Maven 2.1 meets Classworlds 1.2-alpha-13?

[Benjamin Bentmann]

Hi,

if possible, I suggest to update Maven 2.1 to use the latest version of
Classworlds instead of 1.2-alpha-12. The recent Hudson package is still
failing for an encoding like UTF-16, among others due to PLX-367 which
prevents Maven from booting.

In this context: Brian has once setup some ITs to run Maven with
file.encoding=UTF-16. Since Maven 2.0.x is to my knowledge locked to an
old Classworlds version (i.e. there is no chance for the ITs to ever
pass), it might be best to focus on the future and upgrade these ITs to
run for Maven 2.1 instead.


Benjamin

---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@...
For additional commands, e-mail: dev-help@...

[Jason van Zyl-2]

Sure we can coordinate that, I have an OSGi like classloader in a  
branch of ClassWorlds so I can cut a release and upgrade.

But I would still like to work on the Hudson bundle to improve support  
for Window, and then I would like to add

1) platform support, and by that I mean building everything that we  
depend on from source in preparation for
2) trunk/branch support where we can easily try any changes on a the  
trunk of a dependency or a branch of a dependency

This would enable folks (exactly what Shane is doing now) to make  
changes somewhere in a dependency like ClassWorlds, Plexus, or Maven  
itself and easily run through all the standard tests to make sure  
everything works. Once this is done you could easily try this yourself  
and validate it. Then making the suggestion, we agree and it should be  
as easy as pressing a button to upgrade everything and commit the  
changes.

If you want to make the change locally and validate with the Hudson  
bundle I'm fine with the change. Go for it.

On 13-Jul-08, at 7:41 AM, Benjamin Bentmann wrote:

> Hi,
>
> if possible, I suggest to update Maven 2.1 to use the latest version  
> of Classworlds instead of 1.2-alpha-12. The recent Hudson package is  
> still failing for an encoding like UTF-16, among others due to  
> PLX-367 which prevents Maven from booting.
>
> In this context: Brian has once setup some ITs to run Maven with  
> file.encoding=UTF-16. Since Maven 2.0.x is to my knowledge locked to  
> an old Classworlds version (i.e. there is no chance for the ITs to  
> ever pass), it might be best to focus on the future and upgrade  
> these ITs to run for Maven 2.1 instead.
>
>
> Benjamin
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: dev-unsubscribe@...
> For additional commands, e-mail: dev-help@...
>

Thanks,

Jason

----------------------------------------------------------
Jason van Zyl
Founder,  Apache Maven
jason at sonatype dot com
----------------------------------------------------------

People develop abstractions by generalizing from concrete examples.
Every attempt to determine the correct abstraction on paper without
actually developing a running system is doomed to failure. No one
is that smart. A framework is a resuable design, so you develop it by
looking at the things it is supposed to be a design of. The more  
examples
you look at, the more general your framework will be.

   -- Ralph Johnson & Don Roberts, Patterns for Evolving Frameworks


---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@...
For additional commands, e-mail: dev-help@...

[Benjamin Bentmann]

Jason van Zyl wrote:

> If you want to make the change locally and validate with the Hudson
> bundle I'm fine with the change. Go for it.

Committed to trunk, the local Hudson didn't show any difference (i.e.
only the two already known ITs were failing). But know it's at least
possible to run something as simple as "mvn -v" in combination with
MAVEN_OPTS=-Dfile.encoding=UTF-16 without Maven crashing.

Sonatype's Hudson instance just started re-building and also considering
the few changes [0] in plexus-classworlds since 1.2-beta-12 I am optimistic.


Benjamin


[0]
http://fisheye.codehaus.org/changelog/plexus/plexus-classworlds/trunk?todate=1215131485859

---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@...
For additional commands, e-mail: dev-help@...

[Benjamin Bentmann]

Benjamin Bentmann wrote:

> Sonatype's Hudson instance just started re-building

... and happily failed at running the ITs. The bottom line of the build

Exception in thread "main" java.lang.NoClassDefFoundError:
/home/j2ee-hudson/apache-maven-2/1-SNAPSHOT/boot/plexus-classworlds-1/2-alpha-13/jar

indicates a relation with my change.

I believe I tracked this down to some problematic setup: It appears the
bootstrapping is not deleting the previous Maven distribution, i.e.
never deletes files but only adds/updates files. Hence, the contents of
M2_HOME/boot will be
- plexus-classworlds-1.2-alpha-12.jar (from before) and
- plexus-classworlds-1.2-alpha-13.jar (by my change)

Then the line
   -classpath "${M2_HOME}"/boot/plexus-classworlds-*.jar \
from the Unix shell scripts will pick up both these JARs but separates
them with a space, making the later JAR to be interpreted as the main
class for the JVM.


Benjamin

---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@...
For additional commands, e-mail: dev-help@...

