Become a fan of Slashdot on Facebook


Forgot your password?
Graphics Software

Printed Embedded Data GUIs 111

n7lyg writes: "Xerox PARC has come up with a way to embed data in printed images that involves using something called DataGlyphs. A DataGlyph is essentially an oblong pixel that takes the value of zero or one depending on whether it is printed angled to the left or right. Printed at sufficiently fine resolution, this is no different from ordinary offset printing effects using circular pixels, but when scanned by a computer allows recovery of arbitrary data embedded in the images or text of the printed page. An article in this month's IEEE Computer contains a lot of interesting applications of this technology, including a system to allow teachers to create printed tests and lab assignments with embedded DataGlyphs to allow automatic generation of graded and annotated results." I think we've done an article on this before, but I don't see it in the archives...
This discussion has been archived. No new comments can be posted.

Printed Embedded Data GUIs

Comments Filter:
  • by Anonymous Coward
    including a system to allow teachers to create prinnted tests and lab assignments with embedded DataGlyphs to allow automatic generation of graded and annotated results

    This reminds me of those childrens books which came with a small 'watermark' reading light pen that could read the right answers off the page. You would answer the question by filling A, B, C or D in the box and then you would wave the pen over the answer box and it would read the correct answer, thus you could test yourself. It never really cought on and the book/pen sets were quite expensive, but this seems to be targetted (at least partially) at this market. It could also have an application in barcodes that are more pleasant to the eye.

  • by Anonymous Coward
    Wait till the geek kids figure out how to read the answers from the sucker! :)
  • by Anonymous Coward
    YO! does anyone know what the fuck is up with chilliware?
    i bought all 3 of their programs, and now they are just GONE!!!
    no website, no phone, no email!1!!
  • by Wansu ( 846 ) on Wednesday April 04, 2001 @06:54PM (#314141)
    The last couple of years, it's been interesting to watch Xerox's downward spiral. Maybe what's left of PARC will be completely autonomous once the rest of Xerox finally goes around the bowl and down the hole. It's probably too late for PARC. It's a shame. Lots of good ideas came from there. Xerox management was too myopic to do anything with it.
  • Actually "printed embedded data on GUI things" have a name. They're called UPC symbols. They store information on food packaging. You scan them with a light pen. How many generations need to be taught the same technology we've had since 1970?
  • It works better if nobody knows you are using it. If people do find out, you're cooked to a large degree. It'd be much easier to just have a bar code with an encrypted version of the results. Given how little data are in the answers to a test, the encryption algorithm could effectively be the result of XOR'ing a hash of the passphrase. ;-)
  • Depending on the paper and ink used, this could be good way to archive long time storage. With a magnifying glass and sufficient patience, it would be human readable too.

    However, I think your amount per page is off. If a 3x3 pixel block is used, shouldn't there be 4200 lines of data on the page? With 400 bytes/per line, that works out to about 1.6M. Using a 2x2 pixel block, there would be 6300 lines, and about 3.6M. Even if a blank line is put between each data line, that could still be a significant amount of data.

  • by ksheff ( 2406 ) on Thursday April 05, 2001 @09:32AM (#314145) Homepage

    I didn't really see how this would aid in automatic grading of tests or homework. If the kids have to use a computer & printer to generate the results, then why not transfer it electronically? If they are to use the good ole No. 2 pencil, why not just use 'fill in the dot' type forms?

  • And they will have your fingerprints too. And if you lick a stamp on it, some of your DNA.
  • Isn't it old news? I remember reading about something very similar from Xerox about 5 years ago. What's new?
  • Okay, so it's like a barcode, or like those little static-pattern square things, or any other form of scannable printable data. Maybe it's really really high capacity printed data....

    But umm, where'd the "embedded" and "GUI" come from? Were those just buzzword burps as you were trying to type "Printable Data"?

  • DataGlyphs are flexible in that they are robust, have a high data density, and can be cosmetically pleasing. Traditionally people consider (some) barcodes good for the former and (some) watermarks good for the latter.
  • On a more serious note, how does Xerox deal with large white spaces? How do you keep the rows lined up? There doesn't appear to be much room for error. What happens when someone spills coffee (or even water) on the document? How much redundency is built in?

    Actually, we're quite robust. Read the article for details about synchronization in DataGlyphs. Redundancy is adjustable, and I can personally attest that DataGlyphs have survived my coffee spill. Haven't tried Guiness yet.

