Redundancy and NULLS in Xdb
Date: Sun, 7 Nov 2004 17:49:06 -0500
Message-ID: <fdidnSmXa6P0OxPcRVn-2A_at_comcast.com>
For almost a year now, Neo has been "treating" us to presentations on how his "experimental database" represents complex data without using any redundancy or NULLs at the atomic level.
Of course, he's wrong. The data items like "brown" that he claims are "atomic" really aren't atomic at all. Not only can they be broken down into characters, but the individual characters can be further decomposed.
The only form of digital data that can't be further decomposed are the raw binary digits: zero and one.
What Neo has shown us is syntactic sugar. Once we strip all that sugar away, and get down to the real nitty gritty, we'll see that an Xdb database is really a bunch of ones and zeroes.
Right away, Neo is in trouble. The ones are all redundant, except for the first! They all look identical! A truly non redundant system would have only one "one" in it. All the others would be replaced by pointers back to the original one. The same treatment would have to be given to the zeroes: a truly redundant system would have one one "zero" in it. All the others would be replaced by a pointer back to the original zero.
But it doesn't stop there! Not only are the zeroes redundant, but they also have NO VALUE! The PHB once pointed this out to Dilbert! What they really are, are placeholders indicating where a one might have been written, but was not. Functionally, they are NULLS.
So Neo's Xdb. at the truly atomic level, is full of redundancy and NULLS. Hence he hasn't even won his own contest!
In case you are by any chance wondering.... the above is SATIRE!!!! Don't take ANY of this SERIOUSLY, OK? Received on Sun Nov 07 2004 - 23:49:06 CET