Re: [O] Should org-insert-todo-heading be a state change?

2017-07-14 Thread Nicolas Goaziou
Hello,

Allen Li  writes:

> Org mode allows logging timestamps and making notes when changing to a
> TODO keyword:
>
>TODO(!) | DONE(!)
>
> However, org-insert-todo-heading does not respect this.
>
> I'm wondering whether it should.  For example, given the above
> keywords, the intent would most likely be to log when TODO headings
> are created and when they are marked finished.
>
> If one were to create a heading and mark it TODO using C-c C-t, then
> the state change would be logged.  In other words, making a heading a
> TODO entry is considered a state change from " " to "some todo
> keyword".
>
> However, if one were to use C-S-RET, the new TODO heading would not
> have a timestamp logged.
>
> I can't think of a situation where the current behavior is desirable,
> so I am inclined to think that treating the creation of a TODO heading
> as a state change from a normal heading to a TODO heading would be
> more convenient.

My gut feeling is that the current behaviour is also correct, and
possibly less surprising. "changing" and "creating" are two different
concepts.

Besides, I suspect that many users create TODO headings with 

  M-RET T O D O SPC

in which case the behaviour you suggest would create a discrepancy.

Eventually, in contrast to TODO state changes, you need to be at the
future headline location to create it. IOW, it is very easy to add
a note upon creating the headline.

Regards,

-- 
Nicolas Goaziou



[O] Should org-insert-todo-heading be a state change?

2017-07-13 Thread Allen Li
Org mode allows logging timestamps and making notes when changing to a
TODO keyword:

   TODO(!) | DONE(!)

However, org-insert-todo-heading does not respect this.

I'm wondering whether it should.  For example, given the above
keywords, the intent would most likely be to log when TODO headings
are created and when they are marked finished.

If one were to create a heading and mark it TODO using C-c C-t, then
the state change would be logged.  In other words, making a heading a
TODO entry is considered a state change from " " to "some todo
keyword".

However, if one were to use C-S-RET, the new TODO heading would not
have a timestamp logged.

I can't think of a situation where the current behavior is desirable,
so I am inclined to think that treating the creation of a TODO heading
as a state change from a normal heading to a TODO heading would be
more convenient.