cvs commit: ports/games/8kingdoms Makefile ports/misc/airoflash Makefile ports/graphics/autopano-sift Makefile ports/x11/avant-window-navigator-xfce4 Makefile ports/lang/boo Makefile ports/x11/cl-clx-sbcl Makefile ports/palm/coldsync ...

[Chris Rees]

On 10 April 2012 14:56, Pietro Cerutti <gahr at freebsd.org> wrote:
> On 2012-Apr-10, 09:04, Jerry wrote:
>> On Tue, 10 Apr 2012 07:40:09 -0500
>> Mark Linimon articulated:
>>
>> > Finally, I agree that FreeBSD can't "guarantee" any given ports will
>> > work, but I think we owe the users the effort to make sure if a port
>> > is included, it's at least not completely useless.
>>
>> To "guarantee" that the port will work is certainly beyond the scope
>> of the ports system; however, to guarantee that it is fetch-able
>> and build-able is an implied action by the simple fact that it is
>> included in the ports structure. If, after a reasonable amount of time,
>> a solution for a port's inability to properly build or if the port is
>> just plain not able to be fetched, then it should be removed from the
>> port system. There is no upside to keeping ports that will not build,
>> or cannot be fetched.
>
> For the record, the port in question was fetchable / buildable /
> runnable. For some reason, some python class sometimes doesn't get
> byte-compiled, resulting in packaging (PLIST) errors.
>

BUILD_ENV=      PYTHONDONTWRITEBYTECODE=yes

usually helps.

Chris