bug#66357: 29.1; malformed load-history entry when built --with-ns

[Aaron M. Ucko]

Thanks for the quick reply, and sorry for not following up earlier.

Eli Zaretskii <eliz@gnu.org> writes:

> I cannot reproduce this (but I'm not on macOS).

I haven't yet been able to elsewhere either, even with GNUstep libraries
installed and Emacs configured to use them.

> My guess is that you
> have packages installed via package.el, and some of your init files
> are executed (and load package.el) even though the .emacs is commented
> out.

I do indeed have some packages installed, and it looks like their
general presence affects Emacs's startup logic.  I'm not sure how much
the precise set matters, but FTR, I specifically have

  $ ls ~/.emacs.d/elpa/
  \@-readme.txt                     groovy-mode-201203310931/
  aggressive-indent-readme.txt      jq-mode-20220610.1119/
  applescript-mode-20210802.1715/   kotlin-mode-20210917.1911/
  archives/                         kotlin-mode-20230123.1859/
  asn1-mode-20170729.226/           ldap-mode-readme.txt
  auctex-13.0.16/                   magit-20220131.125/
  compat-29.1.3.4/                  magit-20230221.648/
  crontab-mode-20210715.133/        magit-section-20220130.2007/
  dash-20210826.1149/               magit-section-20230213.2018/
  dash-20221013.836/                protobuf-mode-20230216.1713/
  git-commit-20220130.2254/         s-20220902.1511/
  git-commit-20230220.1901/         transient-20220130.1941/
  gnupg/                            transient-20230220.1425/
  gradle-mode-20150313.1905/        with-editor-20220130.1942/
  graphviz-dot-mode-20230213.1945/  with-editor-20230212.2214/

>      Each element in this alist describes one loaded library (including
>      libraries that are preloaded at startup).  It is a list whose CAR
                                                               ^^^^^^^^^
>      is the absolute file name of the library (a string).
       ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

The offending entry is missing a file name (or even a leading nil
placeholder), and consists *solely* of two

>      ‘(require . FEATURE)’

forms, which are valid only later in the list.

-- 
Aaron Ucko <ucko at ncbi>, NCBI C++ Toolkit core development group