Thursday, May 12, 2016

Spring 2016 report to the Computer Conservation Society

This was presented on May 11 to the Computer Conservation Society by Doron Swade.

We have been pecking away at Babbage’s original design drawings for some while now and have found with regret that we are unable to reverse engineer a coherent and consistent understanding of the Analytical Engine from the mechanical drawings alone. There are some 300 drawings and some 2200 Notations – descriptions of the mechanisms using Babbage’s language of signs and symbols. There were three phases of design - early, middle, and late.There is overlap between these, there are ad hoc upgrades, and only fragmentary explanation, where there is explanation at all.

It remains unclear whether any of these three phases is graced by a complete design. This in itself would be unfortunate but not catastrophic as mechanisms can be devised as functional replicas for omissions provided the intended function is fully understood. The immediate problem is that the extent of incompleteness is not clear. The work of the late Allan Bromley in decoding the AE designs in invaluable but he published only a small part of his substantial findings and these are anyway based on only part of the archive. While much is understood about many of the main mechanisms and the general scheme, there remain fundamental aspects control and sequencing that are not yet well understood and have resisted further illumination.

To achieve a more comprehensive understanding of the designs Tim Robinson in the US is going through the entire Babbage archive (over 7000 manuscript sheets) and producing a cross-referenced searchable data base. The purpose of this is to marshal all known sources so that we have a bounded idea of all relevant material. The intention is to reveal any explanations and/or drawings that Babbage might have left that have not yet come to light.

A second line of attack is to the 2200 Notations for the AE using the newly acquired knowledge of the Mechanical Notation. Allan Bromley maintained that the Notations were indispensible to his understanding. However he did not publish how he had used the Notations and he was the last to use them as an interpretative tool. The hope here is that the Notations will provide some of the missing information about logical control and moreover give insights into design strategy. In parallel with Tim’s comprehensive data base index I am going through the twenty volumes of Babbage’s ‘Scribbling Books’ identifying all material on the Mechanical Notation – a fast-track way of accessing this specific material.

Currently the stages envisaged for the project are:

1. Finalising research of the original design
2. Specifying a viable version
3. Computer modelling and simulation/3D printing
4. Manufacture and construction

We need more hands to the pumps and have latterly diverted some effort to fundraising. We have a 3-year plan at the end of which we expect have the requisite understanding of the designs, a platform from which to specify a viable version of an AE that is historically authentic, and to have trialled tools for modelling and simulation. The funding proposal includes pulling in requisite expertise including modellers and mechanical engineers.

In summary, we have had to bite the bullet with the realisation that without a concerted assault on the sources, a fuller understanding of the Engine design will not be forthcoming. We need to understand the intentions of the design well enough to identify missing mechanisms and understand their intended purpose well enough to devise fill-ins that are consistent with Babbage’s design style. Effort is now divided between continued study of the designs and fundraising to ramp up the effort.

Doron Swade


  1. How can we volunteer to help?

  2. I would be very happy to help too. Even for boring stuff ;-)

  3. I've offered help on at least a couple of occasions...I'm sure there are parts of this that would be amenable to "crowd sourcing".

  4. For annotating, categorization and linking of the original documents, putting them up on some kind of Wiki would be a great idea. Even if you just asked people to transcribe the text and put the Notation symbols into some computer-readable form, that's something that crowdsourcing would get you VERY rapidly - and in the process, you'd get a much larger mailing list and potential contributor list.

  5. I wasn't clear whether you were asking for our help, like some of my fellow commenters. Please tell us what it is that you would like us to do!

  6. I would like to run a few statistical regression and machine learning algorithms on the raw manuscripts. I would contribute any findings back to plan 28 to further the endeavour. gregory.james.ray -at- Thank you.

  7. @Herb: No, it's not clear - he says that they need more people - but that he needs to raise funding to get them. That seems unnecessary when there are clearly a BUNCH of people who'd just kill for the chance to be one of the people who helped turn the Analytical Engine into reality. For the cost of adding names to a nice little "thanks to..." plaque on the finished machine and/or an entry on the website - Plan 28 could get an almost unlimited amount of help from enthusiastic volunteers. For myself, I've been working on weird and wonderful machines for decades - and I have 40 years of computer expertise in 3D graphics, 3D printing, rapid manufacturing of all kinds. It's hard to believe that people like me wouldn't flock to the project if given the chance.

