Org-mode release 4.58

Org-mode release 4.58

[Carsten Dominik]

Hi,

I have released a new version of Org-mode, at

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

Don't forget to press the refresh button in your browser, to make
sure it does not show you a cached version of the page.

Enjoy!

- Carsten

Changes in version 4.58
-----------------------

- Full undo support in the agenda.

     You can now use UNDO in the agenda.  This command will reverse 
changes
     to the agenda *and* the linked buffer if this change was make with
     a remove editing command from the agenda.
     I am quite, but not 100% sure that this does always the right
     thing, so serious testing would be very much appreciated.

- Finding stuck projects for the weekly GTD review.

     The idea for this goes back to a mail from Tim O Callaghan:
        http://thread.gmane.org/gmane.emacs.orgmode/518/focus=523

     Projects that are stuck in the GTD sense if they have no
     NEXT ACTIONS defined.
     For example, if in your GTD setup you identify projects with a
     tag PROJ, and if you have a TODO keyword NEXT to indicate next
     actions, you could do

       (setq org-stuck-projects '("+PROJ/-DONE" ("NEXT") nil))

     and then use `C-c a #' to get a list of stuck projets

     This feature is still experimental.  It does seem to work for me, 
but
     it may not yet fit other people's requirements, please give me 
feedback.

     This is implemented using a new extension feature that is described
     in appendix A3 in the manual, so people who know Lisp can hack it.

- Visiting projects (subtrees) in an indirect buffer.

   This uses a dedicated frame to display the tasks.  The command is
   bound to `C-c C-x b' in an Org-mode buffer, and to `b' in the agenda.

   This is based on a discussion with Eddward DeVilla and
   Jason F. McBrayer.

- auto-filling for comments works now under Emacs, I have not yet have
   time to fix the filling stuff under XEmacs.  Coming soon.?


Enjoy!

- Carsten

Re: Org-mode release 4.58

[Eddward DeVilla]

This is just after a quick test.  It looks like there is supposed to
be a menu item for open a subtree in an indirect buffer under
show/hide.  It doesn't do anything for me.  C-c C-x b and C-u C-c C-x
b both work.

I was also wondering if it be reasonable to add a variant to open the
subtree in the current window.  I already tweaked the example you
provided before to do it, but I didn't know if it would be reasonable
for org-mode to allow it.  I think I would also use it a good deal
from agenda.

Edd

On 12/5/06, Carsten Dominik <dominik@science.uva.nl> wrote:
> Hi,
>
> I have released a new version of Org-mode, at
>
> http://www.astro.uva.nl/~dominik/Tools/org/
>
> Don't forget to press the refresh button in your browser, to make
> sure it does not show you a cached version of the page.
>
> Enjoy!
>
> - Carsten
>
> Changes in version 4.58
> -----------------------
>
> - Full undo support in the agenda.
>
>      You can now use UNDO in the agenda.  This command will reverse
> changes
>      to the agenda *and* the linked buffer if this change was make with
>      a remove editing command from the agenda.
>      I am quite, but not 100% sure that this does always the right
>      thing, so serious testing would be very much appreciated.
>
> - Finding stuck projects for the weekly GTD review.
>
>      The idea for this goes back to a mail from Tim O Callaghan:
>         http://thread.gmane.org/gmane.emacs.orgmode/518/focus=523
>
>      Projects that are stuck in the GTD sense if they have no
>      NEXT ACTIONS defined.
>      For example, if in your GTD setup you identify projects with a
>      tag PROJ, and if you have a TODO keyword NEXT to indicate next
>      actions, you could do
>
>        (setq org-stuck-projects '("+PROJ/-DONE" ("NEXT") nil))
>
>      and then use `C-c a #' to get a list of stuck projets
>
>      This feature is still experimental.  It does seem to work for me,
> but
>      it may not yet fit other people's requirements, please give me
> feedback.
>
>      This is implemented using a new extension feature that is described
>      in appendix A3 in the manual, so people who know Lisp can hack it.
>
> - Visiting projects (subtrees) in an indirect buffer.
>
>    This uses a dedicated frame to display the tasks.  The command is
>    bound to `C-c C-x b' in an Org-mode buffer, and to `b' in the agenda.
>
>    This is based on a discussion with Eddward DeVilla and
>    Jason F. McBrayer.
>
> - auto-filling for comments works now under Emacs, I have not yet have
>    time to fix the filling stuff under XEmacs.  Coming soon.?
>
>
> Enjoy!
>
> - Carsten
>
>
>
> _______________________________________________
> Emacs-orgmode mailing list
> Emacs-orgmode@gnu.org
> http://lists.gnu.org/mailman/listinfo/emacs-orgmode
>

