This may not be a problem in most applications, but back when I did some development with the original P89LPC932, we had massive headaches with the reset power requirements. In essence, when the power was removed, the capacitance didn't allow VPOR to be met unless it sat for several seconds.

The result was very unreliable resets (most would reset fine..some would result in a lockup). It wasn't a problem that most would notice during development, however. It was only obvious once we sat there power cycling on the final PCB 100's of times.

At one point, we had one of the Philips application engineers come and take a look. In the end, Philips should have chosen a higher VPOR threshold and this was admitted by the engineer. Unfortunatly, I think even the A1 revision has this limitation.

I believe most of the "re-programming" issues have been improved with ICP (versus ISP). Although you may want to look through flashmagic's forums.

Anyway, that has been my personal experience. It was a costly development (time lost), but once you are aware of all details, it's a great microcontroller.

Tomas