Could Unix happen today?
Connecting state and local government leaders
Almost 40 years ago, a relatively few number of folks working on a shoestring budget came up with the Unix operating system, which in hindsight, could be seen as a revolutionary technology. Could that happen today? Maybe not with operating systems, but perhaps in other fields of software, said Brian Kernighan, one of the original developers of the OS.
Kernighan, now a professor at Princeton University, recounted the early development of Unix in a talk earlier this week hosted by the Washington D.C. chapter of the Association for Computing Machinery.
Because the concepts Unix brought to modern computing are so ubiquitous, it's hard appreciate how much the OS has engendered. In fact, Kernighan argued with admitted rhetorical flourish, nothing in computer science in the past 30+ years has matched Unix's impact. Hierarchical file structures, programmable shells, regular expressions, pipelining, program portability, programmable tools, high-level languages (such as Unix's close-cousin C) and even the very idea of computer security all got their start within Unix.
Even as he talked up Unix, Kernighan was modest about his own contributions (he did awk, among other utilities). At Bell Labs, he worked across the hall from whom he considered to be the two main creators of Unix, Ken Thompson and Dennie Ritchie (who is still at Bell Labs, by the way).
The original Unix only ran on 9,000 lines of code, Kernighan noted. In comparison, the latest Linux kernel has over 6 million lines of code Today, even the comments of a simple operation such as cat (a copy command) can run three MAN pages. 'It shouldn't be like that,' he said. 'Bigger is not necessarily better.'
How did Unix happen? AT&T commissioned the operating system to help with the specific problem of switching billions of telephone calls, but that doesn't explain its subsequent success in pretty much all other domains of computing.
That success could best be attributed to the working climate AT&T offered, Kernighan said. Four factors, almost accidental in nature, played a role: AT&T hired smart programmers, led them with good managers, offered a totally unfettered environment for these programmers and provided plenty of cheap hardware.
Kernighan said he doubted we will see that much innovation spring from another operating system, as the market is quite mature and the existing choices offer most of what people need. But thanks to the Internet and the availability of lots of cheap hardware and free software, we may see similar revolutionary technologies elsewhere. "For the kinds of things we would find fundamentally useful, it could definitely happen again," he said. --Posted by Joab Jackson