Thanks. If there is no objection by end of Sunday, I'll start this work earlier next week.
Lin On Fri, Sep 18, 2009 at 12:42 AM, Jack Cai <[email protected]> wrote: > Agreed, since it won't hurt. > > -Jack > > On Fri, Sep 18, 2009 at 4:13 AM, Lin Sun <[email protected]> wrote: >> >> Hi, >> >> Recently, I opened GERONIMO-4683 in G about the Transaction.commit >> signature is missing the IllegalStateException. The reason why I >> raised this JIRA is because in OSGi RFC 98 (Transaction in OSGi) >> compliance test, we use Geronimo's JTA spec jar as the baseline. >> During OSGi RFC 98 compliance test run with an implementation of RFC >> 98, OSGi signature test currently checks strictly on exceptions throw >> by each of the method to see if it is the same as the baseline's >> signature, which is the Geronimo JTA spec jar. If it is not the same >> the test fail. For example, below is what is specified by the JTA >> java doc and G JTA spec. >> >> >> G JTA spec jar - Transaction.java >> public void commit() >> throws HeuristicMixedException, >> HeuristicRollbackException, >> RollbackException, >> SecurityException, >> SystemException; >> >> >> JTA 1.1 Java doc - Transaction.java >> public void commit() >> throws RollbackException, >> HeuristicMixedException, >> HeuristicRollbackException, >> SecurityException, >> IllegalStateException, >> SystemException >> >> What do you think of adding the missing unchecked exception >> "IllegalStateException" back to our JTA spec and release a newer >> version of the JTA spec jar just to be the same as what is in the Java >> doc? I think it is good for us to be consistent with what is in the >> JTA spec and we should be consistent in declaring the unchecked >> exceptions (we currently declares the SecurityException but not the >> IllegalStateException). >> >> p.s. if you are interested and have access to, the related discussion >> is OSGi alliance can be found here - >> https://www.osgi.org/members/bugzilla/show_bug.cgi?id=1447 >> >> Thanks >> >> Lin > >
