Re: "Encouraging" customers to upgrade was: I need Feature Card 2617 for old Model 500 -- MIDRANGE-L

Good points from Darrell, with comments inline...

Darrell A Martin wrote:...The customer is not always right, but the customer is always 
the customer - as long as the vendor cashes the check. If a software vendor can make 
money supporting customers who are on downlevel hardware, then that is what should be 
done; otherwise not. This assumes that "make money" takes into account:
        - opportunity costs for resources used
        - customer good will for future purchases

- the ethics of putting a customer at *actual* risk if a hardwarefailure occurs

I think a vendor has a certain degree of "duty" both to the customer andto himself, at least over the long run, also to make clear to thecustomer the costs of refusing to upgrade. Backup and recovery are onecost, support for OS "happenings" another. All hardware suffers entropylike anything else in the universe, even the systems that best resist itlike our IBM midrange systems.

It may be ironic, but the legendary reliability of the i/400 OS/hardwarecombination actually suggests that the *actual* risk of not upgrading issignificantly lower than for other platforms.
Note that there is normally extremely little *actual* risk, often none atall, that arises from "missing out on new features". This may be heresy insome parts of our industry, but that doesn't mean it isn't true. Ofcourse, it doesn't mean there isn't value to new features, either. Andwith that we return to the customer, whose prerogative it is to decidewhether the value received exceeds the value paid.

That's absolutely true. But if we would make a better case to ourselves(meaning midrange programmers in general) about "new features", maybe itwould "spill over" to the rest of the decision-makers.

Like you could even make a case that you don't need any stinkin'"high-level programming language", you could always "make do" withAssembler. Of course I like the idea of "getting close to the machine"and there are uses for that, like writing compilers themselves.

I think a good argument for keeping up with the releases is to "keepfrom falling behind" in technology. By this I don't mean for its ownsake, but inevitably a competitor will find that he can open up newfunctionality, cut operational costs in half, recuperate lost equipment,or even open up new markets where there were none, and the "dinosaur" isleft behind choking on the dust.


--Alan