1. John Smith 19 Gold badge
    Happy

    A steam punk VDU ?

    If you know Charles Stross's "Laundry" series you'll know that the narrators boss uses a thing called a Memex to store and query information held on micrfiche cards.

    It's known that Charles Babbage planned to link a printer directly to his Analytical Engine to eliminate typos when producing navigation tables. On that basis I think a "steam punk" teletype would not be that difficult.

    But a VDU? Something that displays (and edits) characters on a screen without electricity?

    I'll accept any technology, mechanical, photographic, chemical, biological, pneumatic, etc.

    To keep it simple(ish) it only needs to support 1 font in both upper and lower case. No graphics, no color but it should allow last line editing.

    I think the keyboard should be fairly easy, it's the display that's tricky. Creating a reusable surface that's erasable and re-writable without semiconductor storage.

    I'll be interested to see what comes up.

    1. TDog

      Re: A steam punk VDU ?

      You don't just need to edit the characters - you need to feed back the changes to the differential engine. It is for that reason that I am proposing Generic Objects Lytic Facades. These would be a series of small balls with characters and numbers exposed on flattened surfaces which could ingeniously rotate to expose a single object (i.e. letter or number).

      By placing a set of similar objects in a row (hence generic) words and or numbers could be displayed. By manually rotating an ball an "edit" could occur and be read. This is similar to dissolving or destroying the first information set and replacing it with a new one - hence lytic (from lysis).

      Because these objects would be a visible interface between the differential engine and the user, they would act as a sort of screen between the engine and the user, or a façade.

      I rather suspect that one of the large industrial conglomerates (probably Whitworth with their excellent approach to precision) will be contacting me shortly with a view to licensing my GOLF ball technological display.

      Simon

    2. Chozo

      Re: A steam punk VDU ?

      Seems do-able...

      Typewriter to punch card data entry, fluidic logic to sort & read data back to a flip board display. If you wanted bleeding edge technology you could add add a magic lantern with a carousel and those new fangled wax cylinders young Edison is playing with.

      PS: I'm having a spot of bother with the Ballistic Express pigeon launcher when sending messages and cannot dial down the cannon pressure, frightful mess on the house across the street please send a man over soonest.

  2. Anonymous Coward
    Anonymous Coward

    Well, there go all my spare thoughts for the next day. Thanks a lot, asshole!

    Are we assuming current levels of manufacturing and materials tech, just applied only mechanically, or are we supposed to limit to levels of practicality that you have without CAD, etc? IIRC one of the things that scuppered Babbage was the inability to manufacture quantities of the internals he needed to sufficient tolerances, so it probably makes a big difference.

    1. John Smith 19 Gold badge
      Happy

      Well, there go all my spare thoughts for the next day. Thanks a lot, asshole!

      If it was easy someone would have written it up on a web site by now for the lulz. :)

      Actually as the London Science Museum proved some years ago 19th century tech was up to the precision task, but Babbage suffered from a bad dose of paralysis-by-analysis. He kept fiddling when he should have frozen the design, put any new ideas into his V 2.0 file and got on with it. The tech level does make a big difference. As does the materials science. Photoetching allows mass production of fairly high resolutions micromechanisms, but requires better emulsion tech than the 19th century could manage (because photography was in its infancy) but I'm not sure how much better (IE improveable by a lone chemist).

      "Super finishing" tech from the car industry of the late 1930's could give a surface finish of 100nm.

      1. the spectacularly refined chap Silver badge

        The 1930's were a full hundred years later than Babbage's work. Things had moved on considerably during that time. In no small part to Joseph Whitworth, who it is fair to say is one of the pioneers (if not the pioneer) of true precision engineering.

        He was also one of the team who were actually building Baggage's engines early in his career. Much has been written as to how this work influenced his later career because he realised Babbage's machine couldn't be made with the processes of the day.

        Don't forget that Babbage's design worked when the Science Museum built it using modern components and processes. There was nothing wrong with the design, it was simply too difficult to construct usign the technology of the day. This was an era when if you brought together e.g. two gear wheels you would then need to manually fettle both until they worked together smoothly. Replace one on them and you would have to repeat the process: the accuracy simply wasn't there. It was only with greater precision (using a tooth-cutting process invented by Whitworth) than such components became standard commodity items that could be relied upon to work together straight from the parts box.

        1. Roger Greenwood

          My toolbox . . .

          . . .still contains Whitworth spanners. A very underated chap.

  3. Anonymous Coward
    Anonymous Coward

    One other thing - are we assuming access to some type of power source? EG, a constant supply of compressed air, or a rotating shaft from somewhere?

    I think I have a concept that might work fairly well. Might even be able to scroll. Can't make any guarantees about refresh rate though... :)

    1. John Smith 19 Gold badge

      "One other thing - are we assuming access to some type of power source? EG, a constant supply of compressed air, or a rotating shaft from somewhere?"

      Either is reasonable.

  4. Jim Lewis

    There are several existing solutions to this.

    1. Etch a sketch, pointer is moved on surface on which a powder adheres until displaced.

    surface is returned to usable form by shaking in original toy, but could be blown perhaps?

    http://www.ebay.co.uk/sch/i.html?_trksid=p2053587.m570.l1313.TR0.TRC0.Xetch+a+sketch&_nkw=etch+a+sketch&_sacat=0&_from=R40

    2. Magnetic writing board. Probably easier to use this, see examples here:

    http://www.ebay.co.uk/bhp/magnetic-writing-board

    3. Or the simplest working principle, where a piece of grey translucent plastic sheet is impressed upon with a backing of a wax coated sheet. Where contact is made a dark colour is seen, Lifting the sheet 'erases' it. Also known as a magic slate.

    http://www.ebay.com/bhp/magic-slate

    They could all be modified in some way to provide the desired functionality, (perhaps not trivial to do so however!)

    1. John Smith 19 Gold badge

      "1. Etch a sketch, pointer is moved on surface on which a powder adheres until displaced.

      surface is returned to usable form by shaking in original toy, but could be blown perhaps?"

      I'd forgotten all about etch-a-sketch and I did think it's viable.

      This would also be the "vector graphics" version of it. :)

      2. Magnetic writing board. Probably easier to use this, see examples here:

      3. Or the simplest working principle, where a piece of grey translucent plastic sheet is impressed upon with a backing of a wax coated sheet.

      I think you've cracked a key issue, which is the erasable memory aspect of the display.

      The down side is likely to be partial erasure.

  5. Anonymous Coward
    Anonymous Coward

    Two more contributions here

    From HippyFreetard

    From TheOtherHobbes

  6. Steve Graham

    Didn't Gibson and Sterling already have this in 'The Difference Engine'?

    It's been a very long time since I read it, but I seem to remember there was a cinema-sized display composed of a large array of pixels, each of which could show a differently-coloured face under program control.

    1. Don Jefe

      The tech in The Difference engine was similar, but closer to a slide projector, as the 'data' was predefined on sets of cards. There wasn't really a way for that system to deal with dynamic data or multiple data sources as the thing being proposed by John Smith 19 would be able to do.

      Then again, I could have completely missed the point, as that book was horrendously bad. I was so disappointed in it.

  7. The lone lurker

    How about something like a Magna Doodle?

    http://en.wikipedia.org/wiki/Magna_Doodle

    I envision it to be similar to a mechanical typewriter but the keys would have magnetic letters and instead of striking an ink ribbon they would be placed upon the surface thus drawing up the ferrous material in the shape of a letter. A user adjustable wiper could be used to erase letters and a carriage return much like a standard typewriter would move the writing surface when a line was fully populated

    Printing might be an issue but some kind of photography could be utilised to store the finished page of text before you had to wipe it all of and move on to the next.

    1. ~mico

      Good one

      I was thinking along the same lines, too.

  8. BongoJoe

    I would imagine a large typewriter striking the back of something which is statically charged which is translucent. Then we have a continual fall of carbon down the back which sticks to the whatever has been struck.

    Some sort of lighting burning behind the whole thing to make the thing stand out and there you have it.

    editing? The same way as in vi or ed.

    Malc

    1. John Smith 19 Gold badge

      "I would imagine a large typewriter striking the back of something which is statically charged which is translucent. Then we have a continual fall of carbon down the back which sticks to the whatever has been struck.

      Some sort of lighting burning behind the whole thing to make the thing stand out and there you have it."

      I think you've come up the world's slowest laser printer :).

      BTW the tricky bit about those was the photoconductor used. Carlon's breakthrough was to use amourphous Selenium, which might have bee available in the 19th century and was by the late 30's,

      Actually the light erases the charge IIRC so the moment you've looked at it you've erased it. :(.

      1. BongoJoe

        Actually the light erases the charge IIRC so the moment you've looked at it you've erased it. :(.

        Well, that ought to confound the boujders and other assorted enemies of Her Imperious Majesty.

        As for the laser printer part; yes you're right. I just imagined something that bangs and crashes with massive sparks from the van der Graaf generators and the flickering limelight behind.

        And, what's more it could be portable. In a fashion.

  9. Berwhale

    Sand...

    http://www.youtube.com/watch?v=3x53fpoLeVc

    http://www.dubno.com/sandtable/

    1. John Smith 19 Gold badge

      "Sand...

      http://www.youtube.com/watch?v=3x53fpoLeVc

      http://www.dubno.com/sandtable/"

      Very cool, but I'd have to admit the response time might be a bit too slow. :).

  10. Cliff

    http://www.skyscanner.net/sites/default/files/image_import/arrivals.board.JPG

    Nothing in this technology is inherently electrical or couldn't be replicated with pneumatic/clockwork.

    Editing of the last line would be a series of dials to advance or regress the currently displayed character

    I'll leave the inner details of the implementation to you ;-)

    1. Pond

      Steam trains

      My mind jumped to the same thing as they use them in railway stations and trains ran on steam.

      1. Great Bu

        Re: Steam trains

        I also thought of this technique, having drawn up a few quick calculations the only thing I can't tie down is how many coal shovellers I will need to keep the hydropneumatic fliper mechanism running at an adequate speed.

        Did the original spec mention what FPS was required when running Crysis ?

  11. FuzzyTheBear
    Pint

    use projection .. have say .. 20 wheels with letter shapes that rotate with a mechanical keyboard or again a set of parallel wheels , backlighted to a frosty piece of glass .. edit is simply rotate one of the wheels . did i miss the point ? :D

  12. Gene

    This <could> work

    The VDU would be 60 characters X 40 lines. Each character would have tiny windows in a 5 X 7 grid to spell out a letter or number. A small red flag would be mechanically moved in the appropriate holes to make each character. Perhaps compressed air would work best. Note: this will not be portable.

    1. Gene

      Re: This <could> work

      Forgot to mention the backspace key that would clear the last character by removing the flags. Repeated backspaces would permit editing.

  13. Adrian 4

    Luminous paint

    A cylinder, coated with a luminous paint that shows an after-image.

    Project each character on the cylinder in turn, stepping the cylinder for the next line.

    Display won't be permanent but will fade .. ideally, as the cylinder turns but in the case of slow linefeeds, it will disappear in the viewed area too. Maximum writing speed limited by the time taken for the display to fade to lowish levels.

