lulue@manta.NOSC.MIL (Dan Lulue ) (01/25/90)
Here is some information (my experience, actually) about upgrading a Mac
Plus with the Radius Accelerator 16 and Full Page Display. I am posting
this both to inform and to ask for a favor.
First, the upgrade costs were as follows:
Accelerator 16 card (w/o 68881 co processor) $600
FPD $690
FPD card $450
Everything looked great, but not everything worked great. The Plus
refused to boot off my admittedly antique Photon 30 (Warp Nine), and
things crashed randomly after booting off floppy. The old disk boot
problem is well known. Some sort of formatting problem, or
maybe it was the interleave factor. Whatever. We chucked the Warped
Nine and got an Ehman (modern formatting). This booted every now and
then, a vast improvement over the W9. Then I started calling Radius for
support, which they graciously provided. Unfortunately, they provided
it in little dibs and dabs. The first dib was that the Accelerator 16 and FPD
roms needed to be upgraded (FPD = 2.8, Accelerator 16 = 2.0). The vendor
we purchased the boards from installed the upgrade under the warranty.
The next call to Radius (you guessed it, the rom upgrade didn't do it),
unveiled the fact that the INITs needed to be v 2.5 for the FPD, and 2.5
for the Accelerator 16. I checked my versions and they were up to date. Back
to Radius. "Oh by the way, if you have both the Accelerator card and the FPD
card, then you should only have one set of roms, the Accelerator roms. And
they should be on the FPD card. Replace the FPD roms with the Accelerator
roms."
By now, I'd learned to ask ahead, so I asked if that would take care of
the problems I was having wih printing. "Oh by the way," they said, "there is
a little problem with AppleTalk on the Plus and we have a software
fix that you can get off AppleLink."
That's where I'm at now. If you've either benefited from, or been amused by
my experience, and you have AppleLink access, could you possibly get this
INIT for me? The file is "laserwriterIINTfix" (something like that) in
the 3rd party s/w directory under "sw updates/radius sw updates", again,
something like that.
I am at an anonymous ftp site, so either e-mail if the file is binhexed,
or ftp if binary (ftp...,anonymous login, cd pub, put file).
I am sending this message Jan 24, at 8:55 PST, so if you read this a
few days from now, you might drop me a note before going to any trouble.
In all fairness, Radius has been very helpful, their product is good,
and the vendor is an angel. And all that. Now if I could just get
the system to work.
Thank you very much in advance.
Dan Lulue
lulue@nosc.mil
------------------------------------------
|Disclaimer| --
------------ --
-- These are my own views, no one could --
-- possibly be forced to write like --
-- this. --
------------------------------------------ts@cup.portal.com (Tim W Smith) (02/04/90)
The Mac ROMs have trouble with SCSI. Asking them to work with accelerators is asking for a miracle. If the author of the disk driver supplied with the hard disk never encountered an accelerated Mac, the code is unlikely to perform the bizarre rituals needed to work correctly. For example, after the driver does SCSIGet and SCSISelect, it will do SCSICmd. Under an accelerated Mac, SCSICmd will fail sometimes. Any normal person would treat this as an error and pass an error code back to the Device Manager. However, someone who has been warped by the unholy combination of the ROM SCSI Manager and an accelerator will either loop issuing SCSICmd calls until it works, or will check the SCSI bus to see just what phase it is in and issue the appropriate SCSI Manager call. I prefer the "loop until it works" approach, because looking at the bus requires either trusting SCSIStat in the ROM, or looking at the hardware directly. When I was experimenting with an SE and a couple different accelerators, I noticed that how the SCSI Manager gets screwed up depends on the accelerator. For example, with the Radius accelerator, it was SCSICmd that tended to screw up. With the Dove accelerator, it was SCSIRead. Tim Smith