bug#69684: Functionality of Fbare_symbol has been lost.

[Alan Mackenzie]

Hello, Eli.

On Sun, Mar 10, 2024 at 07:57:26 +0200, Eli Zaretskii wrote:
> > Cc: acm@muc.de, Paul Eggert <eggert@cs.ucla.edu>
> > Date: Sat, 9 Mar 2024 23:23:50 +0000
> > From: Alan Mackenzie <acm@muc.de>

> > Hello Paul, hello Emacs.

> > Since a recent commit, Fbare_symbol (in src/data.c) no longer works on a
> > symbol with position when symbols-with-pos-enabled is nil, instead
> > signalling an error.

> > This is due to the new CHECK_SYMBOL (sym); statement in Fbare_symbol,
> > which wasn't there before.

> > Since I merged master into my development branch, I can no longer build
> > it because of this change.  A rapid reversion to the previous
> > functionality would be appreciated.  :-)

> Couldn't you fix this on your branch by (what sounds like a trivial)
> change in Fbare_symbol?

I've done that already here, of course.  But I was wanting to make a
commit yesterday evening, which of course I can't with this bug
unresolved.  Paul is quite particular over the exact formulation of these
macros and inline functions, which have an important influence on Emacs's
speed.  So I thought I would give him a chance to fix it neatly first.

> Btw, why is your branch special in this regard?

It's the branch where I'm implementing position information in doc
strings, using symbols with position to get this information.  This is
bug #67455.  Progress is pretty much at the stage where we can start
discussing the exact presentation of the information in backtraces, etc.

-- 
Alan Mackenzie (Nuremberg, Germany).