How a Hardware Genius Turned a 1930s Teletype Into a Linux Terminal

[Peter Flass]

https://www.popularmechanics.com/science/a30647778/linux-teletype-programming-hardware-video/


--
Pete

[Scott Lurndal]

Peter Flass <peter...@yahoo.com> writes:
>https://www.popularmechanics.com/science/a30647778/linux-teletype-programming-hardware-video/
>

Considering that:

1) Unix was developed on an ASR-33 teletype
2) Converting from RS232C to 20ma (or 60ma) current loop requires
a small handful of discrete components and was widely used to
interface unix systems to said teletypes almost fifty years
ago
3) Most uarts still support 5 bit serial transmission at very slow
baud rates (the 'u' in uart does stand for universal :-)
4) Unix worked with all-uppercase terminals just fine, including
the ability to enter lower-case characters via a two character
sequence.

I'm not sure it's such a big deal that linux can also do this.

[robin....@gmail.com]

On Thursday, April 30, 2020 at 11:25:44 PM UTC+10, Peter Flass wrote:
> https://www.popularmechanics.com/science/a30647778/linux-teletype-programming-hardware-video/

That was done in the 1970s by "Electronics Australia" magazine
using a microprocessor.

No hardware alterations were required.

[robin....@gmail.com]

It was in October 1976 that Jim Rowe designed the interface
ASCII-BAUDOT translator.

[Bob Eager]

When I started using UNIX, we only had a couple of mixed case terminals.
The rest were ASR-33s.

We managed.

--
Using UNIX since v6 (1975)...

Use the BIG mirror service in the UK:
http://www.mirrorservice.org

[Quadibloc]

On Thursday, April 30, 2020 at 8:11:10 AM UTC-6, Scott Lurndal wrote:

> I'm not sure it's such a big deal that linux can also do this.

Well...

It's certainly true that Linux has a command-line interface.

Even making that work on an ASR 33 terminal might be problematic, as Unix tends
to have a preference for lower-case. Also, the character | which is not in the
upper-case only subset of ASCII is very useful in that operating system.

So using a 5-level terminal, with an even more restricted character set,
probably would require, at least, including some sort of character translation
layer.

It is true that hooking up 5-level Teletypes to computers was rather common in
the early days of the microcomputer era. Still, I'm not surprised that there is
media attention to their use, if only because they're now almost a forgotten
technology.

John Savard

[Scott Lurndal]

Quadibloc <jsa...@ecn.ab.ca> writes:
>On Thursday, April 30, 2020 at 8:11:10 AM UTC-6, Scott Lurndal wrote:
>
>> I'm not sure it's such a big deal that linux can also do this.
>
>Well...
>
>It's certainly true that Linux has a command-line interface.
>
>Even making that work on an ASR 33 terminal might be problematic, as Unix tends
>to have a preference for lower-case.

The part you snipped, above, stated that unix worked just fine with
upper case. As Bob pointed out, the ASR-33 was the standard terminal
for unix development.

Also, the character | which is not in the
>upper-case only subset of ASCII is very useful in that operating system.

It was, however, a character on the ASR-33 keyboard.

>
>So using a 5-level terminal, with an even more restricted character set,
>probably would require, at least, including some sort of character translation
>layer.

Generally handled by the terminal driver based on 'stty' settings.

[Peter Flass]

Right, I think they made a bit too much of it, but I still thought it was
interesting. Apparently his next project is something to so with Morse
Code.

--
Pete

[John Levine]

In article <6ac9db73-c51c-4299...@googlegroups.com>,

<robin....@gmail.com> wrote:
>It was in October 1976 that Jim Rowe designed the interface
>ASCII-BAUDOT translator.

There have been two way ASCII<->Baudot translators as long as there
has been ASCII, used to forward telex messages between the older
Baudot Telex and TWX networks and the newer ASCII TWX network. I
believe that started in 1959.

Back in the 1980s when I had an MCI Mail account which had a gateway
between their ASCII mail system and the Baudot telex network. I used
it to send joke telexes to hotels in Europe where my father was
staying, most starting with DISREGARD PREVIOUS MESSAGE.

