From: Dave Voorhis

To: Editor

Date: 7 Aug 2004

 

I've been enjoying your material on dbdebunk.com for some time now, but haven't felt inclined to respond until today, when I read No Integrity: A Systemic Problem. You state:

 

Some think that the hope lies with open source/public domain (OS/PD) product designers who, unfettered by commercial considerations, can and will try to “do the right thing.” We’ve been doubtful, though, because such designers operate in the same industry and business culture as commercial vendors, are the product of the same education system, and consume the same information generated by the industry and media.

 

I agree with your views about the current business and media climate and the failings of the educational system, but I fear your comment paints all us Open Source developers with one broad brush.  I'd like to offer a little hope that there ARE some of us who try to "do the right thing." As evidence (and a bit of shameless self-promotion) I offer the URL to my project, Rel: http://dbappbuilder.sourceforge.net/Rel.html.

 

Admittedly, it is a long way from complete and the shortcomings (at this early stage) are profound, but it is a serious and ongoing effort to produce a true relational database. Given some of the positive feedback I've received from a handful of folks who have downloaded Rel--despite it being too embryonic for any practical value beyond teaching or experimenting--there are developers and users out there who "get it."

 

On an unrelated note, I again agree with the point you're making, but I believe your quote of MySQL documentation, "Reasons NOT to Use Foreign Keys constraints," may be out of date.  At least, it doesn't agree with recent versions of MySQL.  Foreign key constraints have been implemented. Perhaps the MySQL crew are learning?

 

 

From: Fabian Pascal

To: Dave Voorhis

 

My comments were on the systemic rule, not this or that rare individual exception. There is a link to Rel on the Links page on DATABASE DEBUNKINGS, but it's not yet anywhere near a usable product. We'll see if it ever becomes a full-fledged product and its acceptance rate by the system.

 

The point is not whether MySQL has added FKs or not, but the state of knowledge of its authors.

 

 

Posted: 10/15/04