From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.2 (2018-09-13) on inbox.vuxu.org X-Spam-Level: X-Spam-Status: No, score=-0.6 required=5.0 tests=DKIM_INVALID,DKIM_SIGNED, HEADER_FROM_DIFFERENT_DOMAINS,HTML_MESSAGE,MAILING_LIST_MULTI, RCVD_IN_DNSWL_NONE autolearn=ham autolearn_force=no version=3.4.2 Received: from minnie.tuhs.org (minnie.tuhs.org [45.79.103.53]) by inbox.vuxu.org (OpenSMTPD) with ESMTP id 8489f341 for ; Wed, 11 Sep 2019 00:49:54 +0000 (UTC) Received: by minnie.tuhs.org (Postfix, from userid 112) id A5F7494790; Wed, 11 Sep 2019 10:49:53 +1000 (AEST) Received: from minnie.tuhs.org (localhost [127.0.0.1]) by minnie.tuhs.org (Postfix) with ESMTP id BF76F9478D; Wed, 11 Sep 2019 10:48:48 +1000 (AEST) Authentication-Results: minnie.tuhs.org; dkim=fail reason="signature verification failed" (1024-bit key; unprotected) header.d=yaccman.com header.i=@yaccman.com header.b="WHD4G33B"; dkim-atps=neutral Received: by minnie.tuhs.org (Postfix, from userid 112) id 1644B9478D; Wed, 11 Sep 2019 10:48:45 +1000 (AEST) X-Greylist: delayed 1195 seconds by postgrey-1.36 at minnie.tuhs.org; Wed, 11 Sep 2019 10:48:43 AEST Received: from baboon.elm.relay.mailchannels.net (baboon.elm.relay.mailchannels.net [23.83.212.8]) by minnie.tuhs.org (Postfix) with ESMTPS id 73FB593D35 for ; Wed, 11 Sep 2019 10:48:43 +1000 (AEST) X-Sender-Id: dreamhost|x-authsender|scj@yaccman.com Received: from relay.mailchannels.net (localhost [127.0.0.1]) by relay.mailchannels.net (Postfix) with ESMTP id 3CAA8500E09; Wed, 11 Sep 2019 00:28:47 +0000 (UTC) Received: from pdx1-sub0-mail-a6.g.dreamhost.com (100-96-6-79.trex.outbound.svc.cluster.local [100.96.6.79]) (Authenticated sender: dreamhost) by relay.mailchannels.net (Postfix) with ESMTPA id DA457500ADE; Wed, 11 Sep 2019 00:28:45 +0000 (UTC) X-Sender-Id: dreamhost|x-authsender|scj@yaccman.com Received: from pdx1-sub0-mail-a6.g.dreamhost.com ([TEMPUNAVAIL]. [64.90.62.162]) (using TLSv1.2 with cipher DHE-RSA-AES256-GCM-SHA384) by 0.0.0.0:2500 (trex/5.17.5); Wed, 11 Sep 2019 00:28:47 +0000 X-MC-Relay: Junk X-MailChannels-SenderId: dreamhost|x-authsender|scj@yaccman.com X-MailChannels-Auth-Id: dreamhost X-Soft-Exultant: 1daec8a521a747c3_1568161726394_2558276878 X-MC-Loop-Signature: 1568161726393:2868626486 X-MC-Ingress-Time: 1568161726393 Received: from pdx1-sub0-mail-a6.g.dreamhost.com (localhost [127.0.0.1]) by pdx1-sub0-mail-a6.g.dreamhost.com (Postfix) with ESMTP id E2FFB7F47E; Tue, 10 Sep 2019 17:28:42 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha1; c=relaxed; d=yaccman.com; h=message-id :from:to:cc:in-reply-to:subject:date:content-type:mime-version; s=yaccman.com; bh=/i7/kp/Y3Nq7NCmULf8tZicZ4Oc=; b=WHD4G33BnwVrH 6U//WvSFNsul9SBBf2i1nbOOQAfFVRgkr2VsBMu+Apy7UyHzdI0Y7rtJp+RTjSFF 6isndfMuX2cGEdcu+ky9U2vJPXWKje47tz0Mj1lBCJ/WAwGUUZyoKHpWg7uADWLQ 6A+UVtcbxDYRzQMvAae06ootyt4diI= Received: from localhost (ip-66-33-200-4.dreamhost.com [66.33.200.4]) (using TLSv1 with cipher ECDHE-RSA-AES256-SHA (256/256 bits)) (No client certificate requested) (Authenticated sender: scj@yaccman.com) by pdx1-sub0-mail-a6.g.dreamhost.com (Postfix) with ESMTPSA id 058C27F47F; Tue, 10 Sep 2019 17:28:41 -0700 (PDT) Message-Id: <99b89a5f55b82dcf67f7bebe92b752a5b0e0a72d@webmail.yaccman.com> X-DH-BACKEND: pdx1-sub0-mail-a6 From: "Steve Johnson" To: "Clem Cole" , "Warner Losh" X-Mailer: Atmail 7.8.0.2 X-Originating-IP: 10.35.43.154 in-reply-to: Date: Tue, 10 Sep 2019 17:28:41 -0700 Content-Type: multipart/alternative; boundary="=_46f365b54b2cc00295962d7cd72f8800" MIME-Version: 1.0 X-VR-OUT-STATUS: OK X-VR-OUT-SCORE: -100 X-VR-OUT-SPAMCAUSE: gggruggvucftvghtrhhoucdtuddrgedufedrtddugdefhecutefuodetggdotefrodftvfcurfhrohhfihhlvgemucggtfgfnhhsuhgsshgtrhhisggvpdfftffgtefojffquffvnecuuegrihhlohhuthemuceftddtnecusecvtfgvtghiphhivghnthhsucdlqddutddtmdenucfjughrpefkhffvofhijgfuffgtggesrgdtreerredtjeenucfhrhhomhepfdfuthgvvhgvucflohhhnhhsohhnfdcuoehstghjseihrggttghmrghnrdgtohhmqeenucfkphepieeirdeffedrvddttddrgedpuddtrdefhedrgeefrdduheegnecurfgrrhgrmhepmhhouggvpehsmhhtphdphhgvlhhopehlohgtrghlhhhoshhtpdhinhgvthepieeirdeffedrvddttddrgedprhgvthhurhhnqdhprghthhepfdfuthgvvhgvucflohhhnhhsohhnfdcuoehstghjseihrggttghmrghnrdgtohhmqedpmhgrihhlfhhrohhmpehstghjseihrggttghmrghnrdgtohhmpdhnrhgtphhtthhopehstghjseihrggttghmrghnrdgtohhmnecuvehluhhsthgvrhfuihiivgeptd Subject: Re: [TUHS] PWB vs Unix/TS X-BeenThere: tuhs@minnie.tuhs.org X-Mailman-Version: 2.1.26 Precedence: list List-Id: The Unix Heritage Society mailing list List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Cc: The Eunuchs Hysterical Society Errors-To: tuhs-bounces@minnie.tuhs.org Sender: "TUHS" --=_46f365b54b2cc00295962d7cd72f8800 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: quoted-printable Well, I can share a bit of the piece of the elephant(s) I saw.=0A=0AWhil= e most of the Unix folks were PDP-11 only (after MULTICS), I had=0Aspent= a couple of years helping to run the MH comp center.=C2=A0=C2=A0 The=0A= system for the IBM 7094 was among the best anywhere -- card decks were= =0Astaged to tape and the overhead of changing from one job to the next= =0Awas minimized.=C2=A0 Most modest-sized job submissions were finished= in=0Aroughly an hour.=C2=A0=C2=A0 When MULTICS started, the 7094 went a= way and it=0Awas a huge jolt.=C2=A0 There were some amazing programs (e.= g., a LISP=0Acompiler written in MACRO FAP that would go 250 levels deep= to produce=0Agood code).=C2=A0 There were acoustic simulation programs= that were way=0Aahead of their time.=C2=A0 And they were all lost.=0A= =0AWith MULTICS still way in the future and no more 7094, a software=0Ap= ackage was put together to allow the MULTICS machines to run the=0Acomme= rcial GCOS system (the software was called the K package, K=0Astanding f= or kluge.=C2=A0 It was well named...)=C2=A0=C2=A0 Anyhow, they needed=0A= help, and asked me to help, and I did so for a couple of years,=0Alargel= y supporting FORTRAN.=C2=A0=C2=A0 Meanwhile, Dennis was inventing B, and= =0Athere was a limping version for the GE.=C2=A0 He may have actually us= ed it=0Aas part of a bootstrap of the language -- I'm not sure.=C2=A0 So= I=0Adiscovered it, and started to use it to write some system admin=0Ap= rograms.=C2=A0 I soon ran into some limitations, and started to lobby=0A= Dennis for some changes.=C2=A0 He listened politely, and then said "Why= =0Adon't you fix it.=C2=A0 The compiler is just a program, after all..."= =C2=A0=C2=A0=0ASo I put some things into B, including support for the GE= native 6-bit=0Acharacter set.=C2=A0 This eventually found its way into= C and CPP.=C2=A0=0ADecades later, if you wrote `abc` in the DEC C compi= ler, it produced=0AGE strings...=C2=A0=C2=A0 I also made it possible to= call FORTRAN from C (the=0Aother way was a bit harder...).=0A=0AEventua= lly they set up an organization to formally own running the=0Acomp cente= r, and I was invited to return to Research and I did.=C2=A0=C2=A0=0ABut= I had gained a lot of sensitivity to the needs of Engineers and=0AScien= tists just wanting to get their work done. =0A=0AYears later, when I was= asked to come over to Summit and head the=0ASystem V language departmen= t, V7 was behind us, and the Interdata port=0Ahad taught us a lot about= portability.=C2=A0 PCC had proved itself on a=0Ahalf dozen machines and= had been well received.=C2=A0 But most of the Bell=0ASystem was using 1= 6-bit DEC machines and Dennis' compiler.=C2=A0 C++ was=0Aunder develpmen= t and looked like a clear winner, but it needed a lot=0Aof work.=C2=A0 T= he debugging of C++ was horrible -- a program, Cfront,=0Aconverted C++ t= o C, but in the process screwed up the line numbers and=0Anames so that= using a debugger was almost impossible. =C2=A0 My three=0Agoals in goin= g to Summit were to make C++ commercially available, to=0Aprovide a debu= gging technology that could handle not just C++ and C=0Abut FORTRAN and= ADA and PASCAL, and to base the code generation on the=0APCC2 model to= gain portability.=C2=A0 Although AT&T's computer business=0Awas pretty= much a disaster, I felt that I had met these goals=C2=A0 -- Elf=0Aand D= warf, C++ debugging, and the first commercial C++ product=0Ahappened on= my watch.