Octave Help Mode?

Octave Help Mode?

[Christian Herenz]

Hi,

I have a problem with Ocatve Mode. C-x C-h (or M-x octave-help) does not work. 
It seems that octave-hlp.el cannot find the entrys, so no symbol can be 
described. However, the variable octave-help-files is set to "octave". Maybe 
this is a bug in my distribution?
I use openSUSE 10.2, and I get no error message, but as stated above I cannot 
get help on anything.

At another system running Debian (Sarge) I get the error message 
"Info-find-node: Info file octave does not exist", but on the bash "info octave" 
shows the info file. However, it is not shown in M-x info (info-mode) inside emacs?

Any tips?

Greets,
Christian

P.S. Sorry, I first postd this wrongly to gnu.emacs - which seems not to be the 
right NG...

Re: Octave Help Mode?

[David Hansen]

[Cc: to emacs-devel]

On Tue, 15 Apr 2008 01:10:17 +0200 Christian Herenz wrote:

> I have a problem with Ocatve Mode. C-x C-h (or M-x octave-help) does
> not work. It seems that octave-hlp.el cannot find the entrys, so no
> symbol can be described. However, the variable octave-help-files is
> set to "octave". Maybe this is a bug in my distribution?
> I use openSUSE 10.2, and I get no error message, but as stated above I
> cannot get help on anything.

It's working here but I use a developer version of emacs.  Is C-h S
working?

To the emacs-devels:

Is there any need for `ocatave-help' if `info-lookup-symbol' is doing
the same (as far as I can see)?

> At another system running Debian (Sarge) I get the error message
> "Info-find-node: Info file octave does not exist", but on the bash
> "info octave" shows the info file. However, it is not shown in M-x
> info (info-mode) inside emacs?

Maybe adjust INFO_PATH but that sounds like a Debilian bug.

David

Re: Octave Help Mode?

[Stefan Monnier]

>> I have a problem with Ocatve Mode. C-x C-h (or M-x octave-help) does
>> not work. It seems that octave-hlp.el cannot find the entrys, so no
>> symbol can be described. However, the variable octave-help-files is
>> set to "octave". Maybe this is a bug in my distribution?
>> I use openSUSE 10.2, and I get no error message, but as stated above I
>> cannot get help on anything.

> It's working here but I use a developer version of emacs.  Is C-h S
> working?

> To the emacs-devels:

> Is there any need for `octave-help' if `info-lookup-symbol' is doing
> the same (as far as I can see)?

I haven't looked at octave-help, but maybe it is indeed redundant.
Can someone check?


        Stefan

Re: Octave Help Mode?

[David Hansen]

[-- Attachment #1: Type: text/plain, Size: 492 bytes --]

On Tue, 15 Apr 2008 11:55:17 -0400 Stefan Monnier wrote:

>> Is there any need for `octave-help' if `info-lookup-symbol' is doing
>> the same (as far as I can see)?
>
> I haven't looked at octave-help, but maybe it is indeed redundant.
> Can someone check?

Well, I just did.  It does indeed nothing more than `info-lookup-symbol'.

I suggest to remove octave-hlp.el and applying the attached patch.  But
it can't hurt to wait a bit and give the maintainer a chance to have his
word.

David


[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #2: Type: text/x-patch, Size: 2606 bytes --]

*** octave-mod.el.~1.44.~	2008-02-17 01:40:24.000000000 +0100
--- octave-mod.el	2008-04-15 20:13:12.000000000 +0200
***************
*** 31,44 ****
  ;; It defines Octave mode, a major mode for editing
  ;; Octave code.
  
- ;; The file octave-hlp.el provides `octave-help', a facility for looking up
- ;; documentation on a symbol in the Octave info files.
- 
  ;; The file octave-inf.el contains code for interacting with an inferior
  ;; Octave process using comint.
  
! ;; See the documentation of `octave-mode', `octave-help' and
! ;; `run-octave' for further information on usage and customization.
  
