The Economist has covered the early days of Unix, searching for why Unix and its companion language, C, became so important. They cover it fairly well.
Each successful technology carries its lessons; indeed it is hard to see how a successful technology cannot bring forth some new lessons, just in order to beat the status quo. The cleanliness of the Java virtual machine over the evils of C's buffer overflows and core dumps, for example.
The article didn't get it all, of course. One of the results of the beauty of the design was that the doco was rather small. I recall an experienced programmer, John, standing up in an early user group meeting holding up the manual. "I can take this home in my briefcase," he said, whereas other operating systems filled boxes and book shelves.
Of course, that bug was eventually fixed, and Unix joined its competitors in shipping with boxes of manuals.
Another artifact of the nominal or free delivery to Universities was that by the time the late 80s came around, there were enough software engineers that had trained at University that Unix was an easy sell. When selecting an OS for some project, it was pretty easy to say, "I want something like Unix." And that's one of the two punches that defeated IBM, the other being the PC.
The article also glosses over the role of BSD. It became the competitor, and it also became the one that fixed the (academic) shortfalls for the (commercial) real world. Virtual memory being the chief of those, but also more features, and less security. It's mostly forgotten these days, but BSD blazed the insecurity trail by mostly ignoring issues that the Bell Labs people thought sacrosanct, in the quest for more features.
Yes, we in the Bell labs religion hated the Berkeley religion, but it became the one to open up the net and practically every other application. Partly because the Californians didn't care about security, and partly because they could see the benefit in fast and furious communications. Of course, there were a dozen other lessons, and that's the story of the Internet, not of Unix, but the parallels with Microsoft's current day dilemma are worth pondering over.
On a closing note,
"Dr Pike says that the thing he misses most from the 1970s at Bell Labs was the terminal room."
It's true, I grew up in those places, and now that I think about it, we are only just now recapturing that spirit. The article seems to think that the open source community has inherited the legacy of Unix's collaborative spirit, but I disclaim that: email is too clunky.
Only with IM and potentially VoIP are we at the verge of re-creating the terminal room.
Posted by iang at July 27, 2004 05:37 AM | TrackBack