NSK said:
- Every article is developed in a software-like fashion, i.e. there
are versions and when the mains (maintainers) approve an article it will be checked by the admins. The admins then publish the article in a non-Test page (i.e. the main namespace) and give to it a unique version number such as 1.4.
Wouldn't this put a tremendous strain on administrators? Every single article would have to be vetted before publication. I think it would also make contributors less likely to produce articles, too. For instance if I'm copy editing an article it's not unusual for me to encounter a redlink and write a brief stub for that link. This enters my watchlist and, believe me, people do notice those stubs and add material to make them, when I go back a few weeks later, often a very good article on the subject. An example of this is my very brief stub for "Portolan chart" which I produced in the course of copy editing an article uploaded by someone else that was later deleted as a copyright violation. The second and third paragraphs were added more recently by someone else. http://en.wikipedia.org/wiki/Portolan_chart
This is one thing that works very well in a Wiki. I don't think I would have bothered if the production process had been more complex.