Thanks. Kind of a picky failure. Funny things happen with indefinites. I
found it as I was thinking about the idea of inplace updates with scalar
primitives. My concern was to see which ones could fail part-way through
processing the verb. Wouldn't want a verb to update part of an array when
it encounters a failure leaving the noun partially updated. That is one of
the nice things about the way J handles arrays and errors. It either
updates the whole noun result or it is left unchanged.

On Sat, May 14, 2016 at 9:10 PM, Marshall Lochbaum <[email protected]>
wrote:

> Yes. Complex gcd checks that neither of the inputs are infinite but does
> not test for NaN. A pretty easy fix.
>
> Marshall
>
> On Sat, May 14, 2016 at 04:07:42PM -0600, Don Guinn wrote:
> > The following statement hangs:
> >
> >    +.~1 2 1j_. 3
> >
> > Hangs in j64_804, j64_805 and J6 32 bit under Windows 10
> > ----------------------------------------------------------------------
> > For information about J forums see http://www.jsoftware.com/forums.htm
> ----------------------------------------------------------------------
> For information about J forums see http://www.jsoftware.com/forums.htm
----------------------------------------------------------------------
For information about J forums see http://www.jsoftware.com/forums.htm

Reply via email to