=C2=A0 I also had a lot of fun working with a bright=0Aand div= erse group of talented people.=0A=0ASadly, as the AT&T computer business= collapsed, I realized that I=0Aloved doing advanced development, but no= t at ATT.=C2=A0 My final straw was=0Awhen an excellent QA team leader at= another location was fired for=0Atelling the truth about how lousy one= of the machines performed.=C2=A0 So=0Ait was off to Silicon Valley for= me.=0A=0ASteve=0A=0A----- Original Message -----=0AFrom:=0A "Clem Cole"= =0A=0ATo:=0A"Warner Losh" =0ACc:=0A"The= Eunuchs Hysterical Society" =0ASent:=0ATue, 10 Sep 2019= 11:16:34 -0400=0ASubject:=0ARe: [TUHS] PWB vs Unix/TS=0A=0ABelow ... fr= om memory - Someone like APS=C2=A0was a little closer to some=0Aof this= than I was, so I might=C2=A0have a few things wrong.=C2=A0 I don't=0Ath= ink so, but it's been quite a few beers=0A=0AOn Mon, Sep 9, 2019 at 2:26= AM Warner Losh wrote:=0A=0AOK. I'm totally confuse= d, and I see contradictory information around.=0ASo I thought I'd ask he= re.=0A=0APWB was started to support unix time sharing at bell labs in 19= 73=0A(around V4 time).=0A=0ANo...=C2=A0 that is not quite right.=C2=A0 P= WB was Mashey's project to build=0Aan RJE system to front end SW develop= ment for the IBM systems, which=0AAT&T had a number [IIRC Call Accountin= g and lot of the 'business' part=0Aof the Bell System was mainframe base= d].=C2=A0 I think Dick Haight was=0Aalso involved.=C2=A0 I've forgotten= the site there were at.=C2=A0 It might=0Ahave been Holmdel or Whippany.= But it was not MH or Summit.=0A=0A=C2=A0=0A=0APWB 1.0 was released just= after V6 "based on" it.=0A=0AWell not so much "right after", but it was= based on V6.=C2=A0 There are=0Adifferences.=C2=A0 IIRC this was the fir= st attempt at redoing how groups=0Aworked.=C2=A0 The biggest=C2=A0=0A=0A= additions were=0A=0A an IBM RJE support, =0ASCCS and a different set of= backup utilities; =0Aincluding some disk to disk (volcpy)=C2=A0=0Aand t= he original binary formatted program for 9-tracks (=0Acpio)=0A to replac= e Ken's assembler based =0Atp.=0A=0ASCCS was important and the RJE suppo= rt was important because that was=0Athe system being used and it made a= huge impression on AT&T staff.=C2=A0=0A=C2=A0A terminal to a UNIX box w= as way cheaper and to the IBM and people=0Awere so much more productive.= =0A=0AAlso remember, that tp(1) was written in assembler had been origin= ally=0Atargeted to DECtape in a very early version of Research UNIX.=C2= =A0 The=0ADECtape nature is why the directory was on the front of the ta= pe.=C2=A0=0AKen moved it 9-track but used the same tape format.=C2=A0= =C2=A0I don't=0Aremember who=0A=0A wrote stp (super-tp - in C), [?? Har= vard ?? it's on the Harvard tape=0Aand is how I got it].=0A=C2=A0=C2=A0= But better peripheral support was really important in Mashey's=0Asettin= g.=C2=A0 In that world, the production computer system was being put=0Ai= n the raised floor computer rooms next to a mainframe and they had=0A'op= erators' so John and team started to think more about what was=0Aneeded= to admin the system.=C2=A0 =C2=A0IMHO: this was the first heavy use of= =0Ashell scripts, while I saw them in MH, it was Mashey's guys that caus= e=0Ame personally to have an ah-ha moment about them.=0A=0AInterestingly= enough, and I have talked to Bourne and Mashey about it,=0AJohn's use o= f the V6 was definitely one of the groups that were asking=0Afor a new s= hell, which Bourne set out to solve; but that is not yet=0Aavailable.=0A= =0AAt some point (and here is where we need Steve Johnson, aps, and I=0A= wish the late Ted Kowalski) to fill in details I can not.=C2=A0 USG/Summ= it=0Awas chartered to "support UNIX for the Bell System."=C2=A0 =C2=A0As= I=0Aunderstand it, the genesis for their system was a kernel from MH th= at=0Awas moving towards V7s but not there yet, the 'Typesetter C' and a= =0Abunch of other utilities that Summit had collected/developed, but=0Aw= hich I do not know.=C2=A0 I think fsdb was around by that time. The new= =0ABourne Shell and adb were being developed although how complete I'm= =0Anot sure.=0A=0ABut accept for the new shell and updated compiler, I r= emember the=0Asystem 'felt' like V6 (Thompson shell) and thinking how mu= ch 'better'=0Adifferent v7 (Bourne Shell was) when we finally got it. Th= is earlier=0Asystem is the one Ted brought to CMU in the fall 1977 (I th= ink that is=0Athe right date) to update the V6 system were then running.= =C2=A0 Anyway,=0ATed always referred to this as a UNIX/TS kernel.=0A=0AA= nother thing we did not have SCCS or the RJE stuff.=C2=A0=C2=A0=0A=0AWha= t I'm not sure of is if there was a formally release of what ted=0Ahad= =0A.=C2=A0 So it may have been that TS had them and sent the release to= =0AMashey, although I don't think there were such releases originally in= =0ATS.=C2=A0 FWIW:=0A=C2=A0I believe that in our (CMU) case,=0ATed=0A=C2= =A0would just grab things as they appeared that he thought =0Awe needed= at CMU =0Aand he pushed things back (like CMU's =0Afsck as he found thi= ngs we had that he thought we would like).=C2=A0=0AInterestingly enough,= RJE and SCCS was needed for the IBM support and=0Awhile Ted (and his un= dergrad roommate, Bill Joy) had worked on the MTS=0Asystem on the IBM's= at UMich, I always felt like Ted looked down on=0Athe mainframes (which= was were I had also emerged but from CMU's TSS=0Ateam).=0A=0AAlso,=0A T= ed was a die-hard original =0Acpio user and I liked the user interface t= o stp, which I remember was=0Aa difference/source of argument=0A.=C2=A0= =C2=A0Tar did not yet exist. TS had some of the PWB tools like volcpy;= =0Abut we were using DOS-=0A11's similar but different backup =0Ascheme= (I've forgotten the name of the format; but the tapes were=0Aboot-able,= which=C2=A0volcpy=C2=A0tapes were not).=0A=0AFWIW:=C2=A0 cu(1) did not= yet exist.=C2=A0 I wrote a program (that I tended to=0Aprefer in some w= ays for many years) called connect(1cmu) that did the=0Asame thing.=C2= =A0 We used it to download images to the Microprocessors=0Alike the KIM-= 1.=C2=A0 =C2=A0It was originally written with the v6 portable C=0Alibrar= y, which is also what the original fsck used (it's what we had=0Aon v6).= =C2=A0 =C2=A0Ted introduced me to what would become stdio and one of my= =0Afirst tasks was using it with connect(1cmu).=C2=A0 The other thing I= =0Aremember about that program is it was the first time I wrote somethin= g=0Athat used two separate processes on a UNIX system that cooperated wi= th=0Aeach=C2=A0other and found it so much easier than on the PDP-10.=0A= =0AAlso, Dennis' stand-alone system for V7 was not yet available BTW.=C2= =A0=0A=C2=A0If I think of anything else about that system I can remember= , I'll=0Asend an update=0A=0APWB 2.0 was released just after V7, also "b= ased on it".=0A=0AI think the confusion is that TS and V7 were done sort= of at the same=0Atime and while the folks working on them talked to eac= h other,=C2=A0it has=0Anever been clear to me who was behind TS. For ins= tance, I would learn=0Athat Bourne was the 'project leader' for Seventh,= in that he was the=0Aperson that collected everything for it.=C2=A0 I n= ever heard of someone=0Ahaving the same role for TS, which is why I some= times think it was a=0Aname inside of Summit, but never actually saw the= light of day as a=0Aformal release.=C2=A0 =C2=A0I really am not sure an= d would love to learn more=0Adetails (I wish Ted were still alive to fil= l us in).=0A=0AAs for V7 itself, Ken wrote tar(1)=0A in response to cpio= (preferring an ASCII based header, but=0A'threading' it like cpio did,= but keeping the user interface that=0Atp/stp had).=C2=A0 As I understan= d it, Dennis built up did the standalone=0Atoolkit stuff.=C2=A0 Ken chan= ged groups and messed with the file =0Asystem =0Ain =0Athe kernel.=C2=A0= Lo=0A=0Ats of new peripheral support, which is why he also added lseek(= ) as=0Adisks overflowed a 16-bit integer for the seek position=0A.=C2=A0= Plus there were a number of other small changes between v6 and=0Av7.