> Message: 4
> Date: Wed, 10 Aug 2022 12:29:24 +0200
> From: Holger Veit <hveit01(a)web.de>
> Subject: [TUHS] PCS Munix kernel source
>
> Hi all,
>
> I have uploaded the kernel source of 32 bit PCS MUNIX 1.2 to
> https://github.com/hveit01/pcs-munix.
Thank you for sharing this work, most impressive!
> MUNIX was an AT&T SVR3.x implementation ...
Are you sure? Could it perhaps be SVR2? (I don’t see any STREAMS stuff that one would expect for R3).
> The interesting feature of this kernel is the integration of the
> Newcastle Connection network
One of my interests is Unix (packet) networking 1975-1985 and that includes Newcastle Connection. I’ve so far not dived deep into this, but your work may be the trigger for some further investigation.
My understanding so far (from reading the paper a few years ago) is that Newcastle Connection works at the level of libc, substituting system calls like open() and exec() with library routines that scan the path, and if it is a network path invokes user mode routines that use remote procedure calls to give the illusion of a networked kernel. I’ve briefly looked at the Git repo, but I do not see that structure in the code. Could you elaborate a bit more on how Newcastle Connection operates in this kernel? Happy to communicate off-list if it goes in too much detail.
I note that the repo Readme says that the kernel only does some basic IP networking as a carrier, but I also see some files in the tree that seem to implement a form of tcp (and that seem unrelated to the early Unix tcp/ip’s that I have seen so far). Or am I reading too much into these files?
===
Re-reading the Newcastle Connection paper also brought up some citations from Bell Labs work that seems to have been lost. There is a reference to “RIDE” which appears to be a system similar to Newcastle Connection. The RIDE paper is from 1979 and it mentions that RIDE is a Datakit re-implementation of earlier an earlier system that ran on Spider. Any recollections about these things among the TUHS readership?
The other citation is for J. C. Kaufeld and D. L. Russell, "Distributed UNIX System", in Workshop on Fundamental Issues in Distributed Computing, ACM SIGOPS and SIGPLAN (15-17 Dec. 1980). It seems contemporaneous with the Luderer/Marshall/Chu work on S/F-Unix. I could not find this paper so far. Here, too, any recollections about this distributed Unix among the TUHS readership?
> And I've received the documents! This is a pastebin with the rough contents of the documentation package.
>
> https://pastebin.com/jAqqBXA4 <https://pastebin.com/jAqqBXA4>
>
> Now for some analysis:
I’m interested in the journey of SysV IPC. So far I have established that these originated in CBUnix, with a lot of thinking on how to optimize these around the time that Unix 3.0/4.0/5.0 happened. They did not appear in Unix 3.0 / SysIII, and from the Unix 4.0 documentation I gather that it was not included there either.
This would make Unix 5.0 / SysV R1 the first release with what is now known as SysV IPC. The PDP11 version of R1 has the CBUnix version of shared memory, as the VAX version did not make sense in the limited address space of the PDP11.
From the pastebin summary, it would seem that IPC is not in this documentation either? That would be surprising, and reopens the possibility that IPC was part of Unix 4.0
Paul
> I've always believed that pic was so well designed
> because it took a day to get the print out (back then),
I'm afraid this belief is urban legend. Credit for pic is due 100% to
Kernighan, not to the contemporary pace of computing practice.
Even in the 1950s, we had one-hour turnaround at Bell Labs. And the
leap from batch processing had happened well before pic. Turnaround on
modest Unix source files and tests has changed little in the past
fifty years.
Doug
Thread fork as we're drifting from documentation research specifically.
One matter that keeps coming to mind for me is the formal history of the runlevel-based init system. It isn't in Research obviously, nor was it in PWB. The first time it shows up in the wild is System III, but this version is slightly different than what was in CB-UNIX at the time, which is what eventually wound up in System V.
The pressing question is whether the version in System III represents an earlier borrowing from CB or if perhaps the runlevel init started in USG, got bumped over to CB and improved, then that improved version came back to supported UNIX in 5.0.
