Re: SQL RAND() help. (my solution) Probably not a solution

[CRPence <CRPbottle@xxxxxxxxx>]

  You will almost surely not want to reseed every time in that manner. 
 That could just as well be replaced by microseconds to get a /random/ 
value.  Investigate a solution that has the original non-seeded RAND() 
giving an expected pseudo-random number versus the same value every 
time.  Release was not mentioned...  Note that on V6R1 a PTF was 
required [maybe a PTF is required to the GA code; for sure pre-GA] which 
corrects a problem [at least somewhat] similar to the problem described 
here.

Regards, Chuck

Michael_Schutte@xxxxxxxxxxxx wrote:
> I am able to solve my issue by inserting an integer value in the Rand
> function.
>
> I converted the current time into a  10I 0  field, then executed.
>
> CurrentTime = %Dec(%Time():*ISO);
> Exec SQL Select Dec(Rand(:CurrentTime)*10000000000,10,0)
>          Into :RandomNumber
>          From SYSIBM/SYSDUMMY1;
>
> But I question why....????
>
> Michael Schutte/Bob Evans wrote on 05/05/2008 10:53:34 AM:
>
> > I know that there's been a big discussion about this before... but I
> > have an issue that I couldn't find in the archives.
> >
> > If I run this statement interactively (in STRSQL).  I get a
> > different value everytime...
> > select dec((rand()* 10000000000),10,0) from sysibm/sysdummy1
>
> > But when I put it in the SQLRPGLE program (which so happens to be a
> > DEV2 web program), I get the same value.
> > select dec((rand()* 10000000000),10,0) into :RandomNumber from
> > sysibm/sysdummy1
> >
> > RandomNumber always equals 5138706625
> >
> > If I put a number in the Rand Function,  Rand(somenumber).  I get a
> > different value than above but it doesn't change on repeated calls.
> >
> > The activation group is *NEW and the Rand function is non-
> > deterministic. So I don't follow why I'm getting the same value.
> >
> > I've even tried doing  Values(...) into :RandomNumber;