Simply put, TODS is in great shape, in terms of its prestige and impact. My first editorial, in the September 2001 issue, listed the convincing evidence of this fact.

However, several challenges confront TODS, and to a large extent scientific journals in general. In the remainder of this note I attempt to characterize these challenges.

DECLINING SUBSCRIPTION BASE

In July 1995 there were 3101 regular TODS subscribers. By March 1999, less than four years later, this had fallen fully by half, to 1686. At that rate, there will be no regular subscribers in a few years.

This is somewhat less alarming that it initially appears, because some of the decrease is attributable to institutional Digital Library (DL) subscriptions, which have dramatically increased over that same period. Those at most research universities and major research labs already have access to the ACM DL, and many of those have dropped their individual subscriptions.

Nevertheless, this large drop is troubling, because it reduces the number of people who receive the print copy every three months (people are less likely to go the the library to look at recent copies, or periodically browse the DL), and it is the canary in the mine, indicating less interest in the publication.

While many EiCs and members of the ACM publications board are sanguine about decreasing subscription levels, I find these decreases alarming. Decreasing subscriptions are not directly the problem; rather, this is a symptom of a more disturbing underlying concern: a lessening interest in the journal itself. Quite frankly, TODS has become a journal of a few (about a dozen per year), very long (average length: 40 pages), old (average age: over three years) papers, which are read by very few researchers. And the move toward electronic publication only exacerbates this situation, because print subscribers who perused the publication when it arrived in the postal mail will no longer have this reminder. TODS needs to include more recent work, more and shorter papers, and other features to entice people to look at the journal, while retaining very high quality and material of lasting value.

We have already taken several steps in this direction, by encouraging shorter submissions and encouraging survey papers, both academic and industrial surveys. But we need to do much more.

INADEQUATE BACKLOG

For the past few years TODS has been coming out later and later. The June 2000 issue was delivered in mid-January 2001. The ACM Publications Board has been working very closely with the ACM publications staff to address late issues for TODS and for most other ACM journals, with impressive results. The March 2001 issue was delayed in the production process by over five months; the December 2001 issue was delayed by a much-improved process by only 22 days. The production process is now back on schedule.

It is now clear that TODS has an inadequate backlog. The publications staff requires that all the papers for a particular issue be delivered by the EiC by the first day of the month four months before the issue date. So the papers for the June 2002 issue are due February 1, 2002; the papers for the September 2002 issue were due May 1 (last week). We have just completed the June issue, but no papers have yet been accepted for the September issue, so the backlog is over four months behind. This was not particularly a problem when the production process was so slow, but it has now become a major concern. (I note that several other ACM Transactions also have inadequate backlogs.)

We've already moved forward in this area, by inviting papers from SIGMOD (I just accepted an invited paper from SIGMOD'01 and have just invited three papers from SIGMOD'02; thanks, Mike Franklin!) and hopefully from PODS (this is still in discussion by the PODS EC, with a final decision scheduled for the SIGMOD conference). We missed EDBT this year but I hope we can get EDBT in 2004. Inviting ICDT papers is on hold.

In any case, an inadequate backlog is another canary in the mind, indicating inadequate interest in this publication.

ACM AS THE PREFERRED PUBLISHER

The ACM Publications Board has prepared a wide-ranging strategic plan to establish ACM as the preferred high-quality computer science publisher. As an initial step, the pubs board has worked with the SIG Governing Board and the EiCs to state explicitly what Rights and Responsibilities it provides and expects. This document is comprehensive and exacting in what it expects, of ACM and of its journals. Currently, neither ACM nor any of its publications fully implement the rights listed in this document. TODS needs to move aggressively to make good on the rights that ACM promises.

IRRELEVANCE IN THIS INTERNET AGE

A related concern is the perceived decrease in relevance of "old school" journals in an age of electronic journals, instant news, and a wide variety of (high- and low-quality) free information sources on the web. For each the papers that appeared in TODS from September 1999 to June 2000, the total processing time, from initial submission to appearance in print, was 25 months. It is difficult for such a slow and cumbersome medium to compete with conferences and newsletters, which have a total processing time on the order of six months to nine months, and often are freely available on the web. I believe this is a driving factor in the declining subscription base, and implies that even with access to the ACM DL, TODS papers are being less frequently read.

ANALYSIS

How serious are these challenges? It is the impression of many on the ACM Publications Board that current journals will cease print publication in just a few years, and may become irrelevant in just a few more years. In the longer term, there is the oft-expressed concern that journals will become the dinosaurs of the information age.

My view is somewhat to the contrary. I feel that with the democratization of information on the web, more is available, but with a concomitant increase in variability. There is much more good information, but also much more poor or simply incorrect information, on the web. In such an environment, indicators of quality are even more critical. So I believe that the careful reviewing associated with TODS, along with its established imprimatur, will be of great value, and will be the basis for the journal's enduring legacy. That said, there will always be the need to innovate the strong journals for them to prosper.