Some where in the coco data base I saw a Atari.arc file to see ATARI Graphics
on the COCO, that is the one I am refering to. It was arced with tc2 then. I
never could de=arced the darn thing on my , thren coco2. I have d/l 3 or 4 arc
file to my COCO and no
go. Excer

Marty, I am having trouble with arc files.I d/l COLOUR KEYS 3 and tc3 and all I
get is probably compress with a newer version. I d/l a second time file by file
and is fine.I d/l to my IBM using procom and to my coco usimg ae.bin. One more
thing if I d/l us
ing MIKEYTERM OR ULTIMATERM , wh how do I have to save them, which
format(ascii, bin,etc.). Thanks Excer

Try downloading Xcom9 from the Os-9 sig and then use TRScopy to move it
over.

Steve Bjork

To get 80 columns at start up do:

MODPATCH
L TERM
C 2C 28 50
C 30 01 02

to get black background and green foreground (in addittion to 80 columns)

C 33 02 03
C 34 03 12
C 35 03 12

after these do:

V

This will modify the Term driver. Then you will have to do a cobbler to make it
permanent.

Hope this helps, Bill Haesslein

As others may have already told you, OS9 Level II for the CoCo 3
is available at any Radio Shack store. IF they don’t
have it in stock, you should be able to make them orderit
for you. If you have trouble there, complain loudly in a letter to Fort Worth,
and let us
here know about it too.

—marty

I have been given to understand that if the 12V is not supplied to
the write pre comp circuit, then write pre comp simply does not
happen. No more, no less. THIS is no big deal, since write pre comp
is unnecessary for modern 5 1/4 in 40 trck drives.

—marty

Question:

Are you using the controller directly IN the CoCo 3 system port,
or with a Multipak?

OT the only chip on the board
that uses +12 volts! The write pre comp circuit chip also
uses +12 volts.

I picked up a bunch of 5 volt only 1793 pin out compatible
Mitsubishi chips (M5W 1793-2 chips) for $1 each at a swap meet.
They work fine at normal speed in a CoCo 1 controller.
When I get my controller back from loan, I may try to see
if they make the thing work any better at 2 Mhz.

One electronic slavage place I knew of was sellig
delsoldered but working Fujitsu 1793 chips for about $10 each
a while back.

—marty

Thank you for the info. I guess I will just have to break down and buy a
multipak and an rs232 much to my chagrin. I don’t have a place for a multipak
right now.
I guess I will have to move my computer station over by the Telephone stand with
the deskphone on it and put the phone on top of the disk drive to have room to
put the multipak on the computer station.
When that is done then I will start using OS-9 L2 communications programs for
all my local BBS that I log on even use it for LD calls and DELPHI.
Thank you again for the info, even though it wasn’t what I wanted to hear. You
told the truth. That is the best thing for all. Have a good day. Dave

Greg-
I just bought OS-9 Level II, and am having big problems. The disk boots
correctly USUALLY, but whenever I try to do something, like FORMAT or BACKUP
a disk, it usually crashes about halfway thru. It either just sits there and
the drive stops changing tracks, or it stops and the screen flickers little
lines across it (crash!). A few times (1/4) it works OK, but the others it
doesn’t. Have any idea why? I thought it might have something to do with my
drives (kinda old, non R/S), but I’m not sure.

-Mike

Due to the nature of the hardware of the CoCo (bit banger port)
and the OS9 software (interrupt driven multitasking multiuser)
it is an EXTREMELY poor idea to use the bit banger port,
and makes VASTLY greater sense to use the RS232 pak, which
can be had for $30 right now on close out sale at RS stores.

Micro Technical Consultants offered Xterm that used the
port on the old CoCo 1, tho… and may have a version
that works under L2 on the CoCo 3.

—marty

The syntax for TRSCOPY is:
TRSCOPY FILENAME PATHNAME #bufferspace (for copying RSDOS to OS9)
TRSCOPY PATHNAME FILENAME #bufferspace (for copying OS9 to RSDOS)
Examples:
TRSCOPY /dd/cmds/FLIGHT FLIGHT/BIN:0 #40K
TRSCOPY LETTER/TXT:0 /d0/data/newletter
etc..
– greg

THE MULTIPAK IS #26-3024 AND IT HAS BEEN UPGRADED FOR THE COCO3.R/S SAID THE
CONTROLLER AND DRIVES MAY NEED TO BE ALIGNED. PERHAPS I NEED A NEW CONTROLLER??

WILD GUESS , DO YOU HAVE THE OLD RADIO SHACK DISK CONTROLLER ? THE ONE THAT
USES 12 VOLT AND WONT WORK IN THE COCO 2 OR 3 WITHOUT THE MULTIPACK ? IF SO ,
THERE’S YOUR PROBLEM .

The OLDEST of disk controllers… that sold with the OLD GRAY
CASE drives and with the CoCo 1… will not, I am told,
work properly with the CoCo 3 at double speed. You may
need a new disk controller.

Also… have you upgraded your Multipak?

What model of multipak do you have (Cat # 26-3024 or cat # 26-3124?)

—marty

-< '...I'll be watching you ..' >-

Try setting ‘t’ (echo cmds) at the top of your startup file, so you can see
how far your procedure gets before bombing out. I suspect it is exiting
before it reaches your xmode baud cmd.

On Level I, if your startup encountered an error it would exit AND leave
the error msg for why it choked. While I haven’t played around with Level
II much yet, my startup file will NOT leave the error msg onscreen (even
if it ever sends it in the first place – my vision isn’t that quick!). It
will just fall through to the OS9: prompt, with nary a peep about things
that it skipped, or why.

With the ‘t’ switch on, you can watch every move your startup file makes.
If my guess is correct and you have an offending cmd preceding your xmode
baud cmd, you will see everything up to and including the one os9 finds
unpalatable. Fix that one and continue.

Hope this helps out,

— drifter —

HELP, I HAVE OS-9 LVL II.
CAN I USE MIKEY TERM 4.3
TO DOWNLOAD OS-9 FILES
AND PROGRAMS ?????
IF SO HOW DO I CONVERT
THEM TO LOAD VIA OS-9
OR IS THERE A NEED TO
CONVERT THEM TO A OS-9
FILE ???

THANKS,
RICH DUNKLEE, WA2ONM
–… …– .
-.-. ..- ..-.

Try config #56K (or #48k if 56 is too much). I haven’t tried, but config
may just allocate the extra memory to the copy buffer, dramatically
decreasing the disk swapping required.

***** Dennis *****

Try, instead, leaving specific questions on the OS9 online special
interest group (type OS at the main CoCo SIG prompt).
That is more likely, I suspect, to get you the help you need.
Totally open ended questions like “I need help with OS 9… meet
me to help me” tend not to be very likely to produce any help.
Folks really can’t do much with such questions. But questions
about a specific problem usually bring fast and accurate help
to the asker.

—marty

Looks like the rs-os9 conversion program is getting it’s moneys worth in the
os-9 programs. Did not relize that it only changed
one track at the time I sent it in. Am glad that at least there are some
smart(er) people
out there that cuaght the slight mistake.
Even though I don’t use OS-9 myself thought that since others do that that
program
would be useful to others that might want to get some OS-9 programs like from
this
sig or other services or from a pal down the street. Use it in good health guy!!
By the way like the previrbiale stumbling over a dead body!! This program was
like that. Just happened across it lie out
begging to be taken!!!

Marty,

I picked up “Rogue” the other day so I could tinker
around with OS-9 L-II a bit and had a dickens of a time.

Seems it didn’t want to load and/or run most of the time.

In the process of messing around with it to get it to go
most of the time I observed several curious things.

The original 35 track single sided disk which the OS9 on it steps
at 33msec doesn’t like to work well at all for me.

I cobbled the disk drivers to work at 6ms and wow, what an
improvement!

I also picked over the loader on track 34 and notice it
kicks the Coco 3 into hi-speed right off the bat.

I have an old 1.0 (vanilla-no mods) disk controller and
it never ever EVER would do disk I/O in the high speed mode.
As a matter of fact the ROM is so slow that typing “DIR” in
hi speed gets me a “? syntax error”.

Naturally the ROM isn’t the problem since that’s all running in
RAM on the Coco 3.

