Keybindings for Emacs with no X?

Keybindings for Emacs with no X?

[Angel de Vicente]

Hi,

I run Emacs in text mode, inside GNUScreen, running in a
gnome-terminal. As a result, many key combinations don't get sent to
Emacs properly. For example, M-SPC gets to Emacs as C-M-j. Since I run
Emacs always like this, and basically I live inside Emacs, I would be
willing to spend some time trying to find a way out of this
problem. Searching the web has not really helped much... 

Any help on where I should be looking? 

Thanks,
-- 
Ángel de Vicente
http://angel-de-vicente.blogspot.com/

Re: Keybindings for Emacs with no X?

[Bob Proulx]

Angel de Vicente wrote:
> I run Emacs in text mode, inside GNUScreen, running in a
> gnome-terminal. As a result, many key combinations don't get sent to
> Emacs properly. For example, M-SPC gets to Emacs as C-M-j. Since I run
> Emacs always like this, and basically I live inside Emacs, I would be
> willing to spend some time trying to find a way out of this
> problem. Searching the web has not really helped much... 
> 
> Any help on where I should be looking? 

Since the problem you described is with the terminal emulator that you
described the place to fix it would seem to be in the terminal
emulator.  It would seem that gnome-terminal is the deficient code in
the path between your keyboard and emacs.

For what it is worth I live inside emacs inside an xterm and M-SPC is
propagated correctly by xterm.  I do have "Meta Sends Escape"
(XTerm*metaSendsEscape:true) configured in my Xresources for xterm.

Perhaps using xterm would be a solution for you?

Bob

Re: Keybindings for Emacs with no X?

[郭晓峰]

I am using xterm-extras.el
http://www.dur.ac.uk/p.j.heslin/Software/Emacs/Download/xterm-extras.el

And there are other mapping to simplify your works.

On Thu, Feb 2, 2012 at 11:10 PM, Bob Proulx <bob@proulx.com> wrote:
> Angel de Vicente wrote:
>> I run Emacs in text mode, inside GNUScreen, running in a
>> gnome-terminal. As a result, many key combinations don't get sent to
>> Emacs properly. For example, M-SPC gets to Emacs as C-M-j. Since I run
>> Emacs always like this, and basically I live inside Emacs, I would be
>> willing to spend some time trying to find a way out of this
>> problem. Searching the web has not really helped much...
>>
>> Any help on where I should be looking?
>
> Since the problem you described is with the terminal emulator that you
> described the place to fix it would seem to be in the terminal
> emulator.  It would seem that gnome-terminal is the deficient code in
> the path between your keyboard and emacs.
>
> For what it is worth I live inside emacs inside an xterm and M-SPC is
> propagated correctly by xterm.  I do have "Meta Sends Escape"
> (XTerm*metaSendsEscape:true) configured in my Xresources for xterm.
>
> Perhaps using xterm would be a solution for you?
>
> Bob
>

Re: Keybindings for Emacs with no X?

[Philipp Haselwarter]

fwiw, here's the section of my key-config dealing with terminal
emulators, cascaded screen etc.

