Lpc2000

> Equally true, but the problem is the same on brownout or even 
possibly
> (more remote) on cold start.

Point taken. Though I understand how the Watchdog generally works, 
I've heard about strange problems (like the ones u mentioned) that 
can happen if used. So, I won't fully trust Watchdogs yet. 

> My app ignores P0.14, because I have a stamp-like RF module , and 
the user
> could be doing all sorts of things on this I/O pin.
> Making this safe inherently implies extra and irritating HW cost 
anyway.

That's the give-and-take nature of the design game. Unfortunately, 
I'm running an LED over P0.14. So, there's no way for me to get 
around that unless I include a jumper & do power cycles.

Right now I'm weighing out IAP/ISP pros & cons. My biggest challenge 
since Summer was to update firmware using IAP calls. I believe this 
would give the application code far better code control & security. 
But right now it just looks quite complex to implement.
 
The thing I like about the previously posted code is that both P0.14 
& reset issues were nullified, making ISP look surprisingly much 
easier to do. But knowing ISP communication is ASCII based, code 
security can be compromised, as Owen previously mentioned.

Leighton