Hi

Making this change

Index: contrib/reasoners/hermit/pom.xml
===================================================================
--- contrib/reasoners/hermit/pom.xml    (revision 1352483)
+++ contrib/reasoners/hermit/pom.xml    (working copy)
@@ -158,7 +158,6 @@
               org.semanticweb.HermiT.*;version=1.3.4,
               org.semanticweb.HermiT.datatypes.*;version=1.3.4,
               org.semanticweb.HermiT.datatypes.rdfplainliteral.*,
-              org.semanticweb.owlapi.*,
               rationals.*,
               uk.ac.manchester.cs.bhig.util,
               uk.ac.manchester.cs.owl.explanation.ordering,

allowed me to start the hermit Bundle in a Stanbol launcher running trunk.

I have only started the Bundle and have not actually tested the reasoner.

However looking at all the exported packages makes me wonder if they are really 
all needed. AFAIK one only exports packages that are directly/indirectly 
referenced by the API of the services provided by a bundle. As I am assuming 
the the hermit bundle is only implementing a reasoning service I would assume 
that it does not need to export any packages.

hope this helps
best
Rupert

On 21.06.2012, at 19:26, Alessandro Adamou wrote:

> More follow up, hoping it will give enough info for someone to be able to 
> help out.
> 
> I've just tried the reasoners.hermit bundle (latest trunk revision) on the 
> 0.9.0 incubating full launcher.
> 
> Of course I had to adapt the parent reference and version numbers first, but 
> then the plugin built and ran fine.
> 
> Something from the post-release must definitely be tampering with it. But I 
> cannot figure what it is.
> 
> Could anyone please at least double-check this? (works on release, does not 
> work on trunk head)
> 
> Best regards
> Alessandro
> 
> 
> On 6/21/12 1:21 PM, Alessandro Adamou wrote:
>> Another update: if I no longer export the OW API packages, I manage to 
>> activate the bundle, but not the 
>> org.apache.stanbol.reasoners.hermit.HermitReasoningService component.
>> 
>> The logged exception is a LinkageError, still on the OWL API
>> 
>> 21.06.2012 13:12:55.933 *WARN* [FelixPackageAdmin] 
>> org.apache.stanbol.reasoners.hermit 
>> [org.apache.stanbol.reasoners.hermit.HermitReasoningService] activate cannot 
>> be found (java.lang.LinkageError: loader constraint violation: loader 
>> (instance of org/apache/felix/framework/ModuleImpl$ModuleClassLoaderJava5) 
>> previously initiated loading for a different type with name 
>> "org/semanticweb/owlapi/model/OWLOntology")
>> 
>> (stack trace is just Felix and classloader stuff)
>> 
>> Mabe if I could somehow tell the reasoners.hermit module NOT to bring in its 
>> own OWL API by not using Embed-Dependency, but that would be a problem as 
>> HermiT requires files that are not just in Java packages.
>> 
>> Best,
>> Alessandro
>> 
>> 
>> On 6/21/12 12:57 PM, Alessandro Adamou wrote:
>>> I am bringing this one up again in case it was lost to the Jenkins problem 
>>> thread.
>>> 
>>> I do know that the HermiT JAR also includes the OWL API, but this has been 
>>> working so far (also during the Workshop that was based on 
>>> 0.9.0-incubating), how can this ever not be working anymore just after 
>>> changing module versions?
>>> 
>>> I've tried to use <_exportcontents>, which does solve the split package 
>>> issue but not the constraint violation.
>>> 
>>> Any ideas?
>>> 
>>> Thanks
>>> 
>>> Alessandro
>>> 
>>> 
>>> On 6/20/12 3:32 PM, Alessandro Adamou wrote:
>>>> Hi,
>>>> 
>>>> I've been trying to install the HermiT reasoner bundle in the full 
>>>> launcher, but ever since it was moved to contrib and version numbers 
>>>> changed I cannot get it to run.
>>>> 
>>>> I get this package conflict when trying to start the bundle on Java 1.6.31 
>>>> Mac OS 10.6.8. Also the message seems a bit different from other package 
>>>> conflict messages in the past (new Felix?):
>>>> 
>>>> =========
>>>> ERROR: Bundle org.apache.stanbol.reasoners.hermit [175]: Error 
>>>> starting/stopping bundle. (org.osgi.framework.BundleException: Constraint 
>>>> violation for package 'org.semanticweb.owlapi.util' when resolving module 
>>>> 175.0 between existing export 175.0.org.semanticweb.owlapi.util NO BLAME 
>>>> and uses constraint 59.0.org.semanticweb.owlapi.util BLAMED ON [[175.0] 
>>>> package; 
>>>> (&(package=org.apache.stanbol.reasoners.owlapi)(version>=0.10.0)), [101.0] 
>>>> package; (&(package=org.semanticweb.owlapi.util)(version>=0.10.0))])
>>>> org.osgi.framework.BundleException: Constraint violation for package 
>>>> 'org.semanticweb.owlapi.util' when resolving module 175.0 between existing 
>>>> export 175.0.org.semanticweb.owlapi.util NO BLAME and uses constraint 
>>>> 59.0.org.semanticweb.owlapi.util BLAMED ON [[175.0] package; 
>>>> (&(package=org.apache.stanbol.reasoners.owlapi)(version>=0.10.0)), [101.0] 
>>>> package; (&(package=org.semanticweb.owlapi.util)(version>=0.10.0))]
>>>>        at org.apache.felix.framework.Felix.resolveBundle(Felix.java:3449)
>>>>        at org.apache.felix.framework.Felix.startBundle(Felix.java:1727)
>>>>        at 
>>>> org.apache.felix.framework.Felix.setBundleStartLevel(Felix.java:1333)
>>>>        at 
>>>> org.apache.felix.framework.StartLevelImpl.run(StartLevelImpl.java:270)
>>>>        at java.lang.Thread.run(Thread.java:680)
>>>> =========
>>>> 
>>>> So apparently it's going berserk on multiple OWL API packages, but 
>>>> reasoners.owlapi is inheriting them from commons.owl. The HermiT jar has 
>>>> its own OWL API packages, but it has been working fine so far. I don't 
>>>> remember the versions of HermiT or the OWL API changing in Stanbol.
>>>> 
>>>> I've tried that with the revision just before my commits from today and 
>>>> keep having the same problem.
>>>> 
>>>> Does anyone have a clue? It would really be appreciated.
>>>> 
>>>> Thanks
>>>> 
>>>> Alessandro
>>>> 
>>> 
>>> 
>> 
>> 
> 
> 
> -- 
> M.Sc. Alessandro Adamou
> 
> Alma Mater Studiorum - Università di Bologna
> Department of Computer Science
> Mura Anteo Zamboni 7, 40127 Bologna - Italy
> 
> Semantic Technology Laboratory (STLab)
> Institute for Cognitive Science and Technology (ISTC)
> National Research Council (CNR)
> Via Nomentana 56, 00161 Rome - Italy
> 
> 
> "I will give you everything, just don't demand anything."
> (Ettore Petrolini, 1917)
> 
> Not sent from my iSnobTechDevice
> 

Reply via email to