Virtual Mat - bit like MapTool, but with 3d

[revinor]

Some time ago, I have decided to try writing online rpg table program, similar to MapTool. I actually did consider starting from Maptool, as I used to work on it a bit in the past, but it is very much 2d oriented, while I wanted to try something 3d (actually, it is more like 2.5d)

After that, I got distracted for few months, but now I'm back to the business and I would like to share the current state of development with you.

You can take a look at the small teaser video at http://youtu.be/6nwAJY3jIEc?hd=1

Trick which allowed me to bootstrap development is to reuse NWN (Neverwinter Nights) models and animations. Both me and my players have original version of NWN, so we can use the models for our play. There is also quite a lot of free models created by fans, so even without NWN, there should be enough content to get something working (said that, full version of nwn with all extensions costs few dollars in bargain bins...).

What is already implemented:

* a lot of framework code which is the foundation for later extensions
* client/server communication allowing to broadcast events between client machines
* model display and animation, partial implementation of particle system (visible in fire elemental for example)
* walk/run around action
* dice rolling using physics, properly displayed on all the clients (physics libraries are NOT deterministic...)
* basic big-jpeg-as-map display, with per-square height levels (editable online)
* shadows, bloom, glow, metallic reflections etc (jme3 engine provides this almost for free)
* basic footstep sounds synchronized to animations



To be implemented:
a LOT of things. It doesn't really make sense to list all the details, just few ideas I'm playing with

* PC model composition from equipment pieces (NWN way)
* movement acknowledgment mode - players are just drawing the path and sending to do GM for validation and execution
* targetting templates - shapes for spells/etc which will automatically highlight friends/foes in the range
* placeables - non-animate objects to be put on the map to make it bit more 3d
* macros - possibility to script certain sequence of actions
* MapTool-like walls/FOV? Only monsters in FOV would be visible to players
* possibility to compose maps out of smaller pieces connecting together
* proper walk path calculation, taking terrain into account (this means also possibility to define terrain types)
* per-model states/effects, to indicate conditions/spells active
* fight/charge/die animations connected to proper actions
* etc, etc, etc


What is NOT in scope:

* full implementation of any particular ruleset. Roll 1d20+modifier, compare versus static number, perform another roll if greater-or-equal -> possible, as script action, but not built-in assumption
* public servers/game finders/etc
* real-time mode
* any kind of AI/versus computer play



Source code is available at http://www.virtual-mat.net, but just for curious people - it requires modified version of jme3 and some resource files (dice models, extracts from nwn, sounds, etc) which are not public. I hope to get to the point where the program will be runnable by other people in few weeks from now.

[aliasmask]

That's pretty sick. Until you get a huge library of 3d figures I could imagine just using a 2d image to stand up and always face the camera, like a card or whatever. That's kind of the dungeon building benefit of 2d. Perhaps you can have virtual pogs with images on them.

There's another guy here working on headless virtual table, agent6. You should check out his stuff and maybe collaborate.

[Full Bleed]

That looks like a lot of fun! I hope you continue to develop it.

But I think I'd have a hard time playing without my MT macro support these days. Too bad you can't rip out the guts of MT's campaign properties, token properties, macro system, and init window.

[revinor]

Until you get a huge library of 3d figures [...]

Idea is to reuse NWN resources. Games is very cheap this days (I have seen 'diamond' edition, with all extensions for 5$) and just by itself it contains like 300 animated monster models + almost infinite combination of PC equipment/models. On top of that, you have 3rd party stuff developed for NWN, another few hundred animated models, plus over a thousand static objects.

There is a lot to be done to have it working, but fortunately 3d resources are kind of covered.

[wolph42]

that is pretty slick indeed. i'm running a sci-fi game however (w40k) so it aint realy for me. Next to that the most important part for me to work with MT is the FoW and VBL without that its 'useless' in my games. Finally but this completely depends on the implementation, creating vbl in this seems like a lot of work (or if you really have a good lib AND a good code implement, a lot less work).

In all it really looks awesome!