I was wondering if there are any parts in my old 1.0 controller
that are just intrinsically slow and that differ in some way
from parts used in the newer 1.1 controllers which seem not to
have this problem.

I find it curious that setting the step rate to 6ms vastly
improved the reliability of disk I/O although it (OS-9 L-II)
still crashes now and then when disk I/O is initiated.

Got any ideas? I bet you’ll hear variations on this theme
as OS-9 L-II becomes available generally.

Oh, I didn’t mention that “Rogue” is an OS-9 “dungeon” game
that has an OS-9 L-II operating system.

Neet game when the disks behave!

If I can help with more info lemme know.

Mike

Special software drivers are needed for given particular
hardware real time clocks. There is no way the
general release of OS9 in any version can be expected to
suport any given particular third party real time clock.
The same is true of IBM PC real time clocks.

Given the great value of having the computer know the
day and date under OS9 or MS DOS (as the OS time stamps
all files, and this is QUITE a valuable thing), it
is disgusting stupidity on the part of Tandy, IBM,
and most clone manufacturers to NOT have included
a real time clock as part of the BASE system on the
mutha board.

Consider… a multifunction stop watch that runs for
about 7 years on a lithium batter.. of name brand
variety (Casio) typically sells for $20. And the
cheap watches, that just give day and date, are
commonly available for $2 to $5. And those have
a DISPLAY, not needed by a mutha board based clock.

I find the failure to put a clock in the CoCo,
IBM PC, Tandy 1000, and such to be evidence of the
relative lack of competence of the deisgn teams
associated with those projects.. or of the sliminess
and greed of the associated marketting division.

—marty

I was reading the messages you left in the database about the CoCo 3 and
Level II. Some comments:

OS-9 certainly wasn’t developed on the Color Computer! As I understand it,
it was developed as an operating system for Basic09, and was to be offered to
Motorola, which didn’t take it. Microware offered it as a product on several
different 6809-based systems and the rest is a minor footnote in computer
history… For a while, it seemed that OS-9 on the Color Computer was more of
an afterthought on Tandys part, although the Level II and Multi-Vue seems to
demonstrate that they’re waiting to stay in the game.

Generally, a 6809 compares well with a 8088 and even a 68000 in the number
of instructions executed per unit of time. Clock speed alone does NOT tell the
whole story! For example, I think the CoCo3 and Level II/Multi-Vue will compare
favorably with a 512K Macintosh and it’s windowing environment. The 68000 in
itmay not be running at 8 MHz, and it’s loaded down with handling the graphics,
soit does run slower than one might expect.

I don’t reasonably expect the CoCo3 can keep up with an existing Amiga,
although it shouldn’t look too shabby either. Possibly the same could be siad
of the Atari ST line (but perhaps not the newer models just announced).

On the other hand, if OS-9/68K becomes available for the Amiga and Atari/ST,
then that might be a very superior system.

Exactly how well Level II and Multi-Vue will perform remains to be seen. I
expect it WILL be a little slow because of the overhead, but there may be ways
around it. I do highly recommend running out of the largeest RAMdisk you can
get going, putting most of your CMDS directory into it. Or, a 10 meg hard disk.
I found the LR Tech host adaptor and a 30 meg fast hard disk (Seagate ST4038)
would run nearly as fast as a RAMdisk.

Some notes on Level II: memory management uses an address translator, not
bank switching as such to allocate memory. Each process has up to 64K or so
for its own exclusive use, and the system has 64K for modules, descriptors and
such which doesn’t have to be shared with individual programs. This largely
gets around the 64K add
ress range of the 6809 apparently with little overhead, though
it would be nice to have a larger address range…

Folks, including me, speak sneeringly of the Intel uP family and MS-DOS.
It DOES have a lot of excellent software and support, and I do use a Leading
edge PC for most of my computing right now. I am hoping that the CoCo3 and
Level II will provide an equally powerful environment, lacking only good
software
support. I think that the CoCo3 does put us up to about that level of
performance, and perhaps a little bit more.

OS-9 is very similar to Unix, except that it is written in assembly language
and is smaller and faster in a given environment (no fair comparing with a VAX!)

What’s the bottom line? Remains to be seen. Much depends on the software
support that the CoCo 3 and Level II will receive in the next couple of years.
I think it has a lot of potential, but I speak as a technically-minded hacker
and OS-9 user/enthusiast, and perhaps my judgement is abit clouded… The
system I REALLY want to have is a 68020/30 with 4-16 megs of memory and
OS-9/68K, but I expect that won’t be down to a reasonable level of price and
software
development for a couple of years yet…I want some SERIOUS graphics, you see…
As it stands now, the CoCo 3 offers the potential of about equal performance
and price with the PC and lower-end 68000 computers, if we don’t get too sticky
about the matter of software and additional hardware support. I look forward to
having a lot of fun and hopefully some good use out of my system.

Now, what’s holding up Level II and Multi-Vue (get cracking, Tandy!!!)

–Damon

Bad Mpaks in that way are rare.. but that is a possible
explanation.

—marty

I would be interested to see the reaction of Mark Siegel when he gets sales
reports on the sale of Level II in the month of Feburuary (February, that is.)

BTW, do youknow if the OS9 ram-disk driver for the 512K Disto RAMdisk emulates
a 40-track drive EXACTLY or does it need to be patched?

– pkw

OS-9 runs only on machines that use 6809′s or 68000 processors.
The IBM PC, with its 8088, cannot run OS-9, other than as
a software emulation of the 6809, which would be hidiously
slow due to the lousey quality of the 8088 compared to the
6809. A 10 MHz 80286 AT might be able to emulate a 6809
at an acceptable speed in software, tho. Alternatively ,
there are one or two hardware cards that have a 6809
on them that fit in the PC, and use some of its resources (its
keyboard and screen display and I/O. At present I know of no
software emulator for the 6809 on the PC or AT, and doubt
anyone would bother to write one. Afterall, emulators
are written for successful and popular microprocessors
to run on less popular machines, usually. Not the other
way around!

—marty

There are two coprocessor boards for the PC bus which use a 6809 and can
run OS-9. One was advertised in a recent issue of RAINBOW and another is
shortly
to be offered on a limited basis locally (Atlanta, GA). Now that’s about the
only way you can reasonably expect OS-9 to run on a PC, by using a 6809 in the
PC to run it! That will only run programs written for the 6809, not the 8088
and other Intel family processors.

Aside from that, there may be versions of UNIX or unix-like operating
systems
that will run on an AT (uses a 80286) or an 80386-based machine. I don’t follow
that very closely, so I’m clear on what products are being offered.

Hope this is of some help.

–Damon

In your January 1987 column in the Rainbow you mentioned some
patches for the OS9 CCDISK to allow double sided drives to be
used. What are the file names and which DL are they in? Can
these patches be used with OS9 version 2?

Here is a copy of the xcom9 help screen.

Prefix commands with ‘Cntl-Y’ (hex $19)
Use ‘H’ cmd for Help, ‘O’ cmd to show option settings
XCom9 prompts for command with Beep?
@ Send a break – 6850 only
$ Shell command
A toggle Ascii mode (strip parity,tabs, etc)
B set Baud rate – needs driver mods
E toggle local Echo mode
F specify Flow ctl (xon/xoff) for modem
G Go begin xmodem sending
H Help – show this menu
M toggle Modem path enable
O show Option settings
Q Quit – end XCom9
R Receive a file thru modem port
S Send a file thru modem port
V toggle View modem data on crt
X start/stop XMODEM protocol
CNTL-R enables/disables writing to Rcv file

This should get you started.

After writing some optimized fast disk duplicators,
I am quite familiar with the matter of tuning the interleave
to specific dedicated applications. But of course this involves
specially formatted disks.

You seem to have missed the basic question:

WHAT interleve should be used for formatting ones general
purpose OS9 disks? BASIC uses an interleve of 4 in its
format program (RS BASICK).

In the course of experimenting with differing interleaves
for READING sectors, I have found that ultra slow
step rates SEEM to cause one to need a different
read interleave for a given application.

I can understand your suggesting one mess with the READ
interleve for a given application…

Bu
BUT, I was asking about what interleave should be
used for laying down the sector headers during FORMATTING
of the disk. THAT should be kept the same, I would think,
if only to keep one from going crazy!

—marty

Dale: Here is some further insight in to one of the problems I
mentioned with T/S Word’s Tsfmt.

