svn commit: r358722 - head/Mk

[Eitan Adler]

On 23 June 2014 05:45, Baptiste Daroussin <bapt at freebsd.org> wrote:
> On Sat, Jun 21, 2014 at 10:17:19AM -0700, Eitan Adler wrote:
>> On 21 June 2014 10:08, Mathieu Arnold <mat at freebsd.org> wrote:
>> > Author: mat
>> > Date: Sat Jun 21 17:08:34 2014
>> > New Revision: 358722
>> > URL: http://svnweb.freebsd.org/changeset/ports/358722
>> > QAT: https://qat.redports.org/buildarchive/r358722/
>> >
>> > Log:
>> >   Add a small notice about maintainership of this file.
>> >
>> >   With hat:     portmgr
>> >   Sponsored by: Absolight
>> >
>> > Modified:
>> >   head/Mk/bsd.options.desc.mk
>> >
>> > Modified: head/Mk/bsd.options.desc.mk
>> > ==============================================================================
>> > --- head/Mk/bsd.options.desc.mk Sat Jun 21 17:03:33 2014        (r358721)
>> > +++ head/Mk/bsd.options.desc.mk Sat Jun 21 17:08:34 2014        (r358722)
>> > @@ -7,6 +7,8 @@
>> >  #  - OPTION_DESC?= must be 19 characters or less
>> >  #  - OPTION description text must be 43 characters or less
>> >  #
>> > +# - This file's MAINTAINER is ports at FreeBSD.org so that entries can be added to
>> > +#   it easily.  Any sweeping changes should be approved by portmgr.
>>
>> IMHO this is silly.  The risk of changing description text is almost
>> zero.  Sweeping changes should be reviewed, but I don't think gating
>> the review on a member of portmgr is useful.
>>
> Given what happened recently on those yes portmgr needs to review those sweep

I'll copy what I wrote on phabricator:

If anything, the recent commits proved that this statement is exactly
opposite of what it should be.

People add entries too easily, not realizing that this file is just a
default. This means that new entries are often too general or useless.

Also, people are too discouraged to change the descriptions when they
are useless.


-- 
Eitan Adler
Source, Ports, Doc committer
Bugmeister, Ports Security teams