Re: Attention XEmacs users

Re: Attention XEmacs users

[Tim O'Callaghan]

Problem with XEMACS noutline mode. This backtrace happens when i try
and cycle from OVERVIEW. 

Debugger entered--Lisp error: (void-variable set-extent-endpoints)
  (if (> (extent-end-position ex) end) (progn (set-extent-endpoints ... ... beg) (set-extent-endpoints ex end ...)) set-extent-endpoints ex (extent-start-position ex) beg)
  (if (< (extent-start-position ex) beg) (if (> ... end) (progn ... ...) set-extent-endpoints ex (extent-start-position ex) beg))
  (lambda (ex ignored) (if (< ... beg) (if ... ... set-extent-endpoints ex ... beg)) (if (> ... end) (set-extent-endpoints ex end ...) (delete-extent ex)))(#<extent [29943, 30105) I isearch-open-invisible outline 0x29d0908 in buffer PROJECTS.ORG> nil)
  map-extents((lambda (ex ignored) (if (< ... beg) (if ... ... set-extent-endpoints ex ... beg)) (if (> ... end) (set-extent-endpoints ex end ...) (delete-extent ex))) #<buffer "PROJECTS.ORG"> 30056 30105 nil end-closed outline)
  (save-excursion (map-extents (function ...) (current-buffer) beg end nil (quote end-closed) (quote outline)))
  outline-discard-extents(30056 30105)
  outline-flag-region(30056 30105 nil)
  outline-show-heading()
  (if (<= (funcall outline-level) level) (outline-show-heading))
  (lambda nil (if (<= ... level) (outline-show-heading)))()
  funcall((lambda nil (if (<= ... level) (outline-show-heading))))
  (while (and (progn ... ...) (not ...)) (funcall fun))
  (progn (goto-char (match-beginning 0)) (funcall fun) (while (and ... ...) (funcall fun)))
  (if (re-search-forward (concat "^\\(?:" outline-regexp "\\)") end t) (progn (goto-char ...) (funcall fun) (while ... ...)))
  (when (re-search-forward (concat "^\\(?:" outline-regexp "\\)") end t) (goto-char (match-beginning 0)) (funcall fun) (while (and ... ...) (funcall fun)))
  (save-excursion (setq end (copy-marker end)) (goto-char beg) (when (re-search-forward ... end t) (goto-char ...) (funcall fun) (while ... ...)))
  outline-map-region((lambda nil (if (<= ... level) (outline-show-heading))) 29294 30105)
  (save-excursion (outline-back-to-heading) (setq level (+ level ...)) (outline-map-region (lambda nil ...) (point) (progn ... ...)))
  (let (outline-view-change-hook) (save-excursion (outline-back-to-heading) (setq level ...) (outline-map-region ... ... ...)))
  show-children(1000)
  show-branches()
  (while (and (progn ... t) (looking-at outline-regexp)) (show-branches) (if (bobp) (throw ... nil)))
  (catch (quote exit) (while (and ... ...) (show-branches) (if ... ...)))
  (save-excursion (goto-char (point-max)) (catch (quote exit) (while ... ... ...)))
  org-content()
  (cond ((and ... ...) (message "CONTENTS...") (org-content) (message "CONTENTS...done") (setq org-cycle-global-status ...) (run-hook-with-args ... ...)) ((and ... ...) (show-all) (message "SHOW ALL") (setq org-cycle-global-status ...) (run-hook-with-args ... ...)) (t (org-overview) (message "OVERVIEW") (setq org-cycle-global-status ...) (run-hook-with-args ... ...)))
  (cond ((org-at-table-p ...) (or ... ...)) ((eq arg t) (cond ... ... ...)) ((integerp arg) (save-excursion ... ... ...)) ((save-excursion ... ...) (org-back-to-heading) (let ... ... ...)) (buffer-read-only (org-back-to-heading)) ((if ... t ...) (if ... ...) (indent-relative)) (t (save-excursion ... ...)))
  (let* ((outline-regexp ...) (bob-special ...) (org-cycle-hook ...) (pos ...)) (if (or bob-special ...) (setq arg t)) (cond (... ...) (... ...) (... ...) (... ... ...) (buffer-read-only ...) (... ... ...) (t ...)))
  org-cycle((4))
  (if (integerp arg) (progn (show-all) (hide-sublevels arg) (setq org-cycle-global-status ...)) (org-cycle (quote ...)))
  org-global-cycle(nil)
  call-interactively(org-global-cycle)
  (cond ((org-at-table-p) (call-interactively ...)) (t (call-interactively ...)))
  org-shifttab(nil)
  call-interactively(org-shifttab)


Tim.

