On Mon, 03 Apr 2017 11:30:38 +0200, Igor Gnatenko wrote:
That wasn't call to change anything,
I know.
I don't want to change anything,
I know.
RPM behavior is not interesting because it is just evaluating
constraints,
nothing else. Interesting thing is depsolving - libsolv.
Maybe, but the level of comments like that is not deep enough. In the end it is
RPM that is given a transaction set to process, and it doesn't know about any
packages you don't tell it about.
Nowhere in this "thread of doom" somebody has explained *why* current
depsolvers
show behaviour that has influenced our packaging guidelines. Reduced/simplified
test-cases aren't worth a penny then. Unfortunately.