Thursday, March 12, 2015

Who is Sharla P. Boehm?

Last November I had the pleasure of attending the Museum Computing Network conference – which I highly recommend. It was at a session there that I watched Rosanna Flouty as she presented the diagram below in her slides, illustrating different models for organizing communication.

Some of you will recognize the source – it comes from the work of Paul Baran in the early 1960’s, when he was part of a small group of people thinking deeply about how to build networks of communication between computers. Baran’s work, completed at Rand Corporation in Santa Monica, was foundational in the creation of what would become the Arpanet and, eventually, the Internet.

The core of Baran’s contribution was what we now know as packet switching – the idea that in order for computer A to send a message to computer B, you don’t need a fixed electronic “pipe” between them. Instead, you can break the message up into small pieces, called packets, and pass them from one computer to another, heading in the direction from A to B. This strategy has tremendous advantages – for example, messages can be rerouted in the event one computer fails or becomes overloaded –  but this wasn’t apparent at the time. In fact, many people questioned whether or not Baran’s idea could work at all.

I’m moderately familiar with Baran’s ideas, at least at a high level, but when Jennifer displayed the cover of one of Baran’s reports on the screen, I was quite astonished by what I saw. Here’s the cover:

This was part of the 11-part series that established the effectiveness of packet switching, or, in the more colorful terminology used here, “hot-potato routing”. As I mentioned, there was far from universal acceptance of Baran’s proposal, and this paper was important because it contained a detailed description of a simulation, written in the programming language Fortran, that demonstrated not only that packet switching could work, but exactly how it could work. And look at the authors of the paper: Sharla P. Boehm and Paul Baran.

Who the heck was Sharla P. Boehm? Paul Baran is generally considered one of the four or five most important “Fathers of the Internet”, but who is Sharla Boehm? I can’t remember ever seeing the name before, and I was determined to try to do some research to find out. Her name is the first on this paper, so obviously she made an important contribution, but one that seems to have been largely forgotten.

It took a bit of Google sleuthing, but I was able to find at least a partial answer. Sharla Boehm was a programmer at Rand in the early 60’s,  married to Barry Boehm, who became a well-known computer science professor. Barry Boehm taught for many years at the University of Southern California, and was an important theorist in the development of the field of Software Engineering. (In fact, the name Boehm seemed familiar to me, and that’s because I had encountered Barry Boehm’s work in my time as a CS professor.)

In 1996, Barry Boehm wrote an interesting memoir about his early days in the software industry, entitled “An Early Application Generator and Other Recollections”.  He mentions, in passing:
Coincidentally, my wife Sharla had developed the original packet-switched network simulation with Paul Baran [Baran-Boehm, 1964]).
So there it is – Sharla Boehm wrote the code that demonstrated the feasibility of packed-switched networks. You can look up the original paper that she and Baran wrote, and read every line of code that she wrote and see the actual output from her program.

I couldn’t find any other references online to the work of Sharla Boehm. Perhaps she went on to write more software, or perhaps she gave that up and moved on to other things. It would not be realistic to equate her contribution to that of Paul Baran, but she was there at an important time doing important, difficult, complex work and apparently doing it well enough that Paul Baran put her name first on the document. (Ironically, even Barry Boehm puts Baran’s name first on the reference, and some references to the article don’t even include Boehm at all.)

So I salute Sharla Boehm and thank her for her contribution. She helped create the online world we live in today, and it’s wrong to overlook or forget what she accomplished.

Once I had found the connection between Barry and Sharla Boehm I searched to see what else I could find. It appears that they are both alive and well and living in Santa Monica, not far from Rand Corporation where they apparently met.  I did find one more thing – I searched Flickr to see if I could find a picture of Sharla, and found the photo below from a retirement party for Barry Boehm. There’s Sharla & Barry in front.  I am glad to have a chance to “put a face” to the name on some 50 year old code that helped to change the world.

(Addendum, 5/12/2015: more on Sharla's story...)