Re: Attention XEmacs users

[Carsten Dominik]

Thanks for the report, I have fixed this.  Please find the modified 
version at

http://www.astro.uva.nl/~dominik/Tools/org/noutline.el

- Carsten

On Jun 22, 2006, at 17:35, Tim O'Callaghan wrote:

> Problem with XEMACS noutline mode. This backtrace happens when i try
> and cycle from OVERVIEW.
>
> Debugger entered--Lisp error: (void-variable set-extent-endpoints)
>   (if (> (extent-end-position ex) end) (progn (set-extent-endpoints 
> ... ... beg) (set-extent-endpoints ex end ...)) set-extent-endpoints 
> ex (extent-start-position ex) beg)
>   (if (< (extent-start-position ex) beg) (if (> ... end) (progn ... 
> ...) set-extent-endpoints ex (extent-start-position ex) beg))
>   (lambda (ex ignored) (if (< ... beg) (if ... ... 
> set-extent-endpoints ex ... beg)) (if (> ... end) 
> (set-extent-endpoints ex end ...) (delete-extent ex)))(#<extent 
> [29943, 30105) I isearch-open-invisible outline 0x29d0908 in buffer 
> PROJECTS.ORG> nil)
>   map-extents((lambda (ex ignored) (if (< ... beg) (if ... ... 
> set-extent-endpoints ex ... beg)) (if (> ... end) 
> (set-extent-endpoints ex end ...) (delete-extent ex))) #<buffer 
> "PROJECTS.ORG"> 30056 30105 nil end-closed outline)
>   (save-excursion (map-extents (function ...) (current-buffer) beg end 
> nil (quote end-closed) (quote outline)))
>   outline-discard-extents(30056 30105)
>   outline-flag-region(30056 30105 nil)
>   outline-show-heading()
>   (if (<= (funcall outline-level) level) (outline-show-heading))
>   (lambda nil (if (<= ... level) (outline-show-heading)))()
>   funcall((lambda nil (if (<= ... level) (outline-show-heading))))
>   (while (and (progn ... ...) (not ...)) (funcall fun))
>   (progn (goto-char (match-beginning 0)) (funcall fun) (while (and ... 
> ...) (funcall fun)))
>   (if (re-search-forward (concat "^\\(?:" outline-regexp "\\)") end t) 
> (progn (goto-char ...) (funcall fun) (while ... ...)))
>   (when (re-search-forward (concat "^\\(?:" outline-regexp "\\)") end 
> t) (goto-char (match-beginning 0)) (funcall fun) (while (and ... ...) 
> (funcall fun)))
>   (save-excursion (setq end (copy-marker end)) (goto-char beg) (when 
> (re-search-forward ... end t) (goto-char ...) (funcall fun) (while ... 
> ...)))
>   outline-map-region((lambda nil (if (<= ... level) 
> (outline-show-heading))) 29294 30105)
>   (save-excursion (outline-back-to-heading) (setq level (+ level ...)) 
> (outline-map-region (lambda nil ...) (point) (progn ... ...)))
>   (let (outline-view-change-hook) (save-excursion 
> (outline-back-to-heading) (setq level ...) (outline-map-region ... ... 
> ...)))
>   show-children(1000)
>   show-branches()
>   (while (and (progn ... t) (looking-at outline-regexp)) 
> (show-branches) (if (bobp) (throw ... nil)))
>   (catch (quote exit) (while (and ... ...) (show-branches) (if ... 
> ...)))
>   (save-excursion (goto-char (point-max)) (catch (quote exit) (while 
> ... ... ...)))
>   org-content()
>   (cond ((and ... ...) (message "CONTENTS...") (org-content) (message 
> "CONTENTS...done") (setq org-cycle-global-status ...) 
> (run-hook-with-args ... ...)) ((and ... ...) (show-all) (message "SHOW 
> ALL") (setq org-cycle-global-status ...) (run-hook-with-args ... ...)) 
> (t (org-overview) (message "OVERVIEW") (setq org-cycle-global-status 
> ...) (run-hook-with-args ... ...)))
>   (cond ((org-at-table-p ...) (or ... ...)) ((eq arg t) (cond ... ... 
> ...)) ((integerp arg) (save-excursion ... ... ...)) ((save-excursion 
> ... ...) (org-back-to-heading) (let ... ... ...)) (buffer-read-only 
> (org-back-to-heading)) ((if ... t ...) (if ... ...) (indent-relative)) 
> (t (save-excursion ... ...)))
>   (let* ((outline-regexp ...) (bob-special ...) (org-cycle-hook ...) 
> (pos ...)) (if (or bob-special ...) (setq arg t)) (cond (... ...) (... 
> ...) (... ...) (... ... ...) (buffer-read-only ...) (... ... ...) (t 
> ...)))
>   org-cycle((4))
>   (if (integerp arg) (progn (show-all) (hide-sublevels arg) (setq 
> org-cycle-global-status ...)) (org-cycle (quote ...)))
>   org-global-cycle(nil)
>   call-interactively(org-global-cycle)
>   (cond ((org-at-table-p) (call-interactively ...)) (t 
> (call-interactively ...)))
>   org-shifttab(nil)
>   call-interactively(org-shifttab)
>
>
> Tim.
>
>
> _______________________________________________
> Emacs-orgmode mailing list
> Emacs-orgmode@gnu.org
> http://lists.gnu.org/mailman/listinfo/emacs-orgmode
>
>