=C2= =A0 Some of this stuff from PWB and Summit went back to MH (fsck as=0Aan= example), but not everything (like cpio/volcpy/SCCS).=C2=A0 I kind of= =0Athink of the kernel and Typesetter C going from MH to Summit and the= =0APWB teams.=0A=0A@Steve Johnson, I need your help here.... at some poi= nt PCC was=0Acreated in MH (along with lint).=C2=A0 Didn't that start on= V6 but was not=0Acomplete until V7? And when did you move to Summit to= lead the=0Acompiler effort there?=C2=A0 My impressions that was yet to= happen, but=0AI'm fuzzy on dates.=0A=0ARemember, there are a number of= teams at BTL hacking on UNIX by=0Athen.=C2=A0 Dale's team in Columbus,= the crew in Indiana=C2=A0Hill,=C2=A0 folks at=0AWestern Electric (the T= eletype folks ported the Ritchie C to the Z80=0Aat some point for instan= ce),=C2=A0_etc._=0A=0AAgain, I don't remember the politics but like any= big company, you can=0Aimagine it was not all that clean and crisp.=C2= =A0 =C2=A0PWB 2.0 & 3.0=0Adefinitely picked up features from other UNIX= systems.=C2=A0 As I=0Aremember, Dale's shared memory hacks would beget= System V Shared Mem,=0ASemaphores and IPC (they are different, but they= started in Columbus).=0A=0AThe other thing I'm not clear on is when the= PWB team was folded into=0AUSG (=0A=0AUnix Support Group)=0A=C2=A0in Su= mmit.=C2=A0 =0A_I believe_ that was after PWB 2.0 was released.=C2=A0=C2= =A0=0ABut at some point, Mashey's team and the USG got interwoven.=C2=A0= I=0Areally don't know/remember many of those details as I watched them= =0Afrom the outside and only knew the results.=C2=A0 The key point is th= e PWB=0A2.0 would eventually be released as the internal, but=C2=A0=0Aof= ficial=0A UNIX for the Bell System.=C2=A0 =C2=A0It was supposed to bring= together the=0Aneeded from the different labs; but it was not >>=0Aoffi= cially=0A<< released _outside of the Bell System_ (it was an internal pr= oduct,=0Aremember at this point, AT&T is not allowed=0A=C2=A0to have com= puter products, etc...)=C2=A0=0A=0ASo PWB 2.0 is basically internal, and= a melding of V7, TS, PWB 1.0 and=0Astarting to take things from differe= nt labs with in=C2=A0BTL -- different=0Afrom all of them but mostly a su= perset.=0A=0A=C2=A0=0A Later Unix TS 3.0 would become System III.=0A=0AN= o --I do not think this is a true statement... not sure where you got=0A= that, m=0Aore in a minute=0A=0AWe know there was no System I or System I= I. =0A=0ACorrect.=C2=A0 =0A=0ABut was there a Unix TS 1.0 and 2.0? =0A= =0AThis is where it gets sticky.=C2=A0 I don't think so.=C2=A0 =C2=A0TS= was the=0Aoriginal work by USG.=C2=A0 =C2=A0What I do not know is if it= ever was=0A'packaged' as PWB had been. _I do not believe it was_.=C2=A0= =C2=A0I think a=0Alittle like the way Research 'bled' out a little a ti= me, pieces of TS=0Amade their way to MIT, CMU,=C2=A0_etc_. but never as= a formal release.=0A=0A=C2=A0=0AAnd were they the same thing as PWB 1.0= and 2.0, or somehow just=0Aclosely related? =0A=0ASee above... I'll exp= lain how PWB 3.0 became System III in a minute.=0A=0A=C2=A0=0A=0AAnd I'v= e seen both Unix/TS and Unix TS. Is there a preferred spelling?=0A=0ADon= 't know.=C2=A0 I remember Ted always called it UNIX/TS all caps.=0A=0ATh= e thing you left out is how PWB 3.0 became System III.=0A=0ATwo importan= t issues.=C2=A0 First with V7, AT&T (Al Arms) wrote the first=0Abinary s= ystem redistribution license.=C2=A0 The commercial folks were=0Ahappy to= have a redistribution license, but the terms were not what=0Athey reall= y needed.=C2=A0 Much of the issue was that AT&T was not the=0Acomputer h= ardware or software business and really did not understand=0Athe issues= that the vendors had.=C2=A0 Professor Dennis Allison of=0AStanford, was= consulting=C2=A0for almost all of us in the computer=0Aindustry at the= time (for those that don't know Dennis, around the=0Asame time he found= ed what is now called the Asilomar Microprocessor=0AWorkshop (check out:= =C2=A0=0Ahttps://www.computerhistory.org/atchm/the-asilomar-microcompute= r-workshop-and-the-billion-dollar-toilet-seat/=0A[2]).=0A=0ADennis arran= ged for a big meeting at Ricki's Hyatt in Palo Alto and=0Ainvited Al Arm= s and team, plus a representatives from his clients. I=0Awas the techie= with a lawyer from Tektronix in the room (as I have=0Asaid in other ema= ils this it is only time I have been in a meeting=0Awith Bill Gates).=C2= =A0 The folks I remember who were there: was Bill=0AMunson and team from= DEC; Fred Clegg and Team from HP; Bob MetCalfe=0Afrom 3Com; Gates and t= he MSFT crew; folks from SCO and DG.=C2=A0 =C2=A0There=0Awere some other= s, about 10 firms in total; although I think if=0Aremember correctly, IB= M was not among them [This is the meeting where=0AGates famously exclaim= ed: "_You guys don't get it.=C2=A0 The only thing=0Athat matters in the= software industry is volume_."].=0A=0ABTW: The bits we were discussing= was the upcoming release from USG, to=0Abe called PWB 3.0 and they were= for the PDP-11 only (which was fine,=0Athat was what we all had been li= censing already.=C2=A0 We could still use=0Athings from other places, be= cause that is what those other places were=0Aall licensed to have -- all= was good in UNIX-land).=0A=0AThus began a series of negotiations for a= new license agreement that=0Awould allow the HW vendors to better ship= UNIX as a binary product:=C2=A0=0AFWIW: Gates wanted to pay $25/copy.= =C2=A0 =C2=A0The DEC, HP and DG folks=0Alaughed.=C2=A0 $1K/copy was fine= by them, since their HW was typically=0A$50-150K/system.=0A=0AEither sh= ortly after or maybe during the negotiations time, Judge=0AGreen ruled a= nd AT&T got broken up.=C2=A0 =C2=A0One of the things that occured=0Ais t= hat AT&T was now allowed to sell SW and more importantly their new=0A3B2= 0 as a product (against IBM and DEC).=C2=A0 From a SW standpoint, AT&T= =0AMarketing did not like the 'Programmers' moniker, feeling that it=0Aw= ould limit who they could sell too.=C2=A0 So they rebranded the new=0Aso= ftware product 'System III.'=0A=0ANote the printing of the manuals had a= lready begun, which is why the=0Acover of the manuals say System III, bu= t the title pages say PWB 3.0.=0A=0AAs other have said a few years later= , another PWB release came out for=0Athe Bell System, _a.k.a._ PWB 4.0;= but this was not licensed outside.=0A=0AAt some point later, negotiatio= ns=C2=A0had restarted on yet another=0Alicense with the System III licen= sees and AT&T.=C2=A0 =C2=A0By the time that=0Acompleted, yet another rel= ease had been finished by USG.=C2=A0 The biggest=0Achange was the additi= on support for HW besides the PDP-11. In=0Aparticular, the official USG= support for the VAX and the 3B20.=C2=A0 What=0AI forget, but I think in= that license you had to declare a system type=0Aand most licensees pick= ed the VAX.=0A=0ABy the time of release and finalization of the license,= AT&T Marketing=0Awhich had already started the '_Consider it Standard_'= campaign,=0Acalled the new release "System V."=0A=0AAT&T Marketing woul= d stay with System V moniker from then on and we=0Aknow have SVR2, SVR3,= SVR4, SVR5 in later years.=0A=0AThanks for all your help with this topi= c and sorting things out. It's=0Abeen quite helpful for my talk in a few= weeks.=0A=0AWarner=0A=0AP.S. Would it be inappropriate to solicit feedb= ack on an early version=0Aof my talk from this group?=0A=0AI would sugge= st sending a pointer to this group to the slides and ask=0Afor people to= send you comments privately.=0A=0A=C2=A0=0A I'm sure they would be rath= er keener on catching errors in my=0Aunderstanding of Unix history than= just about any other forum...=0A=0AIndeed - happy to help.=0A=0AClem=0A= =C2=A0=0A=0A =0A=0ALinks:=0A------=0A[1] mailto:imp@bsdimp.com=0A[2]=0Ah= ttps://www.computerhistory.org/atchm/the-asilomar-microcomputer-workshop= -and-the-billion-dollar-toilet-seat/=0A --=_46f365b54b2cc00295962d7cd72f8800 Content-Type: text/html; charset=UTF-8 Content-Transfer-Encoding: quoted-printable
Well, I can share a bit of the piece of the elephant(s) I s= aw.

