[Help-autogen] running test cases

[Maggie Hamill]

Hi,

I am using autogen 4.5.14 as a case study for research on software
reliability. Since all the test cases that come with version 4.5.14
pass we would like to run the test cases from version 5.6.2 on version
4.5.14. Assuming some functionality has been added, and bugs have been
fixed that may not have been tested in the past this should increase
the number of failed test cases. We do this so we study the behavior
of failures against our theories.

However, when we try to run the newer test cases on the older versions
we are getting errors. We noticed the directory structure is different
for the newer versions, and think this might be contributing to our
problem. However, if we simply copy new test files in the the tests
folder of the older version and edit the make file to contain the
names of these files, we get like 31 out of 32 failed cases, and then
we get errors in [check-TESTS] error 1, [check am] error 2, and [check
recursive] error 1.

I was wondering if someone could tell me what would need to be changed
to run the new testcases on the older version of the software. Maybe
this would be easier if the two versions we used had a similiar
directory structure(ie. 5.1.5 and 5.6.2). However, when we try to make
version 5.1.5 we get errors about cross referencing non existent
nodes. Any ideas? I would really appreciate any help.

Thanks,
Maggie*