Check out these fine folks using a "Selector Light Pen" with the IBM 3270 Data Display System.

#retrocomputing #lightpen

In this case, it wasn't a gimmick. The late Chuck Peddle, a staple of the Vintage Computer Federation forums, had this to say about them:

Selecta pens were an extremely popular accessory on 3270 terminals and worked very well - that is to say flawlessly - and so they were certainly not a "dead end technology".

IBM made the Datamaster, the system from which the IBM PC derives much of its heritage, also compatible with a light pen.

It seems a bit odd then, that we go from a hugely successful product line like the 3270, to basically a half-baked nod to light-pen support in the PC, with IBM not even bothering to sell a pen accessory of their own.

It's not like they didn't know how to make them. They were still selling the damn things!

You could even buy a specially kitted out IBM PC that operated as a 3270 terminal.

The IBM 5151 was initially the only monitor that you could purchase for the PC - it took nearly two years for IBM to get the color 5153 out the door (your other option was a third-party composite monitor in the meantime).

The 5151 had a refresh rate of 50Hz. To avoid eyestrain, it used a high persistence P39 phosphor. This wreaked havoc with light pens circuits.

So much so, that IBM actually removed the light pen header from subsequent versions of the MDA card. It's not just unpopulated - it's gone. No light pen for you, mister!

This had already been a solved problem for nearly a decade or more. Light pen systems often used dual phosphor coatings - one phosphor would be a long persistence phosphor in the visible spectrum, while a much faster phosphor responded in the infrared range, which a pen could pick up.

Best of both worlds! Except, you know, this is incompatible with trying to sell stuff dirt cheap.

The fast responding phosphor wasn't always infrared. One historically famous display was the DEC Type 30 - you may know it as the display used for what many consider to be the game that started the video game industry - Space War.

It had a fast blue phosphor and a yellow very high persistence phosphor, giving the display a very unique two-color appearance, as your blue ship left behind it a hazy yellow trail.

Here's an excellent article on the significance of Space War, with some visualizations of what the game looked like.

https://www.masswerk.at/nowgobang/2021/spacewar1

No light pens were involved, sadly.

(Now Go Bang!) Spacewar 1 and the Beginnings of Video Game Aesthetics

Tales of the early days of the first digital video game.

Now Go Bang! β€” mass:werk / Blog

Here's the Digigraphic Display System 270 from Control Data Corporation, circa 1965.

This was another dual-phosphor system using the same "P7" phosphor as the DEC Type 30.

It could connect to your CDC 3200 (you have one of those, right?)

If you look closely, you can see the light pen laying on the table.

I want you to consider that this is 1965, and this system has an addressable resolution of 4096x4096. Yes, it's vector so its not exactly comparable to a modern 4K display, but that's some hella resolution.

One thing I need to learn more about is how light pens worked with vector displays. Because all of the timing logic I'm familiar with requires a raster scan with regular sync signals. This ain't it.

#retrocomputing

I suppose with a vector display it's actually a lot easier - the system knows where it is currently plotting, so as soon as the photodiode or phototransistor goes boop, we just process an interrupt with the current (x,y) position. Easy peasy, right?
Good thing the system has three separate light pen interrupts.

I will be at VCF East in a few short weeks.

I will be bringing a light pen, and you may be able to play with it at the Seequa Chameleon exhibit, assuming I can fashion an adapter and we can find any light pen software that the Chameleon will run.

If you're attending, stop by and say hi!

@gloriouscow What were the ergonomics of a good light pen system? I assume that software that was specifically designed for it worked well technically, but did it have the same issue of arms getting tired after some use, similar to laptops with touch screen?

@tom_verbeure I don't think there are any really great ergonomic solutions. Using a recessed monitor might help, to allow use of the system without holding your arm up. But then you are forcing the user to look down, which has other ergonomic concerns. I have arthritis in my neck now (yay for getting old) and even drawing on paper while looking down at my desk becomes uncomfortable after while.

There were some applications that were uniquely suited to a pen, though. FTG documents describe a use case of using a pen to control a CRT from behind a transparent protective shield up to 14" away. This was undoubtedly an industrial application - if you have a need for a blast shield like that you probably also don't have a mousepad lying around. Interesting scenario.

