The task of the referee is to evaluate a (research) paper in a timely manner for publication in a specific journal or conference proceedings. This involves determining if the work presented is correct, if the problem studied and the results obtained are new and significant, if the quality of the presentation is satisfactory or can be made so, and what revisions and changes to the paper are necessary and/or desirable. The evaluation must be relative to the degree of selectivity of the specific publication.
In this article, we discuss the problem of how to evaluate a paper for publication, and by inference, how to write one. The role of the editor, and rules and procedures used by most computer science journals are discussed. Brief discussions of referencing proposals and survey and tutorial papers are also given.