Many of us are familiar with the advantages and disadvantages of relational database technology.
There are two problems with finding things in an rdbms environment.
Firstly, you have to build a structure that seems appropriate and then you have to be aware of that structure when you want to carry out searches. The envelope can be pushed somewhat during searches using natural language processors and “cube” software.
The problem of carrying out searches and connecting the dots across multiple different relational databases remains where there are no defined connections between records in one rdbms and other rdbms’.
In our work, we have successfully combined relational and hierarchical database technology in a graphic environment where you can carry out free-form searches across all of the data and visualize the results of a search, particularly where “hits” were found and, for us at least, interestingly, where there were no “hits”.
The latter has been an area of focus for Civerex for the past few years. We now have operational connect-the-dots applications using this technology.
“Finding” things that are not there has practical value – consider McDonalds trying to decide where to put a new outlet. On the one hand, they could look for locations where Wendy’s has outlets but they might at times be more interested in finding locations where Wendy’s does not have outlets.
Any organization with an interest in participating in this initiative should contact Civerex at Civerex@civerex.com. Civerex is starting up a large-scale venture that will make extensive use of the technology we have developed.
KWKeirstead's Blog 
It is interesting to read about Civerex also using a special DB technology, because when we built our platform in 1997, the first thing we realized that a normal relational database was not satisfactory to offer the needed dynamics and adaptiveness. So we developed an OBJECT-RELATIONAL datastore that is the core of each Papyrus Platform node. No need to buy a database license. It is a different kind of distributed NOSQL database built on top of fully licensed Oracle-Berkeley key-sequential (B-tree) technology. Data object definitions can be replicated across any number of servers and searched the same way. Automatic replication reduces latency and provides data backup. NLR queries read the DB-object structures and verify the validity of possible queries. Fuzzy-searches and pattern matching as we use in our AI-capabilities become really easy to do.
Apparently there is a new kind of application capability that can’t be built on old technology such as SQL databases.
LikeLike
I am not surprised that you went the route you took. Smart move!
For a number of years we represented Interactive Software Engineering, inventors of the Eiffel O-O language (Dr Bertrand Meyer). We also represented mdbs inc, inventors of the MDBS IV/Titanium network database management system and GURU, their expert system application development environment.
Our president, Jean Barbara Kelly, organized high tech seminars on satellite communications, military radio, LANs/WANs, programming languages and dbms’, initially out of Singapore and later Canada. We stopped doing these seminars around 1994.
Our current project is to build what we call a 3-D visual dbms for an application that we expect will have 1,500,000 users on various hosted configurations. This is a 5-year initiative.
LikeLike