|
| KenAdam |
UserPreferences |
| Edinburgh Computer History | FrontPage | RecentChanges | TitleIndex | WordIndex | SiteNavigation | HelpContents | moin.sf.net |
I was the first student (along with Francis Lam Shin Cheung) to complete the joint honours in CS and Elect Eng in 1976, and joined Ferranti after graduating. What a surprise to find that they were writing code for fighter aircraft navigation systems in OCTAL!
I was employed as a Digital Design Engineer, but wrote a 2 pass assembler so I could write test programmes faster. With only paper tape, it was a case of loading the first pass, loading the source, punching the intermediate, loading the second pass, loading the intermediate and finally punching the load file.
As you can imagine, there was a complicated "bootstrap" process here till I got it to be self sufficient.
Eventually I persuaded Ferranti to pay Rainer Thonnes and Con Bradley to help me write a Second Pass for Peter Robertson's portable IMP compiler, and then wrote an operating system in IMP, using Tandberg DC300 cartridges, then 8" floppies, then 5 1/4" floppies (another bootstrap process!).
This system was in use for many years as FIST (Ferranti Inertial Systems Test) and was sold in various countries.
Along the way, we wrote a Coral 66 compiler in IMP, and then re-wrote it in Coral so that it could compile itself. Eventually we moved to VAX hosts, and modifed the compiler to compile in VAX Coral (proably one of the least used compilers ever written - eventually we were only the DEC customers still using the compiler (interstgin when they produced new VMS versions).
Earlier this year I was asked to look at a problem withe the Link-editor tool (another IMP to Coral translation) on the VAX. I had written the code in the early 1980's and last modified it in 1986 (according to the config control system (CMS)). Seems that while it worked on a 400 Mb disk, it didn't like the larger clusters on the 3Gb disk they had just installed (screwed up mapping the file into a memory block when the cluster size changed - there were no clusters when the code was written!)
More to my amazement than anyone else's, I fixed the problem in a day (I hadn't logged in to a VAX in several years).
The CS department at EU from the 70's has a lot to answer for!
Does anyone remember playing Snooker on the type 340 display of the PDP-7? That was written by Rainer and I after JPG (John Gray) asked us one night why we didn't make it do something useful. Took two weeks to write (OS, software and graphics in 8K of memory) and was played for about 2 months before the PSU went bang.
These days I design Digital Map displays for air and land vehicles. (takes a bit more than 8K, but I have now got it ported onto my phone)