    I can understand that ORGANIZING that effort might be a little harder. If this were my baby, during this initial investigation phase, I'd be putting out the 7000 pages of goodies online and listing a bunch of "open questions" with a Wiki on which to collect possible answers...let people figure out "The Notation" themselves - find ways to encode it, be able search it, to write more of it to fill in gaps - maybe even to simulate it in some manner.

    For the known mechanical parts, I'm sure you'd get people who'd be happy to model them in 3D and animate the parts of the machine that are known.

    And very likely, all of those enthusiasts would be more than happy to donate a little cash where needed.

    But just asking for donations "cold" is hard. Getting people involved in the actual work will make it easier to extract cash from them where using their brainpower isn't helpful.

  8. As an idea of what's possible, consider the comic series...then look at the effort people put in at in trying to dig up explanations for the comic. Something like that would work *SO* well here.

  9. We are unable to crowd source work on the Babbage documents because they are currently restricted in terms of availability.

    1. Why? They surely are out of copyright!

    2. Protect the interests of the researchers I suspect.

    3. Access to the scans of Babbage's documents is not controlled by Plan 28 or any of the researchers.

    4. Can you tell us who owns the documents?

    5. I could be wrong, about any of this but, this is what I know:

      The main bulk of the corpus of all of Babbage's unpublished writings are with the Science Museum's Library and Archives in Wroughton, and then Doron Swade seemed to have mentioned the Buxton papers at the Museum of the History of Science in Oxford. (I thought I recalled another batch somewhere in the UK, but maybe that was the same one one or it found its way to the Science Museum.)

      I'd managed to get in contact with a Library and Archives assistant through a UK FOI request almost.. 3 years ago now (Aug 2013). Yikes..

      They had what they called a "finding aid" left by Bromley. At the time, it was just updated, and put into some new archive management system but which was not available online.

      She said though if I could give her references from the Bromley list, The Babbage Papers red book now really difficult to find, she could provide specific high-resolution scans at a charge, presumably because of things like high-res capable scanning equipment, oversized documents that are 150 years old, and limited manpower doing that sort of activity.

      I tried to prioritize which few drawings and docs would fill in the most gaps in my understanding, then came down with a case of "real life" and never got back to her.

    6. To my understanding, library science involves trying to balance making past information available with preserving the media that information is originally on. That's already difficult given the overall field seems to have low manpower.

      Then you throw in that Babbage left a huge body of writings and drawings and books, on projects he was constantly revising, redesigning, and superceding, in whole and in part.

      THEN you throw in that Bromley's reference information to it all, which he spent a fair chunk of his on-site time building, was published 25 years ago, had very few consumers (how many people buy a book that's just a huge index to an archive only a museum has access to? for an industrial age appliance few people could afford to build even a significant part of?), and is now long out of print.

      .. and you start to get a sense of the enormity of the task.

      IIR, Bromley's papers and reference indices were produced while on sabbatical (6 mo? 1 yr?) at the Babbage archives and via grants for that purpose.

    7. fromCalifornia: the answer to your question is the Science Museum in London.

  10. Has open access been made available to the Science Museum's technical archive? [Edit: Nm. Forgot to his post and by the time I did, this had been answered.]

    About a year ago or so, I spent some weeks chasing down all the published drawings and technical descriptions I could from here in the SF Bay Area, from the math and science libraries at Berkeley, Stanford, and San Jose State. One of them just seemed to be a giant index list to drawings/docs that aren't publicly available (but presumably in any of the UK archives somewhere).

    I felt I got a general sense of the underlying principle of the mechanisms and linkages from Bromley's papers and the layout of Plan 25. I was still curious about higher-level complex ops like division and general conditional execution or the apparatuses and cards pertinent to them. I guess since they're the least publicly accessible, I now know why.

  11. I fully understand the cost, difficulty and risk of scanning in these large documents - but the point is that this has already been done...and doesn't (or at least, shouldn't) ever need to be done again. So the key question isn't "Why can't the documents be scanned and made available?" -- it's "Why can't the existing digital scans be released?"

