return to main 1401 Restoration Page

RE: CT 1401 successfully runs FORTRAN compiler!

E-mails, in chronological order, some omitted 
--------------------
Subject: Re: Wednesday demo
From:	Marc Verdiell 	 	 
Date:	Wed, Jan 10, 2018 10:10 pm
To:	Ken Ross 
Cc:	...
Glad the CT worked and wasn't too tired. 
We ran the FORTRAN compiler from tape earlier in the day (thanks to Michael Albaugh), 
  that gave it quite a workout! 
As Frank noticed watching the lights, "I had never seen the 1401 so busy"...

Marc
----------------------
From:	Robert Garner 	 	 
Date:	Wed, Jan 10, 2018 10:27 pm
To: ...
Bob,

Quoting from Marc’s email:
> We ran the FORTRAN compiler from tape earlier in the day 
>  (thanks to Michael Albaugh), that gave it quite a workout!
> As Frank noticed watching the lights, "I had never seen the 1401 so busy"

Did not an unflappable restorer ;-) once maintain that 
   “our 1401s are incapable of running the FOTRAN compiler!?”

An astounding counter-proof and accomplishment everyone!  ;-)))

- Robert

p.s.  I suspect Marc is producing the next episode of Mission Impossible...  ;-)
--------------------------------

On Jan 18, 2018, at 3:54 AM, ed@ed-thelen.org wrote:

> Did not an unflappable restorer ;-) once maintain that 
>  “our 1401s are incapable of running the FOTRAN compiler!?”

There is the fact that the 729 Tape System Emulator
   could not  handle the record lengths used in some of the
   system tapes, I forget if it was
     ? assembler
     ? FORTRAN

-Ed Thelen
--------------------------------
From:	Michael Albaugh  	 
Date:	Thu, Jan 18, 2018 9:57 am
To:	Robert Garner 

> Mike Albaugh, cc’d, can describe what he did to 
>   work around the 729 Emulator's block size limitation...

Put simply, I did the simplest thing that could work at all.
Since I did not want to lean too hard on my very rusty autocoder
skills, I used a nearly trivial scheme.

1) Write a program the split the records. Each record gets split
into chunks less than a specified length. The only tricky part is
the make sure the last record is still > 13 characters. Then, if
this is the last (sub) record of the original, I add a '.'. Otherwise,
I add a '&'.

2) Write a 1401 program that reads a "tape" of this form, appending
each record in a buffer (if the previous record ended in '&') until
I get the one ending in '.', then writing the re-merged record.
Of course, I strip out the trailing '&' and '.' in the process.

I am not on the computer that has these files (convenient excuse,
since they are embarrassingly slapdash), but I will be cleaning
up, documenting, and putting them on the web "soonish". Right now
various other emergencies are ahead of this task.

-Mike
--------------------------------
From:	Van Snyder 	 	 
Date:	Thu, Jan 18, 2018 1:29 pm
To:	ed@ed-thelen.org

Ed:

I sent a split/recombine program in 2010, to use to create a "real" tape
by sending split records through the emulator. I tried to run it on the
CT machine but the 729's were not cooperating that day.

The split program writes a recombine program on the output tape. Then
it writes split records. After the splitter runs, you boot from the
output tape and it writes the original tape contents, with their
original record lengths.

I also sent a revised Fortran compiler that had sufficiently short
records that it ought to work through the emulator.

Here it is again. Split_Recombine-VanSnyder

Van
---------------------