--
Carsten Dominik
Sterrenkundig Instituut "Anton Pannekoek"
Universiteit van Amsterdam
Kruislaan 403
NL-1098SJ Amsterdam
phone: +31 20 525 7477

Attention XEmacs users

[Carsten Dominik]

If you are using XEmacs to run org-mode, the following is important for 
you:

Starting with the next release of Org-mode (4.38, only a few days 
away), Org-mode will require the use of noutline.el, a new 
implementation of outline mode on which Org-mode is based.  This new 
implementation uses text overlays with invisibility properties to hide 
text.  This new implementation is already standard on GNU Emacs, so 
nothing needs to be done if you are using GNU Emacs.

XEmacs currently still uses the old implementation which is based on 
selective display to hide text.  This works great, but is much harder 
to program for.  Since there is now a port of noutline.el for XEmacs 
which seems to work stably, I will enforce using it, to make things 
much easier for the Org-mode side, and thus for me.  Therefore, 
starting with the next version of Org-mode you need to install 
noutline.el.  In the near future, I will slowly but surely remove the 
code needed for compatibility with the old outline.el.

The only way to keep me from doing what I just said is to download 
noutline.el *now*, test it and report problems with it or with the 
combination noutline/Org-mode.

You may download it at

    http://www.astro.uva.nl/~dominik/Tools/org/noutline.el

and install it in the usual way:  byte-compile it, put both noutline.el 
and noutline.elc on your load path.  And put (for now) (require 
'noutline) somewhere early into your .xemacs/init.el file.

Thanks for your understanding.

- Carsten

P.S.  All credit for noutline.el for XEmacs goes to Greg Chernev for 
porting it and allowing me to make it available.  Thanks Greg!

- Carsten



--
Carsten Dominik
Sterrenkundig Instituut "Anton Pannekoek"
Universiteit van Amsterdam
Kruislaan 403
NL-1098SJ Amsterdam
phone: +31 20 525 7477

Re: Attention XEmacs users

[Tim O'Callaghan]

On Wed, Jun 14, 2006 at 06:54:13PM +0200, Carsten Dominik wrote:
> 
> If you are using XEmacs to run org-mode, the following is important 
> for you:
> 
> Starting with the next release of Org-mode (4.38, only a few days 
> away), Org-mode will require the use of noutline.el, a new 
> implementation of outline mode on which Org-mode is based.  This new 
> implementation uses text overlays with invisibility properties to 
> hide text.  This new implementation is already standard on GNU Emacs, 
> so nothing needs to be done if you are using GNU Emacs.
> 
> XEmacs currently still uses the old implementation which is based on 
> selective display to hide text.  This works great, but is much harder 
> to program for.  Since there is now a port of noutline.el for XEmacs 
> which seems to work stably, I will enforce using it, to make things 
> much easier for the Org-mode side, and thus for me.  Therefore, 
> starting with the next version of Org-mode you need to install 
> noutline.el.  In the near future, I will slowly but surely remove the 
> code needed for compatibility with the old outline.el.
> 
> The only way to keep me from doing what I just said is to download 
> noutline.el *now*, test it and report problems with it or with the 
> combination noutline/Org-mode.
> 
> You may download it at
> 
>    http://www.astro.uva.nl/~dominik/Tools/org/noutline.el
> 
> and install it in the usual way:  byte-compile it, put both 
> noutline.el and noutline.elc on your load path.  And put (for now) 
> (require 'noutline) somewhere early into your .xemacs/init.el file.
> 
> Thanks for your understanding.
> 

Could you just add it into a 'contrib' type directory in the
distribution file?

I was thinking of suggesting the creation of a contrib directory where
you can throw things like this, as well as contributed perl scripts
and other stuff such as the other org mode extensions....

Tim.

Re: Attention XEmacs users

[Carsten Dominik]

On Jun 15, 2006, at 15:21, Tim O'Callaghan wrote:
>
> Could you just add it into a 'contrib' type directory in the
> distribution file?

