~/.emacs vs ~/.emacs.d/init.el

~/.emacs vs ~/.emacs.d/init.el

[Juanma Barranquero]

With the recent introduction of `user-emacs-directory', I've been
wondering... Could someone please clarify the relation between these
things:

 --user                   (option for Emacs)
  init-file-user
  ~/.emacs.                (also _emacs, .emacs.el, _emacs.el)
  ~USER/.emacs             (is that possible on Windows?)
  ~/.emacs.d/init.el
  ~USER/.emacs.d/init.el   (is that possible at all?)

I've read the info nodes "Init File", "Find Init" and "Initial
Options"; I've also looked at the code. I'm still not sure whether

  emacs --user=another

should make ~another/.emacs.d/init.el possible, though the code
suggests that the answer is not (~/.emacs.d/ was hardcoded, and
`user-emacs-directory' does not change that), and I'm not sure why
(conceptually speaking).

             Juanma

Re: ~/.emacs vs ~/.emacs.d/init.el

[Eli Zaretskii]

> Date: Wed, 13 Jun 2007 12:00:32 +0200
> From: "Juanma Barranquero" <lekktu@gmail.com>
> 
>   ~USER/.emacs             (is that possible on Windows?)

No, we don't support ~USER on Windows yet (although it can be done on
the Windows versions since W2K).  Volunteers welcome.

Re: ~/.emacs vs ~/.emacs.d/init.el

[Juanma Barranquero]

On 6/13/07, Eli Zaretskii <eliz@gnu.org> wrote:

> No, we don't support ~USER on Windows yet (although it can be done on
> the Windows versions since W2K).  Volunteers welcome.

I wouldn't dismiss the idea of implementing it myself, once the task
is clear enough.

What about my other questions? Should --user=myuser influence whether
init.el is found (I mean, ~myuser/.emacs.d/init.el), or is ~/.emacs.d/
always valid only for the current user?

             Juanma

Re: ~/.emacs vs ~/.emacs.d/init.el

[Eli Zaretskii]

> Date: Wed, 13 Jun 2007 19:34:00 +0200
> From: "Juanma Barranquero" <lekktu@gmail.com>
> Cc: emacs-devel@gnu.org
> 
> On 6/13/07, Eli Zaretskii <eliz@gnu.org> wrote:
> 
> > No, we don't support ~USER on Windows yet (although it can be done on
> > the Windows versions since W2K).  Volunteers welcome.
> 
> I wouldn't dismiss the idea of implementing it myself, once the task
> is clear enough.

The task is to map ~USER to

  C:/Documents and Settings/USER/Application Data

For this, we need to extend w32.c:getpwnam to support not just the
current user, but also the other users on the local machine (ask me
about the relevant Windows APIs if you need help finding them), and
then it should Just Work (the actuall expansion happens in
expand-file-name and other similar primitives defined on fileio.c).

> What about my other questions? Should --user=myuser influence whether
> init.el is found (I mean, ~myuser/.emacs.d/init.el), or is ~/.emacs.d/
> always valid only for the current user?

I don't have anything intelligent to say on that, I'm confused myself.

Re: ~/.emacs vs ~/.emacs.d/init.el

[Juanma Barranquero]

On 6/13/07, Eli Zaretskii <eliz@gnu.org> wrote:

> The task is to map ~USER to
>
>   C:/Documents and Settings/USER/Application Data
>
> For this, we need to extend w32.c:getpwnam to support not just the
> current user, but also the other users on the local machine (ask me
> about the relevant Windows APIs if you need help finding them), and
> then it should Just Work (the actuall expansion happens in
> expand-file-name and other similar primitives defined on fileio.c).

OK, I'll put this issue in my ToDo list.

> I don't have anything intelligent to say on that, I'm confused myself.

Glad to know I'm in good company, then ;-)

             Juanma

Re: ~/.emacs vs ~/.emacs.d/init.el

[Davis Herring]

> I've read the info nodes "Init File", "Find Init" and "Initial
> Options"; I've also looked at the code. I'm still not sure whether
>
>   emacs --user=another
>
> should make ~another/.emacs.d/init.el possible, though the code
> suggests that the answer is not (~/.emacs.d/ was hardcoded, and
> `user-emacs-directory' does not change that), and I'm not sure why
> (conceptually speaking).

It certainly should!  --user picks a _directory_ (the home directory of a
user), and .emacs, .emacs.d/init.el, etc. are all conceptually relative
filenames in that directory.  I hate to duplicate the
`user-emacs-directory' variable (or whatever it ends up being called), but
those filenames should really get `expand-file-name'd against a home
directory chosen by that option (or by $HOME otherwise).

Davis

-- 
This product is sold by volume, not by mass.  If it appears too dense or
too sparse, it is because mass-energy conversion has occurred during
shipping.

Re: ~/.emacs vs ~/.emacs.d/init.el

[Juanma Barranquero]

On 6/21/07, Davis Herring <herring@lanl.gov> wrote:

> It certainly should!  --user picks a _directory_ (the home directory of a
> user), and .emacs, .emacs.d/init.el, etc. are all conceptually relative
> filenames in that directory.

Yes, that's my interpretation, too. But ~/.emacs.d/ appeared at least
seven years ago (there's a reference in the ChangeLog from
2000-04-24). It's hard to believe nobody has worried about -u in all
that time. <conspiranoia>Something Else Is At Work...</conspiranoia>

> I hate to duplicate the
> `user-emacs-directory' variable (or whatever it ends up being called), but
> those filenames should really get `expand-file-name'd against a home
> directory chosen by that option (or by $HOME otherwise).

The Function Without a Name can take care of that easily, I think.

             Juanma