Re: Org-mode release 4.58

[Carsten Dominik]

On Dec 5, 2006, at 21:59, Eddward DeVilla wrote:

> This is just after a quick test.  It looks like there is supposed to
> be a menu item for open a subtree in an indirect buffer under
> show/hide.  It doesn't do anything for me.

Yes, stupid bug, fixed, thanks.


> C-c C-x b and C-u C-c C-x
> b both work.
>
> I was also wondering if it be reasonable to add a variant to open the
> subtree in the current window.  I already tweaked the example you
> provided before to do it, but I didn't know if it would be reasonable
> for org-mode to allow it.  I think I would also use it a good deal
> from agenda.

There is a variable to configure how this should be done:
org-indirect-tree-new-frame
However, it does not yet include `current-window', I'll add this, makes 
sense.

Thanks

- Carsten

Re: Org-mode release 4.58

[Jason F. McBrayer]

On Tue, Dec 05, 2006 at 02:59:48PM -0600, Eddward DeVilla wrote:
> I was also wondering if it be reasonable to add a variant to open the
> subtree in the current window.  I already tweaked the example you
> provided before to do it, but I didn't know if it would be reasonable
> for org-mode to allow it.  I think I would also use it a good deal
> from agenda.

I've tried this with the example provided before, and also replaced a
lot of the functions that switch from the agenda buffer to the
corresponding org-mode buffer with functions that switch to the
subtree indirect buffer.

Now I've switched back to the original, default behaviour.  The
problem with what I tried was that I tended to clock-in to the
indirect buffer, with the result that org-clock-marker tended to get
lost, making switching tasks or clocking out break.  I consider this
my fault, though, not an org-mode issue.

-- 
+-----------------------------------------------------------+
| Jason F. McBrayer                    jmcbray@carcosa.net  |
| A flower falls, even though we love it; and a weed grows, |
| even though we do not love it.            -- Dogen        |

Re: Org-mode release 4.58

[Carsten Dominik]

On Dec 5, 2006, at 22:14, Jason F. McBrayer wrote:
> I've tried this with the example provided before, and also replaced a
> lot of the functions that switch from the agenda buffer to the
> corresponding org-mode buffer with functions that switch to the
> subtree indirect buffer.
>
> Now I've switched back to the original, default behaviour.  The
> problem with what I tried was that I tended to clock-in to the
> indirect buffer, with the result that org-clock-marker tended to get
> lost, making switching tasks or clocking out break.  I consider this
> my fault, though, not an org-mode issue.

Yes, in this case clocking out will only be possible as long as the
indirect buffer lives.  However, we can easily put the clock marker
into the base buffer.  Try the following patch:


--- org.el.orig 2006-12-06 03:45:40.000000000 +0100
+++ org.el      2006-12-06 03:46:45.000000000 +0100
@@ -6805,7 +6805,7 @@
        (indent-relative)
        (insert org-clock-string " ")
        (setq ts (org-insert-time-stamp (current-time) 'with-hm 
'inactive))
-      (move-marker org-clock-marker (point))
+      (move-marker org-clock-marker (point) (buffer-base-buffer))
        (message "Clock started at %s" ts))))

  (defun org-clock-out (&optional fail-quietly)

Re: Org-mode release 4.58

[Jason F. McBrayer]

On Wed, 2006-12-06 at 03:52 +0100, Carsten Dominik wrote:
> Yes, in this case clocking out will only be possible as long as the
> indirect buffer lives.  However, we can easily put the clock marker
> into the base buffer.  Try the following patch:

I've tested the patch, and it appears to work correctly.

