I think you could turn almost all arguments in this text the other way around. A team of 4 developers spend 12 months building a system and then a DBA works wit

I think you could turn almost all arguments in this text the other way around. A team of 4 developers spend 12 months building a system and then a DBA works with it for 4 hours a week. Which part should you optimize? Long time ago there were many more Oracle and DB2 users. No-one should ever have switched to SQL Server accoring to the logic in the text. Accoring to the text Oracle must be better since it is more expensive in almost all configurations I've seen (in this argument I don't care if it really is better or not). Should all projects that don't need the extra features/quality/performance still pay the premium? And so on. I agree that there is a cost to esoteric choices but I think is really wrong to say that you should always choose the dominant product/technology. If that were the case we would still use switches on big "mainframes".