[TUHS] DECtapes under the UNIX room floor
Warren Toomey via TUHS
tuhs at tuhs.org
Wed May 6 15:57:51 AEST 2026
On Wed, May 06, 2026 at 05:26:01AM +0000, Thalia Archibald via TUHS wrote:
> In Pirzada’s 1988 thesis, he mentions some DECtapes that were found under the
> UNIX room floor while he was searching for early distributions. Are these the
> Dennis_Tapes?
Here's what I've got. In Jan 1999 I sent this e-mail to dmr:
Now, about that V2 DECtape. In the July 1984 issue of ;login: there is a
Unix trivia quiz. At the top, it says:
The following quiz was distributed at the Salt Lake City conference
by Rob Pike... Jim McKie had the best score for an individual (57) and was
awarded an authenticated 1972 DECtape containing UNIX Version 2.
In 1995, I emailed Jim and asked him what had happened to the tape? He
replied:
a couple of years ago we sent that dectape along with a bunch of
others we found under to floor to keith bostic at berkeley who
had resurrected a dectape drive. i don't remember getting it back.
i'll enquire.
i loved pdp11's.
So I assume that it must have been one of the DECtapes you had Keith read.
However, somehow it got lost because you haven't mentioned it.
So: do you guys still have those DECtapes, or does Keith have them?
But earlier, in Sept 1997, Dennis has sent me this e-mail:
The files here were read from old DECtapes made in the early 1970s.
Paul Vixie and Keith Bostic unearthed a DECtape drive and
made it work, we unearthed the tapes.
The tapes were written in either the 'tap' or 'tp' format,
which are similar in that they have a directory of up to 192
entries at the start with names and other information including
the size and tape address of the files. 'tp' was the later
format, and was in use by November 1973, the date of the 4th edition manual.
With `tap', the times associated with the files were recorded
in pre-modern units: sixtieths of a second, from an origin that
changed. The first three editions of the manual had BUGS sections
noting that 32 bits can represent only about 2.5 years in this
unit, and this implied continuing crises as the time overflowed.
I believe that the change to use seconds for Unix time took place
along with the change to the C version of the operating system, which
occurred about the end of the summer of 1973, and also that the change
from `tap' to `tp' took place at the same time. (This is consistent
with the dates of the 3rd and 4th edition manuals).
Thus the dates recorded with the `tp' tapes probably correspond
reliably to the modification dates of the files at the time
of saving them (of course, this gives only a upper bound on
their creation, since they might have been copied or trivially
touched just before saving them).
Recovering the proper dates for the `tap' tapes is less reliable,
because there was at least one change of epoch (from 1971 to 1972)
during the period they could possibly have been produced.
I believe that the 1972 epoch is most likely the correct one
for the tapes here.
apl
Ken's apl program, together with his rendition of
astro (translated from Morris's Fortran).
Probably OK dates, 1975-76.
config
Ken's program for configuring PDP11s. It told you
price, maintenance, checked for bus loading.
Interesting for DEC price lists in 1974.
dmr
Random stuff from my directory. Most probable dates: 1972.
The `paper' directory contains a version
of the original SOSP Unix paper (haven't compared
it with the CACM version).
cgd appears to be an experiment in converting
Fortran threaded code to machine language,
using a warmed over version of the earliest C
code generator. It's written in NB, not C.
fd is some fortran programs, in particular a
polynomial root-finder I found somewhere.
aman is some version of the PDP-11 assembler manual.
fd is a fortran program for plotting functions
of 2 variables.
notes1 and notes2 are evidently notes I made for myself
for a talk on unix. They are quite interesting.
tty.s is evidently the terminal processing routine for an
assembly-language version of unix. note that it handles
IBM 2741 terminals with two kinds of typeballs
(938 and correspondence).
crypt.c encrypts using a variant of the Hagelin machine.
pig.b is an interesting artifact: it is a B program
that echoes what you type in Pig latin. (Incidentally,
there is a translation of this program into C, dated 1978,
in a subdirectory that still spins on a disk attached to
the Unix machine where I get my mail.)
dmr2
the let directory contains drafts of a bunch of letters
to people who asked about unix in early days. (lett6
is to andy tanenbaum).
The let directory contains drafts of a bunch of letters
to people who asked about unix in early days. (lett6
is to andy tanenbaum).
Other bits: restric discusses some early thoughts
about types in C. I don't know who it was addressed
to (perhaps even to myself). ct is an even briefer
list about what I though was important.
iosys is a manual for writing system device drivers.
e-pi
Our programs to compute e and pi to a million places.
The one for e worked; it took some months on a time-shared
PDP-11 without memory mapping.
Incidentally the <SO>J<SI> directory name is correct:
this sequence printed the greek letter pi on a Teletype
model 37. Plan 9 won't let me create this. It wants
me to use Unicode!
games
Ken's work on various games. Check out chomp/c0.c,
which has a briefly-existing form of structure
declaration using parentheses instead of braces.
ken
check out the values of the AT&T Savings plan in the
early 70s (plan), and what interested Ken's son (corey/*)
then.
ken-du
Here, you can read our wtmp files from 1974 and 1975.
ken-sky
A bunch of interesting old ken stuff (eg a version of
the units program from the days when the dollar fetched
302.7 yen)
last1120c
The source for a version of C that compiled for the
PDP-11/20. It didn't have structures, it didn't have
#define, and pointers were declared p[]. But look!
The initialization routine says
init("float", 2);
init("double", 3);
/* init("long", 4); */
init("auto", 5);
init("extern", 6);
nsys
If the decoding of the date on the tape is correct (Aug 1973),
this is the source for the earliest C version of Unix
likely to be recovered. `nsys' meant the C version
rather than the assembly version, and Aug 73 is plausible for
a running instance of this system.
prestruct-c
This is approximately the same era as `last1120c,'
but 6 months later; despite the name, this compiler now
handles structures. (The struct keyword
has displaced `long' as keyword number 4!) The name of
the tape indicates that the state of the compiler
was saved just before converting it to use structures
inside itself.
The compiler is much less complete--just
the c??.c files, not the tables for code generation.
s1
I haven't cracked this yet.
s2
Is not source, but a dump of (parts of) /bin, /etc,
/usr/lib, and bits of a few other directories.
Caution! The tape uses absolute pathnames,
and is dangerous to extract unless you want to install
old PDP-11 binaries. (tap format).
sys-dsu
This system didn't come directly from us. After
poking around it for a while I found a file containing
this (in cr.h):
/*****************************************************************
** **
** U C L A Data Secure Unix **
** **
** Copyright 1977 **
** **
** Mark Kampe, Charles Kline, Gerald Popek, Evelyn Walton **
** **
*****************************************************************/
So, it appears that `dsu' is `data-secure unix' and this is a
record of one of the early security projects! They must
have sent it to us. Might be fun to examine it; I suppose
we should tell Gerry Popek of its existence.
unix
This is probably a `boot' tape; it has binary images of
the system and some raw utility programs for use when booting
(memory testers, loaders for diagnostics, and the like).
Cheers, Warren
More information about the TUHS
mailing list