bug#63988: 30.0.50; Recent header line format changes cause spin/seg fault with format-mode-line

[Spencer Baugh <sbaugh@janestreet.com>]

Eli Zaretskii <eliz@gnu.org> writes:
>> From: Eshel Yaron <me@eshelyaron.com>
>> Cc: aaronjensen@gmail.com,  63988@debbugs.gnu.org,  Stefan Monnier
>>  <monnier@iro.umontreal.ca>
>> Date: Sat, 10 Jun 2023 12:07:22 +0300
>> 
>> > So maybe we should declare this feature a failed experiment and remove
>> > it?
>> 
>> FWIW I think that make sense, but IMO it'd be best to only remove the
>> treatment of `:eval` in `window_wants_header_line`, and keep the new
>> treatment of `header-line-format` being a cons cell with a void or
>> nil-valued variable car.  That's still useful because it works well with
>> minor mode variables, and it's less risky as it doesn't involve
>> evaluating arbitrary lisp, just inspecting a variable.
>
> Why would it make sense to leave the non-nil car case?
>
> It sounds like the consensus here is that indeed this feature is not
> worth the complications, and so, unless I hear some good reasons not
> to do so, I intend to delete it in a week's time.

Ugh, I am only just now seeing this... I would have appreciated a CC
since I'm the one who requested this feature, or maybe a mail sent to
the original bug report...

Anyway, the reason for this feature remains what is described in
bug#63825: displaying a header line only when there is information to be
displayed in the header line.

Concretely, for which-function-mode: it is nice to have
which-func-format displayed in the header line, but in some kinds of
buffers it does not have any information to display.  When a window is
displaying such a buffer, I'd prefer to not have a header line, because
the header line will just be empty.

This use case has no need for the full-fledged :eval mode of this
feature.  It works fine with the non-nil car case.

How else should I achieve this?

I could add this functionality to which-func-mode directly: it could
learn to set header-line-format buffer-locally only in buffers that
support which-func-mode.  Would that be acceptable?  If so I'll send a
patch for that.