-- 
+-----------------------------------------------------------+
| Jason F. McBrayer                    jmcbray@carcosa.net  |
| A flower falls, even though we love it; and a weed grows, |
| even though we do not love it.            -- Dogen        |

Re: Org-mode release 4.58

[Bastien]

Hi Carsten,

Thanks for this new release.  I noticed some strange auto-fill related
bugs in 4.57, but it seems okay now.  Here is a little report:

- '*word* is not fontified because of leading single quote (')

- http:///this,url is not fontified because of the coma (,)

- org-agenda-list-stuck-projects contains a call to (match) which
  leads C-c a # to produce a "Wrong number of arguments" error.
  Removing the (match) is okay.

- Maybe a naive question because i don't know the .ics format, but are
  the warnings like "This is a deadline" mandatory ?  DL and S already
  indicate the status of the event, don't they?

- Again on .ics: my scheduled headlines often have links, those are
  not well displayed in any iCal i know.  What about striping links
  from their first part (i.e. [[A][B]] -> B)?

Some feature suggestion (for org v5 ??):

- What about C-u C-c C-e to force ignoring timestamps when
  (re)publishing  a project?

- Cycling through priorities does not allow to delete the priority.
  My guess is that [#A] -> [#B] -> [#C] -> nothing -> [#A] -> ...
  is more convenient.

- What about diary inclusion in timeline (C-c a L)?

- In org-mode each task is potentially a project, since each headline
  is potentially at the top of a new tree.  This default behavior is
  perfect but sometimes i need what i would call a frozen inline task.
  Something that is a task without being (un)foldable, something that
  is not at the top of a potential subtree.

  For example:

  ====================================================================
  * Project 1

  Text for project one.

  **: Frozen inline task for project one.

  Text for project one (won't be (un)folded by TABing the frozen task
  aboce)
  ====================================================================

  See my point?

  Don't know if anyone need it as well.  A natural syntax for this
  would be to add a column like in **: - C-c : would "freeze" a task.

Enough for today!  But i wouldn't be able to nagg you with such tiny
things if i were not using org-mode :)

Cheers,

-- 
Bastien

Re: Org-mode release 4.58

[Carsten Dominik]

On Dec 6, 2006, at 11:45, Bastien wrote:

> Hi Carsten,
>
> Thanks for this new release.  I noticed some strange auto-fill related
> bugs in 4.57, but it seems okay now.  Here is a little report:
>
> - '*word* is not fontified because of leading single quote (')

Emphasis, this is a never-ending story.  I'll check if it makes sense 
to change
this case, but there will always be special cases where things do not 
work...

>
> - http:///this,url is not fontified because of the coma (,)

This is on purpose, please use <...> or [[...]]] around URL's with 
special characters.

>
> - org-agenda-list-stuck-projects contains a call to (match) which
>   leads C-c a # to produce a "Wrong number of arguments" error.
>   Removing the (match) is okay.

Arrrgh, yes this is a bug.  Will be fixed soon.  Dammit, I messed up a
release again - and I thought I had implemented something cool :-(

> - Maybe a naive question because i don't know the .ics format, but are
>   the warnings like "This is a deadline" mandatory ?  DL and S already
>   indicate the status of the event, don't they?

Yes.  I don't remember why I put these, maybe to make things more clear.
Can you suggest a better output format?

>
> - Again on .ics: my scheduled headlines often have links, those are
>   not well displayed in any iCal i know.  What about striping links
>   from their first part (i.e. [[A][B]] -> B)?

Hmmmm, would it not be useful to click on them?  I am not sure what 
calendar programs would support that.  Need to make a few test myself.


>
> Some feature suggestion (for org v5 ??):
>
> - What about C-u C-c C-e to force ignoring timestamps when
>   (re)publishing  a project?

