non noisy dired

non noisy dired

[Rustom Mody]

[-- Attachment #1: Type: text/plain, Size: 341 bytes --]

I keep trying to use dired but get put off by the noisy output

I am on windows so every file being prefixed by my name and root is useless.

(setq ls-lisp-verbosity nil)
removed that
but the rwxrwxrwx
(which is useless on windows) I cant remove.
If I use
(setq ls-lisp-use-insert-directory-program  t)
the useless name and group re-appears

[-- Attachment #2: Type: text/html, Size: 475 bytes --]

Re: non noisy dired

[Eli Zaretskii]

> Date: Thu, 4 Jun 2009 19:48:25 +0530
> From: Rustom Mody <rustompmody@gmail.com>
> 
> I am on windows so every file being prefixed by my name and root is useless.

Not anymore: Emacs 23.1, to be released soon, will display the _real_
owner and group of every file (if you are using a sufficiently new
version of Windows).

> (setq ls-lisp-verbosity nil)
> removed that
> but the rwxrwxrwx
> (which is useless on windows) I cant remove.

It is on my TODO to teach Emacs on Windows to display the _real_
access rights, based on native Windows file permissions.  So this,
too, is going to be _real_ attributes of the files.  Stay tuned.

> If I use
> (setq ls-lisp-use-insert-directory-program  t)
> the useless name and group re-appears

When you set this option non-nil, Emacs uses ls.exe to generate the
directory listing, so the control of which file's attributes are
displayed is via the command-line switches passed to
insert-directory-program.  You can use non-default switches by
invoking Dired with a numeric argument -- Dired will then prompt for
the switches.

RE: non noisy dired

[Drew Adams]

	I keep trying to use dired but get put off by the noisy output
	I am on windows so every file being prefixed by my name and root is
useless. 
	(setq ls-lisp-verbosity nil)
	removed that
	but the rwxrwxrwx
	(which is useless on windows) I cant remove.
	If I use
	(setq ls-lisp-use-insert-directory-program  t)
	the useless name and group re-appears

http://www.emacswiki.org/emacs/DiredDetails

You'll never have to see anything other than the file name, unless you want to,
and you can do that by hitting a toggle key.

Other than that solution, you can control the ls that switches used, for the
expanded (unhidden) listing, as Eli mentioned.

Re: non noisy dired

[Suvayu Ali]

Drew Adams wrote:
> http://www.emacswiki.org/emacs/DiredDetails
> 
> You'll never have to see anything other than the file name, unless you want to,
> and you can do that by hitting a toggle key.
> 
> Other than that solution, you can control the ls that switches used, for the
> expanded (unhidden) listing, as Eli mentioned.
> 

Hi Drew,

This is probably OT, but still. When you mentioned this extension I 
tried out dired-details+.el. It works perfectly as expected when I load 
it using `M-x load-library dired-details+'.

However if I put (require 'dired-details+) in my .emacs, I get "Symbol's 
value as variable is void: dired-mode-map" in the Messages buffer. On 
starting emacs with the --debug-init flag I get,

> Debugger entered--Lisp error: (void-variable dired-mode-map)
>   byte-code( <some gibberish> [dired-mode-map fboundp dired-details-show dired-details-install define-key "(" dired-details-toggle ")" ad-add-advice (fit-dired-frame nil t (advice lambda nil "Save `dired-details-last-state'.  Fit Dired frame if `one-window-p'." ... ...)) after nil ad-activate dired-details-hide (fit-dired-frame nil t (advice lambda nil "Save `dired-details-last-state'.  Fit Dired frame if `one-window-p'." ... ...)) provide dired-details+] 5)
>   require(dired-details+)
>   eval-buffer(#<buffer  *load*> nil "/home/jallad/.emacs" nil t)  ; Reading at buffer position 2406
>   load-with-code-conversion("/home/jallad/.emacs" "/home/jallad/.emacs" t t)
>   load("~/.emacs" t t)
>   #[nil <some gibberish> " [init-file-user system-type user-init-file-1 user-init-file otherfile source ms-dos "~" "/_emacs" windows-nt directory-files nil "^\\.emacs\\(\\.elc?\\)?$" "~/.emacs" "^_emacs\\(\\.elc?\\)?$" "~/_emacs" vax-vms "sys$login:.emacs" "/.emacs" t load expand-file-name "init" file-name-as-directory "/.emacs.d" file-name-extension "elc" file-name-sans-extension ".el" file-exists-p file-newer-than-file-p message "Warning: %s is newer than %s" sit-for 1 "default" alt inhibit-default-init inhibit-startup-screen] 7]()
>   command-line()
>   normal-top-level()

