27 Nov
2018
27 Nov
'18
9:47 p.m.
We lost J.F. Ossanna on this day in 1977; he had a hand in developing
Unix, and was responsible for "roff" and its descendants. Remember him,
the next time you see "jfo" in Unix documentation.
-- Dave
27 Nov
27 Nov
11:08 p.m.
joe was much more than that. he knew how
to play the system. example:
out of whole cloth, he invented a form to
order a teletype and opx (bell labs extension)
installed in the home. he then filled out the
form for each of the unix-room dennisons.
there was a phone call from a confused
clerk, and then we all got teletypes and
data sets at home. as an aside, the opx
came with free watts (long distance which
was very expensive in those days.)
On Tue, Nov 27, 2018 at 1:47 PM, Dave Horsfall <dave(a)horsfall.org> wrote:
We lost J.F. Ossanna on this day in 1977; he had a
hand in developing Unix,
and was responsible for "roff" and its descendants. Remember him, the next
time you see "jfo" in Unix documentation.
-- Dave
28 Nov
28 Nov
2:57 a.m.
As a long time roff fan (I still use it, yes, I've learned LaTex, I much
prefer roff), I'm hugely bummed that Joe left us so early. I feel like
there would be more fun stories, like Ken's story.
If I remember correctly, he wrote the first (Unix *) version of roff in
PDP-11 assembly, right? Granted, PDP-11 assembly is perhaps the most
pleasant assembly ever, but it is still assembly. Roff is a non-trivial
program, I can't say that I've every written anything remotely that big
in assembly (the only thing I'm proud of is writing swtch() in VAX, 68K,
and some other CPU that I can't remember, but that was tiny, hard to get
right, but tiny). I've got mad respect for what he did, I feel like the
whole roff thing doesn't get enough respect. It wasn't just roff, though
that started it, it was pic (I *love* pic), eqn, all the other filters
that go down to roff. For lmbench I wrote my own grap like tools
because grap wasn't open source.
I was talking to Marc Donner, a Morgan Stanley techy (since moved on
to google and who knows where) about why I liked roff. At the time
I had built webroff which took roff -ms input and made websites.
Marc pointed out that the reason I liked roff was, for the most part,
it didn't say how to do something (that was buried in the macros),
it said what you wanted to do.
Ken, if you have more Joe stories I'd love to hear them, I feel like
I missed out on a cool person.
(*) I know that nroff was "new run off" and it came from somewhere,
MIT? Some old system, but it wasn't invented in Unix. That said,
I've never seen docs for the previous system and I kinda think Joe
took it to the next level. If you haven't studied the docs and
written macros, you should. It's a pretty neat system.
On Tue, Nov 27, 2018 at 03:08:36PM -0800, Ken Thompson via TUHS wrote:
joe was much more than that. he knew how
to play the system. example:
out of whole cloth, he invented a form to
order a teletype and opx (bell labs extension)
installed in the home. he then filled out the
form for each of the unix-room dennisons.
there was a phone call from a confused
clerk, and then we all got teletypes and
data sets at home. as an aside, the opx
came with free watts (long distance which
was very expensive in those days.)
On Tue, Nov 27, 2018 at 1:47 PM, Dave Horsfall <dave(a)horsfall.org> wrote:
> We lost J.F. Ossanna on this day in 1977; he had a hand in developing Unix,
> and was responsible for "roff" and its descendants. Remember him, the
next
> time you see "jfo" in Unix documentation.
