Re: /srv/bzr/emacs/trunk r112148: Added eshell-tramp module

Re: /srv/bzr/emacs/trunk r112148: Added eshell-tramp module

[Aidan Gauland]

Stefan Monnier <monnier@iro.umontreal.ca> writes:

> So IIUC the way it's enabled is by loading the package and then calling
> eshell-tramp-initialize.

I think the modules' `initialize' functions are called by whatever part
of Eshell loads the modules, but I'm not certain.  This is one of the
yet-to-be-documented parts of Eshell.

> This said, IIUC just loading eshell-tramp.el will also enable it (tho
> not "completely").  That's a problem (and yes, this is not a new
> problem, it's just a problem in most/all eshell modules).

Yes, if the feature `em-tramp' is loaded but the module is not enabled,
then it is just taking up memory and doing nothing.  And, yes, this is a
problem with Eshell modules in general, and ERC modules, from what I can
tell.  I really have no idea what a better solution would be.

(I CC'd this message to emacs-devel, because it seems more general now
and less about one particular commit.)

--Aidan

Re: /srv/bzr/emacs/trunk r112148: Added eshell-tramp module

[Michael Albinus]

Aidan Gauland <aidalgol@no8wireless.co.nz> writes:

Hi Aidan,

> (I CC'd this message to emacs-devel, because it seems more general now
> and less about one particular commit.)

Beside the discussion, I believe your commit is also worth a ChangeLog
entry and some few words in /etc/NEWS.

> --Aidan

Best regards, Michael.

Re: [Emacs-diffs] /srv/bzr/emacs/trunk r112148: Added eshell-tramp module

[Stefan Monnier]

>> This said, IIUC just loading eshell-tramp.el will also enable it (tho
>> not "completely").  That's a problem (and yes, this is not a new
>> problem, it's just a problem in most/all eshell modules).
> Yes, if the feature `em-tramp' is loaded but the module is not enabled,
> then it is just taking up memory and doing nothing.

Just taking up memory is not a problem.  But IIUC since it defines
eshell/su, it will have an impact on eshell buffers, even if
eshell-tramp-initialize is not called, right?


        Stefan

Re: /srv/bzr/emacs/trunk r112148: Added eshell-tramp module

[Aidan Gauland]

Michael Albinus <michael.albinus@gmx.de> writes:
>
> Beside the discussion, I believe your commit is also worth a ChangeLog
> entry and some few words in /etc/NEWS.

I did commit with a ChangeLog entry, but not anything to NEWS.  What's
the policy regarding when to mention a change in NEWS?

--Aidan

(Sorry for the duplicate; I forgot to include the CC header.)

Re: /srv/bzr/emacs/trunk r112148: Added eshell-tramp module

[Aidan Gauland]

Stefan Monnier <monnier@iro.umontreal.ca> writes:

>>> This said, IIUC just loading eshell-tramp.el will also enable it (tho
>>> not "completely").  That's a problem (and yes, this is not a new
>>> problem, it's just a problem in most/all eshell modules).
>> Yes, if the feature `em-tramp' is loaded but the module is not enabled,
>> then it is just taking up memory and doing nothing.
>
> Just taking up memory is not a problem.  But IIUC since it defines
> eshell/su, it will have an impact on eshell buffers, even if
> eshell-tramp-initialize is not called, right?

Oh, no, that won't be a problem, because eshell-tramp-initialize has to
be called (for each buffer) to register the functions.

em-tramp.el line 51:

  (make-local-variable 'eshell-complex-commands)
  (setq eshell-complex-commands
        (append '("su" "sudo")
	        	    eshell-complex-commands))

So if the module is enabled (and loaded, of course), and you disable it
and then create a new eshell buffer, running the commands su and sudo
will invoke the external commands, not the internal eshell/su and
eshell/sudo.

--Aidan

(Sorry for the duplicate; I keep hitting R instead of F in Gnus.)

Re: /srv/bzr/emacs/trunk r112148: Added eshell-tramp module

[Eli Zaretskii]

> From: Aidan Gauland <aidalgol@no8wireless.co.nz>
> Date: Fri, 29 Mar 2013 09:00:58 +1300
> Cc: emacs-devel@gnu.org, Stefan Monnier <monnier@iro.umontreal.ca>,
> 	emacs-diffs@gnu.org
> 
> What's the policy regarding when to mention a change in NEWS?

When it is something users will want to know about.

Re: /srv/bzr/emacs/trunk r112148: Added eshell-tramp module

[Michael Albinus]

Aidan Gauland <aidalgol@no8wireless.co.nz> writes:

> I did commit with a ChangeLog entry

I don't see it.

Best regards, Michael.

Re: /srv/bzr/emacs/trunk r112148: Added eshell-tramp module

[Aidan Gauland]

Michael Albinus <michael.albinus@gmx.de> writes:

> Aidan Gauland <aidalgol@no8wireless.co.nz> writes:
>
>> I did commit with a ChangeLog entry
>
> I don't see it.

