[X3D-Public] interest in network sensor / client based server software
info at 3dnetproductions.com
Thu Jan 9 08:49:51 PST 2014
It is good to hear from you. I truly applaud and appreciate your continued efforts and desire to improve open standards with regard to multi-user systems. As a matter of fact, if you look deep into the archives, you will find that long ago I also considered the approach you are suggesting involving clients acting as servers for specific tasks. And actually, in a way, that is kind of how it works by necessity. Actions taken on local clients become drivers for the others. The difference in approach though, is that everything must go through at least one central server in order to be distributed across a large number of clients, as I assume you want a system that can be used by more than a few users. A true peer-to-peer system could certainly work for a few users, but otherwise, I'd be concerned about the memory and processing capacity of the clients. Could we imagine the Internet without servers? One of the primary concerns with 3D multiuser systems, is to keep the smallest memory foot print possible, and at the lowest hardware level possible (closest to the CPU). You truly need to get the pedal to the metal as they say. That is why it is preferable to use dedicated servers in order to be able to accommodate more users and the fastest way possible to avoid latency. Then you begin to think in terms of servers-to-servers in order to scale the entire system past the limitation of individual servers. And once you look into that, you realize that an even better approach is to use multiple servers that do not only talk to each other but also, clients talk to multiple servers. And the way to do that is.... NetworkSensor, right back to square one. For the X3Daemon system we have implemented something very close to it via scripting nodes within the current networking limitations of Web browsers and X3D viewers. More can always be done naturally. We are not stopping here either. But as far as re-inventing the wheel as you put it, I would invite you again to work with us or with any other existing system such as BS Collaborate which you are already using. Once you fully grasp how wheels work, then chances are you could make a better one.
From: Christoph Valentin [mailto:christoph.valentin at gmx.at]
Sent: Thursday, January 09, 2014 10:21 AM
To: info at 3dnetproductions.com
Cc: 'doug sanden'; x3d-public at web3d.org
Subject: Aw: RE: [X3D-Public] interest in network sensor / client based server software
No doubt, we are trying to re-invent the wheel here.
However, even the wheel has undergone an evolution from the wooden wheel, to solid tire, to pneumatic tire, and so on......
It's good to know people, who know the *original* wheel, the *real* wheel.
Because those people know also the deficiencies of the *original* wheel and can help making the *new* wheel even better.
Just my humble opinion.
All the best
P.S.: first we should clarify, if there is need for a *new* wheel, or if the *original* wheel is sufficient, otherwise nobody will finance the *new* wheel
P.P.S.: I have no answer to the question in "P.S.", therefore my project is a hobby project :-)
> Just a simple reminder. You can begin giving your
> X3D world multiuser features almost instantly. Extended
> feature set and customization are also available.
More information about the X3D-Public