When adjusting and filling text, if there is a space at the end
of a paragraph, an extra blank line is printed.

If within a paragraph, a line ends and has an extra space
following a period at the end, this space is interpreted as the
end of the paragraph and the justification ends there.

These two problems may be due to the way text is sent to
the formatter (Via a pipe) which in turn limits the ability to
“look ahead” in the text to see what is coming next.

I am using DynaStar to compose my text. I set the wrap margin to
what ever I will be setting the Tsfmt line length to, and do not
do any justification in DynaStar.

Would I be having the problems of these extra spaces if I was
using T/S Edit?

So here I am again, filling out another Tandy Software Registration card,
and once again wondering… does filling these things out do me any good?
I’ve never gotten anything sent back to me in response when I fill out the
registration cards for OS-9 software… this could, of course, mean that
there is never any bugs in OS-9 software for them to get back to me on…
(har har chortle snicker). On the OTHER HAND….!

Dear friends:

I need some information about OS-9.

(1) WHAT is the optimum Interleave (“Skip”) factor
for formatting OS-9 disks?

(2) Does this number depend on the STEP RATE at which you
are running your drives? (I would guess it does to
some extent).

(3) A friend of mine running his 40 track single sided
drives at 30 ms step rate says the 2 is the optimum
interleave factor for formatting OS-9 disks. Is
this so? And what would be the optimum interleave
factor for those of up running OS-9 at 6 ms step rate?

(4) Would the question of whether one is running single
or double sided drives also affect the choice of
optimum interleave to be used in formatting an OS-9 disk?

I ask these combined hardware/software questions for several
reasons: On general principles I want to learn the answers,
but more specifically I am about to modify an existing
commercial mas dupe program for ultra fast copying
of 40 track (single and double sided) OS-9 disks
(using either a 512K CoCo 3 OR a CoCo 1 or 2 equipped
with a 512K Disto RAM disk card).

But need to know what interleave (skip) factor to use
in my format program.. and will have to adjust my
sector read and write order table accordingly too.

Thanks much for what help you can give me, in advance!

—marty

You ask a lot of questions for a christmas eve! OK… here goes:

(1) OS9 Level 1 version 1 will not boot up on the Coco 3.
You would have to upgrade to OS9 Level 1 version 2 (a $25 or so upgrade after
presenting
proof of ownership of Level 1 ver 1 at RS Stores, I am told).

(2) However, there will be NO upgrade policy for OS9 Level II
when it comes out, and that will pretty much immediately become
the operating system of choice for OS9 types. This is because
OS9 Level II can make use of all of the memory and extra
graphics capabilities of the Coco 3. But, as it is in effect
a massive new product and not really an “upgrade”, Tandy and
Microware decided (not entirely unreasonably) that it should
be sold as a separate product and not as an upgrade.
(My own feeling is they should give it away with each disk system,
but I doubt Microware or Tandy felt they could afford that).

(3) I’m not one of the best folks here to ask about OS9, tho
I’ve picked up a little about it. Our better OS9 authorities
include Dale Lear (DALELEAR), Greg Law (GREGL), and Rick Adams
RICKADAMS). Don Hutchison (DONHUTCHISON) seems to be getting
rather experienced in OS9 too. DRIFTER and DWHILL also seem
to be rather savvy OS9 users who hang out here often.

(4) As for OS9 and the RS232 pak, OS9 Level 1 Version 2 and
OS9 Level II directly support the RS232 pak via T2 drivers.
If you don’t have such a driver I strongly suspect you are
using the soon to be totally obsolete OS9 Level 1 ver 1.
In general, the RS232 pak requires support software
in order to be used. The ROM priveded with is is total
garbage. Numerous RS DOS programs use the pak, and
OS9 directly supports it now via the T2 driver. Several OS9
products are available that use it thru that driver.

—marty

Steve,

I consider Basic09 to be one of the few “usable” Basic’s around.
But then again, when you’re hooked on structured languages, you’re
HOOKED. My biggest gripe with most Basic’s is that you can’t express
yourself with it and the listing look like junk… But with Basic09
you can actually “flow” through a program like water flows through a
pipe.

I’m using some “tricks” with my terminal program to save space,
mainly memory… Yeah, even for Level II. I’m gonna keep the kermit,
and xmodem file transfer stuff in a separate program so that if you
want to download via xmodem, it
‘ll load and execute “xmodem” with the proper parameters and if
you then change your mind after performing an xmodem transfer and
use kermit instead, it’ll unlink xmodem and replace it with
“kermit”. Of course, under Level II I plan on having them both
loaded and linked in memory for even faster execution. The way I
figure it is, under Level II, each file transfer utility can be
allocated up to about 56k, so without the mainline terminal program
in the way you can download direct to disk using a much larger buffer.
If you look at it like the terminal itself is 13k, xmodem is 13k, and
kermit is 13k then that’s 39k with only roughly 21k left for the variables
and buffers. But, with each a separate utility you can have roughly 43k
for the variables and buffers for each. And then again, under Level I
how can you do anything with the full code being 39k? I don’t know
many people running Level I on the CoCo with 39k, and I only have 41k
myself with a STRIPPED boot file.

You’ll love it once you get running at 1200 baud, especially when
you don’t have to pay more for it on Delphi! Boy, I just wish I had a
2400 baud capable Telenet down here! Yeah Marty, I know, I know, 2400
may not be much faster with xmodem but still, it IS faster, especially
with BIG uploads. I just uploaded 2000 blocks to a local system here
in about 20 minutes compared to about 1 1/2 hours it would take at 300.

This issue has come up many times over the last two years or more:

The port address of the RS232 pak’s UART chip is fully decoded
and as such that chip can be addressed regardless of what slot
the RS232 pak is placed in. To THAT extent the RS232 pak is
fully slot independent.

BUT!!!

The RS232 pak’s interrupt control logic is NOT slot independent
and the RS232 pak’s slot MUST be selected for that aspect of it
to work (due to incompetent design of the CoCo system bus when
the first CoCo was standardized: They did NOT bring out the
IRQ or FIRQ lines to the bus, and instead created this
thoroughly obnoxiously kludgy CART line nonsense as a
Rube Goldberg way of handling interrupts via the system bus.)
Now the CART line is SWITCHED by the slot selector in the Mpak.

Thus, programs that use the RS232 pak under INTERRUPT control
will require that pak to be in a given specific slot.
Programs that do NOT make use of that interrupt (like
MOST RS DOS CoCo term programs such as Mikeyterm and Greg E Term)
will not care at all what slot the pack is in.

—marty

Hey, Greg, I was just thinking. You know all that extra memory on the
Coco 3, which is hanging out there unavailable to us Level I, Version
2.0.0? Wouldn’t it be nice if someone would write a version of PRINTERR
that would somehow load the whole ERRMSG file into unused memory (outside
the usual memory map), and then jimmied the system to look up numeric
error codes in RAM, instead of churning the disk?

Sure, there are problems. You’d have to make sure that the memory usage
didn’t conflict with Mike’s 80-column driver. And how would you make sure
that some other process wouldn’t get scheduled to run while the GIME MMU
registers were in a wierd state? (Turn off interrupts briefly while you
accessed the other segment(s)? Voluntarily sleep for a tick?)

Could it be done? Let’s kick some ideas around. Sure, Level II will be
here Real Soon Now, but why wait? And what about those people who opt NOT
to go for Level II? Who wants to tackle this problem? Note that I don’t
necessarily think that YOU have to do it… but does anyone want to
volunteer? (Don’t look at ME… I am too blasted BUSY, alas!)

Of course, maybe what I am suggesting is impossible, in which I will
slink back into my cave… and await some bugger doing it anyway!

That would have to be a pretty BIG bug in VERIFY that would make it
mess up on its whole purpose in life… Note that EVERY module that
Ernie modified had a bad CRC.

Yeah, he might have used (inadvertently) the on-disk modules with the
bad CRC’s, rather than the fixed versions that he used VERIFY to create.
It is possible, I suppose.

Does the percent sign at the end of your message mean that your message
was originally composed on a UNIX system? My messages sometimes have
that “feature” for that very reason. Makes me feel pretty foolish…
I really ought to go back and edit them… but sometimes I’m just too
rushed.

