trb@ima.ISC.COM (Andrew Tannenbaum) (07/09/88)
> I've recently acquired a copy of the PWB/UNIX User's Manual in two volumes, > (one section looks like the current User's Manual and the other like the > Programmer's Manual) printed in May 1977. Can anyone tell me where PWB/UNIX > fits in to the UNIX history? I've never heard of it before. There were two major versions of PWB/UNIX, 1.0 was V6 based, 2.0 was v7 based. The PWB (Programmer's Workbench) group were UNIX developers whose major charter was to develop economical UNIX systems as front ends to real computers (like IBM, Univac, Honeywell), so that a system analyst could have one terminal on his desk instead of three terminals (or cardreaders), he could use one editor - one integrated set of tools. Some of the major contributions of PWB are the MM macros for troff and the UNIX RJE stuff (which was heavily used in Bell Labs, at least when I was there - until 1983). See the paper "The Programmer's Workbench" by Dolotta, Haight, and Mashey, in the classic UNIX BSTJ from the summer of 1978. While at MASSCOMP, I drew up this timeline of UNIX history until AT&T SVR2/4.2BSD. Andrew Tannenbaum Interactive Boston, MA +1 617 247 1155 <<>><<>><<>><<>><<>><<>><<>><<>><<>><<>><<>><<>><<>><<>><<>> '\" tbl | troff '\" need a blank line to init troff state .sp .5i .TS box; r l lbp+2 r l||l. Great Moments in UNIX System History = 1969 Thompson fiddles with filesystem ideas on Multics Thompson ports "space travel" from MULTICS to GECOS to PDP-7, hacks UNIX filesystem and... UNIX V1 written in assembly by Thompson for PDP-7&9 _ 1970 Thompson intends to write FORTRAN using TMG, writes B instead. After being denied for $.5M for a research DEC-10, Ossanna asks for $65K for a new PDP-11/20 for text processing research. Ordered in May, comes in August Runs UNIX in core until disk comes in December Kernighan suggests name UNIX _ 1971 Feb UNIX V2 PDP-11/20 in assembly language on unprotected hardware Work begins on C Summer BTL patent dept runs text processing on 11/20 UNIX Nov UNIX V1 manual _ 1972 Jun UNIX V2 manual UNIX gets pipes _ 1973 Feb V3 Manual Feb UNIX V3 PDP-11/34 40 45 60 70 multiprogramming C runs on Honeywell 6000, IBM 370 Oct UNIX first mentioned in public at ACM 4th SOSP Snyder Portable C MIT Masters Thesis PWB/UNIX conceived - RJE SCCS PWB/MM MERT designed for PDP-11/45 _ 1974 June V4 Manual July CACM UNIX paper "The UNIX Timesharing System" UNIX V5 university release _ 1975 May UNIX V6 manual _ 1976 UNIX V6 commercial release LSX LSI-11 UNIX (Minimum Configuration ~$7000) Feasibility study for UNIX under VM/370 at Princeton _ 1977 early Work begins on Interdata 8/32 UNIX port at Bell Labs Interdata 7/32 UNIX port at at U Wollongong Australia UNIX V6 commercial sublicenses USG UNIX Generic Issue 3 (UNIX V6 based) May PWB/UNIX 1.0 (UNIX V6 based) Bell Release Jun John Lions' commentary on UNIX V6 late MERT Release 0 _ 1978 Spring Johnson and Ritchie Interdata (later PE) port complete Jul Bell System Tech Journal UNIX issue, Part 2 Vol. 57 No. 6 current: USG UNIX, PWB/UNIX, UNIX V6, Research UNIX UNIX TS 1.0 (UNIX V7 based) Bell release UNIX 32V VAX port (UNIX V7 based) _ 1979 Jan UNIX V7 manual Apr UNIX RT (supersedes MERT) Jun PWB/UNIX 2.0 (UNIX TS based) Bell release Dec 3BSD UNIX VAX virtual memory _ 1980 Feb UNIX TS 1.3 VAX & PDP-11 Bell release Apr UNIVAC 1100 UNIX TS 2.0+ Bell release Jun UNIX TS 2.0 VAX only Bell release Jun UNIX 3.0 (replaces TS and PWB) Bell release Oct 4.0BSD UNIX _ 1981 March UNIX/370 Bell Release Mar 3B UNIX 3.0 Bell Release Jun 4.1BSD UNIX _ 1982 Jun UNIX 5.0 Bell release _ 1983 Aug 4.2BSD manual Dec UNIX System V Release 2 manual _ .T& r r r. Andy Tannenbaum MASSCOMP 9/84 .TE
jfh@rpp386.UUCP (John F. Haugh II) (07/10/88)
andy didn't include release 4.0, which as i recall was released in the summer of 1981. as far as i know it was only released internally and to some universities which had professors on loan from the labs. i had the misfortune of having to deal with it for a short time, prior to xavier (where i hung out) getting release 5.0 running on the 11/45. i can't say much about 4.0, except that it appeared to be closer to 5.0 than to 3.1 (did i get that right??? was it 3.2???) we were having hardware trouble so no one got to use the system much. as i recall, 4.0 still fit in 128K or so of memory. release 5.0 had to be trimmed down quite a ways, and it still took more than 128K. which leads to the next question. has anyone every managed to fit s5r2 on a pdp-11? - john. -- John F. Haugh II +--------- Cute Chocolate Quote --------- HASA, "S" Division | "USENET should not be confused with UUCP: killer!rpp386!jfh | something that matters, like CHOCOLATE" DOMAIN: jfh@rpp386.uucp | -- with my apologizes
kjk@pbhyf.PacBell.COM (Ken Keirnan) (07/11/88)
In article <3776@rpp386.UUCP> jfh@rpp386.UUCP (The Beach Bum) writes: >andy didn't include release 4.0, which as i recall was released in the >summer of 1981. as far as i know it was only released internally and to >some universities which had professors on loan from the labs. > >i had the misfortune of having to deal with it for a short time, prior >to xavier (where i hung out) getting release 5.0 running on the 11/45. >i can't say much about 4.0, except that it appeared to be closer to >5.0 than to 3.1 (did i get that right??? was it 3.2???) we were having >hardware trouble so no one got to use the system much. > >as i recall, 4.0 still fit in 128K or so of memory. release 5.0 had to >be trimmed down quite a ways, and it still took more than 128K. which >leads to the next question. has anyone every managed to fit s5r2 on a >pdp-11? First, you are correct about UNIX 4.0. This version was something of a "pre-release" of System V that was missing several major enhancements that eventually appeared in 5.0 (I don't remember the specifics), and yes, this version was only released within the Bell System (I still have 4.0 running on 2 pdp-11/70s). Second, System V release 2 was the last release of UNIX from Bell Labs to run on the pdp-11/70 (I don't know if it runs on the 11/45). The pdp-11 version of release 2 is missing several operating system features and utilities that are available on larger machines, but the "Labs" managed to fit the kernel text into 64K. Kernel data is larger than 64K, but is handled by using a combination of pdp-11's "kernel" and "supervisory" modes to map the necessary data space. In spite of all the memory management manipulation required, Sys VR2 runs great on the pdp-11/70. Yes, I still have an 11/70 running SVR2, and I still find it a challange to shoehorn large programs on to the system (Documentors Work Bench 2.0 was the most recent). Ken Keirnan -- Ken Keirnan - Pacific Bell - {att,bellcore,sun,ames,pyramid}!pacbell!pbhyf!kjk San Ramon, California kjk@pbhyf.PacBell.COM