Dr. Frederick Wolfe
> A little more on this. xi: variables are automatically
> dropped. If you want to keep them, then they need other
> names. r() are discarded.
I'm glad Fred and I seem to be converging rapidly here.
On the principle that the exception proves the
rule -- meaning strictly that the exception _tests_
the rule, as in the original Latin -- how far is the
practice of -xi- consistent here?
First, the documentation of -xi- clearly warns that
its products get over-written.
Second, -xi- variables contain information deducible
from other variables. So long as those variables remain,
they can be re-created willy-nilly.
Hence this situation is pretty close to the opposite
extreme, I think, and no precedent for -merge-.
r-class results, again, are advertised as temporary
results for the user's convenience, which can be
re-created at will, so long as the data remain the
same.
Nick
[email protected]
<<attachment: winmail.dat>>