As for the notable differences:
SysIII init allows for runlevels 1-9. It will call /etc/rc with the current state, the prior state, and the number of times the current state has been entered. If the script is called for instance due to a powerfailure, then the current state is passed suffixed with an 'x'. The inittab entries are in a different format:
state:id:flags:process
Where state is the runlevel, id is a two-character identifier, flags can be either 'c' (like respawn) or 'o' (like off I think). No flag then indicates to run once. Flags 't' or 'k' will terminate or kill a process before it is started again if a given runlevel is entered and it is already running.
This of course is in contrast to SysV init which instead offers runlevels 0-6 as well as a, b, and c. Init itself can be called with runlevels S|s or Q|q additionally and these act as calls to enter single user mode or rerun the current init state if I'm understanding correctly. Neither S nor Q options appear to be valid for the inittab runlevel. Init tab entries here are:
id:rstate:action:process
Where id and rstate are essentially just the same fields from SysIII swapped. Action replaces the flags field with the more well known respawn, wait, once, initdefault, etc. behaviors.
All in all, different enough that inittabs between the two wouldn't be compatible. SysV also includes the telinit command which appears to be able to handle those a, b, and c runlevels.
Anywho, that's my understanding of the init changes, with the pertinent question remaining whether the SysIII-style init ultimately started from the same place as SysV, or if the general design idea was there between USG and CB, and they got to a similar answer from different directions. Colon-delimited /etc files aren't uncommon, so while unlikely, it could be entirely possible the two inittab formats arose relatively independently, but the truth remains obscure in my mind at least. I don't really blame Research for not picking up this init system, it seems like there were a few parallel streams of development around the turn of the 80s, and the easier answer was probably to just stay the course. I seem to recall reading in a thread somewhere Rob Pike discussing the resistance in the Research group regarding sucking up each and every little feature USG tried to promulgate as standard, and this init system got specific mention.
- Matt G.
And I've received the documents! This is a pastebin with the rough contents of the documentation package.
https://pastebin.com/jAqqBXA4
Now for some analysis:
The User's Manual is branded System V but also displays a Western Electric Bell logo. I've seen Release 5.0 manuals displaying the Bell logo and System V manuals without, but never a System V with. That implies the publication of the manual had to change a few times, one to switch from internal Release 5.0 to commercial System V and another time to remove the Bell logo due to divestiture. I would have to wonder if similar transition can be seen with different revisions of these documents?
The Release Description manual has a list of System V relevant documents and they all appear to be accounted for here, so this should represent the wealth of documentation available to a user of System V Gold in 1983.
Most documents are traceable to documents in the Unix 4.0 collection. I've suffixed various documents here with the coordinate to the same in the 4.0 collection. Changes of note:
- The System V documentation includes instructions for 3B20S machines as well as the instructions for DEC equipment. PDP-11 and VAX guidance have been combined into a single document.
- The System V documentation adds documents concerning an "Auto Call" feature. Didn't see this anywhere in 4.0, so should be new circa System V.
- This documentation refers to the last version as System III rather than making any mention of 4.0. Given that the specific documents mentioning this are System V-branded, and there are comparable documents that are Release 5.0 branded, this implies there may be a document floating around out there somewhere equivalent to the Release Description manual but that actually covers the transition from 4.0 to 5.0.
- The documentation package drops the updated CACM paper, likely because it's available all sorts of other places.
- The summary and documentation roadmap documents appear to have been synthesized and combined into the Release Description.
- Snyder and Mashey's shell tutorial was either dropped or combined with Bourne's shell introduction
- No evidence of an MM foldout like was distributed with 4.0 (and before, there are sources around implying these foldouts started with the PWB group, may have been printed as early as 1977)
- Either the original EQN paper is dropped or relevant bits mashed together with the user's guide
- EFL documentation seems to be dropped, or is merged into one of the other Fortran documents somewhere down in there. The processor is still in the man pages though.
