... an extreme case, isn't it.

Of course, not knowing the exact details I cannot come up with valid alternative solutions, but I'd be most surprised if I could not do either of
- disabling all interrupts for the time of "measurement" and let the master recover from this
- use a timer which is capable of (hardware) capturing
- use an interrupt (of a high priority) to capture the timer...

My point is, if the address compare logic is there the better, but if it's not there, in the vast majority of cases it does not really matter. On the other hand, I can imagine 10 other features which would be more useful than this - just one for example, a FIFO for the UART - but those tend to be more expensive and the headlines less impressive, isn't it.

Nevertheless, I'd be glad to discuss the topic - and all the other topics we might have been arguing over in the last few years - personally, either in RDU (which I haven't the faintest idea where it might be) or at any other place, provided I'll have the time and funding to get there; but instead of a whiteboard, I'd suggest a chequered table-cloth and two glasses with a content of your choice as the main tools... :-)



Oh, thank you... :-)

JW


PS. Anybody organizing the 1st European 8052.com conference?