minimum opengl version

[bjlittle]

Hey @marcomusy

Thanks for such an awesome package 🍻

Simple question: what is the minimum recommended OpenGL version required by vedo?

Many thanks 👍

[marcomusy]

Hi Bill, thanks!
vedo builds on top of vtk, so i guess it depends on the vtk version. For version 8.1.2 the minimum required opengl seems to be 1.1 (but might need to recompile vtk using option VTK_RENDER_BACKEND).

[bjlittle]

Thanks Marco for the super fast response 🏃💨

If I have spare capacity, I'd love to be able to contribute to vedo... is there an easy way in to the codebase for me?

Do you have some super, super simple issues that I could try to contribute towards and move forwards?

That might be a tangible and beneficial way for me to begin to understand what's going on under the awesome hood of vedo

[bjlittle]

BTW please do close this issue now. Many thanks! 👍

[marcomusy]

Thanks Bill for your interest! Obviously you're hired! :)
I'm planning to make a new release today if if manage to fix a couple of things.. so probably the best way to start could be to run the already existing examples and check if you can spot anything that seems weird/wrong or simply counter intuitive.
It would be extremely useful to have a "fresh look" from someone who is relatively new to the package!
To upgrade to the dev version:
pip install -U git+https://github.com/marcomusy/vedo.git
then

cd examples
./run_all.sh

feel free to focus only on anything specific you like though! no pressure! :)
Let me know if you need any help,
M.

PS: btw have you used vedo for some project already?

[marcomusy]

@bjlittle
I migrated this to discussions - it's probably the right place...

[bjlittle]

Good luck with the release Marco 👍

Okay, thanks for the guidance, I'll try to make time to dip my toes 😉

I realise how valuable it is to have a fresh perspective... regardless of experience (clearly I'm spinning up and looking for an excuse to become more aware of VTK at a deeper level... yup, I'm a noob to VTK, I'm not going to pretend otherwise, but we've all got to start somewhere 😉)

I'm particularly keen on using a VTK based technology that gives us interoperability between Python, either through the likes of your lovely vedo and/or pyvista, and paraview.

My focus is meteorologically based cubed-sphere quad mesh analytics and visualisation. Matplotlib is super awesome, but sadly just doesn't scale for the high resolution meshes that I (and my team) want to work with i.e., quad meshes containing up to ~6.5-7M nodes. Hence the paradigm shift away from the likes of matplotlib to a technology that can leverage the obvious benefits of OpenGL and GPU acceleration. Hence why I've landed on your doorstep, so to speak.

I started looking around the Scientific Python ecosystem for suitable packages about a year ago and pyvista and vtkplotter very quickly caught my eye... and I've just been tracking progress since then really, when I can. Now we're at the stage of wanting to build and integrate with a suitable package.

So in summary, that's roughly the context of where I'm coming from, HTH.

Since I've now got your attention... why the rebrand of vtkplotter to vedo? What was the motivation behind that? Also, how are you using it in a professional capacity? Clearly you're fully vested in vedo, given the level of your single handed contributions to the package (which is rather insane BTW 😲)... are you curating vedo full-time? Is it underpinning other scientific package/s that you're involved with?

[marcomusy]

Thanks Bill for the info.
7M vertices sounds a bit scary also for vtk probably. In vedo there's already something like that but would surely need a more clever implementation to handle that many nodes!

from vedo import *
show(TessellatedBox(), Sphere(quads=True), N=2, sharecam=False)

why the rebrand of vtkplotter to vedo? What was the motivation behind that? Also, how are you using it in a professional capacity?

i liked the name game with the 3d inverted e :) besides I didn't want it to be strictly linked to vtk in the name.
also pyvista went through to same (i think it was originally called vtkInterface if i'm not wrong)

Clearly you're fully vested in vedo, given the level of your single handed contributions to the package (which is rather insane BTW astonished)... are you curating vedo full-time?

not at all, vedo is just one side activity, my main focus is the embryology of limb development.
I've been developing it because we made , and still make, extensive use of 3d data (volumetric and polygonal surface data) so it all started by trying to put a whole load of sparse scripts into a coherent framework. Then I realized that this could be useful to others for other purposes..

[bjlittle]

This is an example of using pyvista to show a C48 cube-sphere Sea Surface Temperature timeseries interactively.

C48 simply means 6 panels x 48 cells x 48 cells = 13,824 node mesh

I'm keen to play with vedo to see what's the experience is like