perfsonar development work

[Nina Jeliazkova <>]

Nina Jeliazkova wrote:
> Michael,
>
> Could you advise what jars needs to be used in order to compile the 
> code? Just using perfsonar base still gives plenty of unknown classes.
Sorry, errors were due to eclipse misconfiguration .. I was too quick in 
complaining.

Regards,
Nina
> As you probably know, perfsonarUI doesn' really use perfsonar base, 
> except for AA, so I would like to avoid additional dependencies.
>
> Best regards,
> Nina
>
> Michael Bischoff wrote:
> > Hello all,
> >
> > It still needs some testing but level 0 is pretty finished.
> >
> > API specification
> > http://anonsvn.internet2.edu/svn/perfsonar/trunk/perfsonar_java-lsclient-api/ 
> >
> > API implementation
> > http://anonsvn.internet2.edu/svn/perfsonar/trunk/surfnet_java-lsclient-api-impl/ 
> >
> >
> > Some information
> >
> > ----------------------------
> > API specification
> > is a set of interfaces + classes that are used by the interfaces 
> > model objects and exceptions.
> >
> > LsCLient client = LsClientProvider.getClient();
> > should give you a client which implements all levels.
> >
> > You should only program against the API specification and not the 
> > implementation
> > (So don't import anything from nl.surfnet.*)
> >
> > ----------------------------
> > API implementation
> > Sending of soapmessages is based on SAAJ codebase result wrapped in a 
> > inputstream that
> > removes SOAP elements and attachments and thus only returns the 
> > contents of a SOAPbody as per
> > specification. The BufferedSoapStrippingStream is homegrown. I should 
> > write more tests for
> > it.
> >
> > Only level 0 is implemented calling other methods(level 1+2) will 
> > result in a
> > UnsupportedException,
> >
> > To use the implementation simply add the jar to the classpath.
> >
> > ========================================================================
> >
> > Jason could you check if the InputStream is ok (as opposed to 
> > returning a array of elements).
> > Sticking to a array of elements ties you down to dowm, proposed 
> > approach allows for SAX/StAX
> >
> > Greetings,
> >
> > Michael Bischoff
> >