RE: dired-mark

[Drew Adams]

> > But what if you really intend to mark "." or ".."?
>
> That's precisely the question.

Yes.

> I can't imagine any case I would need it. However, as
> experience tells our imagination runs short occasionally...

It doesn't matter if you can't imagine a case where you would need it.

As long as users can operate on marked files and directories, someone will
want to, well, mark `.' or `..' and then operate on it.

Madame Lambda or Mister X might define a command that prints each marked
file and all files in each marked directory. Or search them. Or do whatever
to/with them. And s?he might use that command on `.' or `..' as well as on
other marked directories.

There is no reason that users shouldn't be able to use `dired-mark' to mark
`.' and `..'. If you don't want to mark them, then don't. If you have code
that does something to all marked files and dirs, but you don't want to
treat `.' and `..', then exclude those in your special-purpose code.

Wrt subdirs, it is only when point is on a subdir header line that
`dired-mark' does not mark the subdir's `.' or `..' - in that case, it marks
all files and directories _contained_ in the subdir. `dired-mark' always
marks `.' and `..' whenever they are targeted (e.g. cursor on that line).

That's TRT, IMO. Acting on a directory is not (necessarily) the same thing
as acting on everything in it. It is useful to be able to mark and act on
any directory, including `.' and `..'.