IBM mainframe printer typeface - equivalent font?
Mike Coley
using what I think must be a high speed belt (impact) (IBM?) printer.
(BTW, the characters are all uppercase.) I want to be able to print
these documents from a desktop using a truetype or type 1 font.
Moreover, I want the output to look **exactly** the same (or reeeeealy
close.) But I am having a devil of a time identifying the original
typeface, or its origin. It is a monospaced font that looks a lot like
OCR-B, but the capital B and D have little pointy serifs, and the
capital C and G look more octagonal than rounded (especially noticable
at large type sizes.) I found a sample at one site called "OCR-B BON"
that looked really close, but it was hard to tell from the low res
image. Does this ring a bell with anyone in the mainframe world, and
has anyone ever taken these old-time typefaces and converted them to
modern fonts? Any bright ideas are appreciated.
Thanks,
Mike
Mxsmanic
Many line printers can have a number of different types of print trains
(belts) mounted, so the exact appearance of the characters depends on
the train chosen. You really need to put a sample up somewhere to see
how it looks. It does sound like an OCR-compatible belt but it might
not be.
--
Transpose hotmail and mxsmanic in my e-mail address to reach me directly.
us...@hotmail.com
Tim Murray
(in article <da06e6b4.04090...@posting.google.com>):
> I have documents that are being printed on a big mainframe system
> using what I think must be a high speed belt (impact) (IBM?) printer.
> (BTW, the characters are all uppercase.) I want to be able to print
> these documents from a desktop using a truetype or type 1 font.
I used to have, but have since lost it, a font aptly named Line Printer that
matched a common belt from IBM. Maybe it's still floating around out there.
Dick Margulis
Tim Murray wrote:
Just a bit of historical accuracy.
There were never belts. The line printers from the fifties and early
sixties (up through the 1400 and 7000 series computers) used chains.
Starting with the System 360, they used trains.
A chain had a complete character set in alphanumeric sort order,
followed by another such set, followed by another. These characters were
affixed to links in a roller chain. I don't recall the actual number of
sets, but we're talking about uppercase letters, 0-9, and a few
punctuation marks and special characters, plus control codes. IIRC the
whole BCD set was octal coding and therefore consisted of 64 characters
and control codes. The chain circulated rapidly, and there was a hammer
at each column position (120 columns). When the next instance of the
desired character whizzed by a given column, the hammer would strike.
These matchups were done independently, so the hammers would not strike
in left to right order. Speeds got up over 1200 lines per minute IIRC.
A train consisted of unattached blocks of sintered metal, one character
to a block, running around a race. There were three copies of each
character adjacent to each other, followed by three of the next
character, etc. This was the EBCDIC character set, (Extended Binary
Coded Decimal Interchange Code), which had lowercase letters and
additional punctuation. I'm pretty sure the character set was the same
size as ASCII (128 total available positions, including control codes),
but distributed among the 256 possible hexadecimal numbers. EBCDIC could
be instantly converted to ASCII and vice versa, but they were not the
same encodings.
As the train traversed the print head, again there was a hammer for each
column on the printer. The enhancement to print quality came from not
hitting the same character twice in a row. (A solid line of one
character really slows down either kind of line printer.) So if columns
1, 3, and 5 all required an A, the first A in the block of three would
be printed in column 1, the second one in column 3, and the third one in
column 5. The next column that needed an A would get the first one
again. This strategy compensated for the recoil after being struck, when
hitting the same character again would produce a poor impression (and
probably shorten the life of the character).
The big advantage of the train over the chain was that it greatly
reduced friction, allowing for much higher printing speeds.
Character
-- Here's part of a response I wrote yesterday and decided not to send
... but since you've brought this all up ...
IBM's print trains (chains before that) were available in a (small)
variety of type styles and character sets, including both upper and
lower case. The 1403 printer trains had 240 characters in 5 sections of
48 each. If you wanted a larger character selecton (lower case, for
instance), the printing would be slower because it had to wait longer
for the right characters to be in the right positions (of which there
were 132 to a line).
(In fact, it's a chain of "Special Characters" that my usenet nym is
from - it was the name of our computer department bowling team).
On the print trains, the styles and character positions were modified to
minimize ink build-up and wear-and-tear brought on by the repetitive
heavy impacts.
The available styles corresponded very roughly to those on IBM
typewriters, and type balls. I remember that Orator, Script, and APL
sets were available.
- Character
Character
That rang a bell, and I found it in my archives. It's not too bad a match.
