[emacs-wiki-discuss] Re: Broken Planner, cause unknown

[Edgar Gonçalves]

Yesterday afternoon, Sacha Chua wrote:
> address@hidden (Edgar Gonçalves) writes:
>
>> The patch worked fine on keeping cursor position. Now I have another strange
>> behaviour, (since my updates from patches 432-435), when updating plan pages,
>> planner opens other plan files. then modifies them, and asks me if I want to
>> kill their buffers, since they got opened. maybe a save isn't being done,
>> somewhere?
>
> I merged a patch from the Muse port of planner that mentioned touching that
> behavior. Aha! GNU Emacs behavior broke because it passed a double list to
> derived-mode-p. Patch-437 should do the trick.
I'm so sorry to inform that that last patch completelly ruined my \M-x plan
<RET> function... It went back to the old behaviour, an infinite recursion
showing this in the *messages* buffer (until I quit):

Wrote c:/home/wiki-src/plans/2005.07.01
Preparing diary...done
Wrote c:/home/wiki-src/plans/2005.07.01
Preparing diary...done
Wrote c:/home/wiki-src/plans/2005.07.01
Preparing diary...


So I set debug-on-quit to t, and found out:
(...)
  (when planner-id-update-automatically (planner-id-update-tasks-on-page))
  planner-id-update-tasks-maybe()
  run-hook-with-args-until-success(planner-id-update-tasks-maybe)
  basic-save-buffer()
  save-buffer()
  (progn (save-buffer))
  (if (buffer-modified-p) (progn (save-buffer)))
  (when (buffer-modified-p) (save-buffer))
  (let ((planner-sort-tasks-automatically ...) 
(planner-renumber-tasks-automatically ...) (planner-align-tasks-automatically 
...) (planner-renumber-notes-automatically ...) (planner-tasks-save-behavior 
nil)) (when (buffer-modified-p) (save-buffer)))
  (progn (let (... ... ... ... ...) (when ... ...)))
  (if (and (planner-derived-mode-p ...) (planner-page-name) (not ...)) (progn 
(let ... ...)))
  (when (and (planner-derived-mode-p ...) (planner-page-name) (not ...)) (let 
(... ... ... ... ...) (when ... ...)))
  (save-current-buffer (set-buffer buffer) (when (and ... ... ...) (let ... 
...)) (when (and buffer-list ...) (kill-buffer nil)))
  (with-current-buffer buffer (when (and ... ... ...) (let ... ...)) (when (and 
buffer-list ...) (kill-buffer nil)))
  (lambda (buffer) (with-current-buffer buffer (when ... ...) (when ... 
...)))(#<buffer 2005.07.01>)
  mapcar((lambda (buffer) (with-current-buffer buffer (when ... ...) (when ... 
...))) (#<buffer 2005.07.01> #<buffer planner-config.el> #<buffer  *Minibuf-1*> 
#<buffer *followup to Sacha Chua on gmane.emacs.wiki.general*> #<buffer 
*Article*> #<buffer *Summary nntp+news.gmane.org:gmane.emacs.wiki.general*> 
#<buffer *Tree*> #<buffer *Group*> #<buffer *BBDB*> #<buffer *scratch*> 
#<buffer  *Minibuf-0*> #<buffer  *Echo Area 0*> #<buffer  *Echo Area 1*> 
#<buffer  *code-conversion-work*> #<buffer Map_Sym.txt> #<buffer *JDEE bsh*> 
#<buffer .bbdb> #<buffer  *nntpd*> #<buffer  *Gnus agent overview*> #<buffer  
*gnus work*> #<buffer  *nnmail message-id cache*> #<buffer  *nnimap* mega> 
#<buffer  *nnimap* evaluator> #<buffer  *nnimap* tagus> #<buffer  *nnimap* esw> 
#<buffer  *canonical address*> #<buffer  *extract address components*> #<buffer 
 *Async Prefetch Article*> #<buffer  *Original Article*> #<buffer  *Gnus 
Backlog*> #<buffer  *split*> #<buffer  *server news.gmane.org nntp  *nntpd**> 
#<buffer .newsrc-dribble> #<buffer diary> #<buffer  *gnus article copy*> 
#<buffer *Messages*>))
  planner-save-buffers((#<buffer 2005.07.01> #<buffer planner-config.el> 
#<buffer  *Minibuf-1*> #<buffer *followup to Sacha Chua on 
gmane.emacs.wiki.general*> #<buffer *Article*> #<buffer *Summary 
nntp+news.gmane.org:gmane.emacs.wiki.general*> #<buffer *Tree*> #<buffer 
*Group*> #<buffer *BBDB*> #<buffer *scratch*> #<buffer  *Minibuf-0*> #<buffer  
*Echo Area 0*> #<buffer  *Echo Area 1*> #<buffer  *code-conversion-work*> 
#<buffer Map_Sym.txt> #<buffer *JDEE bsh*> #<buffer .bbdb> #<buffer  *nntpd*> 
#<buffer  *Gnus agent overview*> #<buffer  *gnus work*> #<buffer  *nnmail 
message-id cache*> #<buffer  *nnimap* mega> #<buffer  *nnimap* evaluator> 
#<buffer  *nnimap* tagus> #<buffer  *nnimap* esw> #<buffer  *canonical 
address*> #<buffer  *extract address components*> #<buffer  *Async Prefetch 
Article*> #<buffer  *Original Article*> #<buffer  *Gnus Backlog*> #<buffer  
*split*> #<buffer  *server news.gmane.org nntp  *nntpd**> #<buffer 
.newsrc-dribble> #<buffer diary> #<buffer  *gnus article copy*> #<buffer 
*Messages*>) t)
  byte-code(".Ã=.      .Ä .Å.ŠÆ .Å.Ç .)È .*.#.É       Ê\".*Å." 
[planner-tasks-file-behavior live-buffers 
planner-appt-task-use-appointments-section-flag close buffer-list nil 
planner-goto-today planner-appt-update planner-appt-update-appt-section-maybe 
planner-save-buffers t] 3)
  planner-id-update-tasks-on-page()
  (progn (planner-id-update-tasks-on-page))
  (if planner-id-update-automatically (progn (planner-id-update-tasks-on-page)))
  (when planner-id-update-automatically (planner-id-update-tasks-on-page))
  planner-id-update-tasks-maybe()
  run-hook-with-args-until-success(planner-id-update-tasks-maybe)
  basic-save-buffer()
(...)