The OS9BOOT that COBBLER writes is from memory. Rest assured that I think ANY
question having to do with generating a new OS-9 boot is a good question.
It’s a very difficult topic to understand and/or master, even for the
initiated, and thus this exchange will no doubt help more people than
just yourself… of that you can be sure!

Dale Lear and I were discussing your problem
over lunch today, and we both would really like to help you get
to the bottom of this. If you do an IDENT on each module (on disk)
referenced in the OS9GEN list, and the CRC’s for those modules
are all correct, then the CRC’s in the resulting OS9BOOT file
really SHOULD all be correct. So your problem truly has us
puzzled, and we love a good puzzle… especially if we can
help it come out all right in the end…!

One thing you might try is to verify (generate good CRC’S) the ENTIRE
OS9BOOT file by doing your patches in memory, saving modules in memory,
etc, then using OS9GEN to generate a new OS9BOOT file on your second
drive (onto a freshly formatted disk). Don’t worry about the verify step
until this point. When the OS9BOOT (with admittedly bad CRC’S) has been
written to /D1, do the following:

VERIFY /D1/TEMP U
DEL /D1/OS9BOOT
COPY /D1/TEMP /D1/OS9BOOT

That DEL and then COPY thingie looks wierd, true, but it has to be done that
way… it is a trick to make sure the OS9BOOT file is still contiguous.

Dale and/or myself will stay in touch with you. It is truly a sad thing that
the nastiest, most touchy part about OS9 is the generation of an OS9 boot disk,
and yet this is the very thing that all OS9 neophytes are asked to do as their
first foray into the (otherwise wonderful, in my opinion) world of OS9. Not
that you are a neophyte, tho… you sound like you know what you are doing. As
Dan Rather says… “Courage!”

Actually, if I were to write them, I’d try to be polite
and constructive. I reserve my most scorching style
for forums like this, where it has some entertainment
value, both to me and, I hope, to others.

—marty

We have been told again and again that the ability of OS9 Level II
to support windowing, and the further enhancements that
Multiview will bring, will allow some DARN fine user interfaces
to be easily written under OS9 Level II. Let’s hope
(1) we actually do eventually SEE those products
and (2) what we have heard is so.

—marty

As best I can tell, Damon, there is NOTHING in the design of the
current GIME chip that would preclude the use
of 44256 (4 by 256K) chips for the upgrade. Indeed, there is
every reason to believe those chips would work just fine.

The only problem is that Tandy did not support this as
a drop in upgrade on the mother board, much as it did
not support the upgrade using 41256 chips, because 44256′s
are 20 pin devices, and cannot fit in the 18 pin sockets
of the 4464 chips currently used on the 128K CoCo 3.

Tho of course using bent up pins and the like, a hacker
could probably install same… if the 44256 pin out is
similar enough to that of the 4464.

—marty

Hello Mike,
Its unbelievable that the display is so capable of ‘expansion’
and yet TANDY stopped where they did. I have been gathering POKES
that make this screen do wonders ! When I do an IDENT -m co80
everything checks okay, but if I do an IDENT os9boot, I get
**Module header is incorrect ! ** No big thing… the disk
still boots and I’m F,D,and H with this display, just wondering
what might cause that ? Is it because I dumped to the modules
directory with a name different than what the os9 loads into
memory ?
Stephen (DRACMAN)

Greg,

I’ve also tried xterm. I wouldn’t be quite so harsh about its exterior
as Marty, but it is a little terse. On the other hand, I don’t think it’s
fair to compare it to a comm program on a 640K machine. A menu interface
costs memory, or annoying disk accesses. The developers probably wanted
to conserve memory for the buffer area.

The manual changing of word length for xmodem is kinda dumb, but then
again, xmodem was added after-the-fact in an update. But there is one
other BIG fault with the program which no one has yet mentioned. That is
the fact that when a buffer is attached to a file, it does not automatically
get flushed out when you exit the program normally. That is to say,
you have to go to the buffer menu and manually deattach any buffers that
are open. Otherwise, poof when you quit. Now that’s annoying!

—– Dennis —–

Don,

You need a special cable to run Xterm with XRS232 and XT1 on the
serial port. This is a simple cable and the contruction can be made
from an existing serial cable and the pinout is as follows:

COCO MODEM
GND ————— GND
Txd ————— Txd
Rcd ————— Rcd
Rdy ——-/ Rdy

The easiest way to do this would be to pull apart the end of the cable at the
modem, remove pin 1 (yellow) and strap it to pin 2 (green). You’ll notice that
with this setup, every time a character comes in to the serial port it will
cause the Ready line to change states which causes a FIRQ interrupt. This is how
XTerm can run higher baud rates on the bit-banger port. I will tell you for a
fact that it is well worth the effort to make this cable but it SHOULD have been
include with the package. You might want to give Microtech a call and see if
they goofed with your order because the DID include the special cable when I
ordered mine. That was a long time ago and they may not be including the cable
with the package now but I’d check just to be sure.

I have my shopping list all made out… (in this order)

CoCo 3
Multi-pak upgrade
OS-9 Level II
Developers package
Multi-vue (if a separate item)
Magnovox 8515 monitor

As you can see, the RGB monitor is my last purchase as well. I’ve got a BMC
green screen right now that is super sharp for 80 column text and monochrome
graphics and I’ll live with that until I finish feeding the bill collectors.

I use procomm on my IBM PC clone, and like it immensely.

On my CoCo I use Mikeyterm currently, but when
I get it fixed up I may switch later to Greggie term
or Ricky-term

—marty

The packaging on IBM software often involves high quality printed
manuals of many hundreds of pages. The software is often
many times more sophisticated, and therefore more COSTLY
to DEVELOP, than CoCo Software. Finally, advertising
costs in the IBM PC world are easily ten fold greater
than costs in the CoCo world. There is also the
factor that IBM PC owners went out and spent $1000
to $3000 for their initial investment in their machine..
as opposed to $100 to $500 in the case of CoCo owners.
Therefore they are more LIKELY to be willing and able to
spend more on software.

Thus, IBM software costs more in part because much of it
IS better than CoCo software, and in part because
it costs more to be in the market, and in part because
the buyers are willing to pay a bit more.

Overall, tho, IBM software tends to be far more
professional than the stuff we have for our CoCo..
certainly in most of the major categoriers (Terminal
software, Wordprocessing software, Data Base software
and Spread Sheet software).

But IBM hardware from Taiwan is FAR cheaper than
CoCo hardware accessories… and usually of better
quality too.

—marty

All of the assembled and tested 512K upgrades that I know of
that either are released (the Radio Shack one
and the PBJ one, sold by Spectrum Projects and a host
of other CoCo suppliers) cost $130 to $150.
The two that I know of that are to be released later
(J&R board and Disto CRC board) will also be in this price
range.

Note… the 16 41256 chips alone cost $40 or more
to the supplier. Then they have to pay for the board,
the sockets, the caps, the connectors, and the labor
of assembling the thing and TESTING it too. Finally,
the thing has to be sold at a sufficient mark up
to pay for the advertisements, and allow for
a dealer mark up. All of this considered, I would
say it is unlikely the cost of the 512K upgrade
will come down much in the immediate future.

Hackers, tho, will eventually be able to
buy a BARE 512K upgrade board (from J&R, and
possibly from Disto too… Tony was 50-50
on this the last time I spoke to him).
The price will be $40 or less for this.

Overall it is the obscenely nasty design of the
CoCo 3, its requirement for a plug in board to upgrade
the memory… that forces the cost of an upgrade
to be three times what it would be if the thing
had been desiged with the needed sockets on the mother
board. I am disgusted with Tandy’s decision in this
regard.

The oft-repeated “Tandy Party Line” “explanation”
of why they did not put sockets on the mother
board is “the board was too small to fit them all”.
This is a lie. The board could have been extended
an extra inch or so toward the keyboard with
no design problems at all, and if this had
been done they could have fit all the sockets
they’d need.

Their making the upgrade so it has to cost three to four
times as much as a drop in chip upgrade would is one
more example, I feel, of the incompetence of the
folks at Fort Worth. In saying this I am being
charitable to them… for the alternative
explanation is that they deliberately and cynically
wanted to soak their customers.

