unofficial mirror of help-gnu-emacs@gnu.org
 help / color / mirror / Atom feed
* confused about emacs 23 in daemon mode
@ 2009-09-13  1:35 Rodrigo Amestica
  2009-09-13  2:19 ` Tim X
  2009-09-13  9:18 ` Peter Dyballa
  0 siblings, 2 replies; 9+ messages in thread
From: Rodrigo Amestica @ 2009-09-13  1:35 UTC (permalink / raw)
  To: help-gnu-emacs

it is so nice to have emacs starting so blazingly fast when using the
daemon and emacsclient combination. However it seems that every other
emacsclient I open is not an independent version of emacs and things
get intermixed. I think that in emacs parlance the terminology is
'frame', all emacsclient instances are emacs frames.

The main problems I have noticed under these circumstances is that 'C-
h-i' seems to be shared by all frames such that browsing in one frame
will make the others to move away from their own current info page.
And the mayor show stopper is ECB. It seems that I cannot have 2
emacsclients running their own ECB session.

I'm missing something or is the daemon mode not that usable as it
promised at first glance?

thanks in advance for any hint here,
 Rodrigo


^ permalink raw reply	[flat|nested] 9+ messages in thread

* Re: confused about emacs 23 in daemon mode
  2009-09-13  1:35 confused about emacs 23 in daemon mode Rodrigo Amestica
@ 2009-09-13  2:19 ` Tim X
  2009-09-13  3:24   ` Daniel Pittman
                     ` (2 more replies)
  2009-09-13  9:18 ` Peter Dyballa
  1 sibling, 3 replies; 9+ messages in thread
From: Tim X @ 2009-09-13  2:19 UTC (permalink / raw)
  To: help-gnu-emacs

Rodrigo Amestica <ramestica@gmail.com> writes:

> it is so nice to have emacs starting so blazingly fast when using the
> daemon and emacsclient combination. However it seems that every other
> emacsclient I open is not an independent version of emacs and things
> get intermixed. I think that in emacs parlance the terminology is
> 'frame', all emacsclient instances are emacs frames.
>
> The main problems I have noticed under these circumstances is that 'C-
> h-i' seems to be shared by all frames such that browsing in one frame
> will make the others to move away from their own current info page.
> And the mayor show stopper is ECB. It seems that I cannot have 2
> emacsclients running their own ECB session.
>
> I'm missing something or is the daemon mode not that usable as it
> promised at first glance?
>

As I understand it, this is normal behavior. 

Running emacs in daemon mode is not that different from running normal
emacs. The benefits are that you only have frames when you need them and
you have fast opening when you do want one. However, it is all still
running in a single emacs instance. So, the behavior you mention is the
same behavior you would get if you were running emacs in non-daemon
mode. For example, when you open an info buffer, that is a single info
buffer shared/accessible from all buffers. If you do a buffer listing,
you will see the info buffer as normal and you can switch to that buffer
in any of the frames you have running. If you have two frames open and
both showing the info buffer, going to one frame and selecting a new
infor page will be reflected in the other frame because it is showing
the same buffer. 

You can possibly change this behavior by renaming the buffers. I've not
changed this, but have used a similar technique when I want multiple
shell buffers. Once the buffer is renamed, you probably can open another
info buffer and the two will be independent. 

I hope this helps.

Tim


-- 
tcross (at) rapttech dot com dot au


^ permalink raw reply	[flat|nested] 9+ messages in thread

* Re: confused about emacs 23 in daemon mode
  2009-09-13  2:19 ` Tim X
@ 2009-09-13  3:24   ` Daniel Pittman
       [not found]   ` <mailman.6556.1252812928.2239.help-gnu-emacs@gnu.org>
  2009-10-12  2:19   ` David Combs
  2 siblings, 0 replies; 9+ messages in thread
From: Daniel Pittman @ 2009-09-13  3:24 UTC (permalink / raw)
  To: help-gnu-emacs

Tim X <timx@nospam.dev.null> writes:
> Rodrigo Amestica <ramestica@gmail.com> writes:
>
>> it is so nice to have emacs starting so blazingly fast when using the
>> daemon and emacsclient combination. However it seems that every other
>> emacsclient I open is not an independent version of emacs and things get
>> intermixed. I think that in emacs parlance the terminology is 'frame', all
>> emacsclient instances are emacs frames.

