My first encounter with IRIS was in October 1974, when I went to work for Hagen Systems in Golden Valley, MN (a suburb of Minneapolis). By 1976, I had learned the Nova Assembly language and begun writing drivers and discsubs to support our on-going efforts to provide high quality enterprise systems to our customers.
From 1981 to 1986 I worked on and off as an independent consultant, providing support and operating system level tools to the IRIS community. Some of my most memorable items were:
- Print Spooler driver that would dump everything to a textfile and then print it as an independent task, freeing up the port for other work.
- Trapfault analyzer that would determine the cause of traps.
- Dynamic discsub loader that could hold disc-resident routines in memory to speed up execution.
- Custom drivers that ran cranes in a warehouse from a Basic program.
- Reducing response time on an extremely large system (over 60 terminals) from nearly 2 minutes down to about 3 seconds by rewriting critical system routines.
- Interactive stand-alone disk backup routine that helped prevent inexperienced users from copying over the live system.
- Migrated IRIS to an 8-inch floppy system to see if we could configure a cheaper machine. (Had to disable swapping to keep it from destroying itself !)
- Network driver that connected multiple Nova machines and shared files between them, by working with an engineer who wired the machines together with AppleTalk hardware.
It was the most fun I ever had in the computer industry. IRIS was small enough to reverse engineer major portions of it, yet big enough to support a serious software base.
No one was more disappointed than me when UniBasic arrived on the scene and made IRIS obsolete overnight.