[ On Fri, September 25, 1998 at 15:40:34 (-0700), Jonathan Stone wrote: ]
> Subject: Re: texinfo files 
>
> Primarily, the people who do the work.  if there's disagreement
> between the people who do the work, we thrash it out and arrive at
> some decision -- if the developers dont reach a clear consensus, then
> from the Core group.

OK, that's good.  It would be nice though if policies that the Core
group does decide upon get recorded in public for "us" developers, and
that the the individuals in the Core group could try and refrain from
giving out un-written policies.  I'm not trying to point fingers, but
issues like this have come up in the past, and often there's a whole lot
of hand waving and excited talk before we even find out there's been
background discussion and descisions made.  I realize there are
sometimes timing issues, but it would be really nice if policy
decsisions could be published, hopefully with a rationale, on both the
apprpriate mailing lists and on the WWW for posterity and later review.

I did try and find some statement on the WWW to the effect that all
components the base system must be fully and 100% functional within the
context of the base system only, but I couldn't.  Perhaps I missed it,
and if so I'd appreciate a pointer.  The closest I see is mention that
the base system provides all the functionality expected of a BSD system.
I don't think that implies that all components must be 100% functional.
Certainly that's not the expectaion I have of the "average" traditional
BSD system.

> I dont think Todd is completely correct here, btw: we have XNS support
> in the kernel but we no longer have xnsrouted, making it damn hard to
> acutally _do_ anything with it.

I thought there might be other examples, though I wasn't sure....  Of
course now you've gone and spilled the beans about XNS and it'll be
ripped out until xnsrouted is fixed and made available!  ;-)

> There are many people with different ideas about NetBSD's target
> audience, where we should go, what's "right". When you start doing
> some of the work and get experience in a particular area, that at
> least gets you a voice...

That's exactly my point.  I'm trying to offer a solution that I think is
all-encompassing (i.e. it doesn't leave anyone out, it doesn't prevent
local modification in any fashion, it's as inocuous as possible and
appears to be invisible when not needed, and it lets folks know what
they might need to do in order to replace the stub with the fully
functional program of the same name).

-- 
							Greg A. Woods

+1 416 218-0098      VE3TCP      <gwoods@acm.org>      <robohack!woods>
Planix, Inc. <woods@planix.com>; Secrets of the Weird <woods@weird.com>