gerard.vermeulen@posteo.net writes:

>>> Still failing on my side (when running tests non-interactively from
>>> command line). To fix the problem, please use the approach from
>>> `test-org-list/indent-item':
>>> 
>>> 	    (transient-mark-mode 1)
>>> 	    (push-mark (point) t t)
>>> 
>>> Instead of (set-mark-command nil)
>> 
>> Gerard, may I know if you had a chance to look into my comments?
>
> I think that I have addressed this particular comment.

Not really.
In any case, see the attached updated patch with my suggestion
incorporated.

> However, I am confused by your comments concerning this example
> #+begin_src emacs-lisp
>    <mark>'(1 2 3)
>    '(1 2 <point>3)
> #+end_src
> since it breaks my patch as well as main in the sense that such examples
> can generate three blocks with invalid code.
> I think there is in general no way to protect a user against bad 
> selections
> for splitting by demarcation.

Fair.

> Secondly, I see (saw) sometimes Org emitting warnings with backtraces
> starting from my patch.  I think the warning may be due either to a 
> mistake
> on my side or a bug in Org, but I am not sure.

May you please provide more details?