I'll post the set (there's a bold, italic, and bold italic, too) in
alt.binaries.fonts and e-mail it to the OP.
- Character
Alan J. Flavell
> Dick Margulis wrote:
>
> > Just a bit of historical accuracy.
> >
> > There were never belts.
We certainly had a kind of belt technology on an IBM remote job entry
and print station. But I wouldn't have called that a "mainframe"
printer.
> > A train consisted of unattached blocks of sintered metal, one
> > character to a block, running around a race.
Surely the 1403 print trains had 3 (different) characters per slug,
didn't they?
> IBM's print trains (chains before that) were available in a (small)
> variety of type styles and character sets, including both upper and
> lower case.
Yes, although they could be customised. We had one that we had
customised ourselves, starting from a TN layout (of course the printer
driver software had to be told about it too, by loading a print train
image which matched the reality). (This was with VM/CMS, to be
specific). I -think- the driver actually downloaded the train image
into the printer itself.
TN was the one with upper and lower case characters, and some
technical symbols (but not quite enough for our requirements at the
time, hence the customisation).
Google finds this page http://www.columbia.edu/acis/history/chain.html
- it seems not quite to have made up its mind whether it's a chain or
a train. Sloppy terminology, I suppose.
cheers
Mxsmanic
trains, drums, combs, etc. I don't know how many of these were used in
IBM printers, however.
Dick Margulis
Mxsmanic wrote:
> There were (and are) many different printing methods: chains, belts,
> trains, drums, combs, etc. I don't know how many of these were used in
> IBM printers, however.
>
Possibly all of them. However AFAIK the high-speed line printers we
associate with mainframes were chains followed by trains, at least in
the period I was talking about. There may well be newer technologies
since I left the mainframe world that I'm unaware of.
Dick Margulis
Character wrote:
>>
> IBM's print trains (chains before that) were available in a (small)
> variety of type styles and character sets, including both upper and
> lower case. The 1403 printer trains had 240 characters in 5 sections of
> 48 each. If you wanted a larger character selecton (lower case, for
> instance), the printing would be slower because it had to wait longer
> for the right characters to be in the right positions (of which there
> were 132 to a line).
Thanks for the correction. It was the old card listers that had 120,
although I think there may also have been some line printer models that
did--now I'm not sure.
>
> (In fact, it's a chain of "Special Characters" that my usenet nym is
> from - it was the name of our computer department bowling team).
Does your wife know?
> On the print trains, the styles and character positions were modified to
> minimize ink build-up and wear-and-tear brought on by the repetitive
> heavy impacts.
Modified on the basis of character frequency, à la etaoin shrdlu? Or
some other way?
Mxsmanic
> Possibly all of them. However AFAIK the high-speed line printers we
> associate with mainframes were chains followed by trains, at least in
> the period I was talking about. There may well be newer technologies
> since I left the mainframe world that I'm unaware of.
I believe trains were the last of the impact printers. Of course, they
are still in use, for applications with low volume and a need for impact
printing. I've seen mostly trains and drums. If the characters are
unevenly spaced horizontally, it's probably a train; if they jump up and
down with respect to the baseline, it's probably a drum printer. Drums
are slower.
Nowadays, of course, very high-volume applications (and even
not-so-high-volume applications) use laser printing technologies and
page printers.
Character
>
>
> Character wrote:
>
>>>
>> IBM's print trains (chains before that) were available in a (small)
>> variety of type styles and character sets, including both upper and
>> lower case. The 1403 printer trains had 240 characters in 5 sections
>> of 48 each. If you wanted a larger character selecton (lower case,
>> for instance), the printing would be slower because it had to wait
>> longer for the right characters to be in the right positions (of which
>> there were 132 to a line).
>
>
> Thanks for the correction. It was the old card listers that had 120,
> although I think there may also have been some line printer models that
> did--now I'm not sure.
Yes, first there were 120-wide printers (maybe the later printers could
handle both). They were the same width maximum line; the 132/line type
was smaller.
>> (In fact, it's a chain of "Special Characters" that my usenet nym is
>> from - it was the name of our computer department bowling team).
>
>
> Does your wife know?
Yep - but I think I'm missing the point?
>> On the print trains, the styles and character positions were modified
>> to minimize ink build-up and wear-and-tear brought on by the
>> repetitive heavy impacts.
> Modified on the basis of character frequency, à la etaoin shrdlu? Or
> some other way?
I think they were just in alphabetic order, one to a set. I was
referring to the design and physical structure of the metal characters
rather than their placement.
I also have a vague memory of what looked like a flat brass loop on some
printer or other.
- Character
Tim Murray
> Just a bit of historical accuracy.
>
> There were never belts. The line printers from the fifties and early
> sixties (up through the 1400 and 7000 series computers) used chains.
> Starting with the System 360, they used trains.
The IBM System/32 used a belt. It was about the same thickness a very thin
band saw. The impact mechanism consisted of hammers; the movement was driven
by two rubber wheels, one of them driven at a constant speed. I'm pretty
sure all product literature (both customer and internal) referred to the word
"belt". I carried a took lit for IBM for seven years, but that was a long
time ago.
Mike Coley
don't match what I have. I have uploaded a sample here:
https://www.ase.com/regbatches/uploads/ASE20040907123955mikesfontsample.bmp
As you can see, the B has pointy serifs (the D is the same way),
rounded characters (O, C, and G) are sort of clipped on the corners
(semi-octagonal rather than rounded), and the outside bars of the M
and W both run straight up and down. As I said initially, it looks a
lot like OCR-B, but has a few distinct characteristics.
Now does this ring a bell with anyone?
Thanks!
Mike Coley
Mxsmanic
> Now does this ring a bell with anyone?
It looks like output from a page printing system, not an impact printer.
The outlines of the characters are characteristic of the scattering of
toner on high-speed page printers, and so are the shapes of the
characters.
Are you _certain_ that this came from an impact printer? Does the
original page have indentations in the paper from the impact of the
hammers? Do you see any signs of the fabric weave in the ribbons in the
printed characters (especially obvious when the ribbons are drying out)?
Mike Coley
> Are you _certain_ that this came from an impact printer? Does the
> original page have indentations in the paper from the impact of the
> hammers? Do you see any signs of the fabric weave in the ribbons in the
> printed characters (especially obvious when the ribbons are drying out)?
I am not certain of too much of anything. I do know that the system
has been around since the early 1980's, I have been told that the
paper used comes in big rolls, and they print thousands of documents
in a day. I assumed it was some sort of impact printer, but when I
can't feel or see any indentations in the paper, front or back. The
letters are all uniformly dark; no indication of ribbon wear.
What is a page printer? Is it using a soft font (through something
like IBM's APF)? The document does have other text in this same
font/typeface at three different point sizes (approx. 18, 12, and 10
pt.), and some of the text is oriented upside down relative to the
other parts of the text. I suppose this would all be impossible for a
single pass impact printer? But I am not aware of laser or inkjets
that run at high speed or use paper in enormous rolls. What manner of
beast is this? And does that help to point us toward the source of the
font/typeface?
Thanks,
Mike
Mike Coley
printed in the 1980s and 1990s. I still can't detect any impact
indentations anywhere, but the font/typeface changed around 1991-92
from a similar monospaced font to the one that is currently used
(sample available at the address a couple of messages up in the
thread). I don't know whether this indicates a change in the
underlying printer technology, but I thought I would throw it out
there.
Thanks,
Mike
Mxsmanic
> I am not certain of too much of anything. I do know that the system
> has been around since the early 1980's, I have been told that the
> paper used comes in big rolls, and they print thousands of documents
> in a day.
It's a page printer. They were just starting to become popular in the
early 1980s, and the paper did indeed come in big rolls (and the
original ones used the old-style smelly photocopying technology, so the
paper was thick and slick and curled easily, with a distinctive odor).
I thought I recognized it. The systems used a high-speed charging and
toner deposition system that put lots of very tiny "specks" of tonor
around characters. Still, it was razor sharp compared to impact
printing, and it was always aligned, and it was wickedly fast even in
the early printers (at least a page a second, and much faster very soon
after).
> I assumed it was some sort of impact printer, but when I
> can't feel or see any indentations in the paper, front or back. The
> letters are all uniformly dark; no indication of ribbon wear.
Page printers are xerographic printers that paint the the characters
onto paper with laser light, just like today's laser printers and
photocopiers. Which means that there's a software version of their
fonts somewhere inside the printers, although I'm sure it's not a type 1
font or anything like that (it would be something proprietary, probably
a bitmapped font).
> What is a page printer?
Any printer that prints a page at a time, instead of a line at a time or
a character at a time. The only practical way to print a page at a time
is with non-impact methods, like laser printers do (page printers are
essentially laser printers that are designed for speed rather than
aesthetics).
> Is it using a soft font (through something like IBM's APF)?
Yes, the fonts are encoded in software somehow, with the details
dependint on the printer. I don't remember what APF is, though.
