[bobwdude]

Hi Eric,

Would either your proposed system or Goku's procedural track system allow for multiple types of track models and textures to coexist in the same route? At least in the US, a lot of tracks, even with concrete ties, still use wooden ties for turnouts. Likewise, different weights of rail are used for sidings vs. yards vs. mainlines, etc.

Both your proposed system and Goku's work both sound very promising!

[ErickC]

I actually had no idea procedural track was in the works, what I was going for was basically a substitute for spline track using dynamic track as flex-track (which is how a lot of hardcore N-scalers do it, they don't use sectional track). The idea was that route developers could just re-lay all of their track to retrofit existing routes, and it'd be faster because you wouldn't be dinking around with sections (bear in mind, I am not a route builder, so I have no idea how the mechanics of route-building work). In my head, my approach seemed like a good analog to the way track is laid in TS2018 or Trainz, it seemed like it might speed the process up a bit.

If the problem is the need for track shapes, I can build track shapes. What would you need? I can slice a generic piece of straight track into practically anything, turnout, crossover, and so on. My idea was to keep things as simple as possible to reduce the time requirement, hence not wanting to bother with sectional track. I also need to go out and take some photographs, and come up with an efficient profile. My original goal was to create something not too dissimilar from the track used in TS2018, still very simple, but with the correct rail width and a proper arch to the ballast. I'd create a basic track system and then upload the source files so other developers could build whatever switches or crossovers they might need.

I use GMax primarily, because nothing else I've used (besides 3DS) really meets my mapping needs, but I use MCX and 3DC for file interchange and texture baking, so I can import or export in any format. I'd have to hammer out the particulars, but it's certainly doable. I just need to know what's needed, and whatever references will make modelling it quickly easier.

[ebnertra000]

Erick,

One wouldn't necessarily have to relay all their track. If the shape names are the same, they'll just overwrite the originals wherever they're found. If you do what Norbert did for DB/USTracks, you could change the shape name slightly (easy to do in TSRE) to get the improved shape. Or, you could do it scalerail style and get dedicated entries in the tsection. Of course, none of this matters if you do flex-track.

I would recommend the first or second methods, as those would be feasible for retrofitting. The third, whille useful for new routes, would be a nighmare for existing routes. In order to remove and replace track, all interactives would have to be removed from the area (eventually from all parts of the route you'd want changed). I don't know about you, but that's not something I'm prepared to do.

If you're not familiar with DB/USTracks, you can find his work at http://dbtracks.com. He got a good bit done before he passed away a few years back...

[Genma Saotome]

ErickC, on 19 March 2018 - 07:58 PM, said:

The idea was that route developers could just re-lay all of their track to retrofit existing routes, and it'd be faster because you wouldn't be dinking around with sections (bear in mind, I am not a route builder, so I have no idea how the mechanics of route-building work).

Erick, you are not understanding: DT uses radian math to create their curves. Radians are mathematically incapable in RE to recreate sectional track curves whose length and curvature are calculated using angles. IOW you cannot replace any section curve with DT. Either the curvature will be wrong or the length, or both. In some cases it may be mathematically possible to do if you have something like a dozen decimal places of precision but RE doesn't allow for that. Perhaps TSRE does. Either way, it is not a drop in replacement.

Also, do understand it's not just the shapes -- the shapes are highly irrelevant to the sim itself -- it's the .tdb that counts and the notion of just yanking out a bunch of sectional tracks and dropping in DT has huge negative implications for the .tdb; Once again, it is not about the shape itself, it is about the ShapeIdx() value. Remove an extant value and you will break extant interactives.

There may be something in your idea that I'm not grasping but on the surface it would be entirely a disaster to apply to any finished route.

[Goku]

Radians are used by MSTS too. And the precision is "only" 6 digits. You can't use angles in degrees for math, so the values are converted to radians when tsection file is loaded.
I know route builders that succesfuly replaced standard tracks with DT.

[ErickC]

Bear in mind that I am more future-oriented than anything else. I'm less concerned with how to make old content better and more concerned with how to create an easier way of laying better track that anyone is free to use to encourage the production of new content.

