Well, I'm glad you can "whip up" a schematic in minutes ... but that isn't rocket science. We expect high-school students to do that all the time.

What you don't seem to grasp is that a plane figure that is round is not square. If the 805x architecture contained an I/O space, you could map into it. It does not, so you can't, and therefore you can't legitimately refer to anything in the 805x architecture as being I/O mapped. You can, of course, call it port-mapped, since the architecture does provide ports, but it does not provide I/O space. If you don't have it, you can't map into it. Since you can map functions into port bits, calling it port-mapped would work, both logically and linguistically.

Also, please don't misquote me or take my remarks out of context for your own convenience. We're not running for elected office, so we don't need to do that.

What I said was,


I did not say that any of this is I/O mapped, nor should you, since you must agree that the 805x has no I/O space into which to map.

I don't know what it is that makes you insist that a newer MCU is "better" in any way, than an old, "steam-driven" MCU, so long as it is capable of meeting the requirements of an assigned task. Nowhere does it say that it's better to perform a task at 500 times the required speed rather than at only 5 times the required speed. Nowhere does it say that it's better to use a part that YOU like, if some other part, or combination of parts will also do the job. You may like it better, but then, you have tons of money to waste on making new PCB's for at your boss'/clients' expense, so you can use the latest/greatest technology. I can't see how a fellow with limited resources who's building ONE of something can be expected to take your chosen path. I find it difficult to solder to 20-mil pin-pitch parts in order to use the latest parts. The guy who bought the eval kit that his instructor told him to use might find it difficult to justify buying a new $15 MCU just because Erik likes it. He might find it costly to have a new board made. He might find it disappointing to learn that his new latest-greatest won't do what the old, steam-driven part might have done, e.g allow him to use an external data memory large enough to buffer his data.

Anyone who's read your posts knows you'd never, Never, NEVER use an 8255 for anything, no matter how sensible it might seem, despite the fact that the specified MCU easily allows for its use. Anyone who's read your posts knows you'd never use any component that's been on the market for more than half a day. However, you've completely overlooked the basic purpose of microcontrollers, namely to replace a lot of logic where it would have otherwise been required. Today, there are lots of applications into which one places a general-purpose board with an MCU and some peripheral hardware instead of a board of logic, in order to build one or two of something. That doesn't warrant a new PCB, and a new part, in high-density packaging, with enough I/O's to support your arguments, just won't fit into such an application.

RE