I would go as far and say that we are not limited by anything there are off course constraints that we need to work with (time being one of the biggest ones) but its the same constraints that everyone has to deal with, regardless of which engine/tool they use.

As for the environments, the vision hasn't changed. we still have tens off thousands of planets in EVE that you will get to fight on, one way or another We don't have the technology yet to create a unique environment for each one of them (nobody has). but with every iteration we get closer.

Its a balance between art, design and tech. We could randomly generate maps and say that we have unique environments everywhere but they would probably look bad and not be very fun. (It would be fun for me to create them though) The other end of the spectrum is to hyper design every corner for optimal gameplay and polish every bit of rock and blade of grass, but then you'll endup with a game that only has 1 map. So its all about balance.

The socket system that you mentioned is a good example of this. This system, along with a few others we have is designed from the ground up be amazingly flexible and allow for huge amounts of variation in our environment. Its used in Dust 514 in a relatively simple form because that how it gives us the most out of the tech,art and design. It is something that we will continue to build on top of so that ultimately we'll end-up with unique environments on every planet

I would go as far and say that we are not limited by anything there are off course constraints that we need to work with (time being one of the biggest ones) but its the same constraints that everyone has to deal with, regardless of which engine/tool they use.

As for the environments, the vision hasn't changed. we still have tens off thousands of planets in EVE that you will get to fight on, one way or another We don't have the technology yet to create a unique environment for each one of them (nobody has). but with every iteration we get closer.

Its a balance between art, design and tech. We could randomly generate maps and say that we have unique environments everywhere but they would probably look bad and not be very fun. (It would be fun for me to create them though) The other end of the spectrum is to hyper design every corner for optimal gameplay and polish every bit of rock and blade of grass, but then you'll endup with a game that only has 1 map. So its all about balance.

The socket system that you mentioned is a good example of this. This system, along with a few others we have is designed from the ground up be amazingly flexible and allow for huge amounts of variation in our environment. Its used in Dust 514 in a relatively simple form because that how it gives us the most out of the tech,art and design. It is something that we will continue to build on top of so that ultimately we'll end-up with unique environments on every planet

In the distant past, some players thought EVE players would build the maps...

The vision was that Planetary Interaction (PI) players in EVE would have the option of laying out the shape of the on-planet facilities they put down. That was never supported by CCP that I know of but the work done since on sockets and pre-built socket components combined with the new platform might allow the idea to be revisited.

The way it would work would be that a PI player would lay down his facilities and have the option of doing detailed facility design using the components the Legion team put together or just allowing the the Legion defaults to define the maps.

Pluses:

• Players would relieve some of the burden of populating all the worlds
• No two planets (and at least potentially no two maps) would be the same
• The EVE players and Legion players have a point of interaction
- - Legion players fight over components EVE player care about
- - EVE players can have a hand in protecting components they really want to preserve
- - The nature of PI components moving as the planet is mined changes the tactical pattern on the planet over time

• There is potential for Legion players to hold and mine territory on their own

Minuses:

• There has to be a mechanism in EVE that allows for facility layout
• All the components / terrain / mechanics has to be created / coordinated with the EVE side
• Integration with the rest of the game (Planetary Conquest especially) has to be taken into account

In the distant past, some players thought EVE players would build the maps...

The vision was that Planetary Interaction (PI) players in EVE would have the option of laying out the shape of the on-planet facilities they put down. That was never supported by CCP that I know of but the work done since on sockets and pre-built socket components combined with the new platform might allow the idea to be revisited.

The way it would work would be that a PI player would lay down his facilities and have the option of doing detailed facility design using the components the Legion team put together or just allowing the the Legion defaults to define the maps.

Pluses:

• Players would relieve some of the burden of populating all the worlds
• No two planets (and at least potentially no two maps) would be the same
• The EVE players and Legion players have a point of interaction
- - Legion players fight over components EVE player care about
- - EVE players can have a hand in protecting components they really want to preserve
- - The nature of PI components moving as the planet is mined changes the tactical pattern on the planet over time

• There is potential for Legion players to hold and mine territory on their own

Minuses:

• There has to be a mechanism in EVE that allows for facility layout
• All the components / terrain / mechanics has to be created / coordinated with the EVE side
• Integration with the rest of the game (Planetary Conquest especially) has to be taken into account

The only reason not to move would be licensing, aka saving some cash. Although the $20/month/seat seems cheap enough to those on the outside. Unless another $1600/month might break the studio, or not. Probably less considering that HR, admin and PR don't need licenses.

Some of us do follow the trades and these statements about sticking to UE3 is very difficult to believe as you might well be the only studio in the world to make that decision. In particular when the game that currently uses it has not been very well received.

I'd rather see form follow function...

Designate sockets as power generators, dormitories, ground-to-space resource movers, research labs, mining infrastructure, shield generators, vehicle production facilities, spaceports, storage hangars, armories, drone bays, etc and allow district owners to build, configure, and fill structures to achieve the bonus/function they want to achieve:

-Planetary Shielding
-Vehicle Factory
-Clone Farm
-Military Installation
-Advanced weapon research
-Fuel/Power depot

Then allow dust bunnies to fight in and capture/destroy them and make them susceptible to damage from orbitals, vehicle damage, and small arms fire...but not destructible. Instead let damage linearly decrease efficiency of the sockets.