    1. phuzz Silver badge

      Re: Luminous paint

      That's not a million miles away from how a CRT works, and if you've every played an original Asteroids machine, the slow fade of the phosphor is used to great effect as a little trail behind each 'bullet'.

  14. Graham Dawson Silver badge

    I believe that what you need is something like a pin screen (http://en.wikipedia.org/wiki/Pin_Art), with the pins hidden within and only visible when they're pushed out.

    Pins could be manipulated with an x-y traversing arm that carries the character head.

    The character head is an n*n array of pneumatic or hydraulic hammers, activated in different combinations by the keys of the keyboard. The hammers push the pins out in various combinations; the pins are a high contrast to the screen (white on black perhaps) that then appear as a character. Each key press advances the head one space.

    Removing characters could be difficult. Perhaps the head can make a seal with the back of the character and simply suck the pins in again. Clearing the display would be as simple as tipping it back, or running a shallow-sloped wedge across the front to slide all the pins back into place.

    What this offers: the display becomes its own memory buffer. Each character is "stored" in the display and can potentially be read back by a second head - or even the same head, if there's some way of shifting the hydraulic connection from the keyboard to a memory device. It'd be slow, one character at a time, but it could be done relatively simply by extending all the hammers on the write head and then pressing it against the character, and recording which hammers were pushed back and which remained in place. That would give you a record of the pins that had been stored in the display.

    There'd need to be some way of preventing the pins moving at that point. Perhaps the pin screen could be made with a plate on the back that slides to one side when reading, narrowing the holes the pins pass through and wedging them in place.

    Other possibilities: given you're working with a matrix of pins, you can go back and edit whatever you want by shifting the head back and forth with a set of control keys. With a little tinkering you could even have different font sizes, typefaces and even simple graphics.

    No electricity or magnets required.

    I hope that makes sense.

    1. Chris G

      Not sure whether to vote you up or down as your suggestion is close to what I was thinking.

      The pin mountings would need to have just enough friction to resist movement under the influence of vibration in the machine but easy to register or cancel.

      With sufficient thought I bet it could be made to reproduce images in 3D but how to scan it in would be difficult short of shoving ( for example someone's head into it.

      As for actuation pneumatics could work.

      This would not be portable on any thing much less than a rail wagon

      1. Graham Dawson Silver badge

        I It could be miniaturised quite handily, I reckon. You could get the individual characters down to, oh, maybe 10 or 15mm on a side. Have the screen take up the majority of one side of a box with the mechanics in it and attach the keyboard via a cable containing the hydraulic lines.

        The major potential problems would be in the durability of the tiny pins and the reliability of their motion. The clearances required to make a small display wouldn't be beyond the capabilities of the technology of the day, given how fine they made their clockwork.

        I'm tempted to go off and scribble up some rough plans for this...

    2. John Smith 19 Gold badge

      @Graham Dawson

      "I hope that makes sense."

      It does.

      In fact a version of this has been used as readout device for blind people on a single line basis, with electro magnets tapping the pins.

      I've only ever seen the full 2d version in animations made by the National Film Board of Canada. :)

  15. Anonymous Coward
    Anonymous Coward

    Lots of scope for imagination here

    I wrote a number of stories with a Steam-Punk theme a few years ago. Mine had a lot of what Mills & Boon readers know as 'Bodice Ripping' in it. However, there was a large screen display (10ftx10ft) (paper that scrolled upwards) that displayed characters in perfect copper plate writing all driven by a 'Heath Robbinson' contraption. Communications was by Semaphore over Optical cables.

    Just let your imaginaiton run riot. Add a few delectable babes with very small waists and lots of ... well you get the hint. Oh and electrically powered Penny Farthings were around in abundance.

    Don't ask about a link to the story. I'm trying to find a publisher for a collection of stories on this theme.

    That also why I'm posting anon.

    1. Dave 126 Silver badge

      Re: Lots of scope for imagination here

      >that displayed characters in perfect copper plate writing all driven by a 'Heath Robbinson' contraption.

      Okay, having 'joined-up' writing (copperplate) is a whole extra level of complication, as we know from computer fonts... that's just showing off! However, automatic 'hand'-written text has been around for a long time:

      http://www.youtube.com/watch?v=FUa7oBsSDk8

      - clip from the BBC programme Mechanical Marvels: Clockwork Dreams, Professor Simon Schaffer examines a clockwork creation of Pierre Jaquet-Droz ((1721–1790))

      http://en.wikipedia.org/wiki/Pierre_Jaquet-Droz

      The recent Martin Scorcese film 'Hugo' contains an automaton based on Jaquet-Droz's.

      For the purposes of having a display unit that can be reset, a wax tablet could be used, and a sharp point in place of a pen. You could delete the display by heating it from below- the surface would be horizontally mounted, and displayed vertically by means of a mirror just like an old PacMan arcade cabinet. Obviously the mechanism would be constructed so as to output 'mirror writing'. )

  16. David Paul Morgan

    how about a matrix of...

    ... pneumatic rods.

    like the old screw-in shutter release buttons.

    each rod pushes (and pulls) a black 'pixel' to the front or rear of a small box. Box is filled with a 'milky' fluid so that, when black plate is towards the front, then it is 'black', but towards the rear, the 'milky' fluid makes it look white. each row of the display could be controlled by Jaquard (sp) cards like carpet weaving,

    editing achieved by changing the punch-card.

  17. Peter Gathercole Silver badge

    I thought this would be difficult, until I realised that you could use a pulse-coded dial (like an old-fashioned telephone dial), linked up to something similar to a Strowger exchange and a Solari split-flap display.

    I know I'm cheating a bit using pulse-encoders and electric motors, but I'm sure that you could use rachets, rotating shafts and slip-clutches everywhere that the modern displays uses electric motors.

    If we take 7-bit ASCII as the character set, that would mean 96 different displayable characters which include all upper and lower case English characters and numbers plus sufficient punctuation. This could be encoded using a 32 place dial like a rotary telephone dial, together with two shift keys shifting to different rachets to generate upper and lower case, and numbers, together with the punctuation. These work well with strowger type gear, and all you would need to do would be to pulse each successive position in the split-flap.

    1. Peter Gathercole Silver badge

      Forgot editing

      All you need is some way of changing the strowger selector to a particular position, and then rotating the split-flap character to the new character.

      I'm sure that there is a 'return to space' operation that can be applied to all character positions at the same time to clear the display.

      1. Terry Barnes

        Re: Forgot editing

        A strowger final selector has 100 selectable outlets (as opposed to a group selector which only has 10 selectable levels, horizontal rotation stopping at the first free outlet).

        But if you're going to cheat and do it electrically there are easier ways and you might as well use your strowger kit to drive nixie tubes.

        1. Peter Gathercole Silver badge

          Re: Forgot editing

          I think basic electricity use was discovered in the same general timeframe as steam. Michael Faraday was credited with inventing the electric motor in 1821.

          Nixie tubes are much later. Wikipedia suggests 1955.

          So I contend that basic electricity (not electronics, mind) is totally consistent with Steampunk.

    2. Steve Davies 3 Silver badge
      Boffin

      No, please not 7bit AScii

      It has to be Holerith punched tape and 6bit Characters (as used on ICL 24bit machines)

      You gotta go to extremes here.

      Now where is my paper tape repair kit?

      1. Peter Gathercole Silver badge

        Re: No, please not 7bit AScii @Steve Davies 3

        OK, I was using 7-bit ASCII as it allows upper and lower case characters (one of the requirements). 6-bit ICL code only contains upper case characters, although I understand (I only briefly used an ICL 1904 machine in the late '70s, and never got to grips with the available character set) that one of the characters was used as a shift, to provide lower-case characters.

        I admit that using an American standard was a bit low, but I could not think of a suitable non-US one. In any case, it would have to have been invented, because ASCII did not exist before 1960. If you wanted it to be authentic Steampunk, you would probably have to use the Cooke and Whetstone telegraph system!

  18. jubtastic1
    Happy

    Here's one someone made earlier, out of Lego

    http://m.youtube.com/watch?v=tyijTDuotu4

    Just requires the control cylinder to be rotated to the correct spot, which was something the Difference engine was capable of. For a larger alphabet replace the cylinder with a belt.

    1. Peter Gathercole Silver badge

      Re: Here's one someone made earlier, out of Lego

      Wow!

      That's awesome!

      1. Anonymous Coward
        Anonymous Coward

        Re: Here's one someone made earlier, out of Lego

        That is awesome, some damned fine Lego construction that. What is also amazing is the Youtube comments for the above video don't feature anyone calling anyone else a 'faggot'.

  19. This post has been deleted by its author

  20. keithpeter Silver badge
    Windows

    Rolodex

    I'm thinking a 'rolodex' for each character. Cards with the appropriate type on blown using compressed air to advance or retard one character. A series of air pulses to each position in a 70 by 40 array would allow text to be displayed.

    Once there is display, we can imagine valves or stopcocks at the end of each row or column to identify a character whose state could then be changed by a keyboard.

    Perhaps the keyboard could be set up pneumatically to advance one position so allowing over-typing?

    Serious brass and engineering here, nice idea.

    PS: Stross for Parliament (country left blank)

  21. John Doe 6

    dot matrix display...

    ...like the yellow ones used on buses but with more lines, could be powered either by compressed air, steam, hydraulic or mechanical.

    The correct name is "flip-disc display" https://en.wikipedia.org/wiki/Flip-disc_display

  22. John Smith 19 Gold badge

    Just to tighten up the parameters a little...

    Portability is not an issue but the display (screen is not always relevant with some of the proposals) should be person sized IE that you it can sit in front of you while you work.

    I'll note several of you came up with a version of the split flap display which would be reasonable.

    A few tidbits that might help stimulate creativity.

    1/3 of all pen nibs in the world during the 19th century were made in a 5 sq mile area around Birmingham. These nibs were high precision sheet metal components with tolerances in the thousandths of an inch range, mfgd by a multiple stage pressing process.

    OTOH devices that rely on controlled levels of friction seem more doubtful

    At around that time Reuters where transmitting dispatches around the world using microfilm messages attached to pigeon tails, so photo-etching would appear to be a viable technology for this era after all.

    Rather more doubtful is a thing called a Pockels Read Out Memory (PROM). This is based on Bismuth Silicon Oxide. It's possible to read and write it with light. The crystals are fairly small so some kind of projection system (like a microfiche reader) would probably be needed. A focused light light spot on the crystal at the right color could do selective erasure

    When I first thought of this challenge I pictured the back end being something like a the hardware in side a 1970's photo typesetter, where the various fonts were stored on microfilm cards and the font size varied with an optical system.

    Those who pointed out something like this was described in "The Difference Engine" should remember that was for a cinema sized projection TV type system, no good for interactive I/O. Standard I/O was by punched cards and printer listings, not really what I was thinking of.

    The etch-a-sketch ideas might also be viable as a graphics terminal for CAD work (sort of), although I'd guess more as an out put device with most of the "design" being by processing a command language, rather than tracking a mouse or light pen.

    The real challenge is the temporary and partially erasable image storage problem.

    1. Peter Gathercole Silver badge

      Re: Just to tighten up the parameters a little...

