Re: So what's null then if it's not nothing?
Date: Tue, 13 Dec 2005 15:34:41 GMT
Message-ID: <lSBnf.98945$ki.37687_at_pd7tw2no>
David Cressey wrote:
> "Jon Heggland" <heggland_at_idi.ntnu.no> wrote in message
> news:MPG.1e07ca1b45bb3a9f989744_at_news.ntnu.no...
>
>
>>I don't see why '<' is something to get worked up about. Why is it bad >>to define some (arbitrary) order for truth values? As for defining a >>strange '=', they do so for the other domains (and I think Codd does it >>too), and Codd's article doesn't mention Kleene either. It seems strange >>if Codd didn't bother to correct the SQL committee if it got things so >>very wrong.
>
>
> Strongly agreed.
>
> In Pascal, where FALSE < TRUE, you can use all the comparison operators as
> Boolean operators.
> In particular, <= can be used in place of "implies". In other words, if
> A implies B then A<=B will be true.
>
> People get worked up about order in this forum, because they think that the
> next point you are going to make is that lists are more expressive than
> sets.
>
>
Speaking only for myself, I get worked up about the last comment. I thought it was just that (most of) the RM foundation doesn't depend on order (at root some definitions do seem to depend on ordered pairs and triples).
I think it is more that supporting an ordering operator is up to the domain in question, not the RM. I think some people mean that one ought to get the same logical result (ie. same relation) from two programs regardless of ordering but this doesn't forbid an implementation from imposing an order on all domains (easier indexing) or always presenting a result in some implicit order.
p Received on Tue Dec 13 2005 - 16:34:41 CET