engits / engrid Goto Github PK

Original Report:

Original Report:

Original Report:

Hello!
I can not build a package in OBS.
OBS has a RPMLINT check and its output gave me the following errors:

[ 935s] E: engrid no-return-in-nonvoid-function /home/abuild/rpmbuild/BUILD/engrid-2.0.0git.20150324/src/build/libengrid/../../libengrid/cadinterface.h:125, 126
[ 935s] E: engrid no-return-in-nonvoid-function /home/abuild/rpmbuild/BUILD/engrid-2.0.0git.20150324/src/build/libengrid/../../libengrid/edgelengthsource.h:43
[ 935s] E: engrid no-return-in-nonvoid-function /home/abuild/rpmbuild/BUILD/engrid-2.0.0git.20150324/src/libengrid/cadinterface.h:125, 126
[ 935s] E: engrid no-return-in-nonvoid-function /home/abuild/rpmbuild/BUILD/engrid-2.0.0git.20150324/src/libengrid/cgaltricadinterface.cpp:292, 297
[ 935s] E: engrid no-return-in-nonvoid-function /home/abuild/rpmbuild/BUILD/engrid-2.0.0git.20150324/src/libengrid/edgelengthsource.h:43

Thank you for the attention, hope this report helps development.

Original Report:

Is BRLCAD import meant to be present in the import menu on enGrid 1.4.0?

Old bug tracker: http://sourceforge.net/apps/mantisbt/engrid/my_view_page.php

Original Report:

Hello,

I have tried to mesh a propeller blade with engrid. The surface mesh is created outside engrid with gmsh ("bamg" surface mesh algorithm).
I did the steps:

1. import Gmsh ascii v2
2. create prism boundary layer (default settings except "stretching factor" which is set to 1.7)
3. create/improve volume mesh
4. divide prismatic boundary layer
5. export OpenFOAM (grid only)
   Everything works without errors, but the final mesh contains negative volume cells where the trailing edge hits the hub. The number of problem cells increases, if the "stretching factor" in step 2 is decreased.
   Perhaps this case is of interest as testcase for further code development. I put it here: http://kroegeronline.net/exchange/rotor_domain.tgz

I used engrid built from the current commit of the "release-1.4" branch.

Regards, Hannes

Please consider writing and installing an AppData file with the application description and some screenshots, else enGrid looks really bad in the GNOME and KDE Software Centers. We'd love to showcase more applications, but without the extra data file we can't.

See http://people.freedesktop.org/~hughsient/appdata/ for details; thanks!

Richard

Original Report:

Seen in discussion http://engits.eu/vanilla/index.php?p=/discussion/287/compiling-error/ - there is a patch there: http://engits.eu/vanilla/index.php?p=/discussion/287/compiling-error/#Item_8

Two errors were triggered by Gcc 4.7.0:

1. page defined in MultiPageWidgetPage* page is a redefinition made of another variable a few lines above. I have no idea how any other compiler allowed that to happen, even MSVC! Either way, Gcc 4.7.0 is a lot more stricter.
2. File src/libengrid/guimainwindow.cpp - missing reference to dup and dup2, since they moved from their usual place onto a more proper place, namely unistd.h.

This problem was found on the master branch.

This has been fixed in commit aa67af7 on the windows branch only.

Since I don't have permissions to modify the release-1.4 branch, here's the report ;) :

The latest build.bash (for reference, this one: https://github.com/enGits/engrid/blob/f9a76cc450d0610e015b540ef45bb1ff138cc90e/build.bash) is relying on the https address for the git clone.

The problem is that the https port doesn't always work, since some firewalls have issues with it. Therefore, I suggest the following code instead of a single clone:

for url_address in git://github.com/enGits/engrid.git \
    https://github.com/enGits/engrid.git \
    git://repo.or.cz/engrid-github.git \
    http://repo.or.cz/r/engrid-github.git; do

  if git clone $url_address engrid ; then
    break;
  else
    echo "Repository $url_address failed. Trying the next one..."
  fi
done

The ones at repo.or.cz were mirrors setup by me, because we already know from OpenFOAM that Github doesn't always work as expected, while repo.or.cz has been more reliable for some companies to use (due to firewall restrictions).

