I think we hashed this to death in this thread
http://archive.midrange.com/midrange-l/201203/msg01120.html
But in summary, yes, there have been various magazine articles and other
sources that showed you how to create your own RUNSQL, EXESQL, etc years
before IBM came out with one.
Suggestions were given to prepare for the arrival of this new command.
These suggestions included:
- Setting up auditing on your command to find out all the places which
used it.
- Searching all your source for imbedded uses of that command.
- Using a command exit point to monitor for usage of that command.
Basically the process was:
1) Create a new command that was a copy of your existing RUNSQL command.
2) Using one or more of the above techniques find any existing usages of
your custom RUNSQL command and change them to your new command.
3) Use CHGCMD on RUNSQL to only allow it's use in interactive or batch
programs (programs, not command line, etc). This is to wean people over
to your new command if they have command line access to a RUNSQL type of
utility (security considerations a separate thread).
4) Consider deleting your RUNSQL command to prepare for IBM's RUNSQL
command. I can hear you "what if I missed this year end program which
uses my RUNSQL program?". Tough toenails. You probably didn't library
qualify it anyway and it's going to try to run IBM's version which I
suspect has all different parameter names than what you're using. IOW,
it's going to blow anyway.
Rob Berendt
midrange.com
