return to main page

Connecticut Tape Inventory
and data from those tapes

Robert Garner sent 19 of the tapes acquired with the Connecticut 1401 to Paul Pierce in Portland Oregon. Paul read these tape with his seven track reading equipment and placed the results on his web site. The most convenient way to access these tapes on his web site is to view his "What's New" page.

The page you are viewing in an informal mirror of that part of Paul's web site, plus the data from the tapes converted from BCD to ASCII.

These are conversion routines, written in C from Paul Pierce, default prints ^ rather than spaces
conversion
  • bcd2txt.c
    		• .h files
  • asciibcd.h
  • bitcount.h
  • cvtpar.h
  • parity.h
  • prsf2.h
    		• Van Snyder e-mailed
    "I can make "dumps" of program tapes for anybody who wants them -- 100 characters per line, with word marks. The dump program is attached if you want to use it yourself. This is for SimH-format "tapes."

    The following is a modified mirror of Paul Pierce's information he gathered and read from 19 "Connecticut" 1401 tapes.

    • Organized below from his "What's New"
    • ASCII conversions do NOT use the ^ for space default - blanks are blanks.
    • The mirrored material is in a flat subdirectory /PPierceMirror
    • The 1+ megabyte pictures of the label side of the tape reels are not included -
    • (The Reference Date is listed below to differentiate between the various "1401 SORT 7" tapes.)
    • The ASCII files were converted from the .bcd files using the software provided above
    • files/zones are guesses

    Paul's page w comments & .bcd file .bcd file & size ASCII file files/zones? on ASCII file, comments by Van Snyder
    1401 SORT 7 (09 Jul) adc00249.bcd 151 KBadc00249.txt 1: Operational program, looks like some kind of system generator, followed by an object deck. Doesn't work, but fails differently from adc00236. 2: Object deck, sans boot cards.
    1401 AUTOCODER Mod by OK DATA adc00244.bcd 54 KBadc00244.txt Operational Autocoder tape, with several macros
    1401-FO-052 FORTRAN IV adc00250.bcd 183 KBadc00250.txt Operational compiler
    1401 CLUB PROG adc00240.bcd 36 KBadc00240.txt Operational program, 6 (!) overlays, for "Innis Arden Golf Club"
    1401 DFTS adc00238.bcd 609 KBadc00238.txt 1: Diagnostics tape, looks like several copies of each.
    1401 IOCS adc00237.bcd 271 KBadc00237.txt 1: Program to deblock deck, followed by deck of IOCS macros. 2: Trailer label.
    1401 Sort7 Backup adc00236.bcd 152 KBadc00236.txt 1: Operational program, looks like some kind of system generator, followed by an object deck. Doesn't work. 2: Object deck, sans boot cards.
    1401 Merge 7 adc00234.bcd 512 KBadc00234.txt 1: Operational program. 2: trailer label. 3: a different trailer label.
    1401 AUTOCODER COMPILER adc00232.bcd 68 KBadc00232.txt Operational Autocoder tape, with several macros
    1401 AUTOCODER W/MACROS adc00239.bcd 96 KBadc00239.txt Operational Autocoder tape, with several macros
    1401 FORTRAN IV adc00233.bcd 224 KBadc00233.txt 1: Operational compiler. 2: trailer label. 3: a different trailer label.
    1401 Sort7 (08 Jul ) adc00231.bcd 673 KBadc00231.txt 1: Program to deblock deck, followed by blocked source deck. 2: trailer label. 3: a different trailer label.
    1401-SE-12X adc00229.bcd 246 KBadc00229.txt Very different from adc00227-SE12X.dmp. Looks like a core dump with disassembler 1: Object deck, clear core routine. 2-4: Object decks, unblocked. 5-8: Source decks. 9: trailer label. 10: a different trailer label.
    1401 F104X adc00228.bcd 1,211 KBadc00228.txt 1: Autocoder-format bootable. 2: Source deck "Portfolio Selection." 3: Listing, probably of assembly of file 2, with object cards interleaved (& in CC 1). 4: trailer label. 5: a different trailer label.
    1401 SE12X adc00227.bcd 6,200 KBadc00227.txt Probably "Documentation Aids" program 1: Operational program. 2: Decks, blocked 30 per record, including source for "Documentation aids sample program". 3: Source deck, probably for Documentation Aids system, blocked 30. 4: Deck, blocked 30, not Autocoder source. 5: Listing, probably of assembly of file 3. 6: trailer label. 7: a different trailer label.
    1401 AUTOCODER adc00226.bcd 1,545 KBadc00226.txt 1: Autocoder-format bootable, but probably not useful. 2: Listing. 3: trailer label. 4: a different trailer label.
    1401 KWIC adc00225.bcd 326 KBadc00225.txt ADC00225-SOURCE.TXT, ADC00225-OBJ.TXT 1: Operational program, Autocoder-format tape. 2: Operational program deck. 3: Source deck. 4: trailer label. 5: a different trailer label.
    1401 SORT 7 (25 Jun) adc00224.bcd 1,570 KBadc00224.txt 1: Operational program. 2: trailer label. 3: a different trailer label.
    1401 COBOL adc00223.bcd 4,643 KBadc00223.txt 1: operational compiler. 2: program to dump listings and decks, followed by listing, Autocoder macros that are the compiler's run-time library, and a sample program. 3: trailer label. 4: a different trailer label.

