Re: proofreading man/dired.texi

Re: proofreading man/dired.texi

[Richard Stallman]

    I have attempted to proofread man/dired.texi.

    L22: dired-x is mentioned. I think "wdired" should be mentioned as
    well.

    L16: Not consistent with the "wdired" minor mode, in which characters
    do insert.

If wdired is inconsistent with dired, I would rather change wdired.
It is less well known and less often used.

Why does wdired allow self-insertion?
Is there a good reason for it?
If not, it would seem to be a bug.

Re: proofreading man/dired.texi

[Stefan Monnier]

> Why does wdired allow self-insertion?
> Is there a good reason for it?
> If not, it would seem to be a bug.

The whole purpose of wdired is to allow you to edit your directory (file
names, permissions, ...) directly with self-insert-command, so it's not at
all a bug.


        Stefan

Re: proofreading man/dired.texi

[Miles Bader]

On Tue, 15 Feb 2005 15:12:45 -0500, Stefan Monnier
<monnier@iro.umontreal.ca> wrote:
> > Why does wdired allow self-insertion?
> > Is there a good reason for it?
> > If not, it would seem to be a bug.
> 
> The whole purpose of wdired is to allow you to edit your directory (file
> names, permissions, ...) directly with self-insert-command, so it's not at
> all a bug.

BTW, I noticed that wdired seems to be a bit unintuitive:

 (1) The function name "wdired-change-to-wdired-mode" is bizarre; why
not just `wdired-mode'????  Is there another way one is supposed to
enter it?

 (2) One must hit `C-c C-c' for changes to take effect; this seems a
bit heavy-weight -- why not just RET (which seems to be what other
systems use)?

 (3) Similarly, in analogy to other systems, it might be nice if there
was some obvious visual indication that wdired was active and
filenames "open for editting" -- e.g., if the filenames looked like a
customize-style "input field" (one might even use the same abstract
face for both, e.g. define an `input-field' face, and have customize
inherit from it).

Thoughts?

-Miles
-- 
Do not taunt Happy Fun Ball.

Re: proofreading man/dired.texi

[Stefan Monnier]

>  (1) The function name "wdired-change-to-wdired-mode" is bizarre; why
> not just `wdired-mode'????  Is there another way one is supposed to
> enter it?

If it behaved like a minor mode, that would make sense.  But it behaves more
like a major mode right now, so it makes sense to have a separate command
which does the pre-processing needed before entering the wdired-mode
major mode.  In any case, it's probably more accidental than anything else.

>  (2) One must hit `C-c C-c' for changes to take effect; this seems a
> bit heavy-weight -- why not just RET (which seems to be what other
> systems use)?

I don't know of any other system that allows you to do what wdired does.
Most others I know are more like dired-efap: they allow you to edit just one
file name.

>  (3) Similarly, in analogy to other systems, it might be nice if there
> was some obvious visual indication that wdired was active and
> filenames "open for editting" -- e.g., if the filenames looked like a
> customize-style "input field" (one might even use the same abstract
> face for both, e.g. define an `input-field' face, and have customize
> inherit from it).

Similarly, other systems only provide "one field at a time".  But since
wdired only handles modifications in some fields and not in others, it might
indeed make sense to "highlight" the parts that can be changed.  Maybe with
an `input-field' face.


        Stefan

Re: proofreading man/dired.texi

[Richard Stallman]

    BTW, I noticed that wdired seems to be a bit unintuitive:

     (1) The function name "wdired-change-to-wdired-mode" is bizarre; why
    not just `wdired-mode'????  Is there another way one is supposed to
    enter it?

     (2) One must hit `C-c C-c' for changes to take effect; this seems a
    bit heavy-weight -- why not just RET (which seems to be what other
    systems use)?

     (3) Similarly, in analogy to other systems, it might be nice if there
    was some obvious visual indication that wdired was active and
    filenames "open for editting" -- e.g., if the filenames looked like a
    customize-style "input field" (one might even use the same abstract
    face for both, e.g. define an `input-field' face, and have customize
    inherit from it).

    Thoughts?

They all seem like good ideas--I think you should change them.
But it would be good to talk with the wdired maintainer first.

Re: proofreading man/dired.texi

[Matt Hodges]

>>>>> Miles Bader writes:

 > (2) One must hit `C-c C-c' for changes to take effect; this seems a
 > bit heavy-weight -- why not just RET (which seems to be what other
 > systems use)?

I agree that this makes sense if you want to edit one filename at a
time, and then commit this change immediately.  However, if you edit
multiple filenames using, e.g., replace-string, replace-regexp,
string-rectangle (or whatever), then I think it's reasonable to have
something like C-c C-c to commit these changes.

 > (3) Similarly, in analogy to other systems, it might be nice if
 > there was some obvious visual indication that wdired was active and
 > filenames "open for editting" -- e.g., if the filenames looked like
 > a customize-style "input field" (one might even use the same
 > abstract face for both, e.g. define an `input-field' face, and have
 > customize inherit from it).

I think it would be useful to have some visual cue for filenames that
have been changed but not committed.  Even more so if it becomes
possible to commit changes for single filenames (with RET) without
exiting wdired-mode.

Matt