This particular thread
This particular thread highlights one of the problems I have with other computer programs for car forwarding - they don't take the Car Loading Rules into account. But it gets more complicated for somebody creating such a program, since the rules adherence, as wel as the rules themselves changed over the years.
From about the '20s through at least the '50s, the rules indicated that you were to load an available foreign car that was:
- Destined for the home road; or
- Destined for a location between the loading road and the home road; or
- Destined for a location beyond the home road, so the home road could participate in the movement.
Sounds simple enough. But prior to WWII, adherence was around 90% if I recall. Post WWII it had dipped to around 60%. Also, during specific times (such as the annual grain rush), there were shortages of cars. During the grain rush, it was box cars. That altered the movements of the cars too.
Railroads wanted to get empties off of their road by midnight to avoid having to pay another day's worth of Per Diem. On the New Haven, (and probably most roads in the era) the rules required them to send empties back to the junction point where they were received, unless the car was owned by a railroad with a direct connection to the New Haven, in which case it was routed to the nearest junction with that railroad. There were further exceptions, such as all CN cars were routed to the CV in New London.
Another factor that is a problem for model railroads, but not the real railroads, is the mix of cars themselves. Some cars are much more common than others. For example, if I had a load coming from Chicago to the New Haven, it would be fairly common for it to be a Pennsy car. I have a lot of them in my roster: X25, X29, X35, X37, and X38 box cars for example. But some of those cars are more common than others. For example the PRR only had one X35. So I'd want that to show up maybe one out of every 10 or 20 ops sessions, not every one.
The mix of cars is important as well. In 1950, 26% of the box car fleet were single sheathed composite box cars. Box cars shorter than 40' accounted for about 8%, and of these, 71% were CN and CP Fowler patent single sheathed composite box cars.
This is a list started by George Losse that I've expanded a bit. It shows the largest classes of box cars in 1952 and roughly how many are likely to show up in a group of 100 cars:
Prototype | 1952 | Per 100 | Best HO model |
USRA SS | 9756 | 2 | Westerfield |
USRA DS | 5215 | 1 | Westerfield or modifier Ertl |
23 ARA | 66,125 | 9 | Red Caboose (x29) |
32 ARA | 11,854 | 2 | Atlas |
37 AAR | 60,077 | 8 | IMWX/Red Caboose/Intermountain/TLT |
37 AAR mod | 41,094 | 5 | Intermountain |
War Emergency Boxcars | 5342 | 1 | Intermountain |
44 AAR | 34,065 | 4 | Branchline |
PS-1 | 46,271 | 6 | Kadee/Intermountain |
Fowler Patent SS Box Car | 47,554 | 6 | Westerfield/TLT (P1k) |
USRA Steel (NYC) | 20,000ish | 3 | BLI/Westerfield |
MILW Rib Side (all classes) | 15,000ish | 2 | Exactrail/Intermountain/Rib Side Cars |
B&O Wagontop | 5,000ish | 1 | Exactrail/Fox Valley |
Other SS box cars | 162,700 | 21 | Various |
Other short box cars | 20,000ish | 3 | Various |
CP Minibox | 7,000ish | 1 | True Line Trains |
PRR Wagon-top cars | 14,000ish | 2 | Bowser |
Steel Rebuilt USRA DS Box Car | 11,000ish | 2 | Sunshine |
1937 AAR DS Box Cars | 13,000ish | 2 | Sunshine |
Totals from the 1/52 ORER | 279,079 | 81 | |
I've also been collecting railroad Routing Guides that provide the approved routes from a given railroad to other destinations. These routes are as specific as from a given station on one road to the specific station on another and can run hundreds of pages. This can affect the operations on the railroad. I model a single city, but outbound cars can be routed in one of three directions. A car headed to Chicago can actually be headed in either of those three directions - north, south or west.
I'm using a spreadsheet that Mike Rosenberg created, that Chris also adopted and we've been tweaking it for some time with Mike. What I like about it is that it handles the non-railroad aspects of the car movement. Industries ordering cars. Then I can fill those movements by selecting the cars myself, so I can take into account the car loading and movement rules, the New Haven car movement rules, build a representative mix on the layout, and account for common and rare cars too. While I'm sure it's possible to create a computer program that will account for all of this, it would be very complicated.
Something else that is very important to me is that the program not only print prototypical paperwork, but that it allow me to decide which paperwork to print, and which paperwork I will do by hand. I love the railroad paperwork. That's part of the fun of running the railroad for us. At my sessions, I have a freight agent that receives the waybills from the conductors, then (prototypically) writes out the switchlists for the crews. He then provides the waybills for the outbound trains. So I'm OK with a system that lets you use it with "little or no paperwork" but that's not a primary goal for me, and I would like to make sure I can use "all of the paperwork" if desired.
The other factor that is usually missed in programs like this is that sometimes I want to just run the railroad. Any computer program that tracks the movements of specific cars, needs to know the location of that specific car at all times. But if a crew makes a mistake, or a car misses its movement, that can throw the system off. Mike's spreadsheet does away with this. All we do is mark whether there is a car at a given location or not. It doesn't care what car, or even what kind of car. All it needs to know is whether there is a car there, so it can decide whether to move it or not. All of the car tracking is done with the waybill that can move with the car, whether I'm moving it via an ops session or not. If I choose not to use the waybills, that's not a problem either. The waybills are generated (or pulled, since I save them) prior to the next operating session when setting it up anyway. So a car without a bill is addressed then. Or left without a bill, since that occurred in real life too. The Agent will received one in the "mail" during the session, then the movement of the car will be known.
Another problem is that cars are typically assigned trains throughout the session. On a large layout, with many trains run, a car may be scheduled to move from an inbound freight (from staging) to a through train to another yard, to a local freight to be delivered to an industry. All prototypical in a sense. Except that we often have people waiting for a late train to receive their car. So I would like to see a computer program that designates the routing (off-layout> yard #1> yard #2> industry) that allows the agents and yardmasters to build the trains as the cars come in. A given car may be expected at 10:00, but if it comes at 2:00 it's irrelevant. No train will wait for it, and it will go on the next train headed in the right direction, just like the prototype.
I'm happy to help in any way I can. It would be very cool to see a program that can handle prototypical car movements. I get that not everybody may be as interested in many of the things I'd like to see. However, even those that aren't concerned about prototypical car movements, mix, etc., would benefit from a system that does that anyway.
Randy