new game engine for uru?

[andylegate]

Okay, well for right now, I'm just going ahead with the "proof of concept" and get Ahnonay Linking Room in here:



What I've gotten done so far. Had a lot of nasty surprises with CryEngine:

1) Diffused Textures (your main texture) in your materials MUST be in .tif or .dds format (can't use jpg, tga or png).
2) Normal maps and bump maps MUST be in .dds format.
3) All meshes exported from Max, must have a smoothing group applied to them (CryEngine's import will try to do it for you, but it's not recommended as it always comes out horrible).
4) CryEngine is 1 Unit = 1 Meter (Ahhhhhggggg! Not metric! no no no no no no.....), and I'm having issues with scaling as the way everyone recommends exporting is not working. The hallway was only 4.53 centimeters....okay, so would work if you are a mouse! I finally found something that will work fine and keep the meshes consistent: leave Max's system units set to 1 unit = 1 foot, and use the exporter for CryEngine, then when I get it into CryEngine, I just scale the meshes up by a factor of 32.8083 (as 1 meter is 3.28083 feet). I tested it out, and the places looks the correct scale.
5) I'm snapping the meshes into place since their pivot points were retained, by giving their world location at 0, 0, 0.

NOW that I'm finally starting it, it's going fast. I hope to have the basics of it done by tomorrow afternoon (if I don't kill my son for not doing his school work first!).

After this, I'll try my age Serene since it's an outside age, has wind and snow falling, only instead of the dark purple fog all the time, I'm going to give it a day cycle and allow the sun to stream in through the trees, etc, etc. Should be pretty cool (I hope!).

@ A MOUL "reboot": I like the idea. Any ideas on how to start it? I mean they did it for Star Trek, why not Uru? heheheheh. Maybe I can sweet talk Tony into letting us use just 1 Age, changed of course, let's say Relto but without the pilars and different books? A broken Nexus? The Cleft, only re-done with our own meshes and textures? (I think we could make it look better, but then that's just MHO on that).

Let me hear your ideas, I say we really don't have anything to loose here, and this could be a wonderful experiment (I don't care if it works or not, at least it's better to try SOMETHING than to be sitting around with my thumb up my......well, you get the picture......).

[GPNMilano]

I'm confused as all get out. If you're so keen on upgrading Uru why not just upgrade the Plasma engine so it's up to date with the current engines already out there. Plasma was ahead of it's time when it was made, it shouldn't take that much effort to bring the engine in line graphically with all the others out there. Especially considering we have the source code to do that. Plus I'd tend to lean on the idea that Cyan would be much more willing to let us use their assets for the existing ages if we're to make advanced steps to updating the engine, then they would if we want to port stuff over to an entirely new engine. Plus the amount of time it would take to do one versus the other is a factor. To update the engine and port the existing content to the newer updated versions would take much shorter time then rebuilding it all from the ground up around a new engine. We have access to Cyan's original ages already, and even if Cyan doesn't wish to give us their asset files from Max, we can simply import the ages into blender and export from there or bring them to max and use the PlasmaPlugin.

[Deledrius]

I'm confused as all get out. If you're so keen on upgrading Uru why not just upgrade the Plasma engine so it's up to date with the current engines already out there. Plasma was ahead of it's time when it was made, it shouldn't take that much effort to bring the engine in line graphically with all the others out there. Especially considering we have the source code to do that. Plus I'd tend to lean on the idea that Cyan would be much more willing to let us use their assets for the existing ages if we're to make advanced steps to updating the engine, then they would if we want to port stuff over to an entirely new engine. Plus the amount of time it would take to do one versus the other is a factor. To update the engine and port the existing content to the newer updated versions would take much shorter time then rebuilding it all from the ground up around a new engine. We have access to Cyan's original ages already, and even if Cyan doesn't wish to give us their asset files from Max, we can simply import the ages into blender and export from there or bring them to max and use the PlasmaPlugin.