It's a cute hack but it's not technically very interesting.
--
Regards,
John Levine, jo...@taugh.com, Primary Perpetrator of "The Internet for Dummies",
Please consider the environment before reading this e-mail. https://jl.ly

[John Varela]

On Thu, 30 Apr 2020 14:58:48 UTC, Quadibloc <jsa...@ecn.ab.ca>
wrote:

We had KSR- and ASR-28s talking to the AN/FSQ-7 at Lincoln Labs in
1961. A 1962 photo of SAGE consoles that were used for experiments
in applying air defense technology to air traffic control is at
<https://flickr.com/photos/184595836/498378291/in/>

--
John Varela

[Quadibloc]

On Thursday, April 30, 2020 at 10:33:51 AM UTC-6, Scott Lurndal wrote:

> Also, the character | which is not in the
> >upper-case only subset of ASCII is very useful in that operating system.

> It was, however, a character on the ASR-33 keyboard.

No it wasn't.

I mean, you could get { by pressing ALT MODE, but there was absolutely no way to
generate ` | } and ~ on an ASR 33 ... unless you punched one on a piece of paper tape.

You did have the ASCII characters [ \ ] ^ and _.

John Savard

[Quadibloc]

On Thursday, April 30, 2020 at 10:33:51 AM UTC-6, Scott Lurndal wrote:

> It was, however, a character on the ASR-33 keyboard.

Up-arrow is ^ not |.

John Savard

[Bob Eager]

It was used as | in the same way that upper case letters were generated;
by prefixing with a \.

[robin....@gmail.com]

On Friday, May 1, 2020 at 4:32:15 AM UTC+10, John Levine wrote:
> In article <6......@googlegroups.com>,

> <r......@gmail.com> wrote:
> >It was in October 1976 that Jim Rowe designed the interface
> >ASCII-BAUDOT translator.
>
> There have been two way ASCII<->Baudot translators as long as there
> has been ASCII, used to forward telex messages between the older
> Baudot Telex and TWX networks and the newer ASCII TWX network. I
> believe that started in 1959.

But not using a microprocessor, and with the express purpose of using
a Baudot teleprinter as an ASCII computer printer.

[robin....@gmail.com]

There are no [ and ] characters on the keyboard of the ASR 33 (see Wiki).

Can't see the ^ and _ either.

To get those, you'd need the ASR 37 or 38.

[Quadibloc]

On Thursday, April 30, 2020 at 8:28:56 PM UTC-6, robin...@gmail.com wrote:

> There are no [ and ] characters on the keyboard of the ASR 33 (see Wiki).

> Can't see the ^ and _ either.

> To get those, you'd need the ASR 37 or 38.

^ is up-arrow on an ASR 33, and _ is back-arrow on an ASR 33. Same character
codes, just different graphics.

[ is shift-K, \ is shift-L, ] is shift-M, ^ is shift-N, and _ is shift-O.

John Savard

[robin....@gmail.com]

None of those is shown in the keyboard of the ASR-33 on wiki.

[Ahem A Rivet's Shot]

On Thu, 30 Apr 2020 07:13:07 -0700 (PDT)
robin....@gmail.com wrote:

> On Thursday, April 30, 2020 at 11:25:44 PM UTC+10, Peter Flass wrote:
> > https://www.popularmechanics.com/science/a30647778/linux-teletype-programming-hardware-video/
>
> That was done in the 1970s by "Electronics Australia" magazine
> using a microprocessor.

Pretty much anything with a keyboard and a print head got hooked up
to a microprocessor in the 1970s - people hooked up whatever they could lay
their hands on, getting it working was what the hobby was all about.

--
Steve O'Hara-Smith | Directable Mirror Arrays
C:\>WIN | A better way to focus the sun
The computer obeys and wins. | licences available see
You lose and Bill collects. | http://www.sohara.org/

[Quadibloc]

So? There were a number of variations of the keyboard over its lifetime. And, in
fact, some illustrations of the ASR 33 keyboard even on Wikipedia do show those
characters.

I know this, because I was _there_.

