The codes themselves will stay the same, but they have no real meaning
without context which the description gives. Codes like ERROR_OPEN,
ERROR_CLOSE, ERROR_READ, etc.. have no meaning without knowing the file
which the log gives. ERROR_DEVICE, ERROR_DEVICE_READ, ERROR_DEVICE_WRITE
also have no meaning if you don't know the device. ERROR_NOT_FOUND doesn't
mean anything unless knowing what wasn't found. The error codes are the
same for all products (only very old may have something different) as well,
they are not product specific (except those above 199) and may be a
condition reported deep in to the common libraries.
You can use the sample batch file 1 which has a list of common errors via
http://www.terabyteunlimited.com/downloads-image-for-windows.htm and can
test other conditions (pass bad parameters, etc..) you wan to recovery from
and look in the log to find what it was. Treat unknown (non-zero) as an
error that can't be recovered.
"tas3086" wrote in message news:
[email protected]...
All your programs seem to pass a return code that matches the error code
reported in the log. Automated recovery actions can and are being made
based on this return code. If, as you say. that the return codes are just
generic numbers, and have no relationship to the problem that it relates to,
then I would suggest that you design your programs to pass a return code to
the calling program that has some meaning that is related to the text in the
log. I hope that you did not mean what you said: "knowing the generic
numbers won't help much, you can get the same generic code but not know what
it relates to."
And These codes need to be published.