—marty

I have fI have found the bug in co380 for the home control
code. To correct change offset 201 from 80 to 08. This is for v1.1 and
v1.1b. You can do this with the following command:
debug
l co80
. .+201
=08
q
I’ll upload an update by the end of the week.
Sorry for any problems.
Have fun, Mike Dziedzic

Stylograph III (latest version that I have) is configurable for their own
hires 51×24 screen, O-Pak, WordPak running WordPak control codes and now
for Co80 using Radio Shack control codes. There you have, 4 different way
to run Stylo, not just the one as you had heard. You may want to wait if
you do buy Stylo and get the Level II version instead which can be configured
for just about any terminal on the market, including Co80.

Seems that the only real difference is that you got frustrated by trying to
run OS-9 off one drive and I went out and purchased two 80 track DSQD drives
and then finally a hard disk. You’ll find OS-9 much easier off of double-sided
drives but you must watch out for one very critical point:

You CANNOT boot off of a double sided drive UNLESS you run some type of program
such as bootfix on that drive.

Very typical of situations would be you finally got the 40 track DSDD patches
in and you now want to create a fully capable boot disk with all of your
CMDS and etc. directories on it. What you would do is usually as follows:

OS9:format /D0 R “OS-9 System Master”

OS9:cobbler /D0

OS9:bootfix /D0; * VERY IMPORTANT

If you fail to do that last step you will find yourself with a bootdisk that
will not boot. Worse yet, if you have already copied all of your files over
to this disk you must now start all over again from scratch. This is the only
pitfall you may run into. It happened to me until I figured it out by reading
the documentation, hehe.

I’m not sure at the moment. Let me look through the manual real quick and
see just how it’s ‘sposed to be done…
Ok Penrose, the first thing you will want to go is to get into the editor via
a slash command, ‘/EDIT’ and then use FETCH, ‘/FETCH’ and it will ask you
for the filename to get. A quick scenario would be as follows:
FORUM> reply

TO:

Enter text now. Use ^Z to Exit….
/EDIT
/FETCH
Enter filename: reply.txt

You should then be reported with the number of lines in the message. Notic
if you include a semicolon in the filename, as in ‘reply.txt;1′ you are
actually specifying the revision number if there are more than one file with
the same filename. Generally, don’t use revision numbers if there is only one
such filename in you workspace or if you want the last one uploaded of the
same filename.

I agree that Last Word is a totally useless word processor for anything
besides laughing at… Here’s the biggest quirk that I dislike, and the
one I use QUITE OFTEN. Want to Mark a block for removal, moving somewhere
else or copying? Well, you are limited to approximately one 80×24 screen
full at one time!

Dennis, have you scrolled backward? Do this next time you fire it up.
Get into a good sized file (10k or so) and move all the way to the
end of the file. Then do a ‘Top of File’… It takes quite a bit of
time seeing as how LWord copies the original file BACKWARDS! That’s
right, the first byte in the file is now the last byte in the file!
You don’t see this on the screen, but you will if you list out the
‘lw.scratch’ file.

I’m a humping on my word processor and hope to have it running before
too much longer. Ie, another year as it would appear…. But don’t get
discouraged as I’m continually at it. As soon as I get the user-interface
all figured out, the rest should flow along nicely.

I’m switching between the word process and the BBS at the moment as I’m
trying to get them both finished by 1988, hehe.

I’m currently working on the message bases of the BBS. It’ll have up to
25 separate message bases and each message is variable length. Meaning
that if you type in a message that’s 10 characters long, it’ll only
take up 11 bytes of disk space with the extra one being a NULL for a
terminator. I find this to be very pleasing especially since there are
a lot of people who like to type in messages like Whiffle does with
stuff like ‘Ok, see you there.’ or ‘Have a happy.’ or ‘No’… Hehe,
short, sweet, and too the point… Then there are a few that are like
me and Marty… Each message is like a novel. I thought you’d agree
with that…

There are reasons for getting ’003 Warnings’ but they are NOT significant.
If you assembly the file with listings turned on ‘asm file.asm l’ then you
will see a column in between the line numbers and the hex values that will
contain information characters like ‘E’ for Error followed by a description
or ‘W’ for Warning. There are two reasons for getting a Warning:

1) You used a LBRA where a BRA will work
2) You are accessing data outside of your program area.

These are meant to be programmer messages to kind of flag your mind that
you may be able to shorted the code by using BRA, BSR, BEQ, and etc.
instead of the Long (4 byte) versions. Also, it gives you a hand when
you attempt to access data outside of your programming area. For example,
if you were writing a device driver you would probably use something
like:

LDA #$00 Get value
STA $FF69 Clear the status register

Since $FF69 is not in your definitions of the data memory area Asm will
flag an error that simply means, ‘Are you sure you want to store data
at this address as you MIGHT destroy something important.’ so kind of
take these Warnings in stride but be glad that they are there in case you
do really make a typo and overwrite the direct page or such and cause
OS-9 to crash. In these cases the warning may save hours of debugging
to find out why to find out that you used ‘Buffer,pcr’ instead of
‘Buffer,u’ which are two TOTALLY different things. Oh yes, it will
most certainly flag an error on any attempt to store data into PCR
offsets instead of U offsets.

A very hand set of utilities you will be looking for to use all of the 40
tracks and two heads is SDisk available from D.P. Johnson. I’m using SDisk
with my two 80 track DSQD drives and it is excellent. It can read and write
CoCo OS-9, Standard OS-9, Japanese OS-9, etc. although you probably need
never worry about that. I think this kit of software is $35 and they have
done me a world of good and I don’t have any complaints with it.

Just to be fair, Computerware is also selling Disk Fix and Utilities that
can read/write 40 track DSDD drives and includes one VERY HANDY utility
called Patch. This utility allows you to do just that to any disk file
that you have. You can dump the file or a selected portion, modify bytes
in hex, ASCII, or ASCII with the high bit set, search for a two byte hex
value, and even update the CRC of a module without ever copying the file.

You will also need one other utility to allow the double sided disk to
be bootable since DOS expects the boot file to be on one side of the
disk only. DP Johnson seels BootFix to do this and if I remember correctly
such a utility is included with the Computerware utilities.

If you want my opinion I would buy both packages, I did, and found that
Patch on the Computerware disk was worth the price of the package even
though I use the SDisk drivers.

Which 400 pages of essential informatio will be trashed? The CoCo 3 manual,
e.g. reveals little about the internals of that machine wheras the CoCo I
manual at least gave some infomation “for assembly language users”.

Here, look at the manual and figure it out…

type=h For external devices, type is used for ACIA initialization
values (hexadecimal). The default=00. Bits 5-7 se either
MARK, SPACE, or no parity on all devices. Codes for these are:

000 = no parity
101 = MARK parity
111 = SPACE parity
011 = even parity
001 = odd parity

Bits 0-3 are reserved for future use.

Bit 4 selects auto-answer modem support features:
1 = on
0 = off

baud=h Sets baud rate, word length, and stop bits for software
controllable interface. The numeric coes for baud rate are:
0=6E, 1=12C, 2=258, 3=4B0, 4=960, 5=12C0, 6=2580, 7=4B00

Bits 0-3 determine baud rate
Bit 4 is reserved for future use
Bits 5-6 determine word length
00 = 8 bits
01 = 7 bits
Bit 7 determines the number of stop bits:
0 = 1 stop bit
1 = 2 stop bits

Look at this (exactly as in the 02.00.00 update manual) and tell me what
the value is for 2400 baud… From this idiotic table you’ll never guess
that it’s supposed to be 4! Fortunately I have these values memorized
and didn’t have any trouble with all of those typos. Next problem, once
you get all the values figured out, in binary, convert them to hexadecimal.
This is asking TOO much for a novice, and even some experienced users!
It is this kind of trash that is going to create OS-9 converts, like the
kind that used OS-9 and thought it was too hostile. Mind you, I love
OS-9 very dearly but there’s just got to be a better way.

