun.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:25)
at java.lang.reflect.Method.invoke(Method.java:597)
at
org.codehaus.plexus.classworlds.launcher.Launcher.launchStandard(Launcher.java:329)
at
org.codehaus.plexus.classworlds.launcher.Launcher.launch(Launcher.java:239)
at org.jvnet.hudson.maven3.agent.Maven3Main.launch(Maven3Main.java:158)
at hudson.maven.Maven3Builder.call(Maven3Builder.java:112)
at hudson.maven.Maven3Builder.call(Maven3Builder.java:70)
at hudson.remoting.UserRequest.perform(UserRequest.java:118)
at hudson.remoting.UserRequest.perform(UserRequest.java:48)
at hudson.remoting.Request$2.run(Request.java:287)
at 
java.util.concurrent.Executors$RunnableAdapter.call(Executors.java:441)
at java.util.concurrent.FutureTask$Sync.innerRun(FutureTask.java:303)
at java.util.concurrent.FutureTask.run(FutureTask.java:138)
at
java.util.concurrent.ThreadPoolExecutor$Worker.runTask(ThreadPoolExecutor.java:886)
at
java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:908)



--
View this message in context: 
http://jenkins.361315.n4.nabble.com/Another-Maven-Release-plugin-error-tp4639422.html
Sent from the Jenkins users mailing list archive at Nabble.com.

MSTest Plugin doesn't fail the job on failing test.

[bearrito]

I would like the MSTest plugin to fail the job when a test fails. Is this 
feasible ?

Re: MSTestRunner plugin unable to launch mstest.exe

[bearrito]

I had the same issue. This is what works for me


foreach($testFolder in $testFolders)
{
Set-Location "C:\Program Files (x86)\Microsoft Visual Studio 
10.0\Common7\IDE"
$testPath = Join-Path $testFolder.FullName ( $testFolder.Name  + ".dll" )
$resultPath = Join-Path $PRIMARYWORKSPACE ( $testFolder.Name + ".trx")
$testSettingsPath = Join-Path $PRIMARYWORKSPACE (join-path 
"CoreBuildToolsPackage" "Jenkins.testSettings")
$command =  'mstest.exe /testcontainer:' + $testPath + " 
/resultsfile:"+$resultPath + " /testsettings:"+$testSettingsPath
cmd.exe /c $command

}


Here the $testFolders are just all the folders that contain test assemblies.

The key part is to set the location and then use the cmd.exe





On Tuesday, September 4, 2012 8:34:48 AM UTC-4, Arve Knudsen wrote:
>
> Hi guys
>
> I've configured the MSTestRunner plugin to use the following path to 
> MSTest (sans quotes): 'C:\Program Files (x86)\Microsoft Visual Studio 
> 11.0\Common7\IDE\mstest.exe'. When building, MSTest gets invoked as follows:
>
> cmd.exe /C "C:\Program Files (x86)\Microsoft Visual Studio 
> 11.0\Common7\IDE\mstest.exe" "/resultsfile:/resultsfile testresults" 
> /testcontainer:Project.Tests/bin/Debug/Project.Tests.dll && exit 
> %%ERRORLEVEL%%
>
> This leads to the following error:
>
> 'C:\Program' is not recognized as an internal or external 
> command, operable program or batch file.
>
> This must be a problem with MSTestRunner surely, as I have configured it 
> with the correct path to MSTest? Is there any way to work around this? I've 
> also asked on stackoverflow: 
> http://stackoverflow.com/questions/12263430/jenkins-mstestrunner-plugin-is-unable-to-launch-mstest-exe
> .
>
> Thanks!
> Arve
>

Re: Sharing files between master and slave (dropbox like)

[Les Mikesell]

On Thu, Sep 6, 2012 at 8:50 AM, Jerome Lacoste  wrote:
> I am trying to centralize on my master the list of files required for my
> various slaves.
>
[...]

