No, it's not. Please be specific. You have 40 I/O's in what sort of package?


Thta's not what I said, but you're welcome to correct me if I did. What I said was that the timing of the cycle-stretch doesn't affect anything but the stretched cycle, and, in another instance, I said that the 82C55 doesn't do anything that would otherwise impact the behavior of your MCU.


You've done nothing of the sort. You've done lots of bobbing and weaving to avoid the real issue, and, ultimately, you've admitted that the interrupt latency in the example you chose makes no difference at all. In fact, you seem to have claimed that was your point all along.


You're not wrong in having your preferences as to device types and as to packaging. You're not wrong in your choices of how to design and debug your code. Perhaps you are wrong in attempting to impose those preferences on others, though.

However, you're totally wrong in telling someone else they're wrong in making choices that they've made until you know why they've made those choices and what they can reasonbly be expected to do about it. If they're writing code for a device that's already in place, just adding a function, you can't expect them to change the hardware. Just let 'em do the homework they've been assigned.

RE