FOL/HOL: is there a middle ground?
From: Marshall Spight <mspight_at_dnai.com>
Date: Sun, 18 Jul 2004 19:26:14 GMT
Message-ID: <q9AKc.114259$IQ4.31025_at_attbi_s02>
Hi all,
Date: Sun, 18 Jul 2004 19:26:14 GMT
Message-ID: <q9AKc.114259$IQ4.31025_at_attbi_s02>
Hi all,
Lately we've been discussing 1NF and "a question for Mr. Celko." There are clear incompleteness problems associated with allowing the definition of set A to references the definition of set A. What are some of the other problems with using HOL as a basis for a relational algebra? Is there a middle ground, above FOL but below fully-recursive definitions, that is more expressing than FOL but still consistent?
Could we call it "fractal order relations?"
What are the various pitfalls one runs into with recursively defined relations? Can we enumerate the "holes" and plug them all with restrictions on our fractal order logic?
Marshall Received on Sun Jul 18 2004 - 21:26:14 CEST