> Possible "simple" implementations:
> * a directory on the slave that automatically gets synced with the master on
> a particular location e.g. JENKINS_HOME/slaveContent/$slaveName/
> * a file upload mechanism from Jenkin's web GUI ? (where I could upload
> files into userContent - or a perSlave directory)

We just use a common nfs/samba read-only file share for the tools and
common libraries needed on the slaves.  It's not particularly secure
against reading (just IP based restrictions), but files can only be
changed on the master which happens to be the host.  If network speed
is an issue for runtime access, you could rsync copies locally in a
job that could be scheduled or manual.  For our own library
components, we use svn externals to pull the right versions into the
right places which works the same whether it is a manual checkout or a
jenkins build.

There shouldn't be anything 'new' on a slave unless jenkins builds it,
in which case you can archive the build artifacts as part of the job
so they end up versioned back on the master.

-- 
   Les Mikesell
 lesmikes...@gmail.com

Subversion plugin failing with

[John R Cary]

I have been trying to get jenkins (v1.478) to svn checkout
my repos.  I keep getting the error

org.tmatesoft.svn.core.SVNErrorMessage: svn: OPTIONS ...
...
bad_record_mac

Intensive searching has not found the answer to this
problem.  From http://issues.tmatesoft.com/issue/SVNKIT-176
I tried
JENKINS_JAVA_OPTIONS="-Djava.awt.headless=true -Dsvnkit.http.sslProtocols=SSLv3"
to no avail.

If anyone has any hints on how to fix this problem, I would
be most grateful, but in the absence of that...

The last time I had this working was with the svn plugin
1.34, so I would like to try downgrading to that, but
the jenkins GUI does not appear to give me that option.
The plugin manager gives me the update to 1.42 or the
downgrade to 1.39.

Is there a way to downgrade the subversion server
to version 1.42 in jenkins 1.478?

Thx.John Cary

Sharing files between master and slave (dropbox like)

[Jerome Lacoste]

I am trying to centralize on my master the list of files required for my 
various slaves.

My current solution(s) including drawbacks are described below. Feel free 
to comment if you have a better alternative.

Because of security reasons, I want to keep a minimal number of connections 
between my slaves and my master. My slaves are in the cloud and I treat 
them as less trusted than my LAN master.
Ideally I would have liked some kind of dropbox like feature in jenkins. A 
folder I can drop files on my master or slave and have them automatically 
available on all machines (pull or push).

I don't have that so I've used a 2 way process:

* sharing files from Master to slave(s)

I've placed a Software/ catalog under the JENKINS_HOME/userContent/.
I pull the files when I need them.

DONE

* sharing files from slave to Master

Solution #1

Using the copy to slave plugin.
I created a simple free style job called infra_share_files that has a post 
build action that copies **/* from the slave's workspace to the master
On the server I also ran from JENKINS_HOME the following command:
  ln -s ~/jobs/infra_share_files/workspace/ userContent/slavesContent 

Usage:
1. I edit my job to run on the specific slave. 
2. [if first time for that particular slave] run the job
3. copy the relevant files on my slave under the job's workspace (I create 
a directory per slave)
4. run the job
5. after completion go to master file system and organize the files found 
on the master's workspace for the job into Jenkin's userContent directory

Drawbacks:
* things aren't fully automated
* this blocks my job queue... 
* the copy operation isn't very verbose
* the copy operation doesn't do a smart copying (like rsync would for 
example)

Alternative using a matrix job.

I created a similar matrix job called infra_share_files that has the same 
post build action as above. Select nodes per label and select all slaves 
but the master
On the server I also ran from JENKINS_HOME the following command:
ln -s ~/jobs/infra_share_files2/configurations/axis-label/ 
userContent/slavesContent

Changes from previous:
* advantage: no need to edit the job to choose which slave to run on
* drawback: given the copy to slave post build isn't smart, it takes a 
longer time (I have few slaves so it's OK)
* drawback the userContent/slavesContent/ navigation is uglier (not just 
the workspace is shared) (would be nice if on the server the matrix job was 
creating node specific links  under the workspace/ directory.