  • For starters, "Printed Embedded Data Graphical
    User Interfaces." Read the article.
  • by jab ( 9153 ) on Wednesday April 04, 2001 @07:36PM (#314152) Homepage
    I work at Xerox PARC on this project, and am very happy to see all the interest. Have fun! I'm willing to take a stab at answering any questions posted under this comment.
  • Here I go with my conspiracy theories, but I'm sure that the NSA and CIA have been doing this kind of stuff for years.. but i'm sure the data isn't just hidden, but also encrypted :)

    What is nice about this is that instead of watermarking, the pixels are shaped differently. Watermarking doesn't scan right, but this does.. I don't really see how this can be considered new at all.. and certainly not patentable! But you never know what they will patent these days :(

    Obvious concept with very little to no more realworld applications then a barcode has, only difference is that the untrained eye may not notice it's existance. What is good about this is, instead of putting an obvious and easily readable barcode on an identification card.. one can print the id card with glyphs, this isn't secure.. but it would prevent others from being able to read it without scanning it into a computer. (bar codes can be easily read by the trained eye). The identification card issue is brought up because there have been schools putting social security numbers on identification cards in barcode, or worse.. plain text..

    speaking of plaintext info on identication cards, I wish they would either remove my social security number from my University id... or put encode it in a slightly more secure fashion!

  • by Christopher Thomas ( 11717 ) on Wednesday April 04, 2001 @07:46PM (#314154)
    It really got me to thinking about the implications of DataGlyphs. Suppose each of those survey forms is "pre-numbered" using DataGlyphs. Now they just need to tie the fact that you took the "blank" form from the top of the stack with your identity -- and poof the survey is no longer anonymous.

    A lower-tech solution is to just write on a serial number in UV-fluorescent transparent ink. This stuff's been around for years.

    Or laser-print it at stupid-DPI in flyspeck 3.

    Or just remember the order in which people go into the room and surveys go into the slot.

    Now, unless you're carding the people for other purposes or otherwise already have their ID, this will avail you nothing...
  • ...this is no different from ordinary offset printing effects using circular pixels...

    when printed, the dots that make up an image are typically referred to as.. um.. dots.

    pixels are picture elements, usually on some kind of computer screen. dots of ink are just dots.

    just so's you know...
  • Merge Technologies had a license to use Xerox's "Data Glyphs" several years ago (at least). They use them to tag medical images.

    Their only advantage over other 2-D barcodes was supposed to be how they can masquerade as innocent grey background. That way you don't have to mess up your product's graphic layout with an unsightly bar-code symbol.

    It must have sounded really "Hi-Tech" back then but checkers have enough trouble finding UPC symbols as it is, without hiding the things on purpose.

    Whoopee... not.

  • Actually, paper will outlast most disks and tapes for storage purposes. Good paper and ink will last hundreds to THOUSANDS of years, while disks and tapes might make it to 100.

    In this case, it is good for data density. Encode information two ways. Great for encryption, embedded data (like XML for paper) and steganography.
    Charles E. Hill
  • It's a high-density barcode. That's it. Move along. Nothing to see here.
  • Dude, as someone who has worked in the laser-print and mail-processing industry, let me tell you that people don't see normal barcodes, either. Or micronumbers, or OMR marks, or any of the technology that's added to the printed document to enable it to be processed / tracked automatically. If it's not human readable and obvious, your average person just doesn't process it.

    I always notice this stuff, because I'm attuned to it. None of my non-industry friends do. If I ever happen to point it out, the reaction is always "Hey, wow, I never noticed that before". It's not an intelligence thing, either. I've got some smart friends, but that sort of stuff is just filtered out of the image by your brain if it doesn't mean anything to you.
  • That might be true of Xerox, but it sure isn't for a number of spin-offs, including Fuji Xerox []. The broken X will survive in one form or another.
  • No, it's nothing like a Turing machine. A turning machine is a state machine which can read & write to a long tape of data. A better description is here [].
  • other 2d bar codes are hideously ugly. Notice that
    the use of 45 degree bars gives a uniform
    grayscale look, regardless of runs of zeros and ones in the next.
  • Your idea sounds a bit like the Cauzin SoftStrip that made a splash, then tanked in the mid 80s.
    I still own one.
  • Ok, so they've created a pretty small barcode like scheme. So? Now you just have to be aware that bar codes can have this look (\/\/\/\/\///\\\/\/\/) rather than the usual vertical lines. Granted, they can be made smaller than Code39 or UPC codes, but they're still just bar codes.
  • ala the companion technique, allowing for emailable,
    but not printable, subrosa material. see

    for further elucidation. embedding scientology
    "scripture" within slashdot archives also comes
    to mind...
  • But the whole point of this is that it doesn't destroy the image.

    If you start fiddling with the colour, then the image will be corrupted by the data.
  • What if the individual bills had a unique number imprinted with this method, like a public key. Then the copier would only be able to duplicate the original code and could be picked up when multiple codes appeared too many times (kinda like Ultima Online).

    "I'll take the red pill, no, blue. AAAHHHHHHHHHHHHHHHH........"

  • From what I understand, this sounds somewhat like a project that I had dreamed of around 8 years ago. Basically, my idea was to eventually create a system that archived my data on a sheet of paper to be scanned back in for recovery. (Regular home printers and scanners would do just fine.)

