Re: Emacs 27.1 released

[Eli Zaretskii <eliz@gnu.org>]

> From: Mattias Engdegård <mattiase@acm.org>
> Date: Thu, 13 Aug 2020 21:50:33 +0200
> Cc: emacs-devel@gnu.org
> 
> 12 aug. 2020 kl. 21.21 skrev Eli Zaretskii <eliz@gnu.org>:
> 
> > I just counted its callers, that's all.  A change in a function that
> > has many callers could have unintended consequences beyond the
> > function itself.
> 
> I'm very happy that you are taking an interest in Calc again, and especially that you took the trouble of digging into the ramifications of this little correction! I'm most curious to hear what you found out. You see, when the change to calcFunc-gcd was made, naturally I did look at its callers carefully to assess the impact. Now since you brought it up I've done so again, and found nothing new: all users of that function are either indifferent to the change, or actively benefit from it.
> 
> It seems clear that the code is less buggy now than before the change and should therefore be preferred for the sake of correctness. However, I could be wrong, and would very much benefit from your findings so that my mistakes can be fixed on master!

We are assessing this from two very different points of view: you are
thinking about fixing the bug, and I'm thinking about the dangers of
some unintended consequences of the fix destabilizing Emacs 27.2's
version of Calc.  By their very nature, unintended consequences are
unknown to you and to me, so asking for any specific details is not
useful.

(And could you please drop the thinly-veiled sarcasm, here and
elsewhere?  It makes it harder for me to discuss serious issues with
you, and I don't think I deserve the implied attitude, while doing my
job of keeping the release branch stable.)