Correct.

>> The main problems I have noticed under these circumstances is that 'C- h-i'
>> seems to be shared by all frames such that browsing in one frame will make
>> the others to move away from their own current info page.  And the mayor
>> show stopper is ECB. It seems that I cannot have 2 emacsclients running
>> their own ECB session.
>>
>> I'm missing something or is the daemon mode not that usable as it promised
>> at first glance?
>
> As I understand it, this is normal behavior. 

*nod*  You can see the same behaviour by starting a non-daemon Emacs, then
using emacsclient to open additional frames as well.

Regards,
        Daniel
-- 
✣ Daniel Pittman            ✉ daniel@rimspace.net            ☎ +61 401 155 707
               ♽ made with 100 percent post-consumer electrons
   Looking for work?  Love Perl?  In Melbourne, Australia?  We are hiring.





^ permalink raw reply	[flat|nested] 9+ messages in thread

* Re: confused about emacs 23 in daemon mode
       [not found]   ` <mailman.6556.1252812928.2239.help-gnu-emacs@gnu.org>
@ 2009-09-13  7:08     ` Bruno Barbier
  2009-10-12  2:34       ` David Combs
  0 siblings, 1 reply; 9+ messages in thread
From: Bruno Barbier @ 2009-09-13  7:08 UTC (permalink / raw)
  To: help-gnu-emacs

On 2009-09-13, Daniel Pittman <daniel@rimspace.net> wrote:
> Tim X <timx@nospam.dev.null> writes:
>> Rodrigo Amestica <ramestica@gmail.com> writes:
>>
>>> it is so nice to have emacs starting so blazingly fast when using the
>>> daemon and emacsclient combination. However it seems that every other
>>> emacsclient I open is not an independent version of emacs and things get
>>> intermixed. I think that in emacs parlance the terminology is 'frame', all
>>> emacsclient instances are emacs frames.

Thanks to emacsclient, you only have one emacs. The fact that you get
a new frame or not is just a configuration setting.

   (describe-variable 'server-window)

I mean, the fact that things get intermixed is one goal of emacsclient
(the other one being fast start up).

>>> The main problems I have noticed under these circumstances is that 'C- h-i'
>>> seems to be shared by all frames such that browsing in one frame will make
>>> the others to move away from their own current info page.  

Type M-n to open a new independent info buffer.
And most jumping commands have a 'C-u' that jump into a new buffer.
      
    (info "(info)Create Info buffer")

>>> And the mayor
>>> show stopper is ECB. It seems that I cannot have 2 emacsclients running
>>> their own ECB session.

I would use one emacs by ECB and an other one for planning, taking
notes, etc.  You can then make emacsclient use the one you want or
setup only the third emacs as a server depending on your needs.

    From the emacs info manual:

    |    If there is more than one Emacs server running, you can specify a
    | server name with the `-s NAME' or `--socket-name=NAME' option to
    | `emacsclient'.  (This option is not supported on MS-Windows.)

    (info "(emacs)Invoking emacsclient")




^ permalink raw reply	[flat|nested] 9+ messages in thread

* Re: confused about emacs 23 in daemon mode
  2009-09-13  1:35 confused about emacs 23 in daemon mode Rodrigo Amestica
  2009-09-13  2:19 ` Tim X
@ 2009-09-13  9:18 ` Peter Dyballa
  1 sibling, 0 replies; 9+ messages in thread
From: Peter Dyballa @ 2009-09-13  9:18 UTC (permalink / raw)
  To: Rodrigo Amestica; +Cc: help-gnu-emacs


Am 13.09.2009 um 03:35 schrieb Rodrigo Amestica:

> I'm missing something or is the daemon mode not that usable as it
> promised at first glance?