      The etch-a-sketch would be like a mechanical version of a Tektronix Storage tube terminal (Tek 4010 or 4014).

      1. John Smith 19 Gold badge
        Unhappy

        Re: Just to tighten up the parameters a little...

        "The etch-a-sketch would be like a mechanical version of a Tektronix Storage tube terminal (Tek 4010 or 4014)."

        True. Which is sort of the problem. It's a little too stable. Editing below whole pages was tricky, when it was not just impossible.

    2. Will Godfrey Silver badge

      Re: Just to tighten up the parameters a little...

      The split flap display is an out-and-out winner. It is indeed the Roll-Royce of character displays. Not only does the display look awesome as it's changing, the sound of all the flaps slipping over is fantastic too. As a kid I remember watching, and listening to the route displays at Charing Cross station - totally dumb-struck.

      Also, being such a simple mechanical device it can be powered by just about any means you can think of.

    3. Cliff

      Re: Just to tighten up the parameters a little...

      Split flap also solves your graphics with temporary storage and erasable requirements. Not just A-Z, add more flaps with graphics primitives, job done, including the power-down storage. The e-ink of its day.

      If you were insane you could probably knock up some pretty cool Lisajou patterns with rotors and string, vectorscope-style.

  23. Terry Barnes

    Praxinoscope

    A praxinoscope with 40 slots of different height, each one making up a row of 80 pixels. The display is monochrome and each pixel is either white or black, achieved by a shutter being raised or not. As the drum rotates all the pixels in the slot are knocked down to display white when in a non-visible position and then a series of 80 punch pins that are 'scanning' the display activate the pixel to black (or not) by their position.

    The whole thing is controlled by a series of gears to keep it in sync. The refresh rate is as fast as you can make the thing work reliably - the drum will need to rotate to at least 50RPM I think to make the illusion work properly but it's probably only realistic to refresh one or two rows per revolution.

    If I could upload a picture I could sketch how this thing works - try this though; Imagine a clock face. At 12 is the aperture through which the illusion is viewed. The drum spins clockwise. At 2 is a bar that knock all the pixels in a given row back to white. At 5 is the pin mechanism to set the chosen pixels to black. The bar and mechanism are attached to a stepping mechanism that starts at row 1 and increments once with each revolution - returning to the top after row 40.

    It's sort of time appropriate - the praxinoscope (using still photos) was invented in the 1870's.

  24. tirk

    Minecraft

    Has various examples of this, for example https://www.youtube.com/watch?v=h0e9fZ0242Y

  25. Captain DaFt

    OK, trying to think like they did back then...

    A lightly frosted glass pane for a display.

    Each character consists of a long belt made of gutta-percha with raised alphanumeric characters for each position. (lots of them, 800 for a 40X20 display.)

    The belt rides on a long track, with a geared cog bringing the desired character to the screen position.

    An eccentric cam then rotates, moving the apparatus forward, pressing the character against the screen, rendering it visible.

    The whole is controlled by an alphanumeric keyboard, whose key presses are interpreted by the the analytic engine, which controls the VDU.

    Whether this is the main engine, or a simpler one designed solely to drive the VDU is left as an exercise of the pocketbook of the builder.

    A crude rendition of a single character unit: http://i.imgur.com/aAI1QOQ.png

    And I realise that I'm probably as clear as mud describing this.

    Edit: Just wanted to add that this would probably be deafening when you reached the bottom of the screen, and the whole started to scroll!

    1. John Smith 19 Gold badge
      Unhappy

      "An eccentric cam then rotates, moving the apparatus forward, pressing the character against the screen, rendering it visible."

      The key problem is how to drive the individual cams that the moment when the belt is showing the right character with your approach.

      It's simple enough with a row of separately driven cams but how to get to the other rows?

      1. John Smith 19 Gold badge

        "The key problem is how to drive the individual cams that the moment when the belt is showing the right character with your approach.

        It's simple enough with a row of separately driven cams but how to get to the other rows?"

        This reminded me of a mainframe line printer mechanism. Thinking further I guess you could use a line screen hitting a rotating faceted mirror, creating one line at a time.

        The problem remain the storage requirements :( .

      2. Captain DaFt

        Actually, it would be simpler to leave out the screen, pushrods and cams entirely.

        Just have the face of the characters coloured so that they are visible against the belt, and viewed directly.

        But it occurred to me that the length of the belt would be impractical.

        52 letters, upper and lower case, 10 digits, space, math symbols and punctuation, plus the spacing between characters, the belt would probably be at least two meters long just to show characters a centimeter wide!

        You'd always be replacing broken belts!

        Nah, I'll just throw my vote in with the split flap display, a much better solution than mine!

  26. GloriousVictoryForThePeople

    Just use a CRT its not an anachronism

    Actually the Cathode Ray Tube is Victorian, dating from 1897.

    The HellSchrieber mechinism ( http://en.wikipedia.org/wiki/Hellschreiber ) is a simple dot matrix generation and display system.

    1. John Smith 19 Gold badge
      Happy

      Re: Just use a CRT its not an anachronism

      "Actually the Cathode Ray Tube is Victorian, dating from 1897."

      True.

      But the core requirement is no flowing electricity. IE no DC, no AC.

      Static charge (generated within the mechanism) is just about OK.

      But preferred primer movers and drivers are mechanical, pneumatic, hydraulic with light generated by gas, oil or lime light (that's gas flame exciting a lump of lime)

      1. Mage Silver badge

        Re: Just use a CRT its not an anachronism

        But DC Electricity dates from some time after 1798, at least in 1878 they celebrated a Volta centenary in France. (Not sure if his Birth, death or the Battery). I'm too lazy to look it up.

        Actually there is even the alleged Bagdad Battery,

        Two modern Steam Punk displays with memory ...

        Actually modern DLP projectors use an array of hinged mirrors, doing 40 x 20 characters each made of 11 segments (starburst display normally uses 14, but 11 is possible) with mirrors or the eInk concept ( using rotating 1/2 white & black segment shape beads instead of balls of Kindle etc).

        So we need to operate 40 x 20 x 11 = 8800 segments (either elongated beads 1/2 white & black or hinged mirrors). The beads need no additional light source, only need two positions, could have four faces, and are very stable.

        The first electrical matrix displays may be 1908. Victoria died 3 years earlier?

        12 segments need 3 columns x 4 rows . making the 40 x 20 text panel 120 x 80 X / Y multiplexed.

        One set of 60 Punched metal cards can do the character encoding for each 3 x 4 character cell. The minimum is to reuse 1 & 0 for I and O (as old typewriters did) and only have one case = 26 + 8 = 34 cards / characters.

        We maybe can share 5 and S and also 8 and B reducing it to 32.

        Update of entire display is slow, maybe up to 5 minutes if a 1/2 second per cell is possible. Erase would be fast (1/10th sec) as that can be a single operation without character cell. Per cell access perhaps 1/2 second.

        Each bead about 1/32" diameter and 1/4" long I think?

        I actually have 4 x 3 matrixed 11 segment + dp (= 12) LED panels. Curiously most starbursts are 14, but it looks like a 14 segment. (some segments are tied together).

        BTW LED was discovered about 1912 by a Marconi Employee but ignored. They were looking for a better RF detection crystal and a weak glow was of no interest. It was discovered again in 1930s by a Russian killed during WWII.

        Victorian CRT (on Continent the Braun Tube) was VERY dim as it was "cold cathode", no heater/filament. It wasn't till after thermionic emission was understood (Fleming and De Forest, in contrast Edison was technically clueless, he was a commercial person) that the CRT got a heater. Electronic TV using a CRT with photo target as camera as well as display was proposed about 1906. Baird was very steam punk with his mechanical TV based on Victorian era Nipkow Television.

        The Disk mechanical TV is limited to under 30 lines! or about 12 readable characters. Using stacked plates as a set of 200 to 300 mirrors (the plate edges) then about 20 wide by 20 high text is just about possible.

        Aconite (?) gas lamp illumination and mechanical drive is possible. But it has no memory and needs refresh of at least 11 fps. The static beads, hinged mirrors or flip panels all have memory and the beads and panels use ambient light.

        http://www.medwelljournals.com/abstract/?doi=ajit.2005.692.693

        http://en.wikipedia.org/wiki/Starburst_display

        http://en.wikipedia.org/wiki/Baghdad_Battery

        http://en.wikipedia.org/wiki/History_of_the_battery

  27. Mike 16

    Dates for SteamPunk

    There seems to be a wide variety of assumed time periods. Most of the steampunk I see these days amount to glueing fake cogs on an iphone case. Anyway, on to the technical stuff:

    1) The Difference Engine(s) built by the Science Musem were _not_ built of modern materails and processes. They had "the Beautiful Fragment" (Demo model of the calculation section of Difference Engine 1), so could analyze both the materials and the variation in tolerance. Difference engine #2 was built from the same materials and parts were (if need be) hand-filed to final profile. The only "cheat" that I know of was the use of Laser etching for the numbers on the wheels, which are only used for setting and debug. Unless you want to assert that Clements and Whitworth would have been unable to find someone to etch numerals in 1830s London, I don't think that "hack" is germane. The main problems, to the best of my knowledge had to do with the design being too ambitious (DE1 would have had less numeric precision and yet used three times as many parts), and some "personality issues", not uncommon in startups.

    Before Babbage was born, at about the beginning of the age of Steam (Boulton/Watt engine) Pierre Jaquet-Droz. built the Writer Automaton. There's a nice video of it in operation at

    http://www.youtube.com/watch?v=FUa7oBsSDk8

    including enough to see the basic principles. Substitute chalk for ink and Robert is your parent's sibling.

    Although The Writer wrote a message from "ROM", that ROM was changeable (as can be seen in the video), and basically served to position the cam-stack for the each character. Substitute a receiver/selector like that on the Teletype (last major change in 1925, first plausible "single wire serial" implementation about 1900). If you've ever taken a Teletype from the Model 15 to the 37 (1925 to early 1970s), you can see that there is very little "electrical" about them. Just about anything could wiggle the "magnet" arm, and it's all mechanical from there on.

    Or consider the Linotype, which you can see in action in this trailer:

    http://vimeo.com/15032988

    Clearly, fully formed characters are an advantage, and the main cost of extending to multiple rows would be enlarged matrix magazines. But since one would be reading the characters directly, the "matrices" could be less detailed and the whole "hot lead in your lap" aspect could be skipped.

    Anyway my main point was that "the age of Steam" lasted long enough to overlap with the age of nifty automata in the 18th century, and the age of Electricity in the late 19th. And of course a lot of electricity is still generated by steam, produced by Coal, Nuclear, Solar, etc. energy. How about a Steampunk Breeder reactor to drive this thing?

    1. Cliff

      Re: Dates for SteamPunk

      I love those automata. There's one, slightly less complex but using the same basic principal which can draw a dog, it's a wonder to watch. Engineering could be so precise back then, makes me weep to see the clumsy mass produced rubbish we make and throw away now.

  28. antia

    Ink on Glass

    You could print with ink (or stamp?), on the back of a glass screen, and have a mechanical wiper. This would enable simple-in-concept character-by-character deletion, possibly simultaneously on multiple lines.