#+begin_src elisp
;; {{{2 input quirks

;; gnome-terminal 3.0.1 + screen 4.01.00devel (GNU8cf5efc) 2-May-06 on nzebook
(define-key input-decode-map (kbd "M-[ 1 ; 3 A") [M-up])
(define-key input-decode-map (kbd "M-[ 1 ; 3 B") [M-down])
(define-key input-decode-map (kbd "M-[ 1 ; 3 C") [M-right])
(define-key input-decode-map (kbd "M-[ 1 ; 3 D") [M-left])
(define-key input-decode-map (kbd "M-[ 1 ; 7 A") [C-M-up])
(define-key input-decode-map (kbd "M-[ 1 ; 7 B") [C-M-down])
(define-key input-decode-map (kbd "M-[ 1 ; 7 C") [C-M-right])
(define-key input-decode-map (kbd "M-[ 1 ; 7 D") [C-M-left])
;; terminal 0.4.7 (Xfce 4.8.0) +
;; screen 4.01.00devel (GNU8cf5efc) 2-May-06 on nzebook
(define-key input-decode-map (kbd "M-O a") (kbd "C-<up>"))
(define-key input-decode-map (kbd "M-O b") (kbd "C-<down>"))
(define-key input-decode-map (kbd "M-O c") (kbd "C-<right>"))
(define-key input-decode-map (kbd "M-O d") (kbd "C-<left>"))
(define-key input-decode-map (kbd "M-[ 1 ; 5 A") [C-up])
(define-key input-decode-map (kbd "M-[ 1 ; 5 B") [C-down])
(define-key input-decode-map (kbd "M-[ 1 ; 5 C") [C-right])
(define-key input-decode-map (kbd "M-[ 1 ; 5 D") [C-left])
(define-key input-decode-map (kbd "M-[ a") [S-up])
(define-key input-decode-map (kbd "M-[ b") [S-down])
(define-key input-decode-map (kbd "M-[ c") [S-right])
(define-key input-decode-map (kbd "M-[ d") [S-left])
;; urxvt v9.11 - released: 2011-05-02 +
;; screen 4.01.00devel (GNU8cf5efc) 2-May-06 on nzebook
(define-key input-decode-map (kbd "ESC M-O A") [M-up])
(define-key input-decode-map (kbd "ESC M-O B") [M-down])
(define-key input-decode-map (kbd "ESC M-O C") [M-right])
(define-key input-decode-map (kbd "ESC M-O D") [M-left])
;; urxvt v9.11 - released: 2011-05-02
(define-key input-decode-map (kbd "ESC M-[ A") [M-up])
(define-key input-decode-map (kbd "ESC M-[ B") [M-down])
(define-key input-decode-map (kbd "ESC M-[ C") [M-right])
(define-key input-decode-map (kbd "ESC M-[ D") [M-left])
(define-key input-decode-map (kbd "ESC M-O a") [C-M-up])
(define-key input-decode-map (kbd "ESC M-O b") [C-M-down])
(define-key input-decode-map (kbd "ESC M-O c") [C-M-right])
(define-key input-decode-map (kbd "ESC M-O d") [C-M-left])

;; urxvt v9.12 - released: 2011-06-29, ssh
(define-key input-decode-map (kbd "M-[ a") [S-up])
(define-key input-decode-map (kbd "M-[ b") [S-down])
(define-key input-decode-map (kbd "M-[ c") [S-right])
(define-key input-decode-map (kbd "M-[ d") [S-left])

;; (define-key input-decode-map (kbd "M-O 3 A") [M-up])
;; (define-key input-decode-map (kbd "M-O 3 B") [M-down])
;; (define-key input-decode-map (kbd "M-O 3 C") [M-right])
;; (define-key input-decode-map (kbd "M-O 3 D") [M-left])
#+end_src

It feels a bit kludgy, but well, it works…

-- 
Philipp Haselwarter

Re: Keybindings for Emacs with no X?

[Tassilo Horn]

Angel de Vicente <angelv@iac.es> writes:

Hi Angel,

> I run Emacs in text mode, inside GNUScreen, running in a
> gnome-terminal. As a result, many key combinations don't get sent to
> Emacs properly. For example, M-SPC gets to Emacs as C-M-j. Since I run
> Emacs always like this, and basically I live inside Emacs, I would be
> willing to spend some time trying to find a way out of this
> problem. Searching the web has not really helped much...
>
> Any help on where I should be looking? 

I use GNOME terminal (3.2.1), too, and M-SPC works just fine.
Basically, mostly all keybindings work just fine, except some that
contain non-alphanumeric characters, e.g., C-< is interpreted as just <,
C-? is interpreted as DEL (!)...

Bye,
Tassilo

RE: Keybindings for Emacs with no X?

[Jiaxin Cao]

Why not use a GUI based emacs?

-----Original Message-----
From: help-gnu-emacs-bounces+jiaxin.cao=gmail.com@gnu.org [mailto:help-gnu-emacs-bounces+jiaxin.cao=gmail.com@gnu.org] On Behalf Of Angel de Vicente
Sent: Friday, February 03, 2012 12:38 PM
To: help-gnu-emacs@gnu.org
Subject: Keybindings for Emacs with no X?

Hi,

I run Emacs in text mode, inside GNUScreen, running in a
gnome-terminal. As a result, many key combinations don't get sent to
Emacs properly. For example, M-SPC gets to Emacs as C-M-j. Since I run
Emacs always like this, and basically I live inside Emacs, I would be
willing to spend some time trying to find a way out of this
problem. Searching the web has not really helped much... 

