OK, let me try this one more time with links to get around the
restrictions on the message size.
I was cleaning out my basement and I
found a box of stuff from my office at BRL (where I left in 1987).
Most
of it was buttons I'd picked up at trade shows and SF cons (I had a
fabric partition next to my desk that I had them all stuck to.
Of
course in the box (among a couple of later editions) was Armando's
original UNIX license (note no reference to DEC):
Also in the box were
some buttons from various UNIX conferences. I particularly remember the
Sex, Drugs, and Unix one. Some of you will also remember the year I was
giving out the No Ada shirts. There's a picture of dmr wearing one
floating around somewhere.
Sun was giving out these one year:
Peter
Langston thought this was a little conceited on Bill Joy's part, so the
next show he arrived with buttons to hand out that said things like "The
psl of UNIX" and "The dmr of UNIX". I had a "The ron of UNIX" somewhere
but I couldn't find it in the box.
Finally there was this wooden
nickle courtesy of Bill Yost...
> The wikipedia description
> <http://en.wikipedia.org/wiki/CAT_(phototypesetter)>
> seems pretty accurate although I have never seen the beast myself.
I can confirm the wikipedia description. At Bell Labs, however, we
did not use paper tape input. As soon as the machine arrived, Joe
Ossanna bypassed the tape reader so the C/A/T could be driven
directly from the PDP-11. The manufacturer was astonished.
The only operational difficulty we had was with the separate
developer. If you didn't hand feed the end of the paper perfectly
straight into that machine, the paper would tear. Joe Condon
fixed that by arranging for the canister to sit on rollers so
it could give when the paper pulled sideways.
The first technical paper that came off the C/A/T drew a query
from the journal editor, who'd never seen a phototypeset
manuscript before: had it been published elsewhere?
Doug
> Didn't the BSTJ get phototypeset on your typesetter??
Not that I know of. But Charlie Brown's office did use it.
Brown, the CEO of AT&T, did not like wearing reading glasses
when he made a speech, so his speechwriters produced the
text in large type. Once they were into document preparation
they began to use our machine for other things. When I
discovered that confidential minutes of AT&T board meetings
were in our file system, I told them the facts of life
about computer security, especially at the center of the
UUCP universe, and persuaded the executive suite to get
its own machine.
Doug
the one at WH was directly connected to a vax 11/780, no paper tape either.
so that now finally explains why /dev/cat was write only, it was substituted
for a paper tape reader. it was always a curiosity that you could write
to it, yet never read it (i.e. get a status). a "cat /dev/cat" would
get you a "cat: cannot open /dev/cat" while a "cat /some/file > /dev/cat"
would succeed, but act like you used /dev/null instead
(as /some/file was not valid phototypeseter input)
> From: Doug McIlroy
>
> > The wikipedia description
> > <http://en.wikipedia.org/wiki/CAT_(phototypesetter)>
> > seems pretty accurate although I have never seen the beast myself.
>
> I can confirm the wikipedia description. At Bell Labs, however, we
> did not use paper tape input. As soon as the machine arrived, Joe
> Ossanna bypassed the tape reader so the C/A/T could be driven
> directly from the PDP-11. The manufacturer was astonished.
>
> The only operational difficulty we had was with the separate
> developer. If you didn't hand feed the end of the paper perfectly
> straight into that machine, the paper would tear. Joe Condon
> fixed that by arranging for the canister to sit on rollers so
> it could give when the paper pulled sideways.
>
> The first technical paper that came off the C/A/T drew a query
> from the journal editor, who'd never seen a phototypeset
> manuscript before: had it been published elsewhere?
>
> Doug
>
Look at United States Patent 4074285
http://patentimages.storage.googleapis.com/pdfs/US4074285.pdf
Figure 1 is identical to the machine I ran at Whippany Bell Labs
in the early to mid 80s. It was about 4 1/2 feet tall
Figure 4 is the font wheel (seen as 16 in Fig 1) there were 4 distinct
sectors, each with a different font. One with Times Roman, one with
Times Roman Bold, one with Times Roman Italic and the last was the
symbol fonts (math, greek chars, left hand (\lh right hand \(rh etc. and
this one was made specifically for the Labs as it had a Bell Logo \(bs on it)
The paper was a roll of photo paper, glossy on the text side, rough on
the reverse, it was thick. It would end up going into the cassette
(20 in Fig 1) and would need to be developed. Not shown in the patent
figures was the developing and drying apparatus. At the end of
a job the exposed paper was in the cassette you'd remove
it from the typesetter and put into a device with rollers that would pull
it out and run it thru developer and fixer liquid chemicals. Exiting
that it would go into a dryer drum.
After it was completly dry, as it was still a continuous roll, you
would need to cut all the pages apart by hand (that is why there was
the cut mark macro (.CM) is -ms so you could tell where to cut)
As it came from a roll, no pages ever layed completely flat.
The checmical baths were nasty smelling and it gummed up the rollers.
You'd needed to regularly take the developer roller and gear guts into
the janitor's closet and scrub it with a toothbrush in the slop sink
under running water.
