27 January 2013


My first day working at Interactive Applications Ltd witnessed a miracle - or at least a marvel: I managed to stay (more or less) awake all day.

Our flight had arrived at 10:30 the night before.  I think we were asleep by about 1:30AM - and awake at 6:30.  Ross drove the two of us to the office - somewhere about here.  Ian Hardcastle was my supervisor - and he was not at there.  He was visiting a customer - I think it was called "Auckland Byproducts" and was, in fact, a company that processed the leftovers from slaughtering beasts and processed them, somehow - perhaps into glue.  Ian and I later referred to it as "The Shambles."  He had, however, left a very warm welcome note on my desk (written in marking pen on a sheet of fanfold paper) - and had left, also, a copy of a manual for Microsoft BASIC there.

I had never before that seen a computer with the DOS Operating System - the precursor to modern Windows.  All the computing I had done had been in a sort of ancestor of DOS: CP/M.

I recall that first day vividly.  I think I probably read through the first three or four pages of that manual at least a dozen times - retaining, in fact, nothing.  DOS was close enough to CP/M that I was able to start up a computer (booting, of course, off 5.25" floppies.  DOS was IAL's principal growing market, although much of our installed base was still CP/M (many of which machines booted from 8" floppies).  New machines in the office, though, were DOS - and there were even one or two IBM XTs, with 10MB hard discs - and during my time there, we acquired an early IBM AT.  This latter was an amazing machine.  It had a 20MB hard disc, 512KB of RAM, and a 6MHz 80286 CPU (the XT and the floppy-only PCs had 4.77MHz 8086 CPUs and 64KB RAM - though ours were mostly beefed up to 128Kb).

Within a day or two I was beginning actually to work.

Our software was all written in BASIC - either a CP/M version or a DOS version.  Various actions required the computer to search through collections of information - and even although the collections in question were small by today's standards, the machines were slow.  Searching for an item in a list is tedious.  You have to examine each item in the list in turn.  If the item is the last in the list, you have checked each item for each search.

Our software used a method for storing searchable collections called a B-Tree.  In effect, looking for an item you can first decide if it is in the first or second half of the list; then in the first or second half of the resulting sub-list; and so forth.  This means your search time is much less (for those who care, linear searching of N items takes N lookups; B-tree searching can be done on the order of log-to-the-base-2 of N.  1024 items takes, on average, 10 lookups instead of 1024).

The trick was that the B-tree software was not part of BASIC.  It lived in a special part of memory - and BASIC had to be able to talk to it.  My principal job, for quite some time, was writing the bits of programme that connected BASIC to B-Tree.

I worked at IAL from that day - 21 May, 1984 - through to my last day - 29 November, 1985.  I had a wonderful time there.  I learned a lot more of actual practical (as opposed to theoretical) computing.  I made one friend who is a dear friend to this day - the same Ian Hardcastle.  IAL was fun.  The people were very good to me.  Every now and then I encounter one or another of them.

When we moved to Pukekohe that May, it was our and the Jacksons' intention that we should build a house for joint dwelling of the two families on Ross's property in Glenbrook.  I do not know precisely at what point it became clear that this would not do, but it may have had not a little to do with the - unsurprising! - stresses created by the two large families living together in a house meant for one.

No comments: