Alan Penn wrote:
> This in itself adds another layer of complexity to the whole project
> of coding, and one which (I think) may only be amenable to formal
> code management, systems engineering, top down control, rather than
> the open market place of open source - everyone 'doing their thang' -
> perhaps I am wrong though...
Well, many open-source projects are run on the basis of having a stable
official distribution of the core, and and a modular structure.
Typically these do have formal code management, and a central
At the same time, they also develop within the open market place of open
source: for example, anyone can take the R-project core code, muck about
with it, and release it as their own version - in that sense, the core
remains open. Few people would use *that* version of the core unless it
was stable, reliable and proven, at which time it might merge into the
official stable core. The other way to get things into the core is to
make significant contributions in the open borderlands of the closed
official core, and build a reputation as being able to meet a need.
So, open source in an open market is not incompatible with a central
formal distribution that has tight source code quality control: AIUI,
Linux (OS), the R Project (stats), GTD-PHP (task management) and many
others, all work in this way. Typically, that structure emerges over
time, often concentrating around the nucleus of the prime mover(s) of
With the R Project, there are indeed innovations in methods that go in
in the add-on packages; in areas where there aren't necessarily
well-defined right answers.
An open-source, modular space syntax project is imaginable. Such things
often start from a brand new generation of users, who start a project to
meet their own needs and who won't / can't afford to pay for any
existing tools that would do the job; and/or from within academia.