Yes, I will distribute the file with Org-mode, but in an XEmacs 
directory, not in contrib.  If I put it in contrib, too many people 
will install it also under Emacs, which will break it and create too 
many emails for me :-)

Yes, we could have a contrib directory.  Any suggestions on how this 
should be structured?  It would need to have a mechanism to carry 
documentation, and to clearly mark things as not under my maintenance.

- Carsten

>
> I was thinking of suggesting the creation of a contrib directory where
> you can throw things like this, as well as contributed perl scripts
> and other stuff such as the other org mode extensions....
>
> Tim.
>
>
> _______________________________________________
> Emacs-orgmode mailing list
> Emacs-orgmode@gnu.org
> http://lists.gnu.org/mailman/listinfo/emacs-orgmode
>
>

--
Carsten Dominik
Sterrenkundig Instituut "Anton Pannekoek"
Universiteit van Amsterdam
Kruislaan 403
NL-1098SJ Amsterdam
phone: +31 20 525 7477

Re: Attention XEmacs users

[Tim O'Callaghan]

On Thu, Jun 15, 2006 at 07:38:03PM +0200, Carsten Dominik wrote:
> 
> On Jun 15, 2006, at 15:21, Tim O'Callaghan wrote:
> >
> >Could you just add it into a 'contrib' type directory in the
> >distribution file?
> 
> Yes, I will distribute the file with Org-mode, but in an XEmacs 
> directory, not in contrib.  If I put it in contrib, too many people 
> will install it also under Emacs, which will break it and create too 
> many emails for me :-)
> 
> Yes, we could have a contrib directory.  Any suggestions on how this 
> should be structured?  It would need to have a mechanism to carry 
> documentation, and to clearly mark things as not under my 
> maintenance.
> 

A simple Readme-<contribution>.txt possibly? if you published a
CVS/whatever tree as a 'src' distribution alongside the normal distro,
it would then be up to the contrib maintainers to send you patches
against it for distribution. This should work if you use the same type
of pre-release warning email you sent for each release. 

Tim.

Re: Attention XEmacs users

[Tim O'Callaghan]

I experimented with automating the noutline usage with a patch for org-install:

(if (featurep 'xemacs)
    (progn (setq dir (expand-file-name (concat default-directory "xemacs")))
           (if (file-directory-p dir)
               (setq load-path (cons dir load-path)))
           )
  )

The problem here is that default-directory was not set correctly at
runtime for some reason, so just added the xemacs dir to my load-path
instead. Its something you might consider though. 

Also ":" needed to be added to org-link-escape-chars, as below:
(defconst org-link-escape-chars '((":" . ":") ("[" . "%5B") ("]" . "%5D") (" " . "%20"))
  "Association list of escapes for some characters problematic in links.")

The fix is for a problem that occurs, when trying to open a remote
file link.  org-open-at-point mangles the EFS file links, possibly
tramp ones as well, i've not tried. It by replaces every ':' with '!'.

The link file:/user@machine:/disk:/home/dir/WRK/ tries to open the
file "/user@machine!/disk!/home/dir/WRK/"

Its probably not the best way of going about it, but it works.

Tim.

Re: Attention XEmacs users

[Carsten Dominik]

Hi Tim,


On Jun 20, 2006, at 14:09, Tim O'Callaghan wrote:
>
> Also ":" needed to be added to org-link-escape-chars, as below:
> (defconst org-link-escape-chars '((":" . ":") ("[" . "%5B") ("]" . 
> "%5D") (" " . "%20"))
>   "Association list of escapes for some characters problematic in 
> links.")

Is there a typo in this setting?  The above setting should lead to an 
infinite loop, because you keep replacing ":" with ":" when trying to 
escape the link characters.  And as far as I can see, org-mode passes 
the file name right through to `find-file'.  Could anyone try to 
reporduce this?

- Carsten

>
> The fix is for a problem that occurs, when trying to open a remote
> file link.  org-open-at-point mangles the EFS file links, possibly
> tramp ones as well, i've not tried. It by replaces every ':' with '!'.
>
> The link file:/user@machine:/disk:/home/dir/WRK/ tries to open the
> file "/user@machine!/disk!/home/dir/WRK/"
>
> Its probably not the best way of going about it, but it works.
>
> Tim.
>
>
> _______________________________________________
> Emacs-orgmode mailing list
> Emacs-orgmode@gnu.org
> http://lists.gnu.org/mailman/listinfo/emacs-orgmode
>
>

--
Carsten Dominik
Sterrenkundig Instituut "Anton Pannekoek"
Universiteit van Amsterdam
Kruislaan 403
NL-1098SJ Amsterdam
phone: +31 20 525 7477