[Jason van Zyl-2]

On 13-Jul-08, at 10:56 AM, Benjamin Bentmann wrote:

> Benjamin Bentmann wrote:
>
>> Sonatype's Hudson instance just started re-building
>
> ... and happily failed at running the ITs. The bottom line of the  
> build
>
> Exception in thread "main" java.lang.NoClassDefFoundError: /home/
> j2ee-hudson/apache-maven-2/1-SNAPSHOT/boot/plexus-classworlds-1/2-
> alpha-13/jar
>
> indicates a relation with my change.
>
> I believe I tracked this down to some problematic setup: It appears  
> the bootstrapping is not deleting the previous Maven distribution,  
> i.e. never deletes files but only adds/updates files. Hence, the  
> contents of M2_HOME/boot will be
> - plexus-classworlds-1.2-alpha-12.jar (from before) and
> - plexus-classworlds-1.2-alpha-13.jar (by my change)

I'll check this as at the bottom of the bootstrap it should delete all  
the installed files for the installation.

Can I see a tree of the structure it left there? (You can also take a  
look at the build.xml file in the bootstrap, maybe I mucked up the Ant).

>
>
> Then the line
>  -classpath "${M2_HOME}"/boot/plexus-classworlds-*.jar \
> from the Unix shell scripts will pick up both these JARs but  
> separates them with a space, making the later JAR to be interpreted  
> as the main class for the JVM.
>
>
> Benjamin
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: dev-unsubscribe@...
> For additional commands, e-mail: dev-help@...
>

Thanks,

Jason

----------------------------------------------------------
Jason van Zyl
Founder,  Apache Maven
jason at sonatype dot com
----------------------------------------------------------

You are never dedicated to something you have complete confidence in.
No one is fanatically shouting that the sun is going to rise tomorrow.
They know it is going to rise tomorrow. When people are fanatically
dedicated to political or religious faiths or any other kind of
dogmas or goals, it's always because these dogmas or
goals are in doubt.

   -- Robert Pirzig, Zen and the Art of Motorcycle Maintenance


---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@...
For additional commands, e-mail: dev-help@...

[Jason van Zyl-2]

Thanks for tracking that down. I've fixed it in the bootstrap. The  
build just produced on the CI machine doesn't duplicate the files  
anymore.

On 13-Jul-08, at 11:39 AM, Jason van Zyl wrote:

>
> On 13-Jul-08, at 10:56 AM, Benjamin Bentmann wrote:
>
>> Benjamin Bentmann wrote:
>>
>>> Sonatype's Hudson instance just started re-building
>>
>> ... and happily failed at running the ITs. The bottom line of the  
>> build
>>
>> Exception in thread "main" java.lang.NoClassDefFoundError: /home/
>> j2ee-hudson/apache-maven-2/1-SNAPSHOT/boot/plexus-classworlds-1/2-
>> alpha-13/jar
>>
>> indicates a relation with my change.
>>
>> I believe I tracked this down to some problematic setup: It appears  
>> the bootstrapping is not deleting the previous Maven distribution,  
>> i.e. never deletes files but only adds/updates files. Hence, the  
>> contents of M2_HOME/boot will be
>> - plexus-classworlds-1.2-alpha-12.jar (from before) and
>> - plexus-classworlds-1.2-alpha-13.jar (by my change)
>
> I'll check this as at the bottom of the bootstrap it should delete  
> all the installed files for the installation.
>
> Can I see a tree of the structure it left there? (You can also take  
> a look at the build.xml file in the bootstrap, maybe I mucked up the  
> Ant).
>
>>
>>
>> Then the line
>> -classpath "${M2_HOME}"/boot/plexus-classworlds-*.jar \
>> from the Unix shell scripts will pick up both these JARs but  
>> separates them with a space, making the later JAR to be interpreted  
>> as the main class for the JVM.
>>
>>
>> Benjamin
>>
>> ---------------------------------------------------------------------
>> To unsubscribe, e-mail: dev-unsubscribe@...
>> For additional commands, e-mail: dev-help@...
>>
>
> Thanks,
>
> Jason
>
> ----------------------------------------------------------
> Jason van Zyl
> Founder,  Apache Maven
> jason at sonatype dot com
> ----------------------------------------------------------
>
> You are never dedicated to something you have complete confidence in.
> No one is fanatically shouting that the sun is going to rise tomorrow.
> They know it is going to rise tomorrow. When people are fanatically
> dedicated to political or religious faiths or any other kind of
> dogmas or goals, it's always because these dogmas or
> goals are in doubt.
>
>  -- Robert Pirzig, Zen and the Art of Motorcycle Maintenance
>
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: dev-unsubscribe@...
> For additional commands, e-mail: dev-help@...
>

Thanks,

Jason

----------------------------------------------------------
Jason van Zyl
Founder,  Apache Maven
jason at sonatype dot com
----------------------------------------------------------

believe nothing, no matter where you read it,
or who has said it,
not even if i have said it,
unless it agrees with your own reason
and your own common sense.

  -- Buddha


---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@...
For additional commands, e-mail: dev-help@...

[Benjamin Bentmann]

Jason van Zyl wrote:

> The build just produced on the CI machine doesn't duplicate the files anymore.

