Re: Inspecting behaviour of `make-overlay', is this expected?

[Jean Louis <bugs@gnu.support>]

* Marcin Borkowski <mbork@mbork.pl> [2022-12-25 11:05]:
> On 2022-12-24, at 09:36, Eli Zaretskii <eliz@gnu.org> wrote:
> 
> >> Date: Sat, 24 Dec 2022 10:52:47 +0300
> >> From: Jean Louis <bugs@gnu.support>
> >> 
> >> XXXXXXXXXX
> >> ^
> >> | position 0 in buffer
> >> 
> >> (setq my-overlay (make-overlay 0 10))
> >> (get-buffer-window (overlay-buffer my-overlay))
> >> (overlay-put my-overlay 'display "Hello") then I get:
> >> 
> >> HelloX 
> >> 
> >> instead of "Hello"
> >> 
> >> Is this normal behavior?
> >
> > Yes.  The 10 part says that the overlay _ends_ before position 10.
> > This is standard Emacs behavior with text properties and overlays: END
> > means the first character NOT affected by the property/overlay.
> 
> Somewhat related: https://fhur.me/posts/always-use-closed-open-intervals
> (not my piece, just found it on the internet; also I do not want to
> imply the OP doesn't know/understand that - but I consider this short
> article interesting & worthwhile also for others reading this).

I did not put attention, and can't, because the advise is partially
helpful for me to think of similar issues in future, but not to
discover similar issues in past.

We have in Lisp functions like:

(elt '(1 2 3) 0) ➜ 1
(nth 0 '(1 2 3)) ➜ 1
(nth 1 '(1 2 3)) ➜ 2

then for that reason I always think that first char in buffer is 0 and
not 1, and to be conclusive then I use (goto-char (point-min)) because
I never inspected multiple times (point-min) ➜ 1, aha, now I get it.

If the function `make-list' makes 3 elements:

(make-list 3 1) ➜ (1 1 1)

then I want first element be designated as `1' representng "1st" or
"first", but first is second element. It does not give me feeling of
consistency.

(nth 1 '(1 2 3)) ➜ 2

Things already programmed one has to inspect and determine what is
going on.

k
-- 
Jean

Take action in Free Software Foundation campaigns:
https://www.fsf.org/campaigns

In support of Richard M. Stallman
https://stallmansupport.org/