On Mon, 2 Nov 2020 at 03:34, Eli Zaretskii <eliz@gnu.org> wrote:

Thanks, but I'm not interested in working around installation problems
of programs we invoke.  It's a slippery slope, with never-ending
additional requests we will have to honor.

The reporter and I seem to have swapped positions on this!

I believe that the ispell.el code (which I wrote) is buggy: it should not be incorporating warnings into its output.

Also, the patch I offered is a simplification of the original code. So, I don't think we are losing here.

Eli, I quite agree with your sentiment, and I would certainly not advocate installing a workaround in Emacs unless there were compelling reasons. However, I do not see this as a workaround, and as it is also a simplification, I don't see a problem.

--
https://rrt.sc3d.org