[revinor]

Some kind of macro language is going to be there for sure, together with per-model properties etc. I wonder if it would be possibly to write maptool compatibility layer for macro system... will look into that when I reach that stage.

FoW/VBL is more tricky. Map is created based on tiles (you can see the grooves in the video), so it is perfectly possible to do some kind of per-tile FOW. It would be considerable amount of work, but doable (with models being either fully visible or invisible). Actually having nice, Maptool-like light shining through the small holes in exact manner... I don't think it is possible without major effort.
I'm still not sure if I want 3d walls to be there. This opens a lot of problems modelling/resource wise and probably mean that point-light shadows cannot be used to emulate the missing part of the equation.

As far as sci-fi mode is concerned... unless we could find a good supply of sci-fi 3d models (animations are not critical, even static models sliding around would be pretty cool, it is battlemat emulation, not realtime game) it is not really possible.

[Full Bleed]

I think what you need is a way for the program to automatically turn a 2d image into a virtual 3d pog... or even just have the image stand "upright"/perpendicular to the battlemap. That way 2d assets for any genre of gaming could still be used in the 3d environment.

I'd stay away from 3d walls, too (at least for your version 1.) That would just make map building a lot more complicated.

As for FoW and Vision Blocking... you're going to be hard-pressed to find people on the MT forums that wouldn't want that functionality. I know those features are what brought me to MT in the first place.

[Azhrei]

FoW/VBL is more tricky.

Heh, that's an understatement.

[Raoden]

Impressive. So why hasn't WotC hired you to make their long-overdue 3d 4E VTT?

[revinor]

FoW/VBL is more tricky.

Heh, that's an understatement.

I did spend a lot of time with Maptool FoW implementation, trying to fix shine-through corners and some other small issues and back then I got a full understanding of how it was implemented. It was over 3 years ago unfortunately...

I suppose that tile-resolution FoW could be enough for first cut, even if not looking very nice - but I will probably need tile-based algorithms anyway to determine partial/full cover.

[Azhrei]

And the amount of processing will be much higher since you have to handle 3d (i.e. elevations) and we don't have to.

Your app looks pretty nice and I think people who want their physical battlemat to move to the 'net and become electronic will find it very useful!

I know you've got a long list of desired features but one thing I noticed about the video is how hard it is to tell where a larger than Medium creature will be moving to, i.e. which squares on the grid it will occupy.

[revinor]

And here comes the token support You can probably recognize the token images themselves...

[Full Bleed]

I know you've got a long list of desired features but one thing I noticed about the video is how hard it is to tell where a larger than Medium creature will be moving to, i.e. which squares on the grid it will occupy.

Yeah, having a token base of some type would probably be nice even if it's just highlighted squares.

Speaking of which... it does look like he's already got the bones for bases in his current 2d token support....

And here comes the token support You can probably recognize the token images themselves...

Nifty.

Though I think I'd prefer it if they stood upright/perpendicular to the battle mat (like coins standing on edge.) Then you could do other cool things with them:

1) When the dice hit them you could have them spin a bit.
2) When "Prone/Dead/Unconscious" you could flip the token down.
3) You could have a "back" image of the 2d token (though default would just mirror the front.)

Etc.


Edit: Even though I'm with you on avoiding the 3d walls for now, I was just thinking that it might be nice to be able to place 3d doors. It might look a little strange at first, but I find that when using 2d maps, sometimes it's hard to "see" doors (especially when using FoW/VBL.)

[Full Bleed]

I suppose that tile-resolution FoW could be enough for first cut, even if not looking very nice - but I will probably need tile-based algorithms anyway to determine partial/full cover.

Could you explain what you mean by "tile-resolution FoW" a little better? I'm not sure what that would look like.

[wolph42]

I suppose that tile-resolution FoW could be enough for first cut, even if not looking very nice - but I will probably need tile-based algorithms anyway to determine partial/full cover.

Could you explain what you mean by "tile-resolution FoW" a little better? I'm not sure what that would look like.

my guess would be that a square (grid) would be either fully covered with FoW or not covered at all. This would keep things simple to start with (though a bit clunky).