Re: Attention XEmacs users

[Tim O'Callaghan]

On Tue, Jun 20, 2006 at 03:47:13PM +0200, Carsten Dominik wrote:
> Hi Tim,
> 
> 
> On Jun 20, 2006, at 14:09, Tim O'Callaghan wrote:
> >
> >Also ":" needed to be added to org-link-escape-chars, as below:
> >(defconst org-link-escape-chars '((":" . ":") ("[" . "%5B") ("]" . 
> >"%5D") (" " . "%20"))
> >  "Association list of escapes for some characters problematic in 
> >links.")
> 
> Is there a typo in this setting?  The above setting should lead to an 
> infinite loop, because you keep replacing ":" with ":" when trying to 
> escape the link characters.  And as far as I can see, org-mode passes 
> the file name right through to `find-file'.  Could anyone try to 
> reporduce this?
> 

Sorry, it was a quick hack that seemed to work, and now for some
reason does not. The issue itself still exists though. 

I've tracked it down to the convert-standard-filename call in
org-open-file. If i remove the convert-standard-filename, it does not
recognize it as an efs type file.

Anyway convert-standard-filename is what mangles efs type file
links. My guess is that it may be related to the C:\\filename problem
too.

Tim.

Re: Attention XEmacs users

[Carsten Dominik]

Yes, looks like this function does not do at all what I think it should 
be doing.  I'll remove the call, thanks for tracking this down.

- Carsten

On Jun 20, 2006, at 17:03, Tim O'Callaghan wrote:

> On Tue, Jun 20, 2006 at 03:47:13PM +0200, Carsten Dominik wrote:
>> Hi Tim,
>>
>>
>> On Jun 20, 2006, at 14:09, Tim O'Callaghan wrote:
>>>
>>> Also ":" needed to be added to org-link-escape-chars, as below:
>>> (defconst org-link-escape-chars '((":" . ":") ("[" . "%5B") ("]" .
>>> "%5D") (" " . "%20"))
>>>  "Association list of escapes for some characters problematic in
>>> links.")
>>
>> Is there a typo in this setting?  The above setting should lead to an
>> infinite loop, because you keep replacing ":" with ":" when trying to
>> escape the link characters.  And as far as I can see, org-mode passes
>> the file name right through to `find-file'.  Could anyone try to
>> reporduce this?
>>
>
> Sorry, it was a quick hack that seemed to work, and now for some
> reason does not. The issue itself still exists though.
>
> I've tracked it down to the convert-standard-filename call in
> org-open-file. If i remove the convert-standard-filename, it does not
> recognize it as an efs type file.
>
> Anyway convert-standard-filename is what mangles efs type file
> links. My guess is that it may be related to the C:\\filename problem
> too.
>
> Tim.
>
>
> _______________________________________________
> Emacs-orgmode mailing list
> Emacs-orgmode@gnu.org
> http://lists.gnu.org/mailman/listinfo/emacs-orgmode
>
>

--
Carsten Dominik
Sterrenkundig Instituut "Anton Pannekoek"
Universiteit van Amsterdam
Kruislaan 403
NL-1098SJ Amsterdam
phone: +31 20 525 7477

Re: Attention XEmacs users

[Carsten Dominik]

On Jun 20, 2006, at 17:11, Carsten Dominik wrote:

> Yes, looks like this function does not do at all what I think it 
> should be doing.  I'll remove the call, thanks for tracking this down.

Correction:  ...what I *thought* it was doing.  The function is fine, I 
was wrong putting the call there.... :-)

- Carsten

Re: Attention XEmacs users

[Carsten Dominik]

Maybe you are not trying to open the file in emacs?
Following a link picks an application to open the file, depending on 
the extension.  Under windows, the default is to use "open" for files, 
which is just like double-clicking them.  However, "open" may choke on 
an efs path.

You can force the link to open inside emacs with C-u C-c C-o.  Have you 
tried this?

- Carsten

On Jun 20, 2006, at 18:24, Tim O'Callaghan wrote:

