actually will do it.

What I see, all over the place, and not just in this forum, is that people are all too willing to stake the quality of their products on unverfiable claims from chip manufacturers. Whta puzzles me about this particular problem, or set of problems, is that, whereas chip makers generally use something such as the "... well, you can get around that by using this (supervisor, in this case) but we'll fix the problem in the next revision " response to a problem like this one-or-more, in this case they did nothing of the sort. Dallas went ahead and built the "supervisor/watchdog" functions into their DS89C4x0 series, and have never, AFAIK, claimed that their parts needed a supervisor, while nearly every other 805x-maker has used the availability of supervisor IC's as an excuse not to test their current products in this respect, and further, as a justification for ignoring the problem completely.

If they'd run some exhaustive tests, and, more importantly, published the results, one could, at least say, "OK ... they've done their due diligence," and made choices on the basis of their published result. Instead, we've got essentially nothing.

Back in the days of ROM/EPROM-based parts, this reset matter didn't exist, since there really wasn't a reset problem. That's how the 805x got to be so popular. Integrating a CPU, FLASH, and RAM on a common die is not an easy trick, though the chip makers do manage to get it done to some extent. Most people fail to appreciate the difficulty of designing solid clock and system reset logic, though, and that's not easy. That's also a contributing factor to these mfg's unwillingness or inability to tackle this problem.

I guess it's easier to sell higher speed and lots of "bells and whistles" than it is to sell parts that function properly. Not one of the chip makers advertises that their parts only work correctly with an external supervisor. Why would that be? Is it just because it's too much work, or could it be that they're unwilling to admit they've all, collectively, been pushing defective hardware onto the market, and getting away with it for all-too-long?

RE