DEVELOPMENTS AT THE R.A.P.C. COMPUTER CENTRE 1966

The following article appeared in in the 1966 RAPC Corps Journal.

A new computer, an IBM 360, is now in use at the R.A.P.C. Computer Centre. Ordered to replace the elderly mechanical card listing and sorting machines, to increase printing capacity and provide further computer potential, it is so far advanced in design and performance on our older machines that it is being referred to as a third generation model. The big change is the use of micro-circuitry, which has resulted in a very considerable reduction in size of components as compared with our "second generation" transistorised 1401 and our old "first generation steam era" valve type 705s. Purists may dispute these generalised definitions but not the fact that the 360 is more compact, is cheaper to buy and maintain and that it works faster in more sophisticated ways.

By Dave Ross - Flickr: IBM System/360

What does this mean to the layman who, whether he likes it or not, is living in a world where computers are becoming of increasing importance? Internally computers work at such unimaginable speeds that a statement that the new one clips millionths of a second off operations hardly registers. What is of more significance is the very real increase in speed in the partly mechanical peripheral devices. These, operating at speeds thousands of time slower than the computer itself, are normally the main limiting factor in ADP processing. The new printer attached to the 360 is at least 50 per cent. faster than the one we used before. The word printer is, of course, a misnomer as it is really a giant typewriter, not a press. It can type out data at a rate of 1,100 lines a minute. Or in practical terms it can produce three AFs N1559 per second, 180 per minute, but not, unfortunately, 10,800 per hour. There are limits to the length of continuous stationery which can be used and, as the printer appears to eat paper, paper changing time becomes an important factor in processing. In fact, it takes much longer to change paper on the new printer so some nice problems in , comparative economics arise in order to find out when it is "quicker" to use the "slower" printer.

Of particular interest are the disk packs, a new type of direct access storage. Resembling nothing so much as a number of gramophone records spaced on a spindle, a pack consists of 10 surfaces each holding 200 concentric tracks. The total capacity of a disk pack varies according to the size of each block of characters, but in the optimum case it can hold up to seven million characters. In operation, the pack is loaded on to a disk drive and it rotates continuously some 40 times per second. Ten "read heads" on arms can be positioned simultaneously over selected tracks within 75-thousandths of a second. From that point, data is transferred into the computer at a rate of 150,000 digits per second. Not only is this transfer rate 11 times as fast as transfer from a tape unit, but the system allows direct access to a block of data without the laborious task of reading down (possibly) thousands of feet of tape to find the one you want. Fascinating, too, is the electric typewriter attached to the computer to provide direct information on the progress of runs. To obtain a speed three times as fast as that of the average human operator, the conventional system of arms and linkages has been replaced by a small sphere, studded with type, which pirouettes on the end of an electronically controlled stalk, stabbing unerringly at the paper as it shuttles over it.

These are the obvious advantages of the 360 computer. From the programmer's angle the system provides a very much larger number of usable machine instructions which give him far greater scope in developing his logic. This, of course, is a two-sided factor. The programmer now has to know a lot more. In every respect in fact, this machine makes greater demands on the ability and knowledge of its handlers. Some measure of the intricacy of the computer can be obtained from the fact that at the latest count some 180 different instruction manuals have been published, each a massive volume written in the strange vernacular of the computer world. One of our very first tasks was to compile a compendium of terms so that we could converse intelligibly amongst ourselves. This was a job which would have challenged the ingenuity of Doctor Johnson himself. In the event it was not, perhaps, surprising when certain programmers said that this was all they needed-now they were completely confused! In saying this they were somewhat less than fair. For example, what could be more lucid than (quoting one definition from our compendium) :

" D Sect (Dummy Section)-This is a means of referring to and using an area of storage not defined in the problem program but defined in some other assembly. It represents the control section that is assembled with the problem program but which does not reserve storage in that program."

Well, perhaps you're right. But few can deny the aptness of the definition of " Problem Program-the program written by the user."

Despite the complications we are making good use of the new machinery. Already most of the AFs N1558 and AFs N1559 now being sent out to Regimental Paymasters and Units are printed by it. There IS a lot to learn as we move forward to use it for more intricate processes, but we cannot learn faster than the manufacturers who are themselves still struggling with the manifold problems laid on them by their own technicians. We may occasionally hanker back .to good old days when we had only our steady and, In retrospect, simple 705s to deal with but we have no option but to press onwards. In the computer world one must do this or fall by the wayside ! Already looming ahead are bigger, better and, without a doubt, more intricate machines, which the next generation of programmers will have to master.

If you are contemplating working in the Computer Centre do not let these remarks put you off. The work is interesting, challenging and exciting. It is rarely dull. As in a newspaper office, we work to deadlines. On the operations side there is a continual struggle to keep the machines fed, to organise and reorganise production to get output off in time. On the programming side there is the relentless pressure to prepare programs and test their accuracy by target dates. However, the acquired, experience and skills of six years of running are having their effect. Our motto, "Non sine sudore et lacrimis" (Ed: "Not without sweat and tears"), is unchanged but no longer does the great rallying cry of the early days ring through the building when all seems lost. Yet sometimes on the night shift, when the console lights twinkle and the only humans present are the white-coated acolytes tending the machines, then one may imagine the ghostly shades of the first programmers whispering it down the deserted corridors - "Never mind, chaps, there's all of Sunday night left!"