Precisely. What confuses me is: why now? If you're going to make Uru (or an Uru-like) in another engine, why wait until we can finally make improvements to Plasma? Other engines and SDKs have been available for many years (although these particular ones I admit are relatively new). What you are doing here is impressive-looking and obviously has taken some effort. But it seems a strange split of energy; it would make more sense either to make original content in a new engines for which you can optimize things, or if you're intent on sticking with Uru content, why not help to improve Plasma where the content already is and the conversion work is minimal?

As for the Riven content, this worries me. There's no reason that data can't or in fact shouldn't be updated to still be readable. Data is data, and it's always convertible/upgradable. There are lots of people capable of doing this in the fan community, although I'm sure it's unlikely Cyan would go that route, and they obviously don't have the resources or expertise to do it themselves. Meanwhile, having it bitrot on a single machine is troubling. That's how data dies. Not by being in an old format, but on old hardware that eventually ceases to function, without mirrors and backups.

[Wamduskasapa]

@ A MOUL "reboot": I like the idea. Any ideas on how to start it? I mean they did it for Star Trek, why not Uru? heheheheh. Maybe I can sweet talk Tony into letting us use just 1 Age, changed of course, let's say Relto but without the pilars and different books? A broken Nexus? The Cleft, only re-done with our own meshes and textures? (I think we could make it look better, but then that's just MHO on that).

Let me hear your ideas, I say we really don't have anything to loose here, and this could be a wonderful experiment (I don't care if it works or not, at least it's better to try SOMETHING than to be sitting around with my thumb up my......well, you get the picture......).

Cyan has given us Aitrus' Maps (BEWARE: LARGE in excess of 2.5 MB ea) click icon for full size image

The 1st is the original -- the 2nd is translated and -- the 3rd has the angles.

Why confine ourselves to what has already been rendered?? There is still a whole world to explore...

[andylegate]

@ GPNMilano and Deledrius: Because I'm an artist, 3D modeler and texture maker. I'm NOT a programmer, all though I have some passing knowledge of scripting here and there (and I do quite remember my BASIC from high school that I learned on teletypes and then Atari 400's).

I can not modify Plasma because I do not know how. I do not know how to make Plasma use Normal Maps for rendering. I do not know how to make Plasma use Bump Offsets so that the motar in a 2D brick texture will actually change aspect ratios, giving a flat plane the look at each and every brick was actually modeled. I do not know how to change Plasma's lighting system so that it can use Lightmass and provide some of the most perfect atmospheric lighting, bounce and scatter I have ever scene. I do not know how to change Plasma so that it has a very intelligent AI system for NPC's. I do not know how to change Plasma to do many, many other things.

If the tool I'm using isn't enough for the job: get a new tool that will work.

I could wait, sure. I could sit here and hope all these license issues are taken care of. I can sit here and wait, and hope, and doodle on a note pad. But I know that many of the people here that DO know how to do the stuff I'm talking about above, are quite busy as it is. They have RL to take care of too. And that's okay with me, I understand that just fine.

In the mean time, I can use another game engine and have some wild fun with it.

As I posted earlier: Uru can not be LEGALLY ported over to another game engine. Mark warned me off on that. Messing with Plasma is one thing. Porting over all of Uru's content to another game engine will get me into deep kimchie with the Fryman. That won't be tolerated by them.

That's okay too.

So I'll say it again: I can't "improve Plasma", I'm not a programmer. I'm a 3D modeler. BIG difference.

Anyways, we're just experimenting. I'm doing the work right now, so I'm cool. Meanwhile I'm learning at lightspeed here, heheh:

More meshes added, with the afternoon sun streaming in from the roof, since I don't have it on yet!

[GPNMilano]

@Andy, Sorry for the confusion. No one said YOU had to improve Plasma. There are many here quite versed in that knowledge. My point was that, while yes Plasma has it's drawbacks. There are things about it that need to be fixed. There are things about it that need to be improved. However, if you want to continue to build for Uru, my suggestion would be to fix the engine Uru is built upon. Continue to develop content for Uru, and work with the programmers here make your suggestions and your desires for the engine known on the Plasma section of the GoW's Github account. H-uru. I believe Kaelis submitted an issue on the wishlist for Arist controlled shader support. If there's other stuff on your wishlist you can submit an issue with it and, eventually, it'll be addressed at some point.

