new development?

[bnolsen]

First disclaimer: Not saying I'm going to do this, just exploring.

I really enjoyed SP1 when it was first released and I played SP2 also. I noticed control of the code was released but it didn't change much. I was really hoping that a linux port could happen.

Of course now we're looking at a new crop of very powerful handhelds (dingoo A320, GPH Wiz, Pandora). And I think steel panthers could find new life as a handheld title on these platforms. And there's the added bonus of support for multiple modern operating systems with the ability to port to new ones.

The absolute biggest problem for game development tends to be art development, accuracy and balance. Just looking around recently I found that a ton of people have contributed a lot of this information to make steel panthers better. Possibly to the point that this data is no longer owned by matrix and could be reused in a new game that leverages the same type of data without infringing on whatever copyrights are held by the original SP series.

Resources:
Armor penetration spreadsheet:
http://forum.shrapnelgames.com/showthread.php?t=43977

Game sounds:
http://forum.shrapnelgames.com/showthread.php?t=41246

Vehicle art:
http://ww2drawings.jexiste.fr/Files/1-Vehicles/Vehicles.htm

And apparently there's tons of OOB stuff out there.

I'm guessing if this game "engine" were open source this could be a fairly painless success? The hardest part I see is actually doing the "glue" work and incorporating all the necessary data out there without stepping on anyone's toes. I'm guess the "open source" part is key here since this would wouldn't work unless it were a community effort. I do believe there should still be the ability for folks to "sell" module content for this game (big campaigns, etc). I believe the value of something like this is the content (story, art, data) and not necessarily the engine.

 

[Dion]

About 5 years ago, there was talk about someone making Sreel Panthers IV. Anybody hear anything about it? I guess it was just a rumor.

 

[Redwolf]

The absolute biggest problem for game development tends to be art development, accuracy and balance.
[...]
And apparently there's tons of OOB stuff out there.

This is exactly why I am still sniffing around implementing an OpenSource wargame. If you manage to get it off the ground to a point that the modders and the grogs start contributing (no matter how ugly and limited that might be), you win. Big time.

The communities of (potential) contributors for wargames are very impressive.

I aim at a CMx1 clone, though, with combat mechanics more taken from TacOps than CMx1 (SOPs instead of most TacAI etc).

I'm guessing if this game "engine" were open source this could be a fairly painless success? The hardest part I see is actually doing the "glue" work and incorporating all the necessary data out there without stepping on anyone's toes. I'm guess the "open source" part is key here since this would wouldn't work unless it were a community effort. I do believe there should still be the ability for folks to "sell" module content for this game (big campaigns, etc). I believe the value of something like this is the content (story, art, data) and not necessarily the engine.

Well, I wouldn't underestimate how difficult a working engine is. I should know, the terrain managers in Ogre and OpenScenegraph were huge disappointments.

But you make valid points.

 

[bnolsen]

This is exactly why I am still sniffing around implementing an OpenSource wargame. If you manage to get it off the ground to a point that the modders and the grogs start contributing (no matter how ugly and limited that might be), you win. Big time.

The communities of (potential) contributors for wargames are very impressive.

I aim at a CMx1 clone, though, with combat mechanics more taken from TacOps than CMx1 (SOPs instead of most TacAI etc).

Well, I wouldn't underestimate how difficult a working engine is. I should know, the terrain managers in Ogre and OpenScenegraph were huge disappointments.

But you make valid points.

Engines that try to be everything to everyone end up generally being deficient in some way.

To me chasing something like Steel Panthers is a good start since the hexagonal geometry is fairly simple and the game should scale quite well to the handheld world which I believe desparately needs a solid, mature wargaming experience.

I've got this nasty habit of recoding things at a somewhat low level. I work with photogrammetry and geometry, including software to handle viewing and visualizing point clouds in the billions without using opengl (it doesn't scale very well and OSG just didn't have any of the features I needed). I'm always open to stealing good libraries for basic stuff. The engine should be pretty bare to the metal and definitely cross platform (down to mobile devices even).

 

[Redwolf]

Engines that try to be everything to everyone end up generally being deficient in some way.

Believe me, I know

I am currently surveying the Ogre codebase, piece by piece, making a list of what is actually useful for me (such as rendering backend, 3D model importers, animation support) and what is useless (terrain, raytracing etc.) to then estimate whether it is overall a win.

To me chasing something like Steel Panthers is a good start since the hexagonal geometry is fairly simple and the game should scale quite well to the handheld world which I believe desparately needs a solid, mature wargaming experience.

I don't think you have to go hexes, and I won't do IGOYGO.

I do, however, consider basing an implementation on 2D. Namely, use a bitmap mixing engine like GEGL to build a game like TacOps or CMx1 in views 6-9 (straight up), while displaying terrain with height contours and shadows.

