• Announcements

    • HEAVY265

      Attention Soldiers Operation Fury Needs you!   02/20/2020

      Attention All Soldiers, Operation Fury needs you.  You need to choose a side and sign up.  
      For more intel on Operation Fury Please click HERE Please go to Special Event Forum (here), And sign up for allied or axis.
      This will be a CRS Lead event on both sides.  Xoom will be heading up the axis side and Heavy265 will be heading up the Allied side. This will be for bragging rights.
      Why are we asking players to sign up you ask. We are trying for a role play experience.   We want this to be a true realistic event.  
      So get up and sign up and let's make this the best event ever!!!!!!!!!!
      Give me your war cry, grrrrrrrrrrrrr
      Heavy265 **out**
sublimesw

Unreal 3 / 4 Engine?

98 posts in this topic

This thread started in a interesting way.... now it seems the definitive FTP negative posters Club.:popcorn:

 

 

Share this post


Link to post
Share on other sites
Quote

4. Decision point: Game engine selection, what is most valuable to you? (Assume our team can only do "ONE" effectively)

"CURRENT" GAME ENGINE: Upgrading the technology of our existing game engine

(Includes, and is not limited to: Roadmap completion (1.36, 64-bit, VOIP, content), New tech like SpeedTree, Multi-threading, bug fixes, performance improvements, combined arms game play and scale guaranteed, continued production speed and quality, retains fidelity in models / physics)

Current game engine would remain operational, standard patching as you're familiar with now.
"NEW" GAME ENGINE: Using another game engine, such as UE4 to conduct ground-up development

(Includes, and is not limited to: Places heavy resources to start game development over, rebuild dated technologies, 2+ (or more) years until product released, new modern game models, new animations, faster production speeds, larger work force, export to more platforms - huge scale is not currently guaranteed, but possible in theory according to our senior developers, several existing systems would be required to be ported over, latest version of Direct X / SpeedTree / other technologies, leverage marketplace assets which include dynamic weather patterns and realistic lighting and texturing techniques, characters entering and exiting vehicles, possibility of inventory system)

Current game engine would remain operational during this stage but would receive limited development to emphasize development on new engine.

What is worrying is the caveat that with the UE4 "huge scale is not currently guaranteed...."

Edited by agenda21

Share this post


Link to post
Share on other sites

Huge scale isn’t needed, at least to start with. A more focused approach will allow for greater realism and fidelity in map design. And they didn’t say small, only that it wouldn’t be as big as now let’s just say. To be honest, do we really even use a lot of the terrain we have now? I’d much prefer a smaller, more detailed, 1:1 scale terrain than what we have now. 

Share this post


Link to post
Share on other sites
4 hours ago, raptor34 said:

To be honest, do we really even use a lot of the terrain we have now?

You use everything that has a town in it, kind of have to in order to win the campaign?

Eventually you are going to have to use even more

Share this post


Link to post
Share on other sites
10 hours ago, raptor34 said:

Huge scale isn’t needed, at least to start with. A more focused approach will allow for greater realism and fidelity in map design. And they didn’t say small, only that it wouldn’t be as big as now let’s just say. To be honest, do we really even use a lot of the terrain we have now? I’d much prefer a smaller, more detailed, 1:1 scale terrain than what we have now. 

As big as now? The map we have covers only a fraction of WW2. Scale is the only thingh that stands out. If they go the other way, they compete with all the other games.

Share this post


Link to post
Share on other sites
6 hours ago, ebert100 said:

As big as now? The map we have covers only a fraction of WW2. Scale is the only thingh that stands out. If they go the other way, they compete with all the other games.

WW2OL does compete with other games/sims in the market, just maybe not one for one given its history and design. More to the point, yes, I agree, the map we currently have only covers a small area of WW2 but if you made it larger without giving it the historical content (ToE) that should be included with it are you gaining anything? Even out 1940 Battle of France campaign is missing equipment that should be represented within our current theatre of war. There is potential for new theatres with most of our current ToE, (Limited North Africa) but I guess that I am trying to say is that I feel expanding the map with no plan to accompany it would not add anything to the experience or realism.

