Re: Property for startup visibility?

[Carsten Dominik <dominik@science.uva.nl>]

Hi Ben,

On May 7, 2008, at 8:29 PM, bva@alexanderonline.org wrote:

> Mostly I want the Short and Long Term projects visible and I'd like  
> to hide the calendars.  I like to keep all this in one file because  
> I'm most familiar with simple emacs searching, and I'm not yet good  
> at org-mode/agenda searches (or global TODO lists, or tag matching,  
> or all that)

I am starting to like the idea.

> Since I don't work on "Solve World Hunger" everyday, I find it a  
> distraction (as so many people do) to look at the TODOs under that  
> project, so a simple TODO tree search hasn't helped me much.  I know  
> there is more in org-mode that would fix this, but one of the big  
> pleasures of org-mode is how easy (easy, EASY!) it is to get started  
> on SOMETHING productive, without having to invest a huge shift in my  
> thinking.  (I am working on that shift in thinking, but it has been  
> slow to take hold.  Time to re-read the Allen book; maybe this time  
> I'll excise some more personal disorganizing demons)
>
> So I'd suggest that the 'VISIBILITY' property does get added.  I'd  
> even suggest adding another stop on the org-mode-tab-cycle chain  
> (FOLDED VISIBILITY-PROPERTY CHILDREN SUBTREE).

Definitely not.  Visibility cycling works because there is a clear  
sequence in it from showing little to showing everything.  Such an  
extra stop would kill intuition, I believe.  Instead, I could make a  
command, say `C-u C-u TAB' to switch back the entire buffer to the  
startup visibility.  And by the way, during C-c C-c on an option, org- 
mode-hook is not run again.

> I can even see feature-creep with this, as someone may want org-mode  
> to update the VISIBILITY property automatically on killing/saving  
> the buffer, so when the file is revisited, any hand-tuned folding is  
> recreated.

That has been requested before, but storing the visibility permanently  
in properties sound like overkill to me.  So no for now.


> Thought-experiment: Should the VISIBILITY property be automatically  
> inherited?

I dont think so.  Better to make possible values for the visibility  
that affect the entire tree below.  For example:

folded
children  (like the local "children" state)
content   (like the global "content" state, but local tree only
all

- Carsten