At 9:33 PM -0800 4/23/01, Joshua J. Kugler wrote:
>Yes, you can, but if someone else inserted a row between the time you insert
>and the time you select, it won't be accurate. You run into a race condition.
That's incorrect, actually. LAST_INSERT_ID() is returned on a
connection-specific basis. You get back your own previous AUTO_INCREMENT
value, even if someone else creates a new record in the meantime. So
the original advice (SELECT LAST_INSERT_ID()) is correct.
>Yeah, I know, a pain. Let me know if you figure anything out. :)
>On Monday 23 April 2001 20:56, Andrés Becerra wrote:
>> how about this then?
>> insert into tbWhatever (...) values (...)
>> select max(auto_increment_field) from tbWhatever
>> -----Original Message-----
>> From: Joshua J. Kugler [mailto:isd@stripped]
>> Sent: Tuesday, April 24, 2001 12:36 AM
>> To: Andrés Becerra
>> Subject: Re: Results from Insert
>> The problem is, this doesn't wokr in a Microsoft Jet workspace, only in an
>> ODBCDirect work space (if using Access). But using an ODBCDirect work
>> space breaks other things. So, it's a catch 22.
>> j----- k-----
>> On Monday 23 April 2001 20:00, you wrote:
>> > select LAST_INSERT_ID() as NewCustomerID
>> > and then use the 1st field in the recordset that gets returned
>> > -----Original Message-----
> > > From: Julian Lynch [mailto:julian.lynch@stripped]
>> > Sent: Monday, April 23, 2001 11:39 PM
>> > To: myodbc@stripped
>> > Subject: Results from Insert
>> > When performing an insert on a table that has an auto_increment primary
>> > field, I would like to know what the primary key value of the row is. I
>> > do not know hoe to do this without the support of something like stored
>> > procedures etc. Does anybody know how this may be done?
>> > Thanks
> > > Julian Lynch
Paul DuBois, paul@stripped