22 Mar
2020
22 Mar
'20
3:58 p.m.
Greetings,
I was reviewing a manual page update and came across an ambiguous answer to
a question that came up, so I thought I'd ask here.
execl and execv first appear in our extant unix man pages in V4. The v3 and
v2 man pages don't have this listed at all. Case closed, right? It appeared
in V4.
However, the Dennis_v1/unix72 tree has execl.s and execv.s in them. Diving
back into the history on Warren's github account,
https://github.com/DoctorWkt/unix-jun72/tree/master/src says they come from:
"The files in lib/ come from the libc.sa file which is on the
last1120c.tar.gz
tap(I) tape image, also at the same URL, and form the C library for the
above compiler."
and
"from a working C compiler for 2nd Edition UNIX."
which suggests that it may have been in V2 or maybe even V1. It's
first use in Unix appears to be V5, but the extant pre-v5 code is so
fragmentary it's hard to know for sure. It's not mentioned in section
II or section III of v1, v2 or v3.
Does anybody know for sure, or can provide more insight into the
last1120c.tar.gz file to help disambiguate?
Warner
22 Mar
22 Mar
8:20 p.m.
On Sun, Mar 22, 2020 at 09:58:29AM -0600, Warner Losh wrote:
"from a working C compiler for 2nd Edition
UNIX."
Does anybody know for sure, or can provide more
insight into the last1120c.tar.g
z file to help disambiguate?
Yes, all the details are on Dennis' web page:
https://www.bell-labs.com/usr/dmr/www/primevalC.html
Cheers! Warren
8:49 p.m.
On Sun, Mar 22, 2020 at 2:20 PM Warren Toomey <wkt(a)tuhs.org> wrote:
On Sun, Mar 22, 2020 at 09:58:29AM -0600, Warner Losh
wrote:
Magic historical 8-ball sez: situation unclear, try again.
So could be 2nd, 3rd or 4th edition given the date range.
Warner
"from a working C compiler for 2nd Edition
UNIX."
Does anybody know for sure, or can provide more
insight into the
last1120c.tar.g
z file to help disambiguate?
Yes, all the details are on Dennis' web page:
https://www.bell-labs.com/usr/dmr/www/primevalC.html
9:44 p.m.
On Sun, Mar 22, 2020 at 02:49:03PM -0600, Warner Losh wrote:
On Sun, Mar 22, 2020 at 2:20 PM Warren Toomey
<[1]wkt(a)tuhs.org> wrote:
On Sun, Mar 22, 2020 at 09:58:29AM -0600, Warner Losh wrote:
Someone from the Labs here on the list might be able to clarify. I'm sure
I heard that they start with the PDP-11/20 and got an 11/45 around the time
of Second Edition. So the "last1120c" was the C compiler around the time of
Second Edition.
Cheers, Warren
"from a working C compiler for 2nd Edition
UNIX."
Does anybody know for sure, or can provide more insight into the
last1120c.tar.g
z file to help disambiguate?
Yes, all
the details are on Dennis' web page:
[2]https://www.bell-labs.com/usr/dmr/www/primevalC.html
Magic historical 8-ball sez: situation unclear, try again.
So could be 2nd, 3rd or 4th edition given the date range.
10:03 p.m.
Hey Warren,
On Sun, Mar 22, 2020 at 3:44 PM Warren Toomey <wkt(a)tuhs.org> wrote:
On Sun, Mar 22, 2020 at 02:49:03PM -0600, Warner Losh
wrote:
According to my history of early unix talk, they started to transition to
11/45 in the 2nd edition, and had completed the rewrite by the 4th
edition. The 2nd and 3rd editions had, at least according to their intros,
some support for the 11/20, but it is ambiguous now that I re-read it:
2nd edition (June '72)
"Also, not all installations have the latest version of all the software.
In particular, the assembler and loader have just undergone major
reorganizations in anticipation of a UNIX for the PDP-11/45"
3rd Edition goes farther (Mar '73):
"In particular, any system which uses a PDP-11/20 processor will not
include all the software described herein, nor will the software behave in
the same way. The second, or even the first edition of this manual is
likely to be more appropriate."
By the 4th edition (Nov '73), The PDP-11/20 is gone, not even referenced....
"In particular, the information in this manual applies only to UNIX systems
which operate under the C language versions of the system. Installations
which use older versions of UNIX will find earlier editions of this manual
more appropriate to their situation."
So the outward messaging of the manual, at least, was "We're running away
from the 11/20, and you should too if you want to run the latest code." But
it's not clear to me, since we don't have good sources from the actual 2nd
or 3rd edition to trace where this came in, or if the 11/20 lived on a
little past the 4th edition and had sources freshly loaded on it sometime
after the compiler was written...
Warner
On Sun, Mar 22, 2020 at 2:20 PM Warren Toomey
<[1]wkt(a)tuhs.org>
wrote:
On Sun, Mar 22, 2020 at 09:58:29AM -0600, Warner Losh wrote:
Someone from the Labs here on the list might be able to clarify. I'm sure
I heard that they start with the PDP-11/20 and got an 11/45 around the time
of Second Edition. So the "last1120c" was the C compiler around the time of
Second Edition.
"from a working C compiler for 2nd Edition
UNIX."
Does anybody know for sure, or can provide more insight into the
last1120c.tar.g
z file to help disambiguate?
Yes, all
the details are on Dennis' web page:
[2]https://www.bell-labs.com/usr/dmr/www/primevalC.html
Magic historical 8-ball sez: situation unclear, try again.
So could be 2nd, 3rd or 4th edition given the date range.
1502
days inactive
1502
days old
4 comments
2 participants
participants (2)
-
Warner Losh -
Warren Toomey