    The initial testing stage was to simply compress and then uuencode (possibly with some adjustments to the character set) any random file, print it out, and then OCR it back in and do the deuuencoding. This can theoretically be done with common tools available for years now, but 1) I've never owned a scanner in my life 2) Any large amount of data would take up lots of paper, and would require a lot of work as well.

    Eventually, I would move further towards my own designs: namely inventing an encoding scheme that would allow a lot of information to be saved on the paper, perhaps up to or exceeding a megabyte. This would, of course, also require new software to decipher the data that was encoded onto the paper.

    I used to dream of a world where software developers shipped the latest version of their programs on a few sheets of paper rather than a floppy disk or CD-ROM...

    (It also would occur to me how funny and ridiculous it would be to store one's encoded pr0n archives in a regular binder...)

  • CDROMs basically use the same technology. Optically scanning a flat surface. Sorry, but there doesn't seem to be any need to reinvent the wheel... or disc as it may be.

    That might be the case, but then you can make the same jump of logic by saying that CD-ROMs are "basically" the same as a floppy disk, because they store data on a disc, reading in a circular pattern, and then you can say a floppy disk is "basically" the same as magnetic core memory because they both work off storing bits magnetically. But when you step back and look at the whole picture, you begin to realize that data printed to a sheet of paper really has nothing to do with magnetic core memory, other than they both store data in one way or another.

    What I'm looking for is a storage medium that doesn't decay with time. Paper burns and yellows, floppies and hard drives lose their can get erased. Maybe CDs are ok, but I don't know what the lifespan is.

    My idea of archiving information on paper was not borne from the intention of solving any of those problems. In fact, I can only think of two primary advantages to this method over that of a floppy disk.

    1) Ability to store, transmit, or copy the data as easily as any document (examples: store it in a binder, fax it to someone, or xerox it)

    2) Ability to hide information in a format that most people wouldn't suspect.

  • I've never heard of this Cauzin SoftStrip... perhaps you might provide a picture or link?
  • Imagine the scenario - "send in this coupon and you will receive a free widget". You the consumer look at it, and say, "Hey, I don't have to give an account number or anything! They don't have any codes or anything! I could even give my work address, or a fake name, and they won't be able to tie it back to me!" Wrong, we have your data codes embedded IN OUR OWN LOGO. We know exactly who you are, we know what you respond to, and you better believe, you will get a ton more of this stuff if you do respond. We've been doing this for years... For real.
  • How is this different than having a serial number on the paper? They still have to do something external for them to tie it back to you. I would generally consider either one anonymous enough, unless they ask for a credit card or name or something, of course.
  • Looks like we have yet another way to distribute the DeCSS source code
  • But remember, you have to have the equipment to be able to print something like that. If you limit the access to that equipment, you limit the problem.
  • i suppose this could be used in conjunction with anti-forgery devices such as money-counterfeiting (sp?) and document verification as well. "it says its form the desk of the president" well, scan it for the digital signature" presto. i suppose with complex bits you could make it difficult to copy a digital signature embedded in paper. especially if its a rotating code of some kind. i see all sorts of applications for this beyond adding simple data. i don't see it working with art though as the art (original)would be damaged by the addition of the code.
  • I will be the first to re take the SAT test with the answers inlayed into the question form. Dont mind those Coke Bottle Bottem Glasses, I need them to uhh .. read.
  • There are already watermarking technologies for hiding data in images, and if you just want to put machine-readable data on paper, there are 2D bar codes. The UNIBAR web site has a demo online bar code generator [], so you can see what various types of bar codes look like. (Obligatory Slashdot Linux remark: UNIBAR's product runs under Linux.)
  • other 2d bar codes are hideously ugly.

    That's a good point. And it's one that's been brought up in connection with the USPS's "electronic postage" system, which encodes source, destination, and postage info in an ugly bar code block. Maybe this would work for that application. The Xerox scheme, though, requires high-resolution scanning, while the postal system has to read moving envelopes at very high speed.

  • If instead of using it as only 1 or 0, why not have the color of the pixel represent a value too and allow for more combinations. Could store more data through different colors representing different meanings
  • Can you fax with Napster? :-) [mailto]
  • You can place meaningful data inside the pixels of a printed page? Is that what this story is about? That's quite a logical and cool way of data distribution. I would imagine that the technology would be fairly quick to find uses, one being as mention previously, anti forgery measures. I would imagine such technology however would be regulated fiarly quickly, especially devices capable of decoding, by governemts wanting to keep the technology for themselves, or cripple it so it's not really of huge use to us. Got to keep those terrorists from sending encoded messages now, dont we?

    Now if we were to go scan existing pictures or print for code, what type of meaningful information would there be in... say a typical picture in Playboy?
  • Sure, but if you use say two colors _and_ two different rotations of the pixel, you can encode two bits with one pixel. Or, if you prefer, the pixel can have more than two states thus creating some sort of quasi-quantnum data storage.