>
> -- Dave
--
---
Larry McVoy lm at mcvoy.com http://www.mcvoy.com/lm
4:48 a.m.
another joe:
echo 1 was a 100 foot balloon that was
launched into space in the early 60s. this
was the first satellite that was easily visible
to the naked eye.
joe wrote a set of fortran programs that
tracked the orbit of echo and calculated
the direction to look from a point on earth.
to do this, he had to learn fortran and
orbital dynamics.
the programs were used to point antennas
to send emf from california. bouncing off
echo and received at bell labs in
new jersey. thus, thanks to joe, echo was
the first communications satellite.
by the time i came to bell labs (1966) the
program, azel, for azimuth/elevation, was
expanded to track planets, moons, satellites,
etc. moreover, it tracked the shadow of the
earth cast by the sun (night). it could predict
within a few seconds when echo would wink
on or off as it passed through the shadow.
a version of azel was maintained all the time
i was at bell labs. we used it to predict
eclipses, transits, occultations etc. when
we first got a voice synthesizer, the day's
predictions were spoken at 5pm in case
there was anything interesting.
anyway, at 5pm on june 8, 1983 the voice
announced an "occultation of mercury"
for early the next morning.
no one had heard of such a thing. it was
extremely rare. mercury had to be at
about its max elongation; the moon had
to be only a few hours old (or young);
it had to be dark; the moon and mercury
had to be above the horizon; and lastly,
the moon had to occult mercury.
we all (me, lee mcmahon, dennis ritchie,
rob pike, and bob morris) frantically tried
to verify that it was real. it was, but it
would only be about 5 degrees above
the horizon facing right into new york city.
not a chance. we all went home.
later that night we were writing to each
other and calculating that in an airplane
at 10,000 feet, the event moved up to 10
or 15 degrees above the horizon. also,
in an airplane, we could avoid nyc.
so at 3am, we (me, rob pike, rae mclellan)
went to the airport equipped with cameras
and binoculars. we flew north as high as the
plane would go. we might be the only
people in the world who have seen an
occultation of mercury. thank you joe.
On Tue, Nov 27, 2018 at 6:57 PM, Larry McVoy <lm(a)mcvoy.com> wrote:
As a long time roff fan (I still use it, yes,
I've learned LaTex, I much
prefer roff), I'm hugely bummed that Joe left us so early. I feel like
there would be more fun stories, like Ken's story.
If I remember correctly, he wrote the first (Unix *) version of roff in
PDP-11 assembly, right? Granted, PDP-11 assembly is perhaps the most
pleasant assembly ever, but it is still assembly. Roff is a non-trivial
program, I can't say that I've every written anything remotely that big
in assembly (the only thing I'm proud of is writing swtch() in VAX, 68K,
and some other CPU that I can't remember, but that was tiny, hard to get
right, but tiny). I've got mad respect for what he did, I feel like the
whole roff thing doesn't get enough respect. It wasn't just roff, though
that started it, it was pic (I *love* pic), eqn, all the other filters
that go down to roff. For lmbench I wrote my own grap like tools
because grap wasn't open source.
I was talking to Marc Donner, a Morgan Stanley techy (since moved on
to google and who knows where) about why I liked roff. At the time
I had built webroff which took roff -ms input and made websites.
Marc pointed out that the reason I liked roff was, for the most part,
it didn't say how to do something (that was buried in the macros),
it said what you wanted to do.
Ken, if you have more Joe stories I'd love to hear them, I feel like
I missed out on a cool person.
(*) I know that nroff was "new run off" and it came from somewhere,
MIT? Some old system, but it wasn't invented in Unix. That said,
I've never seen docs for the previous system and I kinda think Joe
took it to the next level. If you haven't studied the docs and
written macros, you should. It's a pretty neat system.
On Tue, Nov 27, 2018 at 03:08:36PM -0800, Ken Thompson via TUHS wrote:
joe was much more than that. he knew how
to play the system. example:
out of whole cloth, he invented a form to
order a teletype and opx (bell labs extension)
installed in the home. he then filled out the
form for each of the unix-room dennisons.
there was a phone call from a confused
clerk, and then we all got teletypes and
data sets at home. as an aside, the opx
came with free watts (long distance which
was very expensive in those days.)
On Tue, Nov 27, 2018 at 1:47 PM, Dave Horsfall <dave(a)horsfall.org> wrote:
> We lost J.F. Ossanna on this day in 1977; he had a hand in developing Unix,
> and was responsible for "roff" and its descendants. Remember him, the
next
> time you see "jfo" in Unix documentation.
