Recent books by Chris Date
In case anybody is still looking in here, Chris Date has just told me about a number of books he's had published over the last five years. I've posted the complete list here at the TTM website, where I also give this link to the publisher's price list for the ebooks and print copies.
Chris did send me a copy of one of them, titled "Database Technology: Nulls Considered Harmful", which has been fun (sort of) to dip into from time to time.
Hugh
Thanks!
I've got 5 of the first 6 of those. (Not the "Codd's earliest writings" one.)
For some reason, I kind of thought he had stopped after that (the two "Dreaming" books) ...
Somewhat in connection with the "deconstructing RM/V2" book : I recall a Codd writing where it was pointed out that "tuples with nulls should not be considered a member of the relation for purposes of constraint enforcement. If the attribute that has the null plays a role in the constraint at hand, that is. Such tuples should only be considered member of the relation once the nuil is replaced with the real actual value that should be there.". But I can no longer find where he wrote that. (I thought it was the RM/V2 book but that doesn't seem to be the case.) Has anybody ever noticed that remark by Codd ? And still remembers where it was made ?
Are you quoting or paraphrasing?
Missing Informatlon (Applicable and Inapplicable) in Relational Databases
SIGMOD RECORD, Vol. 15, No. 4, December 1986
https://sigmodrecord.org/publications/sigmodRecord/8612/pdfs/16301.16303.pdf
page 69:
When an attempt is made to insert a new tuple into a relation and a certain component db-value is missing, it is pointless for the system to base acceptance or rejection of this tuple on whether the missing db-value does meet or might meet or fails to meet certain integrity constraints based on a dependence in which the pertinent attribute is involved. The proper time for the system to make this determination is when an attempt is made to replace the pertinent mark by an actual db-value.
This is in the context of I-mark & A-mark (4VL) nulls.
Although the article says "This material is planned for inclusion in a book"--presumably RM/V2.
PS
Just come across in the ACM TODS RM/T paper ie where Codd first proposed (3VL) nulls:
Extending the Database Relational Model to Capture More Meaning 1979
http://www.esp.org/foundations/database-theory/holdings/codd-1979.pdf
Undisciplined application of predicate logic in designing a database could yield an incomprehensible and unmanageable set of assertions.
ha ha ha
Thank you anonymous ... I was quoting from remembrance as best as I could, so I guess this means "paraphrasing" anyway ...
Quote from Anon8 on September 6, 2024, 4:53 amWhen an attempt is made to insert a new tuple into a relation and a certain component db-value is missing, it is pointless
Although the article says "This material is planned for inclusion in a book"--presumably RM/V2.
8.21 Normalization p. 193, right after the 2 bullets.