[x3d-public] X3D/VRML Multiuser - WebRTC/AJAX

[Christoph Valentin]

Hi Vince, Hi all,

Thank you for this additional information.

Finally I have found the original e-mail I was referring to: it was written by Leonard D. on Mon Mar 11 07:48:58 PDT 2019.

Citing Leonard D.:

Subject: Network Sensor & Protocols

[...]I chaired the Networking Sensor WG many years ago. There was a 
fundamental disagreement as to what should be passed between clients -- 
whether it was real simple data (e.g., position, orientation) in a data 
structure understood by agreement of the clients (mostly because the one 
person wrote all code) or a more general structure that handled a 
variety of data types. Other contentious points were binary vs. ASCII 
(now UTF-8/16) or perhaps something else; and whether the communication 
required a server or was strictly between clients (browsers). The 
simplest case amounted to the Network Sensor being a wrapper around the 
operating system's socket interface.

However it is decided, any X3D Network Sensor will need to be compatible 
with WebRTC when X3D is running in the web browser if the Network Sensor 
is to run between clients or AJAX if running between client and server. 
There are many restrictions and/or conventions that need to followed or 
handled. Two of them are: 1) The port needs to be compatible with 
corporate firewalls (pretty much restricts to 80 and 443, though some 
others are available). 2) If the communication is to the server, then 
the design needs to take into account CORS (Cross-Origin Resource Sharing).

Leonard Daly[...]

Question: is it still valid?

 
Thanks,
Christoph
 

Gesendet: Freitag, 15. April 2022 um 16:33 Uhr
Von: "vmarchetti at kshell.com" <vmarchetti at kshell.com>
An: "Christoph Valentin" <christoph.valentin at gmx.at>, "X3D-Public" <x3d-public at web3d.org>
Cc: "Kevin" <klw71 at yahoo.com>
Betreff: Re: [x3d-public] X3D/VRML Multiuser - WebRTC/AJAX

Joe Williams posted this to x3d-public on Mar 10 2022.
It references an implementation rather than a standards decision by the X3D Working Group
 
Vince Marchetti
 

 

Found from 06, 
 
https://web.archive.org/web/20160328040324/http://www.hypermultimedia.com/ajax3d/index.htm#Start[https://web.archive.org/web/20160328040324/http:/www.hypermultimedia.com/ajax3d/index.htm#Start]
 
to
 
https://web.archive.org/web/20160331112316/http://hypermultimedia.com/ajax3d/T0.htm[https://web.archive.org/web/20160331112316/http:/hypermultimedia.com/ajax3d/T0.htm]
 
and the rest, 
where x3d in object could use XMLHttpRequest client/server interface when hosted by DOM browser. 
 
Since Tony did such a great exercise in the original out of MediaMachines, I “made it mine” and did a Joe version. 
 
Amazing, looks like it still all there and almost works except no Hello because no Flux Player™, and looks like canvas has solid support over object or iframe? Is the XMLHR a thing? 
Old Fun,
Joe _______________________________________________
x3d-public mailing list
x3d-public at web3d.org[mailto:x3d-public at web3d.org]
http://web3d.org/mailman/listinfo/x3d-public_web3d.org
 
On Apr 15, 2022, at 3:20 AM, Christoph Valentin <christoph.valentin at gmx.at[mailto:christoph.valentin at gmx.at]> wrote: 

Dear all,
 
I am sorry to bother you again with this topic.
 
I remember someone (was it Joe?) wrote at the mailing list about a decision of the consortium:
If a Web3D scene is hosted by a browser, then the communictation for multiuser features shall be done via WebRTC and AJAX.
 
Unfortunately, I cannot find the mail any more in the archives. Could someone point me to that mail? Or could you confirm that remembrance?
 
Thanks,
Christoph

_______________________________________________
x3d-public mailing list
x3d-public at web3d.org[mailto:x3d-public at web3d.org]
http://web3d.org/mailman/listinfo/x3d-public_web3d.org