This is Interesting: Free Magazines for Graphics designers and webmasters
Home > Archive > VRML > December 2004 > seeking open source VR community software
You are viewing an archived Text-only version of the thread.
To view this thread in it's original format and/or if you want to reply to
this thread please [click here]
| Author |
seeking open source VR community software
|
|
|
| I am looking for an Open Source VR community meeting space software,
not unlike Blaxxun's Cybertown. I am considering a range of
technologies, including vrml/x3d, and Open Source game engines. I
would also love to hear of any 'out of the box' solutions, ie software
designed to this purpose, as I have found only slim pickings so far (
eg http://www.vrspace.org/ seems to be a mostly-windows client ).
= The Requirements =
Critical features:
* Client-server networking architecture - persistent world that can
be visited by multiple people at once
* Any software components we build on must be open source, with non
restrictive licensing, eg LGPL
* Client must be cross platform ( Linux + Windows ), server must run
on Linux
* Avatars representing users
* Does *not* have to be true 3D, eg 'The Sims' style isometric
semi-3D would be fine...
Ideal features:
* Ease of use. The end users will not be gamers, so point and click
navigation, for example, is preferable to FPS-style navigation.
* Built-in chat
* Server
* Runs on debian ( read our current web host )
* Low-profile process, so it doesn't get squashed
* Writes world states to db in case of crash etc
* CGI access ( I'm dreaming ), as sockets are not guarenteed on
our shared host
* HTML rendering, eg forum postings can appear within the 3D world
= Some Possible Solutions, Pros and Cons =
Open Source Game Engines
+ High performance, great graphics
+ Totally customizable
+ Could be a "more final" solution - ie give us what we really want -
good 3D/VR now
+ Some good ones ( Crystal Space and OGRE ) are just reaching maturity
/ 1.0 status
+ Cross platform clients + servers available
+ Client-server networking available
- Could lock us into one engine / format - may have to redo everything
if we change engines
- Greater complexity
VRML/X3D
+ Web3D standard format
+ Probably simpler implementation than game engine
+ Good clients freely available ( Windows )
- No networking/multiplayer built in
- Seems ( to me ) to be not really designed for "game logic",
interactivity kind of tacked on to a modelling/scene description
language (?)
- Might not be as extensible for larger projects as a game engine
- Game logic is in Java or JavaScript ( ie not C++ )
- Graphics less great
- Clients for Linux currently limited
Any other ideas, or feedback on these pros and cons, much appreciated
: )
Thanks,
Harlan
| |
| R. Russell Kinter 2004-12-04, 7:14 am |
| Hi Shane, er I mean sam
Thanks!
I figured, 90's retro is kinda in fashion lately.
Aside from its "age", "race" works great!
tc
Russ
"sam" <sam3dus@yahoo.com> wrote in message news:<5c8sd.413$0r.134@newsread1.news.pas.earthlink.net>...[color=darkred]
> Works fine...Nice to see it all working, Russell.
>
> But funny see that old content, like my race car wrl...
> those were the days... lol.
>
>
>
> "R. Russell Kinter" <Pyth7@comcast.net> wrote in message
> news:1a90d129.0412031401.1880832e@posting.google.com...
| |
| R. Russell Kinter 2004-12-04, 12:15 pm |
| Yes I am going what your experience says over what:
http://www.geometrek.com/developers/faq.html says.
However, I just noticed (probably because I am windows oriented)
on this faq it says: Irix 6.5 Netscape 4.07 Cosmoplayer 2.1
So for Irix, Netscape seems to be limited to one version and
Irix "maybe" limited to 6.5 as well.
tc
Russ
Joerg Scheurich aka MUFTI <rusmufti@helpdesk.rus.uni-stuttgart.de> wrote in message news:<cokcqr$ru6$1@infosun2.rus.uni-stuttgart.de>...
>
> I tested with Netscape 4 and Cosmoplayer 2.1 and it did
> not work. Maybe there is more luck with Cosmoplayer 1.X ?
