Subject RE: [IBO] IBS_Base more questions...
Author Jason Wharton
It may be the transaction in the schema cache for when metadata queries are
It will eventually timeout or you can explicitly cause it to close if you
really want to.

No worries about the tone, you did just great.

Jason Wharton

> -----Original Message-----
> From: []On
> Behalf Of Dany M
> Sent: Wednesday, December 20, 2006 3:32 PM
> To:
> Subject: Re: [IBO] IBS_Base more questions...
> Jason Wharton wrote:
> > The reason it sets the IsWaiting to true when a transaction
> is ( Started =
> > true ) is because the service app should remain involved to
> a higher degree
> > until all transactions are closed. If there is something
> causing your
> > transaction to remain ( Started = true ) then you should
> figure out what it
> > is and resolve it.
> Figuring out is what I do for *several* hours, always, before
> posting :)
> My transaction was never started outside of the
> DoServiceItems. Maybe I
> should have been clearer about that.
> After a some more prodding I have found out that it is an internal
> transaction of the connection (or maybe session) that is
> getting started
> after the new connection has been established. What I can not
> figure out
> is a) why it is left opened and b) why it does not happen when the
> service is started in the first place.
> If I put the following code after the call to connect in then
> EstablishConnections procedure:
> <code>
> cn.CloseTransactions;
> </code>
> or
> <code>
> for I := 0 to cn.TransactionCount - 1 do
> cn.Transactions[I].Commit;
> </code>
> then all is fine and the annoying "hang" of the IsWaiting
> flag dissapears!
> >
> > If you have modified versions of IBO source files I prefer
> to receive the
> > entire file. If it is based on a previous version of IBO
> then I also prefer
> > to receive both the modified file and the file it was based
> on. I prefer to
> > use my DIFF tool to handle these types of changes.
> >
> I regret that the discussion about this was in another
> thread, but you
> replied that you would get back on this but that it will take
> some time.
> That is all fine. I just included this in the message if
> there would be
> any chance that the two problems were related (which I do not
> think they
> are). I was merely trying to clarify myself.
> The sample states that it is not intended to run. The reason for this
> eludes me, but since I have only used it to look at and never
> ran it I
> cannot trace in the sample and thus it is very difficult for me to
> ascertain whether I messed anything up or the sample has bugs.
> I cannot possibly claim to have "correct code" for such
> elaborate parts
> of IBO.
> I apologize again if I'm not clear enough.
> Regards,
> /Dany
> Oh, and I really hope this post does not bear a bad tone. That is
> absolutely not my intention. English is not my first language.