As well:

https://www.flickr.com/photos/twylo/9795932916
https://commons.wikimedia.org/wiki/File:Mappa_Teletype_ASR-33.jpg
https://vintagecomputer.net/teletype/asr33/ASR-33_keyboard.JPG

Can't quite make out the keys on this one...

https://www.alternatehistory.com/forum/threads/internet-doesnt-become-mainstream.425625/page-2

John Savard

[Bob Eager]

From tty(4), Sixth Edition:

When desired, all upper-case letters are mapped into the
corresponding lower-case letter. The upper-case letter may
be generated by preceding it by `\'. In addition, the fol-
lowing escape sequences are generated on output and accepted
on input:

for use
` \'
| \!
~ \^
{ \(
} \)

[Peter Flass]

Somhow could they develop C on a 33?

--
Pete

[Bob Eager]

Exactly. See my earlier post. Although I'm sure Bell bought them
something better.

[Radey Shouman]

The magic of C trigraphs:

??( == [
??) == ]

[Peter Flass]

Aha! A light dawns. PL/I(F) also initially had facilities for programming
in a 48-character set, minus the characters ><|(not), etc. Likewise Algol
had problems with the “back arrow”, except on Burroughs systems, which
supported it.

--
Pete

[Bob Eager]

Those hadn't been invented then. They used the \ and [ escape.

[J. Clarke]

On Fri, 01 May 2020 13:27:47 -0400, Radey Shouman

For certain values of "magic". I have an abiding hatred for those
things.

[robin....@gmail.com]

On Friday, May 1, 2020 at 6:58:35 PM UTC+10, Quadibloc wrote:

> On Friday, May 1, 2020 at 1:47:42 AM UTC-6, r.....@gmail.com wrote:
> > On Friday, May 1, 2020 at 5:07:38 PM UTC+10, Quadibloc wrote:

> > > On Thursday, April 30, 2020 at 8:28:56 PM UTC-6, r......@gmail.com wrote:
>
> > > > There are no [ and ] characters on the keyboard of the ASR 33 (see Wiki).
>
> > > > Can't see the ^ and _ either.
>
> > > > To get those, you'd need the ASR 37 or 38.
>
> > > ^ is up-arrow on an ASR 33, and _ is back-arrow on an ASR 33. Same character
> > > codes, just different graphics.
>
> > > [ is shift-K, \ is shift-L, ] is shift-M, ^ is shift-N, and _ is shift-O.
>
> > None of those is shown in the keyboard of the ASR-33 on wiki.
>
> So? There were a number of variations of the keyboard over its lifetime.

It means that ASR-33's did not all have those characters.

There were various print cylinders that could be fitted.

> And, in
> fact, some illustrations of the ASR 33 keyboard even on Wikipedia do show those
> characters.

Not on the one that I looked at.

> I know this, because I was _there_.

You are not the only person who has used an ASR 33.

[robin....@gmail.com]

On Saturday, May 2, 2020 at 4:19:20 AM UTC+10, Peter Flass wrote:
> Radey Shouman <s......@comcast.net> wrote:
> > Peter Flass <p......@yahoo.com> writes:

> >
> >> <......@gmail.com> wrote:
> >>> On Friday, May 1, 2020 at 5:39:23 AM UTC+10, Quadibloc wrote:
> >>>> On Thursday, April 30, 2020 at 10:33:51 AM UTC-6, Scott Lurndal wrote:
> >>>>
> >>>>> Also, the character | which is not in the
> >>>>>> upper-case only subset of ASCII is very useful in that operating system.
> >>>>
> >>>>> It was, however, a character on the ASR-33 keyboard.
> >>>>
> >>>> No it wasn't.
> >>>>
> >>>> I mean, you could get { by pressing ALT MODE, but there was
> >>>> absolutely no way to
> >>>> generate ` | } and ~ on an ASR 33 ... unless you punched one on a
> >>>> piece of paper tape.
> >>>>
> >>>> You did have the ASCII characters [ \ ] ^ and _.
> >>>
> >>> There are no [ and ] characters on the keyboard of the ASR 33 (see Wiki).
> >>>
> >>> Can't see the ^ and _ either.
> >>>
> >>> To get those, you'd need the ASR 37 or 38.
> >>>
> >>
> >> Somhow could they develop C on a 33?
> >
> > The magic of C trigraphs:
> >
> > ??( == [
> > ??) == ]
> >
>
> Aha! A light dawns. PL/I(F) also initially had facilities for programming
> in a 48-character set,

That character set used substitutions for some characters such as
semicolon, and keywords for others such as GT, LT, CAT, etc.

