Eleven years is a long time to be in the same job. Could I perform as part of a team? Would my coding skills be up to scratch? And what was the beer like in Manchester?
The last few weeks of my time at the brewery should have been a period of handover, but it didn’t really happen. There was no-one to hand over to! And, of course, it was a brewery; every day was a party!
The excitement of my first few days at the brewery were in stark contrast to the same period at the fashion retailer. There was a lot of new stuff to pick up, but the learning method was to read it off green screens. Tedious! There were a few strange rules too, such as no hot food at the desk, and no oranges either! Although my job title was Senior Programmer I still had to pass the basic tests before I was let loose on actual systems. This was done as quickly as process would allow, and I was allocated a place in the warehouse team. This team looked after all of the distribution systems, which were just about to be rewritten in a new language. I got my first specification, and I was appalled. A well-meaning analyst had written it in what can only be described as beyond pseudo-code. She might as well have coded the program herself. I put a stop to that by letting it be known that if this was analysis I was proficient in that too, so I became one of their few analyst / programmers.
This was a much better situation, and when we were moved from central Manchester to the mill town of Shaw, near Oldham, with full overtime and expenses, things were going very well. The IBM technologies were not so different to the Univac ones, with tweaks to JCL, utilities such as SORT, CICS for online processing, and ISPF as my editor being the only changes. However, it wasn’t as simple to get reports or program listings. These resided in the hallowed operations room, and they were brought down once a day. If you wanted them more often you had to go and fetch them, which involved knocking on the door until someone let you in. I became quite used to having to stroke a few egos to get what I wanted, as DBAs, systems programmers, operators and other ‘experts’ on the non-development side of the fence seemed to reside in their very own, and I’m sure well-deserved, ivory tower.
Anyway, the warehouse development was brand new, and I was just about to learn Mantis, a 4GL that seemed to bypass COBOL in its evolution. Other skills followed, such as Easytrieve, but nothing that was very challenging. There was also the assembler debuggers, Oliver and Simon. These tools allowed a programmer to step through the code and even change instructions mid-execution. I asked the internal teacher how it dealt with different length instructions, and he looked at me quite blank. I didn’t ask any more ‘hard’ questions, and researched them myself instead!
The warehouse development was in full swing until a JCB fell through the floor of the mill, setting back the infrastructure work. We had a team of contract Mantis programmers, I was responsible for all of the batch interfaces, and we were well ahead. Then, gradually, things began to slip. It was clear that the mill would not be ready for ‘go-live’, but it was also clear that the software would not be ready either. But who would blink first? If we could produce just one picking note we would be absolved and the infrastructure team would cop for all of the blame. There was a simple solution, the chap who interviewed me, and who had been seconded to lead the warehouse project, asked me to print a picking note. Not program a picking note you understand! He literally wanted me to send a print file to a printer to cross the finish line. It was above my pay grade to argue, so I did it. It was no big deal either. Eventually both sides got their house in order and the project was a huge success, until the Mantis started to cause all kinds of performance issues, but more of that elsewhere.
Next I moved back to Manchester to concentrate on integrating the new with the old. Now, there were some very clever people in the IT department, including the country’s foremost expert on the Supra database. However, there were also a lot of Assembler programmers who tend to do, well, what they can. It does not matter if a task can be completed with a simple subset of six instructions. If an Assembler programmer can use dozens of instructions and play with registers beyond the comprehension of a merely average programmer, then that is what they will do. Why? Because they can. I cut through as much of this crap as I could, but I was now starting to wonder whether the IBM experience that I had gained at the warehouse, together with my new-found knowledge of the contracting profession gleaned from several trips to the pub with the Mantis contractors, would be enough to enter the world of the freelance.
What happened next was quite unexpected. The fashion retailer decided to move into the world of CDs, videos and pre-school education packs. They purchased a company from Wellingborough whose entire system was written in COBOL. There was only one person in the entire organisation that understood COBOL, me! I was put in charge of their systems and their transition, and after several early morning trips down the M1 it was decided to move their machine up to Manchester. This was a simple enough task, and their computer department, that consisted of three people, came with it. I was in charge, so at the annual results presentation at Granada Studios I made sure they took in the whole tour, including a beer in the Rovers Return. Actually, it wasn’t in the Rovers, and it wasn’t just a beer. We were quite inebriated when we returned to the office to pick up our bags! I was quite used to operating in a small environment such as this, and I then built my own team with a view to migrating their data and functionality to the IBM mainframe.
