They may have updated it since, but when I left CSX in '08 most of the more modern locomotives were running Windows XP, which is certainly not exotic (also probably no longer supported).

My guess is that like B&O 51, most preserved power from the era is going to end up being "stuffed and mounted" displays. And I really don't know how much would even be saved anyway. We're not doing too well with some formerly common units right now. When you've got units costing $5 mil. + new, getting them donated, especially by quasi-governmental agencies, is a near-nightmare if not impossible. And while a company like LTEX might be at least a little sympathetic, they are still in business to make money.

Curiosity killed the cat so I went straight to the horse’s mouth. I know a guy that works in a Class 1 locomotive shop maintaining the software. This is what he said:

“...to the best of my knowledge neither GE or EMD software is designed to work for x amount of time then fail. I will agree that if some point in the future an EMD Ace or GE EVO is donated to an operating museum, the museum would need to get a copy of the current software being used on the locomotive. Now I can’t give you a lot of specifics, but I can tell you EMD software runs on the Windows platform and GE software runs on the Unix platform. The software is proprietary to each locomotive, i.e. you cannot run EMD software on GE locomotives or vice versa. The software is loaded via a USB thumb drive and the thumb drives are loaded from programs that run on laptop computers. Just my opinion, but I don’t see the software as being an issue, I would think the bigger issue might be the IO cards in the GE computer panels and finding replacement computer screens for both locomotives. Again I cannot go into specifics, but the screens on both types of locomotives are computers as well. Also both locomotives run networks to communicate with all the different components. EMD’s use the Controller Area Network aka CAN bus and GE locomotives use the Attached Resource Computer Network or ARCNet. So any heritage railroad would also need to find and stock these components as well for the respective locomotive.”

I'm dealing with this now.

Because of the speeds we operate at, we are required to have event recorders and alerters on our diesel locomotives.

The Wabtec event recorders we used to use were fairly reliable and we did not have a lot of trouble with them. But when we did, getting technical assistance was tricky. They were outdated and no longer officially supported by Wabtec. However, we had a few contacts in the organization that were old enough to remember the system and could provide limited technical support, but parts were no longer manufactured.

I think it was 2017 when the FRA required us to have alerters equipped on all of our diesels. Modern alerter systems are packaged with the event recorder and the speedometer. So we had to replace all of these components on each loco in the fleet, at a cost of about $20k each. At the time there were 2 manufacturers, Wabtec and Qtron. I was told both of them were widely used in the industry. We chose Qtron Little did I know that they were getting close to going under and bought out by Siemens.

The Qtron system has been nothing but a can of worms. We had event recorders go bad within months of install. Speedometers as well. The interface to a laptop requires a serial port. Try finding a laptop today with a serial port! We had to make our own specialized adaptor cord.

We're currently struggling to get our laptops to fully read any of our event recorders. We have 2 laptops, 2 download cords, and 7 locomotives. No combination seems to work. We can get a download, but if we had to reprogram an event recorder due to change in engine number or wheel size, we could not do it. Right now we're having a different laptop shipped to us to see if it behaves differently.

Eric Hadder
Grand Canyon Railway

Last time I checked, modern locos have no issues operating on Natural Gas/Propane with some modifications, albeit not as efficiently. Same question applies to what’ll happen to SRC when the world coal reserves run out...

This issue has come up before on this board but in the context of preserving modern LRVs where the propulsion control is solid state with considerable software/firmware complexity.

The problem is the word "proprietary"! In almost all cases, the source code and internal documentation for these control systems for rail vehicles is a proprietary trade secret of the manufacturer. Even the railroad or transit agency which purchased the equipment doesn't have access to it. This touches on the larger political/consumer rights issue of "right to repair" except it doesn't [directly] affect hundreds of millions of consumers.

