Re: [DISCUSSION] org-capture.el vs remember.el (was: [ELPA] New package: jami-bot and org-jami-bot)

[João Távora]

On Sat, Dec 30, 2023 at 5:21 PM Stefan Kangas <stefankangas@gmail.com> wrote:
>
> Ihor Radchenko <yantar92@posteo.net> writes:
>
> > The refactoring de-coupled what used to be org-remember.el into
> > completely rewritten org-capture.el that added important features that
> > could not be implemented within remember.el framework:
> >
> > 1. org-capture arranges writing the text to remember directly into the
> >    target location where it is to be stored. This addressed an important
> >    problem of remember.el - remember.el always uses temporary buffer for
> >    user input and does not save it until remember-finalize is executed
> >    explicitly. A number of users sometimes forgot to call
> >    remember-finalize, leading to annoying data losses.
> >
> > 2. remember.el only allows a single text input at a time. org-capture
> >    makes it possible for users to write multiple notes in parallel.
> >
> > I am wondering if it would be possible to have the above two key
> > features supported by remember.el. Then, org-capture.el could go back to
> > re-using remember.el infrastructure and thus reduce code duplication.
>
> This is, of course, desirable.  The real question is how much work it
> will take and who will volunteer to do it.

Not volunteering but those two goals seems good, especially
the first one

> On the face of things, remember.el does not comprise a huge amount of
> code, and I don't see why the above two features couldn't be retrofitted
> into it.

Seems like the first one could be done with narrowing.
But would also need some kind of indirect buffer, too,
right?  How does org-capture.el do it?  It would
also mean "commiting to remember" is just C-x C-s and file
recovery works as normal, which is how it should be IMO.

Maybe after-save-hook in this indirect buffer specifically
would have the effect of quitting the window displaying the
indirect buffer? ... Well I just checked and the hook approach
doesn't work, because save-buffer always saves the base buffer.
But surely some other hook or some remapped key would do the
trick.

If this works (big if since given all this is vapourware), we'll
finally have found good use for indirect buffers :-)

The second one is just multiple such setups, which the
window-management code should in theory already allow.

João