The November 2009 issue of CACM published my letter to the editor entitled Abolish Conference Proceedings (Digital Edition). Here is the published text (which was edited for brevity from my longer letter).
As program chair of an ACM conference (Hypertext 2009), I agree with both Lance Fortnow’s Viewpoint “Time for Computer Science to Grow Up” (Aug. 2009) and Moshe Vardi’s Editor’s Letter “Conferences vs. Journals in Computing Research” (May 2009). Moreover, as an interdisciplinary researcher, I experience firsthand how conference-driven publication practices hurt CS in terms of potential interdisciplinary collaboration, reach, and visibility.
That’s why I propose the abolition of conference proceedings altogether. Submissions should instead go to journals, which would receive more and better ones, with refereeing resources shifting naturally from conferences to journals. As a result, journals would improve their quality and speed up their processes. With the CS community’s full attention, the review process would be more rigorous and timely. Deadlines would no longer be so concentrated, and scientists would submit better work, revise as needed, and profit immediately from reviewer feedback; the same referee would judge improvements to a particular submission.
In many cases where conferences and journals are aligned, presentations could be invited from among the best papers published in the previous year. For newer areas and groundbreaking work, a conference or workshop could still accept submissions but would not publish proceedings. Publishing would be the job of journals.
ACM should shepherd such a transition as publisher of both the proceedings of most top computing conferences and of many top computing journals.
After writing my letter to the editor, it was brought to my attention that there already exists a model for the approach I proposed, envisioned by the VLDB Endowment as a transition from the VLDB conference proceedings to the PVLDB journal and ultimately to a Journal of Data Management Research.