>
> so long
> MUFTI
| |
| R. Russell Kinter 2004-12-05, 12:16 pm |
| Braden wrote:
> AFAICT, DeepMatrix is most certainly not available under a "GPL type"
> license. While it claims to be "opensource", its terms are not
> compatible with the Open Source Definition
> <http://opensource.org/docs/definition> (which, whether some folks like
> it or not, is the conventionally accepted definition of the term).
Yes this is true.
Here is the pre-2001 reality:
My understanding of the license
is it's free for non commercial use ($500 otherwise)
and any modifications you do to it "they" i.e. the original
members of Geometrek (where ever they maybe hiding)
get to adopt into the package, if they so desire.
and you must include the license with whatever changes
you make.
Here is the post-2001 reality:
Geometrek as such is a ghost site ('cept for the forum
run by myself and John). The original
members are almost impossible to get a hold of.
They won't answer any e-mails. You would basically
have to hunt them down to pay the fee if you wanted to
sleep better and be legally free and clear for commercial use.
On the other hand, why not let them just come to you,
if it really matters to them anymore.
Currently Geometrek is hosted strictly by the
courtesy of http://www.lucidspace.com,
Because the original server blew. If lucid
space decided not to host, it's safe to say
Geometrek would be no more.
I have taken over developing the applet on my own
and have improved it both stability and feature wise
and stick the bloody license into the archive when
ever I can remember.
The latest and greatest version can be found here:
http://www.sevenemeraldtablets.net/...merald_Tablets/
CortonatestEntryEAI2.html
tc
cyber squatter and backwoods country lawyer
Russ Kinter, ESQ.
Braden McDaniel <braden@endoframe.com> wrote in message news:<1101696153.3278.10.camel@localhost.localdomain>...
> On Sun, 2004-11-28 at 21:01 -0500, Muaddib wrote:
>
>
> AFAICT, DeepMatrix is most certainly not available under a "GPL type"
> license. While it claims to be "opensource", its terms are not
> compatible with the Open Source Definition
> <http://opensource.org/docs/definition> (which, whether some folks like
> it or not, is the conventionally accepted definition of the term).
| |
| R. Russell Kinter 2004-12-05, 11:17 pm |
| Hi Harlan,
I'd like to add something about Deep Matrix's "new" features to
Rob's post.
The Geometrek demo Deep Matrix applet is not the latest version.
The latest can be tried out here:
http://www.sevenemeraldtablets.net/...merald_Tablets/
CortonatestEntryEAI2.html
We recently successfully tested this particular applet
with 25 simultaneous users.
The twenty five breaking down as:
20 individual constantly moving
avatars belonging to a bot.
(movement being sent accross the network,
not local animation)
and five living users,
3 using Cortona and 2 Contact.
The applet can use either I.E. or
the Netscape 4* series as the html browser.
It will also work quite successfully with
Cosmoplayer and Netscape 4* combination.
Cosmoplayer and I.E. combination is not
supported any longer.
It also has a capability for I.E.
and WorldView 2.1
The Irix OS and Cosmoplayer ability can
no longer be claimed to work.
(Isn't that right MUFTI?)
The applet/server has persistent storage as
long as the world is populated.
Some other features are:
Java server can run on linux.
Simple chat and/or moving bot capability.
Beamto and whisper.
Text strings to and from the vrml world and chat.
Personal chat text strings to a user's avatar.
Persistent while populated shared events
Avatar gestures (feature going back to V-net)
Will be glad to run the stress test bot demo for you,
or anyone else for just fun'n glory.
thanx
Russ Kinter
Muaddib <rob_nospam@vrspace.org> wrote in message news:<41AA82E7.20104@vrspace.org>...[color=darkred]
> HW wrote:
>
>
> Hey there HW,
>
> I can speak to the possibilities for VRML (take note, I am a
> designer for VRSpace ;) ). There are essentially two actively gpl type
> vrml servers out there: Deep Matrix (http://www.geometrek.com/) and
> VRSpace. Both work off the same tech: EAI which uses the MS JVM for IE,
> or Netscape java for Netscape. Neither has a usable linux client.
