New Era in Unreal Engine or Cobra Engine?

[Kelster87]

I still don't understand what is Cobra doing now that other engines can't do. Why is that so "special"? And where do you get the info that UE can't do such things (genuine question) ?

It's the scale at which Cobra can do it. I'll use UE for my examples, as it's the one I have the most recent experince with. As you've seen, UE can do large terrains, but it won't wrap that large, beautiful terrain around a planet that is one of several hanging in the same system - the way Cobra handles LOD's is a little mad. Now add a moon, a nicely cratered one. About now, UE will choke, it's not optimized for this kind of use. You can rewrite the code to optimise it, but that kind of defeats the reason for using an off the shelf engine. And this is before you add flight (will require work), trading (more work) and all the other game stuff. Never mind scaling it to 400 million systems. UE has grown from the FPS-centic engine it once was, but it's roots are still very clear.

But it really boils down to scale. Cobra was written from the outset to deal with a game universe this big, UE wasn't really and would require work to make it so. UE has examples and base functionality included for much of what is in Elite, but again, scale is the issue.

I've also ignored the need for the devs to all learn a new set of tools, a new dev process and all that goes with it.

 

[Ian Skippy]

Licensing another engine also means you create dependencies that you avoided with proprietary solutions. If stuff doesn't work as it is supposed to, you can handle it in-house, if you are a license holder your company relies heaviliy on external service .... well, I can tell you, I know how THAT feels if stuff goes wrong over a prolonged period.

Proprietary lacks the option of external code reviews, on the other hand, there are no compliance issues to worry about.

Biggest draw back of proprietary I can pull out of my head, some fixes may never happen, hence using constant workarounds, which in the worst case can cause "bloatware" and more.

And again Star Citizen provides a great example. They started with CryEngine. As that engine was wholly unsuited for a space game they had to modify the engine a lot. It took a lot of time and still doesn't really work well, if at all. But by now it changed so much that many of the solutions to new issues that are native in new versions of CryEngine cannot be used in Star Citizen due to their frankenstein-engine. So you end up with all the drawbacks of the engine you started with, all the band-aid solutions you can throw at it and none of the real solutions more recent iterations have. Joy!

 

[VR SenseiMatty]

So when a game running UE does something not-so-well it is the devs fault, and if it is a game running Cobra it is the engine's fault? And when people explain some of the benefits of Cobra you yell "But I dont care about that, here is the game I want!", and then turn around and claim nobody explained what the advantage of Cobra was.

You seem very invested in your idea that UE is awesome and Cobra is not, while also aware you dont quite know much about any of this. Wouldn't it make sense to just drop this narrative of yours, and if you are really interested start learning about this stuff before starting these discussions?

Your answer demonstrates more an interest in defending FDEV rather than explaining something, which is useless because I'm not attacking FDEV.

 

[VR SenseiMatty]

Not being 100% sure that these are in development is quite different from stating as fact, without basis (as the full capabilities of the COBRA engine are "unknown" to most), that "the Cobra engine still lacks of core features that are instead already present in the other one".

We'll see, in time, what functionality Frontier adds to COBRA.

But you're not answering to my question...

 

[Robert Maynard]

But you're not answering to my question...

The answer to the question is that few know enough about the relative capabilities of the two game engines in question to answer it - and that's not us.

 

[VR SenseiMatty]

Licensing another engine also means you create dependencies that you avoided with proprietary solutions. If stuff doesn't work as it is supposed to, you can handle it in-house, if you are a license holder your company relies heaviliy on external service .... well, I can tell you, I know how THAT feels if stuff goes wrong over a prolonged period.

Ok so it means that if you want to develop a game on an existing engine you first need to be 100% sure that it can handle what you want to develop or you will end up with big troubles

 

[VR SenseiMatty]

The answer to the question is that few know enough about the relative capabilities of the two game engines in question to answer it - and that's not us.

My question is pros and cons of developing a game on an external engine that already have "x" features available rather than developing the game on a propertary engine that still lack those features. Feel free to leave FDEV and ED names out of the discussions. It's just for my understanding.

 

[M00ka]

Your answer demonstrates more an interest in defending FDEV rather than explaining something, which is useless because I'm not attacking FDEV.

Matty, in a way you are. You are accusing FDev of not using the best (in your opinion) software for the game.

Now please have a think about what you are requesting and the repercussions. For starters, everything stops, development and bug rectification wise for ED. No more skins, no more bug fixes, no more minor enhancements, no more Fleet Carriers and most of all, no big DLC in 12 months time. It would take FD at least 2 to 3 years to rewrite the game under UE and then make sure it works, if it works at all. And since FD would be required to pay an ongoing licencing fee to UE then they would probably be forced to either introduce a subscription system or just scrap the current game (i.e. turn off the servers) and release a new ED V2 at full price to pay for the licencing and extra work.

Do you really want all that to happen?

 

[VR SenseiMatty]

Matty, in a way you are. You are accusing FDev of not using the best (in your opinion) software for the game.

Now please have a think about what you are requesting and the repercussions. For starters, everything stops, development and bug rectification wise for ED. No more skins, no more bug fixes, no more minor enhancements, no more Fleet Carriers and most of all, no big DLC in 12 months time. It would take FD at least 2 to 3 years to rewrite the game under UE and then make sure it works, if it works at all. And since FD would be required to pay an ongoing licencing fee to UE then they would probably be forced to either introduce a subscription system or just scrap the current game (i.e. turn off the servers) and release a new ED V2 at full price to pay for the licencing and extra work.