    After all, what on earth is the point of hiding mere copies of esoteric documents like these? The only reasonable purpose to scanning them is to help us to understand them - and what better way to do that than to throw them out to a large group of volunteers?

    I can't conceive of any reason to hold onto these scans.

    1. > So the key question isn't "Why can't the documents be scanned and made available?" -- it's "Why can't the existing digital scans be released?"

      Because Plan 28 does not control them. We do not have copyright in them and cannot disseminate them.

    2. Not 100% sure about UK copyright law, but US law the plans are are long out of copyright and the original being in the public domain so are any scans... the scans do not have a separate copyright from the original work.

      If you have access to the documents, and you're in the states, you should be able to share them, unless they're under some kind of special legislation.

      Scans of works in the public domain are also in the public domain. Also, the plans are owned by a public institution... that institution is granting you access but wouldn't let them be shared more widely? Why? Sorry I know this is an old post but I just found it...

    3. The key legal case here is the National Portrait Gallery versus Wikipedia dispute. The portrait gallery in London photographed all of their collection of many thousands of paintings and put them up on their website for the public to view. An enthusiastic Wikipedia supporter wrote software to upload them all to Wikipedias' "WikiCommons" site - on the grounds that these were mere copies of works that are out of copyright.

      The case went to court in both the UK and the USA.

      The UK decision was that in photographing them, the gallery had added some small artistic value (they did the lighting, framed the image, adjusted the camera settings, etc) - and that therefore they have copyrights to the scans, the guy who copied them is in trouble and WikiCommons must take down the images ASAP.

      In the US, the legal decision was that photographing the images was explicitly intended to produce a near-perfect copy and therefore DIDN'T add any significant art to the content. So the scanes are not copyrightable.

      Since WikiCommon's servers are in Florida - they still have those portraits up on their site.

      So...the Science Museum has said they'll put high resolution versions of their scans up online in the new year - and when they do, it seems that we can legally download them all onto a server in the USA - but **NOT** in the UK - and there should be no problems with doing that.

      It's possible that they might put restrictions on access to their website that say something like "You agree not to download/copy these images" - but if someone ignores that rule and copies them anyway (which could get them into trouble) - the copies would legally unencumbered. Since it would be easy for someone to do that via an anonymizing service and thereby avoid any consequences - we can assume that (as in the National Portrait Gallery case) someone, somewhere would probably do it. (NOTE: As a UK citizen, it wouldn't be me!)

      We know that Plan 28 have copies of the scans - but they had to sign some sort of non-disclosure agreement to get them - which overrides copyright concerns. So (frustratingly), they can't legally share their copies to the world...although if they did, it would be legal to disseminate those through US servers.

      So - bottom line - we have to wait until next year, when the Science Museum (hopefully) does what they promised to do - and then we can get on with crowd-sourcing anything that can usefully be crowd-sourced. Meanwhile, the Science Museum site hosts low-resolution copies of the scans - that are just *tantilizingly* too blurry to make the text readable!

    4. It's sounding like a US-based server will have less legal encumbrances than a UK-based one. Are there jurisdictions that might be better?

    5. > We know that Plan 28 have copies of the scans - but they had to sign some sort of non-disclosure agreement to get them - which overrides copyright concerns. So (frustratingly), they can't legally share their copies to the world...although if they did, it would be legal to disseminate those through US servers.

      Just to be clear: Plan 28 would never releases those images and breach the contract with the Science Museum.

      > So - bottom line - we have to wait until next year, when the Science Museum (hopefully) does what they promised to do - and then we can get on with crowd-sourcing anything that can usefully be crowd-sourced. Meanwhile, the Science Museum site hosts low-resolution copies of the scans - that are just *tantilizingly* too blurry to make the text readable!

      Yes, and last meeting with the museum indicates that this is being actively worked on.

  12. I agree with the previous comments. If the objective of Plan 28 is to reverse engineer the Analytical Engine and the existing team has concluded after three years (or is it five?) of study they can't figure it out, the Plan 28 organization has no more reason for being.

    I've gotten the feeling, and it may just be a result of limited information provided by Plan 28, that making the Babbage papers available to the public has never been their priority. The three major items on the website are 1) T-shirts for sale, 2) details on how to donate cash, and, 3) instructions for subscribing to the mailing list.

