Re: S.O.D.A. database Query API - call for comments

On Thu, 3 May 2001 01:42:06 +0200, "Carl Rosenberger" <carl_at_db4o.com> wrote:
>> Step-by-step with many small functions, is the old way. Wrapping that
>> around objects, is old wine in new bottles. You may like old wine,
>> but the new bottle is unlikely to improve it.
>
>O.K. those are the "have done that, been there" old man arguments.
>Where are the technical ones?

Fundamentals of Database Systems, C.J. Date. You have to ask?

>Is SQL the end of the road forever?
>Are you so sure that we are not doing some things different?
>Did you take a look at all?

I looked. Been there, done that. Know better now.

I've seen different projects do stuff like this a bunch of times. It usually is done by object guys who don't know squat about relational. I've been careful not to say that that's your situation, only asked what you were doing about views and joins. I'm unclear from what I've read, and from your answers, to what degree you understand the role that these play in databases.

SQL can and should and is being expanded in various directions, I'm all for user-extensible SQL and replacing tables by classes, deriving new classes, providing a clean object/relational interface. I'm not sure who, if anyone, ships that today, but surely it's on its way.

>I am 100% positive that we provide the simplest way to *store* objects,
>which was definitely not possible in the same way 15 years ago. Programming
>languages just did not have features like reflection.

If you're worried about storing and retrieving objects, well and good, but you seemed to be proposing this as a better way to query even normal relational info. That is what I have responded to.

>Why should we not come up with a likewise good approach to query objects?
>We only started the discussion.

Those who do not learn from history are doomed to repeat it.

>Evolution always has something to do with picking up the good parts in old
>ideas.

I'm all for that, but it does not seem to me the old ideas have been sufficiently looked at.

Joshua Stern
JRStern_at_gte.net Received on Sat Jul 21 2001 - 20:00:42 CEST