>
> -- Dave
--
---
Larry McVoy lm at mcvoy.com http://www.mcvoy.com/lm 
2:44 p.m.
On Tue, Nov 27, 2018 at 11:50 PM Ken Thompson via TUHS <tuhs(a)minnie.tuhs.org>
wrote:
http://www.mcvoy.com/lm
another joe:
echo 1 was a 100 foot balloon that was
launched into space in the early 60s. this
was the first satellite that was easily visible
to the naked eye.
joe wrote a set of fortran programs that
tracked the orbit of echo and calculated
the direction to look from a point on earth.
to do this, he had to learn fortran and
orbital dynamics.
the programs were used to point antennas
to send emf from california. bouncing off
echo and received at bell labs in
new jersey. thus, thanks to joe, echo was
the first communications satellite.
by the time i came to bell labs (1966) the
program, azel, for azimuth/elevation, was
expanded to track planets, moons, satellites,
etc. moreover, it tracked the shadow of the
earth cast by the sun (night). it could predict
within a few seconds when echo would wink
on or off as it passed through the shadow.
This is an amazing story; thanks for sharing, Ken.
There is an interesting film about project ECHO on youtube:
https://www.youtube.com/watch?v=sY8MeZ6lpwI
While it doesn't mention Joe Ossanna directly, there is a small part in the
film where the satellite is located after being launched. Given your story,
one might reasonably assume that that part of the narrative refers
to Ossanna's program, albeit indirectly.
Btw: I've heard that interference detected through the horn antenna at the
Holmdel site lead was explained by cosmic background radiation that was
attributed to the Big Bang; this apparently provided critical observational
evidence that led to the acceptance of the Big Bang theory.
a version of azel was maintained all the time
i was at bell labs. we used it to predict
eclipses, transits, occultations etc. when
we first got a voice synthesizer, the day's
predictions were spoken at 5pm in case
there was anything interesting.
anyway, at 5pm on june 8, 1983 the voice
announced an "occultation of mercury"
for early the next morning.
no one had heard of such a thing. it was
extremely rare. mercury had to be at
about its max elongation; the moon had
to be only a few hours old (or young);
it had to be dark; the moon and mercury
had to be above the horizon; and lastly,
the moon had to occult mercury.
we all (me, lee mcmahon, dennis ritchie,
rob pike, and bob morris) frantically tried
to verify that it was real. it was, but it
would only be about 5 degrees above
the horizon facing right into new york city.
not a chance. we all went home.
later that night we were writing to each
other and calculating that in an airplane
at 10,000 feet, the event moved up to 10
or 15 degrees above the horizon. also,
in an airplane, we could avoid nyc.
so at 3am, we (me, rob pike, rae mclellan)
went to the airport equipped with cameras
and binoculars. we flew north as high as the
plane would go. we might be the only
people in the world who have seen an
occultation of mercury. thank you joe.
!! That's neat.
- Dan C.
On Tue, Nov 27, 2018 at 6:57 PM, Larry McVoy <lm(a)mcvoy.com> wrote:
As a long time
roff fan (I still use it, yes, I've learned LaTex, I much
prefer roff), I'm hugely bummed that Joe left us so early. I feel like
there would be more fun stories, like Ken's story.
If I remember correctly, he wrote the first (Unix *) version of roff in
PDP-11 assembly, right? Granted, PDP-11 assembly is perhaps the most
pleasant assembly ever, but it is still assembly. Roff is a non-trivial
program, I can't say that I've every written anything remotely that big
in assembly (the only thing I'm proud of is writing swtch() in VAX, 68K,
and some other CPU that I can't remember, but that was tiny, hard to get
right, but tiny). I've got mad respect for what he did, I feel like the
whole roff thing doesn't get enough respect. It wasn't just roff, though
that started it, it was pic (I *love* pic), eqn, all the other filters
that go down to roff. For lmbench I wrote my own grap like tools
because grap wasn't open source.