    The mailing list tells us their new 3-year plan is "continued study of the designs and fundraising to ramp up the effort".

    Frankly, the comments on this page are right: ramping up the effort is as simple as creating a wiki site and posting the documents. One also wonders what the funds raised now are going towards? Is it just to pay the few experts who will continue studying? After all, there is no hardware being built now, nor any realistic plans to do so. Why is fundraising the top priority?

    Part of the new 3 year plan is to trial "tools for modelling and simulation" Perhaps I can dig out the 3D CAD model I did about four years ago of the carry mechanism and its dynamic simulation. Honestly, any solid modeling package will be fine for modelling, and they all read each other's files. The dynamic simulation module in each package solves the same multibody equations of motion in the same way. Why pay people to play with software when the answers are already known?

    This may be a harsh critique, but the current approach to the Analytical Engine hasn't gotten very far. More of the same will likely be just as successful. In any event, more transparency from the Plan 28 organization can only help their mission.

    Tim Nye

    1. Unfortunately, Plan 28 does not control public dissemination of the scanned Babbage documents. That is entirely out of our hands. Believe me that if we had the right to do so they would have been made public and we would have crowd sourced everything. We simply do not have copyright on the scans.

      Also, no one is being paid (or has been paid) at any time on this project. Everyone has volunteered.

      We have gathered together everyone who is an expert on Babbage and the AE and they have been working together.

    2. Also, Tim Nye: if you would like to email me (plan28 @ jgc dot org) I'd be happy to put you in contact with the rest of the team as the modeling help may be very useful.

    3. Thank you for pointing out how horribly out of date the web site is. I have updated it.

    4. So, to summarize: The science museum have digital scans of all 7000 pages of Babbages' documents...right? The universal opinion of everyone here (including, it seems, John) are of the opinion that we need more eyes on those documents in order to proceed. It's pretty clear that nobody is going to pay for that to happen.

      In the most recent "Corporate Plan" the "Mission Statement" for the museum says:

      Inspiring, engaging and motivating the
      widest audience about the development
      of the modern world and its relevance
      to the future, through the best use of
      our collections.

      In support of this mission:
      The Science Museum will make sense of
      the science which shapes our lives.

      So the Science Museum has a clear choice:

      1) Keep the document scans locked away - where they'll collect digital dust forever. Nobody will ever understand Babbage's invention of the computer - which is insane given the importance of computers in every facet of human existence these days.

      2) Release the documents and see whether a crowd-sourced effort can produce some insight. Maybe we'll never make a physical Analytical Engine - maybe there will be another Allan Bromley out there who can crack it open and reveal the secrets. they think that locking away these digital scans will achieve their mission statement?

      I don't think so. Release them to the world - and they will WITHOUT DOUBT inspire, engage and motivate the many, many people who'd love nothing more than to try to figure out this machine - the "widest possible audience". And they'd be doing so
      "through the best use of our collections".

      So - who do we have to convince? How many names do you want on the petition? How much publicity do you want through places like Digg and Slashdot?

      If it's more widely known that the situation is that a public museum of the standing of the London Science Museum is preventing an effort to understand these important documents - the outcry will be overwhelming.

      On the other hand, if they do it right - put copies on a Wiki and have the public come and work on them - they'll get an incredible amount of credit and adulation from the tech community.

      I don't see any advantage in locking away these scans.

    5. The Science Museum has consistently said that they plan to open the archive up. I do not believe that they are trying to prevent understanding.

      To the contrary in the period that Plan 28 has existed the archive has gone from very hard to access (because it wasn't scanned) to digital and the museum has been working on a system to make the scans public.

    6. OK - so what they have is of the order of 7,000 PNG or TIFF files or something? The "system to make them public" can be as simple as "Upload them onto a Wiki"...that's something I can do in an evening!

      My wife's business wiki has far more than 7000 digital drawings on it - and I got that going incredibly quickly.

      We're not looking for a museum-quality online exhibit here. We just need a raw dump of all of the material in a form where we can work with it.

      The beauty of crowd-sourcing is that any grunt-work involved in organizing something will get done. There are people out there who will happily categorize, organize, annotate and make beautiful viewers for these things.

