question on survey question

["Bob Cozzi \(RPGIV\)" <cozzi@xxxxxxxxx>]

In next week's Midrange Developer, my 10-second survey is going to be
something like this:

Should IBM be required to base all future versions of the compilers on
V5R1, thus allowing future code to be compiled as TGTRLS(*V5R1M0)?


My question to you all is this:
Is this a viable question, and is it clear enough?

My thinking is that this will change the way IBM does things, so that if
they add a built-in function in V5R2, you and I who are still on V5R1
can use that built-in function for our V5R1 applications.

Today, if you want to compile to *PRV you have to have actually
installed the V4R5 compiler on your V5R1 machine. I'd like to see IBM
move away from that model, and instead, create enhancements to the
existing compiler but require that enhanced compiler to run on the base
Version release. Meaning for the entire life of V5 the V5R1 compiler
would both compile and be able to target V5R1. Even if there were a V5R7
the identical "binaries" would be installed on V5R7 and V5R1 and
everything in between.

So basically this means that the compile would allow you to produce code
that runs on V5R7 or V5R1 and the only time a source member would not
compile is if there were an operating system-dependant feature used in
your code. Such as a new data type that may be in the database and in
RPG but not back-release compatible. Otherwise, if they add a new opcode
or a new built-in function it would work on V5R1 and later.

Comments


Bob