Re: Bug: Export dispatcher scrolling with hidden top [9.3.6 (9.3.6-17-g389288-elpaplus @ /home/gustavo/.emacs.d/elpa/org-plus-contrib-20200224/)]

[Gustavo Barros <gusbrs.2016@gmail.com>]

[-- Attachment #1: Type: text/plain, Size: 2090 bytes --]

Hi Kyle,

thank you for looking into this.

On Tue, Jun 09 2020 at 01:48, Kyle Meyer <kyle@kyleam.com> wrote:

> Gustavo Barros writes:
>
>> this is a friendly ping on this export dispatcher regression.
>> (Considering 
>> https://lists.gnu.org/archive/html/emacs-orgmode/2020-06/msg00038.html).
>
> Hmm regression?  Do you suspect this worked for you at one time?  If 
> so,
> it'd be helpful if you tried to bisect the issue in the Git repo,
> especially if others can't reproduce it.
>

Hmm.  You are correct here.  "Regression" was a presumption, as it used 
to work for me, but stopped doing so eventually.  But I'd have a hard 
time telling you when, if you asked me.

Still, I did try to narrow it down as you asked.  And my first attempt 
was Emacs 26.3 with shipped-in Org version 9.1.9, and the behavior is 
already there.  So, something else triggered this for me, but the 
behavior itself is old, not a regression.

> I've tried to trigger it with your minimal configuration and haven't
> observed the displacement you describe, even when making my frame 
> height
> very small.

I was afraid there might be things in the environment which could affect 
this (DE? distro? monitor size? Emacs configure options? I don't know), 
but it does reproduce every single time for me with the recipe I 
provided.

Someone marked this as confirmed soon after I pinged, probably Bastien. 
Whoever did it, could you reproduce it?

Anyway, I attach a couple of images of what happens when I follow the 
provided recipe.  "figure 1" shows the state of the export dispatcher 
when it's called, the top part is hidden, and there is empty space 
below.  Indeed, the dispatcher would fit the space, except that it is 
displaced.  "figure 2" shows state after I hit "DEL", which is the 
expected initial state.  I can then no longer go back to the state in 
"figure 1", as the dispatcher fits, and "SPC", "DEL", "C-n", "C-p" echo 
"beginning/end of buffer", as expected.

If you'd like me to further investigate locally some possibility which 
might be relevant, just let me know.

Best,
Gustavo.


[-- Attachment #2: figure 1.png --]
[-- Type: image/png, Size: 45623 bytes --]

[-- Attachment #3: figure 2.png --]
[-- Type: image/png, Size: 68454 bytes --]