[X3D-Public] The Network Sensor is not enough

Christoph Valentin christoph.valentin at gmx.at
Sat Feb 12 12:03:06 PST 2011

Hello Don

Just some additional remarks regarding my experiences with existing 
implementation of network sensor.

1) The network sensor fires "initialized" = "true" after successful 
initialization. If the network sensor holds a state (and not only events), 
then the first update of the local state will be done, BEFORE the 
successful initialization will be indicated with "initialized" = "true" 
(unless there is no state present at the server at all, in which case the 
local state will not be updated).

This behaviour is very useful. It should be required by the standard.

2) My "SRR Framework" depends heavily on a "sessionId", which is delivered 
by the browser and identifies (non-negative SFInt32) the session (I call it 
"scene instance" instead of "session").

3) I define a "controller role" (one scene instance is responsible for 
state updates). The automatic selection of the scene instance for the 
"controller role" would be easier, if the browser/server provided some 
"network semaphore service" or if the browser/server selected the 
"controller role" on his own.

4) I depend on the browser indicating remote sessions, which left the 
multiplayer session. Alternatively this dependency could maybe avoided, if 
it were guaranteed by the network sensor to send "a last user-defined 
state/event in case of shutdown" to the server.

OK, all this is not very helpful in the current situation, but better to 
tell it than to forget it :)

>> if there is sufficient interest to review past work and discuss
>> potential future work, it would be great to schedule a teleconference.
Unfortunately, this is not up to me :-( , but if someone will indicate 
interest to me, I will get back to you

Regarding X3D-Earth:
Unfortunately, I did not yet find the time to get deeper into this very 
interesting topic, I'm just thinking about one James Bond's theme "The 
world is not enough", and I cannot forget this song :-)

Kind regards

-------- Original-Nachricht --------Datum: Mon, 07 Feb 2011 09:46:57 -0800

> Von: Don Brutzman <brutzman at nps.edu>
> An: x3d-public at web3d.org, Christoph Valentin <christoph.valentin at gmx.at>
> CC: Michael Aratow <maratow at noegenesis.com>, Chris Thorne 
> <dragonmagi at gmail.com>
> Betreff: Re: [X3D-Public] The Network Sensor is not enough
>         thanks for your efforts Christoph.  yes i agree that there are 
> some
> excellent ideas.
> there is past work preserved at
> http://www.web3d.org/x3d/content/examples/Basic/Networking/
> http://www.web3d.org/x3d/content/examples/Basic/Networking/NetworkSensorConnectionPrototypes.x3d
> http://www.web3d.org/x3d/content/examples/Basic/Networking/NetworkSensorConnectionNodes.html
> this work got fairly far but had to stop.  quoting:
> "Work on these nodes has been discontinued, at least temporarily. Because 
> the NetworkSensor node allows definition of new fields by name, a proper 
> implementation requires native X3D-player support and a full 
> Prototype-based implementation is not possible. No X3D browser companies 
> have yet committed to produce such an implementation."
> if there is sufficient interest to review past work and discuss
> potential future work, it would be great to schedule a teleconference.
> I think that there is a lot of great work on networking that can
> be capitalized on.  Deep Matrix, RayGun, BitManagement multi-user
> server, proposal for Anchor/Inline url refresh, etc.
> meanwhile NPS continues to offer our open-source DIS library to
> integrate in other browsers, available in C++ C# Objective-C and Java.
> already integrated in Xj3D.  numerous additional information about DIS
> available at
> http://x3dgraphics.com/slidesets/X3dForAdvancedModeling/DistributedInteractiveSimulation.pdf
> plus proposed additions for X3D v3.3 on member wiki, excerpt follows
> http://www.web3d.org/membership/login/memberwiki/index.php/X3D_v3.3_Specification_Changes#Distributed_Interactive_Simulation_.28DIS.29_Component
> > Distributed Interactive Simulation (DIS) Component
> >
> > Latitude and longitude need to be added to the position-dependent nodes 
> in the X3D DIS Component in a manner consistent with the X3D Geospatial 
> Component.
> >
> > The packets defined in the IEEE DIS Specification do support 
> high-resolution geospatial coordinates.
> >
> > Thus the following fields need to be added to EspduTransform, 
> ReceiverPdu, SignalPdu and TransmitterPdu:
> >
> >       SFVec3d  [in,out] geoCoords      0 0 0       (-∞,∞)
> >       MFString []       geoSystem      ["GD","WE"] [see 25.2.3]
> >
> > Existing translation fields in the DIS nodes are still be retained and 
> relevant as before. Thus translation values simply reposition grouped 
> children nodes within the geospatial coordinate frame, rather than a 
> local coordinate frame.
> url refresh is also on the member wiki:
> http://www.web3d.org/membership/login/memberwiki/index.php/X3D_v3.3_Specification_Changes#X3DUrlObject_abstract_interface:_refresh_for_scene_reload
> > X3DUrlObject abstract interface: refresh for scene reload
> >
> > A major factor for improving the interactivity of X3D scenes might be 
> to allow the deliberate reloading or refreshing of external content. This 
> would facilitate a great deal of client-server functionality without 
> requiring some kind of networking protocol.
> >
> > Similar functionality is provided by the http-refresh capability of 
> HTML, for example <meta http-equiv="refresh" content="5" />. However 
> there are some related problems with user accessibility that might be 
> relevant.
> >
> >     9.3.2 X3DUrlObject
> >
> >
> >     X3DUrlObject {
> >       MFString [in,out] url [] [URI]
> >       SFBool   [in,out] load TRUE
> >       SFTime   [in,out] refresh -1
> >     }
> >     The refresh field specifies a refresh interval in seconds.  At each 
> refresh interval,
> >     the current url resource gets shutdown (if appropriate) and then 
> reloaded.  A
> >     refresh value of 0 or less means that no reload occurs.  If the 
> prior reload is incomplete,
> >     it is interrupted and superceded by the new refresh request.  
> Cached content is no longer
> >     used, unless none of the refreshed content url values remain 
> retrievable.  Time-dependent
> >     media (such as AudioClip or MovieTexture) are handled as if they 
> are again initially reloaded.
> thoughts?
> the more we do, the more we will do.  looking forward to future progress.
> all the best, Don
> -- 
> Don Brutzman  Naval Postgraduate School, Code USW/Br      
> brutzman at nps.edu
> Watkins 270   MOVES Institute, Monterey CA 93943-5000 USA  
> +1.831.656.2149
> X3D, virtual worlds, underwater robots     
> http://faculty.nps.edu/brutzman
GMX DSL Doppel-Flat ab 19,99 Euro/mtl.! Jetzt mit 
gratis Handy-Flat! http://portal.gmx.net/de/go/dsl
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://web3d.org/pipermail/x3d-public_web3d.org/attachments/20110212/13f1544d/attachment.html>

More information about the X3D-Public mailing list