Any help on where I should be looking? 

Thanks,
-- 
Ángel de Vicente
http://angel-de-vicente.blogspot.com/

Re: Keybindings for Emacs with no X?

[Angel de Vicente]

Hi,

Bob Proulx <bob@proulx.com> writes:
> For what it is worth I live inside emacs inside an xterm and M-SPC is
> propagated correctly by xterm.  I do have "Meta Sends Escape"
> (XTerm*metaSendsEscape:true) configured in my Xresources for xterm.
>
> Perhaps using xterm would be a solution for you?

I tried, and it certainly was more promising (out-of-the-box) than
gnome-terminal, but not all combinations were propagated correctly, and
when I added gnu screen, then things got pretty bad again, so I was
hoping for a way to correct all the keymappings in some other way.

Thanks,
-- 
Ángel de Vicente
http://angel-de-vicente.blogspot.com/

Re: Keybindings for Emacs with no X?

[Angel de Vicente]

Hi,

郭晓峰 <lamuguo@gmail.com> writes:

> I am using xterm-extras.el
> http://www.dur.ac.uk/p.j.heslin/Software/Emacs/Download/xterm-extras.el
>
> And there are other mapping to simplify your works.

thanks, I will give it a try. Do you find that all key combinations work
this way? Did you ever try inside a gnu screen session?

Thanks,
-- 
Ángel de Vicente
http://angel-de-vicente.blogspot.com/

Re: Keybindings for Emacs with no X?

[Angel de Vicente]

Hi,

Philipp Haselwarter <philipp.haselwarter@gmx.de> writes:

> fwiw, here's the section of my key-config dealing with terminal
> emulators, cascaded screen etc.

[...]

> It feels a bit kludgy, but well, it works…

Great. If I don't find any other way, I think this should do it. I guess
I can just incrementally build the list of sequences as the need
arises, and soon I will have all the correct keymappings for my
environment. 

Cheers,
-- 
Ángel de Vicente
http://angel-de-vicente.blogspot.com/

Re: Keybindings for Emacs with no X?

[Angel de Vicente]

Hi,

Tassilo Horn <tassilo@member.fsf.org> writes:
> I use GNOME terminal (3.2.1), too, and M-SPC works just fine.
> Basically, mostly all keybindings work just fine, except some that
> contain non-alphanumeric characters, e.g., C-< is interpreted as just <,
> C-? is interpreted as DEL (!)...

I didn't realize I was using such an outdated gnome-terminal version
(2.32.1). I guess it is time to upgrade...

Thanks,
-- 
Ángel de Vicente
http://angel-de-vicente.blogspot.com/

Re: Keybindings for Emacs with no X?

[Peter Dyballa]

Am 3.2.2012 um 08:48 schrieb Tassilo Horn:

> C-? is interpreted as DEL (!)...

That's correct behaviour. C-? is used as DEL since decades in GNU Emacs. At least when run in real terminals...

--
Mit friedvollen Grüßen

  Pete

If all else fails read the instructions.
				- Donald Knuth

Re: Keybindings for Emacs with no X?

[郭晓峰]

I used tmux, it works. But you need to makes sure your terminal
forward the keybinding as you want. For double checking about this,
you can install "libncurses", and write some code like in my post
(http://setter.blogspot.com/2012/01/macremote-emacs.html) to make
sure.

Good luck

Best Regards,
Xiaofeng

On Fri, Feb 3, 2012 at 1:42 AM, Angel de Vicente <angelv@iac.es> wrote:
> Hi,
>
> 郭晓峰 <lamuguo@gmail.com> writes:
>
>> I am using xterm-extras.el
>> http://www.dur.ac.uk/p.j.heslin/Software/Emacs/Download/xterm-extras.el
>>
>> And there are other mapping to simplify your works.
>
> thanks, I will give it a try. Do you find that all key combinations work
> this way? Did you ever try inside a gnu screen session?
>
> Thanks,
> --
> Ángel de Vicente
> http://angel-de-vicente.blogspot.com/
>
>

Re: Keybindings for Emacs with no X?

