We have talked about this before, Laura, and I find it fascinating. You are the only person I know to have met both Pick and Codd in person, although there might well be others on this list. I had not recalled the information about the presentation by Goodman. Good info. I wish this thread were not told to move to u2-community. I would have hoped that by now u2-users folks could filter their mail, whether automated or with some type of discipline so that any topic of interest to some subset of folks here could be held. I will try not to comment again in this thread as I really don't like getting chided (but, of course, like the Chucks and co anyway). cheers! --dawn
On Mon, Sep 21, 2009 at 3:16 PM, Laura Hirsh <[email protected]> wrote: > While working as the OS Product Manager for The Ultimate Corp. a team of us > met with EF Codd, CJ Date and Dr. Nathan Goodman. The meeting was held at > Ultimate's corporate headquarters in New Jersey and was at Ultimate's > request. The purpose of the meeting was so that we could have an open > discussion about the "relational model" and figure out how and if the > Ultimate model fit in. > > After the meeting, I spent the better part of a week one-on-one with Dr. > Goodman (Senior Vice President of Codd and Date, Inc. at the time) to show > him the database, the toolset, the implementations (Honeywell, VAX, IBM, > LSI...), etc along with some application software. We went through each of > the C&D Rules in detail. > > Ultimate had some features that made a difference in this review - one of > which was Ultimate's Update. Not the update language developed by Pick > Systems, but one that was developed by Chandru Murthi. It used the same > dictionary sets to retrieve and update information using one "view" of the > database. The Update language was built into the OS. You could "LIST > CUSTOMERS NAME ADDRESS CITY STATE ZIP START.DATE OPEN.BALANCE", and also > "UPDATE CUSTOMERS NAME ADDRESS CITY STATE ZIP START.DATE OPEN.BALANCE". > Both > update and retrieval used the same language, syntax, and dictionaries. > Multivalues, index usage and updates, validations, referential integrity, > no-nulls, etc were all handled automatically. > > It was a very interesting exchange. At the end of the exploration task, Dr. > Goodman put together a presentation and in this presentation he concluded > that: > > "This model is the best implementation of what the relational model tries > to > accomplish". > > Basically, the C&D Rules were intended as a guideline, not law, and C&D > volunteered that they weren't perfect. It's was also interesting to find > out > that there are very few, if any, "relational" databases that meet all of > the > rules. Unfortunately, Ultimate's Sr. management changed shortly thereafter, > and the new management decided not to pursue this track. IIRC, EF Codd also > passed away soon after. > > So, the question is, "how do we compete" given the model that we have? And > rather than arguing "relational or not", we should be helping define our > model in positive terms. The Better & Better committee of U2UG has > developed > an application to help funnel ideas into the (now, Rocket Software) > development hopper. There are a ton of good ideas, and as a group, we have > an incredible opportunity to move our group forward. > > Thanks, > Laura Hirsh > > _______________________________________________ > U2-Users mailing list > [email protected] > http://listserver.u2ug.org/mailman/listinfo/u2-users > -- Dawn M. Wolthuis Take and give some delight today _______________________________________________ U2-Users mailing list [email protected] http://listserver.u2ug.org/mailman/listinfo/u2-users