I don't understand this one.

>
> - Cycling through priorities does not allow to delete the priority.
>   My guess is that [#A] -> [#B] -> [#C] -> nothing -> [#A] -> ...
>   is more convenient.

Hmmm, yes, maybe.

>
> - What about diary inclusion in timeline (C-c a L)?

That is a no, it has no place there.

Use the agenda restricted to the current file:

C-c a 1 a, maybe make a custom command where you change the ndays 
parameter to get a longer listing?

>
> - In org-mode each task is potentially a project, since each headline
>   is potentially at the top of a new tree.  This default behavior is
>   perfect but sometimes i need what i would call a frozen inline task.
>   Something that is a task without being (un)foldable, something that
>   is not at the top of a potential subtree.
>
>   For example:
>
>   ====================================================================
>   * Project 1
>
>   Text for project one.
>
>   **: Frozen inline task for project one.
>
>   Text for project one (won't be (un)folded by TABing the frozen task
>   aboce)
>   ====================================================================
>
>   See my point?

I sort-of do.  I guess you would to allow tags for these? And have them
listed in TODO lists etc???????

This would be great, but I cannot do it (don't know how).
See my discussion about plain list items some time ago.

But hmmmm, your proposal does ring a bell somewhere - I'll put this 
into a background job, lets see what it does.

- Carsten

Re: Org-mode release 4.58

[Bastien]

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

>> - '*word* is not fontified because of leading single quote (')
>
> Emphasis, this is a never-ending story.  I'll check if it makes
> sense to change this case, but there will always be special cases
> where things do not work...

I just mentionned this one because french people like me often use
single quotes - but that's not that important (sure).

>> - http:///this,url is not fontified because of the coma (,)
>
> This is on purpose, please use <...> or [[...]]] around URL's with
> special characters.

Oops, i forgot that - sorry.

> Arrrgh, yes this is a bug.  Will be fixed soon.  Dammit, I messed up a
> release again - and I thought I had implemented something cool :-(

This *is* cool!   Except that i've found to many stuck projects...

>> - Maybe a naive question because i don't know the .ics format, but are
>>   the warnings like "This is a deadline" mandatory ?  DL and S already
>>   indicate the status of the event, don't they?
>
> Yes.  I don't remember why I put these, maybe to make things more clear.
> Can you suggest a better output format?

I think DL/S is clear enough.  Maybe another way to implement the
distinction between deadlines and scheduled events is to use VEVENT
and VTODO?  http://tools.ietf.org/html/rfc2445 shows a typical VTODO:

     BEGIN:VTODO
     UID:19970901T130000Z-123404@host.com
     DTSTAMP:19970901T1300Z
     DTSTART:19970415T133000Z
     DUE:19970416T045959Z
     SUMMARY:1996 Income Tax Preparation
     CLASS:CONFIDENTIAL
     CATEGORIES:FAMILY,FINANCE
     PRIORITY:1
     STATUS:NEEDS-ACTION
     END:VTODO

This introduce the DUE: and STATUS: elements.  I'm not sure about this
because it would make a strong difference between ical-scheduled and
ical-"deadlined" items (iCal would display them differently), and this
distinction is more flexible in org-mode.  Just to let you know in
case.

>> - Again on .ics: my scheduled headlines often have links, those are
>>   not well displayed in any iCal i know.  What about striping links
>>   from their first part (i.e. [[A][B]] -> B)?
>
> Hmmmm, would it not be useful to click on them?  I am not sure what
> calendar programs would support that.  Need to make a few test
> myself.

AFAIK, SUMMARY is not very suitable for storing URLs. You might add a
URL=http://... element, but you need to strip all emacs-relevant links
(like gnus: bbdb: vm: etc.).

By the way, if seen the combined ics output insert something like
(category ...) - why not using the CATEGORY: element?  

>> - What about C-u C-c C-e to force ignoring timestamps when
>> (re)publishing a project?
>
> I don't understand this one.

