27 replies to this topic

[Impyrium]

Hold the show; I'm certainly not suggesting they migrate engine, if that's what you think I might be meaning. I'm simply asking, simply because I don't know much about it; is CryEngine holding MWO back in any areas?

I'm always paranoid of premade engines like this because they can be tricky to generate games other than their original 'show' game. Now, MW:O has done a pretty good job in bringing that MechWarrior experience to such an engine. But in terms of hit detection errors and other problems, do you think not working with their own engine is slowing down their development?

Or, am I completely misunderstanding the situation and have no idea how CryEngine works?

[Nauht]

I know nothing of Cryengine either but been following SC which also uses Cryengine.

I remember CR saying that there were definite limitations on the Cryengine netcode and that was a major reason they delayed the dogfighting module - to build their own backend and netcode for dogfighting. They could have released something that uses Cryengines default netcode but would have scrapped it to build their own anyway.

Mind you this is no reflection on Cryengine as the developers who work on it say that it is a powerful, and pretty game engine.

Remember that it's used as a base and is highly customisable both graphically and backend. It really is up the devs on how they want to make their game. For instance we have SC being 4K and OR compatible with their own built backend versus MWO being a minimally viable product.

[ThatBum42]

The devs have chosen to make everything actual size in MWO, and not scaled down. So when your range readout says 500m to a target, it really is 500m to the engine. As in, if you were to use a normal Crysis character, you'd need to run for 4 minutes straight or so to get there. I've heard that this presents some issues with Cryengine because it's not really designed for rendering out to these extreme distances as compared to FPSs, which have a lot smaller maps.

I dunno about that, personally. If it's good enough for Star Citizen, which takes place in space, then it should be good enough for us.

[BLOOD WOLF]

Nauht, on 06 April 2014 - 11:20 PM, said:

I know nothing of Cryengine either but been following SC which also uses Cryengine.

I remember CR saying that there were definite limitations on the Cryengine netcode and that was a major reason they delayed the dogfighting module - to build their own backend and netcode for dogfighting. They could have released something that uses Cryengines default netcode but would have scrapped it to build their own anyway.

Mind you this is no reflection on Cryengine as the developers who work on it say that it is a powerful, and pretty game engine.

Remember that it's used as a base and is highly customisable both graphically and backend. It really is up the devs on how they want to make their game. For instance we have SC being 4K and OR compatible with their own built backend versus MWO being a minimally viable product.

that VS wasnt a fair comparison. It kinda makes it look like good vs sloppy.

MWO is a minamally viable product yes, but that's the reason why it's here now and not SC. That doesnt take away from the quality of work that can be put into it. Try not to be misleading. both games can be worked on and improved. I guess you took the word minimal as a negative and it's a common mistake but that doesnt take away from the potential of a game to be developed.

interesting note, SC failed to meet their release date as well but often was used as a base for comparsion between MWO(deadlines) and SC(deadlines).

Edited by BLOOD WOLF, 06 April 2014 - 11:38 PM.

[Nauht]

BLOOD WOLF, on 06 April 2014 - 11:30 PM, said:

that VS wasnt a fair comparison. It kinda makes it look like good vs sloppy.

MWO is a minamally viable product yes, but that's the reason why it's here now and not SC. That doesnt take away from the quality of work that can be put into it. Try not to be misleading. both games can be worked on and improved. I guess you took the word minimal as a negative and it's a common mistake but that doesnt take away from the potential of a game to be developed.

You're right of course. The question then becomes how long are people willing to wait for it to be developed and how far the devs want it to go.

The point is that there are limitations to Cryengine if you use it at its minimum.

[BLOOD WOLF]

Nauht, on 06 April 2014 - 11:41 PM, said:

You're right of course. The question then becomes how long are people willing to wait for it to be developed and how far the devs want it to go.

The point is that there are limitations to Cryengine if you use it at its minimum.

Yea, but its more like starting at a minimum. Of course we know there is sooooo much more they can do. with it, they are giving us the skeleton but we want the meat. All in due time, things are progressing and moving forward. You waited this long, might as well stick around for the year.

[MeiSooHaityu]

The biggest issue with the Cryengine from what I have gathered is the netcode. The Cryengine relies heavily on the client to dictate what goes on in a multiplayer game. It makes any Cryengine game using the Cryengine multiplayer easy to hack and exploit. It ruined Crysis 1 multiplayer for me since a month after release, people were invincible running around with Nuke Tanks.