The pen was used in medical settings because it could operate via sterile sleeves and the metal versions FTG made could be easily sanitized.

Considering the nature of a light pen requires a CRT and a CRT will never be something you can comfortably hold at an angle of your choosing, I think we just lost the ergonomics battle from the start.

@gloriouscow Thanks for the dual phosphor explanation, BTW. I built a light pen for my C64, and it performed horribly, as one could expect. But I always wondered how they were able to make it work so well on professional systems.

@tom_verbeure It's funny a few people have mentioned that they built pens and they worked fine, but most people just made some jittery mess.

It's probably a crapshoot - if you got lucky with the right resistor values and the right display and the right EMF environment, maybe you could make a homemade pen serviceable - I think the complicated guts of commercial pens make them reliable.

There must be something else going on here, this is the light pen for the Vectrex game console - you can see the user is able to interact with dark areas of the screen, somehow.

Curious!

https://www.youtube.com/watch?v=t9LLEJ63_Rw

VECTREX Accessories - Light Pen and 3D Imager

YouTube
It's a bit hard to spot and she doesn't really call attention to it, but it does look like there is a little crosshair or cursor she is dragging around rather than being able to land the pen in arbitrary dark areas of the screen. It's a clever system.
@gloriouscow IIRC, the vectrex way is to draw a spiral around the last known location of the pen. If the pen is really lost, it'll keep expanding the spiral search until it fills the whole screen.

@gloriouscow

Vector displays! Oh this is amazing! I saw this recently about the Hewlett-Packard HP 1345A vector displays used in WarGames, quite interesting stuff!

https://www.youtube.com/watch?v=JrwvIKK3D2o

WarGames Screens: I Bought Them! Let's get them working and drawing!

YouTube
@gloriouscow Another...
@davefischer that's the sexiest thing anyone's ever sent me on the internet before

@gloriouscow

One 1972 Rolling Stone issue had a big Space War article:

https://archive.org/details/19721207rollingstoneexcerptspacewararticlev02

1972-12-07 Rolling Stone ( Excerpt) Spacewar Article 600DPI : Free Download, Borrow, and Streaming : Internet Archive

An excerpt from the December 7th, 1972 edition of Rolling Stone Magazine titled Spacewar, Fantastic Life and Death Among Computer Bums by Stewart...

Internet Archive

@gloriouscow

I purchased a copy of that issue, based on the back cover (or "page 1") portrait:

@elithebearded @gloriouscow
If I remember correctly, that’s the story in which my friend Ralph Gorin won the tournament, right?

@alderson @gloriouscow

Ralph Gorin is mentioned in the article, but not as a winner. Bruce Baumgart won Free-For-All (see photo) and Tovar (I can't find a full name) and Robert E. Maas (REM) won team.

@gloriouscow That phosphor combination is known as P7. The DEC Type 30 display and its relatives, e.g. Type 340, used a 16ADP7 CRT.
[Sorry, I see that you're mentioned it as P7 earlier.]
@gloriouscow
The DEC Type 370 light pen, used with the Type 30 and Type 340 displays, had a fiber optic cable from the pen to a photomultiplier tube. There weren't yet good enough photodiodes for a solid-state light pen.
@brouhaha good point - it's easy to forget that this stuff is that old

@gloriouscow

That was probably because they used a standard radar-CRT ?

By in the movies those are always green

@gloriouscow

I think the rapid-blue/slow-yellow was mostly for air, because it enabled you to estimate speed and direction visually.

Sea vessels are too slow for that.

Many sea-radars had a plastic overlay sheet you could draw on with grease-pencil, to estimate speed+direction, in that slower environment.

@phloggen both things can be true here.
@gloriouscow lol i have a 5151 and tried and failed to construct a light pen...

@lkundrak It is possible to build your own, but what they do not tell you is that anything you can cobble together with a photodiode, a resistor, and some bubblegum is going to be absolutely terrible, giving you a very negative impression of a technology that frankly needs all the help it can get.

Your average light pen is a rather complicated little thing. It has a waveguide or lens assembly to focus and direct the incoming light, amplification and high-pass circuitry and other filtering.

