Hm... what's the central issue here?

• If it's to check whether the package will install, current code actually does this automatically (in mock) and reports installation errors. This should cover the "Install correctly" part.

• It it's to have the output from yum deplist for other reasons, one could certainly run it and generate output. However, I hesitate to include in the main report for sheer size considerations.

Note that current code does make a rpm -q --requires and --provides, inserting output in report.

Given all this, what's the conclusion? Just close it as 'fixed', create a separate report from 'yum deplist', making the source easily accessible from the review directory for users making such things on their own. Other ideas?

I choose to interpret this bug as a need for the reviewer to determine whether a package will install or not, and that yum deplist is useful in this context.

If so, current code instead makes a fully automatic install in mock, reporting possible errors. The initial commit is 4f0a92a but this code has been revised - use current HEAD when testing.

Thanks for input! Please feel free to reopen bug if I havn't understood it correctly. If you do so, please state if it's about the install check functionality, or the need for yum builddep for other reasons.