It looks as if for Terry, with the option ticked, it zooms around where the
mouse was when he double clicked but for me it zooms around where the mouse
currently is. So perhaps a bug.

On Thu, 24 Feb 2022 at 16:40, Terry terrypingm@gmail.com wrote:

That baffles me even more than the simple 'reversal'!

BTW, one other oddity I noticed was that while panning and rotation can be
applied with the mouse pointer anywhere (in the display window), that's not true
of zooming. At some point about a third of the way in from the left edge of the
display window, the wheel scrolls the editor window instead of zooming the
display.

Terry

On Thu, 24 Feb 2022 16:42:57 +0000, you wrote:

With that option enabled on Mac (220114) single click anywhere on the view screen sets that point as the zoom center.

Btw I never knew that until I saw this thread and it really would have been a great help last year !

Bob

On 2/24/2022 9:23 AM, Bob Carter wrote:

With mouse-centric zoom enabled, clicking shouldn't make any difference
to the zoom.  The zoom should be centered around the current mouse
location; whatever is under the mouse stays in the same position on the
screen.

With mouse-centric zoom disabled, clicking also shouldn't make any
difference to the zoom.  The zoom should be centered around the current
view origin; whatever is at the view origin stays in the same position
on the screen.

With or without mouse-centric zoom enabled, double-clicking anywhere on
the model should that point on the model be the view origin.  It has to
be a point on the model, not just empty space, because it sets the view
origin in three dimensions.

On Thu, 2022-02-24 at 18:34 +0000, Jordan Brown wrote:

Weird.
If I single- or double-click anywhere, on the model or not, nothing
happens.

This is true for Versions 2019.05 and 2021.01 (git 41f58fe)

I am running on Ubuntu MATE, if that makes a difference, and 2021.01 is
an Appimage.

On Thu, Feb 24, 2022 at 10:58 AM Terry terrypingm@gmail.com wrote:

BTW, one other oddity I noticed was that while panning and rotation can be

This and the movie you posted are very much nonstandard behavior.  But I've
never heard of, nor seen it happening to anyone before, so it would be
difficult to reproduce.

It seems like mouse centric zooming (being on by default) is functioning
"as it should", in some small way ... but the actual problem is that
somehow the application can't correctly determine your cursor location
during wheel events.
I have no idea what would cause such a discrepancy, and it almost seems
like a bug in the Qt5 GUI framework that OpenSCAD relies on.

What version of OpenSCAD are you running?  And what operating system
version?
Do you possibly have Display Scaling set in the OS?
Is there anything unique about your mouse device, or the rest of your setup?

Can you maybe fix it by resizing the 3D Pane in relation to the editor (by
clicking and dragging the bar between their borders)?
What happens if you temporarily hide the editor, by checking the box from
the menu bar "Window -> Hide Editor" ?

• Hans

Terry, do I recall you have a Spacemouse?
If so are you Spacemouse zooming or regular mouse scroll-wheel zooming.
Does regular mouse zooming change with the Spacemouse disconnected/connected?

--
This email has been checked for viruses by AVG.
https://www.avg.com

Michael,

Yes, I've had a '3D Connexion Space Navigator' for years. Used exclusively in
(now rare) Google Earth sessions. And first thing I did this morning was unplug
it, with high expectations. Restarted OpenScad, but disappointingly no change.

My optimism was because it definitely sounded implicated in that obscure FreeCAD
issue I discussed in both the FreeCAD forum and here in 'Exporting CSG for work
in FreeCAD', 11th February. Never did satisfactorily resolve that. Abandoned it,
for now anyway. However, there may be fragments of 3DConnexion code that can
still cause trouble even when the Space Navigator is unplugged, what do I know?
So I'll pursue that when I get time. Meanwhile my focus is still on OpenSCAD.

Will shortly respond in some detail to Hans.

Terry

====================

On Fri, 25 Feb 2022 17:21:48 +1100, you wrote:

On Fri, Feb 25, 2022 at 12:15:51PM +0000, Terry wrote:

My guess is that the windows-driver was made while testing with a
certain specific application. That application had a "dead zone" for
the movement that causes zoom in openscad/freecad. Thus they tuned it
wrong and missed it during testing. And apparently the driver continues
to send these non-zero zoom-readings even without the hardware attached.

Roger. 

--
** R.E.Wolff@BitWizard.nl ** https://www.BitWizard.nl/ ** +31-15-2049110 **
**    Delftechpark 11 2628 XJ  Delft, The Netherlands.  KVK: 27239233    **
f equals m times a. When your f is steady, and your m is going down
your a is going up.  -- Chris Hadfield about flying up the space shuttle.

On Thu, 24 Feb 2022 23:49:40 -0600, you wrote:

Hans,

I'm using 2021.01 on a PC and the OS is Win 10 Version 20H2 (OS Build
19042.1526).

Here's a further video (63 MB)  to summarise current status:

https://www.dropbox.com/s/hyywxyktg5bomdg/OpenSCADMouseWheel2.mp4?raw=1

In the hope that it might offer possible clues, I've also attached images of all
my Preference > Settings (apart from 3D Print). Until yesterday's post by
Leonard I'd never looked at these. So they are all at their defaults, apart from
mouse-centric zoom which I've now disabled.

I assume you've also seem my comments about 3D Connexion.