Procedural Track for Open Rails

[R. Steele]

geepster posted this in another thread ( we were all getting off topic so a new thread was suggested - here we are)

At the end of the day, we are currently left with 2 major 'modern' track systems (ScaleRail and DB/US), and neither lend themselves all that well for use by third party US commercial content developers who want either hassle free distribution rights or long term durability of user access without proprietary website logins as a future weak link.

The problem is Open Rails is really suited for an environment where the track system is pretty much 'open' and not proprietorially locked down. It is not built around the concept of track cross section being a copyrighted thing by any one content creator. Any new track system needs to be freely unencumbered and openly sharable as Xtracks has been in the past. I support the protections of copyright for route and rolling stock content, but the people who need to see intellectual property restrictions applied to track systems kindly need not apply. Make your money off the routes, I support it. Make your money off the rolling stock, I support that, too. But leave the track systems freely sharable and openly accessible.

Absent that track system freedom utopia, if I were a commercial developer, the safest course of action for long term durability would be build the route in Xtracks, making sure I only use those track pieces where DB/US companion pieces exist (look at Norbert's chart), sell the Xtracks version of the route (Like what Trainsimulations just did with the ORTS version of Seligman) and then offer a free conversion kit to upgrade to UStracks. These conversion kits can either be produced by and distributed for free by the route developer, or can be created and freely distributed by any power users who have taken it upon themselves to download and preserve all the existing US/DB packages regardless if Norbert's site ever goes down in the future. As long as the US/DB conversion is not being done for money, everyone is in the clear and the upgrade package is freely distributable anywhere, and nobody has to worry about any proprietary track system's website ever going offline. We see that now, with power users freely distributing US/DB upgrades for both payware and freeware routes (Cascades, LS&I, Hook & Eye, Clinton Sub, etc).

Or some commercial developer just takes the risk and incorporates US/DB in one payware route download package and deals with any rights issues of a deceased creator if and when they ever come up. After a thorough pounding on the door of the deceased person's estate beforehand, and without answer, that is.

Aside from the inability of third parties to freely distribute, ScaleRail is also encumbered with version changes that have a tendency to break routes that were built around earlier versions of the packages. Try it yourself sometime - create a temporary Open Rails installation, run the installer for Feather River, and then run the installer for Sherman, both routes sharing the same Global and TrackProfiles subdirectory, and then examine the tracks and roads and see what is now broken an what looks goofy. Any card-carrying member of the tinfoil hat club just might conclude these changes in between versions are being done on purpose as a surreptitious anti-piracy crusade or something.

(Cascades with mosaic and UStracks kit applied)

I replied:

Back on track, good stuff. I'm trying to learn how to convert routes to DB/US, there is a steep learning curve for me, some tricky stuff. I've destroyed my experimental route a couple of times. ( backup, backup, backup )
Goku has a thread at ET about procedural track for Open Rails.
http://www.elvastower.com/forums/ind..._1#entry220776

Do you have any thoughts about this approach.

.... or anybody else have some thoughts, opinions...getting a new track system for OR is an important issue and I think geepster makes some important points.

[NorthernWarrior]

Essentially two styles to look at - how N3V do it with Trainz or how DTG do it with their TS.

Both have advantages and disadvantages and for that matter - currently dividing time between building a route in MSTS/OR and Trainz - IMHO there is still something to be said for the MSTS fixed track section approach. Ideally then you need the flexibility of all three systems.

Trainz problem is their track splines are a very undisciplined system - hard to control the curve radius and all too easy to end up with dogleg curves or points that are so sharp a 1950's model railroader would be ashamed of the tight curvature. The DTG system has the advantage of *some* control over the radius and that you can incorporate transitions. Where DTG hit the jackpot is the ability to select a section of track and copy a parallel item, be it another length of track, lineside fencing, platforms or procedural bridges, viaducts etc.

The $64k question has to be how this will be available in an editor? The MSTS RE is rapidly becoming dead stick and even I have started using TSRE for 95% of my effort. Assuming the OR team have no plans to code up their own RE (though that would be nice) then Goku needs to be involved in what can and can't be achieved under his code.

[Edwin]

I like the fixed layout switches in MSTS ScaleRail and DB Tracks as they incorporate all the necessary hardware and have the proper rusty guard rails making them look far more prototypical than prodecurals rendered from the same simple mould.
I find the procedural switches in TS lacking these important details. Also, most prototype switches have a straight section of track through the frog, instead of the all curved diverging track of TS.

Ideally, either 3D Trains or DB/US tracks should be persuaded to surrender their product to the ORTS dev team in semi public domain so it can be fully developed and integrated in any ORTS route editor.

[eric]

I'm not a fan of procedural track for the same reason as Edwin, mainly the lack of frogs and moving points & switch stands. Then again, I'm also not a fan of dynamic track, and most of my routes have zero sections in it.

If there was a cleaner way to place frogs and points, maybe I could be convinced.

[NorthernWarrior]

Another factor with procedural track is integration in the track database so the AI signaller/despatcher sees the points and crossings. Noteworthy that Trainz does not see diamond (flat) crossings and protect these unless you set up elaborate control measures outside of the basic track and signalling systems. And for all the hype around N3V's new(ish) procedural system is that it doesn't draw frogs and switches on crossings either, only on the points.

I do agree about dynamic track. I *suspect* though can't say for certain, Kuju's original intent was for this to be a truly flexible track joining system but ran out of time, money or programming moxie to make it work beyond what we got. Current route I'm working on is UKFS (with my Scalerail now safely backed up ) so dynamic track not really an option but I while I actually prefer the challenge of using all those interworking pieces to make things line up, would be much more carpal tunnel friendly to click the two nodes I want to join and have a custom track piece spawn between them.

[longiron]

Always felt that a hybrid procedural and fixed track system was the ultimate answer. The best of DTG/N3V while having exquisite crossings, turnouts and other complex track elements as fixed items. Let procedural do what it does best and let fixed items do what they do best.

[haverfordwest]

Best solution Chris and worked quite well in Trainz, years back.

When 3rd party creators were making fixed junctions, points (switches), crossovers etc to use alongside the procedural track system.

Mike.