Two things:

1. My intention was simply to post some of what I'd seen at MacWorld and
the implications thereof: some new things are coming, some old things are
going, be prepared. Until we are prepared to build hardware ourselves,
NetBSD must be concerned with what is, not what we wish were so.


2. I am very pleased to see the commits for iMac support; it is an almost
ideal machine to gain wider exposure for NetBSD on hardware that will not
vary in configuration much, and which includes an awful lot of useful stuff:

We can assume a CD-ROM, and produce a CD for installation (or maybe even to
run from; it is 24x, after all).

We know which modem is in the box, so we can provide pre-packaged PPP
dialing scripts, fax support, and so on.

We know what video chip is there, and what the monitor configuration is, so
we can set up a standard X server configuration (big win here - we could
set it up to boot directly into X; from my own experience, configuring X
for the video chip and monitor is a flaming pain in the butt).

Lots of things become easier when you can count on certain hardware being
there all the time. To use a phrase I learned from the manufacturer, we can
provide a "superior user experience." According to numbers released at
MacWorld by Apple, there are over 800,000 iMacs already out there, and
another sells every 15 seconds. That's an opportunity for exposure that we
should not ignore.

This is not to say that we should abandon our aggressive Machine
Independence and generality; I'm simply suggesting that where hardware
sales volume of a relatively static configuration allows, we try to support
that configuration a little better than is otherwise our norm. It's much
easier than doing the full M x N set of all hardware in IBM PC-land, and
then asking the user to guess what s/he has...

	Erik <fair@clock.org>