I can live with that as I don't do that everyday. Yet maybe there is a 
better solution ?


Possible "simple" implementations:
* a directory on the slave that automatically gets synced with the master 
on a particular location e.g. JENKINS_HOME/slaveContent/$slaveName/
* a file upload mechanism from Jenkin's web GUI ? (where I could upload 
files into userContent - or a perSlave directory)

Re: Jenkins doesn't work with Eclipse Mylyn

[alex ouzounis]

any news for a possible fix ?

Alex

On Tuesday, August 7, 2012 3:31:22 PM UTC+3, Stepan Vavra wrote:
>
> Hello,
> I've found that the problem is with the Winstone web server.
> When I deploy Jenkins to Jetty or Tomcat, everyting is fine.
>
> So the resolution is that Jenkins should have used working version of 
> Winstone.
>
> Regards,
> Stepan
>
> On Mon, Aug 6, 2012 at 3:23 PM, Stepan Vavra 
> > wrote:
>
>> I have some news regarding the Jenkins Mylyn integration.
>>
>> I've found, the last version it worked was: 1.457. Looking at changelog 
>> there is no record saying that any API changed in 1.458.
>> Therefore I think this is a bug.
>>
>> Regards,
>> Stepan
>>
>>
>>
>> On Sun, Aug 5, 2012 at 6:13 PM, Stepan Vavra 
>> 
>> > wrote:
>>
>>>
>>>
>>> Hello,

 I haven't had time to discover where exactly the problem is, but Mylyn 
 Builds Connector (1.0.1 installed from 
 http://download.eclipse.org/mylyn/releases/latest) doesn't work with 
 the latest Jenkins.

 I don't even know if Jenkins supports the API Mylyn is using.
 All I know is that Mylyn is able to communicate with Jeknins 1.447 but 
 is not able to communicate with the latest: 1.476.

 Is this some kind of regression that is intended?

 Best regards, 
>>>
>>> Stepan Vavra

>>>
>>>
>>
>>
>> -- 
>> Stepan Vavra
>>
>
>
>
> -- 
> Stepan Vavra
>

Re: JNLP Slave Behavior upon Server Bounce

[domi]

That I'm not 100% sure, but I think the jobs do get killed - even though the 
saves are not dead.
Domi


On 06.09.2012, at 14:59, "Mandeville, Rob"  wrote:

> Just to be sure: they don’t abort their jobs in this case?
>  
> --Rob
>  
> From: jenkinsci-users@googlegroups.com 
> [mailto:jenkinsci-users@googlegroups.com] On Behalf Of domi
> Sent: Thursday, September 06, 2012 8:58 AM
> To: jenkinsci-users@googlegroups.com
> Subject: Re: JNLP Slave Behavior upon Server Bounce
>  
> Hi Rob,
> yes they are, the slaves automatically reconnect to there master as soon as 
> its available again.
> Domi
>  
> On 06.09.2012, at 13:10, "Mandeville, Rob"  wrote:
> 
> 
> I’m running a fairly extensive Jenkins installation with about 120-140 slave 
> nodes, almost entirely on Linux (as is the server).  The server has been 
> hanging and taking up 100% CPU on its server from time to time, so I’ve had 
> to bounce the server.  With 12-hour test cycles, this can be…disruptive.  I 
> am trying to diagnose that problem, but while I’m doing that, I’m also trying 
> to figure out a way to be able to bounce the server and keep the jobs running.
>  
> Currently, all slaves have a launch method of “Launch slave via execution of 
> command on the Master”, and said commands are SSH jobs.  So, when I kill the 
> Jenkins server, its 120+ SSH jobs die because they’re subordinate processes, 
> killing the slaves and any jobs running on them.  I know that you can launch 
> JNLP slaves so that they aren’t subordinate jobs, and won’t get automatically 
> killed by Linux when you kill the server. 
>  
> So my question is: If I have a JNLP slave running a job and its Jenkins 
> server dies, will it re-establish the connection and continue the job it was 
> running when the server comes back up?
>  
> Thanks in advance,
>  
> --Rob
> The information in this message is for the intended recipient(s) only and may 
> be the proprietary and/or confidential property of Litle & Co., LLC, and thus 
> protected from disclosure. If you are not the intended recipient(s), or an 
> employee or agent responsible for delivering this message to the intended 
> recipient, you are hereby notified that any use, dissemination, distribution 
> or copying of this communication is prohibited. If you have received this 
> communication in error, please notify Litle & Co. immediately by replying to 
> this message and then promptly deleting it and your reply permanently from 
> your computer.

