Part 2
BM: Which part of A-Track was the most difficult to develop? Was it the base unit, the handheld controller system, the software or something else?
TC: Both software and the Command Station hardware were fairly straightforward to develop, although the software took most of the time and, because the size of the application was far larger than could be handled by the 800XL editors, it was necessary to write and edit the code on a PC before transferring it back to the 800XL for assembly. The process is described on my website -
---
Using SpartaDOS (my preferred operating system for the Atari Classic) it is possible to set up SIO2PC 'super' disks on the PC with a capacity of 1MB - which was essential for the A-Track project where the source files amount to around 500 KB and the Assembler list file alone occupies in excess of 650 KB. I have to admit that files of this size are totally beyond the capabilities of any of the Atari text editors, so I actually used the PC to generate and edit all of the Assembler source code. This was then transferred to the Atari 800XL via an RS232 comms link, using a P:R: Connection and BobTerm at the Atari end to handle the communications.
BobTerm, written by Bob Puff of CSS, is by far the best comms program for the Atari Classic machines and translates the PC's CR/LF ASCII characters to Atari 'End-of-Line' characters 'on the fly'. After a little more manipulation on the 800XL, the translated source file (held on an SIO2PC 'super' disk in the PC) can then be accessed by the Assembler/Editor cartridge and assembled to produce the required machine-code object program and the output list file. These are output directly to another SIO2PC 'super' disk file on the PC, since there is no way the very large files involved could be held within the 800XL's 64 KByte memory, or on a real Atari disk drive.
Although it all sounds very complex, it is relatively easy to use in practice - and very satisfying to utilise the PC as an Atari peripheral!
---
Again because of the size of the application, debugging was beyond the capabilities of the standard tools, so I had to write my own debugger and embed it within the code (see the assembler listing for details).
BM: And what about hardware? Were there any problems developing the A-Track hardware?
TC: The only real difficulties encountered were with the handheld controllers (HHCs). The filtering fitted internally to the joystick ports meant that it was impossible to drive the length of cable required with more that one HHC.
It was then necessary to develop some interface hardware (the Network Driver) to connect between a full set of HHCs and the 800XL joystick port. Once that was done, there was a lot of experimentation required to match the HHC and Network Driver electronics to the required cable lengths in various network configurations, and ensure that spurious electrical signals were reduced to an acceptable level.
BM: You use words like "were developed", "was written" etc. Does it mean you developed and you wrote? Or were there others helping you?
TC: Yes - I developed A-Track, wrote every line of code, and designed and developed the related hardware (DIU Command Station and Handheld Controllers) completely on my own. The group in the USA, for whom I was developing A-Track acted solely as customers, answering questions and providing feedback after reviewing my proposals and prototype versions.
BM: Where were your customers from? In your previous answers you mentioned at least one A-Track system running in the US.
TC: A-Track systems were only supplied to Decker McAllister in Long Beach, California and to Charles Cole of the Cochise & Western Model Railroad Club in Sierra Vista, Arizona - where they were in operation for about 8 years (1999 - 2007). You are the only other user with an operational system - and the only one currently, since the US systems were "retired" at the end of 2007.
BM: How many Atari A-Track systems did you make in total?
TC: As far as I can determine, I built hardware for about five systems, including booster units (base units without a direct interface to the Atari 800XL). The units were sold for the cost of components plus shipping.
BM: It's hard to believe that such a great thing like A-Track did not find more users. So there really were only 5 units produced?
TC: I built two systems for my own use, two systems for the group Long Beach, California, and one system for the Cochise & Western Model Railroad Club in Sierra Vista, Arizona. Each system consisted of a DIU, one or two Boosters (PBU) and a set of Handheld Controllers (HHC). The groups paid me the cost of the materials used and for shipping the units to the USA.
I did not advertise A-Track as a commercial product because I would then have to provide full customer support - not something I was prepared to do (and ruin my hobby!) since I had some experience of sorting out hardware problems with the initial systems I shipped to the USA.
BM: I understand. But, if A-Train Systems was not your business, what was your job?
TC: A-Track is totally a hobby - and never a source of income. I am (or was before retiring) a professional electronics engineer specialising in the design and development of large real-time computer systems, and spent most of my career managing a large design team in the UK defence industry, developing sonar systems for the Royal Navy.
