My first years in the development department were still as a seconded employee from EDP or the Systems Department, as it was called then, and I still had a permanent position and some tasks for that department until they switched from Univac to IBM as a supplier of mainframe computers. Univac supported a user association called Univac Users Association/Europe (UUA/E), of which B&O was a member, and I was their representative at the biannual meetings in various European countries. B&O mostly used Cobol as a programming language for administrative computing, and I gave a number of talks at the UUA/E meetings and published related papers on various Cobol topics. For a few years I was also on the board of the UUA/E.
Univac Denmark at one time arranged a lecture by Grace Murray Hopper, one of the pioneers of programming, and instrumental in the birth of the programming language Cobol. One of the themes of her talk was that computers should become much smaller, as long signal lines put significant limits on computing speed. She was known to hand out ‘nanoseconds’ at her lectures, as she did on this occasion. Her nanoseconds were a length of signal wire 30cm long, equivalent to the distance light or electricity could travel in a single nanosecond.
Some years later I was at another presentation where a young team of engineers were going through their proposal for a wired industrial remote control. The connection between the controller and the remote could be up to 30 metres long, and their intention was to send a signal from the remote to the controller and receive an acknowledge response back within 5 nanoseconds. I then asked if they knew how long a nanosecond was. They looked askance at me, and didn’t know, but wouldn’t believe it when I held up two fingers about 30 cm apart. Only after checking it out on a calculator did they have to admit that they probably couldn’t expect a response within the expected 5 nanoseconds and that the travel time in each direction would be at least 100 nanoseconds.
In their spare time there were also some evening classes to be attended. During my time in the computer department, I took a Merkonom course in data processing. They ran a one year trial in Ringkøbing with both year 2 and year 3 cousrse levels combined every Saturday morning. I joined the course and became a merkonom. This was before speed limits were introduced on Danish roads, so the drive from Struer to Ringkøbing could often be done at a reasonably fast pace. In winter, however, it was with caution. A few years later, I also became a merkonom in accountancy.
To learn more about electronics, I took an evening class course in Holstebro to get a license as a radio amateur. Understandably, there were quite a few employees in the product development department who were serious radio amateurs. Denmark had introduced a new category (category D) of amateur licensees, where you were quite limited in frequency ranges and transmission power, but on the other hand there was no requirement for proficiency in Morse code at a given speed. I was licensed as OZ1IQL, but my time as an active amateur did not last very long.
The bookshops in Struer and the surrounding area did not have a large selection of English-language books, either fiction or non-fiction. When I traveled to Copenhagen on rare occasions and had some time to spare, I always visited Politikens Boghandel on Rådhuspladsen, as they had a reasonable selection of books in English. I don’t remember how, but I found that there was a bookseller in a small Jutland town (possibly Bramming, near Esbjerg) who traveled a few times a year to England to buy books for resale. The selling price was somewhat lower and the selection much larger than in the regular bookshops. You could place an order list in advance, and the books would be sent to you after they had been purchased in England. I took advantage of this a number of times.
Another source of books was The Scientific Book Club, run by Foyles in London, which was probably the largest bookshop in the world at the time. They published a special edition of a popular science book each month, and you got it sent straight from the shop. I was a member for a number of years. Not all the books were of interest, but I did read many of them. I think the book club closed about the time I left Denmark. In any case, I did not continue my membership. During a visit to London I spent some hours in the bookshop itself.
The first personal computer I had a chance to play with, I think in 1978, was a Commodore PET 2001. It was a single unit with monitor, keyboard and cassette recorder. The screen was monochrome and displayed 25 lines with 40 characters in each, and the keyboard had small square keys arranged in vertical columns. There was a built-in Basic interpreter. The computer had been taken home on loan by a colleague in B&O’s systems department, Martin Bech. We sat down together and typed and ran John Horton Conway’s Life (or Game of Life) from Byte magazine. The keyboard and small screen were the main reasons I was not interested in purchasing one of these PET computers.
Before long I acquired a Tandy/Radio Shack TRS-80 Model I instead, which also used cassette tapes but had a normal keyboard and a separate screen that was a converted monochrome TV receiver. Internally this computer used both upper and lower case letters, but on screen they were all displayed as upper case. A simple hardware modification made it possible to display both upper and lower case letters, although the letters g, j, p, q and y looked a little odd as they were raised up compared to the other letters due to a limited number of scan lines. There were also no Danish letters in the character set.