While trying to publish all projects from my ~/org/ directory, I
discovered that org-publish uses timestamps.

`org-publish-use-timestamps-flag' let us turn this off/on, but i'd
like to be able to bypass this flag manually and force re-publishing
of all files when needed (since timestamps might not be accurate).

>> - What about diary inclusion in timeline (C-c a L)?
>
> That is a no, it has no place there.
>
> Use the agenda restricted to the current file:
>
> C-c a 1 a, maybe make a custom command where you change the ndays
> parameter to get a longer listing?

Yes, i'll do that.  While playing with custom commands, i found that
tags-todo commands are wrongly associated with ??? Here is a patch.

--- /home/guerry/e/org/org.el           2006-12-07 10:40:08.000000000 +0100
+++ /home/guerry/e/org/org.el.orig	2006-12-07 10:25:40.000000000 +0100
@@ -7525,7 +7525,6 @@
 			   ((stringp type) type)
 			   ((eq type 'tags) "Tags query")
 			   ((eq type 'todo) "TODO keyword")
-			   ((eq type 'tags-todo) "Tags TODO")
 			   ((eq type 'tags-tree) "Tags tree")
 			   ((eq type 'todo-tree) "TODO kwd tree")
 			   ((eq type 'occur-tree) "Occur tree")

> But hmmmm, your proposal does ring a bell somewhere - I'll put this
> into a background job, lets see what it does.

All right!  Thanks.

-- 
Bastien

Re: Org-mode release 4.58

[Carsten Dominik]

On Dec 7, 2006, at 10:46, Bastien wrote:
>
>>> - Again on .ics: my scheduled headlines often have links, those are
>>>   not well displayed in any iCal i know.  What about striping links
>>>   from their first part (i.e. [[A][B]] -> B)?
>>
>> Hmmmm, would it not be useful to click on them?  I am not sure what
>> calendar programs would support that.  Need to make a few test
>> myself.
>
> AFAIK, SUMMARY is not very suitable for storing URLs. You might add a
> URL=http://... element, but you need to strip all emacs-relevant links
> (like gnus: bbdb: vm: etc.).

OK, I guess I'll just throw away the link for now.

>
> By the way, if seen the combined ics output insert something like
> (category ...) - why not using the CATEGORY: element?

Yes, why not.

>
>>> - What about C-u C-c C-e to force ignoring timestamps when
>>> (re)publishing a project?
>>
>> I don't understand this one.
>
> While trying to publish all projects from my ~/org/ directory, I
> discovered that org-publish uses timestamps.
>
> `org-publish-use-timestamps-flag' let us turn this off/on, but i'd
> like to be able to bypass this flag manually and force re-publishing
> of all files when needed (since timestamps might not be accurate).

Have you actually tried that?  I think it might already work.

- Carsten
>
> --- /home/guerry/e/org/org.el           2006-12-07 10:40:08.000000000 
> +0100
> +++ /home/guerry/e/org/org.el.orig	2006-12-07 10:25:40.000000000 +0100
> @@ -7525,7 +7525,6 @@
>  			   ((stringp type) type)
>  			   ((eq type 'tags) "Tags query")
>  			   ((eq type 'todo) "TODO keyword")
> -			   ((eq type 'tags-todo) "Tags TODO")
>  			   ((eq type 'tags-tree) "Tags tree")
>  			   ((eq type 'todo-tree) "TODO kwd tree")
>  			   ((eq type 'occur-tree) "Occur tree")

Thanks, fixed.

- Carsten

Re: Org-mode release 4.58

[Bastien]

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

>> While trying to publish all projects from my ~/org/ directory, I
>> discovered that org-publish uses timestamps.
>>
>> `org-publish-use-timestamps-flag' let us turn this off/on, but i'd
>> like to be able to bypass this flag manually and force re-publishing
>> of all files when needed (since timestamps might not be accurate).
>
> Have you actually tried that?  I think it might already work.

Yes, it works great!  Sorry for the noise.

-- 
Bastien