Niall's Oracle Pages

Thanks to Doug Burns who points out this article that claims to show by deductive logic that SQL Server is a superior platform to Oracle.

Now I don't think there are many who will rush to agree with the article, and as Doug says it is briefly amusing. Unfortunately I find it rather sad in the end. The reason I find it sad is that in fact the author does not employ deductive reasoning at all. His argument goes.

Proposition 1: SQLServer is cheaper than Oracle.
Proposition 2: SQLServer is easier to manage than Oracle.
Fact 3: Documented minimum hardware for SQLServer is lower spec than that for Oracle
Proposition 4: Systems that install on lower spec hardware perform as well on lower spec hardware as systems that install on higher spec hardware all the way up the scale.
Proposition 5: Both systems have all features required for modern business systems.

That is it, there is in fact no deduction made. Presumably in fact the missing premise that he has in mind is summed up as "Cheaper, easier to manage systems that meet a minimum feature list and will install on old, cheap hardware are superior to those that are more expensive, have more features than the minimum feature list and will install on old but slightly more expensive hardware". Since we can deduce that SQL Server would win on this criteria if the propositions are all true. As well as the somewhat dubious unspoken assumption listed above the planks of even this small argument are somewhat rotten.

Proposition 1: SQL Server is cheaper than Oracle.
This is entirely dependent upon edition and the negotiating skills of your organisation.
Proposition 2: SQL Server is easier to manage than Oracle.
These guys might differ
Proposition 4: If something will install on low spec hardware it will always outperform software that requires a higher base install when running on the same hardware. Does anyone truly believe this?
Proposition 5: Both systems have all features required for modern business systems. Given that the article was written in the days of SQL 2000 presumably the development of SQL 2005 has been a waste of time since SQL 2000 already met all the requirements of modern business systems. The reality of course is that an extended feature set allows smart developers to implement functionality more efficiently and possibly more quickly and reliably.

In addition of course SQL 2005 comes with some price increases and much higher minimum install requirements and greater functionality - perhaps it too is inferior to SQL2000 on this 'logical' basis.