[Bob Proulx]

Angel de Vicente wrote:
> Bob Proulx writes:
> > For what it is worth I live inside emacs inside an xterm and M-SPC is
> > propagated correctly by xterm.  I do have "Meta Sends Escape"
> > (XTerm*metaSendsEscape:true) configured in my Xresources for xterm.
> >
> > Perhaps using xterm would be a solution for you?
> 
> I tried, and it certainly was more promising (out-of-the-box) than
> gnome-terminal, but not all combinations were propagated correctly,

Oh well.  Nothing ventured, nothing gained.

> and when I added gnu screen, then things got pretty bad again,

I use screen a lot too but haven't run into key issues.

Out of curiosity, what cases did you encounter that were problematic?

> so I was hoping for a way to correct all the keymappings in some
> other way.

Of course!  Good luck in your search.  I will be interested to hear
of your results.

Bob

Re: Keybindings for Emacs with no X?

[Bob Proulx]

Jiaxin Cao wrote:
> Angel de Vicente wrote:
> > I run Emacs in text mode, inside GNUScreen, running in a
>
> Why not use a GUI based emacs?

One reason is because logging in over a high latency connection to a
server on the other side of the world and trying to throw the X11
display back over the very slow global Internet is a test of patience.
Not to mention other issues such as those of security.  Using a text
terminal is much more efficient of bandwidth making the interactive
response fast and just the right solution.

Bob

Re: Keybindings for Emacs with no X?

[Tassilo Horn]

Peter Dyballa <Peter_Dyballa@Web.DE> writes:

>> C-? is interpreted as DEL (!)...
>
> That's correct behaviour. C-? is used as DEL since decades in GNU
> Emacs. At least when run in real terminals...

That's possible and one major reason I seldomly use emacs(client) in a
terminal for anything more than a quick edit.  I've bound frequently
used commands to C-<non-alnum-key> bindings, because those are mostly
undefined by default.  But then it's a pain for me to have them not
available in terminal emacsen.

Bye,
Tassilo

Re: Keybindings for Emacs with no X?

[Silvio Levy]

> > Why not use a GUI based emacs?

> One reason is because logging in over a high latency connection to a
> server on the other side of the world and trying to throw the X11
> display back over the very slow global Internet is a test of patience.

I can attest that for remote editing -- even across town in one of the
most technologically enabled areas of the US -- X emacs is less
satisfactory than emacs in a text terminal.

Silvio

Re: Keybindings for Emacs with no X?

[Tassilo Horn]

Bob Proulx <bob@proulx.com> writes:

>> Why not use a GUI based emacs?
>
> One reason is because logging in over a high latency connection to a
> server on the other side of the world and trying to throw the X11
> display back over the very slow global Internet is a test of patience.

Yes, X11 forwarding is clearly not the way to go.  But why not open
files on the remote server using TRAMP which comes with emacs?
Similarly, you can have many remote shell buffers as well.  That works
very good for me.  Hey, I even read PDF files and view images on remote
machines using TRAMPed dired, doc-view, and image-mode.

Bye,
Tassilo

Re: Keybindings for Emacs with no X?

[Bob Proulx]

Tassilo Horn wrote:
> Bob Proulx writes:
> >> Why not use a GUI based emacs?
> >
> > One reason is because logging in over a high latency connection to a
> > server on the other side of the world and trying to throw the X11
> > display back over the very slow global Internet is a test of patience.
> 
> Yes, X11 forwarding is clearly not the way to go.

I don't even install X11 libraries on server machines.

And I use text consoles for the machine near my elbow too.  The
overhead of starting up X is significant.  Text mode with a
conservative startup customization will be almost instantaneous.

> But why not open files on the remote server using TRAMP which comes
> with emacs?

First because that would require a second window and I would need to
switch back and forth between my remote logged in terminal window and
my local emacs editing window.  Sure if you want to keep switching
windows all of the time then many people keep an emacs server running
continuously.  That is common.  That's fine.  But if you just need to
log into a system and do some light weight administration then you
don't want to have to set up a continuously running emacs server on
your desktop first.  In that case you just want to edit a file.  Sure
I could [n]vi the file for a quick edit but I want to use emacs to
edit the file not vi or nano or other editor.  Therefore I will log
in, do what I need to do, use emacs to edit files, and then log out.

