Indoor scene management

Hi



What is the current status / future plans for interior scene management in JME?



I can find some old posts about BSP and Portals but is anyone actively working on it?



/Giddy

BSP support was on the agenda for this past release, but got pushed.  Since Mojo and Renanse are getting their orders for what they can work on from "higher up", we may be waiting a bit longer on this.  However, I can imagine "higher up" is going to want something along the lines of BSP support in the near future, so hopefully not too much of a wait…and who knows, they may have already done it and just waiting for approval to give it back. :wink:



darkfrog

Actually, it wasn’t Mojo and renese working on Portals but Batman.ac

http://www.jmonkeyengine.com/jmeforum/index.php?topic=1168.0





It looks like the discussion fades into BSP talk by the end. 

http://www.jmonkeyengine.com/jmeforum/index.php?topic=2216.msg24796#msg24796



darkfrog

The most recent BSP related work was done by bakchuda. Still nothing complete though.

My ears were ringing, and noticed somebody mentioned my name…  :smiley:



I kinda put the BSP work on hold due to Real-life creep, but also wanted to wait until a decision was made on what technology/algorithm should be pursued for interior space partitioning.  And since I have seen some of the demos and ideas posted in the forums have a lot of exterior spaces, I think a discussion on how to handle large outdoor scenes is warranted as well.



Out of curiosity, if Mojo and Renanse are busy in their jobs, how does the timeline and roadmap for 1.0 look at the moment?

I smell a poll or two.

Is there any route to 1.0 defined?  I think we're enroute to 0.11 at the moment.  I think most open-source projects are afraid of ever announcing 1.0 of their APIs. :-p



Then again, I'm defying the norm by working on JGN 2.0 right now. :-p



darkfrog

sorry, I was still thinking about before the .10 days, when .9 was the latest, and the rational progression would have been to 1.0.  :slight_smile:



And yes, the 1.0 fear is there as well.





I'll rephrase and just say is there ANY roadmap or .plan or timeline in place?  :?

Bakchuda said:

I'll rephrase and just say is there ANY roadmap or .plan or timeline in place?

And if you want to read more: try this thread :slight_smile:

To emphasize that… We're more at a point where features are added as they are needed (which has always really been the case, but in the earlier days more features were needed. :slight_smile: )



That said, features/improvements are still being worked on and added as we speak.  If there is a particular feature you are just dying for (such as the topic of this thread) you just need to make the appropriate noise.



(yes, I realize that last thread will spawn a shift in conversation starting with a certain frog member.)

well, i'll speak for the shadowy amphibian and say RIBIT!  LOUD NOISES!!  :smiley:

OK.  A day has gone by and more noise needs to be made on this topic.  Bakchuda, what approaches are you tossing into the hat for voting/discussion/work on indoor space partitioning?  BSP subtracted space?  Portals and Anti-Portals are always cool.  There seems to be a lot of talk already taking place on the wide-open-spaces exteriors but sooner or later you'll want a building slap dab in the middle to explore.  Personal opinion, brace yourself, interior space is something that has to be addressed for us to be taken seriously.  I know, I know, what a monumental statement.  I'm just glad to hear it getting kicked around.

Personal opinion, brace yourself, interior space is something that has to be addressed for us to be taken seriously.


I agree. Indoor scene management is about the only *major* feature left out and that is surprising considering the otherwise full feature set. jME is the only "well developed" *free* engine that I know of that lacks it. Ogre and Irrlicht both have indoor scene support through BSP and/or Collada and I would consider them to be on par with jME.

Personally I would love to see Collada support as it wouild be perfect for a Blender / jME tool chain.

I feel a bit hypocritical in posting this as I haven't contributed anything to jME and I know the devs are hard at work on other aspects...

What…me?  I don't make noise…I'm the "happy-go-lucky-frog" with nothing in my mind but my next fly…mmmm…fly…sorry, what was I saying again?



Oh yes, I'm not currently in need of BSP support, but I do think it's going to be something necessary to jME and it's something that a lot of people around the community seem to want.  I'd say if there's any specific feature that the community is really clammoring for right now it is BSP support.



darkfrog

I haven't read much about Collada before, can anybody post some good links for info?



I've read about portals and BSP, and I would definately say portals are a better way to go, because of the freedom it allows being dynamic (as opposed to pre-compiled).  I can try to wrap up the Quake 3 BSP renderer I had posted a while ago, which did a good job of implementing that specific BSP, while we are discussing other options.

Btw, for anyone curious, this is a great article where I learned most of my base portal knowledge:



http://www.fairyengine.com/articles/portals.htm

@Bakchuda - this is the best Collada tech and spec site I've found: http://www.khronos.org/collada/



The coolest thing about Collada IMO is that it allows the export of scene assets in an editable, parsable XML format. And if I'm not mistaken, Collada-based scene descriptions could be used along with portals too.



BTW, nice link you provided. I read the material and understand portals a bit better than before. Thx.  :slight_smile: