return to main page
More Print Character Wars
- e-mails - 2011Newest e-mails near the top - this point
added to web page April 28, 2011
Robert,
I don't remember the lozenge character's having any special use with the 40x tabulators. They had no means of being programmed -- other than wired control panels. Asterisks (not lozenges) were used for totals -- * for a subtotal and ** for a major total.
FORTRAN and COBOL had problems when there were no (, ), +, =, and ' ... only a %, lozenge, &, # , and @ . The 1403-001 and 1403-002 were the only ones that attached to the 1401. The 1416 print chain came either with an A or H character set. The H set had the scientific characters and the A set had the commercial characters. As you know, the 1401 basically didn't care which character set was used because the internal BCD code was the same.
As a 1401 Autocoder/IOCS programmer, I would have liked to have had an H set because we found the + and ' characters easier to read in the programs, but the bank where I worked was only commercial. So the programmers adjusted. The 407 tabulator that we used to list the programs was equally "commercial" -- as was the 557 interpreter. If the bank had desired, they could have rented an 026 keypunch with a scientific character set, but they felt the need was to have an additional backup commercial keypunch.
With regard to the other characters you mention towards the end of your note, special characters (Greek, Russian, etc.) were allowed but, for a price and they replaced one of the standard characters on the 48 character A/H print chain. Which of the 48 would you have sacrificed of the 26 letters, 10 numbers and 12 special characters?
After S/360's announcement, the print chain/train was able to handle up to 240 print characters with 1403 models that had a Universal Character Set (UCS) feature. Adding more characters slowed down the print speed of the 1403 because there were fewer repetitions of them.
Regards,
Bill
Fran,
> I realize now that the reason I stressed that the use of the Lozenge as a 'check protection' character was because I included it in the "edit' control word for just that purpose
Do you recall if that was because 40X folks already been using it for that purpose or because you thought that would be a good application for it?* Do you recall it being used as a total's indicator, per Dick's reference from Mackenzie text below?
Any thoughts on Dick's other question:
"The 1401 manual's Character Code Chart, p.170, indicates "BCD Code" for the column of binary encodings. But I don't think that's correct. The BCD for zero is 0000 (Mackenzie p.31), the 1401 zero is 1010 (8-2) which is BCDIC. So why does the 1401 manual state BCD?"
I've personally never warmed up to the term "BCD" ("Binary Coded Decimal") to describe the character set, as it only refers to encoding of decimal digits, not the alphabetic and special characters.
Hope all is going well. (I need to edit your transcripts of the 1401 oral history we did last year!! And bounce some more 1401 questions your way.)
- Robert
* I understand these Lozenge questions are trifling and minutiae; perhaps that's what makes them interesting. ;-) For those on cc list, I had been asking: "In the 40X accounting machine era, how was the Lozenge character actually used?" What struck me was the lack of the "+" sign from the accounting machine era (now I understand ;-). Lack of a "+" character sure gave grief to the FORTRAN developers! ;-)
According to Brook's architecture textbook, it was Backus and the FORTRAN designers on the 702 with 704 printer (limited to the conventional 48 "commercial" character set) that defined the so-called "dual" assignments (creating a mess -- for some reason they couldn't/didn't pick a 2nd set of new punch/BCD combinations at the same time):
FORTRAN Commercial
= #
' @
( %
+ &
) lozenge -- gone from today's keyboards. ;-)
I see in the 1403 maintenance manual that there were many special character chains: Greek, Russian, Arabic, etc.
Where all these used by 1401 customers??
IBM Almaden Research, San Jose, CA
Office: 408-927-1739
Mobile: 408-679-0976
robgarn@us.ibm.com
From: "fran" <mrfran1@wildblue.net>
To: <dmanning8@stny.rr.com>, "Gene Darrow" <darrowehdeld@aol.com>, Robert B Garner/Almaden/IBM@IBMUS, <rweaver@ix.netcom.com>
Date: 04/20/2011 06:35 AM
Subject: Re: Lozenge
Thanks for the input. I realize now that the reason I stressed that the use of the Lozenge as a 'check protection' character was because I included it in the "edit' control word for just that purpose This didn't preclude its use for other purposes.
