On Tue, Jun 26, 2007 at 11:49:53PM +0200, Johnny Billquist wrote:
> But this just brings me back to the question if we want to do what 
> others already are doing successfully? (Ie. Microsoft).
> Is that the only goal of NetBSD? To be accepted and used by the most 
> number of people? Trying to accomodate for the largest group of people?
> Using the lowest common denominator?

No.  You're extrapolating way out with little or no basis for that
extrapolation.  I think the goals of NetBSD include being robust
and useful for the people using it and attracting other users.  We
could certainly use some help and it makes sense to expand the pool
of poeple who can and will help.  Proplib was not an element of
the latter, though.  It's an element of the former--of making the
system more robust and useful for the people using it.  I think
it's too early to say definitively that it succeeds or fails for
that, but it's a Good Thing to have a consistent and flexible way
to pass information around and to store and retreive it.  I think
it's reasonable to set a highish bar on bringing in new formats
because that impacts the "consistent" part.

plists have limitations, but frankly, the format isn't that important.
If you're editing plists too often--especially with a system in a
hosed state--then you're doing something wrong or the system tools
are lacking something and should be fixed.  Apart from the readability
and editability, the biggest advantage I've heard for using the
SCN format is that it saves space, but a binary serialization would
do that at least as well.

I think the concerns should be more about where and how plists are
used--not in what format they come.  I'm no fan of XML, but it's
functional, and I can't get too excited about it.

-allen

-- 
Allen Briggs  |  http://www.ninthwonder.com/~briggs/  |  briggs@ninthwonder.com