> On Tue, Jun 20, 2006 at 05:10:27PM +0200, Carsten Dominik wrote:
>> Yes, looks like this function does not do at all what I think it
>> should be doing.  I'll remove the call, thanks for tracking this
>> down.
>>
>
> I've tracked it down to something in org-open file. It gets past the
> "No such file: " clause. Efs seems to be functioning well enough up
> until the point where the file actually gets opened.
>
>> - Carsten
>>
>> On Jun 20, 2006, at 17:03, Tim O'Callaghan wrote:
>>
>>> On Tue, Jun 20, 2006 at 03:47:13PM +0200, Carsten Dominik wrote:
>>>> Hi Tim,
>>>>
>>>>
>>>> On Jun 20, 2006, at 14:09, Tim O'Callaghan wrote:
>>>>>
>>>>> Also ":" needed to be added to org-link-escape-chars, as below:
>>>>> (defconst org-link-escape-chars '((":" . ":") ("[" . "%5B") ("]" .
>>>>> "%5D") (" " . "%20"))
>>>>> "Association list of escapes for some characters problematic in
>>>>> links.")
>>>>
>>>> Is there a typo in this setting?  The above setting should lead to
>>>> an
>>>> infinite loop, because you keep replacing ":" with ":" when trying
>>>> to
>>>> escape the link characters.  And as far as I can see, org-mode
>>>> passes
>>>> the file name right through to `find-file'.  Could anyone try to
>>>> reporduce this?
>>>>
>>>
>>> Sorry, it was a quick hack that seemed to work, and now for some
>>> reason does not. The issue itself still exists though.
>>>
>>> I've tracked it down to the convert-standard-filename call in
>>> org-open-file. If i remove the convert-standard-filename, it does
>>> not
>>> recognize it as an efs type file.
>>>
>>> Anyway convert-standard-filename is what mangles efs type file
>>> links. My guess is that it may be related to the C:\\filename
>>> problem
>>> too.
>>>
>>> Tim.
>>>
>>>
>>> _______________________________________________
>>> Emacs-orgmode mailing list
>>> Emacs-orgmode@gnu.org
>>> http://lists.gnu.org/mailman/listinfo/emacs-orgmode
>>>
>>>
>>
>> --
>> Carsten Dominik
>> Sterrenkundig Instituut "Anton Pannekoek"
>> Universiteit van Amsterdam
>> Kruislaan 403
>> NL-1098SJ Amsterdam
>> phone: +31 20 525 7477
>>
>
> Tim.
>
>

--
Carsten Dominik
Sterrenkundig Instituut "Anton Pannekoek"
Universiteit van Amsterdam
Kruislaan 403
NL-1098SJ Amsterdam
phone: +31 20 525 7477

Re: Attention XEmacs users

[Tim O'Callaghan]

On Tue, Jun 20, 2006 at 06:36:19PM +0200, Carsten Dominik wrote:
> Maybe you are not trying to open the file in emacs?
> Following a link picks an application to open the file, depending on 
> the extension.  Under windows, the default is to use "open" for 
> files, which is just like double-clicking them.  However, "open" may 
> choke on an efs path.
> 
> You can force the link to open inside emacs with C-u C-c C-o.  Have 
> you tried this?
> 

Yep, it looks like this is the problem. 


> - Carsten
> 
> On Jun 20, 2006, at 18:24, Tim O'Callaghan wrote:
> 
> >On Tue, Jun 20, 2006 at 05:10:27PM +0200, Carsten Dominik wrote:
> >>Yes, looks like this function does not do at all what I think it
> >>should be doing.  I'll remove the call, thanks for tracking this
> >>down.
> >>
> >
> >I've tracked it down to something in org-open file. It gets past the
> >"No such file: " clause. Efs seems to be functioning well enough up
> >until the point where the file actually gets opened.
> >
> >>- Carsten
> >>
> >>On Jun 20, 2006, at 17:03, Tim O'Callaghan wrote:
> >>
> >>>On Tue, Jun 20, 2006 at 03:47:13PM +0200, Carsten Dominik wrote:
> >>>>Hi Tim,
> >>>>
> >>>>
> >>>>On Jun 20, 2006, at 14:09, Tim O'Callaghan wrote:
> >>>>>
> >>>>>Also ":" needed to be added to org-link-escape-chars, as below:
> >>>>>(defconst org-link-escape-chars '((":" . ":") ("[" . "%5B") ("]" 
> >>>>>.
> >>>>>"%5D") (" " . "%20"))
> >>>>>"Association list of escapes for some characters problematic in
> >>>>>links.")
> >>>>
> >>>>Is there a typo in this setting?  The above setting should lead 
> >>>>to
> >>>>an
> >>>>infinite loop, because you keep replacing ":" with ":" when 
> >>>>trying
> >>>>to
> >>>>escape the link characters.  And as far as I can see, org-mode
> >>>>passes
> >>>>the file name right through to `find-file'.  Could anyone try to
> >>>>reporduce this?
> >>>>
> >>>
> >>>Sorry, it was a quick hack that seemed to work, and now for some
> >>>reason does not. The issue itself still exists though.
> >>>
> >>>I've tracked it down to the convert-standard-filename call in
> >>>org-open-file. If i remove the convert-standard-filename, it does
> >>>not
> >>>recognize it as an efs type file.
> >>>
> >>>Anyway convert-standard-filename is what mangles efs type file
> >>>links. My guess is that it may be related to the C:\\filename
> >>>problem
> >>>too.
> >>>
> >>>Tim.
> >>>
> >>>
> >>>_______________________________________________
> >>>Emacs-orgmode mailing list
> >>>Emacs-orgmode@gnu.org
> >>>http://lists.gnu.org/mailman/listinfo/emacs-orgmode
> >>>
> >>>
> >>
> >>--
> >>Carsten Dominik
> >>Sterrenkundig Instituut "Anton Pannekoek"
> >>Universiteit van Amsterdam
> >>Kruislaan 403
> >>NL-1098SJ Amsterdam
> >>phone: +31 20 525 7477
> >>
> >
> >Tim.
> >
> >
> 
> --
> Carsten Dominik
> Sterrenkundig Instituut "Anton Pannekoek"
> Universiteit van Amsterdam
> Kruislaan 403
> NL-1098SJ Amsterdam
> phone: +31 20 525 7477
> 
> 
> 
> _______________________________________________
> Emacs-orgmode mailing list
> Emacs-orgmode@gnu.org
> http://lists.gnu.org/mailman/listinfo/emacs-orgmode

