XTrkCad, one of the many model railroad track planning programs--and the only one I know that has both Windows and Linux versions--is going open source. This also means the program is now free.

You can't download the source code yet but you can download binaries, enter a registration code, and play with it. I've been doing just that.

One thing I noticed right away when I started trying to use it to plan a layout using Lionel and Marx O27-profile sectional track--which it doesn't support directly, so I had to enter the track, confusingly, using the "custom turnout" tools--is that the model railroad and toy trains people measure track differently.

A Marx or Lionel O27 curve isn't an O27 curve in XTrkCad. It's a curve of radius 12.5 and angle of 45 degrees.

Here are some measurements. Keep in mind this is what they're supposed to be. Manufacturing tolerances and the effects of age sometimes cause these measurements to be off. Some of my vintage track is off by 1/8 inch or more.

O-27	12.5"
O-34	15.75"
O-42	20.25"
O-54	26.375"
O-72	35.25"

I happen to know that O27 and O34 tubular track use a 45-degree angle, and that Lionel and K-Line O42 tubular use a 30-degree angle. Unfortunately I don't have a piece of K-Line O54 or O72 sectional tubular track to measure.

Since O42 track is supposed to be 12 sections per circle, and 180/12=15, I believe to calculate the angle measurement of O54 and O72 you can divide 180 by the number of O54 or O72 track sections in a full circle, then multiply that result by 2. The math works for the O27, O34, and O42 track sections I have, but since I don't own any K-Line O54 or O72 track I can't be certain.

Also, confusingly, traditional O gauge switches are straight turnouts in XTrkCad, even though one leg of them is curved. Here are the parameters I got by measuring a Lionel 1121. The measurements are close enough to represent Marx or Sakai O27 switches:

Diverging length: 8.5
Diverging angle: 45
Diverging offset: 3.75
Overall length: 8.75

And here's what I measure on a Marx O34 switch.

Diverging length: 11.375
Diverging angle: 45
Diverging offset: 4.5
Overall length: 11.188

I believe the diverging angle would be 30 degrees on a Lionel or K-Line O42 switch, but since my vintage Marx and American Flyer locomotives won't make it through modern O42 switches, I don't have any of those to measure.

I've already used the software to draw a layout using Marx O34 track that will allow two-train operation with room on sidings for three additional trains. It's much easier than setting up track on the floor and measuring it to see if it will fit on the table. And you can do your layout and then print an inventory to compare what you have with what you need. Not having enough track to mock up a layout isn't a problem anymore.

On the computer some of the things I want to do don't quite fit; if the track measurements are slightly off, the solution is to cut a section of track to move things a bit, or, if they're off by less than a quarter inch or so, force it. O27 track has a lot of give, and, like I said, manufacturing tolerances and the effects of time can cause real-world track to not match the published standards.