-----maxima-bounces at math.utexas.edu wrote: -----
>Yeah.?I?cut?it?out?in?a?wider?revision?of?the?summation?code?in
>Nov?2005.?I;m?pretty?sure?I?got?the?idea?from?else?(i.e.?don;t
>blame?me,?or?don't?blame?only?me,?anyway).
I was on belay for parts of this revision, and I wrote some additional
regression tests for the new summation code. Also I may have authored
a few hunks of the new summation code. I'm pretty sure that the
pre-Nov 2005 code had some ugly bugs and that it would fail quite a
few of the new tests. So the revision fixed got many many things
right. The summation code is bewilderingly (possibly overly) complex,
so that's quite an accomplishment. Not that it really matters, but I
don't recall the history of expunging sumhack/prodhack.
Barton