Yep, back to "normal". Is there a reason why the job "maven-2.1.x-ITs"
over at Sonatype invokes Maven in the directory
   /home/j2ee-hudson/.hudson/jobs/maven-2.1.x-ITs/workspace
instead of
 
/home/j2ee-hudson/.hudson/jobs/maven-2.1.x-ITs/workspace/core-integration-tests
? It currently doesn't find the POM.

A related suggestion: Instead of issuing a "cd" command, wouldn't it be
more interesting to invoke Maven using the -f switch? I mean, Maven
builds should not be subject to the current directory so leaving this as
is will add another test aspect to the ITs.


Benjamin

---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@...
For additional commands, e-mail: dev-help@...

[Jason van Zyl-2]

On 13-Jul-08, at 2:33 PM, Benjamin Bentmann wrote:

> Jason van Zyl wrote:
>
>> The build just produced on the CI machine doesn't duplicate the  
>> files anymore.
>
> Yep, back to "normal". Is there a reason why the job "maven-2.1.x-
> ITs" over at Sonatype invokes Maven in the directory
>  /home/j2ee-hudson/.hudson/jobs/maven-2.1.x-ITs/workspace
> instead of
> /home/j2ee-hudson/.hudson/jobs/maven-2.1.x-ITs/workspace/core-
> integration-tests
> ? It currently doesn't find the POM.

Configuration error on my part.

>
>
> A related suggestion: Instead of issuing a "cd" command, wouldn't it  
> be more interesting to invoke Maven using the -f switch? I mean,  
> Maven builds should not be subject to the current directory so  
> leaving this as is will add another test aspect to the ITs.

Try it. I just hacked that together. I'm going to put this in a repo  
where you guys can hack on it but if you send me changes that work  
I'll incorporate them.

>
>
>
> Benjamin
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: dev-unsubscribe@...
> For additional commands, e-mail: dev-help@...
>

Thanks,

Jason

----------------------------------------------------------
Jason van Zyl
Founder,  Apache Maven
jason at sonatype dot com
----------------------------------------------------------

In short, man creates for himself a new religion of a rational
and technical order to justify his work and to be justified in it.

   -- Jacques Ellul, The Technological Society


---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@...
For additional commands, e-mail: dev-help@...

[Jason van Zyl-2]

On 13-Jul-08, at 2:33 PM, Benjamin Bentmann wrote:

> Jason van Zyl wrote:
>
>> The build just produced on the CI machine doesn't duplicate the  
>> files anymore.
>
> Yep, back to "normal". Is there a reason why the job "maven-2.1.x-
> ITs" over at Sonatype invokes Maven in the directory
>  /home/j2ee-hudson/.hudson/jobs/maven-2.1.x-ITs/workspace
> instead of
> /home/j2ee-hudson/.hudson/jobs/maven-2.1.x-ITs/workspace/core-
> integration-tests
> ? It currently doesn't find the POM.

Fixed.

Also happy to set you up as a user there and you can fix those things  
yourself.

>
>
> A related suggestion: Instead of issuing a "cd" command, wouldn't it  
> be more interesting to invoke Maven using the -f switch? I mean,  
> Maven builds should not be subject to the current directory so  
> leaving this as is will add another test aspect to the ITs.
>
>
> Benjamin
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: dev-unsubscribe@...
> For additional commands, e-mail: dev-help@...
>

Thanks,

Jason

----------------------------------------------------------
Jason van Zyl
Founder,  Apache Maven
jason at sonatype dot com
----------------------------------------------------------

Selfish deeds are the shortest path to self destruction.

  -- The Seven Samuari, Akira Kirosawa


---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@...
For additional commands, e-mail: dev-help@...

[Jason van Zyl-2]

On 13-Jul-08, at 2:33 PM, Benjamin Bentmann wrote:

> Jason van Zyl wrote:
>
>> The build just produced on the CI machine doesn't duplicate the  
>> files anymore.
>
> Yep, back to "normal". Is there a reason why the job "maven-2.1.x-
> ITs" over at Sonatype invokes Maven in the directory
>  /home/j2ee-hudson/.hudson/jobs/maven-2.1.x-ITs/workspace
> instead of
> /home/j2ee-hudson/.hudson/jobs/maven-2.1.x-ITs/workspace/core-
> integration-tests
> ? It currently doesn't find the POM.
>
> A related suggestion: Instead of issuing a "cd" command, wouldn't it  
> be more interesting to invoke Maven using the -f switch? I mean,  
> Maven builds should not be subject to the current directory so  
> leaving this as is will add another test aspect to the ITs.

I updated this as well now. The job on the CI  machine now uses this  
and I've updated the bundle.

>
>
>
> Benjamin
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: dev-unsubscribe@...
> For additional commands, e-mail: dev-help@...
>

Thanks,

Jason

----------------------------------------------------------
Jason van Zyl
Founder,  Apache Maven
jason at sonatype dot com
----------------------------------------------------------

First, the taking in of scattered particulars under one Idea,
so that everyone understands what is being talked about ... Second,
the separation of the Idea into parts, by dividing it at the joints,
as nature directs, not breaking any limb in half as a bad carver might.

   -- Plato, Phaedrus (Notes on the Synthesis of Form by C. Alexander)


---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@...
For additional commands, e-mail: dev-help@...