[x3d-public] X3D Text/FontStyle nodes for glTF and OpenUSD

John Carlson yottzumm at gmail.com
Thu Nov 28 23:34:38 PST 2024


This should be read a bit like The Onion.  I believe I posted on this
subject a while back.

Are there X3D Text/FontStyle nodes for glTF and OpenUSD?  Surely Apple has
something for their VisionPro?

If not, how the heck is Google or any AI company going to scrape the
Metaverse for data?  Video?  Metadata?  Cars with cameras and LiDAR in the
Metaverse?

Maybe Spatial Intelligence will swoop in and save the day?  Or just plain
OCR?

Maybe NVIDIA/AMD didn’t think of this when doing 3D acceleration and
omitting symbols?  Is this why Khronos is pushing an OpenVG standard?  Oh,
OpenVG is supported on my graphics card.  Cool!  How do I use it???  Oh,

#include <VG/openvg.h>

Cool!

Does any X3D browser use this? For Layers or Layout?

“
OpenVG 1.1 Lite Provisional was released in May 2020, and relaxed the
accuracy requirements of stroke path rasterization and normalizes the
stencil/blending modes of OpenVG 1.1 so the API can be fully accelerated by
any OpenGL® ES 2.0-compatible 3D GPU for the first time, while maintaining
high rendering quality.
“

So how do I do the equivalent of OpenVG with OpenGL?  Or should I even
worry about it?

Here’s some historical information.  Apparently OpenVG was abandoned in
2011????
https://stackoverflow.com/questions/6287650/rendering-svg-with-opengl-and-opengl-es

Anyone know the state of textual information for Vulkan or WGSL?  I only
know OpenGL fragment and vertex shaders a bit, and I just hear about
texture atlases for text.  Maybe Metal and D3D are better?

https://youtu.be/mGQRV32Nk3s?si=pdL0qLZ43t9KZpso

(Julia says Windows might have a solution—she wanted something
cross-platform)

Travis has 3 videos, starting here:
https://youtu.be/zvGIp-S2mxA?si=yJy9W4x076vKK1v_

So it looks like some heroic programmers are figuring out how to do this,
maybe cross-platform in Vulkan.

Here’s the secret: text made from dots = raster, text made from symbols =
character (typewriter),  text made from curves = vector.

Apparently, we’ve forgotten how to make typewriters.  Remember the printing
press?

Someone get a big stamp and put “DUMMY” on my forehead.

Don’t even get me started about curved 3D text.

Why can’t NVIDIA spend some of their money on cross-platform 3D text APIs
and make all of our lives easier? Are they afraid of competing with Apple
and Microsoft?

Surely Warnock thought of this before PostScript for his 3D language? And
we still don’t have good cross-platform APIs for 3D text?  It’s been, like
60 years!

Yes, I’ve seen the X3DOM demos, how about text along a 3D curve?

Anyone?  An API? A start with new X3D nodes?  I know we can’t standardize
without implementation.  How about starting with doing SVG extrusions?

Maybe take inspiration from Java3D?
https://docs.oracle.com/cd/E17802_01/j2se/javase/technologies/desktop/java3d/forDevelopers/J3D_1_3_API/j3dapi/javax/media/j3d/Text3D.html

Or, more modern:

https://github.com/Birdasaur/FXyz-DeepSpaceBranch/blob/master/FXyz-Core/src/main/java/org/fxyz3d/shapes/primitives/Text3DMesh.java

Here’s the FXyz-Core LICENSE:
https://github.com/Birdasaur/FXyz-DeepSpaceBranch/blob/master/FXyz-Core/LICENSE

Maybe I’ll add these to my 3D asset manager/contacts app!


John
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://web3d.org/pipermail/x3d-public_web3d.org/attachments/20241129/317da7d5/attachment.html>


More information about the x3d-public mailing list