It enabled source produced by a Baudot machine (5-channel tape)
to be compiled. (as well as, of course, from an older BCD card key-punch.

> minus the characters ><|(not), etc. Likewise Algol
> had problems with the “back arrow”, except on Burroughs systems, which
> supported it.

When Algol source was prepared with punch card equipment, substitutions
were used for various characters including the single character "10"
and logical operators.

When prepared on such paper tape equipment as a Friden flexowriter,
substitutions were not always necessary.

[Quadibloc]

On Friday, May 1, 2020 at 8:39:14 PM UTC-6, robin...@gmail.com wrote:

> You are not the only person who has used an ASR 33.

I don't claim to be. You are the one who is in effect making such a claim by
categorically asserting that certain characters aren't found on the ASR 33.

John Savard

[robin....@gmail.com]

On Saturday, May 2, 2020 at 4:57:22 PM UTC+10, Quadibloc wrote:

> On Friday, May 1, 2020 at 8:39:14 PM UTC-6, r.......@gmail.com wrote:

> >>I know this, because I was _there_.

> > You are not the only person who has used an ASR 33.

> I don't claim to be. You are the one who is in effect making such a claim

I'm not making any such claim. Your statement
"I know this because I was _there_"
implies that you are the only person who has used and ASR-33
and that nobody else knows what they are talking about.

> by
> categorically asserting that certain characters aren't found on the ASR 33.

Its true. There's a photo of the keyboard of an ASR-33 in wiki.

I have already stated that 3 posts ago.
You have had ample opportunity to look at it and see for yourself.

[Peter Flass]

“10” was a single character? What genius thought of that?

--
Pete

[robin....@gmail.com]

On Sunday, May 3, 2020 at 2:24:15 AM UTC+10, Peter Flass wrote:

It was intended to make a float number look more like scientific
notation. Thus, 1.23456'10'5
where the '10' represents a single character, small,
and printed below the line, so that it looks like you
have written "10 to the power of".
We're used to writing the letter 'E' for that.

The '10' is part of the Algol language and was designed by the Algol
Committee.

[Quadibloc]

On Saturday, May 2, 2020 at 7:06:57 AM UTC-6, robin...@gmail.com wrote:
> On Saturday, May 2, 2020 at 4:57:22 PM UTC+10, Quadibloc wrote:

> > by
> > categorically asserting that certain characters aren't found on the ASR 33.

> Its true. There's a photo of the keyboard of an ASR-33 in wiki.

There are a *pile* of photos of keyboards of ASR 33 Teletypes all over the
Internet. I posted the URLs of about four of them a few posts back.

I don't _need_ a photo to tell me what the keyboard of an ASR 33 looks like.

The fact that you have seen one picture of the keyboard of an ASR 33 on
Wikipedia which doesn't have these symbols on its keyboard does not mean that
this was typical or usual of them.

Indeed, the picture here

https://en.wikipedia.org/wiki/File:TTY33ASR.jpg

shows one where [, \, and ] aren't printed on the keys. However, they would
still have been on the print head for shift-K, shift-L, and shift-M
respectively.

John Savard

[robin....@gmail.com]

On Sunday, May 3, 2020 at 8:59:29 AM UTC+10, Quadibloc wrote:
> On Saturday, May 2, 2020 at 7:06:57 AM UTC-6, robin...@gmail.com wrote:
> > On Saturday, May 2, 2020 at 4:57:22 PM UTC+10, Quadibloc wrote:
>
> > > by
> > > categorically asserting that certain characters aren't found on the ASR 33.
>
> > Its true. There's a photo of the keyboard of an ASR-33 in wiki.
>
> There are a *pile* of photos of keyboards of ASR 33 Teletypes all over the
> Internet. I posted the URLs of about four of them a few posts back.
>
> I don't _need_ a photo to tell me what the keyboard of an ASR 33 looks like.

