Thoughts on Our Legacy

The theme of this site is designed to resemble the early days of civilian computing at places like Bell Labs and Xerox PARC.

Since it's so long ago as to be alien to 21st century technicians, it's worth remembering and honoring the real origins of computing as we know it.

Before graphical computing, before the personal computer, was the mainframe era. Most technicians would be familiar with the oldest stories of punch-card machines, but that's not so much what I am interested in with this site.

I'm interested more in the slightly later era of computing which arrived with the dawn of interactive systems. A lot of modern computer terminology and functionality dates to this era, especially in the POSIX (read: everything except Windows) space.

The TTY name for a terminal means teletypewriter, which is literally a typewriter with network access, complete with rolls of paper spooling out the top. Terminal control characters like the carriage return, line feed, and bell also date from this era, explaining their anachronistic naming (to be fair, most modern documents describe LF as "newline"). The baud rate of a terminal - which is usually set to the maximum allowed by the standard on modern virtual terminals - used to be more meaningful, since it was the data rate the terminal could output at.

The way files are read in as 'blocks' from start to end - which is fairly baffling on modern hardware - makes a lot more sense when you consider 60s/70s mainframes using magnetic tape reels. My mom remembers waiting for hours for video games to load from magnetic tapes on consumer hardware even into the 1980s.

Along with these fairly obvious technological changes, there have been huge changes in the ideology of computing over the years. It's worth remembering that the early pioneers of computing existed in a totally different cultural moment to that of the post-Snowden, surveillance age we live in now. Technologists in the 1960s would have been familiar with Marshall McLuhan's The Medium is the Message, as well as first and second wave cybernetics literature.

In both cases, the primary motivation of technologists were more aligned with what we'd today call open-source and hacker culture rather than maximizing profitability.

And here I have to confess my own biases: as someone who came into computing from first personal interests, then interdisciplinary academic work, and later the London digital art scene, naturally I prefer the academic + socially activist narratives about computing to the narrative that presents computing as necessarily serving power. But it's important to acknowledge that both forces shaped early computing, and for my money the spirit of what would later become hacker culture was the more important one.

Anyway, the cybernetics field is about formal models of systems considered in the broad sense. The concepts of positive feedback (leading to outsized results from small inputs) and negative feedback (leading to a self-regulating system) are the most familiar examples. Positive feedback is often associated with disaster - like a heater continuing to heat something that's already over its operating capacity leading to nuclear reactor disasters - but it's not always negative. Consider compound interest. Over time, money tends to be worth an exponentially increasing amount, meaning small investments today can lead to big returns in 30 years.

At some point after the mainframe era, discourse turned to how the positive feedback effect could be applied to knowledge systems and learning. And it's here that one of the most important pioneers of early personal computing emerged.

Douglas Engelbart and Bootstrapping Theory

One of my great inspirations, who I wrote about fairly extensively when studying for my MA, was Douglas Engelbart.

Engelbart is today most widely known as the inventor of the computer mouse, and usually receives a small spot in museum exhibits about the history of computing in this capacity. But this is understating his (or more accurately, his team's) importance to the development of personal computing. It also discounts some of his most important ideas around bootstrapping and recursive self-improvement.