Some thoughts on other engines (Unity, Unreal)

[Dairuka 888]

Tossed in the small maul into unreal for fun

 

Edited March 27, 2015 by Dairuka

[Mark 47]

I've been working with Unity a lot lately, and I am a budding indie dev with it. I would love to help port the client to Unity.

 

Here's a canopy bed and apple tree pulled into Unity on a base terrain with the Wurm grass texture in 5 minutes: (Yes, I am using Unity in Ubuntu, I'm just boss like that).

 

 

You could use a voxel/marching cubes asset from the store for your terrain, and that would give everyone that cries for true 3D map what they want and open up so many possibilities for your development.

 

Some examples are: 

https://www.assetstore.unity3d.com/en/#!/content/8131

 

http://forum.unity3d.com/threads/terrainengine-marching-cubes-infinite-procedural-smooth-terrain.174595/

 

TerrainEngine is by far the most sought after and powerful (even boasts flowing volumetric fluids such as water) but has a hefty (around $3000) price tag.

 

Using Unity so so much easier than any engine I have used, including Unreal. With Unity, bridges would have been so much easier to implement. The engine would handle all your physics calculations, lighting, everything. You could focus on game development and content without having to code all of the engine yourselves.

 

If you switched to Unity, not only would Wurm still be available on PC, Linux and Mac, but also be native code. Then, you could port it to Android, iOS, XBox, PS4, etc.

 

I want to help. I made a small arcade game called Base Defender (now I am calling it Ardalon) you should check out: https://www.dropbox.com/s/tsapc0yf04pgtmd/Base%20Defender%20004%20Setup.exe?dl=0

 

I didn't intend on this being an application to join Wurm, but it quickly morphed into it. 

Edited March 27, 2015 by Mark

[Shiraek 464]

Tossed in the small maul into unreal for fun

 

 

That looks great!

 

Something we do where you've got two different materials in an object is to split it into two meshes and apply two different shaders. You could have a metal one for the head, brace and pommel and a "woodlike" one for the haft and handle. I've no idea of course what you've done there so take this post with a bucket of salt as ever

[Planck 22]

Wurm and a very large majority of other "3D" games avoid treating the vertical dimension as an equal for reasons of performance for the most common cases. That means that many features that would be simple in a true 3D engine are very difficult to implement correctly in any of them.

 

This is not a slur on Wurm, or any of the others. It's just the nature of our computing hardware (both client and server) that getting decent performance out of 3D requires simplifying the world model drastically. Wurm makes the simplification that the vast majority of terrain can be described in terms of a heightmap. That means that any terrain that is not a heightmap almost certainly requires lots of special-case coding. The more types of exceptions there are, their interactions with the many types of player actions and other exceptional conditions become complex faster than most people can imagine. That is the sense in which Wurm is "basically a 2D game".

 

That's why I'm not in the slightest surprised that bridges are taking so long. I'm actually surprised they went ahead with introducing them at all.

 

Regarding other game engines: Many simplifications cannot be applied to Wurm, yet are buried deep within the code of those engines. It is drastically misleading to take a tiny client-only example and assume that it must scale up without problems to a server/client architecture with a thousand concurrent players modifying a huge world, some of whom might well cheat if given the opportunity. For those who laud Unity's physics engine, try applying it to a world with tens of millions of terrain tiles, many millions of trees, hundreds of thousands of freely buildable barriers, a thousand or so concurrently moving objects under player control, and feed the results to a thousand clients at once. Then compare its performance to Wurm's.

 

In short, "this game engine makes my tiny standalone game easy and performs wonderfully" is not at all any sort of evidence for "it would make programming Wurm easy and perform wonderfully".

[Ulviirala 431]

A lot of good points why it doesn't seem to be a good idea to change client language and graphics engine. I mean, it's not a "huge" task to port the client to another language, I think that "gargantuan" is more fitting. Coming up with an organised plan alone must be a nightmare.

Programming techniques that work really well in language A, might perform really badly in language B, or not work at all. Perhaps requiring fundamental changes in a software's architecture. Sooner rather than later you'll find yourself crying in front of a mess, so just develop from scratch and implement the protocols and logic.

But that too is a financially hefty investment (because it requires time, and time equals money), that must first be proven to be profitable later down the path, which is done before development even starts.

