[TUHS] Fun with early troff.

[TUHS] Fun with early troff.

[Ron Natalie]

I started with roff (the simplest but utterly frozen) and moved up to nroff.   It was a few years later I was involved with a rather project to make a CAT phototypesetter emulator for the Versatec printer-plotter (similar to the BSD vcat, which we had not seen yet).   My friend George Toth went to the Naval Research Laboratory and printed out the entire typeface on their CAT on transparent film.     Then he set out to figure out a way to digitize it.

Well, next door the the EE building (where the UNIX work took place at JHU) was the biophysics department.   They had Scanning Transmission Electron Microscope there, quite an impressive machine.    The front end of the thing was a PDP-11/20 with some digital to analog converters and vice versa and a frame buffer.    The software would control the positioning of the beam and read back how much came through the material and was detected.     Essentially, you were making a raster picture of the sample in the microscope. 

George comes up with this great idea.    He takes a regular oscilloscope.   He takes the deflection wires from the 11/20 off the microscope and puts them in the X and Y amplifiers of the scope.   He then put a photomultiplier tube in the shell of an old scope camera.    He'd cut out a single character and tape it the front of the scope and hang the camera  on it.   He'd fire up the microscope software and tell it to scan the sample.    It would then put the image in the frame buffer.   We'd pull the microscope RK05 pack out and boot miniunix and read the data from the frame buffer (why we didn't just write software to drive the A2D from miniunix I do not recall).

Eventually, George gets everything scanned in and cleaned up.    It worked somewhat adequately.

Another amusing feature was that Michael John Muuss (my mentor) wrote a macro package tmac.jm.    Some people were someone peeved that we now had a "nroff -mjm" option.

Years later after ditroff was in vogue, my boss was always after me to switch to some modern document prep (Framemaker or the like).   On one rough job I told him I'd do it but I didn't have time to learn framemaker.
I write one page of this proposal, print it and then go on.   My boss would proof it and then my coworker would come behind me and make the corrections.   I ended up rewriting a million dollar (a lot of money back in 1989 or so) proposal in two days, complete with 25 pages of narrative and may be 50 pages of TBL-based tables showing compliance with the RFP.    We won that contract and got several follow ons.

Years later I was reading a published book.   I noted little telltale bumps on the top of some of the tables.   I wrote the author..."Did you use tbl and pic to typeset this book?"    Sure enough he had.   But it was way after I thought anybody was still using  such technology.   Of course, I was happy when Springer-Verlag suddenly learned out to typeset books.    I had a number of their texts in college that didn't even look like the put a new ribbon in the typewriter when setting the book.

[TUHS] Fun with early troff.

[Larry McVoy]

+1.  I've written my own macro packages, I have an outliner that I
can produce troff from, the BK User Guide (dated though it is) looks
like it is troff -ms source, and for the most part it is, you can
roff the whole thing, and I wrote a tool that produces this html:

http://www.mcvoy.com/lm/bkdocs/UG/tmp/

I was talking to Marc Donner (at Morgan Stanley at that time) about how
much I loved roff and said "You love it because the high level macros
are saying _what_ you want not _how to do it_".  The how to do it is
inside the macro definition, if the macro package is complete you rarely
resort to \fBwhatever\fP you just say
.B whatever
which is really really parseable.

The other thing I love is how version control friendly it is.  All those
line breaks are super helpful to version control.

