On 04/18/2011 04:57 PM, Richard Riley wrote: 
Carsten Dominik <carsten.dominik@gmail.com> writes:


On Apr 18, 2011, at 3:21 PM, Santi Villalba wrote:



          

            

              
Hi,

Whenever I invoke org-narrow-to-subtree, I receive a "blocking call to
accept-process-output with quit inhibited" error, the buffer does not
get narrowed down and I cannot edit or move anymore. I'm using a
bleeding-edge emacs 24.0.50.1 and org 7.4.


            

            
Could you please upgrade to org-mode 7.5 and see if the problem persists.


          

          
Done, no change. However, if I open emacs without reading the init files, the problem disappears. So it is about the interaction of org with something in my setup.


          

            

              
Does anybody know if the problem is it my setup or in a bug in org?


            

            
Please provide a backtrace.

M-x toggle-debug-on-error


          

          
That should open a *Backtrace* buffer, right? It does not happen.

I will bisect my .emacs later on and let you know if I find the incompatibility.

Santi



        

        
I am willing to bet that this is something funny with bleeding edge emacs.
Narrowing to a subtree has no reason to call out to a process.....

- Carsten

- Carsten



      

      
That prompt/message is indeed something in emacs 24. Its been like it
for ages and no sign yet of it being suppressed.


    

    I should have reported two different issues:

    

    1- The narrowing error was due to org not playing well with a couple
    of customisations of the scrolling behavior. I substituted these two
    lines with loading smooth-scrolling
    and the problem vanished.

        (setq scroll-conservatively 10)

         (setq scroll-margin 7)

    A better description of the problem I had is: when narrowing
    into a subtree with those settings, the window was getting frozen
    until I pressed tab. It has nothing to do with the error message and
    now it is fixed for me.

    

    2- The "blocking blah blah blah" error - that is only annoying but
    has no other bad consequences - happens when moving the point
    to/from the beginning and the end of the buffer in org-mode. It
    disappears when I disable flyspell-mode.

    

    In both cases I find difficult to send you a test case without
    sending my whole init files. Instructing emacs to start without
    loading the init files and just enabling the scrolling policy or
    enabling flyspell by hand are not enough to reproduce the problems.
    There must be some more faulty interplay in my setup, if I find out
    I will let you know.