Do you really want all that to happen?

That is clear to me, it was the first part of the discussion. Now I'm making another question considering a theoretical new game development.

My question is pros and cons of developing a game on an external engine that already have "x" features available rather than developing the game on a propertary engine that still lack those features. Feel free to leave FDEV and ED names out of the discussions. It's just for my understanding.

 

[Robert Maynard]

My question is pros and cons of developing a game on an external engine that already have "x" features available rather than developing the game on a propertary engine that still lack those features. Feel free to leave FDEV and ED names out of the discussions. It's just for my understanding.

If the game engine being considered for game development is provided and developed by a third party then it would be considered sensible to pick one that is already feature complete in terms of providing all of the requirements of the game as designed. Starting development in the hope that either the engine developer or in-house additions can add desired functionality not present or announced as being developed by the engine developer is a risk (and functionality announced as in-development may not arrive in time for the project or ever).

 

[Ian Skippy]

Your answer demonstrates more an interest in defending FDEV rather than explaining something, which is useless because I'm not attacking FDEV.

Because there are now pages of answers and you ignore them. So now I am personally done explaining things and moving on to wondering what your point is here. So again, why do you feel it is solid logic to blame demonstrable issues with UE games on the devs and what you perceive as issues with ED on the engine? You have made up your mind from the very first post and just rationalize anything away that doesn't suit your obviously false narrative. Since it doesn't matter how many people here explain it I dont see why you should expect people to keep on investing that energy.

Btw, if you understand the question you'd also understand that it is the opposite of 'defending FD', but I'll let you cross that bridge on your own.

 

[Kelster87]

I've given more than a few pointers as to why it's not a wise move, you seem to ignore them.

 

[Ian Skippy]

Ok so it means that if you want to develop a game on an existing engine you first need to be 100% sure that it can handle what you want to develop or you will end up with big troubles

Look at it this way. There have been four primary space games from this generation: ED, NMS, X4 and SC. Three of them created their own engine and successfully released their game. One used a pre-existing engine and has consistently failed at overcoming the basic hurdles they need to overcome.

Can you see some kind of pattern here?

 

[VR SenseiMatty]

Look at it this way. There have been four primary space games from this generation: ED, NMS, X4 and SC. Three of them created their own engine and successfully released their game. One used a pre-existing engine and has consistently failed at overcoming the basic hurdles they need to overcome.

Can you see some kind of pattern here?

This is an interesting point

 

[thistle]

Look at it this way. There have been four primary space games from this generation: ED, NMS, X4 and SC. Three of them created their own engine and successfully released their game. One used a pre-existing engine and has consistently failed at overcoming the basic hurdles they need to overcome.

Can you see some kind of pattern here?

And not just that, NMS and Elite are the only ones that have a procedurally generated galaxy. X4 and SC have hand-crafted system(s). What NMS and Elite have done is start with - can we procedurally generate the galaxy, or galaxies in NMS? How do we do that? Well we need our own proprietary engine to do it. And then add the game loops and bells & whistles later.

I would be doubtful starting with an off the shelf engine would work at all. But hand crafted locations would look nice.

 

[Bortas]

Having an own engine has a lot of merits... They can develop it the way they like instead of being reliant upon the will of someone else to implement, what they need... or not.

If there are things the Cobra Engine in its current incarnation cannot handle, they can expand it and make a new version with downwards compatibility, so they can build upon what's already there and don't have to re-write everything from scratch.

 

[B1rdy]

Graphic quality was never an issue in ED. The only exception is the single light source limitation, but that alone doesn't justify a engine change.

This is incorrect, the engine can and actually displays more than one dynamic light source (s. ship and srv lights), but not for suns. Sandro answered that in one of the live streams.

 

[MaxG10]

Ok so it means that if you want to develop a game on an existing engine you first need to be 100% sure that it can handle what you want to develop or you will end up with big troubles

Nope, not at all. The business of scalability is inherent for every engine, and therenever is 100% that's the nature of the beast, and perhaps the most interesting part (for ambitious coders), to see how far you can go with it.

Look, every software will run into trouble at some stage, regardless what it is. There is a reason that open source is so popular, well, many reasons and good reasons.

C++, whether the above mentioned cry engine ( I hate the history of this , Military and all!) or Unreal, Banshee, Torque and what have you not, even Unity that has C# as it's prog.lang. at it's core is C++, because it is closer to hardware and many many game frameworks written in it.

However, frankly, I think this is going too far here. Such discussion is valuable of course, but rather on github than here, hence appologies if I leave it at that, but I think many contributed and made good points.

Essentially, I see no reason for FD to make such a crazy move and abandon Cobra.

That's all I can contribute.

Best
MaxG10

P.S. .... and.... DO get a new haircut! That avatar looks as if he could not pull a hooked fish from the water. ROFL

 

[B1rdy]

I know there is a fair amount of buzz about the Ray-Tracing technology in the current top-end ranges of graphics cards but such technologies still have their limits and not everyone have cards with that capability baked into the hardware.

Though dedicated hardware is not a requirement for RT, not even DXR is.

 

[govny]

why on Earth they have to license another engine when they have a working engine that they perfectly know because they developed it and that is suited to the tipe of game they built ence they don't have to make compromises?
All this discussion is ridiculous