I have to borrow 2 Transputer cart’s with 2 Tranputters on each cart. I look’s me nice to see if I can connect them together to a netwerk and take a look how a program see this. In my situation it is not always easy to connect a monitor, keybord and mouse to the Window’s 98 PC. Or some time’s I want to work from the living room and the computer stands on the zolder. So I decide to work remote on this PC. But then come’s the “big” problem. Find a program to work remote with a Windows-98 PC. After a lot of searching on internet I found the program Radmin. The old server version of Radmin, version 2.1, run on Windows-98. So i installed it en now I can reach my PC with the selfbuild Transputer inferface connected to 4 transputer processors.
Special for exploring the Transputer netwerk there are some worm-program’s. You load them onto the Transputer network and they give a reslut of the connection’s between the different Transputer processor’s on the network. This iSpy program is original from INMOS, but a lot of people have modified this for tehr one. So found on the website http://www.geekdot.com/category/software/transputer-software/ the TTK, The Transputer Tool Kit. This is a collection with test program’s for the Transputer, also some version’s of iSpy are include. When unzipping en starting the iSpy232 I get a very nice result of my 4 transputer’s connecting to the selfbuild ISA cart. The author of some program’s is Andy, Andy Thanks !!
First i put 4 Transputers in line. This show a nice result. The next network was a tree. One link of the Host-Transputer is connected to the ISA-interface and the other 3 links are connected to 3 different Transputers. This is build as follow.
And give the next result when I start iSpy:
This is a real Transputer netwerk. Oke, now I need a real program to run on it.
On monday 30 Oktober there was a little meeting of the Hackerspace “B+ Westland” at my place. We talk about the result of my work with Transputers, study to a “new purchased” Transputer interface and have a lot of fun.
Next challenge is to understand the OCCAM toolkit workflow. Normally I work with microcontrollers, in that case you don’s have a compiler, linker, configer and a make program. Just compile and load the program in the microcontroller. Now I must follow a different path. I hope I should understand this.
For me this is the -END- of the Retrochallenge 2017/10. I had this month only a spare time for this to reach my goal. So in the future time I go ahead with Transputers to understand the system and how to program them.
I have a lot of fun to work with Transputers and also with the talk’s with the people around me about Transputers.
At an computer meeting of the Compusers group of the HCC (Hobby Computter Club) I meet Albert van der Horts who is a member of the Dutch Forth Usergroup. This Forth Intrested Group have made a version of Forth for the Transputer, called tForth, around the year 1991 – 1995. Albert have done a lot of work on Transputers. On his website: http://home.hccnet.nl/a.w.m.van.der.horst/transputer.html there is a zip-file with al the information, program’s and code to go work with Forth on a Transputer system. I have download the whole stuff and install it on my system with my selfbuild ISA-Transputer interface.
I am not familiar with Forth. It’s not only a computer language but also an development system. And even more then this, it is a whole different way of thinking of working with computers. In fact you maken your one command’s/statemants build from other existing command’s.
I have no experience with Forth but it looks me a nice challenge to try to start tForth on the transputer with my selfbuild interface. I talk to Albert van der Horst and he explain me the way to start the system. So I do, I start the system and now I can program in tForth. This is also working !!
I think it is a long time ago there was a transputer with a running tForth program on it, but the knowledge is stil availble at the Dutch Forth Interest Group’s of the HCC. ( https://forth.hcc.nl/ ) Always nice to visit a regional computer club.
The Transputer system fascinted me very much. After that moment I have contact with this man about the Transputer system. And later he aquire more Transputer board’s and i could borrow some. But my problem was that I have NO interface between the Transputerboard’s and a PC so I decide to build one by myself.
In the handbook of the CSA ISA to Transputerlink interface, Transputer Education Kit User Guide, I found a schema of the CSA-Transputer interface. This circuit around the IMSC012 would I build. Instand of the IMSC012 I bought the IMSC011 IC’s with could have the same function and even more. Together with some TTL-ic’s the project come to a nice print.
I connect my IO-Print to the Transputer and start testing my interface. With the famouse 24 byte [Who-you-are program ?] i get as answer AA AA 00 00 back. This is a 32 bit Transputer !! It is working.
Next step is the Mandelbro program, but when I want to run this at 60 % of the picture the program stop’s.
Each time I start the Mandelbrot I have the same result. Each time !! This point take me a lot of time to find the problem. After a couple week’s a call my friend with the originel CSA-Interface and I could borrow his system to compare this with my selfbuild system. I use the originel CSA IO interface in my IPC system, use the good working program. At the end I start swapping IC’s between the original CSA IO interface and my self build interface. After a lot of debugging and swapping IC’s i found there was aq new 74LS245 withs was wrong ?? At low speed it was working Oke, but at high speed I get a lot of problem’s. Al the 10 new buy pieces were wrong. So i take a look in my storice and found a old 74LS245. Thisa one works very good so I am happy.
Conclusion: even new IC’s could be wrong.
The Transputerboard I use is a Caplan board with 2 T800 Transputers with each 2 Mb memory.
Now I have a selfbuild ISA to Transputer link interface with is working good. Next step is I to understand the propgram’s, how to program and wordt with OCCAM an the Transputer.
As a member of the Dutch Hobby Computer Club, the section Retro Computers (www.retro.hcc.nl sorry, dutch only) I visit almost there meeting’s. Last meeting at 27 September 2017 we talk a lot about the Transputer systems. We planned to go to work with 4 people at the Transputer system’s. Talking, thinking, study, programming and working. So I am not alone.
A couple years ago I visit a meeting of the Hobby Computer Club te Hooglanderveen te Holland (The Netherland’s). This was the Artificiel intelligence user group. An person have a question about a system. He want to work with Transputers. He has a ISA-I/O cart for a PC. This cart was a bridge between a program under DOS to a serial link to a Transputer System. He has problem’s to get his ISA-cart working.
I know that I have at home some “old” PC’s with a ISA-slot in it. That PC’s are running under W98 and DOS. So I told this to hime and we talk together about his problem. He want to run some program’s on a PC, connectes by the ISA-interface to his Transputer board’s.
I took his CSA ISA-interface and a copy of the installation book at home and a couple week’s later I have install the ISA-interface include the necessary software. This was the first time a come across a Transputer system.
A Transputer is a vintage computer system build arount 1985 to 1995, invented by INMOS, an company from Bristol, UK. The Transputer has 4 I/O links for communication between Transputers so you can build big Transputer processor networks to distrubuted your program so your problem can be solved by many processors. So you can split your task in smal part’s. Together with the Transputer hardware INMOS construct the language OCCAM, special for parallel programming.
Around that time this hole system was a big invention and was to call: the poor man’s CRAY. A lot of computerpower for a reasonable prise.
This evening some people of the local hackerspace “B+ Westland” visit my place to look at the 4004 project. We have study some assembly code for the 4004. This was very nice.
We have look at the result of my project and we have also study some other project’s from other people.
This is the end of the RetroChallenge 2016/10. I wish I have more progress the last month, but I am still happy with the result. This week I go to make/finish some little program’s to do the challenge for the Roborama-game’s of the Dutch Computer Club next weekend.
Here is my goal not to win, but to do the challenge.