THEORYNT@YKTVMX.BITNET (02/13/90)
The field of databases has accomplished much during the past two decades, providing technology for addressing important real-world problems and establishing itself as a discipline in computer science. Just as relational databases became a reality, research into the next-generation database technology began, spawning a number of subfields of research, including engineering and design databases, temporal databases, multimedia databases, deductive databases, real-time databases, extensible databases, object-oriented databases, etc. In the view of many database professionals, one of the disappointments the field of databases has suffered during the past several years is the lack of synergy between database theorists and systems people. Systems people have used such conferences as SIGMOD and VLDB as their forum, while theorists have used PODS as their forum; each camp has shied away from the other camp's forum. The subfields of database research pose many problems that require theorists and systems people to address together in order to make the next-generation database technology a reality sooner. Ideally, the systems people should define the problems and offer initial solutions; the theorists should establish rigorous foundations to the problems, and offer solutions that satisfy real-world constraints to varying degrees. This is a proposal to provide a common forum for exchanges of problems and solutions in various subfields of database research among database theorists and systems people. The following are the principal elements of this proposal. 1. The SIGMOD and PODS conferences will be held at the same time and place as essentially independent conferences during the same two and one half day period (or three day period). There will be however several common sessions (e.g., keynote lectures and panels) that should be of interest to both camps. 2. There will be one organizing committee, but two independent program committees for the two conferences. The organizing committee will consist of representatives from both SIGMOD and PODS. There will be a single local arrangement chair for the joint conference. 3. All nontechnical programs, including lunches and the banquet, will be held jointly (i.e., as if it were a single conference). 4. Both the SIGMOD and SIGACT mailing lists will be used for the call for papers and advance program announcements. 5. SIGMOD will take the fiscal responsibilities for the joint conference. In particular, there will be a single registration for the joint conference. There will be two separate proceedings: one for SIGMOD and one for PODS. Each person registering for the conference will will receive both proceedings. The profits/losses from the sale of the PODS proceedings will be split evenly among SIGMOD, SIGACT, and SIGART, the three SIGs which have co-sponsored PODS in the past. 6. We will try this arrangement for two years, beginning in 1991 in Boulder, Colorado. The plans for the SIGMOD `91 (site, PC and general chairs) were firmed up about six months ago. 7. The dates for the combined conference will always be chosen so as not to conflict with the STOC dates. 8. If, after trying for two years, the sentiment among the participants indicates that the cultural gap between the two camps is simply too wide to bridge, then SIGMOD and PODS will go their separate ways, whatever the future holds for them. Comments on this proposal are very welcome. The proposal will be discussed and voted on in the business meeting of the next PODS meeting (Nashville, Tennessee, April 2-4, 1990). Avi Silberschatz (avi@cs.utexas.edu) Won Kim (kim@mcc.com)