[x3d-public] X3D After V3

[doug sanden]

Leonard,
Usage statistics from v3 might help, if anyone collects them.
-Doug
more..
I've been thinking of usage tracking for freewrl. We never know how many real users there are (downloads counts are fattened by mirror-bots), how often they run, or what nodes types in what quantities they use. No one likes being tracked. Unless its anonymous and for a good cause and they have an opt-out choice and the data collected is very generic statistics. Then a server program is needed so a program can call home with the usage statistics. Big Data storage. Then some statistical analysis. Likely there are logarithmic usage curves -like The Long Tail (book)- a few nodes are used very infrequently, and some nodes are used all the time. Transform is my guess for the the most used node. If you had the node usage sorted by frequency as a logrithmic graph, then you could 'cut off' at various support levels.
more..
The line of your thinking these days appears to me to be toward break web3d.org out of the oligopoly mode: is it fair that volunteers for web3d.org do all their work for a dozen interested parties/browser developers? But then those dozen serve 10s of thousands. Having said that, if you can make the specs simpler, so there are more developers making more tools using web3d specs at some level, that would in theory broaden the base. And stop erosion of market share by GLAM and other simpler declarative 3d formats. My own experience: it's hard to differentiate a web3d browser. They all do about the same thing. So in theory a monopoly would work. One browser that runs on all platforms and has all the nodes. It's already broader based than it absolutely needs to be already. 
Except that I keep finding weird/interesting things to do on top of the basic web3d browser.






From: x3d-public <x3d-public-bounces at web3d.org> on behalf of Leonard Daly <Leonard.Daly at realism.com>
Sent: December 14, 2015 12:09 PM
To: 'x3D WG'; X3D Graphics public mailing list
Subject: [x3d-public] X3D After V3
  
I think that the WG needs to seriously consider the usefulness of all nodes in the current version with the intent of deprecating a number of them. Many of these nodes were developed as a means to make it easier to hand-code models or scenes. I feel that  is not needed anymore because there are open-source modeling tools (e.g., Blender) that handle all of those functions. I feel it is a significant disservice to the community to continue with a very expansive set of nodes that are not used much. I would also  recommend that V4 profiles be defined that do not include those nodes so people developing V4 browsers are not required to support them. It would also be necessary to have a conversion tool that updates content using those nodes to profiles that do not use  (or support) them.,

On my (incomplete) short list:
 Extrusion 
ElevationGrid (I know that this provides "gravity" for WALK mode -- a replacement capability would need to be provided)

Text 
FontStyle 
Layout (I think this has already been mentioned) 
Nodes (or node categories) I think need particular investigation 

  Chaser/Damper/Follower (there are some valuable features here, but perhaps the capability can be improved)

Some of Triangle*Set 
2D Geometry (can SVG be used instead?) 


My goal in this is to identify the nodes that are important; those that are used not just for educational purposes; and those that look nice, but are not used. I don't want to remove valuable capabilities, but to allow developers to be able to focus on their  problem without getting distracted by unnecessary or duplicative capabilities. At this time, I am not advocating the removal of those nodes because the long-term and archival use needs to be considered.



-- 
Leonard Daly
3D Systems & Cloud Consultant
X3D Co-Chair on Sabbatical
LA ACM SIGGRAPH Chair
President, Daly Realism - Creating the Future