  ;;; Code:
  (require 'custom)
--- 31,41 ----
  ;; It defines Octave mode, a major mode for editing
  ;; Octave code.
  
  ;; The file octave-inf.el contains code for interacting with an inferior
  ;; Octave process using comint.
  
! ;; See the documentation of `octave-mode' and `run-octave' for further
! ;; information on usage and customization.
  
  ;;; Code:
  (require 'custom)
***************
*** 224,230 ****
      (define-key map "\C-c\M-\C-h" 'octave-mark-block)
      (define-key map "\C-c]" 'octave-close-block)
      (define-key map "\C-c\C-f" 'octave-insert-defun)
!     (define-key map "\C-c\C-h" 'octave-help)
      (define-key map "\C-c\C-il" 'octave-send-line)
      (define-key map "\C-c\C-ib" 'octave-send-block)
      (define-key map "\C-c\C-if" 'octave-send-defun)
--- 221,227 ----
      (define-key map "\C-c\M-\C-h" 'octave-mark-block)
      (define-key map "\C-c]" 'octave-close-block)
      (define-key map "\C-c\C-f" 'octave-insert-defun)
!     (define-key map "\C-c\C-h" 'info-lookup-symbol)
      (define-key map "\C-c\C-il" 'octave-send-line)
      (define-key map "\C-c\C-ib" 'octave-send-block)
      (define-key map "\C-c\C-if" 'octave-send-defun)
***************
*** 283,289 ****
      ["Submit Bug Report"	octave-submit-bug-report t]
      "-"
      ["Describe Octave Mode"	octave-describe-major-mode t]
!     ["Lookup Octave Index"	octave-help t])
    "Menu for Octave mode.")
  
  (defvar octave-mode-syntax-table
--- 280,286 ----
      ["Submit Bug Report"	octave-submit-bug-report t]
      "-"
      ["Describe Octave Mode"	octave-describe-major-mode t]
!     ["Lookup Octave Index"	info-lookup-symbol t])
    "Menu for Octave mode.")
  
  (defvar octave-mode-syntax-table
***************
*** 1517,1523 ****
       'octave-comment-char
       'octave-continuation-offset
       'octave-continuation-string
-      'octave-help-files
       'octave-send-echo-input
       'octave-send-line-auto-forward
       'octave-send-show-buffer))))
--- 1514,1519 ----

Re: Octave Help Mode?

[Christian Herenz]

David Hansen schrieb:
> [Cc: to emacs-devel]
> 
> On Tue, 15 Apr 2008 01:10:17 +0200 Christian Herenz wrote:
> 
>> I have a problem with Ocatve Mode. C-x C-h (or M-x octave-help) does
>> not work. It seems that octave-hlp.el cannot find the entrys, so no
>> symbol can be described. However, the variable octave-help-files is
>> set to "octave". Maybe this is a bug in my distribution?
>> I use openSUSE 10.2, and I get no error message, but as stated above I
>> cannot get help on anything.
> 
> It's working here but I use a developer version of emacs.  Is C-h S
> working?

I am not at my computer at the moment, so I cannot test this. However, I 
  noted yesterday, when browsing the GNU Octave Info File lots of Nodes 
which are shown in the index are actually not present. Is this right, or 
does openSUSE ship a wrong info file. Sorry, this tends to be more an 
octave question, but since I am using it inside Emacs, I ask here.

> 
> To the emacs-devels:
> 
> Is there any need for `ocatave-help' if `info-lookup-symbol' is doing
> the same (as far as I can see)?

Mh, maybe octave-help, if working properly should just search the octave 
manual, so if there is an ambivalence between a octave 'symbol' and 
something which is actually not octave, wrong results could be retrived.

> 
>> At another system running Debian (Sarge) I get the error message
>> "Info-find-node: Info file octave does not exist", but on the bash
>> "info octave" shows the info file. However, it is not shown in M-x
>> info (info-mode) inside emacs?
> 
> Maybe adjust INFO_PATH but that sounds like a Debilian bug.
> 

Do you mean the shell-variable INFO_PATH, which I would export in my 
.profile (or simialar), or do you mean something emacs-specific?
As stated above, just entering "info octave" in a normal terminal works 
fine.

