Bernie,
I'm quite sure this is not a change in the way two versions of R:Base
run. It's just a difference in what programs ran (or didn't run) before your
tests in 6.5++, vs. what programs ran before the problem code in 5.5.
Every single variable in every program that ever gets tested in an IF or
WHILE statement should be declared in a SET V statement. This has
been true since Razzak was a boy. The program you are looking at did
not do that. Sooner or later, it would have bitten you in 5.5, too. (And it
probably did dump somebody at an R> prompt, but they never figured
out why.)
Bill
On Thu, 9 Aug 2001 21:43:17 -0400, Bernard Lis wrote:
>This worked in 5.5 Did 5.5 create the variable when a select clause
fails? Should 6.5++ do the same?
>