I was talking to Marc Donner, a Morgan Stanley techy (since moved on
to google and who knows where) about why I liked roff. At the time
I had built webroff which took roff -ms input and made websites.
Marc pointed out that the reason I liked roff was, for the most part,
it didn't say how to do something (that was buried in the macros),
it said what you wanted to do.
Ken, if you have more Joe stories I'd love to hear them, I feel like
I missed out on a cool person.
(*) I know that nroff was "new run off" and it came from somewhere,
MIT? Some old system, but it wasn't invented in Unix. That said,
I've never seen docs for the previous system and I kinda think Joe
took it to the next level. If you haven't studied the docs and
written macros, you should. It's a pretty neat system.
On Tue, Nov 27, 2018 at 03:08:36PM -0800, Ken Thompson via TUHS wrote:
> joe was much more than that. he knew how
> to play the system. example:
> out of whole cloth, he invented a form to
> order a teletype and opx (bell labs extension)
> installed in the home. he then filled out the
> form for each of the unix-room dennisons.
> there was a phone call from a confused
> clerk, and then we all got teletypes and
> data sets at home. as an aside, the opx
> came with free watts (long distance which
> was very expensive in those days.)
>
>
> On Tue, Nov 27, 2018 at 1:47 PM, Dave Horsfall <dave(a)horsfall.org>
wrote:
> > We lost J.F. Ossanna on this day in
1977; he had a hand in developing
Unix,
> > and was responsible for "roff"
and its descendants. Remember him,
the next
> time
you see "jfo" in Unix documentation.
>
> -- Dave
--
---
Larry McVoy lm at mcvoy.com 
5:08 p.m.
On 11/27/18, Ken Thompson via TUHS <tuhs(a)minnie.tuhs.org> wrote:
another joe:
echo 1 was a 100 foot balloon that was
launched into space in the early 60s. this
was the first satellite that was easily visible
to the naked eye.
joe wrote a set of fortran programs that
tracked the orbit of echo and calculated
the direction to look from a point on earth.
to do this, he had to learn fortran and
orbital dynamics.
by the time i came to bell labs (1966) the
program, azel, for azimuth/elevation, was
expanded to track planets, moons, satellites,
etc. moreover, it tracked the shadow of the
earth cast by the sun (night). it could predict
within a few seconds when echo would wink
on or off as it passed through the shadow.
a version of azel was maintained all the time
i was at bell labs. we used it to predict
eclipses, transits, occultations etc. when
we first got a voice synthesizer, the day's
predictions were spoken at 5pm in case
there was anything interesting.
What a great story. There is today a website (heavens-above.com) that
does the same thing as Joe's azel. Amateur Astronomers visit it
regularly to get the night's predictions for visible satellite
transits, visible passes of the International Space Station, etc. I
had no idea the idea went back that far.
-Paul W.
7:03 p.m.
On Nov 28, 2018, at 12:08 PM, Paul Winalski
<paul.winalski(a)gmail.com> wrote:
> a version of azel was maintained all the time
> i was at bell labs. we used it to predict
> eclipses, transits, occultations etc. when
> we first got a voice synthesizer, the day's
> predictions were spoken at 5pm in case
> there was anything interesting.
Was this the source or inspiration for astro(1)? Ken’s astro was run daily to announce
things in the Unix room, and in my home up until recently to time turning on the lights in
the
evening. (My home still does the Unix-room style astro announcements, but in the morning
at 8:05).
It is available in the Plan 9 for Unix stuff
(https://codeload.github.com/9fans/plan9port/zip/master
<https://codeload.github.com/9fans/plan9port/zip/master>)
ches
5:57 p.m.
I don't know if this was just me, but the inner geek in me first thought
was "How did the pictures turn out"? :-)
(the second thought was "Joe is now a hero to me" even thought I didn't
meet him... and this sounds sooo much like what
I've done with other geeky friends in college, etc... ).