You do not seem to understand that each frame gives you the  
*complete* view of all that is happening in this daemon. There is one  
*info*, one *shell*, one file_a, one dired_b, one ... buffer. All  
frames share the same process and data.

--
Greetings

   Pete

There are three types of people in this world: those who can count,  
and those who cannot.







^ permalink raw reply	[flat|nested] 9+ messages in thread

* Re: confused about emacs 23 in daemon mode
  2009-09-13  2:19 ` Tim X
  2009-09-13  3:24   ` Daniel Pittman
       [not found]   ` <mailman.6556.1252812928.2239.help-gnu-emacs@gnu.org>
@ 2009-10-12  2:19   ` David Combs
  2009-10-12  4:04     ` Juanma Barranquero
  2 siblings, 1 reply; 9+ messages in thread
From: David Combs @ 2009-10-12  2:19 UTC (permalink / raw)
  To: help-gnu-emacs

In article <871vmbh9hy.fsf@lion.rapttech.com.au>,
Tim X  <timx@nospam.dev.null> wrote:
...
>
>You can possibly change this behavior by renaming the buffers. I've not
>changed this, but have used a similar technique when I want multiple
>shell buffers. Once the buffer is renamed, you probably can open another
>info buffer and the two will be independent. 

Renaming "*....*" buffers is really cool.

I myself do it *very* often for *Occur*: *Occur-this,
*Occur-that, ...

But there's a bug, at least in my version: 
  The wonderful "g" command doesn't work right, as
  I recall.  Maybe fixed more recently, but I bet
  not.

  Try it (rename *Occur* to *Occur-this-pat, then
  do M-x occur again, but on "that"-pat, try "g"
  on the renamed one, and see what happens on YOUR
  (newer than mine, probably) version.

THANKS!

David




^ permalink raw reply	[flat|nested] 9+ messages in thread

* Re: confused about emacs 23 in daemon mode
  2009-09-13  7:08     ` Bruno Barbier
@ 2009-10-12  2:34       ` David Combs
  2009-10-12  2:47         ` Richard Riley
  0 siblings, 1 reply; 9+ messages in thread
From: David Combs @ 2009-10-12  2:34 UTC (permalink / raw)
  To: help-gnu-emacs

In article <4aac9a85$0$17751$ba4acef3@news.orange.fr>,
Bruno Barbier  <bruno.barbier.cs@begin_dom.orange.fr.end_dom> wrote:
>On 2009-09-13, Daniel Pittman <daniel@rimspace.net> wrote:
>> Tim X <timx@nospam.dev.null> writes:
>>> Rodrigo Amestica <ramestica@gmail.com> writes:
>>>
>>>> it is so nice to have emacs starting so blazingly fast when using the
>>>> daemon and emacsclient combination. However it seems that every other
>>>> emacsclient I open is not an independent version of emacs and things get
>>>> intermixed. I think that in emacs parlance the terminology is 'frame', all
>>>> emacsclient instances are emacs frames.
>
>Thanks to emacsclient, you only have one emacs. The fact that you get
>a new frame or not is just a configuration setting.
>
>   (describe-variable 'server-window)
>
>I mean, the fact that things get intermixed is one goal of emacsclient
>(the other one being fast start up).
>
>>>> The main problems I have noticed under these circumstances is that 'C- h-i'
>>>> seems to be shared by all frames such that browsing in one frame will make
>>>> the others to move away from their own current info page.  
>
>Type M-n to open a new independent info buffer.
>And most jumping commands have a 'C-u' that jump into a new buffer.
>      
>    (info "(info)Create Info buffer")
>
>>>> And the mayor
>>>> show stopper is ECB. It seems that I cannot have 2 emacsclients running
>>>> their own ECB session.
>
>I would use one emacs by ECB and an other one for planning, taking
>notes, etc.  You can then make emacsclient use the one you want or
>setup only the third emacs as a server depending on your needs.
>
>    From the emacs info manual:
>
>    |    If there is more than one Emacs server running, you can specify a
>    | server name with the `-s NAME' or `--socket-name=NAME' option to
>    | `emacsclient'.  (This option is not supported on MS-Windows.)
>
>    (info "(emacs)Invoking emacsclient")
>
>

Using emacsclient is pretty new to me -- and you seem to
be playing some nice tricks with it.