$ bzr diff -c 112148 ChangeLog
=== modified file 'ChangeLog'
--- ChangeLog   2013-03-25 17:58:35 +0000
+++ ChangeLog   2013-03-26 22:08:58 +0000
@@ -1,3 +1,13 @@
+2013-03-27  Aidan Gauland  <aidalgol@no8wireless.co.nz
+
+       * lisp/eshell/em-unix.el: Moved su and sudo to...
+       * lisp/eshell/em-tramp.el: ...Eshell tramp module
+       * doc/misc/eshell.texi: Updated manual to reflect changes.
+
+       External su and sudo commands are now the default; the internal,
+       TRAMP-using variants can still be used by enabling the eshell-tramp
+       module.
+
 2013-03-25  Jan Djärv  <jan.h.d@swipnet.se>
 
        * configure.ac (HAVE_XKB): Define if Xkb is present.

_______________________________________________
Emacs-diffs mailing list
Emacs-diffs@gnu.org
https://lists.gnu.org/mailman/listinfo/emacs-diffs

Re: /srv/bzr/emacs/trunk r112148: Added eshell-tramp module

[Aidan Gauland]

Eli Zaretskii <eliz@gnu.org> writes:

>> What's the policy regarding when to mention a change in NEWS?
>
> When it is something users will want to know about.

OK, now in which section should I mention this change?  "Changes in
Emacs" or "Changes in Specialized Modes and Packages"?  And for which
version?  24.4?

--Aidan

Re: [Emacs-diffs] /srv/bzr/emacs/trunk r112148: Added eshell-tramp module

[Aidan Gauland]

Stefan Monnier <monnier@IRO.UMontreal.CA> writes:

>> Oh, no, that won't be a problem, because eshell-tramp-initialize has to
>> be called (for each buffer) to register the functions.
>
>> em-tramp.el line 51:
>
>>   (make-local-variable 'eshell-complex-commands)
>>   (setq eshell-complex-commands
>>         (append '("su" "sudo")
>> 	        	    eshell-complex-commands))
>
>> So if the module is enabled (and loaded, of course), and you disable it
>> and then create a new eshell buffer, running the commands su and sudo
>> will invoke the external commands, not the internal eshell/su and
>> eshell/sudo.
>
> That's not my experience.
> After
>    M-: (defun eshell/cat (&rest args) (debug)) RET
> doing "cat ~/.cvsrc" does not show me ~/.cvsrc but jumps into the
> debugger, even though I have not touched eshell-complex-commands.

That would be because eshell/cat is already registered by em-unix

--Aidan

Re: /srv/bzr/emacs/trunk r112148: Added eshell-tramp module

[Eli Zaretskii]

> From: Aidan Gauland <aidalgol@no8wireless.co.nz>
> Cc: michael.albinus@gmx.de,  emacs-devel@gnu.org,  monnier@iro.umontreal.ca,  emacs-diffs@gnu.org
> Date: Fri, 29 Mar 2013 11:02:12 +1300
> 
> Eli Zaretskii <eliz@gnu.org> writes:
> 
> >> What's the policy regarding when to mention a change in NEWS?
> >
> > When it is something users will want to know about.
> 
> OK, now in which section should I mention this change?  "Changes in
> Emacs" or "Changes in Specialized Modes and Packages"?

I understand that this is a change in Eshell, so the latter seems
appropriate.  "Changes in Emacs" is for changes in core functionality.

> And for which version?  24.4?

Yes.

Re: /srv/bzr/emacs/trunk r112148: Added eshell-tramp module

[Michael Albinus]

Aidan Gauland <aidalgol@no8wireless.co.nz> writes:

> $ bzr diff -c 112148 ChangeLog

Ahh, you made it top-level. Lisp changes are documented in
lisp/ChangeLog; could you, please, move the entry there?

Same for the change in etc/NEWS; this shall be mentioned in etc/ChangeLog.

Best regards, Michael.

Re: [Emacs-diffs] /srv/bzr/emacs/trunk r112148: Added eshell-tramp module

[Stefan Monnier]

>> That's not my experience.
>> After
>> M-: (defun eshell/cat (&rest args) (debug)) RET
>> doing "cat ~/.cvsrc" does not show me ~/.cvsrc but jumps into the
>> debugger, even though I have not touched eshell-complex-commands.
> That would be because eshell/cat is already registered by em-unix

I tried again with `partx' instead of `cat' and got the same result, yet
`partx' does not appear in eshell's source code.


        Stefan

Re: [Emacs-diffs] /srv/bzr/emacs/trunk r112148: Added eshell-tramp module

[Aidan Gauland]

Stefan Monnier <monnier@iro.umontreal.ca> writes:

>>> That's not my experience.
>>> After
>>> M-: (defun eshell/cat (&rest args) (debug)) RET
>>> doing "cat ~/.cvsrc" does not show me ~/.cvsrc but jumps into the
>>> debugger, even though I have not touched eshell-complex-commands.
>> That would be because eshell/cat is already registered by em-unix
>
> I tried again with `partx' instead of `cat' and got the same result, yet
> `partx' does not appear in eshell's source code.

Oh... this is a problem, indeed.  I'll draw John Wiegley's attention to
this thread, and hopefully he can explain this.  I *thought* I knew how
this part of Eshell worked. :/

--Aidan