THEN you could let the player base decide where to put the sockets within the sandbox and judiciously decide which district shields the planet, which district powers all the others, which district houses the planet's inventory of vehicles, etc. Force eve players to decide between wearing down the shield to attack on a global scale or send smaller infiltration vessels to bring down the shield, sabotage/capture key strategic resources, or capture the clone inventory and armory and attack from within.

I don't care if the physics played like dust 514 and were built on under the UE3 emblem, I'd play the **** out of Legion and would get an eve account to run my own infiltration missions.

Then add in some space scanning and surface mining like they had back in Starcraft for sega Genesis and maybe work on a viable drone AI....

So long society. Hello basement!

Yep. I've brought up the question numerous times about Unreal 3 vs Unreal 4. Apparently we're using a hugely modified version of Unreal 3 now - to the point where moving to Unreal 4 just doesn't seem realistic. Most of the things we'd want from Unreal 4 we've already created ourselves at this point anyway!

I do know though that the project is currently running DirectX 11.

Edit: I think an example of where we've drastically improved beyond just Unreal 3 is the lighting. CCP Android works on this stuff a great deal. I'll try to get him to chime in with some details later.

Have you guys put together yet that making **** shiny and gleaming does not make it look good? That dark maps are fine, so long as the players have some control to be able to see, that glare DOES NOT give a sense of immersion and that it only makes your players ******* annoyed?

Totally agree. That's what's really being missed in dust, player content.

We are running on a heavily modified UE3 engine. Heavily modified means that we have taken almost all parts of it and extended or changed them. Take our lighting for example, we have completely moved to fully dynamic lighting that supports bounce lighting. Take a look at a presentation one of our programmers did a few years ago on our lighting. The presentation is a little old, but the key concepts are still the same.

http://advances.realtimerendering.com/s2012/CCP/Malan-Dust_514_GI_reflections(Siggraph2012).pptx

This technology has been in active development since then and is getting very advanced (IGÇÖll see if I can get CCP Photon to do a devblog on the current state of it and where we intend to take since it)

We are also using a few very impressive middleware technologies. That means that we get code from companies that are experts in a certain technology and integrate it into our game. The NVIDIA PhysX middleware is a good example. It contains things like advanced particles, advanced physics and GǪ yes destruction. So the reason we are not doing destruction right now is not that our engine does not support it, but rather its not our focus right now.

What engine is chosen at the beginning of development should be thought of more as a starting point for a game rather than a box that the game is developed in.

Well actually it works for me

Trying to battle enemies with the sun at their backs is tough, but a valid tactic nonetheless.

Agree that growing the sandbox is 10x more important than artistic polishing...

Where does all this experience with Unreal 3 come from? Is it soley from Dust 514? I looked at the wiki on eve and it uses some other engine.
Dust has been in production for what a year?
I am guessing that the devs must have experience before dust with unreal 3 otherwise a couple years experience with an engine isnt much.

So far CCP really only has 1 game and thats Eve. Dust is like some short term experiment they couldn't get right and decided to not waste effort on.
I doubt legion will go very far. People only give a game company so many chances.
It would probably be a good idea for CCP to make a sub company and release a legion or dust clone as the sub company so that people dont realize what happening.

Isnt the EVE engine based loosely on the Unreal engine, or rather an extreme modification of it? I thought to hear them mention it, but cant tell exactly where.

I don't want to pretend to know what I'm talking about, but I don't think the Engine has much to do with the game running on a next-gen console or PC. From what I've heard, we've modified the Unreal 3 engine so much that we've already implemented a lot of the additional features we'd be looking to - and at this point it'd be easier to implement whatever else we need rather than to rip out the engine and start again.

by easier you mean cost less money to use the old engine.

==

If only you would understand what the hell you're talking about.

You know Valve?
You know what their Engine is?

Their engine is the Source Engine.

That's based on Gold Source.

That's based on the Quake 1 engine.

And they have been adding features and refining it ever since it came out.
It's actually very different from the original source engine.

It's nicknamed the Tower of Duct tape.
But it works, and I doubt that you will find any resemblance to Quake 1 on the Source Engine, would you?

You're comparing a legend in the business, Valve, to CCP? hahahahaaha...*wipes tears away*

We are running on a heavily modified UE3 engine. Heavily modified means that we have taken almost all parts of it and extended or changed them. Take our lighting for example, we have completely moved to fully dynamic lighting that supports bounce lighting. Take a look at a presentation one of our programmers did a few years ago on our lighting. The presentation is a little old, but the key concepts are still the same.

http://advances.realtimerendering.com/s2012/CCP/Malan-Dust_514_GI_reflections(Siggraph2012).pptx

This technology has been in active development since then and is getting very advanced (IGÇÖll see if I can get CCP Photon to do a devblog on the current state of it and where we intend to take since it)

We are also using a few very impressive middleware technologies. That means that we get code from companies that are experts in a certain technology and integrate it into our game. The NVIDIA PhysX middleware is a good example. It contains things like advanced particles, advanced physics and GǪ yes destruction. So the reason we are not doing destruction right now is not that our engine does not support it, but rather its not our focus right now.

What engine is chosen at the beginning of development should be thought of more as a starting point for a game rather than a box that the game is developed in.

Good, good human.
I hope you killed CCP iOS in his sleep.