That is why PGI re-wrote the netcode from scratch (and why Star Citizen is doing it too from what I have read in this thread). No developer in their right mind would use the in-engine multiplayer from the Cyengine.

Games like MWO need to keep as much of the game dictated on the server side as possible and rely on the client for almost nothing. This is why we have the hit registration issues that we do (missile and ballistic false hits, light mech false position reporting, and the occasional rubber-banding). It is a necessary evil though to keep cheating to a minimum.

TL;DR, In anycase, if the Cryengine has one huge flaw, it is forcing developers to write new netcode from scratch.

[Cromwill]

BLOOD WOLF, on 06 April 2014 - 11:30 PM, said:

that VS wasnt a fair comparison. It kinda makes it look like good vs sloppy.

MWO is a minamally viable product yes, but that's the reason why it's here now and not SC. That doesnt take away from the quality of work that can be put into it. Try not to be misleading. both games can be worked on and improved. I guess you took the word minimal as a negative and it's a common mistake but that doesnt take away from the potential of a game to be developed.

interesting note, SC failed to meet their release date as well but often was used as a base for comparsion between MWO(deadlines) and SC(deadlines).

Okay, let me clarify the difference for you since i have invested heavily in BOTH games (i have nearly every hero mech) and I have en emotional attachment going back decades with MWO that I do not have with SC. That said the developers, how they act, interact with the community, and share information is vastly different.

Yes, SC delayed launch, but they also have press releases EVERY DAY of some kind, one is able to VISIBLE see progress and watch he developers play the dog fighting module. There is therefore, both daily interaction and progress that players can see.
There is honesty and clarity in their decisions, while MWO has most certainly but duplicitous in its dealings with the community at times. One generally has no clear idea when and what is coming with the patches, the monthly updates used to be posted halfway through the month, and then unable to meet that they moved to a general outline of features to expect in the future thus releasing them from meeting or failing to meet deadline after deadline, feature after feature.

I do not need to tick off a litany of MWO's sins, but the primary difference is in how the developers treat the community and their willings to take feedback. Just go look at SC homepage, and look at the level of engagement.

I want MWO to succeed and I think this could have been a full-blown MMO with any vision, but the implementation of MWO development is an abomination to anyone who has worked in the professional sector and understands what the process should look like. This is amateur hour and there is no way around that fact.

Edited by Cromwill, 08 April 2014 - 09:45 AM.

[hideyourkids]

Cryengine isn't "holding us back," incompetent devs are

[DragonsFire]

hideyourkids, on 08 April 2014 - 10:21 AM, said:

Cryengine isn't "holding us back," incompetent devs are

When both PGI and CIG have had to write their own netcode due to the client authoritative nature of the native Cryengine netcode, it does indeed come down to a lack of functionality with the engine.

You may not agree with the direction MWO has taken, or the pace it has proceeded at, but if you've bothered to check out something like Karl Berg's presentation on code complexity in asynchronous distributed server architectures, you'll find that many of the devs are far from incompetent.

[Hans Von Lohman]

hideyourkids, on 08 April 2014 - 10:21 AM, said:

Cryengine isn't "holding us back," incompetent devs are

Then leave. You're so much better at programing than they are, clearly. Why would you still want to be here?

[zagibu]

DragonsFire, on 08 April 2014 - 10:36 AM, said:

When both PGI and CIG have had to write their own netcode due to the client authoritative nature of the native Cryengine netcode, it does indeed come down to a lack of functionality with the engine.

So why did they choose an engine with useless netcode?

[Helsbane]

zagibu, on 08 April 2014 - 11:54 AM, said:

So why did they choose an engine with useless netcode?

Because $. The apparent front end savings of going with Cryengine and spreading the learning curve over the span of development. This coupled with the option to NOT take on a specialist is why we're here today with such a limited game. Not taking anything away from the effort put forth by PGI here, just saying that you learn faster from a teacher than from muddling about on your own.

[DragonsFire]

zagibu, on 08 April 2014 - 11:54 AM, said:

So why did they choose an engine with useless netcode?

I would wager that there were financial//business directives that motivated the choice, much as it motivated Chris Roberts to make the same choice in engine. Unfortunately the best I can do is guess at what those directives are.

In the end, hindsight is 20/20 as they say, and looking back it's entirely possible that they would have picked a different engine. Then again, they may still have stuck with this one anyways. I just know that speaking with multiple developers who have actually worked with CryEngine that it is indeed quite laborious and finicky in many cases.

