This is perfectly sane behaviour IMO, %exclude only having the purpose of excluding file from package and not from unpackaged files check is more explicit. Removing files in %install is sufficient and should help avoid ambiguity and possible packaging mistakes by using %exclude.
For rpm in Mandriva Linux, matching behaviour was recently reintroduced in rpm-4.6.1-1.
Why should a directive that specifies that a file should not be included
_FAIL_ when the file is indeed not included? So that the directive can be removed?
(continued)
FULLSTOP on any deviation from what is specified
BESTEFFORT continue in spite of deviations
In the case of %exclude, I think BESTEFFORT is the better choice. Clearly
someone specified that a file should not be included. And if the file isn't
present at all, that does conform to "expectations" imho.
This is perfectly sane behaviour IMO, %exclude only having the purpose of excluding file from package and not from unpackaged files check is more explicit. Removing files in %install is sufficient and should help avoid ambiguity and possible packaging mistakes by using %exclude.
For rpm in Mandriva Linux, matching behaviour was recently reintroduced in rpm-4.6.1-1.