- ADB documentation seems to be dropped, likewise still in the manuals, listed as DEC only. Since System V seems to treat DEC as PDP-11+VAX, does this imply there was a VAX ADB? My understanding is SDB started on 32V and was *the* debugger for VAX.
- Unix Virtual Protocol papers are dropped, they were marked as 3.0 only in the 4.0 manuals anyhow, so probably not relevant.
- The Standalone I/O Library and SASH (Shell) paper is dropped
- None of the internals nor security papers seem to have made it, so no Unix Implemention, I/O Implementation, PDP and Portable C Compiler Tours, Assembler Manual, PDP-11/23 and 11/34, or Password Security papers.
These will likely be a slower burn than the 4.0 documents since I purchased them myself and am not in a hurry to get them shipped back to someone. That said, if there's anything in the above pastebin that particularly piques any interest, I can try to move those to the top of the stack and get scans done sooner rather than later. I'll also be doing some analysis between these and the 4.0 docs to try and better determine authorship of various documents, my hope is to have a pretty clear picture of whos work went into each manual by the time I'm done with it all.
- Matt G.
> From: Rob Pike
> I still marvel at the productivity and precision of his generatio
We noticed the same thing happening in the IETF, as the number of people
working on networking went up. The explanation is really quite simple, once
you think about it a bit.
If you have a very small group, it is quite possible to have a very high
level. (Not if it's selected randomly, of course; there has to be some
sorting function.) However, as the group gets much larger, it is
_necessarily_ much more 'average' in the skill/etc level of its members.
This rule applies to any group - which includes the members of TUHS,
of course.
Noel
Hi all,
I have uploaded the kernel source of 32 bit PCS MUNIX 1.2 to
https://github.com/hveit01/pcs-munix.
MUNIX was an AT&T SVR3.x implementation for the German PCS Cadmus
workstations in the 80's. They were
based on Motorola 68020 CPUs on a DEC QBUS.
The interesting feature of this kernel is the integration of the
Newcastle Connection network
(https://en.wikipedia.org/wiki/Newcastle_Connection) which I found,
beyond a tech report https://assets.cs.ncl.ac.uk/TRs/175.pdf, no further
references for.
The kernel source was reverse engineered and verified (see readme in the
distribution who this was done) from the binary tape at
ftp.informatik.uni-stuttgart.de/pub/cm/pcs/sw/IS0371P.tap (Computer
museum of the University of Stuttgart), and to my knowledge reveals the
Newcastle connection code for the first time in a commercial Unix.
The Github package includes the kernel sources, i/O drivers, several
standard libraries, the disassembled boot ROM and for reference, two of
my tools, a partial syscall emulator pcsrun which allowed me to run the
C compiler and other native binaries outside the PCS hardware/Unix
environment, and a disassembler pcsdis for the specific COFF dialect
(note that IDA will produce garbage without a specific patch).
Regards
Holger
I've been looking into the history of the nl command lately, which has gotten me curious as to what facilities folks have used at various points in UNIX history for line numbering.
The earliest version of nl I've found is in System III, and it does not derive from Research, PWB, or CB. Neither does it come from BSD, although BSD has the num command which, according to the source commentary, aims to replicate the '#' behavior of ex.
Were there any other facilities for printing back arbitrary lines from a file with line numbers?
Also, would anyone happen to know if the above appearance of nl might have been from the USG line given none of the others feature it? It neither seems to be in V8-V10. nl has managed to find its way into the POSIX standard, so it definitely has some staying power wherever it came from.
- Matt G.
Good morning everyone. Wanted to pose the question since folks here would probably be more likely to know than anyone.
What are the chances that there are surviving tapes of some of the UNIX versions that weren't so well publicized. The versions that come to mind are the CB and USG lines especially, with PWB 2.0 and TS 4.0 getting honorable mention. If folks will recall, we did luck out in that Arnold, a member of this mailing list, did have a documentation trove from TS 4.0, but no binary or source code assets. This had me curious on what trying to unearth these would even look like.