    That was a rather narrow-minded statement of you...

  • so this is just a printable watermark?


  • I never worked on DataGlyphs, but I was at PARC for some of the time that they were being developed. They've been around for quite a while and are actually in use in the real world. Those of you with insurance from State Farm probably get newsletters; each newsletter comes with a tear-off card that says "if you're not interested in receiving this newsletter, just drop this card in the mail". If you look at that card, there's what looks like a gray rectangle, but if you look closely you can see this is made up of a bunch of tiny '/' and '\' marks. That's a DataGlyph block. When State Farm gets one of those cards back, they just drop it onto a scanner, the glyph gets read out, and you get unsubscribed.

    I happened to be standing around when the inventor of DataGlyphs came in with his copy of the State Farm newsletter - he seemed quite pleased to be receiving mail using his invention.

  • There's a little more to it than that: it's an inconspicuous error-tolerant high-density barcode. Inconspicuous means that you can embed data in what looks like just a background gray stipple. Error-tolerant means that you can scan it at low resolutions, at any angle, you can put it through lousy copying or fax it, you can even cut bits of it off or obscure them, and at the end you can still get the original data back. There are limits, of course, but the glyphs are pretty robust. At one point, one team working with DataGlyphs got a T-shirt made up with a glyph block printed on it. The individual marks were, of course, much larger than they would be on paper, and the glyph block had a number of other graphical elements (drawings etc) splattered on top of it. One of the team members, for grins, put the T-shirt on a scanner and fed the result into the glyph decoder - and despite the scale being way different from normal, and the missing glyphs (obscured by the stuff drawn over them), it correctly read back the encoded message.

    Also, DataGlyphs are a couple of orders of magnitude higher density than standard barcodes. Two orders of magnitude is enough to make a qualitative difference in how something is used: applications that could not even be considered before are now easy, and so the way in which something is used can change drastically.

  • I don't have any link or pictures, but I do remember having it for my Apple ][+

    Basically it was just a barcode reader and software that could print barcodes. It worked well on my dot-matrix printer. Some of the hobby magazines printed their BASIC source listings in the barcode format in the margin as well as the human readable form. Like the previous poster said, it made a splash then went away. I thought that the durability and storage density made it an attractive alternative to floppy disks. Especially compared to the old Apple DOS floppy format. Archives were cheap, easy to reproduce, and could (if using acid free paper and good quality ink) last quite a long time.

    But like most things, it's a matter of popularity, not technical quality. It just never caught on.

    Of course nowadays, I just burn a CD when I want to archive or snail-mail somebody a binary. Still, it was a fun toy.

  • >It's a high-density barcode. That's it. Move along. Nothing to see here.

    But that is precisely the great NEW feature of this technology: There's nothing to see! It is invisible to the unaided eye.
    The data is embedded within any full colour image. This sets it totally apart from barcodes (Code3of9, etc), which really stand out and would be somewhat distracting.

    The whole idea is that the data is there if you want to get at it, but if you don't want to know such details, then you don't even see them.

    Also, there are 2D "barcodes" that are in use in manufacturing environments, but they are unsuitable for publication applications for the same reason that barcodes are: they rely on changing the intensity of the area being printed on.


  • I think we've done an article on this before, but I don't see it in the archives...

    Let me be the first to congratulate you with a hearty handshake for being the first Slashdot editor to check the archives before posting! <grin>

  • I have someplace around on an old harddrive a win app that can save a doc into an arbitrary pattern of dots on a sheet of paper. The idea is that with a sufficient high rez scanner you couild recover your binary data from paper backups. It's a win 3.1 app, but I think the folks are still around someplace.

    The other thing I think I have archived achieves hiding data in an image file by subtly shifting the bits. Basically similar to a water mark, but for hiding binary data.

    So while this is interesting, I'm hoping that they don't try for a patent or anything.

    Check out the Vinny the Vampire [] comic strip

  • Jeff works on the DataGlyph technology. I work on applications of it. See [] for my application, which lets you put a sheet with DataGlyphs on your document, drop it in a copier, and get out a paper icon for the document (stored on a WebDAV or other network server). You put the paper icon back in, and press copy to get a copy, type an e-mail address to e-mail it, etc. It's like an electronic version of the paper document, but on paper. We call it a "Document Token". (There has been previous SlashDot discussion of Document Tokens and other applications of DataGlyphs, but I can't find it either).

