> I use SCSI id 0 for the first drive. The SPARC systems
> are a lot different.

I have been using ID0 for everything, and it mostly seems to work,
as long as the boot knows where to go.

> I set my 3/470 up for net boot under SunOS and installed
> the base and etc packages in the net boot root. I also had to 
> put the NetBSD net boot program in /tftpboot. The system came
> right up on the net boot. I used dd to copy the miniroot onto
> the local drive and it ran fine.

I need to play with the netboot thingie.  I have yet to get it to
work well for me.

> You should edit the drive label before you copy the miniroot
> onto the hard drive.

Hmmmm, I have been doing it the other way around, assuming that
the sun format puts a good enough label to allow the swap to boot.
Should I be doing it differently?

> >Ohhh, I would love to try it if it can be made to boot from tape.
> >I tried the 1.3.2 and it died in the miniroot on my machine
> >(3/80-16Mram-2xsun207hd).  I tried loading it from sun3x boot
> >tapes and from miniroot dding to swap from sunos tapes, and it
> >died when entering the miniroot kernel.  What should the physical
> >drive ID's be for the first and second scsi drives on the 3/80?
> >Mine are 0 and 1, and for some reason it did not like bringing up
> >drive 0.  Could it be a label thing or an ID thing?... or something
> >else I should look for?

After two more days of concerted play, I finally have the sun3x port
up and running reliably on my 3/80 box.  Alas, there are some gotchas
that need to be resolved in some way.  On my machine, tapeboot wouldn't.
It never got past reading in the boot kernel reliably.  It also would
not read the miniroot reliably, and stopped at one place in the code.
I have deduced that it is an Archive drive problem, because I have seen
this happen over the years on various archive drives I use.  Funny, on
sunos, the drives work fine.  Also, on a Wangtek drive it reads fine.
I am beginning to suspect that there is a sequence of code bytes passing
through the Archive electroncs that turns it to a screeching halt.

The only way I could get the beast up was to boot on sunos, and format,
then load sunos, then ftp a miniroot in from another machine, then
dd the miniroot to swap and boot it.  Then the only way the I could
get the install tarballs in was to ftp to another machine and pick
them up that way.  That worked fine.  The install by tape wouldn't.
Why I am having problems with such boot/install tapes, I dunno.  It is
the first time I have run it those kinds of problems on tapes in 15
years.

Anyway, after I got the machine up, two other minor gotchas I could not
seem to resolve....

1).  Vi will not access the cursor keys on the sun termcap.  Why? 
     Is there a termcap entry or fix that will allow the cursor keys
     to be read properly by vi?  Everything seemed to reference the
     sun-il termcap eventually.  Is something not right in that?
     Should I borrow a termcap entry out of the sunos machine?

2.)  X would not install and run.  It installed OK, but died looking
     for the libraries.  I linked or cp'd libraries into the appropriate
     /usr/lib, and it came up to the X cursor on the background screen,
     then immediately shut down dumping xclock or xterm core.  Anyone
     have any idea what might be causing that, and is there a proper
     way to install the X suite (I will assume I goofed somewhere?)?
     It should be in the install guide, perhaps?

Kudos, so far though, cuz the port runs like a striped arsed ape, much
faster than sunos, and much much much higher throughput on ethernet I/O.
Once the install was going, it was flawless, and had some nice automated
set up features for ethernet, etc.  Well done....

I don't absolutely have to have X, and I did snatch a copy of the sources
from X.org, so I could build it the hard way, but the existing port should
work, if I get it set right?  I have the necessary rodentia and rodentia
aluminium crosshatched play sheet, so it might be fun to see it run, at
least once. IFF I can get the X running I would like to bring up a local
Apache server on the machine, for some webplay testing.

Anyway, thanks to all for the input, and if we can get the slightly off
bits resolved, it will make a nice suite on the sun3x boxes.

Once it is running to my satisfaction, I will try to port the bootstuff
to the sparc, so I can get my dinosaur sparcs up properly.  They also
can't properly tapeboot, yet.....

Bob Keys