RE: JNLP Slave Behavior upon Server Bounce

[Mandeville, Rob]

Just to be sure: they don't abort their jobs in this case?

--Rob

From: jenkinsci-users@googlegroups.com 
[mailto:jenkinsci-users@googlegroups.com] On Behalf Of domi
Sent: Thursday, September 06, 2012 8:58 AM
To: jenkinsci-users@googlegroups.com
Subject: Re: JNLP Slave Behavior upon Server Bounce

Hi Rob,
yes they are, the slaves automatically reconnect to there master as soon as its 
available again.
Domi

On 06.09.2012, at 13:10, "Mandeville, Rob" 
mailto:rmandevi...@litle.com>> wrote:


I'm running a fairly extensive Jenkins installation with about 120-140 slave 
nodes, almost entirely on Linux (as is the server).  The server has been 
hanging and taking up 100% CPU on its server from time to time, so I've had to 
bounce the server.  With 12-hour test cycles, this can be...disruptive.  I am 
trying to diagnose that problem, but while I'm doing that, I'm also trying to 
figure out a way to be able to bounce the server and keep the jobs running.

Currently, all slaves have a launch method of "Launch slave via execution of 
command on the Master", and said commands are SSH jobs.  So, when I kill the 
Jenkins server, its 120+ SSH jobs die because they're subordinate processes, 
killing the slaves and any jobs running on them.  I know that you can launch 
JNLP slaves so that they aren't subordinate jobs, and won't get automatically 
killed by Linux when you kill the server.

So my question is: If I have a JNLP slave running a job and its Jenkins server 
dies, will it re-establish the connection and continue the job it was running 
when the server comes back up?

Thanks in advance,

--Rob
The information in this message is for the intended recipient(s) only and may 
be the proprietary and/or confidential property of Litle & Co., LLC, and thus 
protected from disclosure. If you are not the intended recipient(s), or an 
employee or agent responsible for delivering this message to the intended 
recipient, you are hereby notified that any use, dissemination, distribution or 
copying of this communication is prohibited. If you have received this 
communication in error, please notify Litle & Co. immediately by replying to 
this message and then promptly deleting it and your reply permanently from your 
computer.

Re: JNLP Slave Behavior upon Server Bounce

[domi]

Hi Rob,
yes they are, the slaves automatically reconnect to there master as soon as its 
available again.
Domi

On 06.09.2012, at 13:10, "Mandeville, Rob"  wrote:

> I’m running a fairly extensive Jenkins installation with about 120-140 slave 
> nodes, almost entirely on Linux (as is the server).  The server has been 
> hanging and taking up 100% CPU on its server from time to time, so I’ve had 
> to bounce the server.  With 12-hour test cycles, this can be…disruptive.  I 
> am trying to diagnose that problem, but while I’m doing that, I’m also trying 
> to figure out a way to be able to bounce the server and keep the jobs running.
>  
> Currently, all slaves have a launch method of “Launch slave via execution of 
> command on the Master”, and said commands are SSH jobs.  So, when I kill the 
> Jenkins server, its 120+ SSH jobs die because they’re subordinate processes, 
> killing the slaves and any jobs running on them.  I know that you can launch 
> JNLP slaves so that they aren’t subordinate jobs, and won’t get automatically 
> killed by Linux when you kill the server. 
>  
> So my question is: If I have a JNLP slave running a job and its Jenkins 
> server dies, will it re-establish the connection and continue the job it was 
> running when the server comes back up?
>  
> Thanks in advance,
>  
> --Rob
> The information in this message is for the intended recipient(s) only and may 
> be the proprietary and/or confidential property of Litle & Co., LLC, and thus 
> protected from disclosure. If you are not the intended recipient(s), or an 
> employee or agent responsible for delivering this message to the intended 
> recipient, you are hereby notified that any use, dissemination, distribution 
> or copying of this communication is prohibited. If you have received this 
> communication in error, please notify Litle & Co. immediately by replying to 
> this message and then promptly deleting it and your reply permanently from 
> your computer.

