Jump to content


Photo

Route Building


This topic has been archived. This means that you cannot reply to this topic.
9 replies to this topic

#1 enterprise_69

enterprise_69

    Conductor

  • Route Beta
  • PipPipPip
  • 315 posts

Posted 04 June 2007 - 02:14 PM

This is not a complaint post I probably like many others dont fully understand what goes on in the route building process, from ground zero till it is out for running freight. Maybe some one could give a indepth process or not so in depth to help us understand why we see and not from this website but many others, it take so long to complete. A route where there is no default items(all new creations, trees, buildings, roads, buildings, trains, cars.........). I have never been impatient about an up coming release of a route, I would just like to know from a creators view.
Thank you

#2 shawnbecher

shawnbecher

    Engineer

  • Valued Customer
  • PipPipPip
  • 258 posts

Posted 04 June 2007 - 02:46 PM

There is nothing simple or quick about building a route. A simplified version of the steps involved in creating a route would go something like this:

1. terrain creation (extraction, downloading DEMs, implementing DEMs)
2. setting up route markers (either using marker files or satellite images)
3. laying track (the more precise, the longer it takes)
4. laying roads
5. object creation (I've never done this, but I would imagine it takes a long time to get it right)
5. object placement
6. interactive placement (signals, crossing gates, car spawners)
7. beta testing
8. error correction
9. packaging for distribution
10. beta testing (you've got to make sure the package works
11. error correction
12. final beta testing
13. route release

That seems simple enough, but I'd estimate that it took me 6-9 months to do a 10 mile short line (26 total miles of track including sidings and industry spurs), and I didn't create any custom objects except the terrain textures. (Shameless plug - Wenatchee & Cashmere RR available at www.train-sim.com)

When creating a commercial route, every object must be created by the company producing the route. I would imagine that would at least double a route's build time.

#3 GaryG

GaryG

    Conductor

  • Valued Customer
  • PipPipPip
  • 342 posts

Posted 04 June 2007 - 03:08 PM

... and don't forget the instabilities of MSTS contributing to restores of versions that worked because that last signal you installed caused crashes when you get to that tile.

That's assuming you still have a saved version that actually was 100% sane and didn't have that undetected hidden bomb in it that showed up only when you actually ran the route.

Perhaps exaggerating a bit but probably not too far from the truth for some developers.

GaryG

#4 SAR704

SAR704

    Dispatcher

  • Valued Customer
  • PipPipPipPip
  • 579 posts

Posted 04 June 2007 - 08:46 PM

QUOTE(GaryG @ Jun 5 2007, 7:19 AM) View Post

... and don't forget the instabilities of MSTS contributing to restores of versions that worked because that last signal you installed caused crashes when you get to that tile.

That's assuming you still have a saved version that actually was 100% sane and didn't have that undetected hidden bomb in it that showed up only when you actually ran the route.

Perhaps exaggerating a bit but probably not too far from the truth for some developers.

GaryG


Ditto, and I dread it later in the year when the route on my machine is a few weeks from finishing. I haven't previously signaled a route, so I think more than one backup will be in order....

#5 Genma Saotome

Genma Saotome

    Dispatcher

  • Valued Customer
  • PipPipPipPip
  • 736 posts

Posted 05 June 2007 - 01:32 AM

Good trackwork takes time. Fer instance... how do you know beforehand you're going to need 7 degrees of 280r curves as the tightest radius? How do you know beforehand you can use 3 different radii as the transition into that curve, with individual radius of 1500r01d, 700r05d, and 500r05d? How do you know beforehand that if you continue to use a grade of .250 ascent that 200m later you'll either be making a big cut where you don't want one or will be using a grade of 1.250?

The answer is you try all sorts of things, pull out each failure, and try again. And again. Until it comes out right. And pull out might include pull out the 20 + track sections you were sure a half hour ago were going to be fine.

Really long tangent? You know where it's supposed to lie but if you can't get that lead-in curve perfect -- and I mean perfect -- the far end of that long tangent is going to plop down anywhere but where it is supposed to go. So "you pull out each failure, and try again. And again. Until it comes out right. And pull out might include pull out the 20 + track sections you were sure a half hour ago were going to be fine".