By the second half of the 80s it was replaced by QMS PostScript
laser printers.
> From: "Jacob Goense" <dugo(a)xs4all.nl>
> All, I'm looking for images of the cat device as mentioned several
> times in the 7th edition manual, see e.g. TROFF(1)and CAT(4).
>
> From what I gathered during my digs is that it should look like a
> GSI 8400, but that didn't help. Anyone here who can help me find out
> what these machines looked like? A picture would be the best, but
> information on what to look for in images of unnamed typesetters will
> do as well.
>
> /Jacob
>
>
All, I'm looking for images of the cat device as mentioned several
times in the 7th edition manual, see e.g. TROFF(1)and CAT(4).
>From what I gathered during my digs is that it should look like a
GSI 8400, but that didn't help. Anyone here who can help me find out
what these machines looked like? A picture would be the best, but
information on what to look for in images of unnamed typesetters will
do as well.
/Jacob
Hi all.
This may be of some interest. From a friend at Utah:
> Date: Sat, 30 Nov 2013 16:06:25 -0700 (MST)
> Subject: [it-professionals] computer history: Arpanet IMPs resurrected
>
> The simh list about simulators for early computers recently carried
> traffic about an effort to reconstruct and resurrect the Arpanet
> Interface Message Processors (IMPs), which were the network boxes that
> connected hosts on the early Arpanet, which later became the Internet.
>
> There is a draft of a paper about the work here:
>
> The ARPANET IMP Program: Retrospective and Resurrection
> http://walden-family.com/bbn/imp-code.pdf
>
> Utah was one of the original gang-of-five hosts on the Arpanet, and we
> received IMP number 4. Utah is mentioned twice in the article, and
> also appears in the map in Figure 3 on page 14.
>
> One amusing remark in the article (bottom of page 7) has to do with
> the fail-safe design of the IMPs:
>
> In addition ``reliability code'' was developed to allow a
> Pluribus IMP to keep functioning as a packet switch in the
> face of various bits of its hardware failing, such as a
> processor or memory [Katsuki78, Walden11 pp. 534-538]. This
> was so successful there was no simple off switch for the
> machine; a program had to be run to shut parts of the machine
> down faster than the machine could ``fix itself'' and keep
> running.
>
> As happened with early Unix releases, machine-readable code for the
> IMPs was lost, but fortunately, some old listings that turned up
> recently allowed its laborious reconstruction, verification, assembly,
> and simulation.
Arnold
Clem Cole <clemc(a)ccc.com> wrote:
>If the original BBN code had
>been left alone, since most people did not have the issues Berkeley did,
>they would never have bothered with sendmail.cf.
Now I might be badly wrong, but nonetheless this strikes me as
badly revisionist history.
The motivation for sendmail.cf was the collision of multiple
namespaces (Arpanet, Bitnet, Usenet, etc.), each implemented
in varying nonstandard ways by different mail clients and servers,
resulting in messes like "IJQ3SRA%UCLAMVS.BITNET%SU-LINDY(a)SU-CSLI.ARPA",
as one of many, many examples, as observed in the famous
"The Hideous Name", Rob Pike & P.J. Weinberger, 1985
http://pdos.csail.mit.edu/~rsc/pike85hideous.pdf
The thing is, although sendmail.cf was/is itself hideous to understand
and therefore make maintenance changes to (although I have), it is
quite capable of actually handling the above kinds of messes, and
being extended to handle new messes as they turn up.
In short, it got the job done, despite its weaknesses.
I may be wrong, but it was my strong impression that, back in the
day, this could not be said of anyone else's code, BBN or otherwise.
Doug Merritt
I am reading the delivermail (later known as postbox and then sendmail)
code from 4.0BSD and from sccs history from June 1980.
Its arpa-mailer(8) manual says it just spools the letter and actual
delivery will be performed by the ARPANET mailer daemon and refers to
mailer(ARPA) manual. The arpa.c code says "is stuck away in the
outgoing arpanet mail queue for delivery by the true arpanet mailer."
Where is this true arpanet mailer? I am guessing it periodically looks
in /usr/spool/netmail/ and delivers the messages using FTP and RFC458.
Where is this mailer(ARPA) manual?
Where is the ftp server code used for the incoming mail? (Example
code mail-dm.c is provided for the ftp server to "handle the MAIL <user>
command over the command connection.")
Also where is the uucp-mailer(8) manpage referenced in delivermail(8)?
Jeremy C. Reed
echo 'EhZ[h ^jjf0%%h[[Zc[Z_W$d[j%Xeeai%ZW[ced#]dk#f[d]k_d%' | \
tr '#-~' '\-.-{'
On Mon, Nov 25, 2013 at 04:55:58PM -0600, Jeremy C. Reed wrote:
> Kashtan's VMS performance comparison paper and Joy's followup from early
> 1980 both refered to the VM/UNIX as Version 2.1 of the Berkeley system;
> this was the "Third" distribution; by April the kernel was known as 3.1.
>
> 2.4BSD was mentioned in the kermit source's Makefile. But maybe a
> mistake.
I stand well corrected!
Thanks,
Warren
