Re: Specifying all biz rules in relational data
From: Kenneth Downs <firstinit.lastname_at_lastnameplusfam.net>
Date: Tue, 14 Sep 2004 14:05:17 -0400
Message-ID: <usb7ic.6p9.ln_at_mercury.downsfam.net>
>
> The answer can always be yes if you want. Algorithms are usually
> written in a source (text)file, which is data. The algorithm could
> also be expressed in a object model, in a xml file or even a
> relational model. But, it is not sure that it would be more readable
> or easy just because it is described in another way, then a source
> file. For some specialized areas - database querying, mathematical
> calculations, etc - behaivor can be expressed very successfully in a
> declarative way. But I am still waiting for a general all-purpose
> declarative language to come. There have been some partial successfull
> attempts, like Prolog, but like I said, I am still waiting. I you have
> a revolutionary idea, please tell me.
>
> /Fredrik
Date: Tue, 14 Sep 2004 14:05:17 -0400
Message-ID: <usb7ic.6p9.ln_at_mercury.downsfam.net>
Fredrik Bertilsson wrote:
>> The real question is, can the algorithm >> be expressed purely in data? My claim here is that it can, and moreover, >> that it is worthwhile to do so.
>
> The answer can always be yes if you want. Algorithms are usually
> written in a source (text)file, which is data. The algorithm could
> also be expressed in a object model, in a xml file or even a
> relational model. But, it is not sure that it would be more readable
> or easy just because it is described in another way, then a source
> file. For some specialized areas - database querying, mathematical
> calculations, etc - behaivor can be expressed very successfully in a
> declarative way. But I am still waiting for a general all-purpose
> declarative language to come. There have been some partial successfull
> attempts, like Prolog, but like I said, I am still waiting. I you have
> a revolutionary idea, please tell me.
>
> /Fredrik
Revolutionary, no. Or at best, that would be for others to say, but I have no pretensions in that direction.
My contention is that the best form for specifying a database app is itself data, because I can use that to create docs, code, tables, etc. It is the gift that keeps on giving.
Seems where we disagree is whether it is worthwhile to do this. I will let you know when I have 0.1 ready to publish, and I will check out your site in the meantime.
Thanks for the chat.
-- Kenneth Downs Use first initial plus last name at last name plus literal "fam.net" to email meReceived on Tue Sep 14 2004 - 20:05:17 CEST