I couldn't get the <some gibberish> parts to paste into thunderbird or 
any other text editor, sorry. I tried this on Fedora 10 and Xubuntu 
Hardy Heron with the emacs 22 installed from the distro supplied packages.

What am I doing wrong here? I am very new to emacs, just a few months. 
So its very likely I could be missing something trivial.

Thanks for any pointers.

-- 
Suvayu

Open source is the future. It sets us free.

RE: non noisy dired

[Drew Adams]

> > http://www.emacswiki.org/emacs/DiredDetails
> > 
> > You'll never have to see anything other than the file name, 
> > unless you want to, and you can do that by hitting a toggle key.
>
> This is probably OT, but still. When you mentioned this extension I 
> tried out dired-details+.el. It works perfectly as expected 
> when I load it using `M-x load-library dired-details+'.
> 
> However if I put (require 'dired-details+) in my .emacs, I 
> get "Symbol's value as variable is void: dired-mode-map"
> 
> What am I doing wrong here? I am very new to emacs, just a 
> few months. So its very likely I could be missing something trivial.

No, you were not doing anything wrong.
Here's the explanation (it was a bug):

(require 'dired-details+) loads dired-details+, which in turn does (require
'dired-details), which loads dired-details.

But neither explicitly loads dired, which is where variable `dired-mode-map' is
defined. After it loads dired-details, dired-details+ tries to use that variable
at load time --> error.

Should be fixed now. The fix is to use the variable not at load time, but after
dired is loaded.

Thx - Drew

P.S. Two years ago, the author of dired-details.el, Rob Giardina, integrated all
of the dired-details.el and dired-details+.el features into Emacs - he submitted
a comprehensive patch. The Emacs developers seemed to agree to this addition,
but they never followed up, so it never happened. I've ping'ed them several
times about it (the last time was 2008-11-12), with zero response. Too bad.
Rob's integration was clean, avoiding the `defadvice' and other hoops that these
two libraries jump through.

Re: non noisy dired

[Thien-Thi Nguyen]

() "Drew Adams" <drew.adams@oracle.com>
() Sun, 7 Jun 2009 07:02:17 -0700

   The Emacs developers seemed to agree to this addition, but
   they never followed up, so it never happened. I've ping'ed
   them several times about it (the last time was 2008-11-12),
   with zero response.

After the next release please bring it up again.  (BTW, i am under the
impression that you have write privs to the Emacs repo (and thus could
do the integration yourself) -- am i missing something?)

thi

RE: non noisy dired

[Drew Adams]

OT -

>    The Emacs developers seemed to agree to this addition, but
>    they never followed up, so it never happened. I've ping'ed
>    them several times about it (the last time was 2008-11-12),
>    with zero response.
> 
> After the next release please bring it up again.

OK, if I remember. Or you can. ;-) When you get tired of pinging, maybe you can
find a third person (fourth, counting Rob) to carry the torch.

> (BTW, i am under the impression that you have write privs
> to the Emacs repo (and thus could do the integration yourself)
> -- am i missing something?)

I might have write privs (dunno), but I don't want to get involved with that,
sorry. I don't mind submitting patches and ideas for discussion, when I have
time.