This led me to conclude that you must not have planner-id-update-tasks-maybe in
the buffer save hook. Or, you don't have this line in your config:
(planner-option-customized 'planner-id-update-automatically t)

So I switched it to nil - which is NOT the default value ;) - and everything
works like a charm. Except, offcourse, the automatic id update - which i don't 
find that
usefull. But I supose there may be who thinks otherwise!

>
> I'm not so sure about 'close being a reasonable default for
> planner-tasks-file-behavior. That could just be me and my ridiculously long 
> task
> lists, though. I keep it at nil because I don't mind having tons of open
> buffers... Hmm.
I like them closed. I mean, if I want a buffer opened, I'd like to be the one to
ask for it! However I don't usually toggle between planner buffers except by
plan links, when I'm working on a certain page...


I've made some suggestions in another post, one of this days I'll jot something
out to be able to have a fine planner the way I need it, and then I'll post it
here. Meanwhile, happy planning - and thanks for your kind help!

>
> -- 
> Sacha Chua <address@hidden> - open source geekette
> http://sacha.free.net.ph/ - PGP Key ID: 0xE7FDF77C
> interests: emacs, gnu/linux, personal information management, CS ed
> sachac on irc.freenode.net#emacs . YM: sachachua83

-- 
Edgar Gonçalves
Software Engineering Group @ INESC-ID
Portugal