Monday, January 24, 2011

Old days of mainframe, many programmers took labor abuse

Is it right for an employer to order salaried people to work uncompensated overtime? That happened back in 1977 when I worked for Bradford on the NYS Medicaid MMIS.  The people in the Claims Processing  area got behind, and management, with all of its usual tact, made that edict.

I was in the reporting area (MARS and SURS) so we got out of that, and we moved to a different office in midtown to avoid the stress. I liked working in midtown Manhattan better than the lower end anyway (“Life” was halfway in between, in the village).  In designing the reporting, we had to make an executive decision on whether to go “quick and dirty” (process the accumulated claims detail every month end) or summarize, with the risk that an error would propagate forever. We did “quick and dirty” for about three months. Those were the days that COBOL programmers had trouble keeping matched sequential files “’Nsync” (before Justin Timberlake). There was a technique called the “Balanced Line Method”.  Batch COBOL (and ALC) production cycles presented real challenges, and a way of life. You really learned a lot only when something went into production and you had to stand by it. 

I don’t know how people code MARS and SURS today, but they sound like good candidates for SAS.  Get through the DATA step and it’s a piece of cake.

In the 1980s, at a credit reporting company, one of the best ALC mainframe programmers (a real expert on "coding out of addressability") worked "part time" (32 hours per week). When the company came under financial pressure from being taken private, management disallowed part time work and required every one to be full time.  In the 1990s, during another crunch at another company, a 37.5 hour official work week became 40, and Friday afternoon off and "short days" were eliminated. 

Indeed, so much has changed in IT since my coming of age. Or has it?  Programmers, guarding their individualism, resist organizing.

No comments: