Forgot your password?
Cyhist Apr 16 1997 E
Document Actions
Date: Wed, 16 Apr 1997 23:12:00 PDT
Reply-To: "CYHIST Community Memory: Discussion list on the History of
Cyberspace" <CYHIST@MAELSTROM.STJOHNS.EDU> Sender: "CYHIST Community Memory: Discussion list on the History of
Cyberspace" <CYHIST@MAELSTROM.STJOHNS.EDU> From: Lauren Weinstein <lauren@VORTEX.COM>
Subject: V6 Unix
______________________________________________________________________
Community Memory: Discussion List on the History of Cyberspace ______________________________________________________________________
Greetings. The single aspect of the V6 Unix C compiler that stands out most in my mind was the lack of any "long" type. This limited ints and unsigned to 16 bits, which frequently became a bit troublesome. PWB Unix ("Programmer's Workbench") which was released by the Labs between V6 and V7, was the first Unix that included a compiler that supported longs, as I recall.
The removable disk cartridges mentioned by a previous poster would have been the ubiquitous RK01's, which were the workhorse of the 11/45, along with DECtape. I actually remember having an 11/45 running (on ARPANET, providing net services) using only a pair of DECtapes (one for swap!) at a time when the disk was down (I still have some DECtapes around). Anyone remember the programming technique for blowing the fuse on the DECtape controller? Oh never mind.
It did indeed seem that the subjective amount of time you sat around waiting for things to happen was similar to what it is now in many cases. But programs (and especially libraries) were usually smaller, network throughputs much less (remember that the early ARPANET backbone was 56Kps), and what we *expected* to get done was much reduced from what it is now. Expectations drive perceptions, and the amount of time the average human is willing to sit around staring at a screen, waiting for something to happen, remains much the same.
Unix (and its presumptive kin, like Linux) are still in many fundamental ways much like the Unix we knew at the beginning. Yes, saying "rm * .o" instead of "rm *.o" to the shell is still a poor idea in many default environments. But as with the Unix of old, virtually any and all of these environments could be customized to the heart's content of the facility or user. That flexibility springs from an elegant system built almost entirely on an elegant high level language (C). Both system and language could be documented in a set of printouts only a few inches thick, instead of with linear feet of manuals weighing down heavy bookshelves for the other systems that ran on those same minicomputer platforms.
These are some of the factors that made Ken and Dennis' design, elaborated over time by many within and outside Bell Labs, a work that fundamentally changed the direction of computing in a widening sphere that continues to this day.
--Lauren--
______________________________________________________________________
Reply-To: "CYHIST Community Memory: Discussion list on the History of
Cyberspace" <CYHIST@MAELSTROM.STJOHNS.EDU> Sender: "CYHIST Community Memory: Discussion list on the History of
Cyberspace" <CYHIST@MAELSTROM.STJOHNS.EDU> From: Lauren Weinstein <lauren@VORTEX.COM>
Subject: V6 Unix
______________________________________________________________________
Community Memory: Discussion List on the History of Cyberspace ______________________________________________________________________
Greetings. The single aspect of the V6 Unix C compiler that stands out most in my mind was the lack of any "long" type. This limited ints and unsigned to 16 bits, which frequently became a bit troublesome. PWB Unix ("Programmer's Workbench") which was released by the Labs between V6 and V7, was the first Unix that included a compiler that supported longs, as I recall.
The removable disk cartridges mentioned by a previous poster would have been the ubiquitous RK01's, which were the workhorse of the 11/45, along with DECtape. I actually remember having an 11/45 running (on ARPANET, providing net services) using only a pair of DECtapes (one for swap!) at a time when the disk was down (I still have some DECtapes around). Anyone remember the programming technique for blowing the fuse on the DECtape controller? Oh never mind.
It did indeed seem that the subjective amount of time you sat around waiting for things to happen was similar to what it is now in many cases. But programs (and especially libraries) were usually smaller, network throughputs much less (remember that the early ARPANET backbone was 56Kps), and what we *expected* to get done was much reduced from what it is now. Expectations drive perceptions, and the amount of time the average human is willing to sit around staring at a screen, waiting for something to happen, remains much the same.
Unix (and its presumptive kin, like Linux) are still in many fundamental ways much like the Unix we knew at the beginning. Yes, saying "rm * .o" instead of "rm *.o" to the shell is still a poor idea in many default environments. But as with the Unix of old, virtually any and all of these environments could be customized to the heart's content of the facility or user. That flexibility springs from an elegant system built almost entirely on an elegant high level language (C). Both system and language could be documented in a set of printouts only a few inches thick, instead of with linear feet of manuals weighing down heavy bookshelves for the other systems that ran on those same minicomputer platforms.
These are some of the factors that made Ken and Dennis' design, elaborated over time by many within and outside Bell Labs, a work that fundamentally changed the direction of computing in a widening sphere that continues to this day.
--Lauren--
______________________________________________________________________