Marty, there was some changes made inbetween 01.00.00 and 02.00.00… Somewhere
along
the way all that got modified, I think in 01.00.00 was where it was
changed. But in the beginning, if you attempted to access a non-existant
drive that you had a descriptor for or to access a valid drive that
did not have a floppy in it, it would time out in about 2 seconds or
so. No more, it will now sit there supposedly waiting for you to insert
a floppy into the drive. This is handy in some cases but I much prefered
to have it time out as I’ve had the same thing happen to me quite a few
times. In order to save the hassles and to free up more memory I finally
completely removed the floppy driver from the boot file. The only drivers
I now have are for H0 and TERM. I keep the other stuff in the CMDS directory
for those moments when I need them and then I load them. This would be a
very big pain if you were running off of floppies. OS-9 indeed is a hard
disk environment and any attempt to run it off floppies is anything but
easy. I tried it for more than a year with 80 track double-sided drives
and I still didn’t have enough room. My CMDS directory alone fills two
80 trackers on both sides…

Are you asking yourself why I aw
always use C for my programming stuff… #include “MODULES/Disk.io”….
#include “MODULES/Keyboard”… #include “MODULES/acia.io” etc. It makes
life a lot easier. I’m trying to keep together a set of utilities that
can be included for such mundane tasks, after all, who wants to keep
re-inventing the wheel. The OS-9 assembler makes it a little easier
as well seeing as how it allows you to include files via “use”.

Bob, Why would Tandy want to kill off a computer that out sold the Amiga (total
sales) in just the first 1.5 weeks????

Why would Tandy Want to kill off a computer that out sold the first year run
of the color computer one in just it’s first three weeks????

True, most of the first color computer 3 where sold to Color Computer 1/2 owners
. But the software will be there when the system needs to stand on it!
It takes time to write this new software. (most of what we are trying to do
has never been done on a computer that cost less then $3000.00)

Beside, if it takes Tandy a little while to come out with the new software
then just what much more $$$ for the third party software houses out there.
Right?

Steve Bjork

If you what to add color to my boucing ball program just change the Shape
file. But will take some time to update this 16k file. (It took a Color
Computer about 3 hours to create it in the first place.)

I will be starting on a new Boucing Ball program for the Color Computer 3
(under Os-9 level two) will a ful rainbow of colors! I hope to have it
done in time for the Color Expo 87 in January.

Steve Bjork

Dale: Just a few quick quick comments on TS Word since you’re taking
notes…

It would be nice (but not essential) if there were some provision to
insert the date and/or time in the header and/or footer. I am
curently working on a paper and a easy time/date stamp would be quite
nice to tell all these versions and drafts apart! Now I just type the
info in by hand.

The capability of two lines of headers and footers would be nice.

When you use .nx to get the next section of the document, the next
file starts at the top of a new page. I think this is a mistake as a
logical break in a document may not be at the top of a new page. I am
breaking up my current file for reasons other than the start of a new
section, and this creates a problem. What I will probably do in this
case is to merge all the files together before I do the final print.
If someone wants a new file to start at the top of a new page they can
insert a .bp directive.

I am entering my text using Dynastar on a Wordpak II. I have DS set
to the line length I will be printing at. I am not justifying with
DS, just word wrapping. I let TSFMT do the filling and justification.
I found one problem/bug with this approach. If there is a period at
the end of a text line that is followed by two spaces, TSFMT
interprets this as the end of a paragraph and does not do any filling
or justification for that line. If one of the two spaces after the
period is deleted, then the filling & justification works just fine
and the line is properly justified and filled using text from the
following line if needed.

Please note that I am not complaining or anything, I can (and am) work
around all of these items, I’m just pointing them out to you for your
information and consideration.

Dale: Got a couple of T/S Word questions.

1. Is there any way to put comments in to a source file? From
my experimenting it appears as though undefined formatting
directives (such as .co) are ignored and replaced with a blank
line in the printed output.

2. It looks like T/S Word always puts two spaces after a
period, which is fine in ordinary text, and is a feature that
is lacking in Dynastar. I’m typing up my bibliography and
there are two spaces after my abbreviation for volume (vol.),
and after author initials (Strecker, W. D. for example) when I
print it through TSFMT. Is there any way to turn this feature
off? I looked through the manual and didn’t find any way to do
this.

Thanks in advance.

Ten years, I want an Xmodem free world now, hehe… By the way Mark, if you
do any patches to Kermit, there are two that I definitely want to see in
there at any cost… Make it possible to perform an R to receive and
an S to send files withut having to go to the shell and forking anew
kermit… Would definitely make my fingers a lot more appreciative.

One more Kermit question for you, actually a Delphi question for any one
that knows. Is it possible to download more than one file at a time from
DATABASES wiht Kermit? I tried using several different means but they all
download one file at a time and Delphi reports “Status: 1 file downloaded”
so it appears to be possible. Perhaps I’ll have to drag out the Delphi
handbook and see if it is in there.

Greg,
ex: “kermit -sl /t2 file” The “S” say send and the “L” is the
line paramater and it means “do th send on the spcified line”, in
this case /t2. Whne kermit is local, you don’t need thatas it sends
and recives from/to the keyboard. Note that the order of the options
to kermit IS important. “kermit -sl /t2 file” and “kermit -sl file /t2″
meab two very different things!!!

2nd, here on delphi, there are 3 kermit commands. The two most
important are “kup” and “kdown” which up and download using kermit.
The single command “kermit” puts you in “server” mode, sort of remote
control kermit. In order to end kermit on delphi’s side, you have to
issue a “kermit -ql /t2″ command to send a “quit” signal to delphi.
in Order to send files, you do the standard “kermit -sl /t2 file”, BUT
in order to get files from a kermit server, you must use the “kermit -gl

Ho
hope thishelps

—Mark S.

Kevin,

certainly two of those RS mini RAM boards COULD be used by
a hacker to kludge up their own 512K upgrade board…
I know a chap (the same feller who got ME my copy of that
CoCo 3 schematic) who tried to kludge one up after
chopping out two rows of 8 DRAM sockets from some defunct
piece of Tandy hardware. But this approach is HIEDOUSLY time
consuming. Considering at least THREE vendors will be supplying
512K upgrades, and at least one of them (and probably soon at least
two) will be making BARE boards available for at most $40 each,
I suspect THAT will become the way to go. Certainly it is not worth
MY time to hassele building one prototpye from scratch.

On the other hand, it probably will not be hard for a technically
oriented chap at any club to run off a few dozen PC boards of his
own for the club. One will assume the chap will lay out his own
board as opposed to pirate the pattern from one of the available
commercial boards. Tho if price competition ever causes the
cost of a bare board to drop to much below $40, it hardly will
pay for ANYONE to make their own.

By the way, in correction of what I just wrote…

I did not mean thre VENDORS, I meant three totally separate
DESIGNERS will be introducing 512K boards… there will
be MANY vendors handling one or another of those three
different 512K boards.

Consider that the cost of producting one totally bare
512K board is certainly under $10 if one is doing it
in quantity of 1000 or more. The usual markup from
production cost to retail sale price is a factor of three to
four, placing the likely sale price of such bare boards in the
$30 to $40 range, assuing anyone will be selling such an item.
And I have reason to believe that not one but perhaps two
differnt third party 512K boards will be sold as such.
But for now I doubt the retail price will drop much below
$40 for the bare board. Note that the chips to populate
it cost $48, and that is not counting the sockets and
the dozen or more needed capacitors. So, even for a hacker,
I suspect the 512K upgrade will not cost much less than $90
for a good long while. Unless the price of chips drops dramatically.
At present the price of chips seems pretty stable.

—marty

Steve, I got 2.0 to boot up with my hard drive! I patched CONFIG to
work with my hard drive. I then created a MODULES directory and copied
all of my drivers, etc. into this directory and ran config. Thanks
for giving me some advice. I have a question for you. Have you or
anyone else done some graphics work under OS-9 on either the old
CoCos or the new CoCo ? I am interested in seeing some new programs
that use this capability on OS-9.
Bert

I hope that’s in source code. I don’t think I can download binary on my
PC. Really do need to get a modem for the CoCo so I can at least readily
download programs…

–Damon

We will very soon (in a matter of a few days at most)
have an 80 column driver for OS-9 level 1 ver 2
posted for public acess in the OS-9 data base section.

—marty

An 80-column driver for OS-9 Version 2.0.0 on the Coco 3 was recently
uploaded to the OS-9 database here on Delphi. I am supposed to check it
out and enable it within the next several days, but I am so busy helping
my wife handle the kids and the turkeys today that it will probably be
a while. But I am looking forward to using it myself…!

