IPB
viagra super force buy buy online

Welcome Guest ( Log In | Register )

 
Reply to this topicStart new topic
Now that MSTS-X is VapourWare, Is there room for improvements to Activity Generator
SurvivorSean
post Jan 23 2009, 10:33 PM
Post #1


Switchman
*

Group: Members
Posts: 9
Joined: 27-May 08
Member No.: 163



Well as you may have already heard, MSTS-X appears to be on hold, or gone. Sounds very similar to both the last promise of MSTS v2 and another one I heard about TrainMaster as well. Ironically I posted a message a couple days ago over on Mophouse Forum of some of the plans I had for my RTS concept (Railroad Transportation Simulator). The reason why I never pushed too hard on RTS was I wanted to see where MSTS-X was going to take us. With recent additions of Bala and GTA from MLT, I have had a renewed interest in both running trains and creating activities in templates for the region.

So without going into too much in what I planned on doing with RTS again, I did want to bring up a couple of things. I like to use 2 programs in assisting me with my activities that I create. Activity Generator and Activity Changer. Of course Activity Generator of course is why we are all here. Activity Changer you may not be aware of however, is an older program that introduces random AI trains into activities. In fact I often run Activity Changer on Activity Generator templates to randomize the AI services that show up while I run the random switchlists created by this fine program.

One thing is for certain, I'm not about to wait for the next simulator to come and grace us. As far as I'm concerned MSTS is quite capable of doing much more than is currently capable. One way or another I either want to see these things implemented, or if necessary program them myself. I'd certainly choose option 1 as I'd much rather leave the programming to more experienced people in the industry. Though I do program, I'm a bit dated when it comes to the latest programming languages and techniques. So this is more of a let me throw this to the wall and see if it sticks. In other words if Steve, or others are interested - please do. I'd also like to offer anyone my assistance in getting this implemented, or if others would like to could assist me in making this happen.

Improved Waybill handling and the ability to run continuous activities: This was mentioned a while back as something Steve would like to see in a version of SLG for MSTS-X. I know the save files have not been cracked, just like so many other things in MSTS with little or no technical documentation to go on. A very good program for model railroading called Freight Yard Manager I find is a good way of handling car movements. It is on the basis of moving freight via waybills versus movement via car types. What this does is force movements that make sense, and makes car logistics much more dynamic. Another feature that this program does well is continued operations. It assumes that car movements are completed as the waybill specifies. Otherwise there are functions of changing the cars location to that of where the actual car resides at the end of a session. I think the addition of these functions would give MSTS an even greater experience, and one way or another I want to see it, or work toward it myself.

Scenario Methodology in Random AI: Now I know this has nothing to do with what SLG or Activity Master does but I thought I'd throw it out there anyways. The program Activity Changer does a good job at generation random AI traffic. It randomly changes the AI times, and selects random services (with some services containing different paths). This works very well on some single track mainlines (though the program actually encourages it's use on double track mainlines). As long as the route is set up correctly (balanced nodes, and not too complex tracks) the passing path optional setting in the AE will allow proper meets for random AI. It does work well for double track mainlines too with one major exception. If one of the services is either a passenger train or fast freight (those with either less work or higher speed limits) overtaking of other services in the random method is impossible. In most of these double track routes there are plenty of opportunity to overtake via crossovers, however passing paths do not work at all when it comes to double track - attempting a passing path that never rejoins actually will crash the AE.

What I'm looking to do is expand on the Activity Changer program by taking a different approach and that is where the term scenario comes into play. With proper route and train databases it doesn't take much to calculate where trains run, how fast they can run, and where they have the ability to overtake. For instance you wouldn't want a passenger to overtake a slow freight and miss the platform. However, you could have that passenger crossover in certain areas where a platform would not be bypassed. For each possibility of where the dispatcher could put you, basically means another scenario. Each scenario would have it's own path set, along with appropriate waiting points (and or double reverse points) to trigger the AI dispatcher to work for you. An improved Activity Changer program would be as smart as the AI and/or VCR mode and know where trains would meet. Thus it would be able to compute where trains would be, how long they need to wait (waiting points), and what paths they would have to take. Once the random generation is complete the scenario would copy the appropriate paths for each service and generate the activity.

Of course what all this means is the continued challenges of what Activity Generator already provides with new switch lists each time. Activity Changer already provides the random AI, and this would only give greater potential for higher capacity lines to take advantage of. The bottom line is for me (and I assume many of you) running the same activity over and over again is boring. As an activity creator I don't like reading my own novels, and enjoy the random factors I already create. These newer features would only add to the potential and provide even more prototypical operations without the need of creating their own activities the old fashioned way.

Thanks

Sean
Go to the top of the page
 
+Quote Post
nbeveridge
post Jan 24 2009, 08:31 AM
Post #2


Train Master
*****

Group: Administrators
Posts: 550
Joined: 2-January 07
From: Texas
Member No.: 10



QUOTE(SurvivorSean @ Jan 23 2009, 10:33 PM) *
Activity Changer already provides the random AI,
Thanks

Sean



Actually, it would be fairly easy for Steve to modify AG to permit randomization of the AI trains. Then maybe Activity Changer would not be needed. But remember that in general we set up the templates with opposing traffic, and only rarely with following traffic for the very reasons you state. My suggestion for randomization would be that the user could set how much the trains vary, such as 10% early to 25% late, and then the randomizer picks something in between those bounds for each AI train independent of the others. Recognizing that the AI dispatcher in MSTS is not perfect, this can certainly have unexpected results. But the user could always pick 0% early and 0% late to get back to the original template.

As to continuation of trains from activity to activity, we know that the save game file has yet to be cracked. Perhaps that will happen sometime in the future. Maybe the bin patch people can tell us how the file is built???


--------------------
Norman
Go to the top of the page
 
+Quote Post
sgdavis
post Jan 29 2009, 08:58 AM
Post #3


Train Master
*****

Group: Root Admin
Posts: 727
Joined: 1-October 03
From: United States
Member No.: 1



QUOTE(nbeveridge @ Jan 24 2009, 08:31 AM) *
Actually, it would be fairly easy for Steve to modify AG to permit randomization of the AI trains. Then maybe Activity Changer would not be needed. But remember that in general we set up the templates with opposing traffic, and only rarely with following traffic for the very reasons you state. My suggestion for randomization would be that the user could set how much the trains vary, such as 10% early to 25% late, and then the randomizer picks something in between those bounds for each AI train independent of the others. Recognizing that the AI dispatcher in MSTS is not perfect, this can certainly have unexpected results. But the user could always pick 0% early and 0% late to get back to the original template.

As to continuation of trains from activity to activity, we know that the save game file has yet to be cracked. Perhaps that will happen sometime in the future. Maybe the bin patch people can tell us how the file is built???


Yes, I have considered randomizing the AI trains. The problem of course is with the current MSTS pathing and AI dispatcher, this will increase the likelihood/possibility of "cornfield meets" (2 trains facing one another on a straight piece of track) and other dispatching problems. If the dispatcher worked correctly this would be a great feature.

I will look at adding this feature as you describe it here Norm, users can always opt not to use it. And if Activity Changer does this and doesn't cause too much AI traffic trouble, than AG will work similarly.

Thanks,

Steve


--------------------
Steve Davis Skyline Computing SteveDavis@SkylineComputing.com
Go to the top of the page
 
+Quote Post

Reply to this topicStart new topic
1 User(s) are reading this topic (1 Guests and 0 Anonymous Users)
0 Members:

 



Lo-Fi Version Time is now: 23rd April 2024 - 03:52 AM
viagra super force bitcoin