Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Posting-Version: version B 2.10.2 9/18/84; site dmsd.UUCP
Path: utzoo!watmath!clyde!burl!ulysses!gamma!epsilon!zeta!sabre!petrus!bellcore!decvax!ittatc!dcdwest!sdcsvax!sdcrdcf!hplabs!hpda!dmsd!bass
From: bass@dmsd.UUCP (John Bass)
Newsgroups: net.micro.mac
Subject: MacSCSI, new roms, and HFS
Message-ID: <209@dmsd.UUCP>
Date: Fri, 10-Jan-86 10:27:10 EST
Article-I.D.: dmsd.209
Posted: Fri Jan 10 10:27:10 1986
Date-Received: Tue, 14-Jan-86 04:11:48 EST
Distribution: net
Organization: DMS Design, San Jose Office, CA
Lines: 66

I just finished testing my MacSCSI board with the new apple roms --
they worked fine dispite the "they probably won't work" bs I was given
during the summer. The new roms use the additional address line
A15 on pin one instead of the A20 line I used to decode the ncr address
space. We have HFS up and running under the HD20 startup system on
the MacSCSI board ... the changes are pretty minimal and affect only
the assembly code in np_dvr taken from the ram disk prototype code.

We are currently helping a mail order dealer get the AP Clip version of
the MacSCSI board into production bundled with the Infosphere server/spooler
package -- it is configured as an internal 20mb priced under $1k -- what
a deal .... .the outragous $2,600 spacey drive ticket is finally broken --
look for the Warp 9 adds.

We have drivers for a couple controllers now, the Xebec S1410, Xebec Owl
drive, omti 3100, wd100? series, DTC 510. An update disk will be available
for those who have our boards in a few weeks.

We have built small batches totalling 50 boards (plus one batch for a dealer).
We have been building to order in small batches since the demand has been
minimal -- I would guess largely do to the assumption the new roms wouldn't
work. We will continue to build to order, holding orders until we can
build a batch of 15-25 ... this has been running about 4-8 weeks.
That will probably improve since it isn't a dead board anymore.

I have helped about the same number of boards get built as the result of
free telephone support to a number of DDJ readers. Several univiersity groups
seem to have coubled togather PCB's and fabed a few units. Several others have
succesfully built the project on perfboard using point to point ... as an
aside it seems that all attempts the bring the rom data/address lines more
than 10" on a ribbion cable failed -- so the chip carrier design seems the
safest. Pulling up the unused inputs on the NCR chips seems to have been
the cause of most folks confusion ... the NCR inputs float low (ACTIVE) for
some reason ... not like nice ttl.

Todate no one has bothered to share back support for the update disk ... and
several have refused to share using some hard to follow logic about the large
amount of effort they spent .... it would be nice to wisk my hard spent hours
from their grasps ... I know that several have mustered togather partition
support and other controller interfaces.

I had thought that putting the source and drawings out in the Mac community
would foster some sharing of learning, code, and problems since little
is documented about Mac device drivers ... with the mac the only internals
documentation seems to be using mac nosy and many wasted hours of disassembly.
With the fabulous end user tech support from the Apple ....

It will be nice when the Ivory Tower guys at apple come down
to earth and start supplying the same support as most ofter MAJOR micro vendors
including IBM, DEC, Zenith, Tandy, Commadore AND Atari. From nearly all these
companies you can get schematics, firmware listings, driver listings, and/or
OS listings (in some cases for a signature and $50 -- in others a product
need, signature, and maybe $2,000 license fee). Apple is the only major
company in the game that withholds EVERYTHING from everyone ... users,
developers, and dealers ALIKE. Apple is losing many VAR accounts over this
issue -PLUS- the high price differential now apparent with machines like
the ST out and shipping at 1/3 the price.

Not everyone at apple has this same view ... the developers support group
has been pretty good at getting seed material out to major vendors ...
It would be nice if smaller guys could get in a little easier...

My thanks to Kin Seto (Apple) and Evan Sulley (Infosphere) for the tech
support to fully realize the MacSCSI project.

now off to sleep and another project ....