While most of the Unix folks were PDP-11 on= ly (after MULTICS), I had spent a couple of years helping to run the MH= comp center.=C2=A0=C2=A0 The system for the IBM 7094 was among the best= anywhere -- card decks were staged to tape and the overhead of changing= from one job to the next was minimized.=C2=A0 Most modest-sized job sub= missions were finished in roughly an hour.=C2=A0=C2=A0 When MULTICS star= ted, the 7094 went away and it was a huge jolt.=C2=A0 There were some am= azing programs (e.g., a LISP compiler written in MACRO FAP that would go= 250 levels deep to produce good code).=C2=A0 There were acoustic simula= tion programs that were way ahead of their time.=C2=A0 And they were all= lost.

With MULTICS still way in the future= and no more 7094, a software package was put together to allow the MULT= ICS machines to run the commercial GCOS system (the software was called= the K package, K standing for kluge.=C2=A0 It was well named...)=C2=A0= =C2=A0 Anyhow, they needed help, and asked me to help, and I did so for= a couple of years, largely supporting FORTRAN.=C2=A0=C2=A0 Meanwhile, D= ennis was inventing B, and there was a limping version for the GE.=C2=A0= He may have actually used it as part of a bootstrap of the language --= I'm not sure.=C2=A0 So I discovered it, and started to use it to write= some system admin programs.=C2=A0 I soon ran into some limitations, and= started to lobby Dennis for some changes.=C2=A0 He listened politely, a= nd then said "Why don't you fix it.=C2=A0 The compiler is just a program= , after all..."=C2=A0=C2=A0 So I put some things into B, including suppo= rt for the GE native 6-bit character set.=C2=A0 This eventually found it= s way into C and CPP.=C2=A0 Decades later, if you wrote `abc` in the DEC= C compiler, it produced GE strings...=C2=A0=C2=A0 I also made it possib= le to call FORTRAN from C (the other way was a bit harder...).

