| ??? 06/15/04 15:24 Read: times |
#72538 - RE: Error logging for purpose of Maintenance Responding to: ???'s previous message |
The one I 'argue against' is that I have often seen error logging as an excuse not to THOROUGHLY design because "the errors will be caught by the log when we test". You are absolutely right, I can’t show a customer the error log to explain why his/her system has crushed :) A little trick I often use is to have a routine called 'crash' that is implemented whereever there is a possibility of something wrong from outside factors such as a reading from and ADC going crazy because of an analog input being overloaded. This routine may be called from many places, it simply record the return address. This is very good idea, actually my current design use a different routine/function for handling each error situation. I will compare your idea with my current implementation to see which one is better for my system. But it sounds very nice. I like it. Thank you very much indeed. By the way, the above remind me, I have seen ever so often designs where the ADC was used full scale i. e. max normal input = max value. That is a BIG mistake. In such a design there is no way to detect an error causing the analog input going wild because of an open or short. You are absolutely right. I never use an ADC in full scale, too. thank you very much indeed |