The thing is, I really want to work with octave, and I just think it is 
good to look something up directly inside the environment I am doing my 
octave stuff (i.e. emacs), however since I am fan of printed works I 
ordered via my libary today the "GNU Octave Manual" by J.W. Eaton 
(Network Theory Ltd.). Since it is not present in my home-libary I have 
to wait a little bit (cause its inter-libary loan), but I will fill out 
some forms today, that more "official" printed GNU Documentation will be 
available in our libary soon. At present, there is actually none :(

Greets,
Christian

Re: Octave Help Mode?

[Peter Dyballa]

Am 15.04.2008 um 10:42 schrieb Christian Herenz:
> does openSUSE ship a wrong info file


It has to be. By default: or do you have *no* choice of packages to  
install? (If this is true, then the dir file(s) can be read-only.)

The dir files in the info directories are the important ones. The  
installation process should invoke install-info to include an entry  
for each info file to be installed. This step can fail because a  
modern install-info requests a particular format of the lines that  
are to be included in the dir file. I've seen that install-info did  
not work when the info file did not tell into which section the entry  
belongs.

--
Greetings

   Pete

The day Microsoft makes something that doesn't suck is the day they  
start selling vacuum cleaners.
				– Ernest Jan Plugge

Re: Octave Help Mode?

[David Hansen]

On Tue, 15 Apr 2008 10:42:41 +0200 Christian Herenz wrote:

> David Hansen schrieb:

>> It's working here but I use a developer version of emacs.  Is C-h S
>> working?
>
> I am not at my computer at the moment, so I cannot test this. However,
> I noted yesterday, when browsing the GNU Octave Info File lots of
> Nodes which are shown in the index are actually not present. Is this
> right, or does openSUSE ship a wrong info file. Sorry, this tends to
> be more an octave question, but since I am using it inside Emacs, I
> ask here.

This seems to be suses fault.  With Debians GNU Octave, version 3.0.0
the manual looks pretty complete here.

>>
>> To the emacs-devels:
>>
>> Is there any need for `ocatave-help' if `info-lookup-symbol' is doing
>> the same (as far as I can see)?
>
> Mh, maybe octave-help, if working properly should just search the
> octave manual,

That's exactly what info-lookup-symbol does.  It takes the info manual
that fits best to the current major mode.  E.g. in c-mode it would just
use the libc manual.

> so if there is an ambivalence between a octave 'symbol'
> and something which is actually not octave, wrong results could be
> retrived.

In this case "symbol" just means "something in the index" and defaults
to the (syntactically) symbol at the current cursor position.

>> Maybe adjust INFO_PATH but that sounds like a Debilian bug.
>>
>
> Do you mean the shell-variable INFO_PATH, which I would export in my
> .profile (or simialar), or do you mean something emacs-specific?

The shell variable

> As stated above, just entering "info octave" in a normal terminal
> works fine.

Hmm, you can try to set `Info-additional-directory-list' then, but this
actually looks more like an Emacs bug to me.

> The thing is, I really want to work with octave, and I just think it
> is good to look something up directly inside the environment I am
> doing my octave stuff (i.e. emacs), however since I am fan of printed
> works I ordered via my libary today the "GNU Octave Manual" by
> J.W. Eaton (Network Theory Ltd.). Since it is not present in my
> home-libary I have to wait a little bit (cause its inter-libary loan),
> but I will fill out some forms today, that more "official" printed GNU
> Documentation will be available in our libary soon. At present, there
> is actually none :(

You can also produce pdf or ps from the texinfo sources if that is good
enough for you.

David

Re: Octave Help Mode?

