IFC Viewer ifcplusplus

This forum section is only for IFC-related issues
Post Reply
User avatar
bernd
Posts: 10908
Joined: Sun Sep 08, 2013 8:07 pm
Location: Zürich, Switzerland

Re: IFC Viewer ifcplusplus

Post by bernd » Sat Apr 12, 2014 8:44 am

If Qt4 is used IfcPlusPlus runs again using Linux for myself. :D

Find a version here at branch linuxbuild-qt4 https://github.com/berndhahnebach/IFCPl ... xbuild-qt4
See file README.Linux for information what I did.

I still would like to find out what could be done regarding the segmantation fault if Qt5 is used.
User avatar
yorik
Site Admin
Posts: 12046
Joined: Tue Feb 17, 2009 9:16 pm
Location: Brussels, Belgium
Contact:

Re: IFC Viewer ifcplusplus

Post by yorik » Mon Apr 14, 2014 3:00 am

Does it build correctly with the internal carve lib for you? It fails for me (qt4 version)... It seems that the carve library is not built? I get this:

Code: Select all

[ 96%] Building CXX object IfcPlusPlusGeometry/CMakeFiles/IfcPlusPlusGeometry.dir/src/ifcppgeometry/StylesConverter.cpp.o
Linking CXX shared library ../Release/libIfcPlusPlusGeometry.so
/usr/bin/ld: cannot find -lcarve
collect2: error: ld returned 1 exit status
make[2]: *** [Release/libIfcPlusPlusGeometry.so] Error 1
make[1]: *** [IfcPlusPlusGeometry/CMakeFiles/IfcPlusPlusGeometry.dir/all] Error 2
make: *** [all] Error 2
it seems that no libcarve.so or libcarve.a is built (required by -lcarve)... I can't find it anywhere in the build dir. Would I be missing something?
User avatar
bernd
Posts: 10908
Joined: Sun Sep 08, 2013 8:07 pm
Location: Zürich, Switzerland

Re: IFC Viewer ifcplusplus

Post by bernd » Mon Apr 14, 2014 10:44 am

yorik wrote:Does it build correctly with the internal carve lib for you? ...

it seems that no libcarve.so or libcarve.a is built (required by -lcarve)... I can't find it anywhere in the build dir. Would I be missing something?
Yes it does with internal carve. Branch linuxbuilt-qt4 was used and it runs smooth on my debian jessie.

I had to tell him explicit where the internal carve is. I don't know why? It should work without telling him this if one uses internal carve. Since I would like to switch to external carve again as soon as possible I didn't look to much in that. I just told him where the internal carve is. I even had to use absolut path.

See https://github.com/berndhahnebach/IFCPl ... ADME.Linux
Last edited by bernd on Mon Apr 14, 2014 10:52 am, edited 5 times in total.
User avatar
bernd
Posts: 10908
Joined: Sun Sep 08, 2013 8:07 pm
Location: Zürich, Switzerland

Re: IFC Viewer ifcplusplus

Post by bernd » Mon Apr 14, 2014 10:48 am

EDIT: Sorry double post, may be someone could delete.
User avatar
yorik
Site Admin
Posts: 12046
Joined: Tue Feb 17, 2009 9:16 pm
Location: Brussels, Belgium
Contact:

Re: IFC Viewer ifcplusplus

Post by yorik » Tue Apr 15, 2014 5:50 pm

Ah I managed to compile. There was a caps problem, when it tries to link to the carve lib, it uses -lcarve (which looks for a libcarve.a or libcarve.so file) but the compiled carve lib is Release/libCarve.so (big C). So I did a simlink named libcarve.so in the same folder and things worked.

But there seem to still be some problem with inverted normals... Is this happening for you too?
screenshot.jpg
screenshot.jpg (134.6 KiB) Viewed 2252 times
User avatar
bernd
Posts: 10908
Joined: Sun Sep 08, 2013 8:07 pm
Location: Zürich, Switzerland

Re: IFC Viewer ifcplusplus

Post by bernd » Tue Apr 15, 2014 9:54 pm

yorik wrote:... But there seem to still be some problem with inverted normals... Is this happening for you too?...
There where lots of bugfixes in Ifc++ but the FreeCAD exported files still have problems for me too. I'll check with FG again regarding this issue. I'm gone report informations as soon as I have new ones.

BTW: I'm really happy about all the progress FreeCAD makes regarding ifc support over the last couple of month and especiallly days.
User avatar
yorik
Site Admin
Posts: 12046
Joined: Tue Feb 17, 2009 9:16 pm
Location: Brussels, Belgium
Contact:

Re: IFC Viewer ifcplusplus

Post by yorik » Tue Apr 15, 2014 10:03 pm

Hmm you are right... This happens only with our exported test house!
I wonder if there wouldn't be an order to follow for the vertices along the wires in an IfcBrepFaceSet... I'll investigate...
Thomas Krijnen indicated me a huge reference manual: http://www.buildingsmart-tech.org/downl ... b.pdf/view

*EDIT* This seems to be the case http://www.buildingsmart-tech.org/ifc/I ... ebound.htm But the documentation is really hard to understand...
User avatar
bernd
Posts: 10908
Joined: Sun Sep 08, 2013 8:07 pm
Location: Zürich, Switzerland

Re: IFC Viewer ifcplusplus

Post by bernd » Tue Apr 15, 2014 10:14 pm

yorik wrote:.., This happens only with our exported test house! ...
Yes, it happens with FeeCAD exported files only and it happens with all FreeCAD exported files. But on the other side Ifc++ ist the only viewer which has such problems.

Thanks for the guide.
User avatar
yorik
Site Admin
Posts: 12046
Joined: Tue Feb 17, 2009 9:16 pm
Location: Brussels, Belgium
Contact:

Re: IFC Viewer ifcplusplus

Post by yorik » Tue Apr 15, 2014 10:30 pm

"This indicated whether (TRUE) or not (FALSE) the loop has the same sense when used to bound the face as when first defined. If sense is FALSE the senses of all its component oriented edges are implicitly reversed when used in the face." this is really a badly written definition... http://www.buildingsmart-tech.org/ifc/I ... ebound.htm No way to know what "first defined" means...
User avatar
bernd
Posts: 10908
Joined: Sun Sep 08, 2013 8:07 pm
Location: Zürich, Switzerland

Re: IFC Viewer ifcplusplus

Post by bernd » Wed Apr 16, 2014 4:43 am

I pointed Fabian to your last Forums post. Waiting and see what he says ...
Post Reply

Who is online

Users browsing this forum: No registered users and 1 guest