Execute the Rexx program by EXEC command without compiling

worthy2005 · Posted: Mon Oct 31, 2005 8:15 pm

Hi everybody,

We can execute the Rexx program by EXEC command and without a compilation. So this could make everybody see the my source program. Everybody could know the logic and revise it without a authority.

So What can I do to prevent this from happening?

priyesh.agrawal · Posted: Mon Oct 31, 2005 9:53 pm

Put you source code in a password protected dataset.

Regards,

Priyesh.

MGIndaco · Posted: Mon Oct 31, 2005 11:14 pm

Why you cannot compile your source?
To run your program with PGM=MYREXX on a OS/390 or z/OS you must compile it with STUB=MVS, and for what I know is the only way....(but as usual I can be wrong.)
Tell us your problem with compiler... We are here to help you if you need...

Ramya A · Active User Joined: 26 Jul 2004 Posts: 104

How could I compile my REXX code? Do I need any utility to be installed in my mainframes in order to get my REXX source compiled?

superk · Posted: Tue Nov 01, 2005 6:25 am

MGIndaco · Posted: Tue Nov 01, 2005 2:47 pm

Are you working on OS/390 or in z/OS or other?
usually you will find in your sys1.proclib a proc named REXXCL
( or in SYS1.SFANPRC)
This is the proc of compiler that will execute the REXXCOMP.
Tell me if you can find this.

Ramya A · Active User Joined: 26 Jul 2004 Posts: 104

Hi MGIndaco

Yes, I could find this compiler REXXCL in my system. I also tried compiling one of my rexx code and it is working!!! This is of great help!! Thank you so much.

Rgds,
Ramya

smileseenu · New User Joined: 27 Sep 2006 Posts: 13 Location: India

i coud't find REXXCL in SYS1.PROCLIB, is there any way to complie the REXX PGM... otherthan this...??

expat · Posted: Thu Nov 12, 2009 3:49 pm

Why respond to a topic which has been dormant for four years. START A NEW ONE.

But, as I am in a good mood, the REXX compiler is a pay for option, so if you can't find it, the chances are that your company never paid for it.

Very few companies that I have worked at haver purchased this option.

vasanthz · Posted: Sat Nov 14, 2009 2:36 am

dick scherrer · Posted: Sat Nov 14, 2009 3:18 am

FWIW/IMHO - if compiled rexx is needed to "survive" the resources required to run certain rexx code, rexx is being misused. . .

While rexx can do many powerful things, it is not meant to be used in high-volume situations. . .

Again, only MHO. . .

d

MBabu · Posted: Sat Nov 14, 2009 4:17 am

The original question (4 yrs ago) was how to implement "security by ignorance" (if I think they can't change it, then they must not be able to change it).

But Dick is right. Rexx is for scripting, not volume processing. Although, IMHO, for infrequent tasks, that distinction is blurred.

dick scherrer · Posted: Sat Nov 14, 2009 6:23 am

PeterHolland · Posted: Sat Nov 14, 2009 8:10 pm

dick scherrer · Posted: Sat Nov 14, 2009 10:15 pm

Hello,

MBabu · Posted: Sun Nov 15, 2009 12:42 am

even a million is OK but it might take a minute or two where SORT might take a few seconds. Depends on the complexity of the task. Since most tasks asked for on this board seem to be moving a field or counting records, rexx is fine for files in with hundreds of thousands of lines, but if the task is more complex such as multi-file compare, sorting, etc, then Rexx may not be the best choice.

Strangely, the use of stems is very inefficient. I never figured out why that is. Rexx on other platforms as well as other languages that implement associative arrays like Java don't have that problem. TSO Rexx also appears to keep and reparse comments at runtime. A program with many comments will run slower than one without them. In these cases, compiling does make a difference. ISPF goes so far as to ship a copy of ISPDTLC (>24000 lines of rexx) which is devoid of comments and leading spaces. That does not sound like a very robust implementation of Rexx to me. In other words, it isn't that Rexx as a language is slow, but the TSO implementation of it is slow.