cvs commit: ports/devel/hs-alex Makefile

[Doug Barton]

Ion-Mihai Tetcu wrote:
> On Fri, 05 Jun 2009 18:35:11 +0200
> Gabor Kovesdan <gabor at FreeBSD.org> wrote:
> 
>> Doug Barton escribió:
>>> Gabor Pali wrote:
>>>   
>>>> pgj         2009-06-05 11:59:16 UTC
>>>>
>>>>   FreeBSD ports repository
>>>>
>>>>   Modified files:
>>>>     devel/hs-alex        Makefile 
>>>>   Log:
>>>>   - Respect custom PREFIX, LOCALBASE
>>>>   - Bump port revision
>>>>   
>>>>   Reported by:    QAT
>>>>   Approved by:    gabor (mentor)
>>>>   
>>>>   Revision  Changes    Path
>>>>   1.27      +7 -0      ports/devel/hs-alex/Makefile
>>>>
>>>> http://www.FreeBSD.org/cgi/cvsweb.cgi/ports/devel/hs-alex/Makefile.diff?&r1=1.26&r2=1.27&f=h
>>>>
>>>>     
>>> I can't see any reason for a PORTREVISION bump here. Kudos for
>>> fixing this issue, but if we do bumps for every time nothing but
>>> this is fixed our users are going to be doing a lot of needless
>>> recompiles for no benefit.
>>>   
>> As approver, I also thought about this but some people might use a 
>> different PREFIX/LOCALBASE configuration constantly and in this case
>> a rebuild is necessary to get this port installed correctly. I don't
>> think using custom PREFIX/LOCALBASE is widespread but theoretically
>> it is possible that there are such people.
> 
> The bump was not needed here since the port didn't build at all with
> custom LOCALBASE/PREFIX. Since it didn't build it couldn't have been
> installed or packaged, so no need to bump.
> 
> In the general case I think we should bump PORTREVISION when fixing
> something, even for non-default configurations since there's no other
> way for a user to know to upgrade. Some disagree.

That would be me. :) As much as I dislike it personally (because we
don't have a better mechanism) I think that there is general agreement
that rule number one is that if a delta changes the package,
PORTREVISION should be bumped. IMO rule number two should be that if a
delta changes something for a user who already has the thing
installed, and therefore it is worth the time/energy spent to
recompile it, then there should be a bump. Other than that, my vote is no.

In any case, it would be good for there to be clear guidance regarding
the specific issue of LOCALBASE/PREFIX changes since with any kind of
luck we have a lot of them coming up.

Doug

-- 

    This .signature sanitized for your protection