RE: RPG IV release levels and complexity

[Scott Klement <klemscot@xxxxxxxxxxxx>]

On Tue, 18 Nov 2003, Bob cozzi wrote:
> In a one or two-person shop this is a perfect scenario.
> But when other things are in play, such as the issue with people who didn't
> need that cool new feature of DB2 in V5R1 so they choose not to upgrade and
> now it would be cheaper for them to buy a new machine than upgrade the OS.
> So if I am a software vendor, my market share is reduced because I have to
> decide to not user V5R1 and later features and thus have the largest
> potential prospect base, or use those features and reduce my prospect base
> by 70 percent. Today, that's probably only about a 50 percent reduction vs
> when V5r1 first came out, but it is still huge.

These are good points.

Clearly, you think that putting the V5R2 compiler on a V4R5 system will
solve the problem? (In this example, anyway)

But then, what happens when you need a new component of OS/400, instead of
an RPG feature?   Say, for example, you need the GSKit support in order to
do SSL, or you need the ability to digitally sign objects?

I guess the thing that frustrates me is that IBM does a very good job of
*discouraging* people from upgrading their OS by dropping support for
things, making it very expensive, making the install process very
difficult, breaking hardware compatibility, etc.   Doesn't that seem
backwards?  Shouldn't IBM be trying to encourage people to upgrade?

But then again, IBM runs into a similar situation as you do.  If they have
to continue to suppor everything forever, it hampers their ability to
improve things and to move forward.  This means that, like you, they have
to cater to the lowest common denominator.  They've decided not to take
that approach, but rather to drop support for things so that they can move
forward.

Sigh... this is all giving me a headache :)