> First what do we mean by a linux client, i.e. is it a browser plugin or
> a standalone app. Browser plugin: no such animal exists. Standalone
> app: consider freewrl, as far as I understand, their EAI is in the
> works. But whatever your linux choice is, it will have to use the EAI.
> There is also abnet which uses sun java, but I never really know if
> the source is open. Check out http://www.vrmlworld.net/
>
> As far as non-vrml solutions, I have recently been taking a look at
> VOS. It actually seems to be very similar architectually to VRS, but
> uses Crystal Space as the client instead.
>
> And since I am an expert in vrs, I will tell you what it has got.
>
>
> yes
>
> GPL
>
> Only Contact for Windows and IE. Won't work with Cortona, their EAI is
> too buggy for our purposes. Deep matrix works with both.
>
> Yes
>
> VRML
>
> Yes
>
> Yes
>
> yes
>
> ? Sure it can be easily managed.
>
> yes
>
> ?
>
> Saw that VOS has this. Pretty cool
>
> VRML eh, you would be suprised. Check out Quake3 conversions, they look
> pretty good.
> http://www.bitmanagement.de/develop.../quake3/q3.html
>
> In as much as the EAI allows. Servers is completely modular.
>
> ?
>
>
> Server runs anywhere java does.
>
> yes
>
> vrml
>
> Networking is built in.
>
> True.
>
> Very modular actually.
>
> Doen't see why Game Logic is a speed concern. Rendering sure, but
> logic? Java is just fine.
>
> Indeed
>
> Good Luck,
> Rob
>
| |
| Braden McDaniel 2004-12-05, 11:17 pm |
| On Tue, 2004-11-30 at 13:43 -0800, R. Russell Kinter wrote:
> Braden wrote:
> Yes this is true.
> Here is the pre-2001 reality:
> My understanding of the license
> is it's free for non commercial use ($500 otherwise)
> and any modifications you do to it "they" i.e. the original
> members of Geometrek (where ever they maybe hiding)
> get to adopt into the package, if they so desire.
> and you must include the license with whatever changes
> you make.
> Here is the post-2001 reality:
> Geometrek as such is a ghost site ('cept for the forum
> run by myself and John). The original
> members are almost impossible to get a hold of.
> They won't answer any e-mails. You would basically
> have to hunt them down to pay the fee if you wanted to
> sleep better and be legally free and clear for commercial use.
> On the other hand, why not let them just come to you,
> if it really matters to them anymore.
Because if I were to use the code in a commercial setting and they did
come to me, I'd be entirely at their mercy legally. If the owners of the
code really have no interest in (or hope for) profiting from licensing
it, it's a shame they won't take the hour or so it would take to make it
available under an OSI-certified license. The potential cost of using
code available under such ambiguous terms is just too great--no matter
how remote you perceive the chances of having to pay the piper.
--
Braden McDaniel e-mail: <braden@endoframe.com>
<http://endoframe.com> Jabber: <braden@jabber.org>
| |
|
| Hi Russ,
Thanks, this is very helpful. Deep Matrix actually looks the most
attractive so far for our needs. Is the source available for the
Seven Emerald Tablets version? Does this use sockets to communicate
between the java applet and the java server?
> Avatar gestures (feature going back to V-net)
Um, so is this a descendant of V-net? How easy is it to implemnt new
gestures, BTW?
> Will be glad to run the stress test bot demo for you,
> or anyone else for just fun'n glory.
That would be great : ) If we end up pursuing this line, I'd love to
take you up on your offer. What timezone are you in? I am GMT+10ish
(Sydney)
Thanks to everyone for your responses, I'm (slowly) getting up to
speed in the VRML client/server jungle...
