Hacker News Reader

Comment by jgrahamc | Hacker News Reader

It's a pity that the BBC focused on that aspect because that wasn't the main point of briefing them (and only came up because they asked): the main point was that the project is actually happening and we need about £250,000 to do the first part.

The first part being... study all the documentation Babbage left behind and build a 3D working simulation of the Analytical Engine. When done that'll be the first time anyone has seen the AE in action, albeit a computer, physical simulation.

And the Kickstarter situation is more complex than just 'costs too much': http://blog.jgc.org/2012/10/why-plan-28-decided-not-to-use.h... And, as I told the BBC, we have discussed using Kickstarter once we come to build the engine itself when it would be more appropriate.

View on HN

Shame about the angle the BBC took, but the publicity worked at least in part - I've just donated :)

Thank you. It is greatly appreciated.

"You could say it was the first failed government IT project."

That is a wonderful quote and good luck with the simulation.

Looks like The Register[1] has picked up your clarifications straight from HN and your blog.

[1] http://www.theregister.co.uk/2012/10/22/3d_simulation_babbag...

But introduced their own errors: see spelling of my name.

I think I was one of the first to sign the pledge you first released some time ago, good to see it's coming along. Donated.

John, what software are you going to use to do the simulation?

We don't have a definitive answer to that yet. We've talked to various companies and some experts. If you have specific knowledge of that field I would be very interested to talk to you.

The Analytical Engine will have something like 40,000 parts...

That is amazing, will the simulation be open source at some point, I would love to run that!

Yes. There are a number of reasons for doing the simulation (not least of which is checking that it will actually work), but one is that I want to make the simulation public.

As a charity Plan 28 is set up to educate the public at large about Charles Babbage and the Analytical Engine while building it. That means publishing openly as much as we can. Study has been ongoing for a while now and first public results will be early in 2013.

Will the simulation assume perfect mechanics or try to simulate friction, flex &c based on materials around then? If latter, then strikes me as very difficult project.

When you say 40000 parts, is that 'just' moving parts or every single nut, bolt, rod, plate etc?

When i was doing some 'blue sky' research at an architectural practice a few years ago I looked at a variety of mech eng packages for modelling skyscrapers at a level of detail which would have exceeded 40000 parts. I was interested because they were more advanced than the AECCad software available at the time and had better options for managing complexity. CATIA seemed to be the gold standard, but its very expensive and not at all intuitive. It seemed like we would need specialist technicians to operate it. Solidworks was most accessible in terms of usability. Spaceclaim was interesting but immature and didn't have much in the way of simulation. I didn't like Inventor as much as Solidworks, but that's maybe just because I tried Solidworks first. I dont know if any of these would be able to simulate the mechanical operation of the whole mechanism in one go though. I'd be very interested in hearing what you choose.

Idea for raising some of that £250k: make and sell to the public replicas of a simpler mechanical computer, the Antikythera mechanism.

There are many mechanical computing devices that could be popular. Slide rules, the (popularised by Gibson) Curta pocket calculator, the Pascaline, Schickard's Calculating Clock, etc etc.

(http://www.computerhistory.org/revolution/calculators/1/44)

(http://history-computer.com/MechanicalCalculators/18Century....)

Reply on news.ycombinator.com