Re: CVS commit: src/games/factor

[Robert Elz <kre%munnari.OZ.AU@localhost>]

    Date:        Wed, 19 May 2010 02:52:21 +0000
    From:        David Holland <dholland-tech%NetBSD.org@localhost>
    Message-ID:  <20100519025221.GA3784%netbsd.org@localhost>

  | I already proposed a resolution that nobody disagreed with, but Joerg
  | decided to ignore it, and I'm tired of arguing.

I did, I just didn't say anything about it, as factor just isn't that
important (it is /usr/games/... after all...)

I also disagree with Joerg's solution, and what Aleksej has been proposing.

My suggestion remains as it started, put the code in factor (with respect
to this issue) back the way it was, cease all the arguments, and just
document how it behaves (ie: put the fix in factor.6 instead of factor.c)

Correct or not, by whoever you pick's definition of what is correct,
just doesn't matter - not for these data values, as no-one who is
going to use /usr/GAMES/factor is going to ever ask to factor 0, or 1
(or for that matter, anything else < 10, and most probably < 100) so it
simply doesn't really matter what would be printed if they did.

I mean, go hunt for all the bug reports about how factor has (for the past 20 
years) treated 0 ?   How many can you find?   Why did anything ever need 
fixing?

All those of you doing research in various areas of number theory need
better tools, and ones that suit your preferred set of axioms.

kre