what next for the pkg_install rewrite

[Russell Jackson]

On 08/19/2010 07:38 AM, Bapt wrote:
> Hi,
>  
> Now that GSoC is over a lot of good work has been done on pkg_install.
>  
> I think it would be great to organize the way the on going work on pkg_install 
> will be done, for that purpose we need someone to officially manage the work, to
> validate what will be implemented/cleanup, and distribute the work among volunteers.
>  
> I think this is the best way for a new pkg_install really appear.
>  
> I'd like to contribute to the work on pkg_install, and would like to 
> help cocordinate the effort to prevent reinventing the wheel, and to be 
> ensure the code is consistent etc.
> 
> I know there are many like minded people out there who might like to help, please identify
> yourself, either publicly or privately so we can put together a team to find a pkg_install
> replacement.
>  

A little late to this discussion, but what would really help with
systems automation (i.e. puppet) is a way to install a package by
specifying the origin like you can do with pkg_info -O (e.g. pkg_add -r
-O databases/mysql51-server).

I've devised a ad hoc way to do this by fetching the INDEX.bz2 for the
correct arch and release and looking up the port name by the path. I
then feed pkg_add -r <baseurl>/All/<portname>.tbz. This works, but it'd
be nice if I didn't have to put such low details in puppet.

Another thing that would help is if pkg_info -O would exit with a
non-zero status if there are no packages installed with the specified
origin.

-- 
Russell A Jackson <raj at csub.edu>
Network Analyst
California State University, Bakersfield