[1401_software] Object deck translation program
Bob Feretich
Bob.Feretich at rafresearch.com
Mon Sep 26 18:33:40 PDT 2005
It seem that we need a translation program written to be able to punch
our cards at Cardmation.
From the "System Operation Reference Manual IBM 1401 Data Processing
System",
GA24-3067-2.
1401 "Full BCD" Encoding:
Numeric b 0 1 2 3 4 5 6 7 8 9 2 3 4 5 6 7
Rows b 8 8 8 8 8 8
==================================================
No zone 0 1 2 3 4 5 6 7 8 9 SB # @ : > TM
12 zone & ? A B C D E F G H I . LO [ < GM
11 zone - ! J K L M N O P Q R $ * ] ; MC
0 zone / S T U V W X Y Z RM , % WS \ SM
Where:
SB = Substitute Blank (Manual says that SB can not be read from cards.)
LO = Lozenge
TM = Tape Mark
GM = Group Mark
MC = Delta
RM = Record Mark
WS = Word Separator
SM = Tape Segment Mark
If I translate this table from "Full BCD" to EBCDIC to ASCII , to
achieve the same punch codes, the below substitutions need to be made.
The translation from "Full BCD" to EBCDIC resulted in all of the changes
except for one. The MC translates to the PL/1 "not" symbol, which is
represented in ASCII as ^.
To generate the above punches, it seems that Cardmation input should be:
Cardmation Encoding:
Numeric b 0 1 2 3 4 5 6 7 8 9 2 3 4 5 6 7
Rows b 8 8 8 8 8 8
==================================================
No zone 0 1 2 3 4 5 6 7 8 9 : # @ ' = "
12 zone & { A B C D E F G H I . < ( + |
11 zone - } J K L M N O P Q R $ * ) ; ^
0 zone / S T U V W X Y Z \ , % _ > ?
This does not match the SIMH encoding.
Can someone:
1) Check my substitution table, to see if I made any mistakes.
2) Write a translation program to convert our objects decks and Van's
assembler output to this encoding.
Regards,
Bob Feretich
-------- Original Message --------
Subject: Re: Card punch services/product - Differences in Characters
printed
Date: Mon, 26 Sep 2005 17:35:26 -0400
From: bob at cardamation.com
Reply-To: tech_sales at cardamation.com
To: Bob Feretich <bob.feretich at prodigy.net>
CC: Robert Garner <us032560 at mindspring.com>
Bob:
Hi Bob - There are some differences in a few characters
For example . a question mark ? which the ASCII translation table for
the IBM ASCII punch machines as punch code 0-8-7. While we can punch
automatically the combination you want, 12-0 will result in a
different character.
The same is true for 6-8. This results in the character printed of
equal = rather than >.
Three of the solutions are -
(a) Find an 026 to punch out your cards and print the characters you want.
(b) For an extra - $ 50- setup ($100 total setup) we can punch
the correct characters in the card and probably arrange that a print
format is provided with substitute ASCII bytes to print the characters
in accordance with your program cards This assumes that are print set
includes all the characters you want printed.
In this solution, one set of characters is punched and with a few
substitutions, another set of characters is printed above the punched
holes.
(c) If there are not too many cards with too many corrections required
for your cards for the differences between the old 026 key punch and
IBM's 129 and 029 format which we use then , we punch the correct
combination and the "wrong" symbols are printed for a few punctuation
marks etc. Then with pen and ink your group can change the "wrong"
printed characters to the ones you want (or we can do that for a modest
fee.)
A) Requested 026 EBCDIC characters
Suggestion # 1
Please provide a fax number and we will fax you an ASCII translation
table for the IBM 3780 terminal or 029 ASCII data recorder..
This will show the print characters and hex as well as the punch
combination.
.....................................................................................
All of the character combinations you specified can be punched, but the
print across the top may be different for the 029, or 129 or our
printing card punch (versus an IBM 026).
General Suggestion - # 2
Send us a file with all of the character combinations you want and let
us see what results. You can then determine if you like it.
We will charge - $ 25- for the effort.
And this will be credited to the $ 100 set up (for the approach where
your company sends us files with changes in the printed combinations
for some punch combinations.
Answer to Question re short record and 80 character printing
There are two possibilities in printing across the top of the card.
1 - All the characters are printed above the punched holes.
2 - For an extra charge, we can print the information punched wherever
you want in the 80 characters. Also, the printed and punched
information can be different - all in one card pass.
Now - the way the card record works is that 80 data characters are
followed by carriage return line feed.
If desired, for a card with less than 80 characters punched, once the
last data character is included, you can provide a carriage return line
feed and the card will punch. In the absence of special interpreting
or printing, a printed character is printed above each column with holes.
As suggested above, we have various ways that we can print the
information punched (or other information) wherever you wish across the
top edge of the card.
Hopefully, this is clear.
Bob Swartz
---------------------------------------------------------------------------------------
Date sent: Sat, 24 Sep 2005 13:22:59 -0700
From: Bob Feretich <bob.feretich at prodigy.net>
To: tech_sales at cardamation.com
Copies to: Robert Garner <us032560 at mindspring.com>
Subject: Re: Card punch services/product
It looks like we are going to proceed with using your card punch
service. Before I send files to be punched, I would like to verify the
ASCII to card hole mappings.
