TSW anyone?

[Bananarama]

"Choices"

I have a new desktop image.

[bpetit]

"Choices"

I have a new desktop image.

It is also serving as my cover for my facebook page.

[NorthernWarrior]

This postulation from MaikG at UKTS makes interesting reading: http://forums.uktrainsim.com/viewtop...47175&start=30 scroll down Page 3.
That kind of implies choosing UE4 was always going to bottleneck a simulation project if the graphics are primarily geared to the CPU, not GPU. If so, DTG have built themselves into a bottleneck it may not be easy to code themselves out of.

The observations borne out by my last experience (not played it since, too busy enjoying Run8) when shortly after adjusting the tree quality upward, the physics went bananas and my previous smooth progress up the hill turned into a stall.

[_o_OOOO_oo-Kanawha]

There has been more discussion on this and it seems EPIC, the owners of UE4, need to move their gaming engine forwards towards automatic and better utilisation of multi core processors, and integrate DX12 into the main build.

There are also custem API that can be used for non-core graphics stuff, but these need to be licensed and integrated into TSW and will make the final outcome dependent on yet more parties.

It is getting very complicated if you aren't into game developing at a professional level.

MaikG postulates that any significant advance in fps can only be achieved by cutting back on the physics.

I for one would be glad if the game can play on constant 30 fps without lag or funny acting physics.

[NorthernWarrior]

I will take better/proper physics over the eye candy any day! Problem is, from their marketing POV it was sold on the massive improvement in graphics over the Railworks engine.

So... egg on face if they concede UE4 doesn't cut it after all and switch to some sort of Railworks+ engine.

Egg on face if they plough on but cannot achieve the level of physics fidelity needed to start catching the attention of Run8 etc. users or make it appealing to the more discerning third party DLC producers so they get their kickback on royalties.

Basically, egg on face.

[_o_OOOO_oo-Kanawha]

More from MaikG of virtualRailroads on PhysX not being of much use with UE4 and the Simugraph physics simulation over at UKTS: http://forums.uktrainsim.com/viewtopic.php?f=386&t=147175&st=0&sk=t&sd=a&start=45#p1813243

Is seems DTG has painted itself into a corner with its choice of game architecture for their ambitious plans the deliver the railroad simulation to end all other railroad simulations.

More performance at the present can only come from further rationalising the assets, and optimising/simplifying the physics.

That is even before catenary, moving vehicles, pedestrians and other goodies make their debut, like authentic steam locomotives.

Perhaps not all is lost when Epic delivers a fully DX12 version of UE4 that can indeed make better use of multi-core CPU's and better utilise the raw power of GPU's.

[imnew]

Interesting reading material. Got me a bit worried though.. I hope DTG finds a solution. Thanks for putting up the link Edwin

[Derek]

Every engine has it's restrictions and we've done our due diligence.
It's a brave new world but it's an adventure.

All will be good, still got both feet and not stuck in a corner.
:)

d

[_o_OOOO_oo-Kanawha]

Of course it's a big adventure. I doubt any game studio has ever started an open world franchise of this ambition, size, scope and complexity using the UE4.

Choices were made like 2 years ago at the onset of TSW, with the programs, tools, API, etc. available at that moment and one has to make do with them during the initial development towards the current beta state.

But technology advances, hardware at a slower pace also.
Is UE4 a progressive, open architecture, so advances like DX12 or a different physics emulation can easily be added in a future release of TSW without breaking DLC?

The key question still remains unanswered: will TSW turn into a game or into a simulation .....

So far I like what I see in TSW, a few personal quibbles aside. I demand it to be more than a pretty version of TSX before I part with more money towards it. Many of the simulation and operation aspects are indeed a step forwards, others like the "9" screen not so much.

[Derek]

We are delivering a great experience that is engaging and entertaining.

The question you ask yourself is subjective.

d

[buzz456]

We are delivering a great experience that is engaging and entertaining.

The question you ask yourself is subjective.

d

Indeed and no matter what you do some (and you know who you are I hope) will be complaining.

[NStrains]

We are delivering a great experience that is engaging and entertaining.

The question you ask yourself is subjective.

d

Indeed and no matter what you do some (and you know who you are I hope) will be complaining.

"Can't please everyone all the time!"

[NorthernWarrior]

We will have to take Derek at his word on that, but I do feel DTG need to acknowledge outside a very small group there is considerable disappointment at what has been delivered with the initial release of TSW. It really wasn't quite ready and with hindsight would it not have been better to release on an early access basis - "This is where we are, not quite what we want but with wider feedback and testing we'll get it there"? Certainly the beta test should have been for longer and less restricted access, maybe issues like the physics wouldn't have slipped so dramatically through the net.

As noted previously I didn't refund because I do have (some) hope that a phoenix may yet emerge, not expecting it to be up there with Run8 but at least passable where train performance is concerned. Can the dev team truly resolve the merging of a complex physics application with the UE4 graphics engine, after all the time already spent trying to make it work?

When reading on UKTS the response people have had from Support about only one save slot (that's all there will ever be) does not garner huge optimism - if the little things can't be attended to, then what hope for the principal items? Don't forget some of us already went through this with TANE - over promised then under delivered and only after two years and numerous patches/service packs is it starting to approach what N3V actually set out to deliver.

Not whinging or being negative here, please, just asking the hard questions a paying customer needs to put to the vendor.

[JPS2K5]

I will avoid this like the plague, untill I can control a train with a HOTAS or Throttle Quadrant. 'Simulations' with a keyboard or Xbox controller are not my thing.

[Bananarama]

I will avoid this like the plague, untill I can control a train with a HOTAS or Throttle Quadrant. 'Simulations' with a keyboard or Xbox controller are not my thing.

All we need is the ability to keymap - why something so basic wasn't in the initial release is odd.