> The document does have other text in this same
> font/typeface at three different point sizes (approx. 18, 12, and 10
> pt.), and some of the text is oriented upside down relative to the
> other parts of the text. I suppose this would all be impossible for a
> single pass impact printer?
Right.
> But I am not aware of laser or inkjets that run at high speed
> or use paper in enormous rolls.
Most of the world of mainframes is unknown to the general public. There
are specialized computer printers that print at hellishly high speeds
using laser-printer-style technology. Today all these printing
technologies are tending to merge, but page printers were especially
distinctive when they first came out, in the early 1980s (perhaps
earlier, I don't remember). They were in use before desktop lasers came
into use.
> What manner of beast is this?
The last time I worked with one, it was about 2-3 times the size of a
line printer (and line printers are already beasts in themselves), and
blazed through an entire heavy roll of paper in no time. I presume they
still use paper rolls today, since it's a lot faster than sheet feeders,
but I'm not sure. Good page printers do a very nice job (they did a
nice job 20 years ago, too). They can often imprint other things on the
page besides variable text, i.e., they can print an entire form on the
page along with the text, allowing plain or nearly plain paper stock to
be used. As far as I know, there aren't any color page
printers--perhaps because speed is more important than pretty colors.
Of course, this all tends to blur with digital presses now. But a
digital press is overkill for simple payroll stubs or electric bills,
and a page printer is too inflexible for the type of color printing a
print shop might need to do.
> And does that help to point us toward the source of the font/typeface?
I'm sure the font is proprietary. You could scan and digitize a full
sample to create a font, though.
Mxsmanic
Probably just a change in printer font. It's pretty easy to distinguish
between impact printing and page printers when looking at a printout.
Impact printing always has the very slightly fuzzy look of something
printed with a hammer hitting a ribbon, whereas page printing has the
sharp, smooth, high-contrast, perfectly-aligned look of a laser printer.
You'll see impact printers today in applications where the printing
volume is relatively low, or where legal or other concerns require
printing in indelible ink with impact, or where carbons are required
(the main case of this today being systems that print through a mask
sheet to an NCR-paper copy underneath, in order to hide things like PINs
inside an envelope), or where highly specialized stock is required
(something that wouldn't fit smoothly into a laser printing engine).
Line printers usually use tractor forms with sprocket holes along the
side of the form--you can often remove the holes, but you can still see
the trace of a perforated edge along the side of the form.
In other applications, with high volumes, unusual requirements for
variable content (line printers can do only horizontal lines of fixed
text), etc., page printers are used.
Mxsmanic
page-printing systems, the Infoprint 4100:
http://www.printers.ibm.com/internet/wwsites.nsf/vwwebpublished/4100home_ww
As you can see, page printers are still huge. But at least they now
support TrueType and OpenType fonts. And they are faster than ever, at
1200 pages per minute (double-sided!).
Trevor Frew
<marg...@comcast.net> wrote:
<chopt>
>> On the print trains, the styles and character positions were modified to
>> minimize ink build-up and wear-and-tear brought on by the repetitive
>> heavy impacts.
>
>Modified on the basis of character frequency, à la etaoin shrdlu? Or
>some other way?
And I thought I'd got away from the linotype keyboard long ago!
cmfwyp
Should I use that as a pseudonym?
Trevor Frew
---
Acupuncture is a jab well done.
Usual Suspect
You were in IBM, weren't you ? I'm SURE I saw your name in the internal
forums ...
Now that your pension has been taken from you - and maybe recompensated
after
years of debates ! - perhaps you can tell us all what you were doing in IBM
?
"Dick Margulis" <marg...@comcast.net> wrote ...
Dick Margulis
Usual Suspect wrote:
> Dick Margulis !
>
> You were in IBM, weren't you ? I'm SURE I saw your name in the internal
> forums ...
> Now that your pension has been taken from you - and maybe recompensated
> after
> years of debates ! - perhaps you can tell us all what you were doing in IBM
> ?
>
>
>
The only way I might have been listed on an IBM internal forum is if
someone from IBM came out to the antiwar picket line I was on and
gathered names for a newsletter. When IBM was _the_ game in town, I
worked on IBM equipment, but never as an IBM employee. In any case, it
was all a long time ago. I think the last time I had anything to do with
IBM-anything was probably around 1971, when I briefly used a Selectric
Composer (not an MTST, unfortunately) to set WIN Magazine (published by
War Resisters League) and the couple years after that when my only
connection was the occasional march or protest in front of one of their
plants or offices. In the thirty years since then, the closest I've come
is meeting IBMers and ex-IBMers socially on occasion.