A good number of people I knew were at the Grace Hopper Celebration of Women in Computing last week. By all reports it was a great event, some cases of altitude sickness not withstanding. Hilary Pike has been blogging about it over at Springboard. I especially enjoyed her experienced take on how to get the most out of GHC. Randy Guthrie wrote about what it was like to be one of the very few male attendees. Several other bloggers I follow have been blogging about the event as well. For example See Jane Compute! and Lynn Langit.
This all makes me reminisce about Grace Hopper herself. I was fortunate enough to hear her talk several times over the years. The first time when I was a student in the early 1970s. I believe she had just made Captain about that time. She was on the CS department advisory of the school (Taylor University) I was attending. Several of us students were able to have a casual lunch and conversation during her visit to campus. It was an unforgettable time. I remember one of my classmates asked her why she had joined the Navy. Captain Hopper looked at her, blinked a couple of times and said “why there was a war on.” as if that explained everything. Clearly it was enough explanation for her. She was as proud of being a Naval officer as anything else in her storied career.
Grace Hopper had a great way of explaining things to people. Most people are familiar with her use of pieces of wire that she called nano seconds. Later in life she talked about pico seconds and gave out samples of those. She acquired those samples in the cafeteria where they came in little paper envelopes helpfully labeled with the letter “P.” But I loved how she explained the need for attacking big computing problems with multiple computers.
She explained that back in the horse and oxen drawn plow days there were limits to how big one could breed horses and oxen. Eventually the answer was to add a second horse or ox. So to would we reach the limits to the amount of processing computer we could put into one computer. We would increasingly have to add computers. This was back in the 1970s were parallel computers were mostly a research idea. Smaller computers were starting to be used as front ends to handle communication as well. But mostly we were still trying to build faster and faster super computers as single processors. Timing and future developments aside this remains a great analogy as to why we need to work on multi processor systems. And just as tack had to change and improve for multiple draft animals so does software have to change and improve to handle multiple processors.
Most people remember Grace Hopper as this kindly grandmotherly person. And she was often that. He crew at the Navy called her grandma in an affectionate way. At the same time she could be as rough and as tough as they come. I talked to people are some of the companies that were trying to get their COBOL compilers passed through her office. They found her demanding and insistent. These people did not look forward to calls with Grace Hopper on the other end. She had a job to do for her country and she let no one and no thing stand in her way. It is quite fitting, and I’ve said this before, that the ship the Navy named after her is a real honest to goodness war ship.
She was an inspiration to me and still probably the most impressive person I have ever met. I can think of no one better to honor with the Celebration of Women in Computing.