Please, could you elaborate a bit (or better, much more) on:

  .  the above that you wrote.

  .  your varied uses (and those of others you've seen or
     heard of) of emacsclient (and other things like it).

  .  You used the term "server".  What with the original
     emacs-run throwing off emacs-clients that seem to
     be NOT independent from each other -- sharing buffers,
     etc.  For vocabulary-creation, would that make it into
     some kind of subtype of "server"?  (Obvious to all: am
     pretty ignorant about servers, clients, etc, which is
     why I'm probing, to learn more.)

  .  Anything else you want to say.  (With the idea that 
     maybe RMS could munge it into extra doc for the manual
     or tutorial or something?)


THANKS!

David





^ permalink raw reply	[flat|nested] 9+ messages in thread

* Re: confused about emacs 23 in daemon mode
  2009-10-12  2:34       ` David Combs
@ 2009-10-12  2:47         ` Richard Riley
  0 siblings, 0 replies; 9+ messages in thread
From: Richard Riley @ 2009-10-12  2:47 UTC (permalink / raw)
  To: help-gnu-emacs

dkcombs@panix.com (David Combs) writes:

> In article <4aac9a85$0$17751$ba4acef3@news.orange.fr>,
> Bruno Barbier  <bruno.barbier.cs@begin_dom.orange.fr.end_dom> wrote:
>>On 2009-09-13, Daniel Pittman <daniel@rimspace.net> wrote:
>>> Tim X <timx@nospam.dev.null> writes:
>>>> Rodrigo Amestica <ramestica@gmail.com> writes:
>>>>
>>>>> it is so nice to have emacs starting so blazingly fast when using the
>>>>> daemon and emacsclient combination. However it seems that every other
>>>>> emacsclient I open is not an independent version of emacs and things get
>>>>> intermixed. I think that in emacs parlance the terminology is 'frame', all
>>>>> emacsclient instances are emacs frames.
>>
>>Thanks to emacsclient, you only have one emacs. The fact that you get
>>a new frame or not is just a configuration setting.
>>
>>   (describe-variable 'server-window)
>>
>>I mean, the fact that things get intermixed is one goal of emacsclient
>>(the other one being fast start up).
>>
>>>>> The main problems I have noticed under these circumstances is that 'C- h-i'
>>>>> seems to be shared by all frames such that browsing in one frame will make
>>>>> the others to move away from their own current info page.  
>>
>>Type M-n to open a new independent info buffer.
>>And most jumping commands have a 'C-u' that jump into a new buffer.
>>      
>>    (info "(info)Create Info buffer")
>>
>>>>> And the mayor
>>>>> show stopper is ECB. It seems that I cannot have 2 emacsclients running
>>>>> their own ECB session.
>>
>>I would use one emacs by ECB and an other one for planning, taking
>>notes, etc.  You can then make emacsclient use the one you want or
>>setup only the third emacs as a server depending on your needs.
>>
>>    From the emacs info manual:
>>
>>    |    If there is more than one Emacs server running, you can specify a
>>    | server name with the `-s NAME' or `--socket-name=NAME' option to
>>    | `emacsclient'.  (This option is not supported on MS-Windows.)
>>
>>    (info "(emacs)Invoking emacsclient")
>>
>>
>
> Using emacsclient is pretty new to me -- and you seem to
> be playing some nice tricks with it.
>
> Please, could you elaborate a bit (or better, much more) on:
>
>   .  the above that you wrote.
>
>   .  your varied uses (and those of others you've seen or
>      heard of) of emacsclient (and other things like it).
>
>   .  You used the term "server".  What with the original
>      emacs-run throwing off emacs-clients that seem to
>      be NOT independent from each other -- sharing buffers,
>      etc.  For vocabulary-creation, would that make it into
>      some kind of subtype of "server"?  (Obvious to all: am
>      pretty ignorant about servers, clients, etc, which is
>      why I'm probing, to learn more.)
>
>   .  Anything else you want to say.  (With the idea that 
>      maybe RMS could munge it into extra doc for the manual
>      or tutorial or something?)

