Just a small example of the fact that people need such functionality... and will devise other ways, albeit inefficient and dangerous, to implement the missing feature.
The success of an RDBMS (or any other product for that matter) depends on how well it strikes the balance between the standards implementation, and what the users need. Gurjeet. On 17 May 2006 02:31:20 -0400, Greg Stark <[EMAIL PROTECTED]> wrote:
Thomas Hallgren <[EMAIL PROTECTED]> writes: > Some users of PL/Java make use of a non-default connection from within a > Trigger in order to do this. In essence, they load the client JDBC package into > the backend to let the backend as such become a client. The second connection > is then maintained for the lifetime of the first. Perhaps not the most > efficient way of doing it but it works. And you can do the same thing with any of the PL languages that have database drivers like Perl or Python. It might be a little less inefficient using one of them -- and probably a lot less code. You should be aware of the risk of deadlocks if you touch the same resources. Because the database is unaware that your main transaction is waiting for this other session to complete it won't be able to detect any deadlocks that depend on this hidden dependency. -- greg
---------------------------(end of broadcast)--------------------------- TIP 2: Don't 'kill -9' the postmaster