Ignoring make.conf when looking up for existing flavors, makes the
test less likely to be polluted by end user defaults, like default
version of python etc.
There is no need to document when a port has been flavored or new
flavor has been added, check if someone added such a line, and prevent
the person it is not needed
This is because half of the time, the previous line is the one that is
wrong, so you get redirected to line X, where X-1 has the error, and you
wonder because line X looks just fine.
Sponsored by: Absolight
- Enable FLAVORS.
- Make make describe flavors aware.
- Add a qa check for unique package names amongst flavors.
- Make MOVEDlint understand flavors.
- Add a bit of sanity check to make sure FLAVORS stay lowercase.
- Various fixes.
Reviewed by: portmgr
Sponsored by: Absolight
Differential Revision: https://reviews.freebsd.org/D12577
ak@ asked that I remove the unused error array, and I went to have a
look at what it was actually used for back in the day. It seemed better
to re-enable the blame feature instead of removing it.
Sponsored by: Absolight
This now warns about these 2 ports as needing to be marked as resurrected:
archivers/brotli|archivers/py-brotli|2016-11-24|Brotli is a python module
devel/libbrotli|archivers/brotli|2017-07-12|meta project no longer required to build libs
www/rubygem-jquery-rails||2017-01-07|Has expired: Depends on deprecated www/rubygem-railties
www/rubygem-jquery-rails4|www/rubygem-jquery-rails|2017-03-27|Remove PKGNAMESUFFIX