Original Report:

Original Report:

Still need to finish updating instructions on the following wiki pages:

• [[enGrid 1.4] Building on Windows with MSVC2008](../wiki/[enGrid 1.4] Building on Windows with MSVC2008)
• [[enGrid 1.4] Building on Windows](../wiki/[enGrid 1.4] Building on Windows)
• [[enGrid 1.4] Building the final installer for Windows with NSIS](../wiki/[enGrid 1.4] Building the final installer for Windows with NSIS)

There are several links that are now outdated in the new wiki:

• Most are pertaining to the git repositories that enGrid has.
• Some are for example the build.bash scripts.

Still have to ask @ogloth what to do here...

Nodes sometime get projected to the wrong side of an object. A good example is a cylinder with just one boundary code for the whole 360 degrees. This is an issue with discrete (triangulated) geometries and might be handled by comparing the local mesh normal with the projected geometry normal.

Sometimes strange behaviour and bugs occur during the NETGEN library call. The problems often lead to memory flooding or crashes. Ideally problems should be detected and the NETGEN call should be aborted.

How this can be achieved is not clear to me at the moment.

While running the build script in Ubuntu 14.04, the only thing I noticed was "/usr/bin/ld: cannot find -ltcl8.5"

The fix is simple: add tcl8.5 to the apt-get line.
A newer version probably gets installed instead of this one.

After installing this package and running the build script again, engrid launches. I don't know if it works yet.

Original Report:

MSVC2008 gives this warning:

  libengrid\meshpartition.cpp(332) : warning C4553: '==' : operator has no effect; did you intend '='?

It refers to "found == true;" from here:

bool MeshPartition::hasBC(vtkIdType id_node, int bc)
{
  bool found = false;
  for (int j = 0; j < n2bcGSize(id_node); ++j) {
    if (n2bcG(id_node, j) == bc) {
      found == true;
      break;
    }
  }
  return found;
}

This is a clear indication that something is not right here...

Hi,
this is no real issue but I'm beginning with that techno and I'd like to know if there is a way to use engrid with command lines only?

Cheers

Original Report:

I am trying to compile the latest release-1.4 branch and I am getting the following errors:

guieditboundaryconditions.cpp: In member function ‘void GuiEditBoundaryConditions::setupSolvers()’:
guieditboundaryconditions.cpp:332:26: error: redeclaration of ‘MultiPageWidgetPage* page’
guieditboundaryconditions.cpp:312:3: error: ‘QString page’ previously declared here
make[2]: *** [release/guieditboundaryconditions.o] Error 1
make[2]: Leaving directory `/home/cfd/engrid/src/libengrid'
make[1]: *** [release] Error 2
make[1]: Leaving directory `/home/cfd/engrid/src/libengrid'
make: *** [sub-libengrid-libengrid-pro-make_default-ordered] Error 2

which branch should I use to compile under Linux?

Best regards,

George

On the wiki page [enGrid-1.4]-Installation, the instructions have been updated to expect the scripts (build.bash and setup_pathes.sh) to be fully merged from the branch window-1.4 to release-1.4.

Additionally, tests are still needed to confirm if these updates will work with the already mentioned Linux Distributions.

Further notes:

• As mentioned, on the branch windows-1.4 you can find the updated scripts build.bash and setup_pathes.sh. I haven't tested these yet, since these are hard-coded in build.bash to work with the branch release-1.4...
• ...And also because I haven't had time to test them either.
• I propose that we use this issue report to divide the test machines into two or more parts and test them on virtual and/or real machines.
• Already added openSUSE 12.2, Fedora 16+17 and Ubuntu 12.04 to the scripts and wiki page.

If the geometry exported from Blender has some loose geometry like detached vertices, Engrid segfaults upon create/improve surface mesh, even though engrid says the surface mesh is OK... so either the Blender export script has to make sure only surface related geometry is exported, or Engrid has to omit unused vertices when reading in...

Hello,

recently I got problems during import of gmsh surface meshes, created with latest version of gmsh. I guess this might be due to changes in the gmsh format.

For my case, the gmsh header contained an additional section between the keywords "$EndMeshFormat" and "$Nodes" with the patch names, which irritated engrid.
The current header looked like this:
$MeshFormat
2.2 0 8
$EndMeshFormat
$PhysicalNames
10
2 1 "rotor_upstream"
...
3 10 "blade"
$EndPhysicalNames
$Nodes
34428
...
.

