Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!linus!decvax!harpo!eagle!mhuxl!houxm!ihnp4!inuxc!pur-ee!uiucdcs!uokvax!emjej
From: emjej@uokvax.UUCP
Newsgroups: net.micro.6809
Subject: Re: OS-9 an TRS CoCo - (nf)
Message-ID: <4281@uiucdcs.UUCP>
Date: Thu, 1-Dec-83 23:34:29 EST
Article-I.D.: uiucdcs.4281
Posted: Thu Dec  1 23:34:29 1983
Date-Received: Sun, 4-Dec-83 04:18:02 EST
Lines: 72

#R:seismo:-42500:uokvax:3500018:000:3523
uokvax!emjej    Nov 30 20:35:00 1983

>Is the "OS-9" system available for the CoCo the same OS-9 which
has been around for a while? 

Yes, it's the same OS-9 as has been out; more precisely, OS-9
Level One, version 1.2. They have done some CoCo-specific things,
especially w.r.t. the display, and alas, they let schedule distract
them from quality concerning the device driver for the disk--it
has the stuff to twiddle RS's bozo 5" floppies wired in. (People
are already distributing patches to use 40 and 80 track floppies
and reasonable step rates, and the folks at Microware say they'll
do it right with the next release. The device descriptors are set
up so that you can read that kind of info and do the right thing
for each device.) Nevertheless, it is the same modulo those things--
a fellow got tired of waiting for BASIC09 to come out for CoCo OS-9
so he copied it over via the serial connection from his Smoke Signal
system, and it came right up and ran fine. *Many* things have to
work right to pull that off, I'd expect.

>Is it usable and does it come with a tolerable C compiler??

That depends on whether you consider C tolerable :-). The C compiler
should be the same as the canonical Microware OS-9 C compiler, which is
K&R except for bitfields and a couple of annoying preprocessor
limitations. As for usability--the problem is that OS-9 shows up the
places where RS cut corners in the CoCo design. A decent disk
controller would go a *long* way toward fixing this. To be more
specific--software does a *lot* of things that hardware should do (vide
programmed data transfer for the disk, software emulating a UART/ACIA
unless you buy the RS-232 cartridge, and software keyboard polling), so
that if you have too much going on at once, things slow down. If
performance problems can be overcome (there are rumors of a super-CoCo,
and Tano is supposed to come out with a 128K Dragon (running Level
Two?), so it might be worth a wait-- even more so, the 68000 version is
due out early next year), it will be quite usable. I did/am doing some
software development for a fellow with a Level Two Smoke Signal
Broadcasting system; it performed quite nicely indeed and was a pleasure
to use.

>The earlier OS-9 made some pretence at being Unix-inspired, if not very
Unix-like. How does this compare??

As mentioned before, this *is* the earlier OS-9. The same shell (primitive
compared to the Bourne shell, or even the Mashey shell, but then, how much
can you put in 1K?).

>Can you really do anything useful with a single 150K flakeydisk?

No. Get more than one (non-RS, at that, for decent # of tracks
and step size); it is, alas, yet another corner-cutting that
the RS controller uses the WD 1793 controller chip, which doesn't
know about double-sided disks. (There's a canonical hack, due to Frank
Hogg, that makes it think that the other side is another single-sided
disk.)

(Also, folks have connected hard disks. Come to think of it, an outfit
called Green Mountain Micro sells a bubble memory attachment for the
CoCo, and I need to pester them about an RBFMAN device driver...)

>Does the CoCo come with any serial ports and/or a parallel printer
interface??

See the above note about serial I/O; there is a connection for serial
I/O on the back of the CoCo, but as noted above it's software emulating
a UART, and slows the machine down rather fiercely. There are some
outfits that sell parallel ports for the CoCo; check out an issue of
*Rainbow* magazine for *lots* of info in that respect.

					Cheers (to steal a phrase),
					James Jones