[Orgmode] Re: Advice sought on managing decision alternatives.

[Carsten Dominik]

Hi Tom,

I have now added org-choose.el, it is part of the current git status.

A few comments:

On Feb 7, 2009, at 9:46 PM, Tom Breton (Tehom) wrote:

> Hi, Carsten.  Here is the new patch to org.el and the new
> org-choose.el
>
> A couple of notes:
>
> * As we talked about, "decisions" and "chosenness" are now called
>   "choose" everywhere.

Great.

> * I was able to add the library-aware customization we talked about.

This is nice, I earned something new! :convert-widget.....

> * I also added new variable `org-todo-normal-interpretations' - see
>   explanation below.

See my comments below

> * New test file.  Essentially the same, with name replacement.

I have not run the tests myself yet.

> * Didn't append the example files; they are all unchanged from before.
>
> ******* About `org-todo-normal-interpretations'
>
> You said your idea was to make a generally useful system.  I noticed
> that one thing was still hard-coded.  It's the part of org-todo that
> finds the next entry:
>
>         (memq interpret '(sequence choose))
>         ...
>         (memq interpret '(type priority))

Yes, this is correct.  I appreciate you noticing this additional
point where changes have to be made.
However, for now I have opted for a different solution:  I made the
sequence interpretation the last test in the cond chain, so that
all interpretations that are not `type' will fall back to this  
mechanism.
I envision that we can add another hook if someone wants an additional
way of handling this.  I would like to minimize the number of variables
where an add-on has to insert itself.

I have commented the corresponding line which tries to add to the
non-existing variable org-todo-normal-interpretations' in org-choose.el

I hope you agree with this solution, if not let me know.

I think what is missing now is documentation.  It seems to me that
there should be some minimal documentation in org-choose.el,
and it would be great to get a tutorial on Worg which describes
this in more detail.

Thanks a lot for this contribution, and for your precision and
attention to detail.

- Carsten