Hi,

Long time lurking. We have two major engineering operations based on Turbine
2.1 and Torque. Several aspects of Turbine and Torque have liberated our
software process, especially in the area of ability to support rapidly
changing customers, markets, requirements and business models. 

> 3) Torque doesn't hold a candle to OJB.

and,

> Torque's three main features are Java code generation, DDL generation,
> and persistence. DDL generation is being handled very nicely by
> commons-sql, persistence is handled nicely by OJB, but as of yet,
> there isn't really a Jakarta project for Java code generation. Instead 
> of making this an argument for keeping Torque around. 

>From my perspective Torque is probably the major reason Turbine is so useful
to us. In our engineering business, we support a direct model, so if the
software cant keep up with the business model, or the software becomes
obsolete to the business model we are unable to capitilise on new revenue
streams. This moves pretty fast as well, and it isnt uncommon for new
functionality to be added and requested with turn around times of one day.
It may be as simple as a customer refusing to pay an invoice until they see
a "received by" field on the invoice, or a complex structural change
requiring that Work Orders show partially billed states to the customer
through their history. Even though our developers differentiate between
trivial changes and structural changes, customers dont. ie the "all I need
is a button which .... ". 

Previous to Torque in our software systems, the database was the rate
determining step. Any change that propogated down to a data structural
change was a pain in the arse and took far too long. It meant the database
schema had to be modified, the data mapping layer and then the business
objects to the view. This caused large differences in the turn around time
of requests to be implemented, if it touched the database, we were subject
to the tyranny of a schema change.

We have a highly fluctuating data model, it has changed from 60 tables to 93
in fourteen months of production, this is despite pushing more transient
data into XML repositories. Our schema.xml has gone through 61 check-ins
since being in production as well. Having a code generation tool that can
efficiently support that type of data structure volatility is pretty
important to us. 

With Torque this type of change is a quick XML edit and running of a shell
script. In essence it has allowed me to make to refactor the database schema
and structure with confidence. Previously we used to get a complex schema
that was designed to be monolithic, with Torque, I dont have to treat the
data schema more carefully or with any more persistance, than any other
layer in the system. The Torque code generation is the main reason for this.


I have had a look at OJB, but as there wasnt any code generation abilities
that I could determine in the same manner as Torque, I saw no personal
reason to switch. Torque is working extremely well for us and I am thankful
to all the people who have worked on Turbine and Torque, it has allowed me
to meet my customers demands with a superior solution. 

Anyway, I thought those on the list might be interested why I use Torque and
why I consider it the best Data Mapping tool currently around for my needs.
I have no opinion on whether it should be dropped or not, but we will be
continuing to use Torque for the above stated reasons. For our purposes
Torque is still the best solution.



Cameron Riley

--
To unsubscribe, e-mail:   <mailto:turbine-dev-unsubscribe@;jakarta.apache.org>
For additional commands, e-mail: <mailto:turbine-dev-help@;jakarta.apache.org>

Reply via email to