A large part of the work in writing a paper in my field (theoretical CS and economics) is writing the introduction, the problem description, the motivation, and the related work. A large part of this writeup repeats in different articles studying a similar problem; thus there is a waste of effort. I thought of initiating a journal in which the authors will be instructed to base this part of their papers on relevant Wikipedia pages. Each article would start with a paragraph such as the following:
"This paper extends the Wikipedia article ABC (version from date of submission). Our contributions are as follows: (a) we prove that the ABC theorem is valid not only for $n=2$ but also for $n=3$. (b) We refute the ABC conjecture for $n=4$. (c) We present an efficient algorithm for solving the ABC problem for the case $n=2$.
The authors are responsible for ensuring that the Wikipedia article at the date of submission is uptodate, contains a complete problem statement, and all relevant related work. The referees will have to verify this. Then, the rest of the paper will be verified as usual. If the paper is published, then the new contributions will be added to the ABC page.
- The literature survey will be done only once, and it will be kept uptodate.
- The notation will be uniform (based on the Wikipedia article).
Current journals will probably not agree to this scheme. But can it work with a new journal? Is this scheme reasonable?