Just think of the daemon as a running program with no human
interface. Emacsclient is then merely a "front end" which displays data
the daemon marshals.. Perhaps you could think of the emacs "server" as
the world wide web and your emacsclient as the browser. If you have two
clients open (web browsers) and the web site changes then both browsers
will show that when you refresh them. Ditto with emacsclient looking at
the emacs process buffers. Not the best analogy but there is oodles out
there in google land about servers and clients.

http://emacs-fu.blogspot.com/2009/02/emacs-daemon.html

The benefits should be self evident: start an emacs process ONCE and
connect to it anytime with negligible start up time - the server is
already started and your emacsclient start is just about instantaneous.

Because the server does not care about the front end too much you can
have different front ends. e.g I have a text mode emacsclient
(emacsclient -nw) running in a linux "screen" session. In addition I
have a hot key set up to fire up an emacsclient on my X desktop. Want to
read my email?  I hit a hot key and the emacsclient is there in a
millisecond and I switch to my gnus buffer. Gnus was already open. No
start time for that either. I thien quite the client or disconnect from
my screen session and emacs server is still there : I can come back to
it later. No screen real estate used. I can keep an emacs server running
on my main development machine and then ssh into it from the other side
of the world and have a fully functioning emacsclient running using
either X or in text mode accessing all my details at my home machine.

Similarly I use it for handling mailto URLs in firefox. See here:

http://www.emacswiki.org/emacs/MailtoHandler#toc4

Particular worthy of being singled out is this line which many people
overlooked when it was added (me included) from earlier emacs server
days :-

emacsclient --alternate-editor="" -c

This will start the server if its not already started and *then* connect
the emacsclient to it.







^ permalink raw reply	[flat|nested] 9+ messages in thread

* Re: confused about emacs 23 in daemon mode
  2009-10-12  2:19   ` David Combs
@ 2009-10-12  4:04     ` Juanma Barranquero
  0 siblings, 0 replies; 9+ messages in thread
From: Juanma Barranquero @ 2009-10-12  4:04 UTC (permalink / raw)
  To: David Combs; +Cc: help-gnu-emacs

On Mon, Oct 12, 2009 at 04:19, David Combs <dkcombs@panix.com> wrote:

> But there's a bug, at least in my version:
>  The wonderful "g" command doesn't work right, as
>  I recall.  Maybe fixed more recently, but I bet
>  not.
>
>  Try it (rename *Occur* to *Occur-this-pat, then
>  do M-x occur again, but on "that"-pat, try "g"
>  on the renamed one, and see what happens on YOUR
>  (newer than mine, probably) version.

What is the problem you're seeing? And, which Emacs version?

    Juanma




^ permalink raw reply	[flat|nested] 9+ messages in thread

end of thread, other threads:[~2009-10-12  4:04 UTC | newest]

Thread overview: 9+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2009-09-13  1:35 confused about emacs 23 in daemon mode Rodrigo Amestica
2009-09-13  2:19 ` Tim X
2009-09-13  3:24   ` Daniel Pittman
     [not found]   ` <mailman.6556.1252812928.2239.help-gnu-emacs@gnu.org>
2009-09-13  7:08     ` Bruno Barbier
2009-10-12  2:34       ` David Combs
2009-10-12  2:47         ` Richard Riley
2009-10-12  2:19   ` David Combs
2009-10-12  4:04     ` Juanma Barranquero
2009-09-13  9:18 ` Peter Dyballa

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).