The only hope for long-term preservation of this stuff in operating condition is if the original vendor eventually releases or licenses the source code and all other internal secret documentation for the product to the preservation organization. You see in general industry it is very hard to get software support for a 20 year old product. Why would a vendor still be supporting software after 40 or 50 years?

Of course the other problem is once you have the code, who can understand and maintain it?

In theory, you could "black box" it in the future: replace the old control system with a contemporary one that does the same thing. If you are talking about engine controls for a diesel, to do so you'd still have to be in possession of basic engineering data about the engine

To date myself; I have in fact used 8" floppy discs while programming computers.

In the beginning they were "one sided" and held a whopping 180 kilobytes of data.

"Large" computer programs would not fit on a disc, so we used "overlays" where portions of the program where on different discs. When running the program it would stop and tell you to remove disc 5 and insert disc 11, etc. etc.. What a pain. Could take hours to run a program that takes seconds today. It was a step up from batch processed punch cards.

Eventually (1979 ?) they introduced 8" floppy discs with data on both sides, IIRC you flipped the disc over to get the data on the other side. Then they had floppy disk drives with two sets of read/write heads so you could get to all of your 360 kilobytes of data.

Back in the old days all we had was ones and zeros, occasionally we'd run out of zeros and had to substitute O's (har har).

Cheers, Kevin.

A regular serial cable is just wire. If your event recorders use nine-pin serial cables, then there are just the nine wires (in the cable) connecting the laptop to the event recorder.

USB to Serial cables are much more complicated. They have chipsets embedded in the cable and software (drivers) running on those chipsets. When you use a USB to Serial cable to connect your event recorder to your laptop, you end up inserting a processor and software in between the laptop and the event recorder. [Murphy likes that.] I'm not surprised that it is not working well; I doubt the cable manufacturers tested their drivers with your event recorder software.

I would try buying an old business-quality laptop that has a built-in serial port. For example, all the Latitude D6x0 series (Wikipedia) have 9-pin serial ports and there are a bunch of reconditioned D630s on ebay for under $130. Your IT department won't want to support it, but you won't really need support if all you do is use it for your locomotives. (Just don't ever connect it to the internet.)

Indeed, but between the wire and the microprocessor (brain of the laptop) is a UART (Universal Asynchronous Receiver / Transmitter) a remarkable piece of semiconductor engineering that needs SOFTWARE support.

I can't tell you the number of times I've connected a "simple" serial (RS-232) cable between a computer and something else expecting to talk to it like making a simple phone call...

It usually takes a number of adjustments to baud rates, xon/xoff, parity, stop bits, hardware handshaking, etc. settings to get it to talk to the thing on the other end of the serial cable (aka "just a wire"). And those setting are controlled by software "drivers" (often obsolete and irreplaceable for the hardware at hand).

One of the most annoying settings is the "auto echo" feature, every letter you type intending to send to the "other end" comes right back at you like this; eevveerryy lleetteerr yyoouu ttyyppee ccoommeess bbaacckk aatt yyoouu... If that drives you nuts imagine what it does to a dumb laptop computer ?



Sounds so simple, if only it was so.

If a serial cable is "just a wire" why are products like this available'

https://www.amazon.com/RS232-DB25-BREAK ... B005H46XS2

I still have a few of these in my tool box.

You are correct, USB is the next level of complexity.

Remember; In Theory there is no difference between Theory and Practice, but in Practice there is usually a heck of a lot of difference....

There are 9 pins in DB-9 connector, but only 3 of them usually matter: ground, transmit and receive.
Most systems that used RS232 for direct communication (as opposed to via a pair of modems) did not bother with RTS/CTS or DTR/DSR/DCD. RS232 is hardly rocket science. I've never encountered a serial device that I could not get working within 2 minutes.

Here's a list of many typical D-sub 9-pin connector configurations:

http://www.nullmodem.com/DB-9.htm

(note that really (before PC users corrupted it) the standard electronic code for this connector is DE-9, not DB-9)- see first paragraph on the link.