If you are referring to version 02.00.00 of level 1 OS9 then yes the
80 col is available if you are able to download the CO380 driver
available on that competing service (OS9 ug on Cserve). As I hear
but cannot substantiate theLevel 2 OS9 will support 80 cols when it
becomes available. I am using 80 col mode as I write this – very!
pleasant to use on my zenith amber monitor. And looks GREAT! on color
too
Bob Palmer

Mark,

I just tried out kermit and I never could get it to transfer a file.
While runing with XTerm on /XT2 I got into workspace and typed ‘Kermit’ and
it asked me if I were sending only text files and I responded ‘yes’ and then
it said type three control-C’s to abort. I then went to shell, and called up
kermit via ‘kermit -sl startup /t2′ and it never transferred. Nary a character
was transmitted though I did recieve some blocks (as viewed from xterm) that
looked like #N every now and again. I figured that maybe xterm was blocking
the way so I completely exited xterm and tried again. Nothing… So I figured
that since I logged at 1200 baud, 7 bit, even parity I changed /T2 from 1200
baud 8 bit no parity to 7 bit even with the same old nothingness.

Am I doing something wrong here or do I have to log in at 8 bit no parity
to get it to work. Although, I was under the assumption that kermit is supposed
to start sending immediately, unlike xmode which must wait for the initial NAK.
I’m a gonna try it one more time and see what happens…

Actually, when you use the OS9->RS converter, you’re just moving the files
from one disk format to another… they’re not being converted in any way.
By the way, Dale Lear (OS9 Section Leader here) suggests copying these
“converted” files onto a TRUE OS9 disk as soon as possible, as the disk
format created by the OS9->RS converter isn’t a true OS9 format.

Anyway, aren’t most of the BASIC09 programs in ASCII format? If they are,
you should be able to strip the trailing stuff (produced by an XMODEM
download) using OS9′s editor. Just delete any lines without text in them,
and you should be fine.

The downloading hint that I posted was for binary files, and there really is
no point in making a text file “executable”. In this case, my hint really
doesn’t apply… I’d try the editor, myself.

Good luck!
Don

Greg,
Kermit IS a terminal program. Though perhaps not the smartest
one. Pete Lyall uses it in combo with xcom9. Whne he wants to
to a file xfer, he uses xcom9 shell command processor to do it.

Yes, Kermit, even the os9 coco version can send and recieve
multiple files. The only probelm is that kermit09 does not
support meta characters. Thus a command line can look like:

OS9: kermit -sl /t2 file1 file2 file3 file4 file5 file6 …

One of the things I am working on in a version with wild card
support. I use kermit all the time between both my CoCo and
my Unix box and my CoCo and delphi. Feel free to ask any more
questions.

—Mark S.

Well guess it’s on of lives little pleasuresd. I had no need for the convert
program since I don’t use OS-9 myself and probably never will, but I did know
that it works well enough to move file to an OS-9
formatted disk. For me it’s good old Microsoft basic and a good assembly
language program and I don’t have to run the dickens out of the floppy drives
everytime I want the computer to do something.
the other choice is to go with a real heavy duty unit with about 600 to 700 of
ram and a hard drive. that’s 600 to 700k of ram!!Taker easy, keep cool.
Dan.

THEN I assume that you have noway of getting regular type of programs if the
terminal program you have is only os-9.
Perhaps there is more then one way to skin a cat so to speak. What other
terminal programs do you have? If I may ask.
Also if you leave me private mail I’ll foward
Darn service!! And if you don’t have any other terminal program could send you
a copy of mikey term. If you do have a good terminal program that does and
ecellent ascii
transfer you could give me a call and I could send you mikeyterm and a
diskconvert program to convert it back to binary on the disk.
what major town are you near? perhaps it might be covered by pc pursuit? could
tell by the area code that you are in.
Take care, have a good holiday.
Dan.

Have you tried kermit to see if it works? I tried using that frigging DTerm
that comes with it and I type A to the modem and get ‘AAAAAAAAAAAAAAAAAAAAA
ad infinitum’ until I type another character and then I get that character
ad infinitum…. So, I have yet to try my binary version of Kermit… I’ll
download it (or at least the rest of it, hehe) and give it another shot.

THANKS FOR THE INFORMATION DALE.
BRUCE WARNER WAS KIND ENOUGH TO HELP ME OUT WITH MY PROBLEM –HE EVEN
CUSTOMIZED THE DS PACKAGE FOR ME AT HIS HOME. MAN, CAN HE MAKE OS-9 SINGG!!!
SAW PIPES AND FILTERS USED WITH SOME OF THE DP JOHNSON FILTER UTILITIES. VERY
IMPRESSED.
MY ONLY PROBLEM IS THAT –INTERESTINGLY ENOUGH — WHILE MY DRIVE 0,2 STEPS AT
12MS AND MY DRIVE 1,3 STEPS AT 20MS UNDER TELEPATCHED TELEWRITER, THEY DO
>>NOT< < FUNCTION RELIABLY AT THESE RATES UNDER BRUCE'S CUSTOMIZED DISK OF DS
THAT HE DID FOR ME! SO I AM
NOW PAINSTAKINGLY COPYING EVERYTHING BUT THE NEW DISK DRIVERS OVER TO ANOTHER
SYSTEM DISK. WE’LL SEE DYNASTAR UP AND RUNNING SOON.
– PKW

PS I HEAR A NICE NEW SPELLER IS COMING OUT FOR LEVEL II ON THE COCO3 REALLY
SOON. ALSO!! ONE OF THE S-100 MAGAZINES HAS AN ANNOUNCEMENT THIS MONTH — ANY
OF THE MULTITASKING OS’S FOR THE S-100 INCLUDING OS-9 WILL BE COVERED IN
GREATER DETAIL IN SUBESEQUENT
MONTHS. THEY ARE SEEKING EXPERTS (OR AT LEAST KNOWLEDGEABLE USERS) OF OS-9 ON
THE S-100 BUS. KNOW ANY? TIME FOR PROSELETYZING!!!

Could that be done!? I seem to be constantly underestimating the CoCo.

I get a lot of use out of my 020… without any graphics — yet. For
a serious programmer the fast compiler is worth a lot. If you want to
run TeX, you will need about a meg of memory and you will yearn for
a fast processor. A PC takes about a minute to format a page.

Figure that very fast means 38KB at least. Wouldn’t that push bits
to the 6809 about as fast as it could take them? Given that kind of
speed a parallel port might be easier.

I was thinking of the 020 as a cycle-server for the CoCo. If it could
actually DMA’d to the video memory it would act like a super CoCo.
If that’s possible it would be SOMETHING!

Peter

I find it VERY hard to imagine why the heck anyone would
want a 68020 home computer that talked out a “very fast serial
port”. Sheesh! What an utter waste of computing power in
a home (=single user in most cases) machine. The 68020
makes NO sense to me for most tasks that folks currently
use a home computer for (home or small business oriented
word processing, spread sheets, and data bases).
The ONLY application where the power of the 68020
really adds anything significant in performance for
the single private user would seem to be in GRAPHICS
manipulators of various sorts. In THAT application
the 68020 should really shine, and allow setting up
truely professional hi res graphics editors that are
impossible to set up practically with any of the
lessor members of the 68XXY family. But for that you
sure don’t want a worthless slothful “very fast” serial port!
No sir… you want DMA of the video!!

Just puttin’ in my two cents, as always!

—marty

NUMERIC KEYBOARD WOULD HAVE DEMANDED A >>TOTAL< < REDESIGN OF THE OUTSIDE CASE.
THAT WOULD BE A GOOD IDEA. THE SEX APPEAL OF THE OUTSIDE OF THE COCO IS NIL.
CONSIDER, COCOJOHN. THE COMMODORE AND THE APPLE ARE POPULAR NOT BECAUSE OF THE
SOFTWARE THAT COMMODORE AND APPLE WROTE FOR THEIR OWN COMPUTERS!!!