Eventually they set up an organization to formally own r= unning the comp center, and I was invited to return to Research and I di= d.=C2=A0=C2=A0 But I had gained a lot of sensitivity to the needs of Eng= ineers and Scientists just wanting to get their work done.

Years later, when I was asked to come over to Summit an= d head the System V language department, V7 was behind us, and the Inter= data port had taught us a lot about portability.=C2=A0 PCC had proved it= self on a half dozen machines and had been well received.=C2=A0 But most= of the Bell System was using 16-bit DEC machines and Dennis' compiler.= =C2=A0 C++ was under develpment and looked like a clear winner, but it n= eeded a lot of work.=C2=A0 The debugging of C++ was horrible -- a progra= m, Cfront, converted C++ to C, but in the process screwed up the line nu= mbers and names so that using a debugger was almost impossible. =C2=A0 M= y three goals in going to Summit were to make C++ commercially available= , to provide a debugging technology that could handle not just C++ and C= but FORTRAN and ADA and PASCAL, and to base the code generation on the= PCC2 model to gain portability.=C2=A0 Although AT&T's computer busi= ness was pretty much a disaster, I felt that I had met these goals=C2=A0= -- Elf and Dwarf, C++ debugging, and the first commercial C++ product h= appened on my watch.=C2=A0 I also had a lot of fun working with a bright= and diverse group of talented people.