This was not without its challenges, but we were given our own dedicated war office, and we adorned the walls with various charts and our ‘implement-ometer’, the kind of pictorial progress diagram that you see on the sides of churches that are raising funds for a new roof. We ran the implementation over a weekend, and it was clear by Saturday lunchtime that we would overrun. By Sunday evening that overrun had turned into a couple of days – the heavy processing of the regular batch suite absorbing all available machine resources. By the end of Tuesday it was complete, and the fashion retailer’s call centre was able to handle calls about CDs etc. It was not a great success, but as so often happens we were driven by deadlines. We had removed a significant licensing fee, and achieved more than 90% success in terms of customer satisfaction, at the cost of a few disgruntled video customers. I’ll take that!
The success was tangible enough for the retailer to buy another video and CD company, based in London. I went down to see their systems, and came back with my report together with an illicit customer extract so that we could try to ascertain how many of their customers were already on our books. Of course, this was still several years before the Data Protection Act, so this kind of industrial espionage went unchallenged! I came back proud of my findings, and I was called before the board. “There is no way I would buy this company,” I announced. “The systems are a nightmare, the Pick operating system is outdated and unfamiliar to almost everyone, and there is evidence of a fair amount of duplication on the database.” The IT Director (before such people were called CIO), looked at me and announced, “Too late, we already have. And we have a month to transfer their systems across to ours.”
This was a real mess, but again it was above my pay grade to argue. We did it, but I reckon the success rate was around 70%! There was a very interesting finding when I discovered two accounts with the same primary key. The business consultant looked perplexed. I called the company. The operations lady explained that they do indeed duplicate account numbers, some are for CD purchasers and some are for video. I asked how they knew the difference, and she replied that they know their customers, so if Mrs Smith rings up, for instance, they automatically know that she is a video customer. The business consultant scratched his head. I explained that unless we can embed Artificial Intelligence into our process that would be impossible to code for. We jumped on a train and headed for London, returning with a solution that was far from perfect, but then again, I did warn them not to buy the company.
The next challenge was to build complementary CICS systems for the CD, video and pre-school education sales. The functionality was quite different to the existing credit accounts, and I persuaded the management to allow the development to be done in COBOL, as that is what the original people understood. The development was going well, but it had become stale. My feet were twitchy again, and the lure of contracting beckoned. I travelled down to Plymouth for an interview. That is a long way. It was Supra and Assembler and I passed the exam with flying colours. It was significantly more than I was on at that time, and despite the travelling I had made up my mind. However, the retail company were less than pleased. They understood my motivation, but they wanted me to stay, and they offered to match my contract rate purely as a temporary contractor. I, of course, accepted. The same money for a Manchester based job as opposed to a several hour train journey and nights in a hotel – no contest. I told my colleagues about the deal, and they were in awe. Unfortunately the management put out an announcement stating that I was not being paid contract rates, I merely had the status of a contractor. A complete fabrication but if that is what they felt they had to say who was I to argue?
It was summer 1993, and I was in the wonderful position of taking my first contract at a place I knew well. It lasted for fifteen months, during which time I relinquished my management responsibility and took on many more technical challenges. I treated myself to my very own PC. It cost about £1800, and it was a 286. Imagine what I could get today for that kind of money. I also bought myself a copy of Microfocus COBOL, complete with dongle, the kind that fitted into the parallel port at the back of your PC. For those of you wondering, the parallel port generally connected the printer, the dongle was an intermediate bit of kit that could have strange effects on the printing operation, but without it you could not compile your Microfocus code. The building that we were in was converted from being a clothing manufacturer. A friend of mine still had a clothing company in another part of the building, and he needed some software writing. I had Microfocus COBOL. Its development and debugging capabilities were so much better than the mainframe, and within a month of working occasional weekends and evenings his system was built, including stock control, invoices, and various other pieces of functionality. The Microfocus COBOL had already paid for itself, and there were plenty of enhancements to the system over the following months. However, in the autumn of 1994 my contract with the fashion retailer was reluctantly terminated. My skills were now in great demand, I had learnt to drive, and I found a new position before my notice period in Manchester was up.