Thomas Lockhart wrote:
>
> Congrats Tom for rising to the challenge ;)
>
> istm that a brute force hack such as is being suggested should be left
> as an exercise for the reader.
>
> These "table links" seem to controvert the ability for a RDBMS to mix
> and match tables in ways which are not hardcoded beforehand.
Excuse me, in what way? All the usual features of SQL are in their
legal place. What you are getting is one more additional ability,
not constraint.
> Regardless of whether "there exist some real servers that offer such
> features I am talking", a departure from the relation model in a
> relational database is likely to lead to undesireable constraints and
> restrictions in our future development.
Until you offer an example of such constraint these words are
groundless fear.
Think of it not merely as index lookup speedup, but as quick and
clever way to fix the optimizer. As we have seen, optimizer now
has troubles choosing the fastest way of doing the query. This
is deep-rooted trouble, because optimizer generally needs some
fine-graded statistics on tables which it is working on, and
gathering such statisctics would require, I am afraid, total
rewrite of the optimizer. In the case of links quiery is always
done the best way.
--
Leon.
