Sunday, May 8, 2016

StarCalc, James Flavin and Chuck Harriger

We found James' name listed as a program author in some of our recovered system LU files:

All rights reserved..V 
Copyright C 1982, Star Computer Group, Inc..V 
This document may not be reproduced without the.V 
prior written permission of Star Computer Group, Inc..V -.V 

STARCALC:VERSION 1,RELEASE 1,PATCH 0,AUGUST 10, 1982..V -.V.V 

STARCALC.06.V.V 

JAMES FLAVIN..V.V 

CALC MODE EXIT..V -.Y.s|..wq.s3wq.s.wq.s|...wq.s.wb.Y.s|.Ewq.s  


We also see the name CHUCK HARRIGER, whom we have recently contacted on LinkedIn.

All rights reserved..V 
Copyright C 1982, Star Computer Group, Inc..V 
This document may not be reproduced without the.V 
prior written permission of Star Computer Group, Inc..V -.V -.V 

MENU..V 
11/19/82..V 
CHUCK HARRIGER..V.V 

MENU SELECTOR.V.V -.Y.s| wbY.s|..wbPAXm.{ y.b\}.}pbV -.V CLOSE ALL CHANNELS..L_.cV.S.m.d.b.F..bT.b.L_.cIi.bV -.V SET ACCOUNTS..W.s.wm}..v..b 

STAR TECHNOLOGIES, INC..W.s.wm.bW.s.wm.bW.s.wm.bW.s.wm.bV.V.V.V.V.W.s|..wm.bV -.V SET PROGRAMS..W.sAr.vwmi1/STARWRTR.DEMO.b.W.sAr.vwmi.bW.sAr.vwmi.b

We've been in contact with James, and here is what he has to say:

05/08/2016:
Hi. Yes I remember working on StarCalc. We thought we could do a better job than Lotus 123. 

I did write some of the disk and tape drivers but I don't recall working on any compression.

I did work with Chuck for several years but I have not talked to him recently

------------------------------------------------------------------

Whenever we get our systems fully reconstructed, and can run StarCalc, we will show a demo here.

So, James, what we would really like to know at this stage, is in the fall of 1982, when you guys were writing StarCalc, were you running this on a straight Point 4 IRIS OS install, using Point 4's hardware, or had you created a proprietary processor and/or hardware for this?

Michael Marsalle from Microtech said that they build their own processor, but could this be after the Star*Technologies merge with Microtech, and this StarCal was prior to that?  Michael mentions that happened in 1983, and your StarCalc dates are from 1982.

For us to understand the architecture of the machine that this was running on will help us adjust the parameters of the emulated environment that we are using.

------------------------------------------------------------------

At that point in time we used Point IV hardware and Microtech hardware interchangeably. We just had to use the correct drivers

Note that IRIS required a Pico-N to operate. 

We also built an add-in board that could plug into an IBM PC and run the IRIS and BITS operating systems. 

James


-----------------------------------------------------


That's great to know, James.  This means that we *should* be able to get an emulator that is set for a Point 4 architecture to work, at least in theory.

We have done some documentation on the Pico-N on our site here:




According to the Computerworld articles, it implies that this was a device that was implemented with IRIS R8 and above.  All of our system backups have IRIS R7.5, so we are hoping this is not an issue.  There is some confusion on this, however.

Dan Paymar, who did a lot of work developing IRIS back in the late 70s, remembers the Pico-N from the late 70s, but it could be that these systems were just running an earlier version of the software even after the Pico-N was available.
  

We see below that the LU0 from this system states that IRIS is R7.5, alongside your company's Star*Writer, which I am guessing was also created by Star*Technologies, alongside your StarCalc.

<IRIS Revision 7.5     STAR*Writer Revision   2.19..............2 7/ 2/ 2/ 2/ 2/ 2/ 2/ 2/ 2/ 2/ 2/ 2/ 2/ 2/        ...(P118 1SWR.TAPECOPY

Do you think we are on the right track here?

-----------------------------------------------------



Yes you are on the right track and an emulator is definitely possible. 