But you do, because there were variations.

> The fact that you have seen one picture of the keyboard of an ASR 33 on
> Wikipedia which doesn't have these symbols on its keyboard does not mean that
> this was typical or usual of them.

There were 69 different print wheels for the ASR-33.
And all with different keyboards.

> Indeed, the picture here
>
> https://en.wikipedia.org/wiki/File:TTY33ASR.jpg
>
> shows one where [, \, and ] aren't printed on the keys.

That's what I posted earlier.
You chose to ignore it.

> However, they would
> still have been on the print head for shift-K, shift-L, and shift-M
> respectively.

Rubbish.
Had you bothered to read the WIKI description to which I referred you,
you would have read that there were 69 DIFFERENT print wheels,
and most of them omitted those characters from the shift-KLM positions.

[Quadibloc]

I did see the reference to 69 different print elements. However, the particular
keyboard in the photo is not of a machine that omitted those characters on its
print element.

John Savard

[robin....@gmail.com]

On Monday, May 4, 2020 at 5:03:07 AM UTC+10, Quadibloc wrote:

> On Sunday, May 3, 2020 at 2:01:52 AM UTC-6, r.......@gmail.com wrote:
> > On Sunday, May 3, 2020 at 8:59:29 AM UTC+10, Quadibloc wrote:
>
> > > However, they would
> > > still have been on the print head for shift-K, shift-L, and shift-M
> > > respectively.
>
> > Rubbish.
> > Had you bothered to read the WIKI description to which I referred you,
> > you would have read that there were 69 DIFFERENT print wheels,
> > and most of them omitted those characters from the shift-KLM positions.
>
> I did see the reference to 69 different print elements.

Good.

> However, the particular
> keyboard in the photo is not of a machine that omitted those characters on its
> print element.

You really don't know what you are talking about.
You obviously don't have the manual.

Of those 69 different type wheels, only a handful of them
had those particular characters [\] on the KLM keys.

The reason for the blank keytops on the KLM keys in the
photo of an actual ASR-33 was because alternative characters
could be provided on most of the 69 type wheels.

And in any case, a type wheel was easily changed for another.

[Quadibloc]

On Monday, May 4, 2020 at 11:43:17 PM UTC-6, robin...@gmail.com wrote:

> And in any case, a type wheel was easily changed for another.

Fairly easily, perhaps, but in practice it was very seldom done. An ASR 33 is not
a Selectric typewriter.

John Savard

[Ahem A Rivet's Shot]

One thought springs to mind, and I could well be wrong if so
please correct me, but isn't it the case that no matter what type wheel is
installed, or what is printed on the key caps the key combinations referred
to up-thread will always put the right ASCII characters on the tape (ASR-33)
and/or serial line (KSR-33 or ASR-33).

[robin....@gmail.com]

Certainly, it's not a Selectric, but type wheels can be changed
for special characters, and because a type wheel can be damaged.
Substitutions can be done rapidly for a temporary change.

Recall that the wheel is hit by a rubber hammer. The rubber
part is fitted over a steel holder. Though the
hammer is made of tough rubber, it can wear completely through.
Once the rubber hammer is worn through, the metal part comes
into direct contact with the aluminium wheel, and characters on the
type wheel rapidly become mutilated.

[Quadibloc]

On Tuesday, May 5, 2020 at 2:30:02 AM UTC-6, Ahem A Rivet's Shot wrote:

> One thought springs to mind, and I could well be wrong if so
> please correct me, but isn't it the case that no matter what type wheel is
> installed, or what is printed on the key caps the key combinations referred
> to up-thread will always put the right ASCII characters on the tape (ASR-33)
> and/or serial line (KSR-33 or ASR-33).

That's true enough. You would have to change the code bars in the keyboard
asembly to change what sequence of bits was sent out for a particular key
combination.

But I don't think that was the issue under discussion.