Secondly for reasons of speed.  Even on fast machines and fast
connections using tramp to edit a file can take many seconds while it
is setting up the connection.  And if this was a quick one-off task
then I may not have set up my ssh-agent for ssh rsa keys yet.  That
will require an additional pause while I enter my ssh rsa passphrase
again.  In the time it takes to do this I may have already finished
using emacs to edit the file in text mode, done whatever else I needed
to do, logged out and moved on.  But if I am going to sit for hours
working on a task then the overhead of starting tramp is insignificant
and the benefits you mention of multiple buffers and so forth work out
in favor of doing that instead.

> Similarly, you can have many remote shell buffers as well.  That
> works very good for me.  Hey, I even read PDF files and view images
> on remote machines using TRAMPed dired, doc-view, and image-mode.

That is all perfectly understandable and reasonable for the types of
tasks that you are doing.  But not everyone has the same work to do.
Other tasks require other work flows.  For example I can't recall a
time that I have ever looked at a pdf document on a remote machine
that I had logged into remotely to do some system administration work.
That case of having pdf documentation to which I would need to refer
to would just would never occur.  I would always be looking at those
types of documents locally on my desktop.  And would also have the web
available for better documentation.

In a different venue some people would suggest using sshfs to mount a
remote system in a local userland filesystem.  Then all remote files
appear to be local.  At the end of the task the filesystem can be
unmounted.  That is yet another different way of working that works
quite well.

Bob

Re: Keybindings for Emacs with no X?

[Silvio Levy]

> Yes, X11 forwarding is clearly not the way to go.  But why not open
> files on the remote server using TRAMP which comes with emacs?

In my case, the answer is that editing files is only part of what I do
remotely. I keep an ssh window open on the remote host anyway; it
makes little sense to then use a *local* invocation of emacs to open a
remote file (with the corresponding overhead of transfer protocols,
mimencode and all that jazz). 

Conceptually, too, it's cleaner if each of my (color-coded) text
windows is acting on a single host.

TRAMP sounds terrific, but it seems to be meant for a different sort
of workflow.

Silvio

Re: Keybindings for Emacs with no X?

[Tassilo Horn]

Silvio Levy <levy@msri.org> writes:

>> Yes, X11 forwarding is clearly not the way to go.  But why not open
>> files on the remote server using TRAMP which comes with emacs?
>
> In my case, the answer is that editing files is only part of what I do
> remotely. I keep an ssh window open on the remote host anyway; it
> makes little sense to then use a *local* invocation of emacs to open a
> remote file (with the corresponding overhead of transfer protocols,
> mimencode and all that jazz).

Yes, I frequently keep a TRAMP dired or shell buffer open on remote
hosts.  In such a buffer, even completion after C-x C-f defaults to the
remote directory of the dired current buffer or the shell buffers cwd.

> Conceptually, too, it's cleaner if each of my (color-coded) text
> windows is acting on a single host.

You could use one frame per host, and also have the frames colored
differently.  But I don't want to evangelize anyone.  I was just curious
why many experienced emacs users don't use TRAMP.  In my experience, it
enormously matured over the last years from a slow and flaky
implementation of a good idea to a highly usable and convenient tool.

Bye,
Tassilo

Re: Keybindings for Emacs with no X?

[Peter Dyballa]

Am 3.2.2012 um 21:03 schrieb Tassilo Horn:

> I've bound frequently
> used commands to C-<non-alnum-key> bindings, because those are mostly
> undefined by default.  But then it's a pain for me to have them not
> available in terminal emacsen.

Because the ASCII "protocol" is so limited. Would 8-bit controls give you more? (Well, another 32...) You would then need to press the Esc key many times...

--
Mit friedvollen Grüßen

  Pete

Wer nichts zu verbergen hat, hat schon alles verloren.
				(Juli Zeh)

Re: Keybindings for Emacs with no X?

[Angel de Vicente]

Hi,

Silvio Levy <levy@msri.org> writes:

>> > Why not use a GUI based emacs?
>
>> One reason is because logging in over a high latency connection to a
>> server on the other side of the world and trying to throw the X11
>> display back over the very slow global Internet is a test of patience.
>
> I can attest that for remote editing -- even across town in one of the
> most technologically enabled areas of the US -- X emacs is less
> satisfactory than emacs in a text terminal.

