• From: Ari Krupnik <ari@c...>
• To: xml-dev@l...
• Date: Sat, 09 Sep 2006 22:57:47 -0700

"Michael Kay" <mike@s...> writes:

>> Aside from the technical benefits, a key challenge is 
>> convicing our business sponsors of the need for such efforts 
>> since, as always, infrastructure is a tough sell. 
>
> I remember Michael Stonebraker saying that if a project isn't spending 20%
> of its effort on tooling then it needs to spend more.

I'm sure you've also seen enterprise projects that are 80% tooling,
frameworks upon frameworks, where most of the custom code is just glue
that tries to make them work together.

Axis <-> glue <-> Tomcat <-> glue <-> Castor <-> "business logic" <->
Hibernate <-> glue <-> Oracle

I once saw a project's LOC count shrink by a factor of 20 after the
frameworks were removed and replaced with "non-reusable" code that
spoke SQL on one end and XML on the other.

Ari.

-- 
Elections only count as free and trials as fair if you can lose money
betting on the outcome.

• References:
  • RE: XML schema Management
    • From: "Michael Kay" <mike@s...>