    In true steampunk style, the device would be bulky, have mechanical arms (to wipe the screen), rapidly fall out of calibration, and need regular servicing (because you would end up with an overly smeared screen due to semi-dried ink).

    It would be able to have a similar look to a PC monitor, and possibly, through mechanisms of ungodly complexity, could support multiple fonts, multiple colours and print-screen functions.

    I had this idea before reading everyone else's posts, and that nobody thought of it first suggests to me that it is fundamentally flawed in a way I have yet to realise.

    1. Anonymous Coward
      Anonymous Coward

      Re: Ink on Glass

      "bulky, have mechanical arms, rapidly fall out of calibration, and need regular servicing"

      Like many 1970s computer disk drives then (e.g. the 2.5MB RK05 from DEC, which occupied a foot or so of 19" rack, and probably a KW or so of electricity).

    2. Anonymous Coward
      Anonymous Coward

      Re: Ink on Glass

      Make the glass screen a cylinder, maybe 400mm or so in diameter, with a big bearing surrounding each end and a gear on the bearing at one end to rotate the cylinder.

      The cylinder rotates between each line, and the precision of the rotation is therefore not critical. A golfball-type typewriter mechanism does the writing, against the inside of the drum. The golfball needs to be actuated by rods and rachets, but I think there is enough room especially as the mechanism can extend as far beyond the cylinder as you like at either end. The rotation between one line and another is accomplished by a steam powered cylinder with a ratchet, of course.

      There is then plenty of space for the cleaning mechanism which can occupy the entire back of the machine.

      You can see that this is simply a topological inversion of a typewriter which, if the paper could be persuaded to stay in line, could be fitted with a continuous loop of paper, so in a reasonable world the patent office would reject the application because the topological inversion of a typewriter would be an algorithm, which cannot be patented. Except in the US and dependencies, of course.

  29. GoingDigital
    Boffin

    I'm not sure all electronics is out. Quite a few electronic parts were around in steampunk 'times'.

    1750s: Leyden jar capacitor

    1840s: Relays

    1880s: Transformers

    1880s: Practical dry cell batteries

    1890s: Neon isolated and detected in the first neon tube.

    Both relays and neon tubes can be used for display, data storage, logic gates and oscillators. It's all there pre-1900. All it needed was hundreds of experts, hundreds of thousands of hours of labour and a computer room the size of a small country.

  30. Oh Homer
    Childcatcher

    Steam Powered

    It seems to me that a SteamPunk VDU should be steam powered and overtly mechanical in nature. My idea is not based on anything I'm aware of, either currently or historically, but is just something that occurred to me in a flight of fancy. I also make no claim to its practicality or efficiency, but it would at least be symbolic.

    I imagine a locomotive-style firebox which heats a sealed tank of air, thus increasing its pressure. That air is then released through a large number of small pipes, controlled by valves, back into the firebox, superheating the fuel at specific locations depending upon the pipe. The firebox door is a special plate that comprises both thermally conductive and insulative material, i.e. metal and ceramics, in a grid arrangement. High pressure air blown onto the burning fuel immediately adjacent to a conductive point on the grid, would cause it to glow red, much like a pixel on a modern display. By precisely controlling the valves, one could cause a specific pattern of points on the grid to glow, thus producing an image, including the image of a word. Closing a valve would allow the point associated with that valve's pipe to cool, thus fading to darkness. With sufficient precision, one could create moving images on the grid, or even various shades of colour, from red and orange to yellow and white.

    An even simpler design would be to use butane instead of solid fuel, pumping it through the pipes and burning it directly onto the plate.

    The valve control system could be something like a typewriter, where each key is connected to a pulley. You could even create a sort of "mouse" input device, in the form of strings tied to a knob in a radial pattern, that tense or relax depending upon the direction the knob is moved within a fixed two dimensional axis, opening and closing the valves as they do so.

  31. raving angry loony

    OK, I've always thought of "steampunk" as "modern world without the electricity to drive it". It's not "victorian", it's "no electricity but all the gadgets..."

    So....

    A display unit that shows various shades of "grey" could be constructed by stretching a rubber sheet across a series of (very small) actuator arms (named "pixels"). Each arm would drive into the sheet, creating a point of "lighter colour" where it presses into the sheet. Very much like a steam-powered typewriter but with 320x200 "keys" (initially, possibly growing to 1024 x 763 some day?)

    So yeah, your VDU would weigh several hundred kilos... but it would display monochromatic images.

  32. jake Silver badge

    Impractical concept.

    Size & mass (given the "steam punk" provision) says no.

    1. jake Silver badge

      Replying to myself (I hate that!) (was: Re: Impractical concept.)

      Purely mechanical plotter[1] and fan-fold paper.

      Walking the dawgs late-night is a great mind-settler.

      [1] Your issue, not mine ;-)

  33. This post has been deleted by its author

  34. poopypants

    Logie Baird

    Logie Baird's television used a rotating Nipkow disk, through which a modulated flashing light was shown.

    A pneumatically activated mechanical shutter placed in front of the light source might be the only variation required on his original design.

    1. Mage Silver badge

      Re: Logie Baird

      His (or Nipkow's really) allows about 3 rows of 3 characters max. It also has no memory.

      The static mechanical displays (hinged mirrors, flip plates/disks, rotating beads) can even be "read back" and used as memory.

      The various Magic Slate & Etch a sketch ideas don't need refreshed but can't be read back.

      The spinning disk needs refreshed at least 11 fps.

      I think the rotatable beads or flip plates/disks are best as:

      1) Ambient light

      2) Memory that can be read back

      3) Fast erase of whole display

      4) X & Y character cell addressable

      5) Possible to use segment bar per cell instead of dot array to reduce cell addressing from 5 x 7 to 3 x 4, much faster and easier to make a desk sized display using "starburst" than 5 x 7 dots.

      So though a CRT or neon dots/.segments (or even under driven tungsten filaments) and batteries are possible in 1890s, both are actually far more difficult and complex to drive than the bead or flip plate arrays.

  35. Anonymous Coward
    Anonymous Coward

    vdu 23,8202;0;0;0;

    See title.

    1. Will Godfrey Silver badge
      Happy

      Re: vdu 23,8202;0;0;0;

      Ha!

      I'm outing you as a BBCist!

  36. Anonymous Coward
    Anonymous Coward

    Why does it have to be reusable?

    Let it print on small pieces of paper that are put together like a jigsaw and define them as consumables. Probably ten elite persons have these things so it's not going to destroy the planet by itself. If you're one of these deviants who believe that paper is a finite resource then just build a pulping and bleaching station to reuse the paper and don't tell me you believe that nonsense about bleach killing fish.

  37. John Smith 19 Gold badge
    Happy

    Note printers and punch cards were viable by the 1850s

    The point of Stirling and Gibon's "The Difference Engine" was a 1950's IT environment could be implemented a century earlier.

    Offline, printed documents and punch card input and coding form input.

    But that was known from Babbages conception. It's the dynamic nature of a VDU that makes it so tough, specifically the storage issues.

    A computer can run with very limited RAM (look at the early mainframes). Data from card or tape, software (potentially) on a punched tape loop. So RAM is needed for temporary storage only.

    But if you need to drive a 80 x 24 VDU? It's not really surprising people went with hard drives for main memory (any 50's mainframe, or airborne computer with more than a few 100 bytes of main memory was likely using a disk or drum to do it). That's probably why the flip display has been so popular a suggestion. Note you could include some short cut signals to speed up its response EG "blank" input, "direction" so it can flip forward and backward and fast"skip" inputs IE skip say the next 16 characters at high speed and then flip 3 more forward (or backward) to display the right character.

    The order of characters could also speed things up. I'd put E,e and space in the middle of the range for example.

    Again I don't expect portability, color, graphics or even grey scale (obviously extra points if you can of course :) ). Just no external electricity.

    BTW the US aircraft industry has over the years tried to implement fluidic computers for backup fly-by-wire systems if the 5 existing computers got toasted by EMP. The approach used stacks of thin "shim" sheets (a few thou thick) with stamped and/or photo etched slots and tapped the hydraulic supply for drive power.

    Not very fast by desktop computer standards (the first FBW experiments were done by NASA with a surplus Apollo computer at 32KHz) but likely a bit brisker than a Babbage system (no need for an accelerator pedal for "turbo" mode either :) )

    I would like to have seen more efforts at some kind of approach using light more directly, but I accept that would be pretty tough. :(.

    1. Anonymous Coward
      Anonymous Coward

      Re: Note printers and punch cards were viable by the 1850s

      "fluidic computers"

      Thank you. Based on the absence of suggestions based on this technology, I was beginning to think I must have imagined the handful of Plessey fluidic logic modules I used to have (not the full Plessey Fluidics Experiments Kit, just a few hand me downs). They were maybe 5mm by 25mm by 40mm for a single function. The FBW controllers you mention sound like a generation or three more advanced/expensive than those.

      A VT05 (from the 1970s ie same era as the aforementioned RK05) is a 20 line, 72 column VDU with characters displayed in a 5x7 matrix. Wikipedia says the storage is a shift register, which in principle can be done with things like mercury delay lines and so on, or a collection of basic (fluidic) logic if preferred.

      1. Mike 16

        Re: Note printers and punch cards were viable by the 1850s

        "fluidic computers"

        My father worked with some fluidic computers, albeit typically analog ones. He was an auto mechanic specializing in automatic transmissions.

        You'd really want to run any steam-punk fluidics with oil or air, rather than actual steam. I shudder at the thought of designing a multi-layer logic block to account for pressure drop along the way due to condensation.

        1. Anonymous Coward
          Anonymous Coward

          Re: Note printers and punch cards were viable by the 1850s

          At one company I worked for, an engineer was given the job of sorting pill-like objects into different bins.

          He designed a working proof of concept which used fluidic logic to take a binary bin number and cause the pills to be directed to the correct bin by a flow of air. The magic of this approach was that the pills never actually came into contact with a surface and so were totally undamaged by their experience. The fluidic logic was amplified so that the signals to each "gate" were used to drive the working air blast that moved the pills.

          It was beautiful, elegant, and the management took one look at it and decided to make him redundant, presumably on the basis that when they said "original thinking" they meant "within, of course, strictly controlled limits".

          1. John Smith 19 Gold badge
            Unhappy

            Re: Note printers and punch cards were viable by the 1850s

            "He designed a working proof of concept which used fluidic logic to take a binary bin number and cause the pills to be directed to the correct bin by a flow of air. The magic of this approach was that the pills never actually came into contact with a surface and so were totally undamaged by their experience"

            Neat. I'm guessing he set it up as a fluidic binary search, with each level diverting the pill (depending on its properties) into a more and more specific tray for collection.

            Fairly simple (and with a clean supply) fairly reliable I would expect.

            Too bad about your friend.

            1. Anonymous Coward
              Anonymous Coward

              Re: Note printers and punch cards were viable by the 1850s

              @john smith 19

              He started a business. He was just the first out - this was the Thatcher period, when British companies were closing down so fast Companies House was probably having to work overtime just to keep track.