    During the pick-up of the Connecticut 1401, Robert Garner sent to Paul Pierce some of the "interesting" tapes to see what is on them and to see if we can use the contents.

    There have been a variety of progress reports e-mailed:

    - this is an attempt to display them as info -
    - collect info for a better catalog of available software 

    

----- Original Message ----- 
From: "Paul Pierce" 
To: "Robert Garner" ; "John" 
Cc: <1401_team-bounces@computerhistory.org>; 
"IBM 1401 Restoration Team List" <1401_team@computerhistory.org>
Sent: Friday, July 11, 2008 7:17 AM
Subject: [1401_team] "Other tape photos"


All,

Here are photos of the other Connecticut tapes that came in the box, except for the 
impressive stack of skew tapes (which I could read, I suppose, but they tend to be very boring):

    p1010001
    p1010002
    p1010003
    p1010004
    p1010005

    p1010006
    p1010007
    p1010008
    p1010009
    p1010010
    [The above are a much reduced file size of the original] 
    http://www.piercefuller.com/misc/tapepix/connecticut.html
(Warning: ugly quick hack web page!)

Paul
_______________________________________________
1401_team mailing list
1401_team@computerhistory.org
http://mail.computerhistory.org/mailman/listinfo/1401_team

-----------------------------------------------------------------------------

Van, Bill,

Good news.

 > COBOL, FORTRAN, RPG all produced an interim Autocoder deck which  
was then compiled using the Autocoder compiler.

Makes sense that the "source code" for all the compilers/utilities was  
in Autocoder.
(But what about comments?  ;-)

So we now have source code for two key 1401 compilers (COBOL, FORTRAN  
IV) and SORT7  and  IOCS.

- Robert

-----------------------------------------------------------------------------

sent July 10

For the record, about a year ago, I discovered I had the original Autocoder
coding for the FORTYRAN IV compiler which I co-wrote. It, of course,
includes the original comments.

Gary Mokotoff
-----------------------------------------------------------------------------
    On Jul 8, 2008, at 6:29 PM, Paul Pierce wrote:

    All,

    Heres an update on the progress of reading Connecticut tapes.

    I've spun all the red tapes and several other small tapes that look like system tapes. Most of the results are up on my web site, see the Whats New page in the Library or check out the 1401 page for everything there is: http://www.piercefuller.com/library/ibm1401.html

    There were more Sort 7 tapes. One of them has a huge dropout, probably a wrinkle, on the track 0 side in the first record, which is most of the program. But its almost the same as the other one so I'm not inclined to fuss with it more unless there is some good reason.

    One of the red tapes (E) labelled 1401CR02X which is a second copy of KWIC has a lot of little read errors. The parity track is weak most of the way through the good data. I'm still working on it. If it turns out to be identical to (I) I'll abandon it.

    The tape "AUTOCODER Modified by OK DATA" has a bad stretched spot, fortunately long after the last written data. It made a noise on the tape cleaning machine, which stopped, and there was a stretched section there and a few turns later a 6-inch section where the oxide and binder came right off the tape (and slipped into a crack in my tape cleaner.) I don't know if the tape cleaner caused the stretch or if the stretch caused the tape cleaner to go wacko, but either way I think its a good argument for having a drive with a good tape cleaner itself instead of using a separate machine like I do.

