Roger Bamford: ... MDS???, right?
Don Slutz: No, that was the second one. The first one ... So that all went away, and then sometime in August or September of 1981, he had another one going with this company in Boston, MDS, who basically wanted a multi-user RSS. Ron Revelle, who left in 1980, went off to Britton-Lee, and was working on a database machine. He actually went there as a hardware guy, even though he had been a software guy. Actually he was with Kapali in System D[69], and so they knew each other pretty well[70]. He wanted to do more in the Britton-Lee machine: more hardware accelerator stuff, and Britton-Lee didn't want to, so he wanted to go do it somewhere else, and he got together with Kapali. So the original Esvel plan was to make a better Britton-Lee database machine. So Roger was going, and Ignatius [Ding]. So we got there in 1981, and ... So we were pretty much staying with the same technology - write-ahead log; bulk-fetch, because it was a database machine, so it was really client-server driven. Data-flow in the optimizer, because view composition was too hard, and a bunch of other reasons. So about six months later, Ron was killed in an accident, and that ended the hardware work. So we went on for a while, another year or so, and got some venture money. I guess they got nothing, right? The venture people never got anything out of it.
Roger Bamford: Everybody got stiffed. [laughter]
Don Slutz: Roger left in late 1983, I think.
Roger Bamford: Yes, it must have been.
Don Slutz: Eventually, the RSS part of it, we delivered that on VM in nine months, starting with an empty office in Campbell. And then there was an MVS version a few months later.
Roger Bamford: You mean the ??? - yes, kind of the RSS equivalent.
Don Slutz: Right.
C. Mohan: That HP bought, right?
Don Slutz: HP bought that, and that became ALLBASE.
Jim Gray: Tektronix was an investor.
Don Slutz: So we made a contract with HP in early 1984, and then things changed a lot and a number of us left - six or so, and then another seven or eight - and HP picked it up with ALLBASE. Finally the company was sold to ...
Jim Gray, C. Mohan: Cullinet. [IDMS/SQL]
Don Slutz: Franco came there in late 1983 or early 1984 for three months.
Franco Putzolu: And then I had some minor conflicts with Kapali.
Don Slutz: And when I decided to leave, I called John Nauman; actually, I sent my resume back to Tandem, and I sent one to Oracle - I never heard from Oracle.
Roger Bamford: I thought you interviewed with Oracle.
Don Slutz: Well, no after I took the job with John, Bob Miner called and he said, "We lost your resume, we're real interested. Come on up anyway." So I went up and talked to him for a while, and I spent a few hours with Larry [Ellison] , which was interesting. We got to talking, and I mentioned that I'd been working on performance at one time. And that's when Oracle had been slammed on the Wisconsin benchmark[71], and Larry all of a sudden stopped talking about interview stuff, and opens his big wood desk. He pulled out all these listings. He said, "We've fixed all that," and he showed me all the Wisconsin benchmark runs that he'd made. We went on and on ... Then I said, "Well, I really decided to go to Tandem." He said, "Don't go there. Come here; get rich." [laughter]
Franco Putzolu: I interviewed with Ellison when I was leaving Esvel. He said, "If you come here, you won't have any problems about money anymore; I promise."
Jim Gray: John, do you want to do the Tandem story?
John Nauman: Sure.
Don Chamberlin: Don, wasn't there some kind of shareholder suit around Esvel at some point?
Don Slutz: Yes, that came later. The suit lasted for years and the final agreement included a gag order.
Mike Blasgen: I always figured that Franco went over there as some sort of double agent. Didn't he go in to Esvel and then come out dragging several people with him?
Jim Gray: Right, we almost lost Andrea Borr, but she decided against it at the last minute.
Don Slutz: You have to understand: I was car-pooling with Franco to Esvel, and he doesn't say much. Actually, he was thinking of going back to Tandem, and I was thinking of going to Tandem, and I don't think we even knew that ...
[69] S. Andler, I. Ding, K. Eswaran, C. Hauser, W. Kim, J. Mehl and R. Williams. "System D: A Distributed System for Availability" Eighth International Conference on Very Large Data Bases, Mexico City (September 8-10, 1982).
[70] Jim Mehl notes: "Ron Revelle may have briefly done some hardware investigations at the beginning of System D, but he was certainly not part of the software work that became known as System D." Don clarifies: "I recall that Ron was working on a processor at first (and I thought it was for System D). When System D started using Series 1's, Ron switched to work on the network connection stuff. When that was stopped I recall Ron quit soon after."
[71] D. Bitton and C. Turbyfill. "Benchmarking Database Systems, a Systematic Approach" Proc. VLDB, Florence, Italy (1983).