In my case, more than a matter of speed is a matter of convenience
(though speed also counts). At work I fire up a gnu screen session, and
inside it I have 5-6 'terminals', all running emacsclients giving me
different aspects of my work (mail, agenda, documentation, programming,
other, etc.). When I go to another office, home, etc. I just connect
through ssh to my workstation, reattach to gnu screen, and I have
*exactly* the same environment I was using at my office. I can do some
work, and next day at my office I'm exactly where I left it at home. For
me this is invaluable... (I have tried doing the same with xpra and
Emacs in X, but then the speed was the problem...).

Cheers,
-- 
Ángel de Vicente
http://angel-de-vicente.blogspot.com/

Re: Keybindings for Emacs with no X?

[Angel de Vicente]

[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1: Type: text/plain; charset=utf-8, Size: 1290 bytes --]

Hi,

Bob Proulx <bob@proulx.com> writes:
>> and when I added gnu screen, then things got pretty bad again,
>
> I use screen a lot too but haven't run into key issues.
>
> Out of curiosity, what cases did you encounter that were problematic?

Right now I'm using a lot org-mode, so I would like to be able to send
M-SPC and C-M-SPC to Emacs.

With my versions of gnome-terminal and screen M-SPC and C-M-SPC get to Emacs as
C-M-j (for the fist one I can use ESC-SPC, but ideally I want to get all
keybindings working, so I don't have to take it into consideration when
reading Emacs/org-mode, etc. documentation.

>> so I was hoping for a way to correct all the keymappings in some
>> other way.
>
> Of course!  Good luck in your search.  I will be interested to hear
> of your results.

Will share if I get somewhere...
-- 
Ángel de Vicente
http://www.iac.es/galeria/angelv/          
---------------------------------------------------------------------------------------------
ADVERTENCIA: Sobre la privacidad y cumplimiento de la Ley de Protección de Datos, acceda a http://www.iac.es/disclaimer.php
WARNING: For more information on privacy and fulfilment of the Law concerning the Protection of Data, consult http://www.iac.es/disclaimer.php?lang=en

How do I bypass Tramp? [was:Re: Keybindings for Emacs with no X?]

[Tami]

On 2012 February 3 Tassilo Horn wrote:

  [brevitized]

> Yes, X11 forwarding is clearly not the way to go.  But why not open
> files on the remote server using TRAMP which comes with emacs?

  ----->> *snip* <<-----

I also live in emacs (no X) on 2 small networks of computers connected via an
in-house meta-network.  My housemate and I regularly transfer files from
machine to machine, subnet to subnet wherever there is disk space or specific
capability.

Recently he acquired some new machines and installed the latest Debian
distributions with emacs21 and/or 22.  Since I've been running emacs 19.34 for
decades (its practically perfect) I was horrified when my usual method of file
transfer - ange-ftp - was replaced by tramp on the new machines (which do nice
things like display video files on a previously-vacant TV channel).

Transfer times between 2 machines in the same room (on the same subnet) now
take eons and for large file transfers, Tramp demands repeated re-logins.

        This is an insanely complex nonsolution to a simple nonproblem.

How do I use just ange-ftp for these in-house transfers without disabling the
ability to use tramp if I ever  need it?

Is tramp actually good for anything useful?  If not, how do I disable it
permanently for myself when using his machines?

If it is useful, can it be sped up?  I've had faster FTP transfers between MIT
and Finland when the 2 countries were connected by one shared 56k line across
an ocean!  The 2 computers I refer to are 4 feet apart!

My housemate is learning emacs, but his default is vi (shudder), netcat, and
packet radio (which takes longer than tramp but lets one sleep through
overnight transfers).

He didn't install the .el files.

Thanks

.Tami
 .signature: syntax error at line 1: `(' unexpected

RE: How do I bypass Tramp? [was:Re: Keybindings for Emacs with no X?]

[Drew Adams]

> I've been running emacs 19.34 for decades (its practically
> perfect) I was horrified when my usual method of file
> transfer - ange-ftp - was replaced by tramp

Just set option `tramp-default-method' to `ftp'.
AFAIK, that causes Tramp to pass through to ange-ftp.  I don't think you will
see any slowdown, but I'm no expert on this.