    Two tapes labelled "Autocoder 1/2" and "Autocoder 2/2" seem to be 9-track. I can try to fire up my 9-track setup if you want them read.

    I'm thinking of photographing the remaining tapes, so you can make requests if you want any of them read.

    Paul Pierce 

    ------------------------------------------------



----- Original Message ----- 
From: "Van Snyder" 
To: "Paul Pierce" 
Cc: lots
Sent: Wednesday, July 09, 2008 3:47 PM
Subject: Re: "Re: reading 1401 tapes"


> I've converted all of the tapes at
> http://www.piercefuller.com/library/ibm1401.html to SimH format using
> bcd2simh, and produced "dumps" of them with word marks.  Other than the
> COBOL tape, the Fortran tapes (1401-FO-050 v3m0 and v3m4, not Fortran
> IV), and the FSU tape from Al Kossow, I haven't split the files on the
> tapes into separate disk files.
> 
> Let me know if you want any of these, and I'll send them.
> 
> Van




-----------------------------------------------------

----- Original Message ----- 
From: "Paul Pierce" 
To: lots
Sent: Wednesday, July 09, 2008 8:21 PM
Subject: "Re: reading 1401 tapes"

    All,

    I convert to ASCII using the bcd2txt program below, an older but equivalent version of which is supposed to be on my web site somewhere. It converts every code to something different (unless you use -b) so that you can edit in ASCII then convert back.

    There are two character sets, one approximates a typical IBM commercial 1403 print chain (or 407 wheel set, or whatever) and the other approximates scientific. The commercial is default, -f (FORTRAN) selects the scientific. Most system programs want -f.

    As many of you know, you can't write a zero (space character) on even parity tape and even parity was normally used for symbolic records, so the spaces were converted to something else. The -b (blank) switch converts them back. Its not the default because that maps two different codes to the ASCII space character meaning you can't convert back without "loss".

    Since many tapes were commonly blocked with multiple logical records (lines of text) per physical record on tape, there is an optional last parameter to unblock to a fixed logical record length. This makes such tapes human readable when otherwise they are a mess of very long lines. You will find that several of the tapes have files with different blocking factors (logical record length) so, for instance, you may need to convert several times with different parameters so you can read the different sections.

    End-of-file marks are converted to ASCII line-feed.

    There was considerable discussion about the conversion tables once many years ago, probably on alt.folklore.computers but maybe on the old computer collectors mailing list. Bob Supnik's SIMH emulators probably have slightly different tables.

    Feel free to borrow code, or just the tables; feel free to enhance. I'm not bothering to stick a GNU license on it here but please pretend it has one.

    Enjoy,

    Paul Pierce

    This is an "rcs diff", whatever that is. It is readable and easy to make. If you don't know either, just look for the lines starting with "diff" and split it all into files. The line after each "diff" is junk.

    [To view these routines and .h files, see the top of this web page.] 

    ---------------------------------------------------------------------
On July 8, 10:09 PM.   Van wrote

Paul Pierce offers tools to read the files he creates when he reads 
tapes.  I've attached them as a gnuzip-compressed tarball.  Winzip can 
handle this on Windoze.  I wrote the program "to_supnik" to convert 
directly to SimH format.

Also see http://www.piercefuller.com/oldibm-shadow/tool.html, which is 
in the tarball too.

My program tpdump, for dumping 1401 "tapes" in SimH format, is also 
attached.

Van


-----------------------------------------------------
On Wed, 2008-07-09 at 18:04 -0700, Van Snyder wrote:
> adc00233-Fortran-IV-FO-052.dmp
> 1: Operational compiler.  2: trailer label. 3: a different trailer
> label.

This is not an operational tape.  It is a distribution tape.  If you
boot it, it punches four decks.  See /C24-3322-2 page 73.

