to go about their rotten business. It is issued without acknowledgment
of any obligation, in response to ports/64393.
The patch does NOT bump PORTREVISION as the change is invisible for the
port's or package's users. $FreeBSD$ is sufficient.
I also refute any "bug" with respect to the the porter's handbook's
pages referenced in ports/64393, my port has not written anything after
bsd.port.mk and portlint has nothing to complain about (only that it
can't figure "patch" is legal in a PATCHFILES variable).
Informational: tools that assume .include <bsd.port.[post.]mk> was the
last line in a port's makefile are broken. They should be comparing
the output of realpath $(make -V MASTERDIR) against the output of
realpath $(pwd) instead:
-bash-2.05b$ realpath $(pwd)
/usr/home/ma/db42
-bash-2.05b$ realpath $(make -V MASTERDIR)
/usr/home/ma/db42
-bash-2.05b$ cd ../db42-nocrypto # switch to slave's port directory
-bash-2.05b$ realpath $(make -V MASTERDIR)
/usr/home/ma/db42
-bash-2.05b$ realpath $(pwd)
/usr/home/ma/db42-nocrypto
PR: 64479
Submitted by: Matthias Andree <matthias.andree@gmx.de> (maintainer)
226faa7c29