[Christian Herenz]

David Hansen schrieb:
> On Tue, 15 Apr 2008 10:42:41 +0200 Christian Herenz wrote:
>>> It's working here but I use a developer version of emacs.  Is C-h S
>>> working?
>> I am not at my computer at the moment, so I cannot test this. However,
>> I noted yesterday, when browsing the GNU Octave Info File lots of
>> Nodes which are shown in the index are actually not present. Is this
>> right, or does openSUSE ship a wrong info file. Sorry, this tends to
>> be more an octave question, but since I am using it inside Emacs, I
>> ask here.
> 
> This seems to be suses fault.  With Debians GNU Octave, version 3.0.0
> the manual looks pretty complete here.

I installed now from SUSEs Science Repo Octave 3.0.0, and now everything works 
as expected. Seems to be a bug in the default octave package which is shipped 
with 10.2.

>>> Maybe adjust INFO_PATH but that sounds like a Debilian bug.
>>>
>> Do you mean the shell-variable INFO_PATH, which I would export in my
>> .profile (or simialar), or do you mean something emacs-specific?
> 
> The shell variable
> 
>> As stated above, just entering "info octave" in a normal terminal
>> works fine.
> 
> Hmm, you can try to set `Info-additional-directory-list' then, but this
> actually looks more like an Emacs bug to me.

This is acutally what happens on the debian-machine I occasionally have to work 
on. But it seems, that sarge ships a rather old version of octave: GNU Octave, 
version 2.1.73. So would be nice, if a sarge user could report on this issue?

I would contact the sysadmin at our university, but it seems that they are 
playing a little bit the devils-advocate by promoting heavily Matlab (they won't 
give me full points, cause I solve my exercises in octave-code rather than with 
matlab (even pointing out the differences)... really pi**** about this atm, but 
this NEEDS to be discussed somewhere else (hints?))


> 
>> The thing is, I really want to work with octave, and I just think it
>> is good to look something up directly inside the environment I am
>> doing my octave stuff (i.e. emacs), however since I am fan of printed
>> works I ordered via my libary today the "GNU Octave Manual" by
>> J.W. Eaton (Network Theory Ltd.). Since it is not present in my
>> home-libary I have to wait a little bit (cause its inter-libary loan),
>> but I will fill out some forms today, that more "official" printed GNU
>> Documentation will be available in our libary soon. At present, there
>> is actually none :(
> 
> You can also produce pdf or ps from the texinfo sources if that is good
> enough for you.

Yeah, and I could print that out. But I really like to have a book. Maybe I am a 
bit old-fashioned in that way, but actually a browsing the information in a book 
than in a pile of printed paper (or even worse, a pdf on the screen) seems to be 
more natural to me. But just for looking up things quickly, the octave-help is best.

Greets,
Christian

Re: Octave Help Mode?

[David Hansen]

On Tue, 15 Apr 2008 16:32:43 +0200 Christian Herenz wrote:

> This is acutally what happens on the debian-machine I occasionally
> have to work on. But it seems, that sarge ships a rather old version
> of octave: GNU Octave, version 2.1.73. So would be nice, if a sarge
> user could report on this issue?

No, I'm running unstable.

> I would contact the sysadmin at our university, but it seems that they
> are playing a little bit the devils-advocate by promoting heavily
> Matlab (they won't give me full points, cause I solve my exercises in
> octave-code rather than with matlab (even pointing out the
> differences)... really pi**** about this atm, but this NEEDS to be
> discussed somewhere else (hints?))

If you don't have fascist disc quota:

./configure --prefix=$HOME && make && make install

David

Re: Octave Help Mode?

[Christian Herenz]

David Hansen schrieb:

>> I would contact the sysadmin at our university, but it seems that they
>> are playing a little bit the devils-advocate by promoting heavily
>> Matlab (they won't give me full points, cause I solve my exercises in
>> octave-code rather than with matlab (even pointing out the
>> differences)... really pi**** about this atm, but this NEEDS to be
>> discussed somewhere else (hints?))
> 
> If you don't have fascist disc quota:
> 
> ./configure --prefix=$HOME && make && make install
> 
> David
> 

That is a nice idea, but would not solve my problem that I was actually forced 
to hand in the results with matlab, otherwise I was told today by a professor my 
results would not be accepted. But I am working on this issue right now, cause 
my course is called "Computational Physics" not "Computational Physics with Matlab".