I've got this nasty habit of recoding things at a somewhat low level. I work with photogrammetry and geometry, including software to handle viewing and visualizing point clouds in the billions without using opengl (it doesn't scale very well and OSG just didn't have any of the features I needed). I'm always open to stealing good libraries for basic stuff. The engine should be pretty bare to the metal and definitely cross platform (down to mobile devices even).

Ogre is reasonable from what I can see. That still doesn't mean it won't pick a fight with me later.

I like GEGL for implementing a 2D top-view game. It would allow me to smoothly mix in terrain, map, units, command overlays, player orders and said height contours and shadows. I haven't actually coded with it yet, who knows whether it holds up to the promises.

 

[bnolsen]

Believe me, I know

I don't think you have to go hexes, and I won't do IGOYGO.

I do, however, consider basing an implementation on 2D. Namely, use a bitmap mixing engine like GEGL to build a game like TacOps or CMx1 in views 6-9 (straight up), while displaying terrain with height contours and shadows.

Picking and choosing from already established good ideas is a good thing.

I'd be curious to do a quick compare between evas and gegl. I know that evas scales to embedded. And there's cairo, agg, etc. But knowing what gimp can do that does likely make gegl a front runner. And if it's all too irritating can always swap out portions. Also need to check which ones of these libraries are friendly towards system level optimizations (fixed point, NEON, etc).

One motivation for redoing SP is the existence of all these known file formats people are used to dealing with. I'm not nearly as familiar with tacops or CMx1 (and I haven't played either of those games anyways).

You primarily have PC in mind here?

 

[Redwolf]

I'd be curious to do a quick compare between evas and gegl. I know that evas scales to embedded. And there's cairo, agg, etc. But knowing what gimp can do that does likely make gegl a front runner. And if it's all too irritating can always swap out portions. Also need to check which ones of these libraries are friendly towards system level optimizations (fixed point, NEON, etc).

One motivation for redoing SP is the existence of all these known file formats people are used to dealing with. I'm not nearly as familiar with tacops or CMx1 (and I haven't played either of those games anyways).

You primarily have PC in mind here?

Myself, I only do Unix derivates. I'll be friendly enough to pick tools that allow a direct compilation on Windows, of course MacOS X shouldn't be a problem, at worst people will have to install the free X11 server.

Never looked into EVAS. Since I am a heavy user of GIMP but never used enlightenment it's pretty safe to say that GEGL would be better for me even if I'd have to hack it up somewhat. I don't think I need any optimizations here, current machines are way overpowered and my game wouldn't be any worse than GIMP's layer mixing.

My hopes here are that Ogre (or OpenSceneGraph) will turn out to be within range of this project.

There is no known file format for CMx1, we would have to OCR the scenarios out of the editor (those we get permission for by the authors). But the CMx1 community has already shown that they are willing and able to do so, and many have translated scenarios from CMBO to CMAK by entirely doing it by hand tile by tile (shame on BFC for not assisting even a bit).

Any OpenSource game that I/we build would be able to use scenarios and OOBs from any other, obviously. Might not play the same when going from hex/IGOYGO to pixels/WEGO, but the terrain and the units would be there.

Makes you wonder why we wargamers are the ones that are stuck with vendors that categorically won't OpenSource anything obsolete, even when it's given away for free for a long time now, while all the FPS kids have all the Doom and Quake sources to build upon. If our wargame vendors end up being eaten by an OpenSource game that gains too much momentum thanks to forcing us to go all-out right from start they wouldn't have anybody but themselves to blame, of course.

I just want to write an AI, why do I have to do a whole game?

 

[bnolsen]

Myself, I only do Unix derivates. I'll be friendly enough to pick tools that allow a direct compilation on Windows, of course MacOS X shouldn't be a problem, at worst people will have to install the free X11 server.

I just want to write an AI, why do I have to do a whole game?

Whatever I might work on right now I'd develop on parallel targeting linux (using cross platform libraries) and psp. I'm not so sure a small screen would handle something too sophisticated so well.

AI would be an area of weakness for me, never tried one yet.

 

[Redwolf]

AI would be an area of weakness for me, never tried one yet.

Well, that's the sad thing. It's what I want to do but it is completely optional and must come after something that is playable at all. Since something that is playable with two players is easier than single vs. AI play the AI would come sometime around the next stone age or something.

 

[bnolsen]

Well I bought myself a viewsonic g-tablet (10", tegra2 chipset) and installed full android 2.2 on it.

Although I think tablets are of limited use I believe they are *the* proper vehicle for resurrecting board games.

I just need to kick my own ass on this. Actually what I would like would be artwork (terrain, tanks, units) and OOB, definitely NOT taken from any of the formal releases (keep this clean of copyright!).

Also I'd like some "specs" or code on how to load/interface with the OOB.

Any hints on where to look for this stuff?