              You are correct about the approach with one caveat; it was more reliable than the previous mechanical sorting system. In an un-steampunk way, there were optical sensors on the outputs to ensure that each pill made it into the right bin.

        2. John Smith 19 Gold badge
          Happy

          Re: Note printers and punch cards were viable by the 1850s

          "My father worked with some fluidic computers, albeit typically analog ones."

          It's a term with several meanings. Some analog, some digital (in fact most are analog IIRC). There was a (brief) period where TTL logic gates (or more likely custom ASICs for hard wired logic) were just too expensive and fluidics were (in principal) a viable option.

          "You'd really want to run any steam-punk fluidics with oil or air, rather than actual steam."

          Absolutely correct. That said I have seen some designs for steam plant (Like the kind that powers warships) control elements which were fluidic controls. Also gas turbines, but more likely using oil.

      2. John Smith 19 Gold badge
        Happy

        Re: Note printers and punch cards were viable by the 1850s

        "The FBW controllers you mention sound like a generation or three more advanced/expensive than those."

        Old Popular Science I saw in a library, but I can't recall the dates on it. Otherwise you'll need to check http://www.dtic.mil/dtic/ as it'll be buried somewhere in the DoD archive, if they haven't sanitized them (like NASA did :( ).

        "Wikipedia says the storage is a shift register, which in principle can be done with things like mercury delay lines and so on, or a collection of basic (fluidic) logic if preferred."

        A fluidic shift register should work quite well. You don't have a fan in / fan out problem with that architecture. You just want to make really sure the system stays sealed. The 70psi air in industrial pneumatics makes a lot of noise when it leaks (At > 2 bar the flow can exceed the speed of sound with little or no nozzle) while liquids can be messy. The aircraft system was running at either 3000-4000psi.

  38. Mage Silver badge

    Rotating elongated Bead

    The simplest is part white and black and rotated by a lever to outer edge at either end. It doesn't need to be hollow, it can clip in to a frame.

    But if it has four sides (a rectangular bead) then you can have two mid shades of grey.

    We can't do more shades without going to very tiny dot matrix beads at 60dpi or higher as with our minimal parts (starburst) 4 x 3 cells, a part rotated bead would look rubbish and it's too complicated.

    I think basically two states is complex enough, on or off. via intermediate mechanical linkage such that only one cell at a time is driven by the 3 x 4 rods from "look up table". The character set wants to be as small as possible. After changing a cell by default the next cell is addressed. Erase can be a plate with rods to engage every bead and push to all white (or all black, which ever is blank)

  39. Christian Berger

    Projection displays

    There's a whole area of projection displays where you have a little wheel containing the symbols and project them into a back-projection screen. You'd just need a whole lot of those, one for every character position. That's also a comparatively cheap way of storing the data.

    Driving it would be a bit hard, but you can probably build a sort of matrix with cross bars just like in certain telephone exchanges. With it you'd need to transmit 2 mechanical signals, one for "reset" which engages a clutch which would turn the wheels until they are in a defined position, the other one would advance the wheel by one position.

    Yes it could probably be done. Unfortunately my background in mechanical engineering is to poor to actually do anything regarding this.

    1. Christian Berger

      Re: Projection displays

      Here's my little design:

      http://casandro.dyndns.org/theregister/steampunk_vdu.png

      Essentially it has 2 axles which go over the whole line.

      Screw A can move the device B left and right, which in turn moves the pawl in between the different ratchet pulley combinations. The pawl is moved forward by the axle D. This axle will move the pawl so it can click the ratched gear one step forward. This movement is transmitted via a pulley to a wheel with stencilled out characters. The characters will be projected to a screen.

      So A addresses the character position while D moves it forward by one position. You can obviously add some clutch to reset all character positions, similar to how it is done with mechanical counters.

      Maybe it is possible to turn the lower part of the system by 90 degrees so you can put the ratchet directly to the character wheel. This would save a lot of parts.

    2. John Smith 19 Gold badge

      Re: Projection displays

      "There's a whole area of projection displays where you have a little wheel containing the symbols and project them into a back-projection screen."

      You've half described the core of a photo typesetter, the sort of machine that NROFF & TROFF were written to drive.

      Basically the film wheels (there were other architectures) were what we'd today call the font ROM, with an inner and outer ring for lower and upper case characters.

      The difference was a single optical path and the stepping across the output (the film for the plate) being optical.

      With some work this would illustrate the different trade offs in design when you shift from electronic to opto/mechanical.

      Instead of a disk per pixel you could have a smaller number with separate drives and 2 postion mirrors. Effectively if the disks would write every copy of a letter (EG all the T's) across the screen, then all occurrences of the next character on the line until you ran out of disk, by which time the 1st disk had moved to the next Nth unique character on the line. 2 position mirrors should move much faster (and be easier on tolerances) than multi-position systems.

      1. Christian Berger

        Re: Projection displays

        I do not think scanning systems would work well in that regard. Where would you store the information?

        Here you can see numerical back projection displays in action:

        http://www.tv-nostalgie.de/seite_220.htm (top right image)

      2. Peter Gathercole Silver badge

        Re: Projection displays @John Smith 19

        Troff (Typesetter roff), not nroff. Nroff used a fixed character set described in a tmac file, and did not have the ability to scale characters to different character sets.

        One of the interesting things is that most people who used nroff assumed that it could only handle fixed-width font devices, because that is all they saw it driving (typically dot-matrix printers). It actually did allow partial character spacing, and I wrote a tmac file to use nroff with a HPLJ compatible OKI laser printer with the advanced character set option, that allowed nroff to produce right-margin justified proportional spaced text using micro-spacing.

        It could not handle pic or grap output, although I got tbl to produce nice solid-box outlines for tables. I believe that it could also do some basic eqn as well.

  40. Alan J. Wylie

    Laserscan HRD-1

    With the exception of lasers (replaced by limelight and lenses?) and the speed advantage of electronics, a lot of the early 1970's Laserscan HRD-1 could be replicated using steampunk technology.

    Here's a link to a brochure:

    http://www.pghardy.net/lsl/brochures/lsl_74_hrd-1.pdf

    A blue argon-ion laser was reflected by two large steerable mirrors set at 90° and then off two very small galvanometer mirrors (used for small movements and to compensate for the inertia of the large mirrors) onto a 100 by 70 mm area on a long roll of photochromic film. The was then projected at 10 times magnification onto a screen. Black lines appeared on an orange background. To erase, you wound the roll of film on one frame. By the time you got to the end, the first frames had faded back to clean orange.

    User display: 1 metre by 0.7

    addressability on screen: 50k by 35k

    Spot size: 200 microns (on screen?)

    http://www.pghardy.net/lsl/lsl_display_hardware.html

    http://www.pghardy.net/lsl/lsl_history.html

    1. John Smith 19 Gold badge
      Happy

      Re: Laserscan HRD-1

      "Spot size: 200 microns (on screen?)"

      The brochure claims 20 micros. 5000x 7000 pixels. That's laser printer territory (in 1973 !).

      Actually I have read about this type of display (Newman & Sproull Principles of Interactive Computer Graphics IIRC). They credit it to IBM as one of the original projection TV systems used for the control rooms at NASA during Apollo. I got the impression it was a case of (IBM) sold a couple (to NASA) and no one else was interested.

      Once again the problem is that huge storage issue

      IBM also did a very high speed/volume photographic data storage system that ran out of one of the US National Laboratories (Lawrence Livermore?). Astonishing, sort of like a digital photo-me booth.

      I guess the obvious question is what happened to LS?

      1. Alan J. Wylie

        Re: Laserscan HRD-1

        > "Spot size: 200 microns (on screen?)"

        > The brochure claims 20 micros. 5000x 7000 pixels. That's laser printer territory (in 1973 !).

        I think the 20 microns quoted was on the photochromic film, then a 10x ,magnification when projected onto the screen.

        > huge storage issue

        ...

        > what happened to LS?

        LSL had a vector approach to drawing and digitising. Once chips became cheap enough, you could just scan the whole document into memory in one go and throw computing power at it. PGH's web page I quoted above has a history of the company. Note the famous people that came to visit: Prince Philip, Maggie T

    2. Mike 16

      Re: Laserscan HRD-1

      Interesting. Sounds similar to, but different in many details from, the Greyhawk Soft Plotter.

      http://books.google.com/books?id=p68XfA0MHpoC&pg=PA226&lpg=PA226&dq=greyhawk+soft+plotter&source=bl&ots=yp4g-CtkOR&sig=wFSVM8DLD5HzeEJ3k8NY2lXzvKo&hl=en&sa=X&ei=r9ZtUvm4CcrkyAHGhIGwBA&ved=0CDYQ6AEwAA#v=onepage&q=greyhawk%20soft%20plotter&f=false

      Sorry for the long URL that might get mangled in the forum, but its the first "natural" result for the three words I used (Greyhawk Soft Plotter). For those disinclined to wade through it, this used lasers and mirror galvanometers like the Laserscan, but "wrote" on an LCD which was then projected. For the "absolutely no electricty" crowd, well, yes, this use electricity for biasing the LCD. I think collimated limelight would sub for the lasers, though. Oh, and Mirror galvanometerss were IIRC, invented by Lord Kelvin back before he was Lord Kelvin, for use on the Transatlantic telegraph. Yeah, the one that was used to send a message from Queen Victorian to Franklin Pierce. To me, "Steampunk" at its best is, pretty much. "Victorian Tech". The whole "no electricity" thing smacks of the various hacks used to _technically_ avoid forbidden activities on the Sabbath. Or the air-driven routers used by Amish woodworkers (for carving wood, not hooking their air-driven tablets to air-driven internet)

      1. Alan J. Wylie

        Re: Laserscan HRD-1

        > Greyhawk Soft Plotter

        Interesting. In the early 80's LSL worked with RSRE Malvern on similar technology. I remember the locked room with the infra-red lasers well: goggles and lots of warning notices. "Do not stare into laser beam with remaining eye" isn't applicable when you can't see the beam.

  41. Pen-y-gors

    Etch-a-sketch

    I'm sure some form of system could be devised to mechanically convert a keypress to move the levers to write on an etch-a-sketch screen. Perhaps use two (or three or four screens) that change position so that you can do 'pages' and then erase the oldest page as it moves back to be the 'current' page

  42. Christian Berger

    Vertically aranged barrels

    http://casandro.dyndns.org/theregister/steampunk_vdu2.png

    This design uses vertically aranged barrels, each one containing the symbols you need (you might want to put a magnifying lens before them.

    Each one contains a gear. A connecting gear transfers the movement of an axle with a gear pattern to one of the barrels. This way you can move one without the others.

    You can add a gear with a gap to each of the barrels. Then you can drop in a set of small gears from above. When you rotate them they will move the barrels until they have hit the gap and lost contact. After a while all barrels will be reset. This is similar to what you have on mechanical counters.

  43. SamBC

    Am I missing something...

    ...or could you not just use the same technique as the older railway station departure boards, some clocks etc - a bank of cards that flip around to show the desired character. One each for each character 'position' on the display.

    You'd have to keep the front clear of obstruction, sure, for the cards to flip around, but not very far. Depending on how you construct it, you wouldn't need that at all.

