I enjoyed the account by a real Lisa power user on PDF pages 48-53 then 57 of the
Fall 1986 report.
My typical analyses involved the computing of median soil erosion factors and erosion rates for many different kinds of land use, soil types, and geographical areas. Each analysis took between 5 and 60 hours of processing time, and I chained them so that a new analysis would start as soon as the previous one finished. While these jobs were running at lowest priority in the "background," I used high-priority interactive processes to screen results, program and test new analyses, and write reports. Thus, the Lisa was running around the clock at close to 100 percent of its CPU capacity for nine months, except for frequent power failures which sometimes shut the system down for a few hours when I was not present. [...]
Power supply in the semi-rural neighborhood was very unreliable. The power interruptions that occurred every five to seven days on the average cost me much time and were especially frustrating when they occurred at a time when a background job running for more than a day was near completion.
(This is all work being done under UniPlus UNIX.)