Alphanumeric characters (A-Z and 0-9) have straight forward ASCII
mappings. But, what ASCII representation do I use for the below
special characters.
The encoding is "026 Commercial":
Character Punch
Lozenge 12-4-8
Group Mark 12-7-8
Delta 11-7-8
Word Separator 0-5-8
Tape Segment Mark 0-7-8
Tape Mark 7-8
Record Mark 0-2-8
Also, the below encodings seem obvious, but it doesn't hurt to check.
Character Punch . 12-3-8 [ 12-5-8 < 12-6-8 & 12 $ 11-3-8 * 11-4-8 ]
11-5-8 ; 11-6-8 - 11 / 0-1 , 0-3-8 % 0-4-8 \ 0-6-8 # 3-8 @ 4-8 : 5-8
> 6-8
? 12-0
The background printing on the card is not important. The 5081 format
is fine. Either of the top corners may be cut, as long as all of the
cards are consistent.
Your email stated that "If a card record is less than 80 characters
long, the carriage return line feed will cause the card to be punched
with one line of characters printed across the top edge of the card."
Is the printing related to the record being less than 80 characters
long? I expected the printing to occur on the top edge of the card
regardless of the length of the ASCII record.
Regards,
Bob Feretich
bob at cardamation.com wrote:
> Dear Bob:
>
> See replies below.
>
> Cardamation Co. Inc.
> Bob Swartz
>
> -------------------------------------------------
>
> Date sent: Wed, 21 Sep 2005 22:53:12 -0700
> From: Bob Feretich <bob.feretich at prodigy.net>
> To: sales at cardamation.com,
> service_bureau at cardamation.com Copies to: Robert
> Garner <us032560 at mindspring.com> Subject: Card
> punch services/product
>
> I am working with the Computer History Museum in San Jose, CA.
> (http://www.computerhistory.org/).
> We are restoring an IBM1401 computer with 1402 card reader/punch.
> The system is now able to read cards from the 1402. We have read our
> master set of diagnostic object decks into PC disk files and would
> to punch duplicate copies of these decks.
>
> 1) What do you charge for punching cards?
> Answer:
> The standard price is $ 100- set up and $0.15 per card. In view of
> the fact that this is for a computer museum, we will lower this to
> $50 set up. If more than 500 cards are requested to be punched in
> one run, we will reduce the price to $0.12 per card or $75.-
> whichever is higher.
>
> We assume that the cards punched would be in standard 5081 format
> (with rows and columns printed on the cards). Other card formats
> might involve a plate charge of $30- to print the cards and a small
> extra charge for cards ($0.02 per card) for punching less than 2,000
> cards.
>
> 2) What file formats do you accept for input?
>
> Answer
> The format we would like to receive is an ASCII text file. We
> prefer receiving enail records with data followed by carriage return
> line feed for each 80 column card. (although we may be able to
> work with the card record data followed by only a carriage return ).
> If the card record is less than 80 characters long, the carriage
> return line feed will cause the card to be punched with one line of
> characters printed across the top edge (row 13) of the card.
>
> Please insert space character for each column in which there is no
> punching.
>
>
> 3) If we had a long term need for punching cards, what product is
> available? Price?
>
> Answer:
> We can provide an IBM 129 data recorder in working condition for
> $1,000 to be shipped from Illinois at the moment This is an off line
> unit with data entry from keyboard. However, that unit is subject
> to prior sale, but we likely can find a bargain working 129.
>
> Our lowest cost models to punch 80 column cards from a pc computer
> file are:
>
> Cardamation RP8791 List price $7,800 (special price $ 4,500)
>
> This machine includes an RS232C interface and DOS software.
> It punches and interprets the cards in one pass at the rate of 900
> cards per hour (15 cards per minute). It also includes a keyboard
> for off line key punching, verifying, reproducing and interpreting
> cards. It has two input hoppers and two output hoppers. It is very
> convenient for reproducing and key punching decks of punched cards
> off line.
>
> 3M 948 Encoder List price $ 4,800 (special price $ 2,500) This
> is for on line punching and printing one line of data under control
> of a pc computer DOS program. ( only in an online mode).
>
> 4) Would you consider donation arrangement with the museum? (If yes,
> I will have museum management contact you, I am just a project
> volunteer.)
>
> We have some obsolete punched card machines we would consider
> donating - but I do not think this includes card punches right now -
> but could later. The museum would need to pay the freight from the
> Philadelphia area. For example, we might donate a Decision Data 9610
> column Data Recorder. It is about 32 years old. (It reads and
> punches the small 96 column cards that were used with the IBM System
> 3.).
>
> Answers by Bob Swartz
>
>
> Regards,
> Bob Feretich
>
>
> bob at cardamation.com
bob at cardamation.com
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mail.computerhistory.org/pipermail/1401_software/attachments/20050926/d4b3641e/attachment-0001.htm
More information about the 1401_software
mailing list