This post has NOT been accepted by the mailing list yet.
I've been running the older version of OpenSCAD that came with Ubunti 14.04 on an OLDER (10 years old) laptop, and have just switched to Ubuntu 16.04 on a "newer" laptop. I was not experiencing this problem before.
When I try to drag and rotate an object, there is severe lag in the graphical redraw. On the magnitude of taking 5 seconds to rotate 5 degrees. and precessing from there, refreshing a new rotation every few seconds until it finally catches up with me 30+ seconds later.
My current setup is:
Ubuntu (Lubuntu) 16.04
OpenSCAD 2015.03-1 (the Ubuntu package
Sony Vaio laptop with
Intel Corporation Mobile 4 Series Chipset Integrated Graphics Controller [8086:2a42] (rev 07)
I have looked into the video card, and from what I've found, this should use the i915 driver. Intel releases drivers for this card, and the latest version (from this spring) is listed on Intel's site as included in Ubuntu 16.04.
Can someone point me in a direction I can go to try to resolve this issue?
I really don't think I should be having a worse experience on new system with a faster processor, more RAM and a never version of openscad.
Hi, welcome to the forum. Your post is still flagged as "This post has NOT been accepted by the mailing list yet", so nobody gets it unless they look.
You need to subscribe to the mailing list, and respond to the registration email.
I have quoted it below for others to see.
Intel video has been problematic.
Until someone with more direct Intel based experience replies, try setting the
Tho I think that has more an effect on the graphics looking wrong, but you never know.
Oh, and posting the content of Help/Library-info will probably help.
Admin - email* me if you need anything,
or if I've done something stupid...
* click on my MichaelAtOz label, there is a link to email me.
Unless specifically shown otherwise above, my contribution is in the Public Domain; to the extent possible under law, I have waived all copyright and related or neighbouring rights to this work.
Obviously inclusion of works of previous authors is not included in the above.
The TPP is no simple “trade agreement.” Fight it! http://www.ourfairdeal.org/ time is running out!
Thanks MichaelAtOz. I'm subscribing to the mailing list now.
I have heard people complaining about Intel graphics in the past, but I was hoping that Intel providing drivers would help in this case. The laptop was free, so I can only complain a little.
I did try changing the Force Goldfeather and Enable for OpenGL 1.x options a few days ago, but it didn't seem to help. I'll try it again.
Here's my Help/Library-info
OpenSCAD Version: 2015.03-1
System information: Linux 4.4.0-42-generic #62-Ubuntu SMP Fri Oct 7 23:10:22 UTC 2016 i686 Ubuntu 16.04.1 LTS 2 CPUs 16777216.00 TB RAM
Compiler: GCC "5.2.1 20151028" 32bit
Boost version: 1_58
Eigen version: 3.2.5
CGAL version, kernels: 4.7, Cartesian, Extended_cartesian, Epeck
OpenCSG version: OpenCSG 1.4.0
Qt version: 5.5.1
QScintilla version: 2.9.1
MingW build: No
GLib version: 2.47.1
Application Path: /usr/bin
Documents Path: /home/leebc/.local/share
Resource Path: /usr/share/openscad
User Library Path: /home/leebc/.local/share/OpenSCAD/libraries
User Config Path: /home/leebc/.config/OpenSCAD
Backup Path: /home/leebc/.local/share/OpenSCAD/backups
OpenSCAD library path:
OpenSCAD font path:
GLEW version: 1.13.0
OpenGL Version: 2.1 Mesa 11.2.0
GL Renderer: Mesa DRI Mobile Intel® GM45 Express Chipset x86/MMX/SSE2
GL Vendor: Intel Open Source Technology Center
RGBA(8880), depth(24), stencil(8)
> I did try changing the Force Goldfeather and Enable
> for OpenGL 1.x options a few days ago, but it didn't
> seem to help. I'll try it again.
From what I've read, the Goldfeather algorithm tends to
be a bit slower than the default SCS algorithm, but helps
in some cases of driver issues. This does not seem to
be relevant in this specific case though.
