El 14/5/25 a las 12:50, Adrian Bunk escribió:
What is a maintainer supposed to do when the package already does "dh --no-parallel" and the upstream Makefiles are basically unfixable?
Whether a Makefile is unfixable or not is subjective. Everything depends on the amount of time that one is willing to spend. It may be the case that a Makefile looks unfixable to you, and at the same time it may be a perfectly legitimate bug to be reported upstream. So, at the very minimum, I think maintainers are supposed to forward those bugs upstream. [...]
How many of the packages that break with "make --shuffle" are currently doing parallel building? I am asking since these might be RC bugs for trixie.
I believed (regarding build failures) that you only cared about RC-ness when the failure happens in the buildds. Have you changed your mind? Thanks.
