Origin of PTYs
Eric Smith
The 1971 edition of the PDP-10 Timesharing Monitors Programmer's
Reference Manual describes them, but I can't find earlier references.
Were they inspired by similar features of another OS? Did PTYs in
Monitor inspire the same feature in TENEX, or vice versa?
jmfbahciv
For a date, take a look at the first couple of pages of PTYSER.MAC.
I don't remember who wrote it.
PTYs were used by any program which needed to login a job that didn't
have a hard-wired TTY line.
/BAH
jmfbahciv
An experienced guess would be when ever MPB was shipped. With
the implementation of spoolers, input and output, a computer center
couldn't have a dedicated CTY for each job nor have the operator
running all over and outside the computer room to deal with
tending printers and mounting/dismounting disk packs and DECtapes.
an operator (or bit god) would want to sit and swivel while
dealing with booting, crashing and user requests.
A lot of that software service started with Level D. Do you
have the Level D spec?
/BAH
Eric Smith
dates earlier than 1973, so perhaps it was rewritten then.
jmfbahciv
> PTYSER.MAC comments don't have an edit history, and don't have any
> dates earlier than 1973, so perhaps it was rewritten then.
I would guess, then, that PTYSER was written in 72 or 73. Do you
have any of MCO files for monitors shipped in that timeframe?
another source which would, or should, have edit histories,
would be any of the MPB sources. So that would be
SPRINT.MAC, SPROUT.MAC, BATCON.MAC (not to be confused with
the BATCON which shipped with GALAXY; GALAXY superceded
all MPB functionality).
/BAH
John Everett
>Eric Smith wrote:
>> PTYSER.MAC comments don't have an edit history, and don't have any
>> dates earlier than 1973, so perhaps it was rewritten then.
>
>I would guess, then, that PTYSER was written in 72 or 73. Do you
>have any of MCO files for monitors shipped in that timeframe?
When I joined DEC in September, 1966, I shared an office with Mauritz
Frederickson, who was working on a "batch processing function" for the
PDP-6 at the time. This used PTYs, so they predate the KA-10.
I can still recall the day I dicovered that typing "TALK PTY" to the
PDP-6 Monitor caused a crash. ;-)
This was late 1966 or early 1967.
--
jeverett3<AT>sbcglobal<DOT>net (John V. Everett)
jmfbahciv
> On 15 Jan 2011 13:46:25 GMT, jmfbahciv <See....@aol.com> wrote:
>
>>Eric Smith wrote:
>>> PTYSER.MAC comments don't have an edit history, and don't have any
>>> dates earlier than 1973, so perhaps it was rewritten then.
>>
>>I would guess, then, that PTYSER was written in 72 or 73. Do you
>>have any of MCO files for monitors shipped in that timeframe?
>
> When I joined DEC in September, 1966, I shared an office with Mauritz
> Frederickson, who was working on a "batch processing function" for the
> PDP-6 at the time. This used PTYs, so they predate the KA-10.
I could not remember when I first encountered PTYs. Thanks.
>
> I can still recall the day I dicovered that typing "TALK PTY" to the
> PDP-6 Monitor caused a crash. ;-)
<GRIN> I also don't remember the command TALK.
>
> This was late 1966 or early 1967.
>
>
That gives us the year.
I liked your story about the reunion. One of my dreams at the end
of the TOPS-10 7.03 development cycle was to make enough money
so all the TOPS-10 developers could join JMF and me on a cruise.
I was going to rent the whole boat.
Hadn't thought about that one in 2 decades.
/BAH
Mark Crispin
> <GRIN> I also don't remember the command TALK.
I do.
TALK was in TOPS-10 up to 4.x series days under the FTTALK conditional.
Apparently, some cretin thought that the new SEND command in 5.x
superceded TALK, and removed it.
TALK was completely different than SEND, and covered a different and
equally value range of functionality. It was real-time, and could involve
multiple TTYs in a talk-ring - a major feature!
We had both TALK and SEND in WAITS. TALK was a monitor intrinsic, and
SEND was an auxillary in the mailsystem (unlike TOPS-10 where SEND was a
monitor intrinsic).
Both TENEX and TOPS-20 had TALK, although the EXEC command was called LINK
in TENEX. It too supported talk-rings. Of course, there was a JSYS to
set up a TALK link, as well as to set up one's TALK receptiveness (accept,
refuse, query).
TENEX and TOPS-20 also had "advise", which was a form of TALK that let you
talk to the other user's programs including EXEC. The closest analog on
TOPS-10 was the feature in X680 where the operator on the PDP-8 console
TTY could connect to some victim's TTY. It was easy for those of us on
300 baud lines to detect when this happened since the console was 110
baud.
SEND in TOPS-20 started as a privileged JSYS, but was presently made
non-privileged. However, it never really worked right. On my systems,
SEND is a mailsystem auxillary as on WAITS.
-- Mark --
http://panda.com/mrc
Democracy is two wolves and a sheep deciding what to eat for lunch.
Liberty is a well-armed sheep contesting the vote.
Mark Crispin
TITLE PTYSRF - FULL DUPLEX PSEUDO TELETYPE SERVICE ROUTINES
SUBTTL M. FREDRIKSEN/RCC TS 25 NOV 68 V005
TITLE PTYSRH - HALF DUPLEX PSEUDO TELETYPE SERVICE ROUTINES
SUBTTL D. WITCRAFT/RAP TS3.19 24 SEP 68 V004
jmfbahciv
> For what it's worth, PTY support is definitely in TOPS-10 3-series:
>
> TITLE PTYSRF - FULL DUPLEX PSEUDO TELETYPE SERVICE ROUTINES
> SUBTTL M. FREDRIKSEN/RCC TS 25 NOV 68 V005
>
> TITLE PTYSRH - HALF DUPLEX PSEUDO TELETYPE SERVICE ROUTINES
> SUBTTL D. WITCRAFT/RAP TS3.19 24 SEP 68 V004
I never touched a listing, while editing the monitor in the early
70s, called PTHSRH nor PTYSRF. PTYSER must have combined the two.
/BAH