Then again, does Wurm really need a different graphics engine, one that has not yet proven its worth in the uniquely dynamic game world of Wurm? I'd say no, but that's a highly subjective opinion of course.

Of course I absolutely love it, when old items get new models that look many times better, but I'm not keen on a brown 10fps experience with bloom, motion blur, and depth of field for example. I don't feel that I'd get any value from that, and I can hardly imagine Unity doing a faster job than the current engine, that's been in development for so many years, and probably being very specialised.

I would believe that it is a minority that is turned away by the current fidelity of the graphics, too. Would better graphics really help the game get significantly more players, while everything else stays the same?

Back in 2003 I was discussing an idea for a game with a friend, and when I told him I wanted dynamic terrain to be terraformable by players, he immediately said that it would be impossible. After much talking, I had come up with a working prototype, it was awesome, though a finished game has never been made.

Imagine my reaction when I started playing Wurm in 2008 .

Wurm. An eternal love that has often been challenged, because drama (omgrolf!?yudothis?!) and stuff (rolf?!nooooooo!!watrudoin) *chuckles*.

Edited March 29, 2015 by Ulviirala

[Kegan 3,926]

The thing that scares me from a player standpoint is that if the game changes to much it might not be as appealing as before. If moving the game to a new engine completely changes things like the building or the terrain modification then we might lose the uniqueness that is wurm.    

[Shiraek 464]

Been biting my tongue on this thread, it's hard not to wade in an argue, but I don't see a lot of point in getting too bogged down in it. However, I will say the following:

• If you're doing all your physics server side, moving to a new engine for your front end purely for better visuals is made less complex, not more
  
• If you're validating every single thing every player does every time instead of a random sample, you're wasting resources. Cheats won't get away with it forever with a random sample.
  
• You can kill off a lot of cheating (macroing) with game design that removes actions that are easy to repeat
  
• Bridge collision could be done with a series of box colliders, since no doubt they're made of components like our existing walls and floors. Ray casts through and collisions with box colliders are not complex to implement, even from scratch in C++
  

 

Queue "what would you know about client-server MMO development?" in 3...2...1...

[Outlaw 697]

One thing to mention about using voxel based terrain like in TerraVol etc...It just doesn't look/feel like Wurm without our current system of tiles.

Like Kegan said, change is good...but it's easy to change too much and remove what makes Wurm...Wurm.

[Tipper 260]

CC is supposed to be working on an overhaul of the engine but remaining in Java, something that will be better optimized and more modern-looking in lighting and such.

[sunsvortex 599]

Really, if WURM were to go to a new engine, id vote for unreal -specifically because of this - This is a MUST WATCH at least the first 5 to 8 minutes. The terrain is modeled on Isle of Skye.

 

[media]https://www.youtube.com/watch?v=clakekAHQx0

Edited March 31, 2015 by sunsvortex

[Ulviirala 431]

[...]

Queue "what would you know about client-server MMO development?" in 3...2...1...

I don't think that a lot of people have confused the client and server side of things unreasonably much, quite the opposite I felt (I've been working to large parts with client-server architectures for a few years), surprisingly.

 

[...]You can kill off a lot of cheating (macroing) with game design that removes actions that are easy to repeat[...]

The vast majority of actions are easy to repeat, I'd say. And it's been made easier with the more recent "select bar" window, as well as the previously added bunch of keybindings. Things added for the honest players' sake, not to make cheating even easier.

 

[...]If you're doing all your physics server side, moving to a new engine for your front end purely for better visuals is made less complex, not more[...]

One thing to mention about using voxel based terrain like in TerraVol etc...It just doesn't look/feel like Wurm without our current system of tiles.

Like Kegan said, change is good...but it's easy to change too much and remove what makes Wurm...Wurm.

I think this, over all technical issues, is one of the most important things to consider. The look and feel of a game, not only gives it a certain characteristic, but it also makes a very strong impression on the players. Would you still like Wurm as much if it looked vastly different? I'd probably have a hard time with it. I don't play some games merely because I receive a bad look and feel from them.

And it's increasingly difficult, the longer a game has already established its look and feel. Wurm has been going for a long time, and even with the updates over the years, has always kept the same look and feel. It also remains to gauge, if a change of this extent would bring considerably more players to the game, to warrant such an investment.

I'm sorry if the thread didn't go into a direction that you did hope it would.