RE: Alternate Platform for RPG

["Joe Pluta" <joepluta@xxxxxxxxxxxxxxxxx>]

> From: Bartell, Aaron L. (TC)
>
> The first
> call is going to be the only one that is slow and after that it
> performs as
> well as RPG.  I have not done any sophisticated timing on this, so it is
> based on human perception and not actual programmatical time-testing.

Aaron, I'm not trying to be argumentative, but I suggest that you actually
do the timings before you state that Java is as fast as RPG.  The last time
I did timings, Java was slower than RPG at pretty much everything, sometimes
considerably slower, especially in the area of transactional I/O.

I posted results to these lists based on real timings, using simple programs
like this:

FJ4PP1   O   E                    DISK
C           *ENTRY    PLIST
C                     PARM           COUNT  100
C                     PARM           TIME1   60
C                     PARM           TIME2   60
C*
C                     TIME           TIME1
C*
C                     DO   COUNT
C                     WRITEJ4PP1R
C                     ENDDO
C*
C                     TIME           TIME2

These programs need to be updated.  They were done in August of last year.
But they showed a considerable performance advantage to RPG native I/O to
both SQL and Java.  They should be run on other versions of the operating
system, and other versions of the JVM, and perhaps other OS configurations
for that matter.

Anybody interested in a joint project to extend this framework and run
additional tests, let me know.  This could be an interesting and informative
venture.

Joe