    1. Christian Berger

      Re: Am I missing something...

      Yes you are missing something. Those are called split flap displays and have been mentioned multiple times already.

  44. Version 1.0 Silver badge

    Dot matrix anyone?

    How about a dot matrix display? Small magnetic dots on the "screen" would be controlled from a mechanism behind the screen using rotating magnetic rods arranged in a matrix - when both rods aligned, the screen dot would be toggled to set (white) or unset (black) and would remain in that position until an update arrived. Overall, viewed from the mechanism side, the arrangement would be not dissimilar from core memory but somewhat larger.

  45. Gazman
    Headmaster

    The Tank

    Imagine a tank of opaque fluid fronted by glass (surrounded by gilt frame perhaps).

    The Tank has a reading light over it, aimed at the glass, which you switch on.

    You start to hit the control rods and an image forms on the Tank wall in front of you.

    Thousands of tiny rods pushed through the fluid by lever action, tipped by black rubber, are pushing up against the screen to form an image of a kitten and the mysterious message 'LOL'.

    That is all.

  46. Anonymous Coward
    Anonymous Coward

    Storage

    I once saw a purely mechanical means of data storage on an automatic assembly machine. It had been designed by a mechanic, not an engineer, and was quite ingenious.

    The mechanical data storage consisted of pins in brass tubes near the periphery of a wheel.The pins could be pushed up or down in the tubes mechanically, and were held in place by a spring and ball mechanism. In this system, each position on the wheel corresponded to a module in the process of assembly. If an assembly station failed,e.g.due to no component at that station, the pin was pushed down, and inhibited further operations at that station (by not pushing an actuator lever as it moved into position). The machinery was pneumatic.

    The VDU backing store could be replicated with 25 of these wheels each with 90 positions 4 degrees apart. Along the radius at each position would be 7 pins, so that an ASCII digit could be spelled out in binary. The writing head would consist of 14 little steam powered cylinders, 7 to push down and 7 to push up, and of course these could be combined into a single 14-cylinder unit for all the ones that lay between two wheels. The read head would consist of tiny steam valves actuated by levers that were operated by the pins as they moved into place. The levers would work on the side of the pin that stuck up above the wheel, thus avoiding the problem of ensuring that read did not destroy a bit state.

    Of course the whole thing would be a bit slow and, if combined with my internal golfball in glass cylinder CRT replacement, a complete screen refresh would probably take around half an hour. But then, I can recall when that was about how long AutoCad took to redraw a screen on a 286, and at the time everybody just accepted it.

    In fact, I can remember when people used to complain that it took about that long to log in to a Windows domain in some large companies. The past is a different country, even when it's familiar stuff.

    1. John Smith 19 Gold badge
      Happy

      Re: Storage

      "The mechanical data storage consisted of pins in brass tubes near the periphery of a wheel.The pins could be pushed up or down in the tubes mechanically, and were held in place by a spring and ball mechanism. In this system, each position on the wheel corresponded to a module in the process of assembly. If an assembly station failed,e.g.due to no component at that station, the pin was pushed down, and inhibited further operations at that station (by not pushing an actuator lever as it moved into position). The machinery was pneumatic."

      OK I think I get roughly what you're talking about. 25 wheels for a line with 7 or 8 bit at each location (IE A row) and 90 locations IE 90 lines, to simplify the indexing problem. With even fairly loose tolerances (say 3-6 thou) this could still be made quite compact.

      Note that the classic way to handle this would be a servo valve arrangement that has a sensitive detector (to sense the pin out or in) releasing fluid which then drives a 2nd more powerful valve.

      That said a sufficiently sensitive system could probably get away with a single level sense/actuate approach..

      It sounds like an adjustable version of the pins in a musical box.

      As I noted above pneumatics and hydraulics are fine for this.

      Any ideas for what the display would be?

      1. Retired Spy

        Re: Storage (Pin Braille Type Display)

        This sounds very much like a braille output device. They are mounted horizontally, typically only display a single or small number of lines, are driven via electricity and output in braille. I love the idea tho'. Now if I were to try to build one...

        (1) 80x24 character pinboard of 7 segment displays (each segment is a hex pin as shown in the current wikipedia entry),

        (2) Pins are set in a plate, plate has cams at corners (and centre), cams push plate back to reset pins, retracting the pins below the "screen" surface, clearing the display,

        (3) A carriage traverses the the display one character at a time, pushing pins out towards the user,

        (4) The carriage carries a set of dies, each of which will push a select group of (7-segment) pins,

        (5) The carriage is positioned so the pusher die of the selected symbol is aligned behind the next character cell,

        (6) An actuator (pneumatic would have the best sound I think) pushes the pinsetter forward which shoves some of the pins forward, out through a perforated "screen",

        (7) A light shining from the side highlights the set pins so you see the letterforms in the contrast between the pins and their shadows,

        (8) For cool factor, Y positioning could be done by moving the screen itself vertically up one line as the carriage slides back to the left. So clearing the screen would included dropping it down so the top line of the screen would be printed first.

        To drive it, you would need to select a symbol (1 to N, N <= 128) and a horizontal position (1-80). If you process one line at a time, reaching the end of the line would trigger the carriage to return to the first character position of the next line (sound familiar?). This could be done by having the carriage slide along two shafts. One rotates to select the character position (X position), the second shaft rotates to select the character. I would use a sliding tube for the pneumatic actuator (shiny brass like a trombone).

        If the die selector was a wheel (daisy wheel printer), you could use modular arithmetic, so the symbol selector shaft would only ever have to rotate in one direction.

        You would have to engineer the whole thing to be self-aligning, which is straight forward, but requires some thought. With high strength materials and adequate power a evolved design might get up to 7 or even 10 cps. If you used a ballistic pin driver, you could probably go much faster, tho' at some point you would run into serious material (elastic rebound, wear, jamming, etc) problems since you are pushing serious mass (pins) as opposed to ink.

        The attached keyboard would be pretty cool too. People would probably pay you just be able to type and see their output appear on the display...

  47. This post has been deleted by its author

  48. harmjschoonhoven
    Angel

    Machina Cryptologica

    Ah, a steampunk VDU. That is the Machina Cryptologica of Athanasius Kircher S.J., described in his book Magnes, sive arte magnetica (The Magnet or the Magnetic Art), Cologne 1643.

    See "Athanasius Kircher, The Last Man Who Knew Everything" by Paula Findlen (ed.), Figure 11.1.

  49. harmjschoonhoven
    Happy

    Re: Machina Cryptologica

    For an illustation of the device go to

    http://diglib.hab.de/drucke/218-25-quod-1/start.htm?image=00425

  50. Long John Brass

    What about a miniature version of this?

    http://en.wikipedia.org/wiki/Split-flap_display

  51. This post has been deleted by its author

  52. Chromatix

    A mechanical seven- or sixteen-segment display is feasible. Here's an electromechanical type which was used nationally as railway clocks for a while: http://www.youtube.com/watch?v=djcaR1tVdVg While this example (designed for numbers only) is merely seven-segmented, a sixteen-segment version could display letters quite competently.

    The obvious way to drive one without electricity is with pneumatics. You would need to build some fluidic logic to store a six-bit character code, and then decode it to drive the individual segments. It doesn't matter if the decoder is a bit slow, since you only need to latch the code into the storage before you can move on to the next display position.

    Input on the bottom row could be done by mechanically actuating the storage valves with a sliding camshaft. Slide the cams under the position to be updated, then rotate the appropriate ones to "set" or "reset" the bits required. A separate "group reset" lever could be arranged to zero the storage registers in the entire row.

    Additional rows above could be driven from the bottom row, and filled by scrolling. A tradeoff could be made here by sending the decoded signal up for latching directly, rather than duplicating the decoders.

  53. yorksranter

    Imagine a flat belt like a conveyor belt, made of an impermeable material. Below it, we have a tank of oil or similar fluid. As the belt travels around its loop, it dips into the fluid, so the top surface of the belt is always covered by a fresh film of oil. Better, the belt passes through a set of brushes that are trickle-lubricated, providing the same effect less messily and with less weight.

    We then have a head that can be moved in two axes - i.e across the belt and along it - within a defined rectangle. This rectangle is one video frame. The head makes a mark in the film of oil with a very sharp tool (like those pen nibs mentioned earlier). As the belt moves on, the frame of graphics is illuminated. The untouched oily surface is reflective, the marks darker. A mirror placed above this section of the belt, angled at 45 degrees, reflects the image onto a screen.

    The image is then plunged back into the oil bath/dragged through the brushes and therefore erased. The frame rate is critical. It shouldn't be a problem to drive the belt at the movies' 24 fps, because the drive is basically a big film projector (actually, not necessarily a very big one - if the head is small and precise enough we can magnify the frame optically).

    Making the write head keep up is hard. This could be tackled in a couple of ways. One would be to segment the frame into subframes drawn by multiple heads, with genuinely steampunky weight and complexity.

    Another way would be to ditch the vector graphic approach in favour of bitmap, which only needs one axis of control. If you did it that way, you could also have multiple pins on the head, making it a dot matrix print head and therefore making it possible to output characters as such rather than drawing them as graphics.

    A machine that lowers a needle onto a passing sheet of flexible material, precisely, really fast would not be at all alien to the Victorians because it's basically a sewing machine or, scaled up, a power loom. This solution is basically a hybrid of textile and cinema technology, so all you need are the Lumiere brothers to walk into the right pub in Leeds in 1895 and talk to some loom engineers;-0.

    The encoding would be basically x-distance, pin number, binary up/down, as the speed of the belt provides the y-distance. You might even be able to map it into Baudot and push it down a telegraph wire or over a wireless telegraph. Alternatively, and probably more robust, imagine the line across the belt as a packet containing a fixed number of sequential pixels that can contain up/down commands, with blanks padded. Now you have the potential of telegraphing into the central loom from a branch office.

    (note to SF writers: the sysadmins on this are going to be scarily tough women from Halifax.)

    1. John Smith 19 Gold badge
      Happy

      @yorksranter

      Neat. You have roughly described an opt/mechanical eidiphor projection TV system.

      Plausible but getting the oil viscosity just right (and either limiting its changes with temperature or temperature compensating the system) will be tricky. Neatly you've got the "display" and the "memory" components combined.

      A very steam punk solution.

      "(note to SF writers: the sysadmins on this are going to be scarily tough women from Halifax.)"

      Yes, I've seen how fit the women in textile factories tend to be. The phrase "fighting for your man" takes on a whole new meaning :) .

      1. yorksranter

        Re: @yorksranter

        I sort-of remembered that Doug Engelbart's 1968 NLS demo used a projection system borrowed from NASA involving a surface coated with oil and an electron beam.

        I actually think vibration might be an even worse problem than temperature.

        1. John Smith 19 Gold badge
          Happy

          Re: @yorksranter

          "I sort-of remembered that Doug Engelbart's 1968 NLS demo used a projection system borrowed from NASA involving a surface coated with oil and an electron beam."

          That would be an Eidiphor.

          "I actually think vibration might be an even worse problem than temperature."

          You could be right. Temperature control is small enclosure is pretty well understood. Obviously the wider the range you need to keep it in the easier, but I'd say a couple of degrees would not be too tough. When you want 1/100 of a C then things get tricky.