FWIW, Rob probably has write privileges too (dunno).
But for some reason his integration work came to nought.

Re: non noisy dired

[Suvayu Ali]

Drew Adams wrote:
>>> http://www.emacswiki.org/emacs/DiredDetails
>>>
>>> You'll never have to see anything other than the file name, 
>>> unless you want to, and you can do that by hitting a toggle key.
>> This is probably OT, but still. When you mentioned this extension I 
>> tried out dired-details+.el. It works perfectly as expected 
>> when I load it using `M-x load-library dired-details+'.
>>
>> However if I put (require 'dired-details+) in my .emacs, I 
>> get "Symbol's value as variable is void: dired-mode-map"
>>
>> What am I doing wrong here? I am very new to emacs, just a 
>> few months. So its very likely I could be missing something trivial.
> 
> No, you were not doing anything wrong.
> Here's the explanation (it was a bug):
> 
> (require 'dired-details+) loads dired-details+, which in turn does (require
> 'dired-details), which loads dired-details.
> 
> But neither explicitly loads dired, which is where variable `dired-mode-map' is
> defined. After it loads dired-details, dired-details+ tries to use that variable
> at load time --> error.
> 
> Should be fixed now. The fix is to use the variable not at load time, but after
> dired is loaded.
> 
> Thx - Drew
> 

Thank you Drew. I really love this library.

> P.S. Two years ago, the author of dired-details.el, Rob Giardina, integrated all
> of the dired-details.el and dired-details+.el features into Emacs - he submitted
> a comprehensive patch. The Emacs developers seemed to agree to this addition,
> but they never followed up, so it never happened. I've ping'ed them several
> times about it (the last time was 2008-11-12), with zero response. Too bad.
> Rob's integration was clean, avoiding the `defadvice' and other hoops that these
> two libraries jump through.
> 

This would be a great addition to Emacs. Hopefully the developers show 
more interest before the next release.

GL

-- 
Suvayu

Open source is the future. It sets us free.

Re: non noisy dired

[Suvayu Ali]

Drew Adams wrote:
> No, you were not doing anything wrong.
> Here's the explanation (it was a bug):
> 
> (require 'dired-details+) loads dired-details+, which in turn does (require
> 'dired-details), which loads dired-details.
> 
> But neither explicitly loads dired, which is where variable `dired-mode-map' is
> defined. After it loads dired-details, dired-details+ tries to use that variable
> at load time --> error.
> 
> Should be fixed now. The fix is to use the variable not at load time, but after
> dired is loaded.
> 
> Thx - Drew
> 

Hi Drew,

Sorry to bother again, I re-installed it and everything seems to work 
except for the neat key-binding, '(' and ')', in dired-details+ where 
pressing either would change the view.

When I put (require 'dired-details+) in my .emacs on invoking dired by 
`C-x d' everything starts up as it should but the key-bindings are ')' 
for more details and '(' for less details, instead of either of them 
working as the toggle. However when I load it using `M-x load-library 
dired-details+' I get the expected behaviour, either keys '(' or ')' 
toggles the details in the dired buffer.

Thank you

-- 
Suvayu

Open source is the future. It sets us free.

RE: non noisy dired

[Drew Adams]

> everything starts up as it should but the key-bindings are
> ')' for more details and '(' for less details,
> instead of either of them working as the toggle.

OK, please try the latest. Sorry for the inconvenience.

Re: non noisy dired

[Suvayu Ali]

Drew Adams wrote:
>> everything starts up as it should but the key-bindings are
>> ')' for more details and '(' for less details,
>> instead of either of them working as the toggle.
> 
> OK, please try the latest. Sorry for the inconvenience.
> 

Thank you Drew for taking the time to fix this. Much appreciated. :)

-- 
Suvayu

Open source is the future. It sets us free.