Re: Jenkins with TestLink

[cjo]

My suggestions were for triggering other jobs on the Jenkins server that 
run on different machines.

i.e 
Project A 
starts either 
Project 1 - tied to windows nodes
or
Project 2 - tied to linux nodes


Where the script in Projects 1 & 2 actually controls the tests, running on 
the node.
e.g. Copies files to test across and then calls a batch script, or shell 
commands.

This means that you label all of the nodes that these projects can run on 
with the approiate labels, 
Note:you can have multiple labels on slaves and projects and multiple 
slaves with the same labels

so you might label the nodes as:
Slave 1 - windows testlink
Slave 2 - linux testlink
Slave 3 - windows build
Slave 4 - linux build
Slave 5 - windows testlink

and the project 1 with windows testlink which would cause it to run on 
slave 1 or 5.
and the project 2 with linux testlink which would cause it to run on slave 2

RE: groovy 
This is basically the same as java with some additions so most valid java 
will work.
Here is an example of what you are trying to acheive using the Enviroment 
variable ABC, 
starting either windows_test or linux_test projects or failing the build if 
it is something else.

import hudson.model.Cause;
def envars = manager.build.getEnvironment(manager.listener);

if (envars.get("ABC") == "WINDOWS"){
// windows bit
  def foo = manager.hudson.getJob( 'windows_test' )
  foo.scheduleBuild( 2, new Cause.UpstreamCause( manager.build ))
} 
else if (envars.get("ABC") == "LINUX") {
// linux bit
  def foo = manager.hudson.getJob( 'linux_test' )
  foo.scheduleBuild( 2, new Cause.UpstreamCause( manager.build ))
} else {
//all others
  manager.createSummary("warning.gif").appendText("Not a valid 
platform", false, false, false, "red")
  manager.buildFailure()
}

And if you want to pass parameters to the downstream builds.
use the following 
import hudson.model.ParametersAction;

def params = []
  params.push( new StringParameterValue( 'MINOR', 2 ) )
  def foo = manager.hudson.getJob( 'Foo' )
  foo.scheduleBuild( 2, new Cause.UpstreamCause( manager.build ), new 
ParametersAction( params ) )
  
