You should work with turbine in cvs. Or wait a few more hours/day for
Jason to build the next tdk.
"Frank W. Nolden" wrote:
>
> maybe this is a stupid question, but I do want to have things completely clear:
>
> I have been using MySQL with autoincrement. This does work when doing inserts. The
>autoincrement field is automatically incremented when inserting a field.
> I want to do the same for Oracle, but Oracle does not support autoincrement fields.
>I have found a triggering mechanism (outside of Turbine and has to be inserted
>manually or added via the templates) that can be used for autoincrement fields in
>combination with the Sequence, but this fails when using it from the peer objects.
>Oracle complains about the unique primary key being null, which is not allowed. It
>looks like Oracle executes the trigger a little too late ;-( This is when using the
>autoincrement keyword in the table definition.
>
> Reading the message below sounds like that Oracle can be used with Sequence and
>without the special trigger by only using the sequence keyword in the table
>definition? Do the peer objects work this out then? Which version of Turbine/Torque
>do I need to get this functionality? The version from CVS like John mentioned?
> Could you please clarify?
> Many thanks in advance,
>
> Frank
> ----- Original Message -----
> From: Turpin, Jay
> To: '[EMAIL PROTECTED]'
> Sent: Sunday, May 13, 2001 01:00
> Subject: RE: Oracle, autoincrment and idtable
>
> I don't know if this might solve your problem with using Oracle sequence,
> but to get autoincrement to work properly, we had to modify the following
> class:
>
> org.apache.turbine.util.db.map.TurbineMapBuilder
>
> It has hard-coded all of the table maps use for the Turbine Security layer
> and specified that each table should use IDBROKER. Once we changed it to
> AUTOINCREMENT for SQL Server 2000, the security classes began working
> properly with the tables we created. For Oracle, you should be able to
> change it to SEQUENCE for each table.
>
> Also, we noticed that this class overwrites the table map for TurbineUser,
> even though it's defined in it's own mapbuilder class. This class should be
> modified to either remove the TurbineUser table map definition, or perform a
> check to only create a tablemap if it doesn't already exist.
>
> Regards,
> Jay Turpin
> Intel Corporation
>
> -----Original Message-----
> From: Frank W. Nolden [mailto:[EMAIL PROTECTED]]
> Sent: Saturday, May 12, 2001 2:01 PM
> To: [EMAIL PROTECTED]
> Subject: Re: Oracle, autoincrment and idtable
>
> Thanks. Although I hoped this not to be true it does confirm my first
> thoughts.
> I will first try and use the idbroker solution (I agree it to be preferreed
> since it is DB independent) and later try the Oracle sequence as it is
> supported in the latest CVS turbine.
>
> BTW I tries to use triggers inserting the autoincrement field using the
> sequence. However, the peers do not like this or do not support this. Still
> get the Oracle error about a unique indexed field not null allowed. Is the
> Sequence support you are referring to in the CVS a fix for this behaviour?
> Regards,
>
> Frank
>
> ----- Original Message -----
> From: "John McNally" <[EMAIL PROTECTED]>
> To: <[EMAIL PROTECTED]>
> Sent: Friday, May 11, 2001 6:22 PM
> Subject: Re: Oracle, autoincrment and idtable
>
> > you will need to add rows for the additional tables to ID_TABLE
> > manually. Also the OM/Peer system can use Oracle sequences, you will
> > want the latest cvs of turbine though. The IDBroker solution is
> > recommended.
> >
> > john mcnally
> >
> > "Frank W. Nolden" wrote:
> > >
> > > I was trying to use Oracle with autoincrement table fields, but
> unfortunately it does not support this. It only supports this with triggers,
> but then we cannot define primary keys being not null. Is there a solution
> for this?
> > >
> > > Now we want to use the idtable broker for use with Oracle, but I cannot
> seem to generate this correctly. Where do I have to configure the use of the
> idbroker using the id_table? I have done it in the XML project definition,
> but the code does not seem to be generated.
> > > Any ideas?
> > > Regards,
> > >
> > > Frank
> >
> > ---------------------------------------------------------------------
> > To unsubscribe, e-mail: [EMAIL PROTECTED]
> > For additional commands, e-mail: [EMAIL PROTECTED]
> >
> >
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: [EMAIL PROTECTED]
> For additional commands, e-mail: [EMAIL PROTECTED]
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: [EMAIL PROTECTED]
> For additional commands, e-mail: [EMAIL PROTECTED]
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