      Perhaps they don't understand...but the largest software project the world has ever seen (Linux) was built by about 4,000 people who worked on it for free - just for the joy of taking part.

      This project is *SO* sexxy. If you open it up, I guarantee you'll have more people willing to do grunt work than you could possibly imagine.

      Sure, we'd need to make sure that the Wiki is customized to make it look like something the museum can be proud of...which means making a 'skin' - that too can be managed in a weekend. We'd need to run it kinda like Wikipedia is run with a few people that have Administrative privilages so we can kick off the vandals and evil-doers - but that too is something you can crowd-source (just as Wikipedia does).

      This doesn't have to be difficult unless someone is trying to make it difficult.

    7. Gregory Ray: I find it difficult to believe someone who is truly motivated to 'open the documents' would, at the same time, disallow the sharing of said documents.

    8. I think the idea of getting transcription of all the Babbage documents crowd sourced is a good one. The problem with any such effort is that it does need organizing. In the examples given of Linux and Wikipedia it's not a true free-for-all.

      For example, for Linux, Linus kept a very strong hand on contributions to the kernel to ensure its quality.

      So I'm sure that having all this stuff public would harness energy from a ton of people, but at the same time it would need some organization. My guess is that the Science Museum (having a ton of other projects to work on) hasn't yet been ready to dedicate staff and time to it.

    9. Gregory Ray: Clearly we're barking up the wrong tree so let's waste no more time. If I may, I'd like to leave with a quote from Babbage himself:

      "Mechanical Notation ... I look upon it as one of the most important additions I have made to human knowledge. It has placed the construction of machinery in the rank of a demonstrative science. The day will arrive when no school of mechanical drawing will be thought complete without teaching it. "

      I believe Babbage would be saddened by the restrictions imposed by the Science Museum on his notations, perhaps we should redirect our frustration to Ian Blatchford?

    10. I'd hold the pictchforks for a minute and not attack the Science Museum. Our understanding is that they are close to making the material available.

    11. I don't think we need to have someone from the Science Museum being dedicated to running a crowd-sourced effort. These things do self-organize once they get started. The MediaWiki software supports several layers of privilages - all that's needed is to have one person in ultimate authority, who immediately delegates to crowd-sourced Admins and lets them do the day-to-day fixing of spammers, trolls and other miscreants. That's the way that Wikipedia runs - and it works quite well.

      Linux is a similar deal. While Torvaulds holds the ultimate key to "What goes into the kernel" - everything else is delegated and there are numerous self-organized working groups and so forth.

      If these 7000 documents were placed on a "Must sign up with email confirm for write-access" MediaWiki - with no more than a one-page document written stating the long and short-term goals - then the system would self-organize from that point.

      I actually think the best analog for this is from an unexpected place. The wiki. The "xkcd" comic is filled with clever nuances that most people miss - and the self-organized "explainxkcd" group set out to explain each new comic as it arrives. From that has grown an incredibly dense annotization of 1,600 comic strips. You can trivially find all of the comics on a particular theme - or all of the ones that have a particular character in them. Each one has layers of explanation. Every document is perfectly formatted with all manner of emergent conventions and rich cross-linking. Jargon has been created to explain certain kinds of events. There is meta-data telling you which comic strips need better explanations - pages that attempt to find threads running through the comics. The people who labor on all of this are obsessive about keeping it all clean and organized. Check out the "Community Portal" pages where technical matters are hashed out, editing efforts are coordinated, tutorials are provided for newbies.

      All of that with how many paid organizers? Zero. How many full-time volunteers? Zero. How much captital outlay? Well - they need a small, relatively low bandwidth web server. I pay $9.99 per month for mine, and it could easily support something 50 times bigger than explain XKCD.

      Please check out explainxkcd...ask people there how it came together and how it stays together.

      It is *EXACTLY* the kind of thing Plan28 needs to have...and it need not cost anyone a penny - and it could be started tomorrow if someone would take the 7000 diagrams and put them on a server from which they can be downloaded.

      Heck, I'll even set up the MediaWiki, buy a domain name and pay to keep it running if the data can be released.

      This doesn't have to be made harder than this.

    12. Their website has said similar things about making content, both Babbage's documents, as well as others more accessible digitally for the past few years I've looked at it.