On Tue, May 08, 2018 at 01:44:37PM -0400, Ron Natalie wrote:
> I started with roff (the simplest but utterly frozen) and moved up to nroff.   It was a few years later I was involved with a rather project to make a CAT phototypesetter emulator for the Versatec printer-plotter (similar to the BSD vcat, which we had not seen yet).   My friend George Toth went to the Naval Research Laboratory and printed out the entire typeface on their CAT on transparent film.     Then he set out to figure out a way to digitize it.
> 
> Well, next door the the EE building (where the UNIX work took place at JHU) was the biophysics department.   They had Scanning Transmission Electron Microscope there, quite an impressive machine.    The front end of the thing was a PDP-11/20 with some digital to analog converters and vice versa and a frame buffer.    The software would control the positioning of the beam and read back how much came through the material and was detected.     Essentially, you were making a raster picture of the sample in the microscope. 
> 
> George comes up with this great idea.    He takes a regular oscilloscope.   He takes the deflection wires from the 11/20 off the microscope and puts them in the X and Y amplifiers of the scope.   He then put a photomultiplier tube in the shell of an old scope camera.    He'd cut out a single character and tape it the front of the scope and hang the camera  on it.   He'd fire up the microscope software and tell it to scan the sample.    It would then put the image in the frame buffer.   We'd pull the microscope RK05 pack out and boot miniunix and read the data from the frame buffer (why we didn't just write software to drive the A2D from miniunix I do not recall).
> 
> Eventually, George gets everything scanned in and cleaned up.    It worked somewhat adequately.
> 
> Another amusing feature was that Michael John Muuss (my mentor) wrote a macro package tmac.jm.    Some people were someone peeved that we now had a "nroff -mjm" option.
> 
> Years later after ditroff was in vogue, my boss was always after me to switch to some modern document prep (Framemaker or the like).   On one rough job I told him I'd do it but I didn't have time to learn framemaker.
> I write one page of this proposal, print it and then go on.   My boss would proof it and then my coworker would come behind me and make the corrections.   I ended up rewriting a million dollar (a lot of money back in 1989 or so) proposal in two days, complete with 25 pages of narrative and may be 50 pages of TBL-based tables showing compliance with the RFP.    We won that contract and got several follow ons.
> 
> Years later I was reading a published book.   I noted little telltale bumps on the top of some of the tables.   I wrote the author..."Did you use tbl and pic to typeset this book?"    Sure enough he had.   But it was way after I thought anybody was still using  such technology.   Of course, I was happy when Springer-Verlag suddenly learned out to typeset books.    I had a number of their texts in college that didn't even look like the put a new ribbon in the typewriter when setting the book.

-- 
---
Larry McVoy            	     lm at mcvoy.com             http://www.mcvoy.com/lm 

[TUHS] Fun with early troff.

[Ron Natalie]

My first job after graduation was to work on a uber-classified project at
Martin Marietta.   These guys had just come up with the idea of using the
PWB tools to manage government projects (this was 1981) so since I had some
time to kill waiting for my security clearance to come through, I helped
them set up the UNIX machines that were to be the source archives, etc.

One of things we needed was some stuff to handle classification markings
both for regular line printer (code) output and also for nroff.    I knew
Dennis Mumaugh at the NSA had written such so I wrote to him and he says
"sure, send me a blank tape."     Well several months later a double wrapped
package comes to my HOUSE (not the office).   Great.    I take it back to
the SCIF and open it up.    Turns out it's just my blank tape (I guess the
NSA can't send anything out if not double wrapped).    So I set forth to add
a field to all the .H directives in -mm to include the security
classification.   Then at the page break I gathered the highest markings and
all the code words and printed them at the top and bottom of page (this
required the "col" filter to be used for nroff output going to our
printers).      I then also buggered the lpr command to look for
classification markigns in the what strings  and print the appropriate
banners there.

When later doing some contract work for IBM, I re-stylized the -mm package
to make the output look just like IBM's UNIX manuals.   I'm not sure what
they actually used internally.

[TUHS] Fun with early troff.

[Dan Cross]

On Tue, May 8, 2018 at 1:44 PM, Ron Natalie <ron at ronnatalie.com> wrote:

> [...]
> Years later I was reading a published book.   I noted little telltale
> bumps on the top of some of the tables.   I wrote the author..."Did you use
> tbl and pic to typeset this book?"    Sure enough he had.   But it was way
> after I thought anybody was still using  such technology.   [...]
>

At the end of 2009 I was recalled to active duty in the Marines to deploy
to Afghanistan, where I would embed with an Afghan Army Unit. Since I was
coming from the reserves, I got to spend a couple of months stateside
getting retrained on the sort basic Marine stuff you forget unless you're
doing it every day. A big portion of that was getting re-familiarized with
radio protocol and the "standard" messages for doing things like calling in
a helicopter casualty evacuation, unexploded ordinance, combat contact,
etc. For example, there is a NATO-standard "10-line" message for calling in
a casualty evacuation that contains things like, "how many people were
injured?", "how were they injured?", "what are the manner of injuries?",
"what is the location of the nearest landing zone?" and "how is the LZ
marked?"; that sort of thing.

As part of that training I had to spend a lot of time on a static-y radio
talking to helicopter pilots sitting on folding chair the next building
(pretending they were flying helicopters), but the part that is relevant is
that we were given a set of "cheat sheets" that contained all the standard
formats in tabular format. I remember looking at them and seeing similar
tell-tale bumps on the lines of the tables and thinking, "hey, this looks
like it was formatted using troff!"