Sadly, a= s the AT&T computer business collapsed, I realized that I loved doin= g advanced development, but not at ATT.=C2=A0 My final straw was when an= excellent QA team leader at another location was fired for telling the= truth about how lousy one of the machines performed.=C2=A0 So it was of= f to Silicon Valley for me.

Steve

=

----- Original Message -----
From:
"Clem Cole" <clemc@ccc.com>

= To:
"Warner Losh" <imp@bsdimp.com>
Cc:
"The= Eunuchs Hysterical Society" <tuhs@tuhs.org>
Sent:
Tue, 10 Sep 2019 11:16:34 -0400=
Subject:
Re: [TUHS] PWB vs Unix/TS


Be= low ... from memory - Someone like APS=C2=A0was a little closer to some= of this than I was, so I might=C2=A0have a few things wrong.=C2=A0 I do= n't think so, but it's been quite a few beers

On Mon, Sep 9, 2019 at 2:26 AM Warner Losh <imp@bsdimp.com> wrote:
OK. I'm totally confused, and I see cont= radictory information around. So I thought I'd ask here.

PWB was started to support unix time sharing at bell labs in 197= 3 (around V4 time).
No...=C2=A0 that is not quite right.=C2=A0 PWB w= as Mashey's project to build an RJE system to front end SW development f= or the IBM systems, which AT&T had a number [IIRC Call Accounting an= d lot of the 'business' part of the Bell System was mainframe based].=C2= =A0 I think Dick Haight was also involved.=C2=A0 I've forgotten the site= there were at.=C2=A0 It might have been Holmdel or Whippany. But it was= not MH or Summit.

=C2=A0
PWB 1.0 was released just after V6 "based on" it= .
Well not so much "right after", but it was based on V6.=C2=A0 The= re are differences.=C2=A0 IIRC this was the first attempt at redoing how= groups worked.=C2=A0 The biggest=C2=A0additions were= an IBM RJE support, SCCS and a different set of backup= utilities; including some disk to disk (volc= py)=C2=A0and the original binary formatted program for= 9-tracks (cpio) to replace Ken's asse= mbler based tp.

SCCS was important and the RJE support was important because that= was the system being used and it made a huge impression on AT&T sta= ff.=C2=A0 =C2=A0A terminal to a UNIX box was way cheaper and to the IBM= and people were so much more productive.