Has anyone tried to dive deep on this sort of stuff before? Would it look more like trying to find old Bell facilities that might have a tape bumping around in a box in a basement somewhere, or is it more likely that if anything survived it would have been due to being nabbed by an employee or contractor before disposal? Or even just in general, what would folks say is the likelihood that there is a recoverable tape of any of this material just waiting to see the light of day? The closest we have on CB is a paper scan of the kernel sources, and I don't know that any assets from USG-proper have ever percolated up, closest thing to any of that would be the kernel routine description bumping around on the archive somewhere. PWB 2.0 is mentioned in several places, but no empirical evidence has surfaced as far as I know, and with 4.0 of course we have the documents Arnold thankfully preserved, but that's it.
Thanks in advance for any insight or thoughts. My concern is that there is a rapidly closing window on ever being able to properly preserve these parts of the UNIX story, although recognition must be paid to all of the hard work folks have done here thus far to keep this valuable part of computing history in the collective consciousness and accessible to researchers and programmers for years and years to come.
- Matt G.
P.S. Even more honorable mention is the Bell Interdata 8/32 work. I've read several places that never saw outside distribution, but I would have to wonder if any of that work survived beyond the visible portability changes in V7.
I didn't expect to have more documents to share this soon, but I've just secured a trove of early System V/5.0 documents, as listed:
System V User's Manual
System V Administrator's Manual
System V Error Message Manual
System V Transition Aids
System V Release Description
User's Guide
Operator's Guide
Administrator's Guide
Programming Guide
Graphics Guide
Support Tools Guide
Document Processing Guide
The System V-prefixed ones are very specifically labeled System V, although I know at least of the User's and Administrator's Manuals with "Release 5.0" branding out in the wild as well. I've got two of the User's Manuals exhibiting this difference. I believe I've seen a scan of the Admin's Manual with 5.0 as well, but I would have to go searching for it, it's on bitsavers perhaps? In any case, this is the documentation series for the initial releases of System V, the ones with "UNIX System" in big letters with grid patterns fading out into the background. I don't know if the second set is considered part of the Release 5.0 or System V version of the document package, or if they made that distinction, but as of present I can positively identify the first 5 as being specifically for the System V version of this release. What is particularly curious is there are documents displaying "System V" but with a Western Electric logo on the front. I've seen a scan of a System V gold User's Manual with the logo removed and a disclaimer on the front page explaining that they can't use the Bell logo anymore due to the divestiture, likewise on bitsavers I'm pretty sure, so this may establish that there were at least three revisions: Release 5.0, System V pre-divestiture, and System V post-divestiture.
Now for a little plug, just because she's been so incredibly helpful, I bought these from Leslie (last name unknown) known as "oldmaddogshop" on eBay. We got chatting for a little while and her husband was a computing professor at the University of Portland for some time as it sounds, and they're currently starting to go through the decades of literature and hardware he's picked up over the years for sale on eBay and perhaps other avenues. She very specifically mentioned a PDP-8 that he happens to have that he's hoping they can coordinate to donate to a museum or some other way to get it into a relatively publicly accessible space rather than winding up in the closet of a private collector. I told her I'd drop a brief mention in letting folks know about the documents in case they'd want the option of perusing some of what they're going to be offloading. She made mention of a stack of USENIX manuals as well, I have a smattering of 4.2 and 4.3 manuals already, so someone may be lucky enough to snag those soon enough. Up currently are an early SVID and some OSF/Motif stuff, but she said they've got plenty of boxes of books to go through.
Anywho, once I receive these documents, I plan on starting the scanning process much like with the UNIX/TS 4.0 stuff, and will be in touch with Warren concerning hosting and a release as time goes on. One bit of input if anyone knows, does the above list represent (aside from Release 5.0 variants) the complete documentation package for System V gold? I can't say I've come across any other titles, and most certainly haven't seen PDFs of anything that isn't included here, but I see plenty of titles I've never seen scanned. If nothing else, I'm hoping that "Release Description" document may have a brief flyover of the published materials, akin to the list of books at the beginning of the SVR4 manuals or the documentation roadmaps of earlier UNIX/TS and PWB releases.
- Matt G.
