Another reason to look for ways to continue products – long term support gets paid for directly that way.
That’s another issue with hardware changes along the way too, you want to not force big software versions changes between those separate hardware changes along the way. Small changes can be dealt with if you have to, by detection with the two versions in the code. All the sorts of things you need to consider in continuation engineering.
Costs are part of it too, as thermionic correctly points out, fortunately the costs of continuation are normally so much less than when initially starting from scratch, that it can very often be worth doing, if you can extend the life of a product substantially.
=-=-=-=-=-=-=
Take-overs by larger conglomerates, means new rules for decision making, like agreements when you need to pass decisions up the chain for review, and when you just make your own choice locally. I’d think dropping something major would normally get reviewed. If the decision came from on high to drop something, that also would get local review and feedback, but ultimately, you could guess who normally makes the final call, in that case.
Same goes to decisions within a group. Often there is some very open discussion, and open differences of opinion.
I’ve only had to break up a fist fight, a few times. You try to keep ego out of decisions, but your work becomes your life sometimes. People also may go elsewhere after a project or product gets canceled.
Talented people leaving often has the affect of changing a companiy’s direction as well. Bob Coover, mostly a software developer left Sequential and started Groove Synthesis. Could the 3rd Wave ( thread ) have been a Sequential product ? Who knows.
Funny aside :
Spell check turns thermionic into thermonuclear !