Tim.

Re: Attention XEmacs users

[Carsten Dominik]

Maybe I should force efs/tramp file names to be opened with Emacs/.

What is the best and most general way to test if a file name is efs or 
tramp or ange-ftp?

- Carsten

On Jun 20, 2006, at 18:53, Tim O'Callaghan wrote:

> On Tue, Jun 20, 2006 at 06:36:19PM +0200, Carsten Dominik wrote:
>> Maybe you are not trying to open the file in emacs?
>> Following a link picks an application to open the file, depending on
>> the extension.  Under windows, the default is to use "open" for
>> files, which is just like double-clicking them.  However, "open" may
>> choke on an efs path.
>>
>> You can force the link to open inside emacs with C-u C-c C-o.  Have
>> you tried this?
>>
>
> Yep, it looks like this is the problem.
>
>
>> - Carsten
>>
>> On Jun 20, 2006, at 18:24, Tim O'Callaghan wrote:
>>
>>> On Tue, Jun 20, 2006 at 05:10:27PM +0200, Carsten Dominik wrote:
>>>> Yes, looks like this function does not do at all what I think it
>>>> should be doing.  I'll remove the call, thanks for tracking this
>>>> down.
>>>>
>>>
>>> I've tracked it down to something in org-open file. It gets past the
>>> "No such file: " clause. Efs seems to be functioning well enough up
>>> until the point where the file actually gets opened.
>>>
>>>> - Carsten
>>>>
>>>> On Jun 20, 2006, at 17:03, Tim O'Callaghan wrote:
>>>>
>>>>> On Tue, Jun 20, 2006 at 03:47:13PM +0200, Carsten Dominik wrote:
>>>>>> Hi Tim,
>>>>>>
>>>>>>
>>>>>> On Jun 20, 2006, at 14:09, Tim O'Callaghan wrote:
>>>>>>>
>>>>>>> Also ":" needed to be added to org-link-escape-chars, as below:
>>>>>>> (defconst org-link-escape-chars '((":" . ":") ("[" . "%5B") ("]"
>>>>>>> .
>>>>>>> "%5D") (" " . "%20"))
>>>>>>> "Association list of escapes for some characters problematic in
>>>>>>> links.")
>>>>>>
>>>>>> Is there a typo in this setting?  The above setting should lead
>>>>>> to
>>>>>> an
>>>>>> infinite loop, because you keep replacing ":" with ":" when
>>>>>> trying
>>>>>> to
>>>>>> escape the link characters.  And as far as I can see, org-mode
>>>>>> passes
>>>>>> the file name right through to `find-file'.  Could anyone try to
>>>>>> reporduce this?
>>>>>>
>>>>>
>>>>> Sorry, it was a quick hack that seemed to work, and now for some
>>>>> reason does not. The issue itself still exists though.
>>>>>
>>>>> I've tracked it down to the convert-standard-filename call in
>>>>> org-open-file. If i remove the convert-standard-filename, it does
>>>>> not
>>>>> recognize it as an efs type file.
>>>>>
>>>>> Anyway convert-standard-filename is what mangles efs type file
>>>>> links. My guess is that it may be related to the C:\\filename
>>>>> problem
>>>>> too.
>>>>>
>>>>> Tim.
>>>>>
>>>>>
>>>>> _______________________________________________
>>>>> Emacs-orgmode mailing list
>>>>> Emacs-orgmode@gnu.org
>>>>> http://lists.gnu.org/mailman/listinfo/emacs-orgmode
>>>>>
>>>>>
>>>>
>>>> --
>>>> Carsten Dominik
>>>> Sterrenkundig Instituut "Anton Pannekoek"
>>>> Universiteit van Amsterdam
>>>> Kruislaan 403
>>>> NL-1098SJ Amsterdam
>>>> phone: +31 20 525 7477
>>>>
>>>
>>> Tim.
>>>
>>>
>>
>> --
>> Carsten Dominik
>> Sterrenkundig Instituut "Anton Pannekoek"
>> Universiteit van Amsterdam
>> Kruislaan 403
>> NL-1098SJ Amsterdam
>> phone: +31 20 525 7477
>>
>>
>>
>> _______________________________________________
>> Emacs-orgmode mailing list
>> Emacs-orgmode@gnu.org
>> http://lists.gnu.org/mailman/listinfo/emacs-orgmode
>
> Tim.
>
>
> _______________________________________________
> Emacs-orgmode mailing list
> Emacs-orgmode@gnu.org
> http://lists.gnu.org/mailman/listinfo/emacs-orgmode
>
>