Hope this helps.
Chris

On Thursday, September 6, 2012 3:18:36 AM UTC+1, bandy wrote:
>
> Hi Chris,
>
> Thanks for the reply. I reviewed all options that you have mentioned. I am 
> trying to write a system groovy script for my use case. But there seems not 
> many resources available in the internet for groovy. I am stuck in place 
> where I want to trigger downstream job (in Windows or in Linux) depending 
> on environment variable from TestLink (which I am able to work with). I 
> dont have any idea so far how can i trigger build in remote machines using 
> groovy script.
>
> Any ideas on that or any other resource that you know might help??
>
> Rgrds
> bandy
>
>
>
> On Monday, September 3, 2012 7:23:10 PM UTC+8, cjo wrote:
>>
>> To run one or other of the downstream jobs, it currently not a simple 
>> case as, but there are are several different ways of doing it.
>>
>> Use a groovy post build script to trigger one of the builds
>>
>> You could use the Parametrized trigger [1] in combination with the 
>> conditional build step plugin,[2] to run the jobs as build steps within a 
>> job.
>>
>> This would mean having a conditional buildstep for each of the platforms, 
>> where the  environment variable is checked
>> which contains the trigger build on other projects for the correct 
>> project.
>>
>>
>> If you want to use the conditional part with the Parametrized trigger as 
>> a trigger for the downstream properly there is a pull request for that on 
>> github,
>> that you can build and test. [3]
>>
>> Chris
>>
>> [1]  
>> https://wiki.jenkins-ci.org/display/JENKINS/Parameterized+Trigger+Plugin 
>> [2]  
>> https://wiki.jenkins-ci.org/display/JENKINS/Conditional+BuildStep+Plugin 
>> [3]  https://github.com/jenkinsci/parameterized-trigger-plugin/pull/17 
>>
>> On Monday, September 3, 2012 10:54:45 AM UTC+1, bandy wrote:
>>>
>>> Hi,
>>>
>>> I am new in Jenkins and wandering for few days to get my head around the 
>>> topic "How to trigger downstream job based on certain conditions ?"
>>>
>>> Details:
>>>
>>> I have testlink integrated with jenkins which manages the test suite and 
>>> test cases. I have a custom field "Test Client Platform Type" in TestLink. 
>>> The tester can specify in the TestLink if the test case has to be run in 
>>> Windows or Linux Platform.
>>> I fetch the custom field in jenkins successfully but can't figure out 
>>> how to trigger downstream jobs in linux and windows client depending on the 
>>> environment variable.
>>>
>>> Any ideas ??
>>>
>>>
>>>

JNLP Slave Behavior upon Server Bounce

[Mandeville, Rob]

I'm running a fairly extensive Jenkins installation with about 120-140 slave 
nodes, almost entirely on Linux (as is the server).  The server has been 
hanging and taking up 100% CPU on its server from time to time, so I've had to 
bounce the server.  With 12-hour test cycles, this can be...disruptive.  I am 
trying to diagnose that problem, but while I'm doing that, I'm also trying to 
figure out a way to be able to bounce the server and keep the jobs running.

Currently, all slaves have a launch method of "Launch slave via execution of 
command on the Master", and said commands are SSH jobs.  So, when I kill the 
Jenkins server, its 120+ SSH jobs die because they're subordinate processes, 
killing the slaves and any jobs running on them.  I know that you can launch 
JNLP slaves so that they aren't subordinate jobs, and won't get automatically 
killed by Linux when you kill the server.

So my question is: If I have a JNLP slave running a job and its Jenkins server 
dies, will it re-establish the connection and continue the job it was running 
when the server comes back up?

Thanks in advance,

--Rob

The information in this message is for the intended recipient(s) only and may 
be the proprietary and/or confidential property of Litle & Co., LLC, and thus 
protected from disclosure. If you are not the intended recipient(s), or an 
employee or agent responsible for delivering this message to the intended 
recipient, you are hereby notified that any use, dissemination, distribution or 
copying of this communication is prohibited. If you have received this 
communication in error, please notify Litle & Co. immediately by replying to 
this message and then promptly deleting it and your reply permanently from your 
computer.

Re: passing parameters with build flow plugin

[ola nilsson]

Use the parameterized build plugin (for the unit-tests job) and setup a 
string variable called MYSVN_REV (SVN_REVISION is a 'built-in'). Pass the 
variable into the build from the flow DSL like you show above. Then, in the 
unit-tests job config use MYSVN_REV (it's an environment variable) in the 
Build Name Setter plugin.

