Re: Why all the max length constraints?

Tony D wrote:
> Dawn wrote :
>
> > That I know. That does not answer my question. I fully understand
> > that one need not specify max length constraints if implementing only
> > "in theory." I wondered if there were anything in the theory that
> > might prompt implementations that have, perhaps as a side-effect, the
> > need for or advantage when individual attributes are specified with max
> > length constraints.
>
> If you know that, then you know the answer to your question,

If I know it, I don't know that I know it.

> which
> means there was no need to ask it,

I wanted to know the answer and still do

> which makes it a silly question.
> I'll say it again: the relational model is orthogonal to type.

I can say that too. That is not my question. My question cannot be answered within the theory, it requires straddling theory (which one could see as one grouping of requirements for a DBMS product) and application of the theory. Requirements do not dictate design, but they can encourage different aspects of design. The simplest design to meet a set of requirements might lead us to select this feature or constraint in a system. If we add in a requirement that such a feature or constraint should not be there, then the design would be different.

So, maybe (and I really don't know) the RM as a set of requirements leads to a design where max length constraints on attributes are the easiest way to implement these requirements. So, if we don't add in the requirement that we do not want any penalty for variable length attributes, tossing that in wtih our RM requirements, then we are most likely to end up with the fixed max lengths.

Did that make any sense?

> That
> means the relational model makes no demands and imposes no restrictions
> on the types available (apart from the obvious requirement for a
> boolean type). It really is that simple. There really is nothing to see
> here.
>
> > It hesitate to report that I still don't know the answer to my
> > question, but I will research it differently and see if I can get
> > further.
>
> I think you do know the answer to your question, but the penny isn't
> dropping.

It is possible, but until I know I know it, I don't know it.

> > It actually thought it would have a simple answer and that
> > someone would have that answer handy enough to pass it along, but it
> > seems you and others misinterpreted the question.
>
> It does have a simple answer. That simple answer has been given on
> several occasions. Perhaps equally you aren't understanding or are
> misinterpreting that simple answer ?
>
> > I wish I knew why, but I'm starting to suspect that others do not have a framework in
> > which they could easily understand my actual questions.
>
> This sentence is perhaps more educational than you truly intended.

I don't know how you took it, but I know that I do not have the same background as most of the cdp readers and writers, given that I do not have a degree in CS and have never been a DBA. I have also spent far more hours and years working with non-SQL DBMS's than SQL DBMS's, although I have a little experience as well as reading knowledge of the latter. So, I know I have to figure out how to ask questions in the language most likely to be understood so my questions can be answered or opinions interpreted correctly (I'm clearly failing at this right now). But perhaps this question has an additional problem. It might be the case that whatever framework (combination of terms, experience, and perspective) I am using that leads me to such a question is not one that is common nor would it be easy for someone without a similar framework to understand. So, it gets interpreted incorrectly.

> > I'm starting to see that those who understand my questions are the ones who do not
> > get angry with me.
>
> Who's angry ? I'm certainly not;

Good. Thanks.

> I'm not even exasperated at having
> seen a simple question answered simply by several people, yet you still
> don't seem to "get it". Ho-hum.
>
> > If you have any interest in doing me a favor, Tony, I would really,
> > really appreciate it if you could try to understand what I am asking
> > and let me know if you are getting closer to understanding that.
>
> I think I understand well enough what you're asking. I will, at risk of
> muddying the waters considerably, suggest that you read up on the
> language Euler, which was invented by Niklaus Wirth, and consider the
> difference between the proposed language and the language that was
> actually implemented;

I am familiar with Euler. I don't see the connection.

> or, consider the difference between the Algol-68
> report's definition of modes and the implementation that most compilers
> actually managed.

This is a different issue, it seems. This might be relevanant to the fact that there is no pure implementation of the RM. I'll need more clues if you really think this is relevant to my question.

> This might help put the difference between what a
> model proposes and what implementers implement (and why) into
> perspective.

I find such stories interesting, and I can see some relevance to the topic, but this will not answer my question. I have no doubt at all that the RM does not dictate that every attribute have a max length constraint. I'm wondering if there are requirements that drive decisions in this area when implementing the RM where those vendors not starting with the RM in their original requirements might not have been driven to similar decisions.

> > It is no doubt an ignorant question but I'm quite sure it is not a stupid
> > question.
>
> I would suggest "irrelevant" and "unnecessary" as alternatives to
> "stupid". Or just plain "silly".

Am I getting any better at explaining my question or am I still getting nowhere? --dawn Received on Mon May 29 2006 - 11:15:30 CEST