Move to JRuby 1.1.3 for implementation.script

[Raymond Feng <en...@gmail.com>]

Hi,

I ran into an issue with jruby-complete 1.0. The jar is pulled in by 
tuscany-implementation-script. Its pom.xml has a bug (an invalid property) 
which prevents the tuscany-maven-bundle-plugin from creating an OSGi bundle 
from the jruby-complete 1.0 artifact.

I did a bit experiment to upgrade to jruby-complete-1.1.3 (whose pom is 
good). I had to copy the latest version of 
com.sun.script.jruby.JRubyScriptEngine to replace TuscanyJRubyScriptEngine. 
The implementation.script module can be built successfully with all the unit 
tests passing. Is it a proof that the global variable issue we try to work 
around is now fixed by the latest version of JRubyScriptEngine?

With this change, I can run tuscany-maven-bundle-plugin to produce bundles 
for tuscany modules.

Please let me know if it's safe to move to JRuby 1.1.3.

Thanks,
Raymond 

Re: Move to JRuby 1.1.3 for implementation.script

[ant elder <an...@gmail.com>]

On Wed, Aug 27, 2008 at 10:44 PM, Raymond Feng <en...@gmail.com> wrote:

> Hi,
>
> I ran into an issue with jruby-complete 1.0. The jar is pulled in by
> tuscany-implementation-script. Its pom.xml has a bug (an invalid property)
> which prevents the tuscany-maven-bundle-plugin from creating an OSGi bundle
> from the jruby-complete 1.0 artifact.
>
> I did a bit experiment to upgrade to jruby-complete-1.1.3 (whose pom is
> good). I had to copy the latest version of
> com.sun.script.jruby.JRubyScriptEngine to replace TuscanyJRubyScriptEngine.
> The implementation.script module can be built successfully with all the unit
> tests passing. Is it a proof that the global variable issue we try to work
> around is now fixed by the latest version of JRubyScriptEngine?
>
> With this change, I can run tuscany-maven-bundle-plugin to produce bundles
> for tuscany modules.
>
> Please let me know if it's safe to move to JRuby 1.1.3.
>
> Thanks,
> Raymond
>

Sounds fine to me if all the tests are parsing.

   ...ant