Author Topic: 4000 software  (Read 5232 times)

pascal.verdet

  • Guest
Re: 4000 software
« Reply #15 on: February 23, 2014, 03:59:49 PM »
Hi,

I agree with both Jim and Arthur.
My primary motivation is to build a new system with components today, and durable in time !!!
However, we must not forget the past, and try to understand and keep track of what has been done
I think it can help us build a new and nice system.

Cheers,
Pascal

Artur D'Assumpção

  • Moderator
  • Hero Member
  • *****
  • Posts: 663
  • Karma: +22/-0
  • Sintra, Portugal
Re: 4000 software
« Reply #16 on: February 23, 2014, 10:08:45 PM »
I'm sure we can make a system that takes on the heritage and 3 decades of acquired experience and catches the SSL nuances, giving it a modern workflow, with today's needs.

Also the idea is to grab common components and affordable, leaving only the interface electronics to design and then the manufacture process. Something that I believe that within the community we can get it at affordable group prices. It's much more cheaper to manufacture 50 cards, than 5. ;)

Of course we have to be realistic and many prototypes will see the light of day before a final 1.0 version. Also it will be best to take a modular and Agile approach, where we develop small and incremental features, with short releases.

retrocores

  • Newbie
  • *
  • Posts: 17
  • Karma: +10/-0
Re: 4000 software
« Reply #17 on: March 08, 2014, 06:19:18 PM »
I have disassembled the boot sector of the G series program disk image kindly donated by ilcaccillo. I acquired part of the NM4 instruction set from a 4/30 disassembler written in BCPL which I ported to Python. After running the disassembler it was fairly straightforward to work out what was going on, I have attached the output along with my interpretation of what’s happening.

The boot program loads two files, one presumably containing configuration data and the other executable code which it then runs, this file is small (7168 bytes) so must load more program files when it gets going.

I did this mainly as an experiment to see if we had enough information to reverse engineer the program disk, just in case we  need to. I have looked at other code on the disk and it all looks very encouraging - it should be possible to figure out the project file format if necessary.

Kind regards,

Steve

Artur D'Assumpção

  • Moderator
  • Hero Member
  • *****
  • Posts: 663
  • Karma: +22/-0
  • Sintra, Portugal
Re: 4000 software
« Reply #18 on: March 08, 2014, 08:40:23 PM »
Very good work Steve! Impressive! :D

Yes indeed it looks very promising. If you want I have the instruction set manuals for the NM4. I've been diving into the Computer Automation documentation, and supposedly there should exist a BUS debug card that would let you debug the instructions of the CA Computer in live mode. If we could find one of these, it could be a great addition if one needed to debug in live mode certain parts of the code. Have you found any interesting strings that might give you a hint what each part of the code does?

Please add me to facebook (https://www.facebook.com/adassumpcao) so we can talk better about this. I have some ideas regarding the SSLMixed computer that you might like to hear and contribute.

Cheers,

Artur



I have disassembled the boot sector of the G series program disk image kindly donated by ilcaccillo. I acquired part of the NM4 instruction set from a 4/30 disassembler written in BCPL which I ported to Python. After running the disassembler it was fairly straightforward to work out what was going on, I have attached the output along with my interpretation of what’s happening.

The boot program loads two files, one presumably containing configuration data and the other executable code which it then runs, this file is small (7168 bytes) so must load more program files when it gets going.

I did this mainly as an experiment to see if we had enough information to reverse engineer the program disk, just in case we  need to. I have looked at other code on the disk and it all looks very encouraging - it should be possible to figure out the project file format if necessary.

Kind regards,

Steve

retrocores

  • Newbie
  • *
  • Posts: 17
  • Karma: +10/-0
Re: 4000 software
« Reply #19 on: March 08, 2014, 10:31:36 PM »
Thanks Artur

It would be great to see the NM4 instruction set manuals, there are lots of gaps to fill in. Most instructions are fairly easy to work out, others like XNX - index next instruction and CSK - 3 way compare less so.

There are  strings to give clues about what does what, although for now I will carry on disassembling linearly rather than jumping ahead, if people are interested I will post the next installment soon, lets see what the next 7168 bytes do ;)

I have been considering writing a emulator, from what I have seen so far this shouldn't be too time consuming and could be fun.

Cheers,

Steve.

marcmozart

  • Sr. Member
  • ****
  • Posts: 480
  • Karma: +28/-0
    • Frankfurt/Germany
    • Mozart & Friends
Re: 4000 software
« Reply #20 on: March 08, 2014, 10:33:13 PM »
Awesome work!!!
1992 SSL 4048 G-Series
Mix Engineer Blog
http://www.mixedbymarcmozart.com

Artur D'Assumpção

  • Moderator
  • Hero Member
  • *****
  • Posts: 663
  • Karma: +22/-0
  • Sintra, Portugal
Re: 4000 software
« Reply #21 on: March 10, 2014, 12:47:08 AM »
Hi Steve,

I've sent an email to Lars from the Computer Automation museum in hope that he has complete Assembly Manuals for the NM4 processors that he can scan and upload to the online site. I'm still waiting for the reply and these seem a nightmare to find. I have a bunch of tables with instructions, but I believe these do not cover the complete instruction set. I've also tried to find manuals for the Forman CPU, since almost for sure these shared the same instruction set, but these are even more impossible to find. I am also exploring other alternative through other old-time contacts to see if get  my hands on these manuals.

If you can code an emulator that would be very interesting, because we could see exactly how key routines between the console and the computer work exactly and able to debug them.