Also, I tend to disagree with players who think WW2OL is just about the big map. For me, it is the combined arms simulation that is a the heart of WW2OL. When I say "Huge scale isn't needed, at least to start with it" I don't mean 2km by 2km maps, and I don't believe the RATs mean that either. I'm simply arguing that a smaller, more detailed map with a new engine wouldn't destroy WW2OL, because it's more than just the mostly empty map. I think there is a real opportunity with 2.0 to move this project forward.

 

Share this post


Link to post
Share on other sites

We're doing several investigations right now and we understand the scale component being a large part of what makes WWII Online unique.

We know for sure, out of the box, we can obtain 20km x 20km maps at a minimum, and we believe (and are attempting to validate) we can work up from there. 20km x 20km is two supercells (or the squares when you turn on AWS). That's a lot of play space to work with. We know we can create tiles like that and get them to load as you move through the world, the key is to do that seamlessly, preferably to avoid having to despawn ~ like we do now.

I think it would be safe to say if we could validate we can have a significantly large scale, people would be all over UE4 as the solution.

The good news is, several of our technologies that we currently have, are capable of being added to UE4 because we have access to their source code, and of course we have our own that we could either replicate (because we have the understanding of how it is achieved) or work to port it over.

6 people like this

Share this post


Link to post
Share on other sites
On 10/8/2018 at 3:24 AM, Nick said:

Both Unity and Unreal use PhysX for their physics but they limit what parameters you can change for the physics objects which is why games made with them tend to 'feel' the same.

The workaround is to mesh servers... host a 50 here, 50 there, as they get closer, mesh them so they see each other. Sea of Thieves and Star Citizen do this, and it's really a good solution to reduce server stress and improve the player experience. Way less choppy.

Imagine a population based set of servers that grows as it needs to  and just sort of poops you around the alternate server universes, but you have no loading, nothing to tell you you've changed servers. Also, SC had a container system that only loads your immediate area as you need it, using your own files to flesh it out as your pov is around. 

Saves a ton of resources. 

 

Things ARE coming, but the game companies that are really pushing the envelope are BIG money games. Record-holders at fund raising.

 

If WWIIOL could ever turn out something in either Unity or UE4, they'd have dumptrucks of money being backed up to their doors. Air, ground and sea... on a current and growing platform. it's gotta happen sooner or later.

Share this post


Link to post
Share on other sites
3 hours ago, madrebel said:

ww2ol has had this since 2001.

No WW2ol has load balance, server meshing uses the cloud to abstract the server location and instead pairs a group of players to the ideal server location for them.

Share this post


Link to post
Share on other sites
10 minutes ago, pbveteran said:

No WW2ol has load balance, server meshing uses the cloud to abstract the server location and instead pairs a group of players to the ideal server location for them.

load balancing, meshing, you're splitting hairs.

 

the backend cell hosts can do all that since 2001. there is no reason too right now as you'd have individual cell hosts with a single player on them. further, latency only improves when you've got groups of players on a single cell host all geo-located within 50ms of each other and the cell hosts. cell hosts in say the EU with players interacting with players on a cell in texas would still experience similar latency issues as they do now. you 'might' shave off a few milliseconds via the cell to cell connection being on better infrastructure vs EU client to US cell host to US client. a us cell talking to a eu cell would have a few MS advantage, not huge, but a little.

ideally you'd want to drive players into cells based on region then have AOs specifically labelled as 'EU AO' so that players know "hey, if i want to play infantry from the US i'm going to be at a disadvantage to the other EU players in this AO".

 

that's all fairly simple however - why? a single cell host for the most part handles all the current load and establishing a colo footprint or cloud instances costs money.

 

now do note, there would need to be some changes/optimizations here and there as the cloud was born long after ww2ol had only a skeleton crew. the underlying design though could pretty easily run in say AWS or GCP or etc etc etc. as long as the links/networks are there, strat updates would still flow between cell hosts and the central DBs in texas and things would just work. if we had the population that required it, this game can scale pretty well considering. how do you think thousands of players could all be in the same world back in the early 2000s?