Re: non noisy dired

[rustom]

Something seems to have changed (or Im not doing somethin' right)
C-Ret (w32-browse) used to work (if dired+ was loaded of course)
Now I need to load it to make it work

RE: non noisy dired

[Drew Adams]

> Something seems to have changed (or Im not doing somethin' right)
> C-Ret (w32-browse) used to work (if dired+ was loaded of course)
> Now I need to load it to make it work

(Subject line is OT - this has nothing to do with dired-details, AFAICT.)

You don't give us much to go on, to help you.
You need to load what, in order to make what work? What are you in fact loading?
What are the symptoms?

You mention `dired+.el'. That library binds C-RET to `dired-w32-browser',
*provided* library `w32-browser.el' is already loaded. So make sure you load
`w32-browser.el' first. Nothing has changed in this regard in `dired+.el' -
perhaps you have changed your load order?

You can check the bindings in any keymap, such as `dired-mode-map', by loading
library `help-fns+.el' and using `C-h C-k' (`describe-keymap').

Re: non noisy dired

[rustom]

On Jun 8, 8:18 pm, "Drew Adams" <drew.ad...@oracle.com> wrote:
> > Something seems to have changed (or Im not doing somethin' right)
> > C-Ret (w32-browse) used to work (if dired+ was loaded of course)
> > Now I need to load it to make it work
>
> (Subject line is OT - this has nothing to do with dired-details, AFAICT.)
>
> You don't give us much to go on, to help you.
> You need to load what, in order to make what work? What are you in fact loading?
> What are the symptoms?

Nothing major -- just that something seemed to have changed (which did
not seem intended)

>
> You mention `dired+.el'. That library binds C-RET to `dired-w32-browser',
> *provided* library `w32-browser.el' is already loaded. So make sure you load
> `w32-browser.el' first. Nothing has changed in this regard in `dired+.el' -
> perhaps you have changed your load order?
Maybe

But I remember that loading dired-details+ (and nothing else) made the
w32-browser keys work.
Now I need to load w32-browser explicitly.  Maybe I loaded it by hand
and forgot :-)
As I said nothing major

>
> You can check the bindings in any keymap, such as `dired-mode-map', by loading
> library `help-fns+.el' and using `C-h C-k' (`describe-keymap').

Re: non noisy dired

[rustom]

On Jun 5, 9:37 am, "Drew Adams" <drew.ad...@oracle.com> wrote:
>         I keep trying to use dired but get put off by the noisy output
>         I am on windows so every file being prefixed by my name and root is
> useless.
>         (setq ls-lisp-verbosity nil)
>         removed that
>         but the rwxrwxrwx
>         (which is useless on windows) I cant remove.
>         If I use
>         (setq ls-lisp-use-insert-directory-program  t)
>         the useless name and group re-appears
>
> http://www.emacswiki.org/emacs/DiredDetails
>
> You'll never have to see anything other than the file name, unless you want to,
> and you can do that by hitting a toggle key.
>
> Other than that solution, you can control the ls that switches used, for the
> expanded (unhidden) listing, as Eli mentioned.

Drew:
autofit seems to need fit --did that (need to put that in the wiki
mention)
After that emacs does resizing of windows in a way that is
disconcerting (to me :-) -- it made a small file into a window the
size of a matchbox!
But I guess I should be able to use it without fit -- need to try more
thoroughly

Eli:
Well progress is always good to see happening but upgrading/
reinstalling is not always easy/possible (for reasons that are too OT
for here)

RE: non noisy dired

[Drew Adams]

> > http://www.emacswiki.org/emacs/DiredDetails
> > You'll never have to see anything other than the file name, 
> > unless you want to, and you can do that by hitting a toggle key.
> 
> Drew:
> autofit seems to need fit --did that (need to put that in the wiki
> mention) After that emacs does resizing of windows in a way that is
> disconcerting (to me :-) -- it made a small file into a window the
> size of a matchbox!
> But I guess I should be able to use it without fit -- need to try more
> thoroughly

Hi Rustom,

I guess you're referring to `dired-details+.el', not just `dired-details.el'
(which does not fit frames). In any case, this has nothing per se to do with the
display or hiding of some Dired columns. I take it as a side issue.