Ola

On Monday, August 20, 2012 8:00:23 PM UTC+2, Tomek Kaczanowski wrote:
>
> Hi All, 
>
> any hints? 
>
> -- 
> Regards, 
> Tomek Kaczanowski 
>
> On 13 Sie, 17:45, Tomek Kaczanowski  
> wrote: 
> > Hi All, 
> > 
> > could someone post an example on how to read parameters passed from 
> > one job to another using Build Flow plugin? Plugin documentationhttps://
> wiki.jenkins-ci.org/display/JENKINS/Build+Flow+Plugingives 
> > information on how to put a parameter to the job, but does not show 
> > how to use them there. 
> > 
> > The scenario I'm trying to implement is that: 
> > a) job "compile" checks sources from SVN and compiles things: 
> > b = build( "compile" ) 
> > b) flow job passes this value to another job, like this: 
> > svn_rev = b.build.properties.environment['SVN_REVISION'] 
> > bUnit = build( "unit-tests" , svn_revision: svn_rev) 
> > c) job "unit-tests" should use this value in setting build name 
> (usinghttps://wiki.jenkins-ci.org/display/JENKINS/Build+Name+Setter+Plugin), 
>
> > however I can't figure out how to do it. Can't seem to be able to 
> > access svn_revision, from "unit-tests" job. 
> > 
> > Could someone enlighten me? I feel like missing something obvious. 
> > 
> > -- 
> > Regards, 
> > Tomek Kaczanowski 
>

EnvInject 1.69 possible bug

[Matthew.Webber]

Having just updated the EnvInject plugin from 1.68 to 1.69, I see some 
additional messages in the console log:

[EnvInject] - Variables injected successfully.
[EnvInject] - Unset unresolved 'GDA_beamline_ESRF_bm26a_developers' variable.
[EnvInject] - Unset unresolved 'GDA_beamline_DLS_b16_developers' variable.
[EnvInject] - Unset unresolved 'module' variable.
[EnvInject] - Unset unresolved 'LS_COLORS' variable.
(some entries snipped)

Previously, the only unresolved message I saw was "Unset unresolved 'module' 
variable"

The variables that the plugin is now unresolving are defined in the main 
Jenkins configuration "Global properties / Environment variables" section. Not 
all are being unresolved, just those which are defined as empty. This is on 
Linux, where there is a difference between an environment variable that exists 
and is empty, and an environment variable that does not exist.

I have to say that although the EnvInject plugin is quite useful, there seem to 
be too many releases which break something or change the behaviour as a 
side-effect of a change. So thanks to the author for writing a useful plugin, 
but please add some more tests to your code!

Thanks


--
This e-mail and any attachments may contain confidential, copyright and or 
privileged material, and are for the use of the intended addressee only. If you 
are not the intended addressee or an authorised recipient of the addressee 
please notify us of receipt by returning the e-mail and do not use, copy, 
retain, distribute or disclose the information in or attached to the e-mail.
Any opinions expressed within this e-mail are those of the individual and not 
necessarily of Diamond Light Source Ltd.
Diamond Light Source Ltd. cannot guarantee that this e-mail or any attachments 
are free from viruses and we cannot accept liability for any damage which you 
may sustain as a result of software viruses which may be transmitted in or with 
the message.
Diamond Light Source Limited (company no. 4375679). Registered in England and 
Wales with its registered office at Diamond House, Harwell Science and 
Innovation Campus, Didcot, Oxfordshire, OX11 0DE, United Kingdom

JENKINS CVS PLUGIN

[Deeps]

Hi,

Is there a way in which i can use a parameter specified at run time for 
CVS_BRANCH?
I'm running Jenkins version 1.475 and CVS Plugin version 1.6.
If not, can it please be included.

I'm doing a cvs checkout and then in my build step i'm using the CVS_BRANCH 
to run some shell commands through SSH Plugin. As i have managed to get 
everything automated through commands and ant scripts and all the input 
required by the user is the required changes in parameters at run-time (if 
required) and clicking. So if this option was available, it would make the 
project more 
 a. user friendly as a user can confirm the details before building the 
job and
 b. robust as configuration errors can be avoided as naive users will 
not have to fiddle with the job configuration page.

Thanks,
Deepak

Re: api to get username who created job or who changed job's configuration.!!!