Fran
----- Original Message -----
From: rweaver@ix.netcom.com
To: Robert B Garner ; fran
Cc: Robert B Garner ; Robert Garner ; Ed Thelen ; Van Snyder
Sent: Wednesday, April 20, 2011 2:46 AM
Subject: Re: Lozenge
Happened to be reading in Mackenzie's "Coded Character Sets, History and Development", 1980, & found the following on pages 66-67 (if he had indexed it, I'd have found it when you 1st asked!)
"During the 1950s, the advent of computers such as the IBM 702, 705, and 1401 saw the expansion of BCDIC into 47 graphics, and also the development of a 6-bit code to represent those graphics. With one exception, the 11 special symbols served an obvious purpose in one or another commercial application. .... The exception was the special symbol lozenge. Because the lozenge appeared on printer chains, it was put to various uses; for example to indicate, in the margin of a tabulation, final totals as contrasted to subtotals."
That doesn't tell us why the lozenge was on the chain, only that it had no specific purpose. Possibly just an arbitrary choice, something to fill an open position?
The reason I was reading such an exciting book? The 1401 manual's Character Code Chart, p.170, indicates "BCD Code" for the column of binary encodings. But I don't think that's correct. The BCD for zero is 0000 (Mackenzie p.31), the 1401 zero is 1010 (8-2) which is BCDIC. So why does the 1401 manual state BCD?
I made a mistake (not really news, I make a lot of them). In an earlier email I'd stated that the 407 had no memory beyond pilot selectors (for one card cycle) and counters. In fact it also had 4 storage units, 16 characters each (numeric, alphabetic - no special characters) that could hold contents indefinitely. Seems I confused "I never used" with "the machine doesn't have", sorry.
My question above re BCD / BCDIC - I would really like to know.
dick w
-----Original Message-----
From: Robert B Garner
Sent: Mar 4, 2011 1:37 PM
To: fran
Cc: Robert B Garner , Robert Garner , rweaver@ix.netcom.com, Ed Thelen , Van Snyder
Subject: Re: Lozenge
Fran,
Yes, I recall your answer, but other old timers/CEs (at the CHM) don't recall seeing lozenge used, even for the purpose you mention.
It must not have been common to fill unprinted numeric positions on checks with lozenges?
- Robert
p.s. You probably didn't think it was odd that 407's, etc didn't have a print symbol for "+"?
I suppose all numbers were assumed to be positive, otherwise there was the negate "-" symbol. ;-)
Imagine that: An era when programs were wired into plug boards, so there was never a program listing on a printer that would need symbols for addition and subtraction! ;-)
According to Brook's architecture textbook, it was Backus and the FORTRAN designers on the 702 (704 printer, limited to the conventional 48 "commercial" character set) that defined the so-called "dual" assignments (creating a mess -- for some reason they couldn't/didn't pick a 2nd set of new punch/BCD combinations at the same time):
704 FORTRAN
# =
@ '
% (
+ &
) lozenge -- not on our standard keyboards anymore. ;-)
IBM Almaden Research, San Jose, CA
Office: 408-927-1739
Mobile: 408-679-0976
robgarn@us.ibm.com
From: "fran" <mrfran1@wildblue.net>
To: Robert B Garner/Almaden/IBM@IBMUS
Date: 03/04/2011 07:00 AM
Subject: Lozenge
Robert:
I was copied on your last email to Fred Brooks, and you ask again about the use of the lozenge in UR equipment.
I've explained several times that the lozenge was a printable character, introduced on the 407, only, as a 'check protection' character for printing checks, in place of leading zeros or asterisks.
Is this not clear, or is there some dispute about this? Chuck Branscomb, Shel Jacobs, Jim Ingram and some others from this era should be able to confirm or deny this.
Fran