    We're also very interested in open standards, and participate in a variety of standards organizations on this and related technologys. For example I'm on the W3C XForms [] committee, designing the next generation of web forms, and paper is one of the new "devices" that is being targeted (along with voice, pdas, phones, etc.). Check it out and send your comments to the www-forms mailing list! As we said when XForms was launched []:

    For the Web to become a truly ubiquitous computing interface, it must move beyond the desktop. XForms and other W3C open standards will make it easy to create and to use rich, interactive Web documents and services on a wide variety of user interfaces -- graphical, voice, and paper.
  • I don't fully understand how this is such a great idea.
    Xerox DataGlyphs provide a solution for embedding digital data in a printed document.

    Using the word 'solution' implies that there was a problem, and now it is solved. I can't remember ever thinking, "Damn, I wish I knew how to imbed binary information hidden in some graphic". Mainly because I've never thought that.

    Obviously now that we have found a 'soultion' we can go out and find a bunch of problems to solve. Don't get me wrong, It's at least a 8 on the cool-meter (Ascii art [] being a 9), I just don't see it as a necessity.

    Not to mention the fact that it isn't all that revolutionary, anyone could have done this (in a much rougher way) years ago. It's 1's and 0's, for crying out loud, we've had those oh, at least a few years now. Ordering and weighting them isn't making me buy Xerox stock.

  • Why would the government regulate this? This is just a way to present data on a physical medium, nothing more. It's not fundamentally more encrypted than any other medium - what you do with the data is what matters.

    If this becomes commonplace, then machines that can read it become commonplace and the government and anyone else just treats it like any other data medium.

    Besides, how could they regulate it? All you'd need to use it is a very high-resolution printer.

  • Ditto. I finally got the SDK for this years ago after jumping through hoops, forking over money and signing an NDA. After all that, it just wasn't worth my time to develop anything, because they were too closed and inflexible about the licensing. It drove me nuts.

    You would think there would be some limit to the number of times a single company could develop a really cool technology, only to waste it. If Xerox had played their cards right, every scanner, fax, copier, and printer manufacturer would be paying them royalties by now. DataGlyphs would be everywhere.

    I first heard about DataGlyphs in this Wired [] story.

  • You don't have a PDF reader? Go read the article.
  • My 2 favorite applications are:
    • Lossless copying - the document contains a URL with its source, so 10,000 generations of copying produce the exact same quality as the original.
    • Document from a page - every page of the document contains the URL for the source, so having a page means being able to reproduce the entire document. This plays very well with version control software.
    Of course to make these really work, you have to build it into copiers and printers and put them on the same network, so they can talk to each other. If only someone who makes printers and copiers had developed this ten years ago...
  • Anybody with the idea, and good background background in error detecting and correcting codes, could have done this. While I consider this at least border-line to "sufficiently obvious", how long until this drops into public domain and anyone can use it? Information on the relevant laws and/or the exact dates involved are appreciated.
  • Here's how I think DataGlyphs should be used:
    • Embed a URL pointing to the document source, in every page of the document when it comes out of the printer or copier.
    • When you copy the document, it can read the source URL, and take advantage of it, to produce a perfect copy, the document from a part of the document, or just what mark-up has been added to the document (the stuff you wrote on it.)
    Will this ever happen? What about if I restrict myself to all Xerox products? When or why not? How about a personal copy, so that I can do this at home, with stuff I print and scan?

    With the infrastructure above in place, now add:

    • This is all tied into a versioning system, so I can take a page of a document, and get the latest version of the document (or a previous version, etc...).
    • When I send a fax, the fax machine reads the DataGlyph URL and takes advantage of it to separate template from markup. This is useless if a standard fax machine is receiving, but an enhanced fax machine could take advantage of the template source to make the fax higher quality. Of course, this is just the tip of the iceberg of how fax machines could be improved to take advantage of the Internet, and most of the improvements have absolutely nothing to do with DataGlyphs.
    Will this ever happen? What about if I restrict myself to all Xerox products? When or why not?
  • by mgkimsal2 ( 200677 ) on Wednesday April 04, 2001 @07:43PM (#314198) Homepage was doing this a while ago, but instead of slashes they used standard squares. Neat idea all the same. Not sure of practical use for most places as a real backup system, and too cumbersome for transporting data, imo. But neat nonetheless. Perhaps people could use this stuff in stationary, and embed company/personal info in the letterhead, scannable by their customers and clients.
  • Slashdot covered a story about Printable Computers [] it coveres printing out motherboards, etc.. not just memmory storage.
  • IIRC, you can scan from any direction.

    Furthermore, it is immune to part of the paper being damaged.

    From what I read about this a year ago, you cold print data in a fine grained "gray" pattern on the background of the paper. The stuff printed on the paper, such as a document, 1040 Long Form, etc. would be what you're paying attention to. Not the pattern on the "safety paper".
  • The big thing I read into it a year ago when it was on Slashdot, was that it was a way to put data onto paper without the general population knowing about it.

