On 02/02/2011 18:30, ant elder wrote:
On Wed, Feb 2, 2011 at 9:43 AM, Mike Edwards
<mike.edwards.inglen...@gmail.com>  wrote:
On 02/02/2011 09:04, ant elder wrote:

I'll guess that Raymond will want to get the pass-by-reference
improvements into this release so what ever happens we'll need to
respin to pick those up,  what else needs to get done for that respin?
Do we need to make further sample improvements right now or can that
be done over later releases?

Florian - have you found why that itest is using the wrong directory
in your environment?

Mike - I'm hoping the dom4j issue will go away if it uses the version
in the central repo, but what about the sample running issues you
first mentioned, do they sound like the same as this issue thats just
come up with the beta1 samples -
http://apache.markmail.org/message/jzu3c2ixbcqwbie7

    ...ant

Ant,

1) dom4j issue

This is a real issue and there is a fix.

The dom4j POM is coming from the intalio repo and this is FORCED by a
<repository>  setting in the POM of implementation-bpel-runtime - if you
start with a clean repo you WILL get the "bad" dom4j POM:

    <repositories>
        <repository>
            <id>intalio.org</id>
            <url>http://www.intalio.org/public/maven2</url>
           <releases>
               <enabled>true</enabled>
           </releases>
           <snapshots>
               <enabled>false</enabled>
           </snapshots>
        </repository>
    </repositories>

The fix is simply to remove this<repository>  setting from the POM of
implementation-bpel-runtime.


The problem with that is that the reason that repository was there was
to get the saxon artifacts so with out it those now aren't found. I'll
go have a hunt around and see if there is anywhere else we could use
for saxon, otherwise i guess we can add it to the tuscany svn repo.

    ...ant

Ant,

OK, works for me.

There is ANOTHER problem with dependencies associated with OSGi that needs fixing too - it is in the dependencies for node-launcher-equinox. The impact means that the build does not work correctly under OSGi and that some tests fail.

I've done a fix and committed it now.

Related, there is a fix in the interface-wsdl module to ensure that Tuscany passes testcase BWS_2016 from OASIS (it was trying to fix THAT testcase that revealed the OSGi problems...).


Yours, Mike.

Reply via email to