Properly designed user-created objects of all types which were created on
earlier versions of the OS [and which have NOT had observability or
creation information removed from them] can usually be restored and
on any newer version of the OS, and in almost all cases, there is no need
recompile the source.
An awful lot of restrictions. Some very serious restrictions.
Restrictions that a lot of companies have spent a lot of hours trying to
get around. Restrictions that a lot of companies who were penny wise and
pound foolish and stopped paying software maintenance to their application
vendor and then had to back up the truck full of money to get versions to
work on newer versions of OS would definitely consider not insignificant.
But accepting these restrictions...
What about CL with commands like EDTDOC buried within them?
What about runtimes affected for saves when IBM changed the default on
access paths? (I think IBM made the right decision to change the
What about CL with commands like homegrown versions of RUNSQL?
What about fools that still use spool file output from IBM supplied
commands and then use CPYSPLF against them, who have found out IBM changed
the spool file layout? (Ok, maybe that doesn't qualify as "Properly
designed user-created objects".)
If you want to throw in a lot of restrictions and what not to get these
rules to work then I have some .bat files, and perhaps even a basic
program or two, that would probably still work way back from DOS to
Windows 7. Let me guess, they don't count because they are interpretive