After the initial retraining, I got sent to an active duty unit to go
through their "workup" and actually deploy with them. Naturally, once I got
to Camp Pendleton to join meet up with the active duty folks I wanted to
give all of my Marines (I also had two sailors and a US Army soldier under
my command too, but that's another story -- suffice it to say that the USMC
doesn't have medical personnel and we get them from the other services)
copies of those same cheatsheets and get them trained up on talking to the
birds and so forth so that any of us could call in an evac if necessary (or
call for fire, call in a TIC, or whatever). But my little cheatsheet was
laminated and didn't photocopy well. So, that evening when I got back to
the "BOQ" I fired up a laptop and typed in the troff to reproduce those
little cards. The result was *exactly* the same output, pixel for pixel (as
near as I could tell). I printed out a bunch of copies, took them to the
local Kinkos and laminated them, and distributed them the next day.

Yup. Troff. I took it to war.

        - Dan C.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://minnie.tuhs.org/pipermail/tuhs/attachments/20180508/e93cd533/attachment.html>

[TUHS] Fun with early troff.

[Kurt H Maier]

On Tue, May 08, 2018 at 04:14:54PM -0400, Dan Cross wrote:
>
> For example, there is a NATO-standard "10-line" message

When did they add the 10th line?  What is in it?  I was trained on the
9-line during my service, and I'd be fascinated to know what extra info
was judged important enough to update the manuals.

> 
> Yup. Troff. I took it to war.
> 

It's been well over a decade, but there were QRF bases in Kabul whose
security manuals and operation maps were generated onsite in TeX on my 
ancient Thinkpad.  I suspect, however, an academic study of combat-zone 
typesetting would be dominated by the inevitability of Powerpoint.

I've often noticed the battlefield is ruled by Windows (e.g. those
pilots were probably poking at Falconview) while the research side is
extremely tied to unix and its ilk.  Obviously BRLCAD and other ARL
involvement here, and just about all of of the DoD HPC program skews
toward unix and linux.

khm

[TUHS] Fun with early troff.

[Dan Cross]

On Tue, May 8, 2018 at 4:43 PM, Kurt H Maier <khm at sciops.net> wrote:

> On Tue, May 08, 2018 at 04:14:54PM -0400, Dan Cross wrote:
> >
> > For example, there is a NATO-standard "10-line" message
>
> When did they add the 10th line?  What is in it?  I was trained on the
> 9-line during my service, and I'd be fascinated to know what extra info
> was judged important enough to update the manuals.


Oh man. I had to go look it up and I found a copy on some website at
Lejeune and another one at training command. It looks like the NATO
standard *9* line request plus a line for patient information (name,
initials, last4, etc). Odd considering that if you've got multiple patients
you'd need more than one "line". Anyway, there's a copy in this document:
http://www.trngcmd.marines.mil/Portals/207/Docs/TBS/THULS.pdf (search for
"CASEVAC").

>
> > Yup. Troff. I took it to war.
> >
>
> It's been well over a decade, but there were QRF bases in Kabul whose
> security manuals and operation maps were generated onsite in TeX on my
> ancient Thinkpad.  I suspect, however, an academic study of combat-zone
> typesetting would be dominated by the inevitability of Powerpoint.


Neat. What branch were you?

I used to get super-annoyed when higher would muck with my manifests which
were, of course, Excel documents. "How do you know these total numbers are
correct?" "Because that cell is a formula that is the sum of all the other
relevant cells and unless you think that Microsoft can't ADD then there's
no reason for it to be incorrect. Please tell Lance Corporal Schmuckatelli
to stop overwriting my shit because he definitely can't add. Oh, and he
should probably get counseled to go take, 'Math for Marines' again; this
time without cheating and copying the answers out of the back fo the MCI."

I've often noticed the battlefield is ruled by Windows (e.g. those
> pilots were probably poking at Falconview)


Those helo pilots were just sitting on chairs eating potato chips in an
otherwise empty shipping container out at MOUTown and making things up as
they went along.

while the research side is
> extremely tied to unix and its ilk.  Obviously BRLCAD and other ARL
> involvement here, and just about all of of the DoD HPC program skews
> toward unix and linux.


Funny, the BFT in my M-ATV ran Linux. I got the guys down at the
maintenance depot to give me login name and password for their account so I
could play solitaire when nothing was going on.

        - Dan C.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://minnie.tuhs.org/pipermail/tuhs/attachments/20180508/fc3722f6/attachment.html>