Later, when you are more comfortable, you might want to give other methods
(i.e., Tramp per se) a try.  And you might even find later that a recent Emacs
version is more convenient and more useful than 19.34.  YMMV, though.

Re: How do I bypass Tramp? [was:Re: Keybindings for Emacs with no X?]

[Michael Albinus]

Tami <kd5rju@wb5aoh.dyndns.org> writes:

> How do I use just ange-ftp for these in-house transfers without disabling the
> ability to use tramp if I ever  need it?

As Drew already said, `tramp-default-method' shall preserve your
ange-ftp settings.

> Is tramp actually good for anything useful?  If not, how do I disable it
> permanently for myself when using his machines?
>
> If it is useful, can it be sped up?  I've had faster FTP transfers between MIT
> and Finland when the 2 countries were connected by one shared 56k line across
> an ocean!  The 2 computers I refer to are 4 feet apart!

Emacs 22 comes with Tramp 2.0, which is way slow. Upgrading to Emacs 23
shall speed up.

There are also lots of settings, which would ease your life (for example
avoiding to type your passwords again and again).

Tramp is useful to access machines which do not run an ftp daemon. It
offers a lot of different access methods, for example you can access
machines running MS Windows, or even Android based cellular phones. (*)

Tramp does more than file transfer. It allows you also to run remote
processes. (*)

(*): depends on the installed Tramp version.

> Thanks
>
> .Tami

Best regards, Michael.

Re: Re: How do I bypass Tramp? [was:Re: Keybindings for Emacs with no X?]

[Tami]

On 2012 February 5 Michael Albinus wrote:

        [brevitized]
> Emacs 22 comes with Tramp 2.0, which is way slow. Upgrading to Emacs 23
> shall speed up.

> There are also lots of settings, which would ease your life (for example
> avoiding to type your passwords again and again).

> Tramp is useful to access machines which do not run an ftp daemon. It
> offers a lot of different access methods, for example you can access
> machines running MS Windows, or even Android based cellular phones. (*)
> Tramp does more than file transfer. It allows you also to run remote
> processes. (*)

> (*): depends on the installed Tramp version.

   Thanks for the info.

   Tramp may be a useful program to keep on my promiscuous netbook, but
   unnecessary for home machines where ftp is enabled behind the firewall. I'll
   upgrade the netbook & keep my comfy emacs.

   19.34 was the first emacs version GNU ported to M$windoz. In the process it
   got thoroughly debugged.  In 1998 I used it to access Windoz-2000 machines
   (on the same lan) using ange-ftp.  I used dired to mark files to transfer.
   It all worked seamlessly on my GNU/Linux 350MB laptop.

   A decade later I tried an early version of emacs23 hoping to be able to run
   packet radio in emacs shell, but the development version was too riddled
   with bugs to use.  I'm still waiting for emacs23 to appear on a Debian
   (stable) distro. (I'm a wait-til-its-stable adopter.)

   Meanwhile we're looking at zenirc with an eye to tweaking the code to run
   unproto packet radio. Packet is just another chat mode, and TNCs speak modem
   to serial ports. 

   Once I have zenpacket there will be no incentive to upgrade emacs :)

                                     * * *
   PS After I sent my query, someone on the list mentioned that they used
   screen to maintain local states on multiple machines.

   I have been doing this for decades.  I have 11 screens on 4 machines, with
   each screen devoted to one or 2 remote-local tasks.  Its fast and I can
   multitask on multiple machines with just a keystroke to switch.  

   And I still have 6 terminals on my own machine for local tasks!

   I really don't need tramp at all except on slut, the promiscuous netbook :)
   
   Glad I asked! 


.Tami
 .signature: syntax error at line 1: `(' unexpected

Re: Re: How do I bypass Tramp? [was:Re: Keybindings for Emacs with no X?]

[PJ Weisberg]

On Tue, Feb 7, 2012 at 4:42 PM, Tami <kd5rju@wb5aoh.dyndns.org> wrote:

>   A decade later I tried an early version of emacs23 hoping to be able to run
>   packet radio in emacs shell, but the development version was too riddled
>   with bugs to use.  I'm still waiting for emacs23 to appear on a Debian
>   (stable) distro. (I'm a wait-til-its-stable adopter.)

