Why is that? If you have to build ONE of something, and you already know that ONE can be made to work with the HMOS 805x and an 8255, code already exists and is on hand, and you want it completed by tomorrow, noon, wouldn't it make sense to use what's already on hand and been shown to work?

I have often, when only one was to be made, been caught with a request to "make one more" or "this is good, let's make 10" and where am I going to get the antique from when that happens.

The "only one will ever be made" is as much BS as "this thing will never need a software glitch fixed"

there's always more to replacing an 8255 than just a decoder and a single register.
True, if you have already designed it in, totally false if you do not use it from the start. The point is do not start with it.

many of these inquiries are from people who have designs they're given by someone else, and they want to build THAT.
"given by someone else" very often means from an old (in electronic terms) book. There is no reason to follow designs in books that were written in 1992.

If they had a boss who was willing to spring for a PCB and the latest technology, do you think they'd be here asking the questions they ask?
YES! some get the base of their design from some old book.

Erik

PS I do not necessarrily advocate "the latest technology", but I do advocate current technology.