Re: Would be really nice if...

From: Serge Rielau <srielau_at_ca.ibm.com>
Date: Tue, 01 Mar 2005 09:19:01 -0500
Message-ID: <38jbtkF5nlpdvU1@individual.net>

Haximus wrote:

> "Joel Garry" <joel-garry_at_home.com> wrote in message 
> news:1109640158.812020.233540_at_z14g2000cwz.googlegroups.com...
> 


>>>I tend to side toward those that wish to dissolve the group if only

>>>because I fail to see that they have brought any value to us. What

>>>they have standardized is so inconsequential that the cost and
>>>difficulty of moving between products is still a substantial barrier.
>>
>>I think that is just more recent. I think standards groups have a life
>>cycle like so many other things, and the first decade or so this one
>>probably had a profound positive effect. The fact that in a commercial
>>world there is going to be a strong economic motivation towards db
>>balkanization really only took over in the latter '90s.
>>
>>I don't really take a position one way or another on dissolution, but I
>>would hope that at least trying for standards is better than
>>unconstrained db lockin. And since Bill and Larry both seem to realize
>>it's the applications over all, they have their own agendas which
>>probably don't agree with ours. Of course, standards as a code word
>>for oligopoly is a danger, too.

> 
> 
> SQL standardization is a farce for the most part, (something I can actually 
> agree with Danny boy on) the vendors should just free themselves from the 
> reigns and some serious product advancement could happen.  I don't 
> particularly like db lockin... I do like choice and portability... but has 
> there been any real leaps made in the SQL world lately?  Nothing 
> significant, advancement is glacial at best. 
> 

I don't think any of the vendor are in reigns and innovation in SQL is quite orthogonal to the SQL Standard.
To use Daniel's example of OO Extensions: 1. Many of the OO extensions are, in fact, standardized. 2. A standard doesn't seize being a standard because vendors chose to not implement certain areas of it.
If Energizer were to choose not to produce 9V Block batteries, that does not invalidate the fact that their form and interfacse is standardized. 3. The innovation rarely lies within the API. It lies in it's implementation.

In terms of recent leaps in the SQL Standard I point out the MERGE statement (now implemented by three products: Oracle, DB2 and XPS), SEQUENCE, IDENTITY, OLAP functions, and SQL/XML. There are other proposed leaps, but I shall not peddle my product agenda here.

Cheers
Serge

-- 
Serge Rielau
DB2 SQL Compiler Development
IBM Toronto Lab

Received on Tue Mar 01 2005 - 08:19:01 CST