[Pellet-users] ANN: Pellet 1.5.2 Bug Fix Release

[Michael Smith]

Pellet Users,

Since you all read the Clark & Parsia blog [1], you've known for hours
that Pellet 1.5.2 was released today as a bug fix update to 1.5.1 and
can be downloaded at http://pellet.owldl.com/download/

If you haven't already, check out the post [2] for a high level release
summary and a description of how Moody's helped make 1.5.2 happen (since
we'd previously said 1.6 would be the next release).

What that post doesn't provide, which I aim to provide here, is a more
complete description of the changes from 1.5.1.  For each of the bugs
mentioned below, you can get even more details from the Pellet issue
tracker [3].


General bugs resolved:
#76 - Explanation cannot be turned off
#77 - Getting all the instances of a concept slowed down
#94 - Incorrect results in the presence of transitive properties and nominals
#96 - Handling guessing rule on the fly is broken
#98 - Default realization strategy slower
#108 - Missing entailments when XSD date/time values are used in hasValue restrictions


Jena binding bugs resolved:
#72 - Model.listStatements(null,p,o) returning triples (s,o,p)
#74 - Model.listStatements(null,p,null) incomplete when p is OWL.sameAs or OWL.differentFrom
#97 - ModelExtractor.extractIndividualModel very inefficient
#99 - Listing direct sub properties always returns empty set
#100 - RDF/XML-ABBREV writer cannot be used with Pellet
#113 - Anonymous inverse properties are not supported through the Jena interface
#120 - Jena loader may incorrectly type properties with multiple typing triples


OWLAPI binding bugs resolved:
#116 - Reasoner.getDescendantProperties returns only direct subproperties


In addition, there were a few changes that were not tied to open issues.

+ In [4], Evren described a change to the branching heuristic used for
cardinality restrictions that we expect will improve typical
performance.

+ In [5], Aditya Kalyanpur noted a bug in the KnowledgeBase copy method
that we resolved.

Finally, there were some tiny bugs quashed in datatype reasoning,
explanation and query for individuals added after consistency checks.

We don't anticipate regressions in this release, but its noteworthy that
two of the changes were heuristic changes based on our understanding of
the "typical" ontology.  So, if you see things noticeably slow down with
this release and you use cardinality restrictions with large numbers or
are doing a lot of instance realization, please let us know.  We
generally tune things for the cases we hear about.


Most of these issues were identified by contributors on pellet-users.
Thanks and keep up the good work.


Finally, for those of you hoping for some of the big ticket items that
we've hinted at in the past, don't worry.  We're progressing well
towards 1.6 and hope to have it out early summer.


Happy Reasoning.
-- 
Mike Smith

Clark & Parsia

[1] http://clarkparsia.com/weblog/
[2] http://clarkparsia.com/weblog/2008/05/01/pellet-152-release/
[3] http://cvsdude.com/trac/clark-parsia/pellet-devel/report/11
[4] http://lists.owldl.com/pipermail/pellet-users/2008-April/002605.html
[5] http://lists.owldl.com/pipermail/pellet-users/2008-April/002598.html
-- 
Mike S