    When I read the Xerox article last year it seemed to work like this. The data is encoded into in a background pattern on the paper. Sort of like the safety paper your personal checks are printed on. If you look closely, it's not a solid color, but white paper with colored wavy lines. The data glyphs is similar. But instead of wavy lines is's a pattern of slashes and backslashes that the eye just picks up as a uniform color pattern. The data is encoded all over the page. It is immune to part of the paper being torn, spindled, stapled, printed over, vomited on, etc.
  • But what if they snail mailed you the survey?

    What if they e-mailed it to you as PDF?

    It doesnt' need to have your identity on the paper, just a serial number that ties it back.
  • Hey cool idea.

    When you vote, you usually have to sign something that you voted. Now if you can just match up the order of signatures in the notebook with the order the blank ballots were removed from the stack, you can identify who voted for the bad guy and go re-educate them about how gravely mistaken they are.
  • Maybe they mailed you the survey?

    Maybe you took it as part of a study, and they pay you $20. Maybe they can tie the order you took the forms off the blank stack to who you are?

    Maybe you take an anonymous survey in a mall, but you have to sign a seperate "guest book", allowing them to tie the card you took with the order of your signature. But you then go into a small room with other people filling out the card, and when finished, drop it in the box as you exit.

    There are other possibilities.
  • serial number in UV-fluorescent transparent ink

    Too likely to be discovered. Most people know about secret inks. Not most people know about DataGlyphs. Furthermore, even of people who know, most wouldn't look for it. The way I read about it (last year) it disappears into the background color of the paper. Well, not really, but you don't notice it any more than the "color" of the safety paper your personal checks are printed on.

    laser-print it at stupid-DPI in flyspeck 3

    But now, you're starting to get into a technology race. Sorta like the government mint is in a technology race with counterfiters. One of the benefits of DataGlyphs was that it can be printed by common printers and is very un-obvious.

    remember the order in which people go into the room and surveys go into the slot.

    It's much more practical to identify people when they take the blank page off the "blank" stack. Even though the forms are pre-numbered, the numbering is invisible. People in a room filling out surveys dropping them into a slotted box as they finish in random order doesn't give you much of a way to identify surveys with individuals.
  • But have you ever looked at a page with DataGlyphs? They virtually disappear. You wouldn't even notice them if you weren't looking for them.
  • Are we willing to give up our freedom?

    The majority doesn't seem to mind so far.
  • Hey You!

    Yeah you!

    Would you like to participate in our anonymous survey? Yeah, it's completely anonymous. You don't write down any personally identifiable information. Just take this printed page and answer the questions on it. Then anonymously drop it through the slot in the box over there.

    Oh, don't worry that it asks lots of deeply personal questions. After all, this is an anonymous survey. It cannot be tied back to you.

    BTW, yes, Slashdot did have a story on this about a year ago. It really got me to thinking about the implications of DataGlyphs. Suppose each of those survey forms is "pre-numbered" using DataGlyphs. Now they just need to tie the fact that you took the "blank" form from the top of the stack with your identity -- and poof the survey is no longer anonymous.
  • by DickBreath ( 207180 ) on Wednesday April 04, 2001 @06:08PM (#314209) Homepage
    What are they going to do about someone sending one time padded messages in a sheet of text?

    The solution is so simple it should be obvious. Why do you even ask such silly questions. Are you dense?

    SImply require all printed material to be government approved. Problem solved. Now wasn't that easy? Think of the benefits in reduced terrorism.
  • but isn't this similar in some aspects to a Turing machine?

    Well, yeah, but every sort of way to encode data in a linear and discrete way is similar to a turing machine tape.

    I don't think that there are any world shaking implications of being able to write executable code on a sheet of paper. :)

  • What I'm looking for is a storage medium that doesn't decay with time.

    I submitted this to slashdot about 3000 years ago, and they rejected it.

    The jist of it was you take this chisel, and a large smooth flat rock, and then another rock that fits nicely in your hand.

    OK, stay with me here, it gets a little technical.

    You create a sort of verbal encoding scheme, a way to represent spoken words in a visual way, then once you assign symbols to words, you take the "hand-rock", and strike the rear of the chisel, while the chisel is resting on the other, flat, rock.

    I created a prototype of this system, it seems that it works best when you use a type of rock that doesn't fracture easily, but yet is still pretty hard.

    Once we get the bugs worked out, I can release more details.

  • by GigsVT ( 208848 ) on Wednesday April 04, 2001 @10:48PM (#314212) Journal
    Now, unless you're carding the people for other purposes or otherwise already have their ID, this will avail you nothing...

    Well, all your ideas have valid rebuttals.

    1. UV Ink

    Risk of discovery. Once one person discovers it, because say, they see it in the sunlight, which contains enough UV to see UV inks on light paper, then everyone will know, because that person probably will submit it to slashdot.

    2. Tiny Fonts

    Same problem, someone might notice and blow your cover.

    3. Order of people in a room

    Well that works OK if you have the people there, but most surveys are mail-in.