[Vincent Latombe]

Hi,

you should check out
https://wiki.jenkins-ci.org/display/JENKINS/JobConfigHistory+Plugin

Vincent


2012/9/6 hitesh kumar 

> Hi,
> Any one knows about api which can get username who created job or last
> username who changed job configuration in last?
>
> regards,
> Hiteswar
>

api to get username who created job or who changed job's configuration.!!!

[hitesh kumar]

Hi,
Any one knows about api which can get username who created job or last
username who changed job configuration in last?

regards,
Hiteswar

Re: Timestamp in log output of a Maven build

[Michaël Pailloncy]

Thank you Domi. I'll test it right away.

Michaël

2012/9/5 domi 

> have a look at this:
> https://wiki.jenkins-ci.org/display/JENKINS/Timestamper
> Domi
>
> On 05.09.2012, at 17:39, mpapo - Michaël Pailloncy 
> wrote:
>
> > Hi,
> >
> > Is there a way to show timestamp in log output of a build ?
> > We are interested to know how many long time take each part of a Maven
> build.
> >
> > Thanks in advance !
> > Michaël
>
>

Re: [VOTE] Bump Jenkins to Java 6

[Stephen Connolly]

On 5 September 2012 23:16, Kohsuke Kawaguchi wrote:

> On 09/05/2012 01:18 PM, Brian Smith wrote:
>
>> Hi
>>
>> I don't think EOL alone is a good reason to upgrade the runtime
>> dependency, anybody concerned about it can run on a newer JVM anyway.
>>
>> It might help if someone were to outline the benefit of upgrading the
>> language version.
>>
>
> The benefits to developers are:
>
>  - We get to use a few APIs that we currently can't rely on.
>(Note that we already do use a number of Java6 APIs in ways that
>gracefully degrades when running on Java5, but this is separate.)
>
>  - Some IDE integrations (apparently) work better when what we tell
>as the compiler language level (1.5) matches with the runtime
>requirement (1.6)
>
>  - Some language level stuff (like @Override on interface methods)
>causes IDE and javac to disagree, which gets fixed with 1.6.
>
> The benefits are admittedly marginal, but the argument is that the cost is
> marginal, too --- just 2% of users on Java5, and I suspect those people
> aren't updating frequently.
>
> And at some point we need to move on, so I suppose it could well be now.
>
>
> I guess what I'm particularly keen on is if there are any minority
> platforms where Java6 isn't available easily, and/or desperate cries from
> users begging us not to require Java6, if any.
>

This is really where the issue is.

If you have Jenkins slaves that are on older OSes which are stuck with Java
1.5 as the "best" JVM they can run, please shout out now.

-Stephen


> (Personally, I'm neutral on this.)
>
>  What's the upside for jenkins dev?  Is there something in java 6 people
>> are hankering to use?
>>
>> Personally the only things in java 6 I've found useful were the new
>> concurrent collections and the ResourceBundle hooks.
>>
>> Cheers
>>
>> Brian
>>
>> On 5 September 2012 20:18, domi > > wrote:
>>
>> Hi all,
>>
>> just after todays meeting on the IRC chat, we startet a discussion
>> about upgrading Jenkins to Java 6.
>> As Java 5 has reached EOL since quite a while, some core developers
>> have asked whether it would be
>> OK to bump Jenkins' runtime dependency from Java5 to Java6.
>> The core is already build on Java6, but until now still backward
>> compatible with Java5.
>> Therefore we would like to know from you (Users) whether you have an
>> issue with this upgrade.
>> This would mean, that in the future you will have to have Java6
>> installed to run Jenkins (for Master and Slave).
>>
>> Here are the current usage numbers (installations we know of):
>> Java 1.5: 655
>> Java 1.6: 29164
>> Java 1.7: 2919
>>
>> So please give us some feedback/votes on this.
>> Domi
>>
>>
>>
>
> --
> Kohsuke Kawaguchi | CloudBees, Inc. | http://cloudbees.com/
> Try Nectar, our professional version of Jenkins
>