--
Carsten Dominik
Sterrenkundig Instituut "Anton Pannekoek"
Universiteit van Amsterdam
Kruislaan 403
NL-1098SJ Amsterdam
phone: +31 20 525 7477

Re: Attention XEmacs users

[Tim O'Callaghan]

On Tue, Jun 20, 2006 at 07:13:17PM +0200, Carsten Dominik wrote:
> Maybe I should force efs/tramp file names to be opened with Emacs/.
> 
> What is the best and most general way to test if a file name is efs 
> or tramp or ange-ftp?
> 

For tramp you'll need to RTM, but for efs its /[user@]host:/[file path]
IIRC ange-ftp uses the same syntax as efs.

You cold change the extension filter to a regexp filter i guess.

Tim.

Re: Attention XEmacs users

[Carsten Dominik]

I just found file-remote-p, which may be the right solution.

- Carsten

On Jun 20, 2006, at 19:29, Tim O'Callaghan wrote:

> On Tue, Jun 20, 2006 at 07:13:17PM +0200, Carsten Dominik wrote:
>> Maybe I should force efs/tramp file names to be opened with Emacs/.
>>
>> What is the best and most general way to test if a file name is efs
>> or tramp or ange-ftp?
>>
>
> For tramp you'll need to RTM, but for efs its /[user@]host:/[file path]
> IIRC ange-ftp uses the same syntax as efs.
>
> You cold change the extension filter to a regexp filter i guess.
>
> Tim.
>
>
> _______________________________________________
> Emacs-orgmode mailing list
> Emacs-orgmode@gnu.org
> http://lists.gnu.org/mailman/listinfo/emacs-orgmode
>
>

--
Carsten Dominik
Sterrenkundig Instituut "Anton Pannekoek"
Universiteit van Amsterdam
Kruislaan 403
NL-1098SJ Amsterdam
phone: +31 20 525 7477

Re: Attention XEmacs users

[Michael Olson]

[-- Attachment #1.1: Type: text/plain, Size: 1168 bytes --]

Carsten Dominik <dominik@science.uva.nl> writes:

> I just found file-remote-p, which may be the right solution.

XEmacs and Emacs22 both have this function, but in case you need
backwards compatibility with Emacs21, feel free to use something like
the following.  This handles both Tramp and ange-ftp file specifiers.

(defun org-file-remote-p (file)
  "Test whether FILE specifies a location on a remote system.
Return non-nil if the location is indeed remote.

For example, the filename \"/user@host:/foo\" specifies a location
on the system \"/user@host:\"."
  (cond ((fboundp 'file-remote-p)
         (file-remote-p file))
        ((fboundp 'tramp-handle-file-remote-p)
         (tramp-handle-file-remote-p file))
        ((and (boundp 'ange-ftp-name-format)
              (string-match ange-ftp-name-format file))
         t)
        (t nil)))

-- 
Michael Olson -- FSF Associate Member #652 -- http://www.mwolson.org/
Interests: Emacs Lisp, text markup, protocols -- Jabber: mwolson_at_hcoop.net
  /` |\ | | | IRC: mwolson on freenode.net: #hcoop, #muse, #PurdueLUG
 |_] | \| |_| Project involvement: Emacs, Muse, Planner, ERC, EMMS

[-- Attachment #1.2: Type: application/pgp-signature, Size: 190 bytes --]

[-- Attachment #2: Type: text/plain, Size: 149 bytes --]

_______________________________________________
Emacs-orgmode mailing list
Emacs-orgmode@gnu.org
http://lists.gnu.org/mailman/listinfo/emacs-orgmode