Re: Making desktop setup hard (Was: Two NetBSDs)

[David Sheryn <david%chromiq.org@localhost>]

Andy Ruhl wrote:

> On Mon, Feb 9, 2009 at 7:27 AM, David Brownlee <abs%netbsd.org@localhost> 
> wrote:
>>        Why is this fanatical focus on stopping people who want
>>        the graphic environment automatically installed so they
>>        can get on with their work?" My natural working environment
> 
> You're spending way too much time answering a nonsensical argument.
> 
> I'm baffled by how much "It's not good for me so it can't be good for
> others" attitudes there are on this subject.
> 
> I have no idea if I will like it, but it seems like progress to me, so
> I'm behind it. 

I tend to agree;

> Those that don't like it can merrily continue on their
> way with the status quo as if nothing ever happened, right? What's the
> problem?

The percieved *problem* is that (1) getting the "new" desktop to work
seamlessly _may_ require changes to the "core" of NetBSD, which may increase
bloat / reduce security or performance / insert other random prediction of
doom.  The concern is legitimate, but I'm sure that the NetBSD release
engineering people will be up to the challenge of ensuring that only the most
minimal changes required will make it through -- everything else will be
optional and on top. (2) it will, it is argued,  dilute engineering effort on
the "core" of NetBSD, i.e. on the bit that the desktop-averse look to NetBSD
for.  Modulo the few changes above, it is unlikely that the majority of the
people working on such a project will be the same people as working on "core",
so this fear is probably unfounded.

As I ranted earlier, what kicking-off this project has done is to raise the
fundamental question of "who is NetBSD for" -- the illuminati and initiated,
for whom the NBD "feature set" has gone as far as is suffcient for them (and
therfore doesn't need to go any further); for a generation of people who
really ought to have something better than Windows (and for which NBD _could_
be an excellent foundation onto which a useable desktop could be built) or
something in between.  The desktop-averse think there is no middle ground;
it's either "austere" NetBSD or the world end when all the noobs start playing
with the "friendly" version and asking dumb questions (to be scupulously fair,
I have seen this happen with other software projects, and it's not pretty.)

That there _is_ a middle gound seems to have escaped them.  It is highly
ironic that the day before the announcement came out, I was trying to remotely
talk a competent network/firewall engineer through a NetBSD install so that he
could have cheap but effective remote network moniting box to gather some
diagnostics (NBD was an obvious choice: ssh for remote access, tcpdump to do
the work, etc.)  He manged to fix the BIOS problems himself, but trying to
remote-tutor an ab-initio vi user through editing /etc/rc.conf was, shall we
say, interesting.  A netbsd-desktop cd that would have simplified getting him
up nad running with wireshark would have been an excellent use of NetBSD...
and it's attracting people like him that could draw in valuable
expertise/resources that might help develop the "core " of NetBSD as well.  I
don't think he's been too put-off yet...

/DHS