      Given the Science Museum's amount of resources and responsibilities, over other history of science material beyond just Babbage's, I'm not remotely surprised that it's taken as long as they have.

      Whoever did the task of digitizing all the material probably wants to organized it all too, as well as defray the cost, which is why that cost matters and isn't just so easily dismissed as "well, it's already been done".

    13. Steve: I'm going to set up MediaWiki on the Plan 28 web server and make it available. Would you like to help administer it? My email is plan28 AT jgc DOT org. Drop me a line.

      As soon as the images are available from the Science Museum we could start one Wiki page per image and get people transcribing them. WDYT?

    14. I would be honored to be given the opportunity to help out at any level. I'll certainly help to get things started and donate what effort I can.

      I'll email you privately to hash out the initial setup - but we should plan to get the community involved at the first opportunity.

      MANY thanks for getting behind this - I know it must be a personal struggle to release the reins a little - but I'm quite sure there will be payback.

    15. Assuming there're no copyright issues with whoever owns the scans (and I really can't imagine this because they'd be copies of Babbage's own works, which should be well out of copyright), a wiki sounds like a great idea.

  13. Gregory Ray: What are we really going to gain from a simulation? Am I wrong that this device is a linear flow of mechanical interactions? Simulations are great for optimising and bug finding but to understand the inner-workings what more can be gained then just using your minds eye? We also have to face the possibility that if Babbage himself nor Clement built the machine it might not be a plausible design.

    1. A simulation can be as simple as "Enter the contents of the punched cards - press 'RUN' - see what comes out" - with no internal mechanical simulation at all. An "emulation" if you like. Or it could have a 3D graphical display of this gargantuan machine with whirling cogs and such...or anything in between.

      But before we can do either of those things, there needs to be a fairly complete understanding of the design...which we don't seem to have.

    2. If you read the history of Babbage, he was VERY bad at getting work done. He'd continually revise and re-revise the plans, making it impossible for work to proceed. He'd be incredibly rude and insulting to the people who'd provide him with the funding he needed. He didn't even get the Difference Engine built - and we know for 100% sure that it was a "plausible design" because the good people at Plan 28 built one - and it works!

      No - the reason Babbage didn't build an Analytical Engine had nothing whatever to do with plausibility of the concept. He simply didn't have what it takes to finish a job or to pick a design and stick with it or to obtain the funding he needed. He was a "computer geek" not a "project manager". He needed a project manager to hold him to a schedule, to say "Good Enough! Build it!"...that kind of thing.

    3. I will add to Steve's reply.

      There are a number of ways simulation will be very helpful.

      The real machine will be made from many metal parts, and not one will be perfect in shape or dimension. For example, Bromley mentoned that in the 40 digit store, if each digit was off by only 0.003" (the thickness of a hair), it adds up to a quarter inch at the last digit. Examining how designs can be made robust against lots of random errors is one issue.

      Another is collision detection. Lots of 3D parts with complex shapes are going to be moving simultaneously. Will jams occur?

      Will all motions be controlled, or will some parts be unconstrained and able to move at random? Seems to me the Difference Engine needed some locks added to keep things from wandering around.

      Then there are the issues with friction, inertial forces and gravity. The machine needs to accelerate and decelerate chunks of metal quickly to work at a reasonable pace. Parts will hit each other and cause force spikes. Friction depends on loads and adds to the forces parts must exert. These issues need to be quantified to design parts of adequate strength, and will need to be iterated to converge to an acceptable solution.

      The Babbage drawings show designs that most likely have not been built and tested. The components may work if made to these drawings, or they may not. Better to find out before construction starts.

      And, the design was never completed. The gaps need to be filled and that will go much faster if people can see how the interfaces work (remember, lots of complex 3D parts moving simultaneously).

      So when it comes to physical hardware there's quite a lot of design work that simulation can help speed.

    4. There are multiple questions to be resolved here. How did Babbage intend the machine to work? Would it actually have worked (with perfectly machined parts)? Would it actually have worked (with Victorian engineering skills)? Could a modern version of it be constructed?

      Simulation is valuable at most of those stages.

