Duey,

I believe you should submit your problem to the RDCC.  If this
behavior occurs it seems to contradict the definition of
STP_RETURN usage.  I could see potential problems if nested
stored procedures (user functions) with different return types
exist.

Jim Bentley

--- "Heffelfinger, Duane" <[EMAIL PROTECTED]> wrote:

> Mike,
> 
> That's exactly what I thought I could do as well, but at least
> when SET
> TRACE ON is running it appears as though the procedure won't
> be called
> at all if the STP_RETURN variable and the procedure RETURN
> types are
> different.  I may add to my enhancement requests some day, but
> for now
> converting RETURN types to text and using the INT function
> works just
> great.
> 
> Thanks for everyone's input.
> 
> Duey
> 
> 
> 
> 
> -----Original Message-----
> From: [email protected] [mailto:[EMAIL PROTECTED] On Behalf
> Of MikeB
> Sent: Wednesday, May 21, 2008 11:33 AM
> To: RBASE-L Mailing List
> Subject: [RBASE-L] - Re: STP_RETURN Variable when running a
> procedure
> 
> Duey,
>   From 6.x days, I have cleared the Store Proc system variable
> (STP_RETURN and 
> MICRORIM_RETURN) as the first line of the procedure.  Perhaps
> it's _not_
> 
> required anymore, but no harm comes of it.
> 
> 
> 
> ----- Original Message ----- 
> From: "Heffelfinger, Duane" <[EMAIL PROTECTED]>
> To: "RBASE-L Mailing List" <[email protected]>
> Sent: Wednesday, May 21, 2008 12:27 PM
> Subject: [RBASE-L] - Re: STP_RETURN Variable when running a
> procedure
> 
> 
> Larry,
> 
> Interesting.  Lately I've been doing this with SET TRACE ON
> (Rbase 7.6)
> and getting a variable type error before the procedure even
> begins when
> the procedure RETURN type is different than the STP_RETURN
> variable type
> from a previously called procedure.  I went ahead and
> converted my
> RETURN value to text in my integer procedures, then changed my
> call
> statements to something like:
> 
> SET VAR vvar = (INT(CALL actconvert (.vactactual)))
> 
> And all is well.  Not sure how we differ in what we're doing. 
> I really
> like procedures as well and have moved a lot of code to this
> method.
> 
> Duey
> 
> -----Original Message-----
> From: [email protected] [mailto:[EMAIL PROTECTED] On Behalf
> Of Lawrence
> Lustig
> Sent: Wednesday, May 21, 2008 10:36 AM
> To: RBASE-L Mailing List
> Subject: [RBASE-L] - Re: STP_RETURN Variable when running a
> procedure
> 
> <<
> As usual I'm probably missing something very simple here, but
> I've had
> some trouble running procedures one after another due to the
> STP_RETURN
> variable.
> >>
> 
> Duey:  I use stored procedures a lot and do not observe the
> problem
> you're reporting.  I can run two stored procedures, one after
> the other,
> returning different types and I receive correctly typed return
> values
> each time.
> 
> For instance, if I run an integer-returning stored procedure
> my variable
> is set to an integer, and I can see STP_RETURN is an integer
> value.  If
> I immediately run a text-returning stored procedure my
> variable (a
> different one, of course) is set to text, and I can see that
> R:Base has
> recreated STP_RETURN as a text variable.
> 
> This is in 7.6, but I've been using SPs almost since they were
> added to
> R:Base and have never seen this problem.
> --
> Larry
> 
> 
> 


Jim Bentley
American Celiac Society
[EMAIL PROTECTED]
tel: 1-504-737-3293


      


Reply via email to