 

do you even understand how ahead of the rest of the world the back end networking in this game really was? if you did you wouldn't be talking about meshing servers as you'd know - its already there. has been since 2001 where a rack full of single core dual socket single core, pentiumD based xeon's, 12GB of RAM, and 100mbit networking managed to support thousands of players simultaneously. If you weren't aware, that same server struggled to handle 32 players in quake, tribes, UT, etc any contemporary shooter struggled on that same hardware.

 

plenty of things wrong with the engine, networking too has issues (hence the recent update) but don't let that fool you. there is some genius in that design and it was done well before anyone else tried it.

Share this post


Link to post
Share on other sites
On 10/19/2018 at 4:00 AM, kase250 said:

This thread started in a interesting way.... now it seems the definitive FTP negative posters Club.:popcorn:

 

 

Oh, are you referring to the only players left here?

Share this post


Link to post
Share on other sites

Maybe I should have said "well-functioning server meshing" to be more clear.

 

This thread went all over the crazy patch. We even saw Jwilly stating that crews getting out of vehicles was "unrealistic".

 

Hilarious.

 

 

The original crew didn't model getting into and out of vehicles as a shortcut. They didn't have the resources and for a long time, nobody cared. Now... 17 years later, the chit is played out. 

Being stuck in a vehicle is dumb, and not realistic.

Share this post


Link to post
Share on other sites
Quote

 

The original crew didn't model getting into and out of vehicles as a shortcut. 

 

Incorrect. It was an intentional CRS design decision, since the intended marketing was toward realism.

While a tank crew obviously could and did exit their disabled tank and run for their lives, a tank crewman getting out and becoming an infantryman, in the general case, would be unrealistic. So, that wasn't made part of the code.

Quote

 

This thread went all over the crazy patch. We even saw Jwilly stating that crews getting out of vehicles was "unrealistic".

 

Getting out of a tank to become infantrymen.

Quote

 

Being stuck in a vehicle is dumb, and not realistic.

 

Getting out of a vehicle to take a leak, or go have dinner, or because you just got stuck or had a wheel blown off and you'd rather run for cover than stay there and die, is quite realistic. It however is not an action that needs to be modeled as part of a combat game.

A tanker jumping out to become an infantryman is not realistic, which...it was said by CRS in the early days...is why it's not part of this game. 

Share this post


Link to post
Share on other sites

@jwilly the problem is that in 20 years other FPS games have set a few standard that might be unrealistic, but that nowadays accounts for the fun (non-frustruting) part of a game.

If we had vehicles and weapons restrictions based on unit category experience and not the rank, that enter-exit might be considered

 

Share this post


Link to post
Share on other sites

Current CRS management decides how to market their game. In earlier times CRS said that other (mostly bigger, mostly much-larger-marketing-budget) FPS games had soaked up the customers who don't care about realism or don't have a correct understanding of it, and greater realism seemed to be CRS's opportunity-niche. 

Share this post


Link to post
Share on other sites
On 10/27/2018 at 0:37 AM, vasduten1 said:

This thread went all over the crazy patch. We even saw Jwilly stating that crews getting out of vehicles was "unrealistic".

That is not what he said.
He said in effect they don't bail out and become an infantry assault force.
They are not even remotely equipped to do so really.

They may very well bail out, and then try to make their way back to their unit to get a new tank etc.
They may also very well likely be raped by enemy fire while trying to exit.

meanwhile, the vehicle they bailed from is still lost from combat.
You can despawn the vehicle, in effect bailing without being shot to death, or having to walk 3km back to the FB to see if there is another tank available.

It isn't a feature people would get much mileage out of once implemented.
They be standing on the ground with limited means to defend themselves probably already shot to death, they wouldn't be able to just run around trying to GTA themselves into someone else's vehicle, they would already be scored as "KILLED" due to loss of their vehicle and they wouldnt be doing the BF42 shuffle of rolling up in a tank, jumping out, capping a flag, jumping in a truck and taking off, bailing out of the truck running to another flag, machine gunning down 7 enemy, capping that flag, then hopping in an airplane, etc etc.

