Subject: Re: proposal: NetBSD system buyer's directory
To: None <current-users@NetBSD.ORG>
From: Erik E. Fair <email@example.com>
Date: 09/26/1997 13:14:29
Well, this discussion has certainly gone far afield of what I had envisioned.
There are two questions I thought about having the answer to on our web site:
1. Does NetBSD run on my hardware?
2. I want to run NetBSD, where can I buy a system that will?
Question Number 1 we've got down for most ports (some simply because there
aren't that many options, e.g. pc532), though we need to continue to be
specific about hardware models, rather than just saying "if it has a foo
processor in it, NetBSD will probably run". In addition, it's best if we
use the official names for systems and peripherals, rather than the code
names that they were developed under. Most of the web pages are in pretty
good shape this way.
Now, life is probably most interesting for the i386 platform, because
people build systems by the component (motherboard, CPU, add-in cards for
disks, graphics, sound, etc). For platforms like this, we really should
have a list of each component known to work (i.e. motherboards, CPUs,
add-in cards) so that when people go mix & match that they have a higher
probability of ending up with a working system. The majority of systems
NetBSD runs on, however, come as finished boxes, possibly with a peripheral
add-in bus, so the documentation problem is not nearly so nasty.
Question Number 2 was the genesis of this discussion (please note the
subject!) - all I was looking for was a web page per platform that gave
contact information for vendors that sell the hardware that we know can run
NetBSD, with no express or implied approval of same. A starting point for
someone who was researching, as it were. Do I hear any volunteers to
collect URLs, telephone numbers, street addresses, and platform types sold?
The "NetBSD user registration" question is a separate matter; that's trying
to keep in touch with our "customers", and let them know when updates are
available, about serious bugs they should know about, about upcoming
releases, etc. It would also be good for the developers to get feedback
about what is the most popular new device that needs a driver, just in case
we run out of motivation to find projects to work on by ourselves.
For the record, I am against automatic registration, and for including a
program that a user could invoke should s/he choose to, provided we tell
the user up front exactly what the program is going to reveal about their
your further comments are solicited,