I'm talking about PURISTS who believe that SQLis not compliant with the relational model (c.f. C.J. Date and the crew at dbdebunk.com).  Likewise I'm talking about people who argue that HTTP should be used to "transfer" resource representations rather than "transport" objects and methods.
 
Obviously SQL and HTTP are successful and will live for decades at least. What is less clear is whether the advice of those who derive recommendations from the abstract principles of the underlying theories  (as opposed to emipircally demonstrated successes) is to be taken particularly seriously.
 
To put it another way:  Most code is not particularly compliant with the principles of OO design; most databases are not highly normalized in acoordance with relational theory, and most successful websites use HTTP in a non-RESTful way.  Is this a problem, i.e. could they do better if they played by "the rules", or do the rules only apply in an idealized world?    Don't we need an answer to this kind of question before worrying about whether Zachman has a nice checklist of things to think about or a rigorous conceptual framework that one should adhere to?
 
p.s.  There's still time to submit your thoughts / experiences on the subject of how all this XML / web services / REST stuff really works in the enterprise-y world for the XML 2006 conference - http://www.idealliance.org/xmlusa/06/call/

 
On 6/16/06, patrickdlogan <[EMAIL PROTECTED]> wrote:

> ...he reminded me of a RESTifarian (or a relational model purist for



 
Not much evidence that the relational model and the HTTP protocol will
be long lived?

Could you please define "long" and provide examples of "the
alternatives that don't struggle too hard"?

I have a feeling we may in fact live on different planets.

-Patrick


__._,_.___


SPONSORED LINKS
Computer software Computer aided design software Computer job
Soa Service-oriented architecture


YAHOO! GROUPS LINKS




__,_._,___

Reply via email to