          Vibration on the other hand, Not so easy.

  54. Spoonsinger

    Sorry the whole 'do something in a "Steam punk" way' is just silly.

    a) Victorians, (as previously mentioned), had electricity.

    b) Steam punk doesn't preclude the use of electricity, and if it does and isn't using 'magic', (maybe something like Space 1889), it just changes the laws of physics anyway.

    c) Just use a large pin art board with appropriate mechanics if you want to go the whole 'not quite Steam Punk VDU route) - maybe just call it Georgian Punk, (but then they knew about electricity as well - but still burnt witches, so will slot in with the whole fanaticism thing).

    1. Cliff

      Re: Sorry the whole 'do something in a "Steam punk" way' is just silly.

      I think it's more of a fun thought experiment with loose boundaries than precise historical re-enactment

      1. Spoonsinger

        Re: Sorry the whole 'do something in a "Steam punk" way' is just silly.

        Point taken, (with added plus), however if it's a 'thought experiment', (i.e. just going a mechanical way), why bring in the whole Steam Punk thing? Steam Punk tends to be Victoriana + magic or Victoriana with wayward physics. If you just want a mechanical solution 'thought experiment', there isn't really a problem - other than the size of the thing you build.

        1. Cliff

          Re: Sorry the whole 'do something in a "Steam punk" way' is just silly.

          Fair point, and +1 to you too. Guess it's a bit of a shorthand for dark wood and mechanical ingenuity. Although it's also a shortcut for sticking fake cogs and brasswork on a fucking iPhone, so not a very useful shortcut...

          1. John Smith 19 Gold badge
            Happy

            @Spoonsinger & Cliff

            As I pointed out up thread

            In the spirit of El Reg headlines the story doesn't really have that much to do with "Steam Punk"

            In fact, hardly anything.

            Which is also in the spirit of El Reg headlines.

            If you read up thread you'd have seen one of my motivations was the idea of a "post WWIII" computer system. One that would not be fried by EMP at any range, because it simply did not use electricity to begin with.

            How does that basis work for you?

  55. yorksranter

    Actually, add more complexity: we need a shutter and a projector-like step drive.

  56. John Smith 19 Gold badge
    Happy

    In the tradition of El Reg headlines the title did not *quite* mean what the story said.

    If you'd read my OP you'd know that idea was to try something resembling Vanavar Bush's "Memex" first described in the Atlantic Monthly in 1945. Well after electricity suppliers were available, but primarily opto-mechanical. Bush's team were quite interested in closed tank liquid processing for microfilm data storage and "dry" film processing (later developed by a different team for the 1960's Lunar Reconnaissance Orbiter mission).

    The real driver was to do it without electricity. I did not expect any trouble with a keyboard but the display/memory was always going to be the long pole in the tent.

    The full origins for this idea (for anyone interested) were:-

    Charles Stross's "Laundry" series and the description of a Memex in that.

    The UL that DARPA had developed a mechanical computer as part of the contingency planning for WWIII

    Sterling & Gibson's "The Difference Engine" and how a society could be run with punch cards, printers and mechanical computers.

    The work on fluidic computers for backup flight control on relaxed static stability aircraft.

    Gordon Dickson's "Dorsai" trilogy and the idea that in a world of starships, nuclear weapons and (IIRC) suppressor field ground combat had reverted to "spring guns" (no combustion) and close range signalling to whistle codes (EMP can cook any radio system).

    S&G describe essentially a 1950's IT environment. I wondered if you could drag it into the 1960's or 70's.

    As always the joker in the pack is not the logic devices, the display or the program storage, it's the RAM.

    So far the answer is looking like "yes."

  57. Jerky Jerk face

    Use an array of steel pins to create the display, like the old timey "Pin Art" desk toy from the 70s80s?

    Brought to life recently by the latest Superman movie and also featured in The Tom Cruise Sci Fi movie oblivion.

    Would 'look' right at home as part of steam punk take

  58. Gavin McMenemy

    I don't know if anyone else has mentioned this but...

    I find it ironic that the memex has been mentioned in an thread about steampunk given Charlie's well documented disdain for the "genre".

  59. srimech

    Ball bearing dot matrix

    I prefer the flip-display solution, but here's another alternative. Metal grids are quite cheaply available and if laid flat, ball bearings will stay in the hole they're placed upon. A ball bearing can block light, or alternatively, glass marbles could be used which can focus light to a display or refract it in other ways.

    The advantage would be that the grid and the ball bearings are cheap stock components, so you wouldn't have to manufacture each bit. You would have to make the device for pushing ball bearings (and reading their position if you want to read out) but that would hopefully only be made once per word, column or so on.

    I'd imagine you would not add or remove ball bearings from the grid, but just shunt one bearing between a '0' row and an '1' row per bit, with the 0 row not being visible to the user.

    1. John Smith 19 Gold badge
      Happy

      Re: Ball bearing dot matrix

      "I'd imagine you would not add or remove ball bearings from the grid, but just shunt one bearing between a '0' row and an '1' row per bit, with the 0 row not being visible to the user."

      I sort of get your idea. The problem then becomes implementing some kind of X, Y mechanism so at the intersection a ball gets moved. Cycling through the X and Y ranges should be (relatively) straight forward.

  60. Chozo

    IF electricity==true

    How far down rabbit hole of technological possabilities would be acceptable to the Steam Punk community?

    Relay logic is a sure bet but what about Thermionics?

    Check out the work of H.P FriedRichs, he builds his own valves & transistors.

    http://hpfriedrichs.com/mybooks/ioa/bks-ioa-gallery2.htm

    Funnily enough there's a lot of work being on done today at the microscopic level with folk etching tiny triodes in silicon wafers so in a steam punk universe the same technique could exist just applied at a larger scale.

    As for the display itself this demo for a DIY LED is pretty crude but its got potential.

    http://makezine.com/2009/04/14/lost-knowledge-homemade-electronic/

    PS: very impressed with my new Röntgen Ray wireless telegraph, excellent reception even through walls.

    1. Chromatix

      Re: IF electricity==true

      I think there are two reasonable technological cutoff points for including electricity in a steampunk aesthetic:

      Liberal: semiconductors are not allowed. Simply assume that the PN junction does not work, thereby disabling the LED, the silicon rectifier, transistors of all kinds, and the microchip (as we know it, anyway).

      Conservative: not only do semiconductors not work, but neither do thermionics. The thermionic valve (or "vacuum tube" for Yanks) was invented in the early part of the 20th century; although cold-cathode effects were known for some time before then, it was difficult to find uses for them.

      Note that even the "conservative" definition still permits "vacuum tube like" technologies which actually rely on striking an arc in a gas medium, eg. the mercury-arc rectifier, the Dekatron and the Nixie tube.

    2. John Smith 19 Gold badge
      Unhappy

      IF electricity==true

      That's pretty much the only ground rule.

      It's not.

      If it helps think of this as re-booting civilization.

      For those wanting to go "home brew" semiconductors you'd have to go far to exceed the thin film and thick film transistors of various Sci Am contributors in "The Amateur Scientist" column back in the 60's and 70's.

  61. Munchausen's proxy
    Pint

    Linotype

    Have we already forgotten this mechanical wonder of information dissemination?

    1. yorksranter

      Re: Linotype

      My top three sources for this were photography/cinematography, textiles, and printing:-)

      1. John Smith 19 Gold badge
        Happy

        @yoursranter

        "My top three sources for this were photography/cinematography, textiles, and printing:-)"

        If you were aware of Eidiphor I'd have expected you to draw the comparison yourself.

        Don't worry about it. I'm just a mine of odd bits and bobs of engineering knowledge and link things together.

  62. Isn't it obvious?

    I actually thought about this years ago...

    ..and the cutest idea I came up with was to use an array of typeballs (from old IBM Selectrics), each behind a mask, plus mirrors and lenses to enlarge the image of the "current" character into a display element. If you're very clever with the optics, you could use different light paths for various typeballs so you could position some typeballs behind others to compress the vertical and horizontal size of the display.

    The Selectrics were entirely mechanical, and you could use any motive source to drive the main shaft (which positioned the head). It even encoded characters in binary, which would simplify using the inputs elsewhere.

    Moving the current (writing) address from character position to position is a problem I didn't dig too far into, but you can at least read back the state of each character from the current set of positioning bits.

  63. Old Handle

    I see the name linotype already came up, but I'm imagining some a bit like that in that the letters would be loose tiles that the machine would physically arrange to display text. This would allow insertion, replacement and deletion as (relatively) simply mechanical operations. The tiles would have unique notches or teeth on the back, allowing the machine to read the text, and also sort the tiles back into the correct trays when they were deleted from the display.

    1. Mike 16

      Re: Linotype

      "The tiles would have unique notches or teeth on the back, allowing the machine to read the text, and also sort the tiles back into the correct trays when they were deleted from the display."

      Linotype matrices already have coded notches (albeit on the top) for precisely the purpose of sorting the them back into the magazine. 19th century tech FTW.

      I agree that combining display with storage is a big win. As our host has mentioned, the hard part is the RAM. On that note, if anyone can help me find info on a keyboard driven Morse Keyer with FIFO buffer (to smooth out sending speed), purely mechanical, I'd appreciate it.

      Please, no pointers to the punched-tape systems or machines implemented with PICs or PCs, or even the rash of surplus-electronics versions in the 1960s. This thing was briefly mentioned in QST in the 1960s or so, and seemed to have implemented the FIFO in some sort of drum, probably like a re-writeable pin-drum (ala music box). Very 1890s, or 1920s at the latest.

      1. John Smith 19 Gold badge

        Re: Linotype

        "On that note, if anyone can help me find info on a keyboard driven Morse Keyer with FIFO buffer (to smooth out sending speed), purely mechanical, I'd appreciate it."

        "This thing was briefly mentioned in QST in the 1960s or so, and seemed to have implemented the FIFO in some sort of drum, probably like a re-writeable pin-drum (ala music box)."

        Weird.

        The "drum with pins" calls to mind a controller for complex chemical synthesis I saw in an old Sci Am for the "Merrifield Technique." Essentially an uncommitted cam timer where you set the speed of the drum rotation and programmed it by inserting or removing "plugs" at each location on each control channel, each segment along the drum being a channel.

        But doing it mechanically? If the input is Morse it could be using a solenoid to pull or push the pins in or out of the drum. How much store? If the storage is substantial a spiral pattern round the drum would make more sense.

        You say a keyboard, do you mean the keyboard writes to the drum in a code like ASCII (or more appropriately Baudot)?

        1. Mike 16

          Re: Linotype

          Actually about the Morse keyboard sender.

          ----------------

          The "drum with pins" calls to mind a controller for complex chemical synthesis I saw in an old Sci Am for the "Merrifield Technique." Essentially an uncommitted cam timer where you set the speed of the drum rotation and programmed it by inserting or removing "plugs" at each location on each control channel, each segment along the drum being a channel.

          -------------

          Sort of the extension into three dimensions of certain light-timers. You place "on" and "off" pegs wherever you want on a disk which is truned by a clock motor. Maybe I should say "4 dimensions", as your device allowed variable speed.