To handle this, e.g. the following change around line 134 in gmshreader.cpp might be useful:

    /*
    f >> word;
    if (word != "$Nodes") EG_ERR_RETURN("$Nodes expected");
    */
    while (word != "$Nodes")
    {
        f >> word;
        if (f.status() != QTextStream::Ok) EG_ERR_RETURN("$Nodes expected");
    }

Regards, Hannes

... in comment under :)

Libraries for Windows available here:

• http://sourceforge.net/projects/brlcad/files/BRL-CAD%20Runtime%20Libraries/7.22.0/

Original Report:

Hi all,
im succesfully installed engrid 1.4 in ubuntu 14.04 even having some troubles.
I would like how to install new concept version in such distro. I assume that I cannot use build.bash script because it is installing version 1.4.
Which installation method do I need to use in order to install these two new branches on my ubuntu?

Thnaks!

While taking care of updating the Debian packaging for the current development line, I stumbled upon the same problem as reported here: http://www.cfd-online.com/Forums/engrid/112127-how-include-cgns-during-build-bash-compilation.html#post403402

The solution seems to be that we need some sort of compatibility put into place for the cgnswriter.cpp file, so that it can handle both CGNS 2.x and 3.x.

Original report:

As reported in this thread: http://www.cfd-online.com/Forums/engrid/138251-error-importing-openfoam-case-engrid.html - there is a new entry "inGroups" for each boundary in the file "constant/polyMesh/boundary".
For example, see the tutorial "incompressible/icoFoam/elbow".

Steps for solving this issue:

1. Write in the FAQ about the workaround: http://www.cfd-online.com/Forums/engrid/138251-error-importing-openfoam-case-engrid.html#post500283
2. Add compatibility in enGrid.

facefinder.cpp:100:7: warning: variable 'ave_num_faces' is uninitialized when used here [-Wuninitialized]

  ave_num_faces += double(m_Faces[cell].size());

Suggest init local variable ave_num_faces to zero before first use.

Original Report:

Hello again,

I ran into problems when creating boundary layers on a geometry with several patches.
After creating the surface mesh, I wanted to create the boundary layers on 14 of the 20 available patches with 25k triangles (these patches are next to each other, if this is important). After my 32 GB of RAM where not enough for this task, I went to another workstation, where the creation of the boundary layers is runnig for about 24 hours now on a 2,4 GHz core using about 140 GB of RAM and not dropping any message to the logfiles or the terminal it was startet from since the following:

creating boundary layer mesh)
preparing prismatic layer

prism quads : 0
prism triangles : 0
stray quads : 0
stray triangles : 28538

WARNING: RestrictLocalH called, creating mesh-size tree

If you need it, I could post the engrid files, so you can work on this case.

Would be nice, if you could help me. Thank you.

Regards
Andreas

I have compiled the latest commit 1a26da5 in the "master" branch on github (in order to use the polyhedral mesh export) on Ubuntu 12.04 64bit.
Everything works so far except the step "divide prismatic boundary layer".
When I start it, I get the following output

dividing boundary layer into 11 layers:

maximal increment : inf
min(y) : 0
max(y) : 0

tol=0
secs. for undefined: 3

I have created the surface mesh with gmsh and imported it using Import >> GMSH (v2 ASCII).

I tracked some variables in GuiDivideBoundaryLayer::createEdges and the problem seems to be that the field "node_meshdensity_desired" contains only zeros. It seems to me that this is only set somewhere during surface meshing, but I'm not sure.

How to include CGNS during build.bash compilation?

Original Report:

Original Report:

Sometime ago it was found out that someone has been maintaining this repository: https://build.opensuse.org/package/show?package=engrid&project=science

All that is left is to either contact them and/or simply update the installation instructions when they update the packages for enGrid 1.4.0.

Note: this issue derives from #29

Original Report:

Original Report:

Original Report:

I'm packaging enGrid for Fedora, and there were a few issues which per Packaging Guidelines ought to be reported upstream. The review is at [1]. The issues pointed out are

• There are calls to exit in the libengrid library. These should be avoided.
• There are files with unknown license:
  • ./src/dialoglineedit/dialoglineeditplugin.cpp
  • ./src/dialoglineedit/dialoglineeditplugin.h
  • ./src/dialoglineedit/test.cpp
• Some changes needed to the qmake pro files might also be of interest: https://smani.fedorapeople.org/review/engrid_build.patch

[1] https://bugzilla.redhat.com/show_bug.cgi?id=1111294

The objective is to bring back the CGNS options back into enGrid's build system.

A hack is provided here: http://www.cfd-online.com/Forums/engrid/124085-cgns-export.html#post454124
But we need a better solution.