John Savard

[David Lesher]

How did an ASR-33 get to be "1930's"?
Try the Model 15.


--
A host is a host from coast to coast.................wb8foz@nrk.com
& no one will talk to a host that's close..........................
Unless the host (that isn't close).........................pob 1433
is busy, hung or dead....................................20915-1433

[Quadibloc]

On Tuesday, May 5, 2020 at 7:34:47 PM UTC-6, David Lesher wrote:
> How did an ASR-33 get to be "1930's"?
> Try the Model 15.

It is the use of a 5-level Teletype as a Linux terminal that started this thread.
But then there was a discussion of how even an upper-case only ASCII Teletype
would have some limitations when used with Unix.

John Savard

[Charlie Gibbs]

On 2020-05-06, David Lesher <wb8...@panix.com> wrote:

> How did an ASR-33 get to be "1930's"?
> Try the Model 15.

That was my first impression. But if you look closely
at the pictures you'll see that it's not a Teletype[tm]
machine but some sort of German equivalent.

A PPOE had two machines: an ASR33 for TWX and a Siemens
machine for Telex which was basically Baudot but with a
couple of different characters, like the "iron cross".
Somewhere I still have a 5-level paper tape I punched
on it - it prints an image of the Madonna and child.

--
/~\ Charlie Gibbs | Microsoft is a dictatorship.
\ / <cgi...@kltpzyxm.invalid> | Apple is a cult.
X I'm really at ac.dekanfrus | Linux is anarchy.
/ \ if you read it the right way. | Pick your poison.

[Ahem A Rivet's Shot]

It seemed to have started about whether an ASR-33 was usable as a
terminal to a modern unix system. Given the above it will work but you might
see blanks on the paper.

[Quadibloc]

On Tuesday, May 5, 2020 at 9:42:55 PM UTC-6, Charlie Gibbs wrote:
> On 2020-05-06, David Lesher <wb8...@panix.com> wrote:

> > How did an ASR-33 get to be "1930's"?
> > Try the Model 15.

> That was my first impression. But if you look closely
> at the pictures you'll see that it's not a Teletype[tm]
> machine but some sort of German equivalent.

In the video, the Teletype trademark is visible clearly on the machine, and it
has the form of a Model 15. It also has a label on it referring to it as
"Printer TG-7-B", which suggests that it is a military model. And indeed it is,
it's described in TM 11-352, for example.

John Savard

[Andy Burns]

Charlie Gibbs wrote:

> a Siemens machine for Telex which was basically Baudot but with a
> couple of different characters, like the "iron cross".

That's the "WRU?" character.

[robin....@gmail.com]

On Wednesday, May 6, 2020 at 1:42:55 PM UTC+10, Charlie Gibbs wrote:

> On 2020-05-06, David Lesher <w.....@panix.com> wrote:
>
> > How did an ASR-33 get to be "1930's"?
> > Try the Model 15.
>
> That was my first impression. But if you look closely
> at the pictures you'll see that it's not a Teletype[tm]
> machine but some sort of German equivalent.
>
> A PPOE had two machines: an ASR33 for TWX and a Siemens
> machine for Telex which was basically Baudot but with a
> couple of different characters, like the "iron cross".

That possibly was a Siemens M100 machine.
Very reliable.

[fred_...@hotmail.com]

You guys are funny. The daisy wheel was invented in 1970. The Model 33 used a rotating cylinder, and was introduced in 1963. I do not know how easy it
is to replace the cylinder, because I have never needed to do so.

FredW

[Peter Flass]

I don’t recall that either. The cylinder was high-grade steel, daisy wheels
were made of some cheap, thin substance, maybe aluminum (as mentioned) or
plastic. This is the crossover point between sturdy and well- made, like
Teletypes and 2741s, and cheap and disposable.