          --------------

          But doing it mechanically? If the input is Morse it could be using a solenoid to pull or push the pins in or out of the drum. How much store? If the storage is substantial a spiral pattern round the drum would make more sense.

          -------------

          No solenoids. No electricity at all, far as I could tell, other than that switched by the output contacts. I don't think it had an electric motor either. Appeared to be clockwork, although I suppose a large office could be run off a line-shaft with belts. :-)

          As far as I could tell, the keyboard pushed a code into a row of bistable pins, then advanced the drum to the next row. Meanwhile, a "reader" of some sort walked around the drum, reading and clearing rows. From the (dim and dimly remmbered at this point) photo, there appeared to be maybe 50 rows of 7-10 (a guess, see below) pins, or maybe invertable springs, or wire-memory like used by Zuse used in Z-1 and Z-2.

          ---------------

          You say a keyboard, do you mean the keyboard writes to the drum in a code like ASCII (or more appropriately Baudot)?

          -------------

          More likely a native representation of Morse. I know of three possibilities. the first is like some tape machines used, punching "on" and "off" holes and advancing a variable number of spaces depending on character length. This makes the reader/sender easy, but complicates the keyboard/punch. Another is to use 10 bits per character. The send-head close the keyer for each "mark" pin, then clears the pin. It stops when it has no more mark pins. The last (which I tend to use in Morse-sending code) is to send a dit for a zero and a dah for a one, except for the last bit, which is ignored, but allows 1000 (read right to left) to be ... (S) and 1111 to be --- (O), etc. There is a little more complexity in the read/send head, but you can store all common characters and digraphs to fit in 7 or 8 bits (pins), for more flexibility and fewer fiddly pins.

          How's that for a digression? To further digress, in the same time-frame as that article, were usually advertisements foe Frederick Electronics, touting their Morse<->Baudot gateways. Electronic, but fascinating.

  64. suranyami

    It already exists: a linotype machine

    http://en.wikipedia.org/wiki/Linotype_machine

    These have been in use in the printing industry to compose a "line of type" at a time, which forms a complete block of text. Editing of the last line was made possible by being able to reject a whole line and start again.

    Although intended for hot metal printing, it doesn't take much of a stretch of the imagination to see this used as a display medium and with some kind of binary encoding physically embedded on the individual character blocks for it to work as digital input medium.

    Later systems in the 1960s used similar techniques but with Microfiche, which would also be an avenue for developing projection displays.

  65. Alan Bourke

    Something like the Entex Adventure Vision / Nintendo Virtual Boy?

    A spinning mirror and a vertical line of LEDs (or bulbs) producing this?

    http://www.youtube.com/watch?v=R4gNP74tqsI

    Or Logie Baird's 'Televisor' ?

  66. John Smith 19 Gold badge
    Unhappy

    You need to be thinking *phototypesetter,* not Linotype if you want to go that way.

    Linotype machines are not desk sized.

    What you've missed about Linotype m/cs is they cope with unlimited copies by casting lead copies of the raw font forms.

    Consider the proverbial 24 * 80 display.

    That's 1920 copies of any character to cover all the possibilities.

    1. Mike 16

      Re: You need to be thinking *phototypesetter,* not Linotype if you want to go that way.

      > Linotype machines are not desk sized.

      You clearly have never seen some of the desks on Mahogany Row. Besides, a good bit of the bulk is the melting pot and casting operation. Not to mention the expanding spaces. I think we can do without full justification.

      > What you've missed about Linotype m/cs is they cope with unlimited copies by casting lead copies of the raw font forms.

      Didn't actually miss that. Met my first Linotype in 1958 or so. Picked up a fresh slug. Anyway, my point ws that the basic ideas behind coded slugs and automated sorting would provide a way of storing the characters in a way that could be read back. Useful for operations like "delete character". I'm not sure how you propose to implement sceen-edit with a phototypsetter. Not to mention how you plan to develop each frame (Although there was a Vctorian-era camera with built-in developer. Sort of like a proto-Polaroid).

      > Consider the proverbial 24 * 80 display.

      My first non-printing display was more like 12x72, or even 8x40, but agreed, the matrix (or character) magazines would need to be large. OTOH:

      > That's 1920 copies of any character to cover all the possibilities.

      Well, a screen ful of AAAAAA would be a bit boring. and I suspect that the 48 characters typical of early (pre 1960 or so) Data Processing equipment would do for a character set. BTW: I've seen older books which substituted characters, possibly when the supply of, e.g. 'Q' ran out while setting a page. Not to mention the common use of lower-case l for numeral one, and occasionally upper-case O for numeral zero.Humans are adaptable.

      1. John Smith 19 Gold badge
        Happy

        Re: You need to be thinking *phototypesetter,* not Linotype if you want to go that way.

        "You clearly have never seen some of the desks on Mahogany Row."

        True.

        Besides, a good bit of the bulk is the melting pot and casting operation. Not to mention the expanding spaces. I think we can do without full justification."

        Also true, but how are you going to produce the multiple copies of a letter?

        " I'm not sure how you propose to implement sceen-edit with a phototypsetter. Not to mention how you plan to develop each frame "

        In my OP I limited it to last line editing.

        "Not to mention how you plan to develop each frame "

        While I'd like to have seen more use of a photographic approach the challenge is tough enough (I think) that no solution should be ruled out unless it's very cumbersome. The trouble was always going to be building an erasable photographic film. That said the USAFRL did work in the mid 90s with bacterial rhodopsin in PVA using 2 color write (1 in the IR) and the IR frequency to erase. Their achieved resolution was 1400 line pairs/mm so a quite small piece could have served.

        "Well, a screen ful of AAAAAA would be a bit boring. and I suspect that the 48 characters typical of early (pre 1960 or so) Data Processing equipment would do for a character set."

        True but worst case planning is that. I chose 24 x 80 because a lot of TTY programs seemed to go with that by default. Regarding a character code well Baudot, used for teletype/telex systems would be more authentic.

        1. Mike 16

          Re: You need to be thinking *phototypesetter,* not Linotype if you want to go that way.

          > how are you going to produce the multiple copies of a letter?

          The letters are exactly analogous to the matrices of a Linotype. If you need 1920 As, then sop be it. Makes the type-magazine pretty large, but it is in the class of problems which can be solved by throwing money at them.

          As opposed to the class of "We have no idea how to do this, even if we had billions at our disposal.

          You just know that this thing is going to be so expensive that it will be installed in the managing director's office, where he will use it to train his philodendron, rather than actually having to work very often.

          > While I'd like to have seen more use of a photographic approach the challenge is tough enough (I think) that no solution should be ruled out unless it's very cumbersome.

          Well, quite a few photographic solutions, especially of the Victorian era are "ruled out" by modern health and safety laws. A friend has been trying to create less toxic (i.e. legal for normal people to use) equivalents to older photographic processes, and had some tales to tell. OTOH, I spent enough time playing bare-handed with mercury as a child that I should be mad as a hatter by now. Oh, wait...

          > Regarding a character code well Baudot, used for teletype/telex systems would be more authentic.

          Baudot was a Johnny come lately in telegraphy (not to mention that what we know as Baudot is actually Murray's later version), pre-dated by at least Morse and Cooke/Wheatstone multi-needle codes. I'm mentally trying to come up with something buildable in Babbage's lifetime, thus from bits and pieces that existed then. Morse makes the cut. Baudot not.

          I'm beginning to think that a "practical" machine would benefit from your notion of phototypesetter, where only one master copy of each glyph is present, but with a pin-screen to store the image.

          Of course, one of my favorite notions would stuff clear or opaque marbles (small as possible) into the bottom of a frame with translucent front and back, and perhaps an optically dense fluid. As each row of these "pixels" is "rendered", it is shoved up in the frame, with older lines emerging from the top. If the opaque "marble" were iron, the two types could be sorted magnetically. for reuse Of course, even a 24x80 screen of 5x7 (in a 6x8 cell) characters would need 92160 of each (actually fewer of the clear ones, unless we do inverse video) .

          Still about the same as the 1920x48 "type matrices" for my Linotype-style machine, but cheaper in bulk, and you could do graphics. can one render a convincing LOL-Cat in 480x192?

          1. John Smith 19 Gold badge
            Happy

            Re: You need to be thinking *phototypesetter,* not Linotype if you want to go that way.

            "You just know that this thing is going to be so expensive that it will be installed in the managing director's office, where he will use it to train his philodendron, rather than actually having to work very often."

            Actually I'm thinking more around a typewriter by 1890. A common (if expensive) piece of "office equipment."

            "Baudot was a Johnny come lately in telegraphy (not to mention that what we know as Baudot is actually Murray's later version),"

            That's tougher. My real core requirement was "No electricity."

            "I'm beginning to think that a "practical" machine would benefit from your notion of phototypesetter, where only one master copy of each glyph is present, but with a pin-screen to store the image."

            Possible. A spinning (or 2 in parallel for upper and lower case) with glyphs that could be pushed out would work provided you can work out how to push the individual glyph into the back of the pin board and index it to the next location.

            "Of course, one of my favorite notions would stuff clear or opaque marbles (small as possible) into the bottom of a frame with translucent front and back, and perhaps an optically dense fluid. As each row of these "pixels" is "rendered", it is shoved up in the frame, with older lines emerging from the top. If the opaque "marble" were iron, the two types could be sorted magnetically. for reuse Of course, even a 24x80 screen of 5x7 (in a 6x8 cell) characters would need 92160 of each (actually fewer of the clear ones, unless we do inverse video) ."

            Not quite sure I get you but the idea of (micro) sized beads, 1/2 black, 1/2 white is actually the core of some eInk displays IIRC from 3M. The correct term is "Electrophoretic."

            eInk breaks the "No electricity" rule, but magneto optic (in the loose sense of the term) would be OK.

  67. Anonymous Coward
    Anonymous Coward

    Re. displays

    A more high tech variant is a beast I call a "PEL" Programmable EL".

    Its based on a combination of copper sulphide memristors and electroluminescent sheet where each EL element has a copper sulphide memristor behind it.

    By changing it from high to low resistance the brightness of each pixel will change allowing for storage.

    Have built the separate parts and connected them, it does work but the finicky part is making the board as to get a pure CuS layer requires precise processing and the EL paint is water based...

    The advantage is multiple colours if SrS:Eu or a variant is used as a phosphor and non volatility.

    Once programmed it can be used as a custom display for magazines etc as the layers are somewhat flexible.

    1. andre 2

      Re: Re. displays

      http://sparkbangbuzz.com/memristor/memristor.htm

      This guy is a genius, I learned about this effect from him and he was the one who originally suggested hooking up a memristor to an EL panel as a brightness control.

  68. John Smith 19 Gold badge
    Meh

    The *only* absolute rule is *no* electricity

    BTW the idea of an EL with a memory device (or more accurately a layer) was discussed in a big book on photoconductivity I cam across some years ago.

    The system used (IIRC) an EL panel backed by some kind of Ferroelectric storage device.

POST COMMENT House rules

Not a member of The Register? Create a new account here.

  • Enter your comment

  • Add an icon

Anonymous cowards cannot choose their icon