Steve Bernard writes:
Steve> The first problem I seem to have fixed by editing on of the Steve> test files (008).
I agree that this is a bug in the text case. I have submitted a patch to fix this to Frank and I guess it will soon show up in the CVS.
Steve> The other problems I'm not too sure on, so I've included all my Steve> debug info in the following attachments. Each file is labeled Steve> to describe what process created it. I'm using an updated Steve> version of VA Linux's RH 7.0.x distro on a K6-2 333MHz. Please Steve> excuse the file naming convention, I couldn't help myself ;)
Steve> The debug info was created AFTER making the changes in Steve> libsmi-0.2.7-changes. In cases where there isn't an .err file Steve> there weren't any errors. The files are in the order that the Steve> operations occured.
I am not able to reproduce this problem. Basically, what the test does is to dump MIB modules in the SMIv2 format and to compare the output with a "blessed" version. This is basically a regression test to see whether we screwed anything up while editing the source code. The output generated by the test driver contains a diff command which you can just past into your shell to see the differences (assuming you are in the test sub-directory:
*** smidump -f smiv2 output differs, for details take a look at *** diff dumps/SNMPv2-MIB.smiv2 smidump-orig-smiv2.out/SNMPv2-MIB.smiv2
It would be helpful to send us the diff so that we can see what the differences are on your box.
/js