I bought a Raspberry Pi this week after ruminating about it for literally years. I finally decided there are a few projects I’d like to attempt that will require the Pi (some of them can be accomplished with an Arduino, but I am more comfortable with a complete single-board computer rather than a controller).

First I bought Eben Upton’s book and read halfway through it and decided it didn’t look too bad, and of course it sure doesn’t cost very much (more on that later). Then I pulled the trigger on a CanaKit starter kit with Pi, case and power supply.

Raspberry Pi 3 B+ installed and running

For years I have wanted my own Unix/Linux system to play with. I have long been fascinated by Unix, and sometimes regret not having got more involved with it back in the 1980s when I first joined the technology industry. But between 1985 and 1994 my employers operated in the DOS/Windows/IPX markets exclusively. I did have a few opportunities to dabble in Unix over the years.

Keep in mind that until Linux came along in the mid 1990s there was really no way economically to use Unix except through a shell account on a shared server (like my old netcom.com account). Unix ran on expensive multiuser minicomputers and workstations, and you were unlikely set up one of these in your spare room.

In early 1985 I briefly worked for Touchstone, a company that produced Unix software, and we had a couple systems in the office for development and testing: a Fortune Systems 32:16 and I believe a machine from Altos Computer Systems, both running Xenix. My access to these machines was fairly restricted. The company also had some kind of relationship with AT&T, which at that time owned Unix, and one day a complete AT&T 3B2 workstation with Unix System V appeared on a desk in the middle of the office. The 3B2 was a very expensive platform as I recall, and I believe this was some kind of freebie for potential development partners. However, no one from the development side of the company had anything to do with the machine as long as I was there, it just sat on an otherwise empty desk in the middle of the marketing department. When I had spare moments, I would sit at it and fool around with Unix System V (this was the first of these mysterious unassigned office machines that I encountered in my career; the second was big Mac workstation sitting in the middle of the Cisco UK office in 1994 that had Mosaic installed, which was the first time I ever saw the World Wide Web).

Once I left Touchstone, however, I remained firmly in the world of DOS, Windows, Netware and IPX/SPX until I joined Cisco a decade later. I first started fooling around with TCP/IP stacks for Ethernet adapters around 1989, but it was still on MS-DOS PCs.

By the early 1990s I was working for Newport Systems, and while we had nothing to do with Unix at that company, one of my colleagues loaned me a copy of The Cuckoo’s Egg by Clifford Stoll, in which he goes into some detail of his work with Unix, which again ignited my interest in the operating system.

A few years later, at Cisco, I was tasked with working on the development of a number of network-related software projects, and found myself researching many RFC-based network utilities and protocols, all of which were heavily connected to Unix and used many Unix standards and techniques, such as plain text configuration files (which were very attractive to anyone who had to suffer with the Windows registry). The Cisco campus on Tasman Drive was festooned with Sun workstations, many of which seemed to orphaned or abandoned like that AT&T 3B2 at Touchstone, and again when I had time I’d find myself noodling around with Solaris, Sun’s version of Unix.

But all this noodling doesn’t really get you anywhere.

In the late 1990s Red Hat popularized Linux distributions that could be installed quickly and easily on basic PCs, and I recall buying a surplus PC from my employer and even installing Red Hat on it on my kitchen table, but I don’t remember doing anything with it after that; I think I moved to San Francisco soon after and never had a chance to play with my Red Hat machine. Then for a while I used a shell account, I think with best.com, for e-mail and Usenet, but that was all.

During the 2000s I would occasionally find myself with an obsolete laptop and I often intended to install Debian or Ubuntu on one, but I just never did it. By then I was no longer involved in technology except as a hobby, and not much of a hobby at that.

With the Raspberry Pi I intend to tackle a number of objectives. First, there are a handful of small motion control and other projects I’d like to attempt for which the Raspberry PI and its component ecosystem would be a good fit. Next, I want to be come more proficient with Python. I’ve spent the last few years trying to learn Python in an admittedly desultory way; I buy a book and work through some of the exercises before I put the book away fro a few months or years. With the Pi I’ll be obliged to employ Python for any of the projects, and for me that’s the best way to learn any new skill: by actually using it to make something I need or want.

Finally, because it’s a Linux machine, I really hope I can finally become familiar with this environment. That’s just gravy, because I was considering the Raspberry Pi for project-related work before I learned it ran Linux natively.

I’m really very impressed with this little computer. Once you use the mouse, keyboard and monitor to set up the Pi and connect it to your WiFi network, you don’t need them anymore; you can connect to the Pi with SSH or VNC. I’ve trained my Pi to connect to both my home and office WiFi networks, so now I can access it at work or at home simply by slipping it into my shirt pocket and taking it with me (I’ll get dedicated power supplies for each location so I don’t have to carry one around).

Headless Pi

Naturally, I’ll want additional units for the various projects I have in mind, but at about $55 a pop for the Pi, case and microSDHC card, it’s cheap and easy to get as many as you need.