Quote:
Originally Posted by kacir
I would bet the programmers were forced to adhere to strict specification put together by some committee where non-programmers and non-technicians vetoed LOTS of interesting suggestions.
|
There's an old story of the days when MS and IBM programmers were working on OS/2 code (circa 1990, before the big split) where the MS designers and coders would code modules for inclusion into the OS package and they would constantly get requests from the IBM mainframers asking them to *remove* features.
The most amusing (from the MS side) was a straight-faced request to remove the entire GDI printer subsystem with its support for typographical fonts (mostly bit-mapped in those pre-Truetype days) and replace it with a single lineprinter interface justified as "improvement; reduces system complexity, increases reliability, and will improve sales of lineprinter hardware". The request, of course, coming from the lineprinter sales dept who were finding their sales lagging those newfangled laser printers introduced six years earlier.
There were also requests that MS stop trying to graft a GUI on top of the "perfectly adequate" Character-based interface of OS/2 1.x.
It is no accident that MS had a campus-wide party when the split was announced a few months later nor that when it came time to design the NT kernel, David Cutler kept his team small and out of reach of marketing or support types.
http://www.microsoft.com/about/techn...id-Cutler.aspx
It is generally understood that a good programming team should be handled like a ballistic missile; you point it at a target and let it do its work without course corrections unless it gets seriously off course. A corollary says that if the team does go seriously off course, you should treat them like a mule and get their attention with a 2x4.