Be that as it may, I understand where you're coming from. I respect it as an artist. However Cyan's stance seems to be pretty clear, Uru's content they've come up with is for the Plasma engine only. So tearing out the entire engine and building Uru around a new one seems to be moot. Uru's content is what makes Uru the game that it is, without this content, it wouldn't be Uru, it would just be an entirely new game built around a new engine.

[Chacal]

Knowledge and experimentation are good per se.
There is no reason why several paths can't be explored at once.
Who can predict, at this stage, what will and won't be useful?

[Aloys]

Andy is just having fun, but as for the potential 'serious' project of a Uru 2 on a new engine.. I don't know. A straight port would be a waste of ressources IMO; it'd be fun, but it'd be so much work, I don't see how it'd be worth it. But as for a whole new game: I don't see how that could be successful commercially speaking; so it'd be just for fun.. And I guess in that regard it could be successful. But creating a new game from scratch, especially a MMO, is just a huge amount of work. And for a volunteer-based community effort, you'd have to count many years before anything comes to fruitition..

Andy: this looks really cool, even at this WIP state. I especially love the bloom effect. It's something quite subtle that's actually very efficient, and really help separates this from Plasma.

To update the engine and port the existing content to the newer updated versions would take much shorter time then rebuilding it all from the ground up around a new engine.

It depends. It can sometimes be faster to start fresh with a clean slate rather than to spend the time to sort through existing code and struggle to update it and keep it relevant with modern features. Updating an 'old' engine (Plasma is what, 9 years old?) is difficult and you often end up with a tangled mess of a codebase, and you end up totally rewriting so much of it anyway that it can be better to simply start from scratch with a different engine. Then you get a much more cleaner and maintainable source.

[andylegate]

@GPNMilano: Well, it's kind of like Mark said, embrace something and go for it. So what I'm doing is experimenting with a very modern game engine, and seeing if I can make "Myst / Uru" types of environments. And it's all for fun. Yah, a lot of work, but you know how the saying goes, if you really enjoy it, then it's not really work. Plasma as you said was very advanced for it's time. And yah, I could make a list for it (ranging from adding shader types, to environmental lighting, to point light beams (bye bye having to make a sprite for your light flare!), to deferred lighting and reflections, to SSDO (Screen Space indirect shadow approximation, which replaced SSAO), S^3, S^4 and S^5 (which is SubSurface Scattering, Screen-Space Self-Shadowing, Screen-Space SubSurface Scattering respectively), and a ton of other stuff. All sorts of goodies.

But then, I can play with all that now.

The old saying "Home is where the heart is." I think of Uru that way.

I wouldn't worry though. It's not like I'm going to burn through and make 35 "Ages" and set up a CryEngine server with a tailor made HUD, etc, etc. Even if I had 10 people from the community join in and help, it would take a LONG time to set anything up. So for now, it's just fun. Like getting out of the Aries K car one day, and test driving a Lamborghini. It's really hard to hand the keys back!

Anyways, I got most of the meshes in, and base materials set up. Did some bump and normal mapping (still have a lot to do on those). But tonight I started playing with the lighting. I've got the rooms in a Vis Area (they use Vis Areas to block outside lighting. You can use Portals too to achieve the same thing, but what is REALLY kick butt is the Irradiance Volumes! You can use these to simulate radiosity and add color to reflected light. So, like if you have a white light hitting a red object, it will produce a pinkish reflection on nearby surrounding objects. Just too coooooooool!

So here's my pic and I'm stopping for the night, got to get up in 5 hours. Oh, I used a Point Light Beam down the hall, hence the nice light bloom. Not only that, but if another player were to be down there, you'd see the light streaking around them!!!

[Chacal]

Looking great!