THE MORAL OF THE STORY IS, APPLE IS POPULAR BECAUSE OF ONE PROGRAM (VISICALC)
AND BECAUSE OF ITS GOOD GRAPHICS. COMMODORE IS POPULAR BECAUSE OF ITS
PRICE/PERFORMANCE AT THE TIME IT WAS ISSUED — AND BECAUSE OF ITS GOOD GRAPHICS.
THUS, TO MAKE THE COCO3 GREAT WILL >>NOT< < TAKE DOING ON TANDY'S PART --
NECESSARILY. IT NEEDS TO BE FAIRLY INEXPENSIVE (IT IS) IT NEEDS TO HAVE GREAT
GRAPHICS (IT DOES — JUST WAIT FOR THE NEW PROGRAMS!!!), IT NEEDS TO PERFORM
WELL (THAT DEPENDS ON THIRD-
PARTY PEOPLE MOSTLY) AND ***********LAST BUT NOT LEAST******** IT NEEDS THAT
>>ONE< < PROGRAM THAT WILL BRING IT TO THE ATTENTION OF EVERYONE ELSE. LIKE
VISICALC DID FOR THE APPLE.

NOW.

I HAVE A DREAM.

THAT ONE PROGRAM WILL RUN UNDER OS9 WITH WINDOWING AND WILL BE >>AS POWERFUL< <
AS ANY PROGRAM FOR ANY OTHER LOW- TO MID-END COMPUTER. IT WILL BE PRETTY. IT
WILL HAVE A GREAT MANUAL. IT WILL HAVE FOUR-COLOR GRAPHICS ON THE COVER OF THE
MANUAL. IT WILL BE WE
LL-SUPPORTED. IT WILL BE FAIRLY INEXPENSIVE (LESS THAN $150). IT WILL >> NOT < <
MAKE UNNECESSARY COMPROMISES.

THATI
THAT IS THE PROGRAM THAT WILL BRING THE COCO OUT FROM THE SHADOWS AND ENSURE A
COCO4.

– PAUL

Yes Jim, TSWORD prints underline-backspace-character in underline mode. This
is supported by many many printers. It is as close to a standard as things
get. Unfortunatly, if your printer doesn’t support it you have to go a
different
route. The best thing to do is define one of your fonts as underline. To do
this
define the control sequence for font 2(underline) as 0F. Then define the
control
sequence for font 1 (standard font) as 0E. To change fonts on a single word or
group of words within a line you have to use the fill mode as well.

Good Luck.
D.

Go for it. I think my 020 is wonderful, but I’m a 6809 fanatic from way
back (the day I read about them in the BYTE articles that talked about
their development). I’ll be cheering for you.

Peter

Marty,
OK you have convinced me. I will break the kermit protocol
manuakl up into about 20k chunks. Then I will need upload time.
Let me calculate how long it will take to do so, and then apply
for the time. I will HAVE to be on a Sunday.

Keep you in touch

—Mark S.

I use a GMX Micro-20 for most of my computing. Mine runs at 12.5
(ZMhz
sometimes at 16Mhz. GMX sells a version that runs at 20MHz. I expect
a 25 MHz version soon, maybe they have demonstrated one
running that fast. As an interesting example of FAST, I can run
the entire C compile sequence on on a mediumm-sized file in less than
ten seconds.

The graphics board isn’t out yet, and when it ddoit does come out it won’t
have the softwre support the CoCo has. Windows isn’t working for
OS-9/68K yet.

Peter

Greg Morse has developed a great terminal package for OS-9 called XCOM9.
I use it all of the time. It supports xon and xoff as well as XMODEM.
You can get it through the OS-9 Users Group. It is distributed under
the pr

emise that you will pay Greg what you think it is worth (another
words, freeware).
Bert

DALE –
AT THE PRINCETON RAINBOWFEST, MY LADY FRIEND WON A FREE COPY OF DYNASPELL. (SEE
HOW YOUR WRK — WORK — TOUCHES MANY FOLKS!) AND SINCE ELIZABETH HAS NO COCO,
SHE GAVE IT TO ME. HOW SWEET! I OR — ORDERED — DYNASTAR WHEN I REDDEMED THE
COUPON FOR DYNAMSP
ELL. THE DOCS SAY THAT DS ONLY WORKS WITH THE HIRES SCREEN IN THE O-PAK PACKAGE.

HORRORS! SAY IT AIN’T TRUE!

CAN I PATCH DYNASSTAR TO WRK ON MY DISTO DISPLAY-80?
I ALREADY HAVE A MODIFIED CO80 MODULE INSTALLED ON MY BASICC09 WORKDISK. LLOKS
GREAT.
– THANKS FOR ANY HELP!!

– PKW IN ARLINGTON VA.

Thanks Dale! I hope that OS-9 users check it out. Enough of graphics demos
and more graphics utilities for OS-9, I say.

The last few days have been warm up here, but the storm raging it’s way to the
West Coast made it pretty cold here (Chico, CA) last night.

I bought an HP7908P Hard Disk with Tape Cartridge backup at a swap for $150.
I intend to write drivers and design and build an HP-IB/GPIB/IEEE-488-1978
interface for the CoCo. Would anyone out there be interested in an HP-IB
interface? I’ve seen lots of printers and plotters and stuff go for real
cheep at swaps just because they only have an HP-IB connector. If I can
get enough response, I may decide to market it. Waddya think? Any demand?

We have been having some trouble with the data base
due to new software being implimented…
hopefully in a few days MEGABYTE’S upload will
be available to you. Sorry about the delay.

—marty

Dale,
I was finally able to upload kermit byfirst, uploading
it to my LAN, and then from the LAN to delphi. Looks like
straight my CoCo to delphi is not possible.

—Mark S.

No, Greg…

A better approach is to use the MAX232 chip. This chip
is a five volt only chip, that has the following functions:

It provides plus and minus 10 volts (using a charge pump…
no external components needed). It also has on board
two TTL to RS232 (+/- 10V) converters, and two
RS232 to TTL converters. Pretty slick, eh?
Price was around $4 or so a chip, if I recall.

—marty

Marty,
It may interest you to know that I used kermit09 to
upload kermit09.I have used the RS-DOS kermit and know
that it does work. Has anyone evry
er, ever enlightened you to the tremendous advantages
kermit has over xmodem? If not, I could…. and there are MANY!

Marty,
Where is a good place to get at the +5 and +12 on the CC3?
I had some do-dads hooked up to my 2 that used these lines, but
I don’t have any kind of schematic at all for the 3. Any help
would be greatly appreciated.
Dan

After booting OS9, place the CONFIG disk in /d0 and a blank, formatted disk
in /d1.

Type ce chd x /d0/cmds. Then type config & follow prompts on screen.

If this does’nt help let me know. By the way, do you have Disk Basic 1.0 or
1.1?

Good luck!

Chuck Hoffman

Dale,
,

Is there any way to use TSEDIT with an 80column display on the COCO 3?
I’ve found an excellent driver on the OS9 SIOS9 SIG on Compuserve for 80
columns.
The only problem I’ve found is not being able to use my favorite editor.

Can TSEDIT or T & TSWORD be defaulted to use the CO80 display driver?
Help!

Chuck Hoffman

Sorry Don, but if you type “unlink module” where module is a module loaded
frorom OS9Boot then it will NOT be removed from the module directory. The only
wayy to physically remove a module from memory when it is the OS9Boot file is to
create a new OS9Boot file without that module.

I understood you to mean that you couldn’t get the 80 column format on your
new new CoCo III. The new basic commands for going from the 32 column screen to
the 40 or 80 column screens is “WIDTH40″ or “WIDTH80″ However if you follow
the sllow
the strings regarding the CoCo III there are lots of complaints about the
color leaching/bleeding on composite monitors. CLS5 is one solution or

messing around with the pallette commands until almost black and white is
a
acheived or better yet getting an Analog RGB monitor. Have you tried
turrning on your computer while holding down the CRTL and ALT keys????

Yep, sure enough, the CRC’s were bad and the fixed modules, when loaded with
the Config utility, create a new boot disk with decent speed and storage. Thanx
for the suggestions and help! If this keeps up, I may actually come to under-
stand OS-9, even though I already love it!

Now, as has been suggested, if we only had the ability to use the 80 columns
and new graphics capabilities…

I know, I know…it’s called Level II…

–Damon

Yup, I whipped up three previous version of that directory utility before
I finally got it the way I wanted with all of the checks that it does. Now
I’ve just discovered a bug in the blame thing… It doesn’t crash or
anything like that but if you do a DIR to a non-existant directory it will
print ‘Directory of . time’ and then sit there. So I need to get that fixed
and I’ll reupload it. I’m hoping to get an assembler version of it going for
too much longer to increase speed and decrease size.