Build On Same Route Instead Of Making Separate Routes

Discussion in 'Suggestions' started by DANNYBOY2487, Jan 13, 2018.

  1. DANNYBOY2487

    DANNYBOY2487 Active Member

    Joined:
    Dec 19, 2017
    Messages:
    144
    Likes Received:
    98
    what i mean is like if Reading to Oxford was made instead of making it as a separate route have it join to the currant route and so on so on with other destinations.
     
    • Like Like x 1
  2. pschlik

    pschlik Well-Known Member

    Joined:
    Dec 8, 2016
    Messages:
    875
    Likes Received:
    1,607
    I get the timpression that DTG has this in mind. Matt did mention (looong ago) that he is interested in a much better way of expanding routes than the awkward "support" of the feature present in TS2018. But it was only hypothetical, as any theoretical expansion is probably not what DTG would do right now...
     
    • Like Like x 2
  3. cActUsjUiCe

    cActUsjUiCe Developer

    Joined:
    Sep 21, 2017
    Messages:
    681
    Likes Received:
    2,179
    Add it yourself once the editor is released
     
  4. Drakoz

    Drakoz Active Member

    Joined:
    Jan 29, 2018
    Messages:
    152
    Likes Received:
    147
    The key factor is TSW needs some way to make the transition from one area seamlessly to a new area without having to exit the current scenario and start a new one. That breaks the realism too much. It should maintain the same train conditions and settings (speed, weather, other traffic on the rails, etc.) and just transition to the new area.

    Yes, they could just pause the sim and do a blank screen load or something like that to load the new area, but it's pretty common now for game engines to handle huge open space environments, loading the entire region or as needed on the fly. I hope the TSW graphics engine can handle that. But I'd be OK with a pause while the new region is loaded as long as the regions overlap just enough that it is seamless otherwise.
     
  5. pschlik

    pschlik Well-Known Member

    Joined:
    Dec 8, 2016
    Messages:
    875
    Likes Received:
    1,607
    I think the idea is to have extensions just add on to the game world...not to be separate routes or even separate loading zones. What you have in mind is worse than what TS2018 does!
     
  6. Drakoz

    Drakoz Active Member

    Joined:
    Jan 29, 2018
    Messages:
    152
    Likes Received:
    147
    Agreed. I"m not advocating separate loading zones because that would suck. Most games today no longer do that, but the game engine has to be designed from the start to be an open world. If DTG didn't do that, then this feature is not likely to be implemented. Or would require loading zones - if they can do it in a seamless way. I didn't know they tried this in TS2018.
     
    • Like Like x 1
  7. pschlik

    pschlik Well-Known Member

    Joined:
    Dec 8, 2016
    Messages:
    875
    Likes Received:
    1,607
    Actually, one of the cool things about Unreal 4 is that it supports world streaming as opposed to loading zones, so it can seamlessly load in and out scenery without a loading screen-and that's a good thing for Train Sim! Basically, that is why a 1000 mile route would work...it just isn't going to happen because it would take way too long to do that. But with extensions, a bunch of routes could connect into one huge route of that length.
     
    • Like Like x 3
  8. Drakoz

    Drakoz Active Member

    Joined:
    Jan 29, 2018
    Messages:
    152
    Likes Received:
    147
    Great news. And as for 1000 mile routes - I'm an iron butt. When I sit down to "go somewhere" I don't like to stop for hours (in real life or sim). But of course, I'd be happy with a save game to break up the coast to coast trips a little bit. :D As if we'll be so lucky to get that kind of route some day.
     
  9. Corvan

    Corvan Well-Known Member

    Joined:
    Sep 23, 2017
    Messages:
    527
    Likes Received:
    734
    I like that idea. Add in the London to Brighton mainline and then do the line between Reading and Redhill. Gatwick Airport here I come. :D
     
    • Like Like x 1
  10. SamYeager270

    SamYeager270 Well-Known Member

    Joined:
    Dec 8, 2016
    Messages:
    633
    Likes Received:
    667
    My worry is that as it stands TSW would grind to a halt in such a case because of the amount of services running in the background. Now I like the services running because it adds presence to TSW but the trouble is that DTG haven't found a way to stop those services slugging the PC. I'm assuming that this is due to the need for services to be fully featured so that users can join services at will.