An example today, I got a Bluetooth water bottle (as part of a reward for
something at work)
and when I set it up it needed a firmware update, which I thought was cool
(and my wife just rolled her eyes...). :-)
Earl
On Tue, Nov 27, 2018 at 11:50 PM Ken Thompson via TUHS <tuhs(a)minnie.tuhs.org>
wrote:
another joe:
echo 1 was a 100 foot balloon that was
launched into space in the early 60s. this
was the first satellite that was easily visible
to the naked eye.
joe wrote a set of fortran programs that
tracked the orbit of echo and calculated
the direction to look from a point on earth.
to do this, he had to learn fortran and
orbital dynamics.
the programs were used to point antennas
to send emf from california. bouncing off
echo and received at bell labs in
new jersey. thus, thanks to joe, echo was
the first communications satellite.
by the time i came to bell labs (1966) the
program, azel, for azimuth/elevation, was
expanded to track planets, moons, satellites,
etc. moreover, it tracked the shadow of the
earth cast by the sun (night). it could predict
within a few seconds when echo would wink
on or off as it passed through the shadow.
a version of azel was maintained all the time
i was at bell labs. we used it to predict
eclipses, transits, occultations etc. when
we first got a voice synthesizer, the day's
predictions were spoken at 5pm in case
there was anything interesting.
anyway, at 5pm on june 8, 1983 the voice
announced an "occultation of mercury"
for early the next morning.
no one had heard of such a thing. it was
extremely rare. mercury had to be at
about its max elongation; the moon had
to be only a few hours old (or young);
it had to be dark; the moon and mercury
had to be above the horizon; and lastly,
the moon had to occult mercury.
we all (me, lee mcmahon, dennis ritchie,
rob pike, and bob morris) frantically tried
to verify that it was real. it was, but it
would only be about 5 degrees above
the horizon facing right into new york city.
not a chance. we all went home.
later that night we were writing to each
other and calculating that in an airplane
at 10,000 feet, the event moved up to 10
or 15 degrees above the horizon. also,
in an airplane, we could avoid nyc.
so at 3am, we (me, rob pike, rae mclellan)
went to the airport equipped with cameras
and binoculars. we flew north as high as the
plane would go. we might be the only
people in the world who have seen an
occultation of mercury. thank you joe.
On Tue, Nov 27, 2018 at 6:57 PM, Larry McVoy <lm(a)mcvoy.com> wrote:
http://www.mcvoy.com/lm
As a long time roff fan (I still use it, yes,
I've learned LaTex, I much
prefer roff), I'm hugely bummed that Joe left us so early. I feel like
there would be more fun stories, like Ken's story.
If I remember correctly, he wrote the first (Unix *) version of roff in
PDP-11 assembly, right? Granted, PDP-11 assembly is perhaps the most
pleasant assembly ever, but it is still assembly. Roff is a non-trivial
program, I can't say that I've every written anything remotely that big
in assembly (the only thing I'm proud of is writing swtch() in VAX, 68K,
and some other CPU that I can't remember, but that was tiny, hard to get
right, but tiny). I've got mad respect for what he did, I feel like the
whole roff thing doesn't get enough respect. It wasn't just roff, though
that started it, it was pic (I *love* pic), eqn, all the other filters
that go down to roff. For lmbench I wrote my own grap like tools
because grap wasn't open source.
I was talking to Marc Donner, a Morgan Stanley techy (since moved on
to google and who knows where) about why I liked roff. At the time
I had built webroff which took roff -ms input and made websites.
Marc pointed out that the reason I liked roff was, for the most part,
it didn't say how to do something (that was buried in the macros),
it said what you wanted to do.
Ken, if you have more Joe stories I'd love to hear them, I feel like
I missed out on a cool person.
(*) I know that nroff was "new run off" and it came from somewhere,
MIT? Some old system, but it wasn't invented in Unix. That said,
I've never seen docs for the previous system and I kinda think Joe
took it to the next level. If you haven't studied the docs and
written macros, you should. It's a pretty neat system.