    4. Unless you are carding them it doesn't matter.

    Well this is my scenario:

    Company mails out surveys with the person's identity printed in this new pixel format in a graphic on the survey or something. You could print it at the same time you print the envelopes, a simple mail merge.

    There is little chance of discovery by the target people, and if they are mailing it back in one of those "business reply" envelopes with no return address, you could probably convince them it is anonymous.

    Sure, this isn't really a new technology per se, there have been ways to covertly embed data in images for a while, but this is probably the first one that can reliably be printed out and rescanned with no data loss.

  • and it's been around for many years. A cool technology but not exactly news. Type "dataglyph" into Google to find out more.
  • New!! Analog Data Glyphs, from Xerox. They look like this when printed:






    When sequenced like such, they can make pretty progress indicators. When printed at sufficiently fine resolutions, computers can not identify them. (Unless anti-aliased using XFree86's Render extension.) When printed at sufficiently coarse resolutions, they use a lot of paper. However, your printer will need to be able to print in grey.

    Prepare to see the DeCSS code hidden within your next Linux-related document, or hacker magazine.

    If you scan it upside down, are all the ones zeros, and vice versa?

  • Hmmm, now that I think of it, you have got a good point. If the Fed really wanted to make counterfeiting as difficult as possible, use these dataglyphs, ran through a steganography algorithm and set up with a public/private key which can be confirmed genuine with a quick scan. Add that to the existing technologies on new bills, like the different colored embedded strips in the bills visible under blacklight. (I've seen 5's with blue, a 10 with a yellow, a 20 with a green, and a 50 with a yellow one)

    on the other hand, I wonder how that would work on beat-up-run-through-the-washer-lying-on-the-ground -etc.-etc. bills? I don't think MNP-4 is going to cut it :P

  • You mean []?
  • This is actually a conspiracy. AOL uses these watermarks in the printed cardboard packaging of the CDs they mail out. It's an infallable method of disseminating the data reported by Carnivore back to the NSA (who can happily declare that Carnivore makes no trasmissions of your e-mail or activities to them over the 'net.) Think about matter what happens, the NSA will always be able to go about their data-gathering with a little help from those AOL cd's which they can pick up anywhere and everywhere.
  • Seems to me this would be a perfect way to implement Optical Character Recognition. No more of this "99% Accuracy" crap. You could even store the font size, color, etc... in these pixels. I wonder if we could somehow take this one step further and make seemingly inert pictures into large pieces of encrypted data. It'd require an enormous number of pixel bits. Theres some old axiom about hiding something in plain sight I wish I could remember.

    "// this is the most hacked, evil, bastardized thing I've ever seen. kjb"

  • CDROMs basically use the same technology. Optically scanning a flat surface. Sorry, but there doesn't seem to be any need to reinvent the wheel... or disc as it may be.

    What I'm looking for is a storage medium that doesn't decay with time. Paper burns and yellows, floppies and hard drives lose their can get erased. Maybe CDs are ok, but I don't know what the lifespan is.

  • Ins't this tech similar (just larger scale) then the stuff embedded in some of the wired magazine ads ?. (Something beginning with D .. I don't have a wired mag here =)) ?
  • Remember the good ole days when computer magazines for the Commodore 64 used to have type-in games and other programs? IIRC, someone once proposed a type of bar & dot code that would be printed along the edges of the magazine and could be read in via a scanner thingy not unlike the Cue Cat (only larger due to mid 80's technology). And since the early 90's there have been various programs that allow you to print a dense dot pattern similar to UPS tracking labels, using an ordinary ink-jet or laser printer, and could be read in using a standard hand scanner. It was proposed that magazines could be distributed with programs without the need for a separate disk (of course, this was back in the days before multi-megabyte games and applications).
  • What would be the purpose? Data on a computer is stored in binary (0,1) format... leaving it in 0,1 format on the paper would allow for it to be read without having to convert it into binary form for the computer to read it.
  • This link to the "article" pointed to a PDF. WTF?
  • CD-Rom I believe has a shelf life of around 10 years. Of course that assumes that you will have a drive and software capable of reading it then. There is another technology, I believe called HD-ROM but I could be wrong, which actually laser-etches a disc made of nickel. It is intended for really long term storage but I don't remember the number of years
  • Or a person could create a font that is all micro-glyphs. Encode "A" into the ASCII or UNICODE binary and the spaces also. Cut the linefeeds to a pixel and the typespace to a pixel. Take your document and display in the glyph-font before printing. Or you could just display the entire binary document in glyph-font before printing (advantage is no unwanted carriage returns).

    Ta-da! Instant encoding (without error correction), but the problem lies in OCR reading off the printed page. I think with quality printing, a good quality scanner and a conversion program, you could easily store lots of data in a super-cheap manner.

    I figure a standard 8.5" x 11" (about 8" x 10.5" with margins) and a printer putting out 1200 DPI.

