Re: Queries and OO
From: Dmitry A. Kazakov (mailbox_at_dmitry-kazakov.de)
Date: 01/17/05
- Next message: Mark Nicholls: "Re: up front designs always useless"
- Previous message: Ilja Preuß: "Re: up front designs always useless"
- In reply to: frebe: "Re: Queries and OO"
- Next in thread: Daniel Parker: "Re: Queries and OO"
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
Date: Mon, 17 Jan 2005 10:38:52 +0100
On 16 Jan 2005 09:30:18 -0800, frebe wrote:
>> I'd say that querying is rather about some specific implementation than
>> modeling.
>
> Lets say we have an application for order administration. Wouldn't it
> be necessary to be able to find orders by customer, by delivery date,
> by status, by product and combinations of these criterias? Not having a
> query mechanism would force you to have a lot of maps and sorted sets
> containing the customer objects. And when combined searches are done,
> decisions about in what order to use the containers have to be made.
I do not argue that there might be no applications requiring this.
> Queries is a way to decouple how objects are mapped into a container,
> from how the objects are retrieved from the container. When you are
> making software, new feature request forces you to retrieve objects in
> new and different ways all the time. A hard-wired solution with
> pre-defined search paths is mostly not possible.
I do not see it so. In my view queries is an interface of a container. No
more than that.
>> I might need no queries even if I do use a database.
>
> True, long time ago there were database types without query
> capabilites. It was possible to make working applications in that era
> too. But maybe the query capabilities was one of the reasons behind the
> success story of relational databases.
I'd cynically say that the success should be rather attributed to the poor
performance of RDBS. You know, people prefer poorly performing things.
Consider Windows and UNIX as examples. When you have some technical
problem, you are not responsible for, then you can always explain where you
had pulverized money and resources...
>> What could be so intriguing in a set of tables?
>
> Many application vendors find relational databases and tables an
> extremly important component in their solution.
Right, bigger servers, more staff, more resources. It is very attractive
for any manager...
> If we want the application to collaborate with a database, we should model this too.
What is the problem the application solves? Is the problem collaboration
with a DB, or it is DB used to solve a problem?
-- Regards, Dmitry A. Kazakov http://www.dmitry-kazakov.de
- Next message: Mark Nicholls: "Re: up front designs always useless"
- Previous message: Ilja Preuß: "Re: up front designs always useless"
- In reply to: frebe: "Re: Queries and OO"
- Next in thread: Daniel Parker: "Re: Queries and OO"
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
Relevant Pages
|