Emacs 23.2 is in Debian Squeeze, which was released almost exactly one year ago.

-PJ

Gehm's Corrollary to Clark's Law: Any technology distinguishable from
magic is insufficiently advanced.

Re: Keybindings for Emacs with no X?

[Ken Goldman]

On 2/3/2012 3:52 PM, Silvio Levy wrote:
>> Yes, X11 forwarding is clearly not the way to go.  But why not open
>> files on the remote server using TRAMP which comes with emacs?
>
> In my case, the answer is that editing files is only part of what I do
> remotely. I keep an ssh window open on the remote host anyway; it
> makes little sense to then use a *local* invocation of emacs to open a
> remote file (with the corresponding overhead of transfer protocols,
> mimencode and all that jazz).
>
> Conceptually, too, it's cleaner if each of my (color-coded) text
> windows is acting on a single host.
>
> TRAMP sounds terrific, but it seems to be meant for a different sort
> of workflow.

Yes.  Your work flow works if you have an account on the remote machine, 
if the machine has emacs, and if the bandwidth and latency are adequate.

If any of these are false, tramp is the solution - pull the file over, 
edit locally, and send it back when you're done.

Re: Keybindings for Emacs with no X?

[Oleksandr Gavenko]

On 2012-02-03, Bob Proulx wrote:

> For what it is worth I live inside emacs inside an xterm and M-SPC is
> propagated correctly by xterm.  I do have "Meta Sends Escape"
> (XTerm*metaSendsEscape:true) configured in my Xresources for xterm.

I have:

  ! Allow Alt as Meta in Emacs:
  XTerm*eightBitInput: false
  XTerm*metaSendsEscape: true

in my '~/.Xdefaults'. Man page xterm(1) say about 3 options related for this
issue:

  eightBitInput
  metaSendsEscape
  altSendsEscape

With this settings I successfully run Emacs under Debian with
'LANG=ru_RU.UTF-8' and Cygwin with 'LANG=ru_RU.cp1251' (in xterm of course).

-- 
Best regards!

Re: Keybindings for Emacs with no X?

[Oleksandr Gavenko]

On 2012-02-04, Angel de Vicente wrote:

> Right now I'm using a lot org-mode, so I would like to be able to send
> M-SPC and C-M-SPC to Emacs.

Right now I run xterm, screen and emacs in it:

  $ xterm
  xterm# screen
  screen# emacs -nw -q

When I type "C-h k M-SPC" I get: M-SPC
When I type "C-h k C-M-SPC" I get: C-M-@

My ~/.Xdefaults settings for xterm:

  ! Allow Alt as Meta in Emacs:
  XTerm*eightBitInput: false
  XTerm*metaSendsEscape: true

-- 
Best regards!

Re: Keybindings for Emacs with no X?

[gtassone]

I use gnome-terminal in Ubuntu and have the same problems with <Meta> in emacs. I have found the following related bugs.
https://bugzilla.gnome.org/show_bug.cgi?id=663779#c13
https://bugs.launchpad.net/ubuntu/+source/vte/+bug/890555

I am not an expert linux user. Can someone describe how to apply the gnome patch?

Also regarding the VTE bug report - it claims it is fixed, but I still experience it with libvte9 1:0.28.2-3ubuntu2. I also find that I have another VTE package installed - libvte-2.90-9 build 1:0.32.1-1ubuntu0. 

Can anyone offer me some advice?

thanks
Gabe

On Thursday, February 2, 2012 11:38:08 PM UTC-5, Angel de Vicente wrote:
> Hi,
> 
> 
> 
> I run Emacs in text mode, inside GNUScreen, running in a
> 
> gnome-terminal. As a result, many key combinations don't get sent to
> 
> Emacs properly. For example, M-SPC gets to Emacs as C-M-j. Since I run
> 
> Emacs always like this, and basically I live inside Emacs, I would be
> 
> willing to spend some time trying to find a way out of this
> 
> problem. Searching the web has not really helped much... 
> 
> 
> 
> Any help on where I should be looking? 
> 
> 
> 
> Thanks,
> 
> -- 
> 
> Ángel de Vicente
> 
> http://angel-de-vicente.blogspot.com/