Later pens that supported higher (or multiple) frequencies had timing circuits to produce a strobe pulse of a minimum specified length.

@gloriouscow Interesting... I was under the impression that IBM created the 5153 _precisely_ because nobody before 1983 was creating a monitor compatible with the connector. So everyone who wanted color had no choice but to use the composite output before then.

But I never looked at old catalogues to confirm/deny this. I just believed it when I read it :P.

@cr1901 sorry, clarified that a bit.
@cr1901 I don't think a whole ton of people were importing Sony KX-1901's from Japan

@cr1901 What I have personally learned is that composite displays aren't nearly as bad as I assumed, if you're okay with monochrome. My 1084D looks pretty brilliant in composite, with crisp white 80 column text. Quite readable.

This is still something I need to fix in my own emulator - I do not automatically disable the color burst in 80 column mode like the CGA does. it gives you the impression that composite is always an unreadable mess.

@gloriouscow But composite 80 column mode _is_ an unreadable mess on contemporary hardware :D!
@cr1901 I'm sure there were a parade of terrible monitors, and if you were hooked up to a television, God help you. But I also have an old composite monochrome monitor that lived its first life as a Delta Airlines terminal - it's pretty crisp as well. although I don't exactly have a firm date for it

@cr1901 I'm fairly sure it's using a P39 phosphor. In any case, I'm going to take oscilloscope measurements of (attempting) to use it with the light pen.

It should be informative to see just why it caused pens so much grief!

@gloriouscow > and if you were hooked up to a television, God help you.

Yea, this is exactly what I did when I got my IBM 5150 in 2011, w/ a 1999 CRT TV. It was awful :).

@cr1901 @gloriouscow
My boss at Apparat bought a fully equipped PC on Wednesday, 1981-08-12, the day of introduction. He had pre-ordered it, which theoretically was impossible, since it didn't exist before that date. I went with him to Computerland to pick it up.
I had long since forgotten that IBM didn't yet offer a CGA monitor. I think we wired up a cable to an RGB monitor we had for something else.
1/
@cr1901 @gloriouscow
The monitor might have been an NEC or Taxan. Not being designed for CGA, it had analog RGB input, so we only got eight colors, and they didn't look right. I wanted to wire up an adapter board with resistors and NPN transistors to produce the 16 colors, but we didn't do that, so all my early PC experience was with the MDA and IBM 5151 monitor. That was definitely the best monochrome text display I'd used by that time.
2/
@cr1901 @gloriouscow
I wanted to write a better OS for it, and Apparat was known for their NewDOS/80 operating system for the TRS-80 models 1 and III. The boss wanted the author of NewDOS/80 to write a PC version, but he apparently wasn't interested. What I had in mind was a small Unix clone.
Somehow, the boss got a very early test copy of QNX, and said there was no point in my writing something like that.
He was absolutely correct, but not for the reason he gave.
3/
@gloriouscow I still have one of those in a drawer somewhere.
@simon_brooke I would be interested in acquiring it from you, if you'd be willing to part with it.
@gloriouscow Sorry, I should clarify β€” it's not an IBM one, it's for a BBC Micro. It worked very well back in the 1980s, but I couldn't say whether it would still work.
@simon_brooke Oh, ha. If I started collecting light pens for 8-bit micros I'll never see the end of it. But I'd very much like an IBM pen - it's just sort of directly in the PC's ancestry.
@gloriouscow I don't have the hair to use the Selector Light Pen for IBM 3270. Bummer.

@gloriouscow *Years* ago, I made a lightpen with a phototransistor and some basic code from a magazine. It worked well on my Apple ][ clone.

I cannot recall what magazine, though.

@AG7GM Magazines of the time were lousy with plans for light pens. I think I've collected six different versions.
@gloriouscow Whew! Thanks for that. Now I *know* that if someone wants to know which rag it was in that *you'll* probably know! πŸ˜‚ 🀣
@AG7GM I know that you should just buy one
@gloriouscow IIRC BBC micro had a light pen input that worked!
@revk that implies someone actually tested it at one point
@gloriouscow The only thing I did not ever do on a BBC micro was econet - so yeh πŸ™‚