I recompiled OCC and pythonOcc, but still the exact same issue. It really
seems to be a bug with intel driver new functionalities (DRI2/UXA).
Unfortunately I don't knw much about those new stuff.
I found a bug report with apparently the same backtrace (from level 0 to 3).
They say they solved it in
I'll try that later (when at home). I hope it will work.
Thanks for your support.
Loïc Simon
On Mon, Dec 7, 2009 at 3:10 PM, Thomas Paviot wrote:
> Loic,
>
> When switching from Ubuntu 9.04 to 9.10, I decided to recompiled
> OpenCASCADE. Let me know if it's better.
>
> Thomas
>
> 2009/12/7 Simo
Loic,
When switching from Ubuntu 9.04 to 9.10, I decided to recompiled
OpenCASCADE. Let me know if it's better.
Thomas
2009/12/7 Simon Loic
> I think I have a very recent version of the driver (as I updated to the
> last kubuntu release 9.10 karmic koala and I maintain it up-to-date quite
> of
I think I have a very recent version of the driver (as I updated to the last
kubuntu release 9.10 karmic koala and I maintain it up-to-date quite often).
In fact I wonder if some thing changed in the new Intel's driver. Indeed,
before I upgraded ubuntu, I could run pythonOcc without any problem of
Hi Loïc,
This issue seems related to your video card driver. I browsed the OCC forum
and found this thread: http://www.opencascade.org/org/forum/thread_14229/.
Paul Jimenez points out that "Intel's OpenGL implementation seems to be
problematic with OpenCascade."
Be sure you use the latest version
I checked at home I still have this segmentation fault. I'm wondering if
it's not a bug in the intel driver. If you have any suggestions?
>> python Level1/Geometry/geometry_demos.py
Here is the backtrace:
"""
#0 0xb182d7c0 in intel_region_buffer () from /usr/lib/dri/i965_dri.so
#1 0xb181e545 in
The patch where I just solved problems relative to sync the samples with the
new SimpleGui. I was obtained via 'svn diff >>patch' run from the pythonOcc
root directory.
Note that I noticed some weird end of line encoding (like in
Level1/Geometry/geomplate.py) : ^M on emacs. I don't know how to hand
Absolutely, I fixed this issue. I will commit changes so that you can test
if it's ok.
Thomas
2009/12/3 Simon Loic
> Actually, I had issue yesterday to buiild smesh! But I'm not much into
> autoconf ...
> I tried it the same way as for salomegeom:
> >> autoreconf --install
> then it fails with
Actually, I had issue yesterday to buiild smesh! But I'm not much into
autoconf ...
I tried it the same way as for salomegeom:
>> autoreconf --install
then it fails with
""
libtoolize: putting auxiliary files in AC_CONFIG_AUX_DIR, `build-aux'.
libtoolize: copying file `build-aux/config.guess'
libto
I will commit changes to the smesh repository this evening or tomorrow
morning. For the precompiled binaries, I think I'll just provide a bundled
version of pythonOCC-0.4 for Windows (the 'all-in-one', including OCC,
salomegeom and salomesh). For Linux/MacOS users, the compilation of geom and
smesh
2009/12/3 Simon Loic
> Thanks, Thomas
> Now everything is all right at work, compilatio went fine and uptodate
> samples don't crash (I'll check at home if everything is ok after update).
> If I understand well, the problem was that one swig interface (SGEOM.i) was
> generated (by py++) using an
Thanks, Thomas
Now everything is all right at work, compilatio went fine and uptodate
samples don't crash (I'll check at home if everything is ok after update).
If I understand well, the problem was that one swig interface (SGEOM.i) was
generated (by py++) using an older revision of salomegeometry
I agree with you on all issues, and i add that as long as i will release
5.1.2.6 (versioning is like a.b.c.d where "a, b, c" is the upstream release
is based on, and d is my release counter), propably you will have to prepare
on your svn the pyocc 0.5 so that it can follow any changes in the
libr
And if it can done within the two coming weeks, we can consider this set of
releases as a Christmas gift to the community ;-)
2009/12/3 Thomas Paviot
> Cool. According to me, salomegeom can be released as-is. I succesfully
> compiled/ran it over different platforms (Windows XP MSVC7/MSVC9, Ubunt
Cool. According to me, salomegeom can be released as-is. I succesfully
compiled/ran it over different platforms (Windows XP MSVC7/MSVC9, Ubuntu
9.04 and 9.10, MacOSX Snow Leopard 64 bits). It works just great.
It's a bit different for salomesmesh. I have modified the automake/autoconf
tools so tha
Geom at the moment has no severe bug or issue ( I just add some features
from time to time ) , so at any time i can create a release.The future plan
is to create a smarter GEOM_Solver class and also a new 2d sketcher driver
in GeomImpl.I am not so sure if it is worth to wait for such a thing, and
Hi Fotis,
No, it's ok. Being sync with your developments is a requirement. However,
when do you plan to release a new version of your work? It would definitely
be much easier for us to build pythonOCC upon a official downloadable
.tar.gz salomegeom package.
Loïc, I updated pythonOCC so that it's
In case it is needed i can leave a copy of the old API in the code.
Thomas let me know in case you need such a thing
Fotis
On Thu, Dec 3, 2009 at 2:16 PM, Thomas Paviot wrote:
> Loïc,
>
> I did not notice that Fotis commited a few changes these last few days. I
> have to update to the latest sv
Loïc,
I did not notice that Fotis commited a few changes these last few days. I
have to update to the latest svn rev. and regenerate the SWIG files. I'll
let you know when it's done.
Best regards,
Thomas
2009/12/3 Simon Loic
> Hi Thomas,
> I wanted to give a try on my work computer (previousl
Hi Thomas,
I wanted to give a try on my work computer (previously I was testing at
home) but I can't compile with --enable-geom.
If I build with the following command :
>> python setup.py build --enable_geom
I get
g++ -pthread -shared -Wl,-O1 -Wl,-Bsymbolic-functions
build/tem
Thanks Thomas,
I updated your latest commit, and launched the samples you updated
(Level1/Geometry/geometry_demos.py and
Level1/TopologyBuilding/topology_building.py - I don't have smesh so far) .
Unfortunately they both end with a seg fault after the viewer is created.
Here is the output.
"""
Dis
Hi Loïc,
The SmpleGui.py module is an improvement over the previous wxSamplesGui that
enables multiple graphical backends. For instance, if you decide whether to
use SimpleGui to manage the display, you first have to set the graphical
backend to use. Fos instance:
set_backend('wx') #if you want t
22 matches
Mail list logo