Programming was my job but after acquiring the TRS-80 personal computer in 1979 it also became a hobby. Ib Hald and S K Pramanik also acquired TRS-80 computers, so we had a small group to exchange information, experiences and programs. As personal computers became cheaper and more common, a number of employees acquired computers of different makes. In addition to various small programs, in my spare time I also wrote a major word processing program for the TRS-80, which was sold commercially in the US under the name Typitall. It was a large program, written in Z80 assembler language, with several overlays for e.g. spell checking to leave as much space as possible in memory for document text. First, however, I had to modify the assembler so that it could translate a program that consisted of several source files. The program was written to be compatible with Tandy’s (the company behind TRS-80) own program, Scripsit. The keystrokes and data formats were the same, so users could easily switch from Scripsit to the Typitall program.
It wasn’t long before we each had our TRS-80s upgraded with extra memory, floppy drives and printers. The first floppy disks were 5.25″ with a capacity of about 80 kilobytes in the 35-track versions. Later they were expanded to 40-track and 80-track and double-sided drives. A receipt from September 1980 shows that two assembled disk drives for the TRS-80, bought from England by B&O’s purchasing department, cost DKK 4,484.31. It was not cheap to be on the cutting edge of personal computers.
Around 1980, Epson launched a small dot-matrix printer, the MX-80, which was cheap by the standards of the time and very reliable, with good print quality compared to the competition. It was controlled by an 8048/8049-compatible processor with external program memory, so I disassembled the program and made some modifications, including a Danish character set. An article about this in an American magazine resulted in contact with a couple in Alaska who needed an Inuit character set for a newsletter they were publishing. I made it and sent them a copy.
For a number of years I subscribed to a whole string of computer magazines such as Byte, Dr Dobb’s Journal, 80-microcomputing, PC Magazine, Computerworld and several others. I also had articles and letters to the editor published occasionally in some of the magazines. It was an exciting time with the early personal computers, and the news mostly came from the USA. There was a small Danish magazine, I think it was called Databit or something similar, which was published for a shorter period.
An apprentice, Jørgen Due, was very interested in personal computers and visited me sometimes to play with the TRS-80. On one occasion he had seen a slot machine with a game whose operation he explained. I later found out it was called Breakout. During the visit I programmed a simple Breakout game in Basic, impressing him greatly that it could be done in such a short time.
Another early TRS-80 user was Peter Bang, grandson of the co-founder of B&O, who later co-founded Navision, which was sold to Microsoft in 2002.
The first IBM PC was launched in August 1981, but didn’t arrive in Denmark until the following year, when I acquired one that became a new pastime with programming. IBM Personal Computer Store prices dated January 1, 1984 indicate prices, exclusive of MOMS, for both hardware and software. Some examples: 8087 co-processor kr. 2.984; 64K memory kit kr. 2.165; 160 KB drive kr. 2.696; 320 KB drive kr. 4.926; 10 MB disk drive kr. 19.929; keyboard DK kr. 2920.
When I left B&O I sold my IBM PC to Poul Nielsen, who later moved to Simtec where he took over the production of sample prints from B&O. With some purchased software, which by today’s standards was very primitive, and a plotter, the PC was used to design prints for the first B&O/JTAS telephone.
It was not long before personal computers were acquired for many different purposes at B&O. Most of them BBC Micro, IBM PC or the more or less compatible ones that came on the market in the years after. Many new programs came onto the market, such as Visicalc and 1-2-3, word processors such as WordStar and WordPerfect, databases such as dBase II, and compilers for various languages.
Shortly before I met his father at Motorola, I had met Anders Hejlsberg at an exhibition in Copenhagen and had bought a copy of his Compas Pascal compiler for IBM PC. A few years later Anders moved to the US, where he worked for Borland on Turbo Pascal and Delphi, before joining Microsoft and being one of the key people behind C# (C sharp), .net and several other programming languages and tools.
Another project in my spare time was a program for an AIM-65, a small computer with a 6502 processor, to accurately measure the water level at some West Jutland harbours. The equipment was developed by the electronics company Gunnar Larsen in Struer. It was designed to give warnings of storm surges on the west coast. Current water levels were sent via modem and telephone line to a central monitoring centre.
The author of the book Microcomputer Architecture and Programming, John F. Wakerly, was so convinced that his students at Stanford University had found all the errors in the book that he offered a $3 reward to the first person to find and report a new bug. I don’t remember how many there were, but I did find quite a few.
Next chapter: B&O Telecom
Microprocessors in B&O consumer products 