    My figures are using 3 pixels per glyph bit in an 8-bit byte gives 50 bytes per inch and 400 bytes/line or with 2 pixels/glyph-bit = 75 bytes/inch or 600 bytes/line. That gives for 3 pixels/glyph-bit a page storage of 4200 bytes/page or for 2 pixels/glyph-bit 6300 bytes/page. Thats a whole 4 K to 6 K per page! I'm going to write a BASIC program and back up my 40 GB harddrive right away (if we tack on a full 4-color range we get 7 times the storage - whoa. I'm throwing out my floppies this instant!)
  • Compute! did that for a while, I remember seeing some of the older issues at the local public library. I don't remember if it was just a single page or multiple pages. I'm going to have to take another look and check the dates.
  • by Gruneun ( 261463 ) on Wednesday April 04, 2001 @07:37PM (#314227)
    interesting applications of this technology, including a system to allow teachers to create prinnted tests and lab assignments with embedded DataGlyphs to allow automatic generation of graded and annotated results

    There's a reason that most encryption and computer security measures are created in a way to suggest that you never, ever trust the client.

    Bonus points to the kid who uses his Palm Pilot or TI-82 to read the answers embedded in his homework. I know I would have spent hours to outsmart the teacher's technology that otherwise could have been spent studying or doing the actual work.

    Interesting technology, though...
  • Would that be possible? To print the dataglyph OF a dataglyph? Some dude decodes it and gets back what he started with, and he's like, "Huh? Maybe I gotta decode it again." So he does, and then gets back what he started with yet again. And again and again and again...

    And again and again and again...

    Dude! Let's do it! For great justice!!

    >> me vs. the end of the world []

  • Intacta has a similiar technology. Of course, so does UPS.

  • Intacta Technologies makes a similar product. It allows any type of multimedia information to be delivered on a piece of paper. The density is high enough for a single sheet of paper to encode the text of more than 50 pages of text. Any computer with a scanner and Intacta software can read the code, which can contain text, images, music or animation. []

  • Someone with more experience please correct me if I'm wrong, but isn't this similar in some aspects to a Turing machine? And considering how flexible one of those is.. this could end up being some very useful technology. (Is technology the right word?)
  • Has there been another posting of this a while ago? I've known about this for about two or three years now, and have seen examples of it. It's really kinda neat, actually. You can have a shaded graphic at the top of a page that contains the data needed to convert the rest of the page into HTML when scanned in. Obviously, other implications of this will arise, and we may not like it so much. Use your imagination.

    Hmm... I just thought of an interesting idea... You encode a MIDI file, or an MP3 on a piece of paper and it plays back when scanned in. Heh. What a way to distribute music, with a photocopier! Oh man, what will I think of next?
  • As I read this article, it seemed that I had heard of this sort of thing before. Here's a link:,4586,25 93 651,00.html

    I grant that this isn't embedded in a "gui", but there's not much gui-looking about a Xerox "X" that looks more like ASCII art.

  • Is to uniquely mark up the paper that is used for 'anonymous' surveys. When you send back your response, it isn't anonymous anymore. They know who you are.

  • by BlueboyX ( 322884 ) on Wednesday April 04, 2001 @05:33PM (#314241)
    The US government is having a fit right now with the ease at which coded messages can be sent (supposedly by evil people everywhere) via the internet. What are they going to do about someone sending one time padded messages in a sheet of text? A love letter could really be a terrorist timeline. A letter from a soldier to his family could actually be a spy message intended for the 'other side.'

    The US gov has already limited net encryption to 128 bits; will they step in and try to regulate this? I would think so...
  • by porter235 ( 413926 ) on Wednesday April 04, 2001 @06:42PM (#314245)
    Check out... []

    Some seem to have got the wrong idea. This is a way to add extra information into a printed page. For instance (taken from the pdf above) a physical blueprint could have extra information imbedded in it such as the pluming layout and the wiring information etc. Then when viewed under a Glyph-O-Scope (magic magnifier type device)the chosen extra information would be superimposed (aligned properly of course) and would move with the page.

    It is not an encryption scheme. If you want to encrypt, you would do so before you encode.

    If you try and add more depth to the encoding by adding colour to the scheme, you destroy the ability to have the message embedded in a full colour image.

    Neat stuff. Check the pdf.

  • This is probably _slightly_ off-topic but...
    I have read once, in my vast perusing of cryptography material, that back in WWI or WWII, a spy in Germany was sending messages out of there by encoding them in paintings. His method was by painting reeds in his pictures with different lengths - long for 1, short for 0. Fortunately for him, the Germans took a long time to figure this out. Unfortunately for him, they did figure it out.
    Hope this is interesting.

In seeking the unattainable, simplicity only gets in the way. -- Epigrams in Programming, ACM SIGPLAN Sept. 1982