Regards,
Harlan
| |
| Joerg Scheurich aka MUFTI 2004-12-06, 4:30 am |
| > - No networking/multiplayer built in
VRML has a networking built in ("Inline", "EXTERN PROTO") but no multiplayer
built in.
> - Graphics less great
The view of graphics (especially for demos) mostly depends from the author 8-)
Some effects (good supported by the current hardware (e.g. real time shadows
environment mapping or shaders)) are missing in the VRML standard (and in the
current X3D Draft), but hopefully the dynamic structure of X3D
http://www.web3d.org/x3d/specificat...onentprofilereg
will solve this problems in the future.
Especially the fact, that the "big" VRML browsers already have extensions
for such effects may help.
> - Clients for Linux currently limited
But the situation gets better each day.
Did you take a look at FreeWRL, Xj3D and octaga ?
so long
MUFTI
--
Klicken Sie diese Akte an, dann klappen sie OK zusammen.
(aus einem Software-Handbuch)
| |
| Joerg Scheurich aka MUFTI 2004-12-06, 4:30 am |
| > The Irix OS and Cosmoplayer ability can
> no longer be claimed to work.
> (Isn't that right MUFTI?)
I tested with Netscape 4 and Cosmoplayer 2.1 and it did
not work. Maybe there is more luck with Cosmoplayer 1.X ?
so long
MUFTI
--
(Press Retry to debug the application - JIT must be enabled)
Abbrechen Wiederholen Ignorieren
aus einer M$Windows "assert" Fehlermeldung
| |
| R. Russell Kinter 2004-12-06, 4:30 am |
| google@thegoldensun.com (HW) wrote in message news:<62e69c0c.0412021957.25ae908c@posting.google.com>...
> Hi Russ,
>
> Thanks, this is very helpful. Deep Matrix actually looks the most
> attractive so far for our needs.
Thanks for your interest Harlan!
> Is the source available for the
> Seven Emerald Tablets version?
Yes, in that I will gladly give it out to anyone
interested, but not right off hand because like a
lot of amateur developers I tend to get caught up
in coding end of things and let the documentation
go to pot.
I will have a basic SDK set up by the end of this
weekend with the source code and my annotated changes
to the original applet, how to run the bots, ect.
Java docs for it will have to come later.
In the mean time you can get the my compiled
version of the applet here:
http://mywebpages.comcast.net/Pyth7/Main/mudeck.eai.zip
> Does this use sockets to communicate
> between the java applet and the java server?
>
Yes
> Um, so is this a descendant of V-net? How easy is it to implemnt new
> gestures, BTW?
Acknowldgement to Steve White and Jeff Solstein (V-Net creators)
are in several places in Deep Matrix's source code. The Deep Matrix
gesture avs are indentical to V-Net's. I think it's fair to call it
a descendent.
Implementation of gestures requires handcoding.
My version revises the avatars somewhat.
>
> That would be great : ) If we end up pursuing this line, I'd love to
> take you up on your offer. What timezone are you in? I am GMT+10ish
> (Sydney)
Am in east coast U.S. GMT-5. Am a night owl with fairly flexible
hours.
We also have a forum set up at http://www.geometrek.com/forums/
Though as you can see, I do alot of talking to myself there :)
>
> Thanks to everyone for your responses, I'm (slowly) getting up to
> speed in the VRML client/server jungle...
>
Ok to be fair (and confuse you more! :), Here is one more VRML EAI
based client/server that a lot of people are not aware of:
http://MindGate.SourceForge.net
I haven't tried it out myself, but it supposedly
is based on Rick Kimball's V-Net+ (a development of V-Net too)
,so it will use blaxxun avs. It claims to work for Cortona now,
which V-Net+ did not do, though I wonder if it has dealt with
Cortona's lackadaisial loading problems like DM has.
Thanx again for your interest!
tc
Russ
|
|
|
| | Copyright 2003 - 2008 forum4designers.com Software forum Computer Hardware reviews |
|