OT: Years of Errors

[Georg Wrede]

Manfred Nowak wrote:
> 2_300_000 Euro, two years too late, still not capable to serve the 
> contracted workload and more then ten severe errors in the bug-tracker.
> 
> http://www.heise.de/newsticker/meldung/87499/from/rss09
> 
> Who can show, that choosing D as implementation language would have 
> avoided such a desaster?
> 
> -manfred

Only having read the referred article, I do suspect that it's all about 
administrative failures at the contractor.

I've been involved in several SW disasters, both as client, independent 
contractor, employee of contractor, and outside (more or less) post 
mortem consultant. When you dig deep enough, IMHO, it very seldom is a 
question of inadequate choice of language or platform. (Although the 
language and the platform are the ones that usually end up as 
scapegoats. That way nobody accuses each other, thus avoiding throwing 
the first stone in a glass house where everyone is guilty and/or a wimp.)

Lack of information exchange between the actual programmer(s) and end 
users, sloppiness of middle management, private agendas, and a general 
disinterest by the formally responsible parties (COOs, Political 
Leaders, or top management in general), are the overwhelmingly most 
prevalent causes of software project disaster -- in my own experience. 
And in that order.

The chosen language would not seem to be the case here.

OTOH, in a number of several minor projects (less than 1 man-year), the 
choice of language actually has made a difference, in my past 
experience. In those cases, the chosen programming language (or, more to 
the point, the libraries or the framework) have drawn enough resources 
to actually suck manpower from management and administration of the 
project.

Middle management's default reaction upon discovering slippage of 
timetables (i.e. smoothing out, covering up, wishful thinking, and 
downrigh lying), usually only exacerbates the brewing disaster.