On Tue, Nov 27, 2018 at 03:08:36PM -0800, Ken Thompson via TUHS wrote:
> joe was much more than that. he knew how
> to play the system. example:
> out of whole cloth, he invented a form to
> order a teletype and opx (bell labs extension)
> installed in the home. he then filled out the
> form for each of the unix-room dennisons.
> there was a phone call from a confused
> clerk, and then we all got teletypes and
> data sets at home. as an aside, the opx
> came with free watts (long distance which
> was very expensive in those days.)
>
>
> On Tue, Nov 27, 2018 at 1:47 PM, Dave Horsfall <dave(a)horsfall.org>
wrote:
> > We lost J.F. Ossanna on this day in
1977; he had a hand in developing
Unix,
> > and was responsible for "roff"
and its descendants. Remember him,
the next
> time
you see "jfo" in Unix documentation.
>
> -- Dave
--
---
Larry McVoy lm at mcvoy.com 
29 Nov
29 Nov
5:54 p.m.
New subject: [OT] azel - was Re: In Memoriam: J.F.Ossanna
On 2018-11-27 11:48 PM, Ken Thompson via TUHS wrote:
...
a version of azel was maintained all the time
i was at bell labs.
As soon as I read this it's been on my mind to ask: Does this program
survive? Presumably it was Fortran? What did it run on?
--Toby
5:57 p.m.
New subject: [OT] azel - was Re: In Memoriam: J.F.Ossanna
On Thu, Nov 29, 2018 at 10:55 AM Toby Thain <toby(a)telegraphics.com.au>
wrote:
On 2018-11-27 11:48 PM, Ken Thompson via TUHS wrote:
References to it are in the unix man pages, but the actual program doesn't
seem to be in the extant unix images.
Warner
...
a version of azel was maintained all the time
i was at bell labs.
As soon as I read this it's been on my mind to ask: Does this program
survive? Presumably it was Fortran? What did it run on?
28 Nov
28 Nov
5:22 a.m.
On 11/28/18 3:57 AM, Larry McVoy wrote:
(*) I know that nroff was "new run off" and
it came from somewhere,
MIT? Some old system, but it wasn't invented in Unix. That said,
I've never seen docs for the previous system and I kinda think Joe
took it to the next level. If you haven't studied the docs and
written macros, you should. It's a pretty neat system.
Kristaps Dzonsons actually went digging for that as part of his "History
of UNIX Manpages" essay[1]. The abridged version is that roff goes back
to a number of rewrites, the original being apparently RUNOFF on CTSS at
MIT. Source code for RUNOFF and documentation for various parts of
troff history have been preserved on that page. It looks
well-researched, though I have no knowledge first hand.
[1] https://manpages.bsd.lv/history.html
9:20 a.m.
Joe sold the (not really existent) UNIX system to the patent department of AT&T, which
in turn bought the urgently needed PDP11. Without that there would be no UNIX. Without Joe
there would be no UNIX.
At 27 Nov 2018 21:48:51 +0000 (+00:00) from Dave Horsfall <dave(a)horsfall.org>rg>:
We lost J.F. Ossanna on this day in 1977; he had a
hand in developing
Unix, and was responsible for "roff" and its descendants. Remember him,
the next time you see "jfo" in Unix documentation.
-- Dave
29 Nov
29 Nov
3:54 a.m.
Blimey! I never thought that I would get such a huge response; there's no
way that I can summarise everything in a single paragraph next year, but
I'll see what I can do.
Thanks, JFO, for everything.
-- Dave
1968
days inactive
1970
days old
14 comments
12 participants
participants (12)
-
Dan Cross -
Dave Horsfall -
Donald ODona -
Earl Baugh -
Fabio Scotoni -
Ken Thompson -
Larry McVoy -
Paul Winalski -
ron minnich -
Toby Thain -
Warner Losh -
WIlliam Cheswick