Re: Why is it so difficult to get a Lisp backtrace?

[Alan Mackenzie]

Hello, Stefan.

On Sat, Jun 25, 2022 at 17:30:30 +0200, Stefan Kangas wrote:
> Alan Mackenzie <acm@muc.de> writes:

> > I mean something like the following, which to a first approximation,
> > works.  To use it, do M-x debug-next-command and the execute the
> > command expected to give errors:

> Good idea, but how about just making it into a manual toggle?  I might
> need to run more than one command to reproduce a bug.

That's another idea.  It has snags, like for example, you set the toggle,
use it, and forget to cancel it.  You then amend one of the controlling
variables (like debug-on-quit) manually, then cancel the toggle, at which
point confusion and irritation happen.

One idea I had was to have an optional numerical argument, saying how
many commands to leave the thing in place for.  Yet another, conflicting,
idea would be to have an optional boolean argument to exclude
debug-on-signal from the mix, so as not to get too many "expected"
backtraces.

-- 
Alan Mackenzie (Nuremberg, Germany).