--
Pete

[Charlie Gibbs]

On 2020-05-16, Peter Flass <peter...@yahoo.com> wrote:

> <fred_...@hotmail.com> wrote:
>
>> On Tuesday, May 5, 2020 at 5:26:55 AM UTC-4, robin...@gmail.com wrote:
>>
>>> On Tuesday, May 5, 2020 at 5:11:35 PM UTC+10, Quadibloc wrote:
>>>
>>>> On Monday, May 4, 2020 at 11:43:17 PM UTC-6, robin...@gmail.com wrote:
>>>>
>>>>> And in any case, a type wheel was easily changed for another.
>>>>
>>>> Fairly easily, perhaps, but in practice it was very seldom done.
>>>> An ASR 33 is not a Selectric typewriter.

<nit>
Nor, for that matter, is a KSR33 or a 33RO.
</nit>

>>> Certainly, it's not a Selectric, but type wheels can be changed
>>> for special characters, and because a type wheel can be damaged.
>>> Substitutions can be done rapidly for a temporary change.
>>>
>>> Recall that the wheel is hit by a rubber hammer. The rubber
>>> part is fitted over a steel holder. Though the
>>> hammer is made of tough rubber, it can wear completely through.
>>> Once the rubber hammer is worn through, the metal part comes
>>> into direct contact with the aluminium wheel, and characters on the
>>> type wheel rapidly become mutilated.
>>
>> You guys are funny. The daisy wheel was invented in 1970. The Model 33
>> used a rotating cylinder, and was introduced in 1963. I do not know how
>> easy it is to replace the cylinder, because I have never needed to do so.

On the other hand, the Model 35's typebox is very easy to replace. :-)

> I don’t recall that either. The cylinder was high-grade steel, daisy wheels
> were made of some cheap, thin substance, maybe aluminum (as mentioned) or
> plastic. This is the crossover point between sturdy and well- made, like
> Teletypes and 2741s, and cheap and disposable.

Most daisy wheels I saw were plastic, although the period tended to
be metal (plated?) to withstand the pounding it would take.

[fred_...@hotmail.com]

On Thursday, April 30, 2020 at 9:25:44 AM UTC-4, Peter Flass wrote:
> https://www.popularmechanics.com/science/a30647778/linux-teletype-programming-hardware-video/
>
>
> --
> Pete

Sorry, completely off-topic

Since I repaired daisy wheel mechanisms "back in the day" -- the operation is:

A metal hammer that moves freely inside an electro-magnet. When the wheel is
about to expose the desired glyph, the magnet is fired, causing the hammer
to drive to the wheel. The hammer hits the wheel, and bounces off. It returns
back. IF IT STICKS, THE WHEEL GETS MANGLED. The spoke then impacts the ribbon, and the paper. The spoke then bounces back, and the wheel is moved to the next
desired glyph. You don't want to operate this mechanism without the wheel.
You don't want the hammer to stick. You need the glyph to move -- many times.
It would be useful to "metallize" the wheel -- probably more important at
the center! of the spokes. Two elastic collisions per impact.

Yes, the character would degrade. But the wheel typically lost spokes sooner
than the type became muddy.

All of this because the print speed of 10cps was too slow. Daisy wheels achieved 30+ cps.

FredW

[robin....@gmail.com]

On Sunday, May 17, 2020 at 12:40:00 AM UTC+10, Peter Flass wrote:
> <f......@hotmail.com> wrote:

> > On Tuesday, May 5, 2020 at 5:26:55 AM UTC-4, r......@gmail.com wrote:
> >> On Tuesday, May 5, 2020 at 5:11:35 PM UTC+10, Quadibloc wrote:

> >>> On Monday, May 4, 2020 at 11:43:17 PM UTC-6, r.......@gmail.com wrote:
> >>>
> >>>> And in any case, a type wheel was easily changed for another.
> >>>
> >>> Fairly easily, perhaps, but in practice it was very seldom done. An ASR 33 is not
> >>> a Selectric typewriter.
> >>
> >> Certainly, it's not a Selectric, but type wheels can be changed
> >> for special characters, and because a type wheel can be damaged.
> >> Substitutions can be done rapidly for a temporary change.
> >>
> >> Recall that the wheel is hit by a rubber hammer. The rubber
> >> part is fitted over a steel holder. Though the
> >> hammer is made of tough rubber, it can wear completely through.
> >> Once the rubber hammer is worn through, the metal part comes
> >> into direct contact with the aluminium wheel, and characters on the
> >> type wheel rapidly become mutilated.
> >
> >
> > You guys are funny. The daisy wheel was invented in 1970.

