On Tue, Jun 22, 2010 at 4:58 AM, Nitro <ni...@dr-code.org> wrote: > Hello, > > I'm writing a custom data manager. The data manager should execute before > the connection's data manager.
I don't know what that means. > So I provided the "sortKey" method. This > works fine for the usual commit(). In particular, sortKey exists to address a deadlock that can occur in 2-phase commit. ... > This means for a sequence such as > > transaction.begin() > transaction.get().join( myDataManager ) > transaction.savepoint() > transaction.rollback() > transaction.commit() > > the data managers are not sorted unless commit is encountered. If I move > the sorting code to the join() method instead, I get the proper search > order. > > Is this a bug? No. sortKey is intentionally used just for commits. Why do you want savepoints to execute in sortKey order? Jim -- Jim Fulton _______________________________________________ For more information about ZODB, see the ZODB Wiki: http://www.zope.org/Wikis/ZODB/ ZODB-Dev mailing list - ZODB-Dev@zope.org https://mail.zope.org/mailman/listinfo/zodb-dev