Either way, I still plan on making any source files public domain, so one is certainly free to create a direct replacement. Actually, if someone has the time for it, that would be great! Before I realized the DB/UStracks author died, I was very tempted to send an email about the notice he had on his page looking for people familiar with GMax who were able to expand the set. Since he had most of the work done, I figured I could have the time for that. When I realized that building on USTracks would be impossible, the idea of creating an all-new system started brewing in my head.

The wisest course of action for anyone interested in creating a replacement, I think, would be to create something that overwrites the default track. There are routes like Blackfoot v3 that supposedly use USTracks, for example, but they require one to have the time and patience to learn how to swap pieces... I don't really have either. I think that most users would probably feel the same way (and there aren't enough route developers around to do conversions). If the end result overwrites the default track, every route is upgraded in one fell swoop without the user needing to wait for someone else to do upgrades.

Anyway... if procedural track is the next big thing, and if I can get a process down to make models in the right format, I'd be more than happy to work on it as time permits. I'd just need to hammer out an efficient but visually-appealing profile.

Genma Saotome, on 20 March 2018 - 10:12 AM, said:

There may be something in your idea that I'm not grasping but on the surface it would be entirely a disaster to apply to any finished route.

Quite possibly. I never had any intention of creating any kind of drop-in replacement. I also did not know that procedural track was in the works. My idea was just to create a system that would be analogous to the spline track found in more modern platforms. I was working on the assumption that OR-only routes would be developed in TSRE and that TSRE would be capable of meeting a route developer's needs in that regard. Bear in mind that I am not considering MSTS in any way.

All I know is Kuju/XTracks instantly kills any route its applied to, people working in other platforms tell me that spline track is much easier to deal with, and I can build models. If it's a bad idea, there's plenty of other things I can do. ;)

[ebnertra000]

I wouldn't bother with MSTS compatibility, either. Whatever some might say, the future is here, and it's called OR. What I am saying is, the only reason I used Xtracks shapes on one of my routes is the massive number of shapes which make nearly any layout possible to do easily. I'd love to dump the old track for something better, but if it means tearing up the track and replacing it, I'll pass...
There are a lot of routes out there today that could benifit from drop-in replacements, and some shapes, especially switches (and probably crossovers, too), could be reused in the future if you wand rigid switch types to go with spline track.
Don't get me wrong, I like the idea of spline track - it would make building this mine complex I'm doing infinitely easier (what were you thinking Erie Mining..?) - but we're not there just yet

[thegrindre]

I personally have 'played' with Trainz for three years and found that the spline track system is a fantastic way to lay track. I loved it and have suggested it a few years ago for Open rails but no response came. It's now time to do something about it. I'm all for a spline track system for OR.
Go for it! http://www.elvastower.com/forums/public/style_emoticons/default/thumbup3.gif

http://www.elvastower.com/forums/public/style_emoticons/default/oldstry.gif

[Genma Saotome]

I suppose it all depends on how it is implemented. In my first route I used KUJU's dynamic track extensively. In all subsequent routes I have used Scalerail. Given only those two choices I would never use KUJU's dynamic track for anything, in part due to the lack of multiple parallel (or congruent) paths. I can't speak for other examples in other sims.

What a whole lot of people don't know about ScaleRail is there are at least 5 congruent curves for each nominal radius -- meaning for something like 700r there are track shapes for the two congruent curve shapes sized for both inside and outside of that original shape. There are such extra shapes to match the track switches too. Of course you don't need to use the original "centerline" curve... you can use any of them as you need. It makes for a huge shape library.

[Stephen Hjellum]

I use primarily ScaleRail for my route building, as it's more or less accurate for my geographic region (i.e. Northern California). But I would prefer a ScaleRail system like in Railworks, which I believe is what Goku is working on. this piece by piece stuff is frustrating and laborious, and while I'm waiting for the procedural track system Goku is working on I bounce between projects - one is a fictional, personal fantasy railroad, and the other is a major undertaking, that is to say Northern and Central California. I started out on the later project as one big mega route, but after seeing the world tile folder size counts, I may have to reduce the route into smaller, regional chunks. But that is getting away from the current discussion, so I'll stow it for now.