Keep up the awesome work! :) 

Cheers,

Artur

Artur D'Assumpção

  • Moderator
  • Hero Member
  • *****
  • Posts: 663
  • Karma: +22/-0
  • Sintra, Portugal
Re: 4000 software
« Reply #22 on: March 12, 2014, 10:40:58 AM »
Steve,

I might have hit the jackpot with my google searching skills! ;)

Check this out: http://bitsavers.informatik.uni-stuttgart.de/pdf/computerAutomation/

There are lots of reference manuals, although not for the NM4 familiy, but for the LSI one. I'm sure there is some similarities between the processors and that you might find some useful info on the programming reference manuals.

I've already downloaded them all, if somehow they disappear from there. :)

Check it out and tell me what you think!

Cheers,

Artur

pascal.verdet

  • Guest
Re: 4000 software
« Reply #23 on: March 12, 2014, 06:02:28 PM »
Hi Steve,

Very nice work !!!
If I can help you, or if we could work together on this subject ?

Cheers,
Pascal

retrocores

  • Newbie
  • *
  • Posts: 17
  • Karma: +10/-0
Re: 4000 software
« Reply #24 on: March 12, 2014, 11:52:36 PM »
Hi Artur

good find! Even though most of this information regards the LSI2 and earlier machines there is considerable similarity that will help. There is some NM4 info on the site, the NM4 general information document has a instruction set (but no opcodes), timing and a list of differences between different models. I think Marc may have uploaded this document to the forum a few months ago, it certainly was a great help.

Pascal,

Good plan to work together on this, I will tidy up what I have and send it on (if could PM me your email address), it may take me a few days.



I think now that between us we have the technology to reverse engineer the workings of the computer we need work out the best way to use our combined skills and knowledge to help develop the new computer.

Is there a specific area that needs attention first? I was going to document the interface between the computer and the desk starting from the IO map. At-least with this we would know what to look for in the code, its is easier to search for IO instructions  mapped to a specific device and work backwards.

Out of interest, will the new computer be required to be  compatible with projects / TR setups from on traditional  systems? 

Cheers,

Steve.


Artur D'Assumpção

  • Moderator
  • Hero Member
  • *****
  • Posts: 663
  • Karma: +22/-0
  • Sintra, Portugal
Re: 4000 software
« Reply #25 on: March 13, 2014, 09:17:09 AM »
Hi Steve,

No, the idea is to break compatibility with old projects and totally forget the old OS system. That would be starting the project with a burden and unnecessary effort in my opinion.

What's being thought out, and you can see the discussion that we already had on this subject, is to design a complete new system from the ground up that uses the current console/computer interface for all console elements. Some immediate things I can remember is:

- VCA faders + electronics
- Ultimation faders + electronics (motor control)
- Solo Mute
- LEDs
- Total Recall
- Console modes of operation
...

Basically is to be able to control everything the SSL computer controlled and work with it to build a more modern computer that adapts to modern needs.

I already have some initial hardware prerequisites aligned up but that's all open for discussion and depending on our findings (I myself in the meanwhile found some reservations regarding certain details on the initial designs).

Some important things to discover through the software that might be useful and ease up the reverse engineer on the hardware side is to find and document how this communication is achieved through the several interfaces the SSL computer provides. This includes bit resolution, byte order, byte protocol, addresses, etc. This would help greatly to the task!

If you don't mind, I will ask Mattia to move this topic into the "Research" sub-forum of the "SSL Mixed Computer" forum, since we are now digging more into this subject and getting important info for the SSL Mixed computer project. Also be free to open any new threads on this forum, so we can have everything organized in there and easy to find.

I will also try to provision during the weekend a Wiki so we can start organizing all the info we get on the old computer and the new one. What do you think?

Cheers,

Artur

arisebag

  • Newbie
  • *
  • Posts: 2
  • Karma: +0/-0
Re: 4000 software
« Reply #26 on: March 19, 2014, 07:11:12 PM »
Hi everybody,

I would like to have the knowledge to help you in this wonderful project!

Anyway I am with you as a supporter and maybe by the time I will be able to bring you some help.

Thanks for that.

Ari

Artur D'Assumpção

  • Moderator
  • Hero Member
  • *****
  • Posts: 663
  • Karma: +22/-0
  • Sintra, Portugal
Re: 4000 software
« Reply #27 on: March 21, 2014, 10:27:23 AM »
Hi everybody,

I would like to have the knowledge to help you in this wonderful project!

Anyway I am with you as a supporter and maybe by the time I will be able to bring you some help.

Thanks for that.

Ari

Hi Ari,

Me and Steve have been doing some preliminary background research on this. We may have some design ideas during the next week, something very very promising. The project will be launched in full force as soon as I have my console in the room so I can start reversing the electronics. Everyone is welcome to come aboard, since this is a project from the community and to the community.

I will post more news very soon!

I will also getting ready a Wiki platform for the project so we can start organize all the info, documentation, research, schematics, etc.

Cheers!

Artur




StefanNowak

  • Full Member
  • ***
  • Posts: 216
  • Karma: +5/-0
    • Sydney, Australia.
Re: 4000 software
« Reply #28 on: April 04, 2014, 06:19:23 AM »
Anyone been in touch with these people?

http://www.dyslexiasoft.com/about.html


sintech

  • Global Moderator
  • Hero Member
  • *****
  • Posts: 829
  • Karma: +59/-0
    • Bristol, UK
Re: 4000 software
« Reply #29 on: April 04, 2014, 08:53:34 AM »
Like :)