Greets,
Christian

Re: Octave Help Mode?

[Richard G Riley]

Christian Herenz <herenz@physik.hu-berlin.de> writes:

> David Hansen schrieb:
>
>>> I would contact the sysadmin at our university, but it seems that they
>>> are playing a little bit the devils-advocate by promoting heavily
>>> Matlab (they won't give me full points, cause I solve my exercises in
>>> octave-code rather than with matlab (even pointing out the
>>> differences)... really pi**** about this atm, but this NEEDS to be
>>> discussed somewhere else (hints?))
>>
>> If you don't have fascist disc quota:
>>
>> ./configure --prefix=$HOME && make && make install
>>
>> David
>>
>
> That is a nice idea, but would not solve my problem that I was
> actually forced to hand in the results with matlab, otherwise I was
> told today by a professor my results would not be accepted. But I am
> working on this issue right now, cause my course is called
> "Computational Physics" not "Computational Physics with Matlab".
>
> Greets,
> Christian

Why on earth do you feel that your Professor should master every
application out there that a student might feel is his right to use? Did
it cross your mind that he does not have the skill in Octave that you
have?

Re: Octave Help Mode?

[David Hansen]

On Tue, 15 Apr 2008 18:41:18 +0200 Richard G. Riley wrote:

> Christian Herenz <herenz@physik.hu-berlin.de> writes:
>
>> That is a nice idea, but would not solve my problem that I was
>> actually forced to hand in the results with matlab, otherwise I was
>> told today by a professor my results would not be accepted. But I am
>> working on this issue right now, cause my course is called
>> "Computational Physics" not "Computational Physics with Matlab".
>
> Why on earth do you feel that your Professor should master every
> application out there that a student might feel is his right to use? Did
> it cross your mind that he does not have the skill in Octave that you
> have?

Erm, if he teaches *computational* physics his skills should be good
enough to read and understand another language that is almost full
compatible to his favorite language.

In my numerical physics course we had completely free choice of the
language (but only got support for C ;).

And by the way, I think it's very questionable if it still counts as
"science" if you use proprietary software, if you don't have at least
full access to the source code.

David

Re: Octave Help Mode?

[Lennart Borgman (gmail)]

David Hansen wrote:
> And by the way, I think it's very questionable if it still counts as
> "science" if you use proprietary software, if you don't have at least
> full access to the source code.


It is good thinking, but I am sorry to say that too much of science 
today is "proprietary" in one sense or another. If you interested then 
there is a lot to study and think about. Some of my favorite questions are:

- Who decide what research to do? Why? Consequences?
- Who decide what framework/paradigm to use? Why? Consequences?
- Who decide what we can read without paying? Why? Consequences?
- Who decide who can do research? Why? Consequences?

Perhaps read Pierre Bourdieu, Michael Focault. There is a lot more than 
positivistic thinking to look into.

Re: Octave Help Mode?

[David Hansen]

On Tue, 15 Apr 2008 20:52:32 +0200 Lennart Borgman wrote:

> David Hansen wrote:
>> And by the way, I think it's very questionable if it still counts as
>> "science" if you use proprietary software, if you don't have at least
>> full access to the source code.
>
>
> It is good thinking, but I am sorry to say that too much of science
> today is "proprietary" in one sense or another. If you interested then
> there is a lot to study and think about. Some of my favorite questions
> are:
>
> - Who decide what research to do? Why? Consequences?
> - Who decide what framework/paradigm to use? Why? Consequences?
> - Who decide what we can read without paying? Why? Consequences?
> - Who decide who can do research? Why? Consequences?

