In my bazqux feed-reader, I have a feed from Hacker News
They post lots of interesting stuff seemingly randomly-picked from all over the WWW and from different time-periods.
They had this today: Asshole driven development (2007)
The author of the article lists a few ADD approaches, but in the comments people have listed maybe a hundred or so more - from their own experiences - and I was very impressed with both the quality and the humour of the descriptions. Many of the ADD methods described I too have personally experienced in real life - they were usually notable in that they contributed in some way to project delay/failure.
One of the ADD methods listed is something like "CYAD"
(Cover Your Ass Development) and that is one which, from memory, I recall was identified in the programming for the development of a nationwide ATM (Automated Teller Machine) network project which was using IBM hardware (ATMs and dumb 3640 network switches) and IBM CICS (a transaction server OS language), driven by an obscure proprietary IBM programming tool/protocol with the acronym CTAM
(Customer Transaction Access Method
), interfacing with IBM host mainframes. However, the analysts working on the project amusingly used the acronym CTAM
as code for Cover The Ass Method
, because that was what they observed had been going on.
The project was closed down and restarted with a "refreshed" (i.e., new) design team and the design objectives were expanded
to include future-proofing the system with full EFT-POS functionality at the outset. The network was re-designed to be non-proprietary and hardware-agnostic (any manufacturer brand of terminal hardware could connect to it), utilising any old ATMs and replacing the dumb 3640 switches with Tandem Non-Stop computers as distributed intelligence/data network processing nodes, using ACI Base24 applications.
It was super-political. I recall the initial designers who had been unceremoniously dumped from the project kept lobbing grenades into the refreshed project, writing long reports as to why the new design was infeasible - it simply couldn't
work, etc.. You'd think they worked for IBM (which I suppose they did, in a way).They only went quiet after the network went online and fully operational, as scheduled. Funny that. Could they have been wrong?