Best CP/M machine?

From: allisonp_at_world.std.com <(allisonp_at_world.std.com)>
Date: Thu Sep 2 11:53:15 1999

> >From what I remember, the Northstar was probably the most widely hated
> "system" around in these parts. There was a local company which produced an
> apparently quite nice accounting package (Champion???) which wouldn't run on
> their most popular model, the "Horizon" because it had an 8K ROM which
> reduced their TPA under CP/M to the point where this package wouldn't run
> properly.

NS* system wasn't the problem it was the MPS-A floppy system, It looked
like 20 of ram/rom at 0E800h through 0EFFFh, It was memory mapped. There
was only 256 bytes of actual rom. This generally meant for CPM users that
the upper 4k of ram was unused.

The fix was simple, CCP and BDOS ending 256bytes belove the controller
and tweek the jump table for 0F000h. Then you put the BIOS above the
FDC in ram. Works well and you end up with a 56-58k system.

> The CompuPro combinations fit in the same category, i.e. the ones who loved
> 'em loved 'em, and the rest of us didn't. The owner of that company had the
> practice of having his people design circuits whith whatever he'd bought for
> cheap this week, and that meant that sometimes they were good, and sometimes
> they weren't. His boards often suffered from compatibility problems, even
> with other boards of his own manufacture. It was, to be sure, spotty.

Stating that is nice but I have about 25-30 of those boards (all the
interfacer models, RAM16/17/20/21/22/23, DISK1A, DISK3, system support,
8/16 cpu, CPU-Z, MPX-1, Mdrive, two crates) NONE support your view. This
may not be true for older boards (I'll bet the early ones were poor).

I know they were considered reliable as they were pulled from 10 s100
crates that were used here before PCs replaced them. They ran CHAMPION,
DBASE, BTRIVE and a few other familiar names using Concurrent-dos on the
8/16 cards.

> S-100 systems, in general, can't be viewed in the same way as, say, a
> single-board machine, because there is too much potential variation in its
> configuration to define it in a specific way. Some manufacturers sold board
> sets, about which they were willing to make certain claims about
> performance, etc, but most of them just wanted to ship their boards and let
> the headaches fall where they may.

They clearly werent PLUG and PLAY. Then again it was an industry wide
issue. The only way out was a one vendor box or do you own system
integration <at your own risk>.


> Computer stores, notably ComputerLand, quite popular in the late '70's-early
> '80's, tended to sell board sets from Cromemco, Vector, and occasionally
> NorthStar because the mfg would stand behind the sets they pushed. The
> Cromemco board sets were often displayed in a desk-enclosure with integral
> (vertically mounted) Persci (very fast, voice-coil-driven) floppy drives,
> into which it was very easy to drop a paper clip or something. Businesses
> tended to buy these because they were sold under a single aegis as opposed
> to letting someone "integrate" a system for them. The theory was that there
> was less risk that way.

Certainly true to my experience on the east coast, the real problem was
the inductry was so volitile that Fly-by-night computer was often common.
and getting support for those older combines was at best iffy. From 1975
through 1980 most every vendor we knew as the "originators" of the
industry either went under or changed names/product multiple times trying
to adapt to the changes that were going on. From 1980 on it only got
worse!

However, this is not a CPM problem.

Allison
Received on Thu Sep 02 1999 - 11:53:15 BST

This archive was generated by hypermail 2.3.0 : Fri Oct 10 2014 - 23:32:35 BST