Re: Impossible Database Design?

From: x <x_at_not-exists.org>
Date: Wed, 17 May 2006 12:07:40 +0300
Message-ID: <e4ep0i$v22$1_at_emma.aioe.org>


"Nikolai Onken" <nikolai.onken_at_gmail.com> wrote in message news:1147855888.068190.203250_at_j55g2000cwa.googlegroups.com...
> > Ask your client what events will want to store in the computer and what
is
> > the best way for him to input that data.
>
> Hehe :) the client wants:
> - 'infinitely repeating dates' (whatever thats supposed to mean in the
> end)

How he would input that data into the computer ? Have you asked him ?

> - being able to predict whether courses overlap
> - having exceptions to series (every monday except xxx)

So what. "My client is my master."
Ask him to provide some examples.
My guess that you will only have to solve some equations. The method you will use will depend on the specific case you have. Experiment with several approaches and ask your client which one he will prefer and if he has the money to pay you for implementing that specific solution.

> My conclusion would be that probably the best way would be to simply
> not support queries for predicting overlaps for dates without ending
> point.

See above.

> This probably makes most sense anyways since the only type of never
> ending event is something like a birthday or memorial or so.. and there
> you probably don't really need to know whether it overlaps with another
> never ending event..

Nothing is never ending. Everything is transformation. :-) Received on Wed May 17 2006 - 11:07:40 CEST

Original text of this message