BM: Interesting! We can make whole interview on this topic. But back to computers. When did you start to work with computers and which one was your first (at home)?
TC: I have been working with computers since about 1963, although I had an interest in them since the late 1950s while still at school. Initially, it was with mainframes, and then later with minicomputers such as the DEC PDP11. Professionally, I started working with microprocessors around 1975, including the design and development of custom processors for Burroughs Corporation. My first home computer was a Sinclair ZX81 in about 1983 - which introduced me to the complexities of Z80 assembler programming.
BM: When and why did you buy your first Atari?
TC: Sometime around 1985 or 1986 when one of the major retailers in the UK were selling off their surplus stock of Atari 800XLs cheaply. I actually bought the system with the idea of adapting the included 1050 disk drive for use with my ZX81 hardware for which I had already developed a printer interface.
BM: What did you use your Atari for? Which peripherals did you have and use?
TC: When I discovered how much easier it was to write both Basic and Assembler software on the Atari compared to the ZX81, I moved over entirely to the 800XL. My interest was in developing utility code, such as handlers for printers and a mouse, and I used the Atari with a modem (and a CompuServe account) to access Atari-oriented bulletin boards before the Internet was properly established.
I collected several Atari 800XLs, a 600XL, and a 130XE, as well as an older Atari 400, together with a set of 1050 disk drives, cassette recorder, a 1020 printer, and a trackball, together with the SpartaDOS P:R: interface and a few other bits and pieces from third-party suppliers.
BM: How popular were Atari computers in UK (in 80s and 90s) compared to other platforms? Did you have Atari friends in UK? Did something like Atari clubs (known in central and eastern Europe) exist in UK?
TC: Not particularly popular, compared to the Sinclair Spectrum, for example. There were a couple of magazines to support the Atari (Page 6 and Atari User - which later merged) but I am not aware of any significant clubs for the Atari being set up in the UK. Personally, I did not know of anyone else who used an Atari computer.
I was aware of groups in Poland, for example, who were very enthusiastic about Atari computers, but their main interests in developing games or very clever graphics did not align with my own.
BM: What were your other Atari projects (besides A-Track)? I.e. what else did you created/developed on/for Atari?
TC: Mainly, as mentioned above, in developing support software. I wrote a couple of articles for the Atari magazines on implementing a mouse driver and how it could be used.
BM: Did you also use your Atari (from time to time) for playing games? If yes, which were your favourite?
TC: Not really - apart from brief sessions with Space Invaders, Breakout, and the like. I was also quite interested in adventure games similar to the original Colossal Cave (which I had played on mainframes), and collected quite a few of the Infocom range, although I did not spend a lot of time playing them.
BM: Why did you choose PC/Windows as the platform for A-Track successor? Why not, for example, Atari ST (TOS) or Linux?
TC: By the time the Atari machines were becoming difficult to support (2002-3) it was obvious that the IBM PC was the principal machine in use throughout the world. The Atari ST had failed to make any great impact (even in the music area) and Linux was in its infancy (seen at the time as a curiosity for amateur computer hackers to play with). PC/Windows was (and is) also an open system (although not quite in the sense of Open Software) so that anyone had access to all levels of the operating system, and it was fully documented by Microsoft - so, at that time, there was no contest.
BM: And the last question. If there was some demand for Atari A-Track these days, would you be willing to produce some more pieces? If not, do you have some spare parts (PCBs, the metal boxes, ...) which you could sell to help an interested person to build his own new A-Track unit?
TC: I am not really interested in building any more hardware for the Atari - I have several hardware project in development at the moment, related to model railroad layout control (and linked to the current Windows A-Track), and there is only so much time available for hobbies.
Unfortunately, I don't have any PCBs or other hardware available - when A-Track was developed, 20 years ago, PCB manufacture was a lot more expensive than it is today, and I only purchased enough to build the systems required - you have the only spare unit that I had - together with all of the circuits and documentation. If you or someone else wants to build a system, then I am happy to answer any technical queries, or try to supply any details that are missing.
BM: Thank you, Terry, for your time and all the interesting answers. But the best thanks to you for developing the whole A-Track system.
Bohdan Milar