A similar principle was used on very early teleprinters, c. 1920s.

> The Model 33
> > used a rotating cylinder, and was introduced in 1963. I do not know how easy it
> > is to replace the cylinder, because I have never needed to do so.

A small spanner and a tool made from piece of fencing wire
to hold the wheel steady while you changed it.

> I don’t recall that either. The cylinder was high-grade steel,

Light-weight aluminium.

> daisy wheels
> were made of some cheap, thin substance, maybe aluminum (as mentioned) or
> plastic.

I think that the Diablo wheel was plastic.

[Peter Flass]

The 2741 was 15, so half.

--
Pete

[robin....@gmail.com]

On Sunday, May 17, 2020 at 2:49:52 AM UTC+10, f..._...@hotmail.com wrote:

> Since I repaired daisy wheel mechanisms "back in the day" -- the operation is:
>
> A metal hammer that moves freely inside an electro-magnet. When the wheel is
> about to expose the desired glyph, the magnet is fired, causing the hammer
> to drive to the wheel. The hammer hits the wheel, and bounces off. It returns
> back. IF IT STICKS, THE WHEEL GETS MANGLED. The spoke then impacts the ribbon,
> and the paper. The spoke then bounces back, and the wheel is moved to the next
> desired glyph. You don't want to operate this mechanism without the wheel.
> You don't want the hammer to stick. You need the glyph to move -- many times.
> It would be useful to "metallize" the wheel -- probably more important at
> the center! of the spokes. Two elastic collisions per impact.

Weren't the characters metallized?

> Yes, the character would degrade. But the wheel typically lost spokes sooner
> than the type became muddy.
>
> All of this because the print speed of 10cps was too slow. Daisy wheels
> achieved 30+ cps.

Nominally, 30 cps, but with a "catch-up" speed more than 30 cps
when the carrier returned

[Quadibloc]

On Saturday, May 16, 2020 at 8:40:00 AM UTC-6, Peter Flass wrote:

> I don’t recall that either. The cylinder was high-grade steel, daisy wheels
> were made of some cheap, thin substance, maybe aluminum (as mentioned) or
> plastic. This is the crossover point between sturdy and well- made, like
> Teletypes and 2741s, and cheap and disposable.

The IBM Selectric typewriter was introduced in 1961.

Selectric elements were made out of plastic, with a metal coating; some
daisywheel printers could also use daisywheels with a metal layer on the
printing side for higher quality printing.

John Savard

[Quadibloc]

On Saturday, May 16, 2020 at 9:14:51 PM UTC-6, robin...@gmail.com wrote:
> On Sunday, May 17, 2020 at 2:49:52 AM UTC+10, f..._...@hotmail.com wrote:

> > It would be useful to "metallize" the wheel -- probably more important at
> > the center! of the spokes. Two elastic collisions per impact.
>
> Weren't the characters metallized?

They _could_ be, and in some cases were, but in general they were not. Thus, the
Diablo 630 and a few other professional daisy-wheel printers could use, as an
alternative to all-plastic daisywheels, ones that were metal at least on the
printing side.

John Savard

[robin....@gmail.com]

It was mainly for longevity of the type faces that a metal
coating was used.

[robin....@gmail.com]

On Monday, May 18, 2020 at 2:48:53 AM UTC+10, Quadibloc wrote:

> On Saturday, May 16, 2020 at 8:40:00 AM UTC-6, Peter Flass wrote:
>
> > I don’t recall that either. The cylinder was high-grade steel, daisy wheels
> > were made of some cheap, thin substance, maybe aluminum (as mentioned) or
> > plastic. This is the crossover point between sturdy and well- made, like
> > Teletypes and 2741s, and cheap and disposable.
>
> The IBM Selectric typewriter was introduced in 1961.
>
> Selectric elements were made out of plastic, with a metal coating;

The weakness of the plastic thimbles was the toothed part;
the teeth regularly broke off.
IBM usually replaced those free of charge.