The issue is that `dired-details+.el' loads `autofit-frame.el', if you have that
library in your `load-path'. Notice the "if", here - this is a user choice. If
you choose to automatically fit single-window Dired frames, then, well, that's
what you get. ;-)

If you do want to load `autofit-frame.el', but you want to turn off
frame-fitting in some particular context, you can bind or set
`fit-frame-inhibit-fitting-flag' to t in that context. If you set or customize
it to t at the top level, then frame-fitting will never take place. Finally, if
you want to be able to manually fit frames (`M-x fit-frame'), but you want to
inhibit automatic frame-fitting of one-window frames, then customize
`autofit-frames-flag' to nil.

So you have several ways to inhibit frame-fitting.

As to the particular frame-fitting behavior you describe - If you are getting a
matchbox size frame, then there are no doubt very few files and subdirectories
in the Dired listing; that is, the buffer to be displayed is small, and with
short lines. If that is not the case, then please report a fit-frame bug - send
me a mail.

You have control over the min and max size for frames that are fit. See the
options available in library `fit-frame.el' (loaded by `autofit-frame.el'),
including these:

fit-frame-min-width
fit-frame-min-height
fit-frame-max-width
fit-frame-max-width-percent
fit-frame-max-height
fit-frame-max-height-percent
fit-frame-empty-width
fit-frame-empty-height
fit-frame-empty-special-display-width
fit-frame-empty-special-display-height

So if a matchbox size frame is large enough to show the entire buffer with no
line wrapping, but you nevertheless want the frame to be bigger and include some
more blank space, then just increase the minimum frame size.

Resizing a frame so that it fits its displayed buffer (e.g. Dired, with no file
details showing) changes the frame width to accommodate the longest line without
wrapping it. You typically do not want to count the length of header lines in
this calculation, however - in particular, for Info and Dired.

It is normally the width of the Dired files and subdirs that you care about;
it's better to let the header line wrap, and so to conserve screen space by
fitting the frame to the longest file/subdir line, ignoring the header line
length. But if you like, you can control this header-skipping (whether it is
skipped, and how many lines are skipped) on a per-mode basis, using option
`fit-frame-skip-header-lines-alist'.

Wrapping header lines adds to the vertical space required, of course. The number
of times a given line is wrapped is taken into account in calculating the needed
frame height (pertinent only if the buffer has few enough lines that all can be
displayed at once).

Frame-fitting also takes into account the number of menu-bar and tool-bar lines
your frame has (as indicated by those frame parameters), when it calculates the
needed frame height. However, if the frame is fit to be narrower than the
menu-bar or tool-bar, there is no way to obtain this info. That is, the code
cannot know whether a menu-bar gets wrapped or how many times it does.

The code just assumes that the menu-bar is wrapped, on average, once. If at some
time the menu-bar is not wrapped at all (but it is showing), and you have few
enough lines in the buffer to show all of them, then the frame will be one line
taller than it really needs to be. If the menu-bar wraps more than once, then
the frame will be slightly shorter than it really needs to be.

In sum:

1. Frame-fitting does some calculation and uses some heuristics to do the best
it can. (It tries, BTW, to adjust to different Emacs releases, which handle
things such as the menu-bar and tool-bar very differently.)

2. You can control how it does this - consult the options in library
`fit-frame.el'. You can also inhibit frame-fitting, either generally or in
particular contexts.

3. Again, this has nothing per se to do with hiding Dired details. Automatically
frame-fitting one-window Dired buffers whenever you toggle hiding/showing the
file details is a feature of `dired-details+.el', not of vanilla
`dired-details.el'.

HTH. Please send a bug report if you encounter some behavior you think is
incorrect.