The previous behavior wasn't a bug. The new special value of 'marked'
was introduced in support of the new user option
dired-mouse-drag-files, which needs to distinguish between the two
cases. Since the special value 'marked' is new in Emacs 29, and used
only in a couple of places, I don't understand why you are saying that
it's "a pattern that's used in many other places". How could that be
true?
The pattern I mentioned is that of a function modifying its behavior
if it receives a "special distinguished value" for an argument; for
example, an argument that is usually an integer may also be some
special symbol value.
Look, long story short, when the first response said the report was
"not a doc bug," I interpreted that to mean that the current docstring
is correct, and dired-map-over-marks is not meant to do anything
special if given the symbol "marked" for ARG, since the docstring
doesn't mention it. The second response still seemed somewhat
ambiguous to me, since it offered a function my-marked-files-list
that I could use instead of dired-get-marked-files. Why would I
want or need that, if the functionality of dired-get-marked-files
weren't going to change?
In any case, I've updated (on the emacs-30 release branch) the doc
strings of the relevant functions with the effect of the special value
'marked'. I think this is all we need to do in this bug.
Yes! Thank you! That is exactly what I wanted and expected to
happen when I filed my bug.