> Can someone point me in a direction I can go to try to
> resolve this issue?
There are two issues related to display performance. One
is caused by changes in the GUI library. This at least
happens for you as the library info shows:
Qt version: 5.5.1
You could try with the nightly builds that still use
Qt4 where that specific issue is not there.
The Debian repository has a version 2015.03-2+dfsg-1
which includes the update for the new Display Widget,
(which would show up as "Using QOpenGLWidget") but is
stuck in the experimental repo for some reason.
The other issue is that the whole display code would
need some improvements to make use of the graphics card
feature available nowdays. I'm not aware someone is
working on that. So if the reason for the slow display
is "simply" a huge amount of vertices, there's no fix
at this time.
OpenSCAD mailing list
-- Update --
Thank you, Torsten.
I've since upgraded to "newer" hardware (circa 2011) with an AMD Mobility
Radeon HD 5430/5450/5470, but since I'm running the exact same OS and
packages, I had the same problem.
Per your suggestion, what I've done this morning is configure the PPA
and installed OpenSCAD 2015.03-2+dfsg-1~xenial as described here:
I now have:
OpenSCAD Version: 2015.03-2
Qt version: 5.5.1
It's DEFINATELY rendering and rotating much better than before.
I've tested 3 SCAD files with good results:
A model with many unions, differences and intersections
An animation with a couple differences
and surface() PNG model.
All work great (though the surface() is slow from a stupidly high number of
Thank you very much for your help.
I just realized that in my original "Lib & Build info" it incorrectly
shows that I had "16777216.00 TB RAM". My updated version correctly shows
1.89 GB RAM.
Should I file a bug on this?
OpenSCAD mailing list
On 10/30/2016 05:49 PM, Bryan Lee wrote:
> It's DEFINATELY rendering and rotating much better than
> All work great (though the surface() is slow from a
> stupidly high number of polygons).
Yep, that's the other issue I mentioned. This needs
quite some work, but that will happen eventually.
> I just realized that in my original "Lib & Build info" it
> incorrectly shows that I had "16777216.00 TB RAM". My
> updated version correctly shows 1.89 GB RAM.
> Should I file a bug on this?
If it shows up fine in the latest version, there's no
need for a bug report anymore.
OpenSCAD mailing list
In reply to this post by leebc
I don't think this is an Intel graphics issue; it is just worse under lower performing graphics cards (eg in laptops, or systems with only processor graphics). I have an NVIDIA GPU in my laptop and see the same problem (although I have a fairly high-end laptop). If you test with a fast desktop GPU or a simple model it may not show up. The model I am using also has a lot of CSG operations (which in preview mode involve multiple passes to render...).
I think what is happening is that the queue of commands to change the view generated when dragging the mouse is growing faster than the graphics system can render. Then even after the mouse is released, the queue drains only as fast as the system can render. Note the problem also occurs with panning, not just rotation (and panning is even more annoying as the model drifts out of view and you have to hunt around, carefully because of the lag, to bring it back).
With this issue the system is only barely usable as you have to reposition the model with very short "bursts". Also, this is not a pure performance issue: I have used exactly the same model and the same GPU to work with models under OpenSCAD in the past with no problem.
This is, simply put, a recently introduced user interface bug. What the system should do is "jump" to the last movement command given in the queue each time it does a render, flushing the event/command queue of any pending requests, to keep it from growing uncontrollably.
I should mention I'm using version 2015.03-1. I'll try upgrading to a more recent development version to see if the issue has been resolved.
I upgraded to 2015.03-2 (from the ppa) and the problem is gone. Unfortunately this version also fails to work with Intel graphics (which I also have on my Skylake laptop) under Ubuntu 16.04. I tried openscad-nightly as well. The problem is (still) fixed there, but that version also does not load properly with Intel graphics (it gets farther along, but fails to find certain OpenGL extensions).
|Free forum by Nabble||Edit this page|