Most would probably decide to despawn after finding themselves sitting there in the midst of a firefight holding a pistol and 12 rounds of ammo and 3km away from the nearest next tank.

 

 

Share this post


Link to post
Share on other sites
4 hours ago, jwilly said:

Current CRS management decides how to market their game. In earlier times CRS said that other (mostly bigger, mostly much-larger-marketing-budget) FPS games had soaked up the customers who don't care about realism or don't have a correct understanding of it, and greater realism seemed to be CRS's opportunity-niche. 

It’s too bad they have been passed by leaps and bounds in the realism niche. Other games seem to have working HE (something WWIIOL has never really had), a better overall damage model, and much more realistic infantry play. Not to mention the potential for friendly fire on the battlefield. In addition more realistic terrain provides for more realistic fire fights.  In all WWIIOL is now getting curb stomped in realism by the competition. 

Share this post


Link to post
Share on other sites

In fairness, other games actually don't have working HE at all as far as a simulation goes.
Not without good reason, doing an actual physical simulation of HE is a pain in the arse to do, both from a coding stand point and from a client processing standpoint.

Our HE, as it stands right now, will actually work very well, and as far as hitting things IS doing very well as far as the HE itself goes.
The 2 legged species are the actual thing at fault presently, they were made initially in a way that they would only interact with very large frags..
They simply don't interact with frags that are small but fatal, the frags do hit them but they take little to nothing from it.
That is being worked out, and when it is fixed i think people will learn to do a lot less running around and a lot more kissing the ground when things are exploding
and if you cant hug the ground, you may have to just be content with dying.

Even the initial HE was good, as far as a design theory goes, but initially the hardware available was simply incapable of running the simulation in full without resulting in single digit FPS, so it was scaled down to what the average game PC could process acceptably, which meant you could not calculate 200 frags let alone 2000, so instead of 200+ frags of various sizes you got a much much lesser amount of frags at a very large size.

That caused 2 problems
1) Coverage, very easy to get missed with such a low frag count
2) Infantry reaction, they got designed to interact with these very large frags and dont know to get hurt from smaller ones.

Coverage is increased now, very much so, and has the potential to increase more (where it is valid to be more) in the future with additional client work.

Infantry will learn to die properly, it is being worked on.

The concussive component is there, and it does function, so it can eventually be evaluated to see how it can more properly function also.

Though it is not terribly exciting, the offline clones have a nifty feature that if hit by frags, they simply die, they are not able to ignore being hit
So you can actually set them up offline in a building for example, and pop a grenade through the window, and then find where you would have live guys even before the HE rebuild, you now have dead guys.
Once infantry type objects get their update, it should be more like what you see with the clones.

 

Friendlyfire does not happen here by design choice, not inability to have it.
The game is very large, at times can be very confusing, and results can be permanent for the duration of a campaign.
Firendlyfire would most likely result in a never ending stream of accidents, and a large amount of ill feelings.
Even now, accidentally killing friendlies by blowing a depot can result in a lot of unhappiness and raised tempers
 

1 person likes this

Share this post


Link to post
Share on other sites
17 hours ago, Merlin51 said:

In fairness, other games actually don't have working HE at all as far as a simulation goes.
Not without good reason, doing an actual physical simulation of HE is a pain in the arse to do, both from a coding stand point and from a client processing standpoint.

Our HE, as it stands right now, will actually work very well, and as far as hitting things IS doing very well as far as the HE itself goes.
The 2 legged species are the actual thing at fault presently, they were made initially in a way that they would only interact with very large frags..
They simply don't interact with frags that are small but fatal, the frags do hit them but they take little to nothing from it.
That is being worked out, and when it is fixed i think people will learn to do a lot less running around and a lot more kissing the ground when things are exploding
and if you cant hug the ground, you may have to just be content with dying.

Even the initial HE was good, as far as a design theory goes, but initially the hardware available was simply incapable of running the simulation in full without resulting in single digit FPS, so it was scaled down to what the average game PC could process acceptably, which meant you could not calculate 200 frags let alone 2000, so instead of 200+ frags of various sizes you got a much much lesser amount of frags at a very large size.

