Results 1 to 5 of 5

Thread: One thing RW got right

  1. #1
    Join Date
    Jun 2005
    Location
    Fairport, NY, USA.
    Posts
    307

    Smile One thing RW got right

    I was surprised ( shocked ) that the destination ( as well as platform and yard track ) markers are now installable as a part of the scenario. Now anyy scenario can have its own markers and will not rely on the route author's placement. A real improvement for those writing scenarios.
    Also I detect an improvement in the behavior of the IA trains, relative to the driver in the standard scenario. In order to get the proper behavior however I found it necessary to place additional destination markers for the driver, especially as x-overs, but as noted above this is now not a problem.

  2. #2

    Default

    Please describe the process for placing scenario specific destination markers. I had looked at the documentation and couldn't find reference to that, so thought it wasn't available.

    Any help would be appreciated.
    Great find.

  3. #3

    Default

    Quote Originally Posted by jimbf View Post
    I was surprised ( shocked ) that the destination ( as well as platform and yard track ) markers are now installable as a part of the scenario. Now any scenario can have its own markers and will not rely on the route author's placement. A real improvement for those writing scenarios.
    Also I detect an improvement in the behavior of the IA trains, relative to the driver in the standard scenario. In order to get the proper behavior however I found it necessary to place additional destination markers for the driver, especially as x-overs, but as noted above this is now not a problem.
    At these cross-overs can the player train join a track which is being used by an AI service ?

    O t t o

  4. #4
    Join Date
    Nov 1999
    Location
    Newark, Delaware
    Posts
    1,293

    Default

    Quote Originally Posted by OTTODAD View Post
    At these cross-overs can the player train join a track which is being used by an AI service ?

    O t t o
    In spite of what Sniper says I know how to operate the switches from the 2D map.

    And using the route by Jims's own hand, I CANNOT operate the MANUAL turnouts to align the route for the player train IF that turnout is needed set the opposite way for an AI train.

    I am not an expert on setting up AI/scenarios, in fact I have never done it. But based on this simple experiment, I would conclude more work is needed on AI interaction.

  5. #5
    Join Date
    Nov 1999
    Location
    Stevens Point, Wisconsin, USA.
    Posts
    14,705

    Talking Trade in 8 of them thumbs for a set of fingers!

    Knowing how and doing it ain't the same thing, especially when your big greasy paws are mashing half the keys at OpilkNbmgjhCxvdfEwsrdf.

    What do we THINK, what do we KNOW, what can we PROVE.

    Need to apply those standards, we THINK the AI traffic is still useless, we don't KNOW that beyond reasonable doubt and we didn't really PROVE it yet. I'm probably gonna have to waste the money (won't have it til next month) to buy this godforsaken abortion and try it myself, because there are some variables involved.

    1. A railsim route with a free roam scenario that includes AI trains has successfully ported over to railworks - but does it work the same way as a route and scenario created with the railworks tools?

    2. Do portals created in the RS world editor behave like the ones in RW?

    3. If someone who knows how (and isn't all thumbs! ) makes a new scenario in that same route following a similar formula using the RW scenario tools instead of RS scenario tools, would the AI trains behave differently?

    That's the thing, I SUSPECT, which isn't the same as knowing for sure and being able to prove it, that they didn't actually fix the AI traffic cuz they don't have the first clue how it's supposed to work and therefore don't recognize it as broken. I have conclusively proven that with RS, but since I don't have RW yet I can only draw conclusions (conclusions are not the same thing as facts) based on what others tell me.

    "Also I detect an improvement in the behavior of the IA trains, relative to the driver in the standard scenario. In order to get the proper behavior however I found it necessary to place additional destination markers for the driver, especially as x-overs, but as noted above this is now not a problem."

    In a perfect world the AI trains would react to the signals instead of that unrealistic invisible path logic - that whole programming makes them nothing more than semi sophisticated animated scenery rather than actual AI traffic. In a perfect world the player wouldn't need ANY destination - path at all, the player is in control of his train and obeys (or doesn't obey) signals, and the AI trains react to the signals with a minimum of intelligence.

    https://www.trainsim.com/vbts/showthread.php?t=255907

    Note in that one, altho it uses the same clumsy path programming, the AI trains react properly not only to the player, but also to each other. It's hard to set up the route, signals, and traffic to do that in MSTS but it's POSSIBLE. I once made the prediction that they couldn't screw it up any worse with railsim, I was dead wrong about that - by leaving out the signal scripts with the signal number clear ahead logic this is a lot worse, since there's no way left to work around it. On MSTS PO&N I can do unpredictable things like blocking the mainline for half an hour and create a traffic jam, when I clear the main the AI trains sort themselves out according to the signal number clear ahead value and resume running their routes. About as intelligent as a trained rat, but a lot better than railsim.

Tags for this Thread

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •