Interesting article - nice job.
Oh how I love the posts stating "don't do this or that" ... you'd think that their authors either 1) got too bloated on Microsoft documentation circa 1998, or 2) they've been taking Joe Celko posts at face value . I often think of a passage from Ken Henderson's second SQL Server 2000 book, when he quotes Steve McConnell in the following;
"To a great designer, not applying knowledge is tantamount to not having obtained the knowledge in the first place" (Henderson, 2002).
I would hope that those same readers would at least post their alternative, stating the reasons for their belief to go along with their rebuttal, rather than a simple “nay”.
Frankly, I enjoy the creativity here. If not for the idea in its form as presented, certainly for the thinking behind it, the very type of creativity that makes us all better at database development. Lest we not forget that someday the relational database will not be as we know it now. In fact, many are surprised that those concepts invented back in the mid to late 1970's are still around! I believe that in the next 20 years, advances in hardware, as example, might lead us to not even think of what were once "data retrieval problems" as we now know them. Dare we venture to think that what was once unacceptable in the relational database model might gain acceptance? Sure, you may never use this type of coding, but who cares? Completely immaterial from my point of view.
Again - nice job here. Thank goodness that we didn’t get another article on SET NOCOUNT ON, ROWCOUNT, or something so exciting as the advantages of SET over SELECT.
Henderson, Ken (2002). The Guru's Guide to SQL Server Stored Procedures, XML, and HTML. Indianapolis: Addison-Wesley.