As long as human beings are involved it's probably always about
money/power/fame and stuff.  But it's especially sad if people don't
even play by their own rules.

It's one of the prerequisites of science that the community can review
and check the results.  It's even worse when a lot of money is involved
e.g. the medical science.

If students try to use free alternatives to proprietary software it's at
least a very small step in the right direction.

And if a professor refuses a nearly full compatible alternative that's
pretty lame.

David

Re: Octave Help Mode?

[Lennart Borgman (gmail)]

Richard G Riley wrote:
> Christian Herenz <herenz@physik.hu-berlin.de> writes:
> 
>> David Hansen schrieb:
>>
>>>> I would contact the sysadmin at our university, but it seems that they
>>>> are playing a little bit the devils-advocate by promoting heavily
>>>> Matlab (they won't give me full points, cause I solve my exercises in
>>>> octave-code rather than with matlab (even pointing out the
>>>> differences)... really pi**** about this atm, but this NEEDS to be
>>>> discussed somewhere else (hints?))
>>> If you don't have fascist disc quota:
>>>
>>> ./configure --prefix=$HOME && make && make install
>>>
>>> David
>>>
>> That is a nice idea, but would not solve my problem that I was
>> actually forced to hand in the results with matlab, otherwise I was
>> told today by a professor my results would not be accepted. But I am
>> working on this issue right now, cause my course is called
>> "Computational Physics" not "Computational Physics with Matlab".
>>
>> Greets,
>> Christian
> 
> Why on earth do you feel that your Professor should master every
> application out there that a student might feel is his right to use? Did
> it cross your mind that he does not have the skill in Octave that you
> have?

Christian, I agree with Richard. Most professors I have seen are human 
beeing with adequate limitations in their capabilities.

But you do not have to give up on using Octave for that. Can't you for 
example talk to your fellow students to see if you as a group can ask 
the institution about the possibility to use Octave instead of Mathlab? 
Then some professor might get time to learn Octave enough.

To convince them you might have to learn some Matchlab and Scilab too 
... ;-)

Re: Octave Help Mode?

[Christian Herenz]

Lennart Borgman (gmail) schrieb:

>>> That is a nice idea, but would not solve my problem that I was
>>> actually forced to hand in the results with matlab, otherwise I was
>>> told today by a professor my results would not be accepted. But I am
>>> working on this issue right now, cause my course is called
>>> "Computational Physics" not "Computational Physics with Matlab".
>>>
>>
>> Why on earth do you feel that your Professor should master every
>> application out there that a student might feel is his right to use? Did
>> it cross your mind that he does not have the skill in Octave that you
>> have?

Afaik Computational Physics should not be about programming in a specific 
langauge. There are only some subtle diffrences in syntax and some commands may 
be better (faster) implemented in matlab.
I would understand if he is a FORTRAN guy and I want to write in C, but some 
Mathlab Scripts do run well in Octave an vice versa.

> 
> Christian, I agree with Richard. Most professors I have seen are human 
> beeing with adequate limitations in their capabilities.
> 
> But you do not have to give up on using Octave for that. Can't you for 
> example talk to your fellow students to see if you as a group can ask 
> the institution about the possibility to use Octave instead of Mathlab? 

Sometimes also fellow students can be limited in ther human capabilities (not 
all, of course). I even think some of mine even don't exactly know about the 
difference between "free" software and pirated software :(
(The understanding about the meaning of free software in general needs to be 
stronger communicated to general public I think, but this is really OT)

However, a small group is supporting me. I also have to talk to some of the 
people who help the professor in correcting the homework. Additionally I work 
togehter with a friend of mine, who is in favour of Matlab, but feels exactly 
like I do about "the need" to do the exercises in Matlab, so in case that there 
are differences between matlab and octave we can point these out.

> Then some professor might get time to learn Octave enough.

Hopefully he is willing to do so...

> 
> To convince them you might have to learn some Matchlab and Scilab too 

Thats actually my plan atm :-)

> ... ;-)

Greets,
Christian