This hub-side change would also fix #4469. It seems reasonable to be forgiving in non-strict mode.
Metadata Update from @mikem: - Pull-request tagged with: discussion
Also probably worth noting here that there are still some other exceptions raised in getRPM that are unaffected by strict, and I think they should probably remain so.
The Invalid type for rpminfo error indicates some nonsense arg. I think callers should at least be passing the correct type of data here. Otoh, I suppose a similar argument applies for invalid NVRAs. I could be swayed here.
Invalid type for rpminfo
The Multiple nondraft rpm matches error falls in the "should not happen" category and indicates a problem. This should not be ignored.
Multiple nondraft rpm matches
This hub-side change would also fix #4469. It seems reasonable to be forgiving in non-strict mode.