Archives of the TeradataForum

Message Posted: Thu, 26 Feb 2009 @ 10:04:22 GMT

	<Prev	Next>		<<First	<Prev	Next>	Last>>

We'll forgive him faster if he keeps on producing more jokes and nostalgic references to things like APL.

SQL does use notation which is too "verbal" for good mathematical notation. Because of the "verbosity" we almost lose all the nice geometric background (ideas based on intersections and unions of sets, projections and sections, etc) in favor of the algebraic-like notation but then without the elegance of the true algebra. In fact, SQL notation is not that different from a very ancient algebraic notation used in the middle ages when they were just writing all formulae in full. Before this French gentleman got bored with it and introduced a new system:

en.wikipedia.org/wiki/Vieta

Vieta's new notation was responsible for the subsequent progress in algebra. People could just write their ideas much faster and post the letters in time for the post to pick them up.

Speaking of nostalgic references: there used to be a different system called QBE, Query by Example.

en.wikipedia.org/wiki/Query_by_Example

I don't think QBE ever evolved to any international standard, and its text version was not very easy to follow. But some vendors (like Borland) did supply nice graphical interfaces for writing queries in QBE. They were good for people thinking in geometric terms, for example. Of course, there are "query builders" for SQL as well but I never saw one which I would want to use twice. Their main problem is that they visualize tables, not the queries, so there is no real benefit for actually writing or discussing statements.

Because of SQL's linguistic bias its usage may correlate with the user's linguistic background. I think this does happen. Having worked in a number of countries I did notice differences in SQL usage.

Victor

	<Prev	Next>		<<First	<Prev	Next>	Last>>