Problems with Monoliths
The issues about monolithic vs. modular ontologies
is closely related to questions about monolithic vs.
modular software in general.
Many articles have been written about the upcoming
version of Windows, formerly called Longhorn and
now officially named Vista. A recurrent theme is
that this is the end of the line for monolithic
development, which even Microsoft cannot afford.
Among the symptoms of their problems are the long
delays, and the abandonment of many features that
had been promised for Longhorn. Even more telling
is the fact that Apple, a company with a tiny
fraction of the MSFT resources, produced a system
with much greater functionality and user friendliness
by building their GUI on top of the notoriously
user-unfriendly, but highly modular Unix (more
precisely, a Unix-clone called Darwin).
Following are some excerpts from one of many articles
on this theme.
John Sowa
______________________________________________________
http://observer.guardian.co.uk/business/story/0,,1868642,00.html
Why Vista will mean the end of the Microsoft monolith
The Vista saga has two interesting lessons for the computer business. It
raises, for example, the question of whether this way of producing
software products of this complexity has reached its natural limit.
Microsoft is an extremely rich, resourceful company - and yet the task
of creating and shipping Vista stretched it to breaking point. A lesser
company would have buckled under the strain. And yet while Microsoft
engineers were trudging through their death march, the open source
community shipped a series of major upgrades to the Linux operating
system. How can hackers, scattered across the globe, working for no pay,
linked only by the net and shared values, apparently outperform the
smartest software company on the planet?
The difficulties in developing Vista stemmed from its monolithic
structure and the need for 'backwards compatibility', ie ensuring that
software used by customers on older versions of Windows will work under
Vista. This vast accumulation of legacy applications acts like an anchor
on innovation. The Vista trauma has convinced some Microsoft engineers
that they will have to adopt a radically different approach.
Its outlines are already visible. It involves, firstly, abandoning the
idea of an operating system as a monolith and breaking it into modules,
and, secondly, adopting 'virtualisation' - a key technology that enables
a single machine to run several operating systems (or modules thereof)
in parallel - to deal with the backwards compatibility problem.