There is a bone of contention latent in this discussion.
Lorb argues that the
value and
functionality of the applications and the installed base does not decline and in fact remains substantial.
I agree. However, that is only part of the picture.
What is important in earning a living or running a business is what someone will actually pay you to make. Today, and going forward. Yesterday just doesn't matter. It's done. It's been delivered. That DOS system that runs a factory was delivered yesterday (or in past years.) There may be funding available to pay a guy or two to maintain and improve it, but there are no more installations. Nothing new. Just bandaging the old.
Just like a Model T.
Obsolescence happens for three principle reasons, as I see it:
1) Functional obsolescence. Example - 300 baud modems. Functional and useful, but not competitive. Simply nowhere as good as modern equivalents for data communication. And in fact, somewhat wasteful of resources (IE, a modem requires an analog phone connection using copper wire.) Another example: iron waste pipe. PVC is the only waste pipe used in plumbing today. Easier to work, to work
with, to cut, and to install. Iron pipe required some gunk called oakum for connections, and was a specialized thing to deal with. Today you use PVC cement in a can with a brush for PVC pipe.
2) Obsolescence due to sheer boredom. Clothing, housing and auto fashions (generally.) New designs must come out in order to induce the customer to buy something.
3) Competitive obsolescence. Technology advancements for the purpose of refreshing existing markets or making new markets. DOS --> Windows, for instance. And in fact, this describes most of the computer and IT industry.
No matter how useful DOS still is in existing applications still in use, all economic activity that is associated with it is stagnant and is at a minimal survival level today.
The one single constant that underlies all of IT
is competitive obsolescence. If you didn't have competitive obsolescence, you would not have had mainframes, then timesharing, then PCs, then whatever else.
I think competitive obsolescence is in a different category entirely than fashion/boredom obsolescence or functional obsolescence. It borrows traits of both, though.
The difference is that the competitive obsolescence is a strong innovator's way of gaining or maintaining market share. And, competitive obsolescence tends to make the old stuff not only obsolete, but perhaps even impractical, costly and/or painful to keep maintain, due to ... obsolescence itself (incompatibility.)
There are many old Model Ts and Model As giving their owner-collectors great pleasure, I'm sure. But they are not a source of economic vitality, except in a tiny way if you run a vintage car shop or you're a collector-merchant.
In like manner there are quite a few businesses and maybe even home users still running DOS and Windows 3.1 and Commodore 64s. In the same way, none command any economic opportunity.
If you want to be where the economic activity is, you have to go there... that now means the newest technology. This has always been the case but in the current tight economy this effect has been exaggerated.
Me - I am still writing C++. How long has it been in wide use, 20 years? So, I'm a hypocrite.
Rather, I blundered into an intersection between what I know already, and a direction that my client wants to proceed in next. I use old techniques to develop new products for a current and future need.
If you
could do this kind of thing with DOS or 16 bit Windows then you could stay on using those technologies...