The version of IRIS that you have with Star*Writer does require a Pico-N

I do not recall what happens if you don't have a Pico-N

The OS load sequence is fairly straight forward. I wrote the BIOS load sequence for several machines after the merger between Star Computer Group and Microtech. 

James

-----------------------------------------------------

We sent James the full LU0 & LU1 files extracted from QIC tape.

-----------------------------------------------------

So are you trying to boot using emulation software?  

What are your plans for a console?

-----------------------------------------------------

Yes, we are.  If I understand your question correctly (which I may not...), the emulator provides its own console options and settings.

For example, SimH will run DG NOVA RDOS well as it is.  It was written by a guy (Bruce Ray) who has written an emulator called reNOVAte, which is purported to run all Point 4 IRIS versions as well.

Here's an example of someone else just installing and running RDOS on it.

http://wconrad.github.io/20151207/simh-nova-hello-world.html

Right now, we are working with SimH, since it is open-source, and reNOVAte is not.  We know will need to modify the LU0 to comply with the SimH DKP environment, or modify the SimH DKP parameters to match the drivers and drive architecture of our existing LUs, or some of both.  We are mid-process with both of these efforts, slowing working toward something that will boot.

We are currently working with some C programmers to modify the nova_dkp.c to get closer to what we see in our LU0 drivers and disk architecture.

http://microtechm1.blogspot.com/2016/05/modifying-c-in-simh-to-run-point-4.html

-----------------------------------------------------

That brings back a lot of memories. 

I was not on the team that created those tapes. I believe you could boot from tape and then select diagnostics or backup and restore. 

I wrote all of that code into an EPROM that was resident on newer hardware. 

From you notes it seems like you won't have any problem with the console. Your simulator seems to have the ability to intercept the TTY instructions. 
....

It seems like you should be able to load Block 0 from the tape into memory and execute it.  

I think it had a menu which should come up on the console. 

It should also have a very rudimentary driver to read in the rest of the tape,  usually into a location above 400 (octal) and then it would jump somewhere into the loaded program. 

The challenge we faced at that time was to keep the tape streaming or it would seasaw back and forth as it read each block.  Most of the work on the drivers was to solve this issue. 

If you can get to the boot menu. You should be able to load and execute any of the programs, including IRIS. 

The challenges will be the Pico-N and the disk driver that is loaded in the OS. 

James

-----------------------------------------------------

And next, we hear from Charles Harriger:

-----------------------------------------------------

Hi, yes I did work with James on Star*Writer in the early eighties. Not sure what help I would be as I am in Michigan currently.
....

The most knowledgeable person I know from that time on IRIS OS would be Mark Lyda. He wrote several assembler calls that then became part of the os. Some of the things he wrote were a call that allowed a program to suspend while another program was run and then return to the original program. Something that Point 4 told us was not possible. 

Star Computer group was buying the mini-computer hardware from Data General and then from point 4. Then as a cheaper alternative they forged a relationship with Microtech. We were buying their computers and they started selling our software packages until the merger became Star Technologies.

Part of the magic of Star*Writer and Star*Calc was the dumb terminals that we added a circuit board to which added 2 pages (screens ) and native word processing capabilities. Recreating the software does not give you Star*Writer or Star*Calc since the program acted as a server to the word processing terminal.

Somewhere I have a newsletter that I will try to find and send to you which talked briefly about the products.

-Charles

....

The program would sit at an input and the terminal would make the request to load a page, save a page, etc. for the processes that needed to happen outside of the terminal.

I am having trouble remembering what the terminal brand was but we put a Star*Writer logo over the terminal brand so if you found one it would be obvious. Also the function key caps were replaced with Help, Load, Save ...

I am thinking Altec but not sure that was the brand. It is funny since we had hundreds of terminals come thru the office that had to be burned in and boards added to but I don't recall the brand. The brand isn't going to help without the added circuit board. I do know that the added code was for a z80 processor so only terminals that had z80 are possibilities. 

Sorry I couldn't be more help. Hope you find one.


-Charles

No comments:

Post a Comment