One of my favorite parts of working with and restoring old computers is learning their story. Old computers are often archaeology sites containing clues about how they were used and where they came from. From the 43MB backup that I took of the hard drive, I was able to recover a significant amount of data that I pieced together to tell the story of this beautiful machine as best as I can.
Preface
There was a significant amount of personal data on this drive. Emails, addresses, working documents, and other sensitive information was recovered. I will not be publishing any sensitive material to protect the identities of the owners and their families.
Names of private individuals, email addresses, phone numbers, personal mail, account information, and similar identifying information have been omitted from this project.
A Railroad Hiding in the Filesystem
As far as I can tell, this 3B2 machine started its main service in the railroad business. According to recovered information and some speculation, this machine was originally owned by Anthracite Railway, which is a now-defunct short line railroad in Pennsylvania. There is a significant amount of data to explain exactly what this machine was used for inside of ATRW. There was information regarding railcar movement records, car-hire accounting, billing, switching settlements, invoices, customer and railroad reporting, and equipment records. There are actually scripts that are for generating switching-settlement reports for operations around Pottstown, Kutztown, and Emmaus. Data regarding Octoraro Railway also appears frequently in data, which makes sense as Octoraro was a sister railway to ATRW.
Running a Railroad, One Shell Script at a Time
One of the more interesting programs I found on my 3B2 is 20/20, a spreadsheet software from Access Technology. A large amount of railroad work seems to have been built around a mixture of UNIX shell scripts, databases of records, and spreadsheets and reports. A user could select the billing operation they needed, process the car records, generate a report, and print it out. There are even scripts that modify themselves or their working templates from month to month, changing date ranges and regenerating monthly reports. I have reason to believe a lot of this software was written in-house, as one of the owners of this machine worked for AT&T and had experience writing C for them, as well as managing a fleet of 3B2s.
When the Office Machine Became Something Else
After several long, fruitful years of office work, this machine started its life as a personal computer. Around the 1990s, the work on this computer had shifted. Now the system is littered with UUCP transfers, including some logs troubleshooting UUCP file-transfer issues. This specific machine's UUCP handle was atrwrail, showing that the owner of this machine had an attachment to the ATRW company. I have reason to believe he was president of ATRW.
Railroad Knowledge Ended Up in the Courtroom
Another user on this 3B2 tells another story of use. There appears to be a large amount of data regarding expert-witness and technical consulting work centered on railroads and other transportation. It contains draft reports, case summaries, qualification material, invoices, correspondence, more 20/20 worksheets, case-status files, and documents prepared for attorneys. It seems after this machine had left ATRW it continued its railroad work, with its owner now using their railroad expertise in a legal setting. The surviving qualification material describes expertise in areas including railroad operations, transportation economics, accounting and cost analysis, rail safety, track and right-of-way matters, rolling stock, signaling, tariffs, and vehicle performance. Organizations appearing in the surviving material include names such as Amtrak, SEPTA, Metro-North, Conrail, CSX, the Long Island Rail Road, and WMATA.
When Registering for Classes Meant Making a Phone Call
This discovery is a complete 180 for this system. Within one of the /usr folders, was about a megabyte of development material relating to a completely different project: the University of Pennsylvania's Penn Automated Registration Information System, or PARIS. PARIS was developed with AT&T to allow students at UPenn to register for classes over the telephone. The recovered tree contains more than a hundred files with source code, shell scripts, tools, diagnostic programs, and tons and tons of documentation on the program. The surviving documentation indicates that the system was designed to work with was designed to work with AT&T CONVERSANT 2 voice-response computers. This system also interfaced with the IBM mainframe running the records system. PARIS would allow for a student to dial into the main system, and they could use touch-tones to register for classes, as well as manage schedules. One of the surviving descriptions says the system expanded advance-registration access to as much as 20 hours per day, seven days per week, and was designed to cope with periods exceeding 1,000 calls per hour. Personally, my working theory is that one of the owners (whom I had previously stated worked with AT&T) had been contributing to this project and used this machine as the development environment.
Fortunes, Games, and the Signs of a Well-Loved Machine
Aside from the expansive collection of corporate documents and source code, the owners of this system also enjoyed using this machine in their free time. The classic UNIX program fortune is installed on this machine, and the majority of users had fortune run at login. This is just one example of the customization of this machine. There are also multiple games on the 3B2, including Adventure, Boggle, Chess, Cribbage, Hack, Monopoly, Sail, Snake, and Trek. There are also personal discussions about railroads, hobby material, customized terminal settings, calendars, and BBS software.
The Story I Can't Tell Just Yet
Within the data there are mentions of another 3B2, which also seemed to be used as a personal computer or server of sorts. I have reason to believe that the second 3B2 I received is that machine. Additionally, there are three more mount points within /etc/fstab which appear to be partitions on the second hard drive that sat within the XM expansion chassis. Unfortunately, neither of the drives containing that information are working at the moment. I hope that I will be able to get both drives back into working order so I can properly recover that data and piece together more of the story of these beautiful systems and share it with all of you. What I can tell you is that this specific 3B2 lived an incredibly long life, with records going from 1984 all the way into 2002. It was a well-loved and well-worn workhorse and I am incredibly lucky that I get to know and tell the story of a machine that might've seemed insignificant from the outside but did some big things.