[ImageJ-devel] Maven puzzle

[Curtis Rueden]

Hi Tobi,

> It turns out, that Jenkins has a problem with it

I see: I could not reproduce because the new "try/catch with resources"
code was not yet on the master branch.

I share your surprise that somehow overriding the scijava.jvm.version
property is not enough. I have run into similar subtle issues with property
overrides for dependency versions, but more recently was unable to
reproduce them. It probably has something to do with how the POMs are
resolved, and would be worth our time to understand better so that we avoid
similar issues in the future.

Regards,
Curtis

On Fri, Feb 13, 2015 at 9:26 AM, Tobias Pietzsch <pietzsch at mpi-cbg.de>
wrote:

> Hi Curtis
>
> On 11 Feb 2015, at 17:45, Curtis Rueden <ctrueden at wisc.edu> wrote:
>
> Hi Tobi,
>
> > [ERROR]
> >
> /Users/pietzsch/workspace/BigDataViewer/bigdataviewer-server/src/main/java/bdv/server/BigDataServer.java:[248,46]
> > multi-catch statement is not supported in -source 1.5
> >
> > However, shouldn’t this be already automatically done by setting
> scijava.jvm.version=1.7???
>
> Whoa, strange. Especially since it says *1.5*, which is the Maven
> default...
>
> Furthermore, I cannot reproduce this problem on my system. Building with
> Java 6, Java 7 or Java 8 (with JAVA_HOME set accordingly), and with only
> the scijava.jvm.version set to 1.7, I am able to build the project
> successfully.
>
>
> I pushed some changes to master, so that we can see whether Jenkins has a
> problem with it.
>
>
> It turns out, that Jenkins has a problem with it:
> http://jenkins.imagej.net/view/BigDataViewer/job/bigdataviewer-server/8/
> has the error I reported earlier.
> Adding int the maven-compiler-plugin snippet fixed it.
> (http://jenkins.imagej.net/view/BigDataViewer/job/bigdataviewer-server/9/)
> Probably there was no code to trigger the behaviour on master before.
>
> all the best,
> Tobias
>
>
> Regards,
> Curtis
>
> On Wed, Feb 11, 2015 at 9:58 AM, Tobias Pietzsch <pietzsch at mpi-cbg.de>
> wrote:
>
>> Hi,
>>
>> maybe one of the maven experts can shed some light on the following
>> behaviour…
>>
>> In my bigdataviewer-server pom.xml file (
>> https://github.com/bigdataviewer/bigdataviewer-server/blob/84b6724ccdefbdcf787f4896a102bc39d415063a/pom.xml
>> )
>> I want to set source compatibility to 1.7.
>> I tried to do this by specifying
>> <properties>
>> <scijava.jvm.version>1.7</scijava.jvm.version>
>> </properties>
>> Strangely, this alone does not solve the problem. mvn on the command line
>> says:
>> ...
>> INFO] -------------------------------------------------------------
>> [ERROR] COMPILATION ERROR :
>> [INFO] -------------------------------------------------------------
>> [ERROR]
>> /Users/pietzsch/workspace/BigDataViewer/bigdataviewer-server/src/main/java/bdv/server/BigDataServer.java:[248,46]
>> multi-catch statement is not supported in -source 1.5
>>   (use -source 7 or higher to enable multi-catch statement)
>> …
>>
>> I can fix it by additionally putting this snippet:
>>   <build>
>> <plugins>
>>         <plugin>
>>           <artifactId>maven-compiler-plugin</artifactId>
>>           <configuration>
>>             <source>1.7</source>
>>             <target>1.7</target>
>>           </configuration>
>>         </plugin>
>> </plugins>
>> </build>
>> into the pom file.
>> However, shouldn’t this be already automatically done by setting
>> scijava.jvm.version=1.7???
>> At least Eclipse seems to think so: I fact, the above snippet to
>> configure maven-compiler-plugin I just copied from the “Effective POM”
>> shown in Eclipse (for the pom without explicitly having the snippet of
>> course…)
>>
>> Does anyone have an explanation? Is this a maven bug?
>>
>> all the best,
>> Tobias
>>
>> _______________________________________________
>> ImageJ-devel mailing list
>> ImageJ-devel at imagej.net
>> http://imagej.net/mailman/listinfo/imagej-devel
>>
>>
> _______________________________________________
> ImageJ-devel mailing list
> ImageJ-devel at imagej.net
> http://imagej.net/mailman/listinfo/imagej-devel
>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://imagej.net/pipermail/imagej-devel/attachments/20150213/88ecad2c/attachment.html>