Re: Unable to clock in on task

[Nick Dokos <nicholas.dokos@hp.com>]

Keith Lancaster <klancaster1957@acm.org> wrote:

> That did it! The patch fixes the problem - it now handles all the
> dangling clocks and then clocks in correctly. Thanks!
> 
> Keith
> 
> On Oct 27, 2009, at 9:36 AM, Nick Dokos wrote:
> 
> > Phil Rooke <phil@yax.org.uk> wrote:
> >
> >> Keith Lancaster <klancaster1957@acm.org> writes:
> >>
> >>> I updated to 6.32 this morning and am now unable to clock in on
> >>> tasks. ...
> >>
> >> I too upgraded this morning and am also having problems clocking in
> >> (using "I" on a task in the daily agenda).  My symptoms are
> >> different,
> >> Emacs doesn't crash but I do consistently get the following:
> >>
> >> Debugger entered--Lisp error: (wrong-type-argument stringp 126)
> >>  expand-file-name(126)
> >>  mapcar(expand-file-name "~/Documents/Org/org-agenda-files")
> >>  org-files-list()
> >>  org-resolve-clocks()
> >>  byte-code....
> >>  org-clock-in(nil)
> >>  org-agenda-clock-in(nil)
> >>  call-interactively(org-agenda-clock-in nil nil)
> >>
> >
> > This does look like a bug: org-resolve-clock calls org-files-list
> > which
> > does not know about the "agenda files list in a file" convention.  It
> > should probably call the function org-agenda-files instead of
> > org-files-list. org-files-list is only used in this one place and
> > should
> > probably be deleted.
> >
> > Try this patch:
> >
> > diff --git a/lisp/org-clock.el b/lisp/org-clock.el
> > index ea23a8d..2ce2f22 100644
> > --- a/lisp/org-clock.el
> > +++ b/lisp/org-clock.el
> > @@ -737,7 +737,7 @@ non-dangling (i.e., currently open and valid)
> > clocks."
> >   (interactive "P")
> >   (unless org-clock-resolving-clocks
> >     (let ((org-clock-resolving-clocks t))
> > -      (dolist (file (org-files-list))
> > +      (dolist (file (org-agenda-files))
> > 	(let ((clocks (org-find-open-clocks file)))
> > 	  (dolist (clock clocks)
> > 	    (let ((dangling (or (not (org-clock-is-active))
> >
> > Thanks,
> > Nick
> >
> 

Let me point out that the difference between Keith's original report and
Phil's report is that Phil included the backtrace - given that, the
patch was easy. So, Keith, you know what to do next time :-)

Even better, using org-submit-bug-report would have shown us that the
"agenda file list in a file" convention was being used - as you can
imagine, that's not the setting I use, so when I tried clocking in
yesterday, things worked without a problem for me.

Moralizingly-but-I-hope-not-overbearingly-so yours,
Nick