[Odins Fist]

And the clock gets louder... Tick Tock... Tick Tock.... Tick Tock...

[AdamBaines]

DragonsFire, on 08 April 2014 - 10:36 AM, said:

When both PGI and CIG have had to write their own netcode due to the client authoritative nature of the native Cryengine netcode, it does indeed come down to a lack of functionality with the engine.

You may not agree with the direction MWO has taken, or the pace it has proceeded at, but if you've bothered to check out something like Karl Berg's presentation on code complexity in asynchronous distributed server architectures, you'll find that many of the devs are far from incompetent.

Dont bother with the troll. Not worth it.

[MeiSooHaityu]

zagibu, on 08 April 2014 - 11:54 AM, said:

So why did they choose an engine with useless netcode?

Like Dragonsfire stated, probably money.

I will say this though, the engine itself (minus the net code portion) always seemed fairly flexible and somewhat impressive. The early Cryengine like Crysis 1 was a resource hog, but looked amazing for the time. I think the newer itterations are better on resources yet still can do some amazing things and is probably pretty flexible.

Also, a game like MWO is so multiplayer dependant, regardless of what engine they went with, they might have had to write new netcode regardless *shrug*.

Edited by MeiSooHaityu, 09 April 2014 - 08:17 AM.

[Mirkk Defwode]

AUSwarrior24, on 06 April 2014 - 07:26 PM, said:

Hold the show; I'm certainly not suggesting they migrate engine, if that's what you think I might be meaning. I'm simply asking, simply because I don't know much about it; is CryEngine holding MWO back in any areas?

I'm always paranoid of premade engines like this because they can be tricky to generate games other than their original 'show' game. Now, MW:O has done a pretty good job in bringing that MechWarrior experience to such an engine. But in terms of hit detection errors and other problems, do you think not working with their own engine is slowing down their development?

Or, am I completely misunderstanding the situation and have no idea how CryEngine works?

Go look at Mechwarrior Living Legends - it was a mod for Crysis that expanded on the base engine and demonstrated the flexibility in this engine.

Though I will say that on the surface this engine is really pretty, under the hood is somewhat of a mess at times - but workable - just takes some time and patience to really kick it in gear.

ThatBum42, on 06 April 2014 - 11:28 PM, said:

The devs have chosen to make everything actual size in MWO, and not scaled down. So when your range readout says 500m to a target, it really is 500m to the engine. As in, if you were to use a normal Crysis character, you'd need to run for 4 minutes straight or so to get there. I've heard that this presents some issues with Cryengine because it's not really designed for rendering out to these extreme distances as compared to FPSs, which have a lot smaller maps.

I dunno about that, personally. If it's good enough for Star Citizen, which takes place in space, then it should be good enough for us.

This is sorta true for the original CryEngine and 1.5 - it's made some improvements. It's more imposters loading at extreme distances that can prove problematic - which that's true for most engines today.

There are only a few engines I think could do this "better" but I don't know the cost associated with those. I can say Ego Engine (Operation Flashpoint Dragon Rising, Forza 2/3/4) can look as pretty and has typically better physics built in, but shoddy netcode like this. UnrealEngine3 (Gears of War 3, Unreal Tournament 3, Red Orchestra 2) is also pretty but it doesn't handle textures as well so a lot of stuff looks like clay in it at times. Part of that could be art and lighting too. But stylized graphics tend to do better for that engine rather than attempting photo realism.

Edited by Mirkk Defwode, 09 April 2014 - 09:36 AM.

[Mechteric]

CryEngine may not be perfect, but as I recall I think the cost for licensing CryEngine versus Unreal engine is much cheaper and simply made much more sense for a Free 2 Play game.

Edited by CapperDeluxe, 09 April 2014 - 10:00 AM.

[Spr1ggan]

ThatBum42, on 06 April 2014 - 11:28 PM, said:

The devs have chosen to make everything actual size in MWO, and not scaled down. So when your range readout says 500m to a target, it really is 500m to the engine. As in, if you were to use a normal Crysis character, you'd need to run for 4 minutes straight or so to get there. I've heard that this presents some issues with Cryengine because it's not really designed for rendering out to these extreme distances as compared to FPSs, which have a lot smaller maps.

I dunno about that, personally. If it's good enough for Star Citizen, which takes place in space, then it should be good enough for us.

The maps in MWO are not large. None of them. There are plenty of FPS games with bigger maps than the ones in MWO.