by redmail.netbsd.org with SMTP; 4 Apr 1999 21:30:15 -0000
	via sendmail with stdio
	id <m10TuT5-000018C@smtp.fundy.net>
	for netbsd-advocacy@netbsd.org; Sun, 4 Apr 1999 18:30:11 -0300 (ADT)
	(Smail-3.2.0.101 1997-Dec-17 #4 built 1998-May-27)
Message-ID: <19990404183011.39632@fundy.ca>
Date: Sun, 4 Apr 1999 18:30:11 -0300
From: David Maxwell <david@fundy.ca>
To: bmike@bigfoot.com, netbsd-advocacy@netbsd.org
Subject: Re: Please give me some reasons.
Mail-Followup-To: bmike@bigfoot.com, netbsd-advocacy@netbsd.org
References: <Pine.NEB.4.05.9904041257160.267-100000@7thsun.piermont.com> <3707C145.36674B04@bigfoot.com>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
In-Reply-To: <3707C145.36674B04@bigfoot.com>; from bmike@bigfoot.com on Mon, Apr 05, 1999 at 03:45:09AM +0800

On Mon, Apr 05, 1999 at 03:45:09AM +0800, bmike@bigfoot.com wrote:
> Sorry, I posted it on the wrong place. But I still have a question.
> The PCs are not good enough but they are cheap. Price decides market
> share. I only hear 'FREE' software but never hear 'FREE' hardware.
> So, how many platforms will survive in the future?
> 
> Sorry again, I have no spite to NetBSD's users. I just want some ideas
> to help me making a decision.
> Maybe the best reason is 'Just like it'.:)

I think it's a perfectly reasonable question.

NetBSD developers tend to take the 'long view' of things. I like that 
I don't see gratuitous changes going into the system, only to be 
replaced by something else later.

In some ways, this is exactly what I dislike about Linux. In Linux
development, lots of people write drivers and patches, and any
particular 'distribution' (red hat/slackware/etc...) picks the 
ones they like and calls it a release. I don't feel confident that
people are looking at the 'big picture'.

FreeBSD, like NetBSD has a 'core' team of developers who review
and approve design decisions, and actual code, before they become
part of the system.

NetBSD however has been focused on being multi-platform from the
beginning, which gives NetBSD an even longer-term view of design
goals. The elegance of a driver supporting a particular chip,
which runs on a Sun, and Intel, an Alpha.... allows an easy
growth path to future hardware. 

PCs may be 'cheap' today, but someday the market will be driven
to drop compatibility with x86, the 640K address space, 64K
segments, and other things. The performance difference of a
486 over a 386, running at the same Mhz is due to design
impovements, and efficiencies, but the 486 is still a holdover
to old design principles. Same with every x86 chip. Take a
look at what an Alpha at the same Mhz can do.

I can't pretend to tell you what system will eventually
replace x86, it may not exist yet, but it will happen.

Linux and FreeBSD are working on porting their systems to 
other platforms, but it will take them a while to integrate
their code bases, they're starting out from a port-hostile
x86-centric environment, and will have to change the way
a lot of core code works.

NetBSD is already there.

-- 
David Maxwell, david@vex.net|david@maxwell.net --> Mastery of UNIX, like
mastery of language, offers real freedom. The price of freedom is always dear,
but there's no substitute. Personally, I'd rather pay for my freedom than live
in a bitmapped, pop-up-happy dungeon like NT. - Thomas Scoville