That's wrong. And clicking mouse-2 on that button takes me to that
wrong file. And so does `C-h f' followed by the name and clicking on
the library, which is shown as bookmark.el.
The debugger correctly traverses the code of the version of the function
that is defined in bookmark+-bmu.el, however.
IOW, the function is defined correctly. It is only the links in Help
and the debugger that are wrong, as well as the file name in Help. But
the doc string in Help is correct - it is the doc string for the
Bookmark+ version of the function.
The redefinition of the function in bookmark+-bmu.el is a
In GNU Emacs 188.8.131.52 (i686-pc-mingw32)
of 2014-10-20 on LEG570
Bzr revision: 118168 [hidden email]-20141020195941-icp42t8ttcnud09g
Windowing system distributor `Microsoft Corp.', version 6.1.7601
`configure --enable-checking=yes,glyphs CPPFLAGS=-DGLYPH_DEBUG=1'
> In my setup, i.e., with Bookmark+ loaded, function `bookmark-bmenu-list'
> is defined in file bookmark+-bmu.el. In the debugger, there is a line
> like this:
> * bookmark-bmenu-list(nil)
> with the function name highlighted with a button. `C-u C-x =' on the
> name shows this:
> Here is a `help-function-def' button labeled `bookmark-bmenu-list'.
> There are text properties here:
> button (t)
> category help-function-def-button
> help-args [Show]
> Clicking [Show] shows this:
>> Do you have a recipe not involving your own code?
> The point is that the help button does not reflect the fact
> that the command was redefined.
> The redefining code (in bookmark+-bmu.el, at top level) is
> simply this:
> (defun bookmark-bmenu-list (&optional filteredp msg-p)
> (interactive "i\np")
> The problem is with the button link - it has apparently
> hard-coded the address of the original definition of
> this command, in standard library bookmark.el.
Could you please give a recipe for how to reproduce this bug using
>> Could you please give a recipe for how to reproduce
>> this bug using vanilla Emacs?
> That bug was reported 5 years ago, for Emacs 25.
> It seems that it's been fixed, in Emacs 26.3 at
> least. So I guess the bug can be closed. Thx.
Thanks for that, and I'm therefore closing this bug.