That caused 2 problems
1) Coverage, very easy to get missed with such a low frag count
2) Infantry reaction, they got designed to interact with these very large frags and dont know to get hurt from smaller ones.

Coverage is increased now, very much so, and has the potential to increase more (where it is valid to be more) in the future with additional client work.

Infantry will learn to die properly, it is being worked on.

The concussive component is there, and it does function, so it can eventually be evaluated to see how it can more properly function also.

Though it is not terribly exciting, the offline clones have a nifty feature that if hit by frags, they simply die, they are not able to ignore being hit
So you can actually set them up offline in a building for example, and pop a grenade through the window, and then find where you would have live guys even before the HE rebuild, you now have dead guys.
Once infantry type objects get their update, it should be more like what you see with the clones.

 

Friendlyfire does not happen here by design choice, not inability to have it.
The game is very large, at times can be very confusing, and results can be permanent for the duration of a campaign.
Firendlyfire would most likely result in a never ending stream of accidents, and a large amount of ill feelings.
Even now, accidentally killing friendlies by blowing a depot can result in a lot of unhappiness and raised tempers
 

This game is 17 years old. There is no excuse at this point for non working HE and poor damage modeling. 

Stop touting the game as the most realistic game like evAr if you can’t even implement friendly fire. It’s a constant defense of unrealism while talking about how real the simulation is. Ridiculous. 

Share this post


Link to post
Share on other sites

Actually friendly fire is easy to implement and during one early intermission in I think 2001 or 2002 it was (didn't end too well). The difference between this game and others is the way it implements damage but physically modelling fragments or projectiles, rather than HE in other games which is statistically/hitpoint based (i.e. you are X meters from a bomb = you always die, whereas in WWIIOL you could be 25 m from a bomb and miraculously survive, while another guy at 50 m gets hit by shrapnel and dies, which is something you would never see in other games). Could it be better? Sure. But I think with the roadmap and the ratchat its clear that progress if being made, which is why old time players like me who have just been lurking around and checking in on the website every 6 months or so are starting to post on the forums.

Share this post


Link to post
Share on other sites
On 11/4/2018 at 6:39 PM, Merlin51 said:

That is not what he said.
He said in effect they don't bail out and become an infantry assault force.
They are not even remotely equipped to do so really.

They may very well bail out, and then try to make their way back to their unit to get a new tank etc.
They may also very well likely be raped by enemy fire while trying to exit.

meanwhile, the vehicle they bailed from is still lost from combat.
You can despawn the vehicle, in effect bailing without being shot to death, or having to walk 3km back to the FB to see if there is another tank available.

It isn't a feature people would get much mileage out of once implemented.
They be standing on the ground with limited means to defend themselves probably already shot to death, they wouldn't be able to just run around trying to GTA themselves into someone else's vehicle, they would already be scored as "KILLED" due to loss of their vehicle and they wouldnt be doing the BF42 shuffle of rolling up in a tank, jumping out, capping a flag, jumping in a truck and taking off, bailing out of the truck running to another flag, machine gunning down 7 enemy, capping that flag, then hopping in an airplane, etc etc.

Most would probably decide to despawn after finding themselves sitting there in the midst of a firefight holding a pistol and 12 rounds of ammo and 3km away from the nearest next tank.

 

 

Next step would be pilots bailing out over towns then capping.

 

Cause pilots often did that in WWII.

Share this post


Link to post
Share on other sites
4 hours ago, minky said:

This game is 17 years old. There is no excuse at this point for non working HE and poor damage modeling. 

Stop touting the game as the most realistic game like evAr if you can’t even implement friendly fire. It’s a constant defense of unrealism while talking about how real the simulation is. Ridiculous. 

Actually given the childish take advantage of coding of a lot of game players, I'm VERY good with the current state of no friendly fire.

 

The one reality buster for most wargamer and actual in combat army types is the lack of artillery.  Forget camping tanks, it was the big guns that did most of the killing.

 

Share this post


Link to post
Share on other sites

  • Recently Browsing   0 members

    No registered users viewing this page.