07-31-2012 02:51 PM
At the moment I think about what is the best way to set a fail
Who can do best practice sharing with me?
I would use the following sample for better understanding.
http://ftp.ni.com/pub/devzone/epd/cvimeas.zip
It is clear for me – how to get some values from a module (called DUT = device under test) and set this value as numeric limit or string test.
But for me is not clear - what is the best way to set fail.
For example if a value cannot read from DUT
In general the error flag can be used like this
testError->errorFlag = TRUE;
testError->errorCode = error;
teststData->replaceStringFuncPtr(&testError->errorMessage, errMsg);
but make it sense to set the test to error on all fail – for example if the temperature cannot be read from DUT should we set fail or error?
On PASS/FAIL Test it is easy to set ERROR on troubles like “RS232 cannot be opened and FAIL if reading from DUT is not successful.
But on numeric limit tests or string tests it is not possible to set fail as result.
What is the best way?
Who can share and describe his solution and the advantages and disadvantages?
Thanks
TODA1980
.
08-02-2012 12:51 PM
Who can help?
Today I discuss this topic very long with my colleque - first we think we use error - later we think this is no good idea.
We are very unsure how to handle it.
For better understanding a very easy example --> programmed as numeric limit test
+++++
open RS232 --> if RS232 can not open set error --> this is clear --> generall fault of tester
send a command "get temperature"
waiting for a reply like "tempeature=xxxx" --> if message "temperatur=" is not received, what should be done??? (error or set a stupid value like -9999)
split response and give xxx back as num --> xxx will be given back to teststand and numeric limit will calculate pass/fail
close RS232
+++++
I hope somebody can help