[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: [Pkg-grass-general] I love this



> However, while we can prepare a lengthy wish-list, we may find our
> resources are a bit limited to address everything and we will need to focus
> on what the folks willing to put in the effort want to package the most.
>
> One question I do have is are we aiming to get a whole bunch of packages
> into the official Debian channel or are we also interested in maintaining
> unofficial debian packages distributed through our own repository for
> packages to big, non-standard or rarely used to officially package?
>

I'm not well-versed in debian project mechanics, but my 2 cents on the issue 
is that we shoot for getting the most bang for our efforts.  For me, that 
means our first concern is getting the software we want working on a Debian 
system.  Inclusion into the main dist should be secondary.  The strange 
compilation issues with php (and any other issues that _must_ be fixed if our 
packages are to make it into the main dist) need to be worked out, but they 
don't need to limit the usefulness of our efforts.

I guess I'd argue for an unofficial repository from which we can work on 100% 
integration with the Debian proper.  Anything short of that is going to leave 
us compiling custom versions for our own purposes.  That would make this 
effort much less interesting, IMHO.

It's ultimately a question of priority and I assert that a debian-based 
gis-friendly OS (even one that starts off with lots of sloppily-constructed 
packages) is much more useful than a handful of expertly crafted packages 
crippled to the point of being useless due to upstream issues that conflict 
with debian packaging guidelines, but which have most-likely never affected a 
single user.

Inclusion into the debian mainstream should be a goal, but not a restraint.

> NOTE: when I say we, I really mean you guys (and gals).  I'm happy to
> provide a bit of supporting advice and help upstream necessary changes to
> packages I have commit access for but my Debian system is ancient. I have
> carefully avoided learning anything about how to prepare .debs or .rpms for
> that matter. You have my moral support of course!
>

I had carefully avoided this as well until very recently.  With the exception 
of the learning curve, I think building debs is technically easier than 
straight compilation if there are more than a few dependencies involved.  
Getting debs to pass muster with lintian may increase this difficulty level.  
I'm not sure.

--Fred

-- 
==========================================================
Fred R. McDavid, III
  540-248-0838
    frm@bitdaddy.com
      BitDaddy Systems, Inc
       * Complex System Design, Management, and Hosting
==========================================================



Reply to: