The microcontroller contains several dozens of different units, like, much of them are relatively hard to use and many questions can be asked about them. They all are documented. I bet many users generate questions on these. Why is then ONE of the units - the serial programming circuit - undocumented? How is it different from the others?

This is NOT the same case as with EPROMs, where underprogramming could result in unreliability. FLASH programming circuits are built into the chips, they self-generate the HV, self-time the process, you can't really screw it up even if you would want it.

Whence even the decision to keep the parallel programming specs is idiotic these days, but let them be - I see that as an incentive to keep up the programmer manufacturers, which is not that bad; besides, knowing that protocol won't give the end user that much of an extra freedom.

But, I repeat, much of the point of serial programming (whatever TLA is used to call it) is negated by not publishing it. The user loses the freedom to implement applications where serial programming is a bonus; and the manufacturer won't gain anything. And that's idiotic.

JW