Van

-----------------


On Tue, 2008-07-08 at 22:10 -0700, Robert Garner wrote:
> # 09 Jul 2008 1401 SORT 7
> # 09 Jul 2008 1401 AUTOCODER Mod by OK DATA
> # 09 Jul 2008 1401-FO-052 FORTRAN IV
> # 08 Jul 2008 1401 CLUB PROG
> # 08 Jul 2008 1401 DFTS
> # 08 Jul 2008 1401 IOCS
> # 08 Jul 2008 1401 Sort7 Backup
> # 08 Jul 2008 1401 Merge 7
> # 08 Jul 2008 1401 AUTOCODER COMPILER
> # 08 Jul 2008 1401 AUTOCODER W/MACROS
> # 08 Jul 2008 1401 FORTRAN IV
> # 08 Jul 2008 1401 Sort7
> # 08 Jul 2008 1401-SE-12X
> # 08 Jul 2008 1401 F104X
> # 25 Jun 2008 1401 SE12X
> # 25 Jun 2008 1401 AUTOCODER
> # 25 Jun 2008 1401 KWIC
> # 25 Jun 2008 1401 SORT 7
> # 25 Jun 2008 1401 COBOL 
> 
> 
> 
> Could you ascertain whether any are source tapes, or are they all
> binaries? (Is there a simple BCD-character-per-byte reader for
> examining your dump files?) 

> p.s.  I agree -- no reason to read the other Sort7 tapes.

Two of the tapes labeled "SORT 7" appear to start with some kind of
system generator program, but neither one works.  It might be useful to
get the other Sort 7 tapes read to see if one of them has a working
system generator on it.

Using Paul's file names, with an explanatory suffix:

adc00223_Cobol.dmp
1: operational compiler. 2: program to dump listings and decks, followed
by listing, Autocoder macros that are the compiler's run-time library,
and a sample program. 3: trailer label. 4: a different trailer label.

adc00224-SORT7.dmp
1: Operational program. 2: trailer label. 3: a different trailer label.

adc00225-KWIC.dmp
1: Operational program, Autocoder-format tape.  2: Operational program
deck.  3: Source deck. 4: trailer label. 5: a different trailer label.

adc00226-AUTOCODER.dmp
1: Autocoder-format bootable, but probably not useful.  2: Listing. 3:
trailer label. 4: a different trailer label.

adc00227-SE12X.dmp
Probably "Documentation Aids" program
1: Operational program.  2: Decks, blocked 30 per record, including
source for "Documentation aids sample program".  3: Source deck,
probably for Documentation Aids system, blocked 30.  4: Deck, blocked
30, not Autocoder source.  5: Listing, probably of assembly of file 3.
6: trailer label. 7: a different trailer label.

adc00228-F104X.dmp
1: Autocoder-format bootable.  2: Source deck "Portfolio Selection."  3:
Listing, probably of assembly of file 2, with object cards interleaved
(& in CC 1).  4: trailer label. 5: a different trailer label.

adc00229-SE-12X.dmp
Very different from adc00227-SE12X.dmp.  Looks like a core dump with
disassembler
1: Object deck, clear core routine.  2-4: Object decks, unblocked.  5-8:
Source decks.  9: trailer label. 10: a different trailer label.

adc00231-SORT7.dmp
1: Program to deblock deck, followed by blocked source deck.  2: trailer
label. 3: a different trailer label.

adc00232-AUTOCODER.dmp
Operational Autocoder tape, with several macros

adc00233-Fortran-IV-FO-052.dmp
1: Operational compiler.  2: trailer label. 3: a different trailer
label.

adc00234-MERGE7.dmp
1: Operational program.  2: trailer label. 3: a different trailer label.

adc00236-SORT7.dmp
1: Operational program, looks like some kind of system generator,
followed by an object deck.  Doesn't work.  2: Object deck, sans boot
cards.

adc00237-IOCS.dmp
1: Program to deblock deck, followed by deck of IOCS macros.  2: Trailer
label.

adc00238-DFTS.dmp
1: Diagnostics tape, looks like several copies of each.

adc00239-AUTOCODER+macros.dmp
Operational Autocoder tape, with several macros

adc00240-CLUB_PROG.dmp
Operational program, 6 (!) overlays, for "Innis Arden Golf Club"

adc00244-AUTOCODER-mod_by_OK_data.dmp
Operational Autocoder tape, with several macros

adc00249-SORT7.dmp
1: Operational program, looks like some kind of system generator,
followed by an object deck.  Doesn't work, but fails differently from
adc00236.  2: Object deck, sans boot cards.

adc00250-Fortran-IV-FO-052.dmp
Operational compiler

    Updated August 25, 2008