Also remember, that tp(1) was written= in assembler had been originally targeted to DECtape in a very early ve= rsion of Research UNIX.=C2=A0 The DECtape nature is why the directory wa= s on the front of the tape.=C2=A0 Ken moved it 9-track but used the same= tape format.=C2=A0 =C2=A0I don't remember who wrote stp (super-tp - in C), [?? Harvard ?? it's on t= he Harvard tape and is how I got it].=C2=A0=C2=A0 But better peripheral support was really= important in Mashey's setting.=C2=A0 In that world, the production comp= uter system was being put in the raised floor computer rooms next to a m= ainframe and they had 'operators' so John and team started to think more= about what was needed to admin the system.=C2=A0 =C2=A0IMHO: this was t= he first heavy use of shell scripts, while I saw them in MH, it was Mash= ey's guys that cause me personally to have an ah-ha moment about them.
<= br>

<= span class=3D"gmail_default" style=3D"font-family:arial, helvetica, sans= -serif;">At some point (and here is where we need Steve Johnson, aps, an= d I wish the late Ted Kowalski) to fill in details I can not.=C2=A0 USG/= Summit was chartered to "support UNIX for the Bell System."=C2=A0 =C2=A0= As I understand it, the genesis for their system was a kernel from MH th= at was moving towards V7s but not there yet, the 'Typesetter C' and a bu= nch of other utilities that Summit had collected/developed, but which I= do not know.=C2=A0 I think fsdb was around by that time. The new Bourne= Shell and adb were being developed although how complete I'm not sure.<= /span>
<= br>

Another thing we did not have= SCCS or the RJE stuff.=C2=A0=C2=A0What I'm not sure of is if the= re was a formally release of what ted had.=C2=A0 So it may have been that TS had them and sent the release to Mashey,= although I don't think there were such releases originally in TS.=C2=A0= FWIW:=C2=A0I believe that in our (CMU) case,Ted=C2=A0would just grab things as they appeared that he thought we needed at CMU and he pushed= things back (like CMU's fsck as he found things we had= that he thought we would like).=C2=A0 Interestingly enough, R= JE and SCCS was needed for the IBM support and while Ted (and his underg= rad roommate, Bill Joy) had worked on the MTS system on the IBM's at UMi= ch, I always felt like Ted looked down on the mainframes (which was were= I had also emerged but from CMU's TSS team).

Also, Ted was a di= e-hard original cpio user and I liked the user interfac= e to stp, which I remember was a difference/source of argument= .=C2=A0 =C2=A0Tar did not yet exist. TS had some of the PWB tools= like volcpy; but we were using DOS-11's similar but= different backup scheme (I've forgotten the name of th= e format; but the tapes were boot-able, which=C2=A0volcpy=C2=A0tapes wer= e not).

FWIW:=C2=A0 cu(1) did= not yet exist.=C2=A0 I wrote a program (that I tended to prefer in some= ways for many years) called connect(1cmu) that did the same thing.=C2= =A0 We used it to download images to the Microprocessors like the KIM-1.= =C2=A0 =C2=A0It was originally written with the v6 portable C library, w= hich is also what the original fsck used (it's what we had on v6).=C2=A0= =C2=A0Ted introduced me to what would become stdio and one of my first= tasks was using it with connect(1cmu).=C2=A0 The other thing I remember= about that program is it was the first time I wrote something that used= two separate processes on a UNIX system that cooperated with each=C2=A0= other and found it so much easier than on the PDP-10.

Also, Dennis' stand-alone system for V7 was not yet= available BTW.=C2=A0 =C2=A0If I think of anything else about that syste= m I can remember, I'll send an update

PWB 2.0 was released just after V7, also "based= on it".
I think the confusion is that TS and V7 were done sort of= at the same time and while the folks working on them talked to each oth= er,=C2=A0it has never been clear to me who was behind TS. For instance,= I would learn that Bourne was the 'project leader' for Seventh, in that= he was the person that collected everything for it.=C2=A0 I never heard= of someone having the same role for TS, which is why I sometimes think= it was a name inside of Summit, but never actually saw the light of day= as a formal release.=C2=A0 =C2=A0I really am not sure and would love to= learn more details (I wish Ted were still alive to fill us in).<= /span>

As for V7 itself, Ken wrote tar(1)= in response to cpio (preferring an ASCII based header, but 'threading'= it like cpio did, but keeping the user interface that tp/stp had).=C2= =A0 As I understand it, Dennis built up did the standalone toolkit stuff= .=C2=A0 Ken changed groups and messed with the file system in the kernel.=C2=A0 Lots of= new peripheral support, which is why he also added lseek()= as disks overflowed a 16-bit integer for the seek position.=C2= =A0 Plus there were a number of other small changes between v6 and v7.= =C2=A0 Some of this stuff from PWB and Summit went back to MH (fsck as a= n example), but not everything (like cpio/volcpy/SCCS).=C2=A0 I kind of= think of the kernel and Typesetter C going from MH to Summit and the PW= B teams.
<= br>
@St= eve Johnson, I need your help here.... at some point PCC was created in= MH (along with lint).=C2=A0 Didn't that start on V6 but was not complet= e until V7? And when did you move to Summit to lead the compiler effort= there?=C2=A0 My impressions that was yet to happen, but I'm fuzzy on da= tes.
Reme= mber, there are a number of teams at BTL hacking on UNIX by then.=C2=A0= Dale's team in Columbus, the crew in Indiana=C2=A0Hill,=C2=A0 folks at= Western Electric (the Teletype folks ported the Ritchie C to the Z80 at= some point for instance),=C2=A0etc.

Again, I don't remember the politics b= ut like any big company, you can imagine it was not all that clean and c= risp.=C2=A0 =C2=A0PWB 2.0 & 3.0 definitely picked up features from o= ther UNIX systems.=C2=A0 As I remember, Dale's shared memory hacks would= beget System V Shared Mem, Semaphores and IPC (they are different, but= they started in Columbus).

