I understand that you all have limited bandwidth. In fact, I'm very grateful
for ALL of your hard efforts. I'm only trying to do my part by submitting
whatever fixes I can.
I'll look through the Castor test cases and see if I can doctor one up to
test this particular scenario.
Still, I think there may be a scalability problem with the Castor patch
process as it stands now. That's why I had asked for a R-W account on CVS.
++jeff
On Wednesday 16 January 2002 05:01 pm, Bruce Snyder wrote:
> Jeffrey Wescott wrote:
> > I haven't heard anything back about this, and I just took a CVS snapshot
> > and it doesn't look like it made it in yet. What's going on?
> >
> > I'm worried about this because we have to submit another patch soon and
> > it's becoming a nightmare to maintain the diffs from the CVS. Any chance
> > we can get a R-W password for CVS for bugfixing? This next one will be
> > our third patch.
> >
> > ++jeff
> >
> > On Tuesday 08 January 2002 05:24 pm, Jeffrey Wescott wrote:
> > > Hi.
> > >
> > > I think I found a bug with RelationCollection.remove() that occurs when
> > > the following conditions are true:
> > >
> > > 1- You have a many-to-many mapping from one object to another, but *NO*
> > > mapping in the other direction.
> > >
> > > 2- Your mapping is a "lazy" collection.
> > >
> > > What happens is that in the ClassMolder.preStore() method,
> > > RelationCollection.find() is called for any objects that were deleted
> > > in the transaction. However, since the reverse mapping was not there,
> > > the _loaded Map in RelationCollection does not contain the object to be
> > > deleted, and thus RelationCollection.find() returns null.
> > >
> > > My fix simply adds the object being deleted to the _loaded map so that
> > > the call to find() later in the transaction will find the correct
> > > object and remove the association.
> > >
> > > cvs diff -u patch is attached.
> > >
> > > Cheers.
>
> Jeff,
>
> Unless you've actually submitted some test cases for the patch, it may
> take a bit for someone to review your patch. The reason being that for
> any of us to sit down and create code for that specific situation takes
> time and effort. All of us are working on Castor part-time, so it would
> be a great help if you could submit some code to test the situations for
> which you are sending patches. If you've already sent test code with
> each patch, please forgive me.
>
> Bruce
>
> -----------------------------------------------------------
> If you wish to unsubscribe from this mailing, send mail to
> [EMAIL PROTECTED] with a subject of:
> unsubscribe castor-dev
-----------------------------------------------------------
If you wish to unsubscribe from this mailing, send mail to
[EMAIL PROTECTED] with a subject of:
unsubscribe castor-dev