      If you read the book "The Cogwheel Brain" you'll see how Babbages' design for the Difference Engine does work - but that he had not attended to some of the practicalities of the design - such as how it would be lubricated or how you could dismantle a small part of it for repairs without taking the whole thing apart or even how you could debug the design.

      Those practicalities are likely to be even less well addressed in the Analytical Engine it seems quite likely that getting beyond a software simulation of the machine may be a tough thing to do.

      Fortunately, we can attack the problem in a step-by-step fashion - first emulate the logical design, then simulate an idealized mechanical system, then do some kind of finite element mesh simulation or something of that nature.

      But one step at a time.

    5. Incidentally, I discovered that the FourmiLabs Analytical Engine site has an emulator for AE! It's available for download - and can be run online if you have Java working in your browser (not many people do these days!).

      I think it's basically reverse-engineered from Ada Lovelaces' description rather than forward-engineered from Babbages' hardware designs.

      But it's interesting.

  14. Gentlemen (I assume, anyway), very good discussions.

    Let me suggest starting a wiki with the handful of items that are clearly in the public domain. Babbage's son's book had a number of figures at the end as well as some information on Mechanical Notation.

    (CAD models I made were from those images. They appear to have been chucked with the old computer, unfortunately.)

    John, could Plan 28 request permission to put some small fraction of images on-line while the Science Museum is working on the archive? Babbage's work has all been public domain for 75 years now. The Science Museum can't copyright his work again, but perhaps they claim rights on the digital images. Maybe they would go for a Creative Commons license that gives them attribution.

    Links to other resources would be helpful, such as some of Bromley's technical reports:

    A wiki site would be great for people like me who don't have the time to commit to a project, but may be able to contribute here or there.

  15. I reached out to the Science Museum, they are tentatively planning to have the images online August 2016.

    1. Well, we have reached the end of August 2016 - and seemingly we only have the low resolution scans. Do your sources at the Science Museum have a revised date?

      This is really becoming most intolerable.

  16. Re: Copyright on the digital images. The one image I've seen is clearly a high res photograph of the document with a wooden ruler placed next to it for scale.

    Copyright law is a bit futzy about recent photographs of out-of-copyright material. If one can claim to have added "artistic" merit to the photograph (eg, by including that ruler) then things get very dicey.

    You should definitely read this: almost identical situation where Wikipedia hosted photographs of a few thousand out-of-copyright photographs from the National Portrait Gallery. The upshot was that doing that was legal under US law - but not under UK law. European law is also involved.

    So it's not as simple as "the drawings are out of copyright".

    1. The reason I don't expect copyright to be a huge problem is because Babbage's papers are subject to Freedom of Information requests, and the Science Museum archives were quite happy to cooperate with making hi-res scans of select papers of his I requested a few years ago.

      The bottleneck really seemed like the cost having to get one of their workers to do it, which is why I'm not surprised to hear the soliciting of donations for an effort that involved scanning tens of thousands of said papers.

      In the worst case where the owner of the scans now made refuses to share citing copyright, we or anyone else in the public could request our own scans. (It'd just be a colossal duplication of effort and hopefully completely avoidable.)

    2. There are already low-resolution scans of a large chunk of the material on the Science Museum archive site. But when you present a 2'x4' sheet of detailed design at 1000x700 pixel resolution - it's not so much use! What we need is free access to the high resolution scans. We know those exist because the Plan 28 people have them - but evidently they signed some kind of non-disclosure agreement and can't release them.

      The problem with FOI requests is that they may be fulfilled with paper printouts - and it's hard to demand a particular resolution. Right now, I feel that it's better to let the good people at Plan 28 lobby the museum. As you can see, above - Plan 28 are agreed that we should take the crowd-sourced approach - and are actively working on getting a Wiki up and running.

      So I'm reasonably confident that this can be achieved...but how rapidly it'll happen is anyone's guess.

      These designs have waited 150 years to get examined...a few weeks/months more isn't so terrible in the grand scheme of things.

  17. Bingo!

    Babbage scans are on-line. E.g.,

    Start at and keep drilling down with the "Contains +" links at the bottom of the page.

    Unfortunately the images are presented as jpegs rather than a lossless format. It looks like they may be generated on the fly by requests. Maybe there is an API to connect to the highest quality version directly.