Teammate build you a bridge or building with an orientation that's at an odd angle to the lie of the track? Gotta use the object rotator spreadsheet and hand edit the world file.

Discover that going down to the far end of the yard, building the throat (with it's curves), and coming back the other way filling in the rest of the tracks gets you... SURPRISE.. there's a bad track joint with the last shape you drop in. Why? Difference of decimal precision plus rounding somewhere and it added up to a noticible problem. You learn the hard way to always lay the tracks all in one direction. So you gotta pull it all out and do it over the right way.

Almost as bad... you lay the yard throat, the tracks, the throat at the far end... and it's all too long. You have to redo the throat.

As bad as figuring out curves on grades: Any wye that's not equal on all sides. Any diverging but generally parallel path that has to come back to the main but the curves of that path are not symetrical. I each case you have to do an extensive trial and error of curvature and the closing of the final gap back into the main. Doing it several times... easy.

Real example: Doing the yard for Port Costa on the Cal-P... I think across the three tiles there are a few hundred track shapes...maybe 600-800 or so. I'll wager I placed 8-10X that before it came out right. Ditto for the trainshed area on the Oakland Mole. I wish I could insert a few images here... I'd show you the track so you could see the nature of the beast.

Now all that sounds bad, but IMO, it's easy compared to doing good artwork. Good art takes time. A lot of time. And I'll leave those details to your imagination.

#6 atsf37l

atsf37l

    Railroad President

  • Route Beta
  • PipPipPipPipPipPip
  • 2,656 posts

Posted 05 June 2007 - 08:40 AM

Let me just say, to all the route builders out there who have contributed so very much to this hobby and have all gotten a little grayer and a little balder for their efforts,

THANK YOU!


#7 muzik4

muzik4

    Ticket Inspector

  • Members
  • PipPip
  • 31 posts

Posted 05 June 2007 - 11:43 AM

Something I don't see mentioned above if you are going to do a prototypical route.

It probably takes almost as long to do the prep work (research) before even starting out with any RE work as it does to do most of the later work.

There are just no shortcuts if you want accuracy.

#8 shawnbecher

shawnbecher

    Engineer

  • Valued Customer
  • PipPipPip
  • 258 posts

Posted 05 June 2007 - 11:59 AM

That is very true. I've been working (off and on) on the track charts (converting elevation from feet to meters) for my next project for about 3 months now.

#9 Genma Saotome

Genma Saotome

    Dispatcher

  • Valued Customer
  • PipPipPipPip
  • 736 posts

Posted 05 June 2007 - 12:15 PM

QUOTE(muzik4 @ Jun 5 2007, 11:24 AM) View Post
Something I don't see mentioned above if you are going to do a prototypical route.

It probably takes almost as long to do the prep work (research) before even starting out with any RE work as it does to do most of the later work.

There are just no shortcuts if you want accuracy.


The numbers are staggering. I've taken over 1200 photographs, none of which are of track or trains. Thank goodness for digital cameras! I have 600 .pdf files of individual pages from Sanborn Fire Insurance Maps, 190 images of period USGS maps, and I don't know how many hundreds of images taken from the web ranging from what a California Highway sign looked like in the 40's to what rice and safflower fields look like today. Fortunately a lot of that can be collected in parallel to building the route. Before hand tho you do need track charts, station plats, markers... speaking of markers, I think I have around 13,000 of those spread across numerous files and then of course hundreds of square miles of photos in USAPhotoMaps... you know the urban scale images are one pixel per ~6-9 inches. I have most of the route imaged at that level of detail.

There's stuff you'd never imagine up front you need. Dirt... what color is it? River water... what color is that? Mud? Same question. Didn't think you needed to know advanced Photoshop? Guess again. Was the Tower Bridge yellow in the 50's? Nope. Gray. Better reskin it then. And on and on and on.

#10 enterprise_69

enterprise_69

    Conductor

  • Route Beta
  • PipPipPip
  • 315 posts

Posted 05 June 2007 - 12:51 PM

Being a helicopter pilot I can say this...
It is harder to build a route for MSTS than it is to learn to fly.
My respects to the route builders.