    Ideally DTG would load in lower impact services in areas where the user character is not present to reduce the load i.e. fully featured services in areas where user character is on foot or stopped at a station, detailed exteriors and fewer featured services where the user character will be passing by in another service and lower featured services and less detailed exteriors elsewhere. Having just typed that I can see that multiplayer might make that idea considerably less feasible. :(

    Perhaps we're back to the idea that DTG need to find an overall way of reducing the impact of services on the PC without losing the existing realism and better graphical quality. I have a horrible feeling that that may only only be possible by re-architecting TSW which is not a trivial matter. :(
     
  11. pschlik

    pschlik Well-Known Member

    Joined:
    Dec 8, 2016
    Messages:
    875
    Likes Received:
    1,607
    You realized that AI 1: take basically no CPU, and only do anything to eprformance when they are nearby, and that 2: longer routes don’t exactly mean more trains, just the same trains staying in the game world longer which doesn’t matter because 3: multithreading SimuGraph (AKA re-architecting TSW) is actually something that has been brought up before. Not sure in what order it would all happen though.
     
    • Helpful Helpful x 1
  12. SamYeager270

    SamYeager270 Well-Known Member

    Joined:
    Dec 8, 2016
    Messages:
    633
    Likes Received:
    667
    pschlik I'm not convinced that your point 1 is necessarily true unless you can show evidence to prove it. Considering TSW seems to use up a fair amount of processor time even when I'm not moving and there are no trains near me would seem to imply the contrary. A longer line, say to Swindon or Bristol, would definitely allow more services to remain in the game for longer. On your point 3 I agree it's desirable, although more work than that might be required, but how easy is it, will it require a complete rewrite of SimuGraph and how long will it take even if DTG decide to do it which is not a foregone conclusion?
     
  13. pschlik

    pschlik Well-Known Member

    Joined:
    Dec 8, 2016
    Messages:
    875
    Likes Received:
    1,607
    GWE at any given time will have 10 to 20 times as many trains as CSX will have in game at any given time, but I see basically the same performance in CSX and GWE with no AI in draw distance, and CSX AI acutally tend to be harder on the game when they are in draw distance than in GWE. GWE is actually a lot less demanding due to the many shorter trains as opposed to the longer trains.

    And on the topic of longer trains, if DTG doesn’t do something about SimuGraph, they won’t be able to porperly represent American freight. And they haven’t. The typical train over Sand Patch will be over 100 cars, and can have anywhere from 3 to 5 locomotives, but in TSW, 50 cars is the longest train length, and many trains could get away with just 2 GP38-2s/1 AC4400CW. Even then, Heavy Haul is still harder on FPS with trains already shorter than real life. If DTG is ever going to even care a little about making American freight realistic and challenging, SimuGraph will need to be reworked. And American freight is a big enough genre to be worth adjusting SimuGraph for.
     
    • Like Like x 1
  14. Drakoz

    Drakoz Active Member

    Joined:
    Jan 29, 2018
    Messages:
    152
    Likes Received:
    147
    My experience is GWE has much lower frame rate than CSX HH. I'm running with graphics settings maxed out (high and ultra) on 1920x1200 and getting 40-60 FPS in CSX HH, but 15-30 in GWE. I'm running on a GTX1080 (1 year old), but about a 5 year old i7 processor. My computer has no issue handling the trains, AI, and graphics in CSX, but I believe the high/ultra graphics settings in GWE is what is killing fps. So I have a great video card, but just an OK CPU/motherboard. Is SimuGraph using nVidia Physics on the graphics card? If so, that might explain my situation.

    Regardless, for the sake of this forum topic (extending a line), they have to implement it in a way that does not take any more compute power for an extended map as it takes for the existing map. Otherwise, it would be impossible to extend the map (if that is DTG's goal to be able to do). Larger and larger maps would just kill the game. So I hope that DTG has set this as a goal from the start, and that the Unreal Engine can do it.

    The bigger point of this discussion is does DTG even want to do it, or will they allow extending lines through the editor seamlessly. Take the NEC NY release coming soon. There is so much rail in NY and surrounding areas it is insane. NEC NY barely scratches the surface. How wonderful would it be if over time, DTG or the community could work to extend all the little sub-lines and tributaries and grow to encompass significant areas of the New York area. Or add the entire Northeast Corridor from DC to north of NY some day and have it interconnect with previous DLC like NEC NY. But DTG has to have created the basic structure to make it possible. Otherwise, this is just an academic discussion.
     

Share This Page