The other thing I'm not clear on is when the= PWB team was folded into USG (Unix Support Group)=C2=A0= in Summit.=C2=A0 I believe that was after PWB 2.0 was released.= =C2=A0=C2=A0But at some point= , Mashey's team and the USG got interwoven.=C2=A0 I really don't know/re= member many of those details as I watched them from the outside and only= knew the results.=C2=A0 The key point is the PWB 2.0 would eventually b= e released as the internal, but=C2=A0official UNIX for the Bell= System.=C2=A0 =C2=A0It was supposed to bring together the needed from t= he different labs; but it was not >>officially<< re= leased outside of the Bell System (it was an internal product, re= member at this point, AT&T is not allowed=C2=A0to have computer products, etc...)=C2=A0

So PWB 2.0 is basically internal, and= a melding of V7, TS, PWB 1.0 and starting to take things from different= labs with in=C2=A0BTL -- different from all of them but mostly a supers= et.


=C2=A0
Later Unix TS 3.0 would become System III.
No -= -I do not think this is a true statement... not sure where you got that,= more in a minute
We know there was no System= I or System II.
Correct.=C2=A0

But was there a Unix TS 1.0 and 2.0?
This i= s where it gets sticky.=C2=A0 I don't think so.=C2=A0 =C2=A0TS was the o= riginal work by USG.=C2=A0 =C2=A0What I do not know is if it ever was 'p= ackaged' as PWB had been. I do not believe it was.=C2=A0 =C2=A0I= think a little like the way Research 'bled' out a little a time, pieces= of TS made their way to MIT, CMU,=C2=A0etc. but never as a forma= l release.
=C2=A0
And were they the same thing as PWB 1.0 and 2.0, or somehow just= closely related?
See above... I'll explain how PWB 3.0 became Syst= em III in a minute.
=C2=A0
And I've seen both Unix/TS and Unix TS. Is there a preferr= ed spelling?
Don't know.=C2=A0 I remember Ted always called it UNIX/= TS all caps.

The thing you left out is how PWB 3.0 became System III.
=

Two important issues.=C2=A0 First with V7, AT&= amp;T (Al Arms) wrote the first binary system redistribution license.=C2= =A0 The commercial folks were happy to have a redistribution license, bu= t the terms were not what they really needed.=C2=A0 Much of the issue wa= s that AT&T was not the computer hardware or software business and r= eally did not understand the issues that the vendors had.=C2=A0 Professo= r Dennis Allison of Stanford, was consulting=C2=A0for almost all of us i= n the computer industry at the time (for those that don't know Dennis, a= round the same time he founded what is now called the Asilomar Microproc= essor Workshop (check out:=C2=A0htt= ps://www.computerhistory.org/atchm/the-asilomar-microcomputer-workshop-a= nd-the-billion-dollar-toilet-seat/).

Dennis arranged for a big meet= ing at Ricki's Hyatt in Palo Alto and invited Al Arms and team, plus a r= epresentatives from his clients. I was the techie with a lawyer from Tek= tronix in the room (as I have said in other emails this it is only time= I have been in a meeting with Bill Gates).=C2=A0 The folks I remember w= ho were there: was Bill Munson and team from DEC; Fred Clegg and Team fr= om HP; Bob MetCalfe from 3Com; Gates and the MSFT crew; folks from SCO a= nd DG.=C2=A0 =C2=A0There were some others, about 10 firms in total; alth= ough I think if remember correctly, IBM was not among them [This is the= meeting where Gates famously exclaimed: "You guys don't get it.=C2= =A0 The only thing that matters in the software industry is volume."= ].

BTW: The bits we were discussin= g was the upcoming release from USG, to be called PWB 3.0 and they were= for the PDP-11 only (which was fine, that was what we all had been lice= nsing already.=C2=A0 We could still use things from other places, becaus= e that is what those other places were all licensed to have -- all was g= ood in UNIX-land).
<= br>
Thus began a ser= ies of negotiations for a new license agreement that would allow the HW= vendors to better ship UNIX as a binary product:=C2=A0 FWIW: Gates want= ed to pay $25/copy.=C2=A0 =C2=A0The DEC, HP and DG folks laughed.=C2=A0= $1K/copy was fine by them, since their HW was typically $50-150K/system= .

<= div class=3D"gmail_default" style=3D"font-family:arial, helvetica, sans-= serif;">Either shortly after or maybe dur= ing the negotiations time, Judge Green ruled and AT&T got broken up.= =C2=A0 =C2=A0One of the things that occured is that AT&T was now all= owed to sell SW and more importantly their new 3B20 as a product (agains= t IBM and DEC).=C2=A0 From a SW standpoint, AT&T Marketing did not l= ike the 'Programmers' moniker, feeling that it would limit who they coul= d sell too.=C2=A0 So they rebranded the new software product 'System III= .'

Note the printing of the manual= s had already begun, which is why the cover of the manuals say System II= I, but the title pages say PWB 3.0.

As other have said a few years later, another PWB release came out for= the Bell System, a.k.a. PWB 4.0; but this was not licensed outsi= de.

At some point later, negotiatio= ns=C2=A0had restarted on yet another license with the System III license= es and AT&T.=C2=A0 =C2=A0By the time that completed, yet another rel= ease had been finished by USG.=C2=A0 The biggest change was the addition= support for HW besides the PDP-11. In particular, the official USG supp= ort for the VAX and the 3B20.=C2=A0 What I forget, but I think in that l= icense you had to declare a system type and most licensees picked the VA= X.

By the time of release and fina= lization of the license, AT&T Marketing which had already started th= e 'Consider it Standard' campaign, called the new release "System= V."

AT&T Marketing would stay= with System V moniker from then on and we know have SVR2, SVR3, SVR4, S= VR5 in later years.


Thanks= for all your help with this topic and sorting things out. It's been qui= te helpful for my talk in a few weeks.

Warn= er

P.S. Would it be inappropriate to solici= t feedback on an early version of my talk from this group?
<= /div>
I would s= uggest sending a pointer to this group to the slides and ask for people= to send you comments privately.

=C2=A0
=
I'm sure they would be rather keener= on catching errors in my understanding of Unix history than just about= any other forum...
Indeed - happy to help.
Clem=C2=A0
=0A=0A=0A --=_46f365b54b2cc00295962d7cd72f8800--