Re: Rounding error
- From: glen herrmannsfeldt <gah@xxxxxxxxxxxxxxxx>
- Date: Mon, 04 Jun 2007 14:36:22 -0800
Richard Maine wrote:
Arno <arnoinperu@xxxxxxxxxxx> wrote:
What is the point of having f7.2 rather than any other f7.x in a read
statement? It seems that indicating the number of decimal digits is
irrelevant, or am I missing something else?
Also, this goes all the way back to the IBM 704, a vacuum tube computer.
(snip)
Today, I'd say that the most common use of the ".2" in contexts like
this is just to be able to use the same format for input and output. It
means something for output, and is mostly irrelevant to input. I've done
that.
Maybe that makes sense today. In days past, one needed carriage
control, at least an initial 1X, on the output format, but not
on the input format. (Unless you want to waste one column.)
That was enough to require different FORMAT statements.
On the subject of computers past and formatted output, the
description of carriage control in the IBM 704 manual includes:
blank (single space before printing)
0 (double space before printing)
+ (no space before printing)
1-9 (skip to channels 1-9)
As implemented for S/360, and maybe not too different for the
704, there is a punched paper tape specifying which row on the
page carriage control codes 1-9 should skip to. It is usual for
1 to be the top of the page, but that isn't required. Skipping
to other parts of the page allows one to print forms much faster
than would otherwise be possible. As an example:
http://www.museumofhoaxes.com/pranks/mcdonalds.html
(I did know some of the people involved, but I wasn't
there at the time.)
Many computer centers charge by the page and also by the number
of actual lines of output. For large print jobs, a custom
carriage control tape would speed up printing and reduce costs.
I believe that one was used for the prank described.
(Though it is well known that the above prank could have been done
just as well on more ordinary printing systems.)
-- glen
.
- Follow-Ups:
- Re: Rounding error
- From: Richard Maine
- Re: Rounding error
- References:
- Rounding error
- From: Alenko
- Re: Rounding error
- From: gsal
- Re: Rounding error
- From: Richard Maine
- Re: Rounding error
- From: gsal
- Re: Rounding error
- From: Richard Maine
- Re: Rounding error
- From: Arno
- Re: Rounding error
- From: Richard Maine
- Rounding error
- Prev by Date: Re: Rounding error
- Next by Date: Re: Rounding error
- Previous by thread: Re: Rounding error
- Next by thread: Re: Rounding error
- Index(es):
Relevant Pages
|