This paper presents CSPub, Wallach’s “proposal for a new cost-free open-access publication model for computer science papers” that could facilitate the work done by authors and reviewers. In explaining his proposal, the author briefly covers how affected groups could benefit from some varieties of his proposal, giving most attention to the academic environment and to academic-oriented conferences and journals. The major affected groups covered are overloaded reviewers, authors of resubmitted papers, authors of conference and journal papers, and authors of short incremental work and technical reports. Wallach characterizes CSPub as a mashup of technical archiving services, review management software, and conference submission processes. In his proposal, any submission handled by CSPub would be open for anyone to see, together with the reviewer feedback. This could aid anti-plagiarism efforts and spam identification. Popular CSPub-received submissions might be picked up for journal publication, which could help reduce journal latency.

Several relevant points are not addressed in this proposal: Should there, or could there, be multiple entities doing the CSPub process concurrently? When, specifically, does “publication” occur for submissions handled by any CSPub entity, and is it the same for all CSPub entities? Is the key determinant for the effectiveness of any CSPub version--a free, built-in comprehensive search process--open to anyone? How well does such a search process cover the in-process and finished submissions handled by other CSPub entities; prior, similar computer science materials, including those of inactive publishers, non-CSPub entities, and CSPub entities; and all CS publications or materials created prior to the initial implementation of CSPub? How well does the search process handle nomenclature changes and additions? For example, in the early years of CS work, we often worked with and on software, but we did not use the term “software” even when the work we did involved machine language executable code.

The author briefly notes some possible negative consequences of CSPub, such as altered ownership of submission content, reduced subscriptions to journals, reduced revenues from digital libraries, and reduced conference registrations. Given the increasing number of computer science papers being authored, CSPub merits the attention of personnel in all of the potentially affected groups.