I'd guess that the vast majority of "bug reports" result from the developer's lack of sufficient equipment to put together a convincing case for hardware bug. Most of the "bugs" I've detected in the past decade have simply been the result of inadequate documentation from the manufacturer.

I inquired, for example, about the timings of the UART in the Maxim/Dallas DS89C4x0 series, and, after two and a half years, have not gotten any information back. They couldn't provide any useable information about the flag timing, interrupt timing, etc. Their blanket statement that it works like the standard UART falls apart in the fact that the chip is a one-clocker. It's not a serious problem, at least for me, since I seldom use the UARTs, but nevertheless, it's a problem for someone who wants to know what he's doing. The UART timings certainly can produce "buggy operation" if one doesn't have a clue about these features.

Nevertheless, if you have "pictures" taken with a logic analyzer and oscilloscope, that show an occurrence clearly in violation of published data, you'll probably get some action, though not necessarily what you want, from the manufacturer.

You can produce good, functioning circuitry and firmware with very little test equipment, but there's no guarantee that you will. You certainly can't verify it. Since many of the people posting on 8052.COM lack logic analyzer, oscillsocope, etc, and some even lack a multimeter, it's hard to imagine that they can provide convincing evidence of a bug. Further, if their code is written in HLL, it's far too easy to blame the "problem" on the compiler or libraries.

RE