    Now if someone can figure out a viewer to add to a wiki that does some image cleanup and a convenient markup system, perhaps crowdsourcing can be used to transcribe the information from these documents. There are probably tools for convenient converting raster drawings to vector for import into CAD.

    After that there is going to be a lot of indexing and analysis to do.

    There are some free 2D CAD packages that anyone could use to work with 2D drawings. 3D solid modelling CAD doesn't appear to be available in free packages, but there are lots of people who have it and can contribute models. There are free viewers available that anyone could use to view the 3D models.

    Then there is the task of documenting all the modules, then identifying what parts of the Analytical Engine design exist and those that don't. Of the components that are designed, has their operation been, or can it be, verified to work?

    Getting back to the issue of project control (e.g., Linux kernel, etc.), let me suggest there is a huge amount of work to do just analyzing what exists before any decisions need to be made. A wiki could be set up with the goal of documenting and understanding the Babbage papers. A project to build a machine is a separate enterprise with different management needs.

    1. Agreed, particularly on analyzing and understanding the material.

      The hierarchy appears to follow Bromley's from The Babbage Papers. While that's an improvement over "just 7000 images raw", that seems adequate for a historian.

      For the purposes of deep comprehension and bringing actual machines to realization though, I definitely feel much further categorization and organization could be discussed and imposed on the available content.

      I found myself squinting a lot and with magnifying glasses attempting to discriminate letters and symbols from similar looking ones, primes vs. double primes in subscripts, and so on. I suspect I'm not the only one, so I feel any transcription that can "modernize" both drawings/visuals and annotations/handwriting would go a long way to improving the legibility problem.

    2. Again, the Science Museums' archive seems to be limited to 1000 pixels of horizontal resolution per photograph - which leaves notebook pages somewhat readable - but the large diagrams are pretty much illegible. No amount of magnifying glasses will help that.

  18. This seems ridiculous. There is no copyright in a scan, because a scan has no creative element. A scan is by definition a non-creative reproduction.

    Unfortunately when institutions need to make money to survive, they restrict access to their "property". A lot of the time, that means unique material rots away unseen.

    1. It's worth reading: which an enthusiastic Wikipedia contributor downloaded a ton of pictures from the UK's National Portrait Gallery - and uploaded them to Wikipedia. There was a court case - and Wikipedia won it...on the grounds that a simple, non-content-adding scan of an out of copyright image cannot be copyrighted.

      That was based on this decision:

      Unfortunately - the various legal decisions make it LEGAL to freely copy scans of out-of-copyright material in the USA, even without permission - but ILLEGAL in the UK.

      This means that Plan 28 (being located in the UK) can't release their copies of the scans without permission from the Science Museum - and unless someone outside the UK has copies, we're doomed unless the Science Museum decides to put them someplace where they can be accessed from the USA.

      Further up this thread, a poster named "Unknown" said: "I reached out to the Science Museum, they are tentatively planning to have the images online August 2016."

      ...but we have no actual confirmation of that.

      Since the Science Museum don't seem to be talking about this - we don't appear to have a way forwards at this point...which is frankly quite ridiculous.

      I can see why the National Portrait Gallery wanted to fight the loss of "ownership" of their scans - they were selling $800,000 worth of reproductions of them every year! But I really don't see the Science Museum making any money at all from thousands of pages of the "Scribbling books" or hundreds of blueprints of gear wheels and such.

      Since their stated goal is to use the artifacts that they own to promote understanding of them - they really, REALLY should release the high-res scans.

      Personally, I'm not optimistic.

  19. Still no sign of any high resolution scans. :-(

  20. Is there any update? I'm moderating a Steampunk panel at the upcoming 2016 World Science Fiction Convention.

    1. I haven't heard anything yet. I just looked at a few random pages on the Science Museum site - and all of the images I saw are still crappy low-rez stuff. Someone earlier in this thread ("anonymous") said that the science museum were planning to release the high rez scans this August...but I guess that could mean the end of August, so all hope is not yet lost!

      Sadly, we don't know who this anonymous poster is - and (s)he hasn't responded to my request to know where this tidbit came it could be bogus.

      I truly don't comprehend the reason why they didn't upload the high-rez images from day #1. But it is what it is.