Login
Username:

Password:

Remember me



Lost Password?

Register now!
Main Menu
H3D.org Feeds

Forum Index


Board index » All Posts (Markus)




Re: getValueByIndex() not compatible with MFVec3f ?
#1
Webmaster
Webmaster


You nailed it. It is about the documentation.

There are two sets of class documenation for H3D. One for C++ and one for the python level.

Please refer to this one for python for 2.3 release:
http://www.h3dapi.org/uploads/api/H3D ... b/doc/html/annotated.html


Or this one if you use trunk:
http://www.h3dapi.org/uploads/api/Tru ... I/lib/doc/html/index.html

If I remember correctly there is no version for it in python. You have to do yourfield.getValue()[your_index]

Posted on: 4/20 10:23
Top


Re: invalid return value from update()
#2
Webmaster
Webmaster


From the python point of view the MFString type of a field contains a list of strings.
So you need to return

["yourstring"]

that is, including the brackets.

[EDIT] I put in a task in our internal system to make the error message better.
So some time in the future

Posted on: 4/18 10:12
Top


Re: Adding a custom force node to H3D
#3
Webmaster
Webmaster


That is indeed the way to do it.

Posted on: 3/19 9:10
Top


Re: Adding a custom force node to H3D
#4
Webmaster
Webmaster


Hello,
Just to correct a potential confusion.
The Vec3(1,1,1)*damping->getValue() is there simply because the underlying HAPI force effect expects a HAPI::Vec3 while the damping field in the Node is declared as a float.
For that reason I simply made a quick solution.
Note that this solution is most likely not what you want simply because Vec3(1,1,1) is not a normalized vector.
I would recommend changing this to something that makes more sense in your case.

Posted on: 3/16 10:14
Top


Re: Adding a custom force node to H3D
#5
Webmaster
Webmaster


Hello.

There were mostly some issues with incorrect include file names (you probably used some old name for the included file and not the actual file name ). Then it seems like there were some variables that were named differently between different files. The constructor in the NegVisSigmoid header and source files did not match up and finally damping was declared as a float in the node but a vec3 in the force effect file.
Please use some tool like tortoisesvn diff or similar so you can see the difference between the files I now upload and your original.

Attach file:


Link only for registered users

Posted on: 3/7 10:02
Top


Re: Recompiling H3D API in Windows 10 "timespec" error
#6
Webmaster
Webmaster


Really hard for me to help you with this without any kind of code.

If you do not want to share the code at this point could you at least just create a simple class that exhibits the behaviour?

Create some new files, name them something like TestClass or similar and add only one member and see if you can get the issue you are having.

Posted on: 2/21 7:39
Top


Re: Recompiling H3D API in Windows 10 "timespec" error
#7
Webmaster
Webmaster


Hello,

if you have put the header files in the same location as the other H3D header files (H3DAPI/H3D/include) then you should use the H3D/X.h variant.

Just make sure that there is not any kind of case sensitivity going on. i.e are they named X.h or x.h? or perhaps Xy.h instead of XY.h.

If you have them in a completely different location than the other H3D headers just make sure that this include directory is added to the project settings.

If you can not figure it out put them with the source file and use "" instead of <>. Although it should just be a matter of writing the correct path.

As a final warning, not related to this issue, adding files to the existing solution will work fine as long as you never touch any of the CMakeLists.txt files or regenerate the solution using CMake because if you do that you have to readd them manually.. When you are set on what the files should be called then I would recommend that you add them in cmake so they are always in the solution file.

Posted on: 2/5 9:59
Top


Re: Recompiling H3D API in Windows 10 "timespec" error
#8
Webmaster
Webmaster


Hello,

Good that my instructions helped you to get it to compile at least.

Quote:
cmake just adds all the files in the source and include folders, right?


No this is not entirely true, it depends on the authors of the CMakeLists.txt. It is actually not recommended (if I remember correclty) to use the commands that add all files in directories to the project. It is much better to list the files. So you have a couple of choices.

1. Add the files manually to the generated project.
2. Modify H3DAPI\build\H3DAPISourceFiles.txt so that your files are added in the list (should be a fairly self explanatory file)
3. Create a completely different project in the style of MedX3D or UI that you can then import into your x3d files using the ImportLibrary nodes url field (see example x3d files in UI or MedX3D).

Posted on: 1/29 9:36
Top


Re: Recompiling H3D API in Windows 10 "timespec" error
#9
Webmaster
Webmaster


Hello,

The latest release will not compile with visual studio 2015 and 2017 unless you:
- Make some changes in the code to compile (fixing timespec issue for example) it is fixed in the latest trunk.
- Compile the external libraries with the desired compiler, or get the ones from subversion repository https://h3dapi.org:8090/External/vs2017

If you are going to use visual studio 2017 and you are comfortable with svn checkouts I would recommend to just checkout latest trunk for H3DWithToolkitsWin (see information in wiki) and replace the checked out External directory with the visual studio 2017 one above.
We have started the process of making a new release but due to time limitations it will not be done any time soon.

You could also get the latest source code from here ftp://www.h3dapi.org/pub/nightlybuilds/source/
but like I said, you still need the Visual studio 2017 externals.

Posted on: 1/25 8:08
Top


Re: 3DVision rendering mode
#10
Webmaster
Webmaster


I do not know of much sadly.

I think nvidia is still selling their quadro cards.

Otherwise various HMDs might be a solution.

And finally, various extend-screen-on-two-monitors solutions using VERTICAL_SPLIT_KEEP_RATIO stereo modes with passive stereo glasses would work.

Posted on: 1/16 13:56
Top



TopTop
(1) 2 3 4 ... 248 »






(C) 2012 SenseGraphics AB    ---    Powered by XOOPS