WI : Babbage and Lovelace complete the Difference Engine and Analytical Engine

So historically, the first computer design was a serious mechanical monster. It also never was completed.

The PoD is that rather than failing to complete these projects, they are in fact built.

What impact does a machine like this have in the mid-1800s?
 
The first uses are probably going to be the census, refining artillery tables, and logarithmic tables. After that, possibly, maybe, banking since one of the machine's major selling points was improved accuracy. But the machine is huge, handmade, and outrageously expensive, all points against anything but niche high value applications. I guess success will depend heavily on the machine's reliability and operating cost. But assuming it gets build and works like everybody thinks it will the census is very likely going to be one of the first uses. Those used to be monsters to calculate.
 
Last edited:
Weren't commercial versions sold in the 1800's?

https://en.m.wikipedia.org/wiki/Per_Georg_Scheutz

Yes, there were a handful of similar machines that were produced in the mid to late 1800's. The biggest thing seems to be that they were never considered anything but huge calculators. They were used to produce things like logarithmic tables, that took tons of manual calculations, but that's about it. There was never much attempt to do more with them.
 
what you are trying to ask is not just about the differential engines, but more about programmable calculating engines
 
The first uses are probably going to be the census, refining artillery tables, and logarithmic tables. After that, possibly, maybe, banking since one of the machine's major selling points was improved accuracy. But the machine is huge, handmade, and outrageously expensive, all points against anything but niche high value applications. I guess success will depend heavily on the machine's reliability and operating cost. But assuming it gets build and works like everybody thinks it will the census is very likely going to be one of the first uses. Those used to be monsters to calculate.

Well, I imagine it might have the same use as early mainframes in that case. Huge processing-intensive pieces of work.

Weren't commercial versions sold in the 1800's?

https://en.m.wikipedia.org/wiki/Per_Georg_Scheutz

I have never heard of these, fantastic - it doesn't appear they were fantastically common - but the fact that they could be made more compact is certainly a factor that would make widespread adoption easier.

Yes, there were a handful of similar machines that were produced in the mid to late 1800's. The biggest thing seems to be that they were never considered anything but huge calculators. They were used to produce things like logarithmic tables, that took tons of manual calculations, but that's about it. There was never much attempt to do more with them.

That seems to be a shame.

what you are trying to ask is not just about the differential engines, but more about programmable calculating engines

Well, the Analytical Engine is the next step if the Difference Engine took off in Babbages workshop and could get the funding for it.

If we look plainly at the Analytical Engine, is this an entirely different story?
 
Well, the Analytical Engine is the next step if the Difference Engine took off in Babbages workshop and could get the funding for it.

If we look plainly at the Analytical Engine, is this an entirely different story?

i think not only funding, but series production is important. otherwise they will stay oddities
and when they get more common, it will become easier to get investors.
 

longsword14

Banned
i think not only funding, but series production is important. otherwise they will stay oddities
and when they get more common, it will become easier to get investors.
It was also too soon. You need electrical actuators,not mechanical.
 

longsword14

Banned
yes, was about to write that babbage was working on the edge what was mechanical possible.
Yes, and mechanical devices have parts that are much more cumbersome to make than later electrical ones. The biggest issue for Babbage was that he had to find somebody who would make the fangled machinery. The costs were prohibitive and the dreamed machine could never justify the ever increasing demands.
It was all too early.early 20th century is a more viable date.
 
i think not only funding, but series production is important. otherwise they will stay oddities
and when they get more common, it will become easier to get investors.

It was also too soon. You need electrical actuators,not mechanical.

yes, was about to write that babbage was working on the edge what was mechanical possible (physical limitations such as friction, mechanical losses etc).

Yes, and mechanical devices have parts that are much more cumbersome to make than later electrical ones. The biggest issue for Babbage was that he had to find somebody who would make the fangled machinery. The costs were prohibitive and the dreamed machine could never justify the ever increasing demands.
It was all too early.early 20th century is a more viable date.

I'm not sold on the claim that you NEED electrical actuators. Why is this? When it was built in 1991 AFAIK they didn't use any electrical ones.

I mean, compared to a modern computer, you're right, electrical components are great, but surely there is a market for calculating machines for big organisations. I.e. The Army, The Government. The Honourable East India Company. etc.

I'm not suggesting that they'd become personal computers, more curious as to the impact. After all, even early mainframes made huge changes in the modern economy. Surely being able to mechanically generate (with the designed printer), account reports is a huge advantage over having loads of clerks involved.

heh. Printed payroll in the 1800s, that is an amusing thought.
 
Yes, and mechanical devices have parts that are much more cumbersome to make than later electrical ones. The biggest issue for Babbage was that he had to find somebody who would make the fangled machinery. The costs were prohibitive and the dreamed machine could never justify the ever increasing demands.
It was all too early.early 20th century is a more viable date.
or she should have kept it simpler, not such big development steps

mid 20th actually, first fully programmable computer was the Zuse Z3, from 1941
 
I'm not sold on the claim that you NEED electrical actuators. Why is this? When it was built in 1991 AFAIK they didn't use any electrical ones.

I mean, compared to a modern computer, you're right, electrical components are great, but surely there is a market for calculating machines for big organisations. I.e. The Army, The Government. The Honourable East India Company. etc.

I'm not suggesting that they'd become personal computers, more curious as to the impact. After all, even early mainframes made huge changes in the modern economy. Surely being able to mechanically generate (with the designed printer), account reports is a huge advantage over having loads of clerks involved.
because of physics, you have problems with corrosion, friction, mechanical losses. the babbage final machine turned out to work, but was huge pain to get started. just because all these problems.
A slightly simpler design would have been better.

there was a big market for accurate logarithm tables, for navigation and such
 
I once read that you would've needed about 50,000 gear-wheels to build the Analytical Engine - simply too much for the current technology.

Did anyone out there read Babbage's autobiography?
 
The costs were prohibitive and the dreamed machine could never justify the ever increasing demands.

I'm not so sure. There are a couple of specific applications where it would be cost effective. I believe that had it been introduced it could have much earlier filled the role of later IBM tabulation machines and their like. Applications where time rather than money is the more important factor and there's a diminishing return for throwing people at it.

I do agree that it's never going to be more than a niche machine in its current form.
 

longsword14

Banned
I'm not so sure. There are a couple of specific applications where it would be cost effective. I believe that had it been introduced it could have much earlier filled the role of later IBM tabulation machines and their like. Applications where time rather than money is the more important factor and there's a diminishing return for throwing people at it.

I do agree that it's never going to be more than a niche machine in its current form.
I meant the money required to make the thing itself.
 
I can't remember where I heard about this, (it might have been here), but one possible path I've heard proposed for a TL with a more successful Babbage and Lovelace is a Victorian proto-internet. Imagine those early machines connected by telegraph, able to communicate via a set standard. You could have finance and industry driven by the rapid tabulation of prices and delivery schedules, increasing efficiency. You could see consumer finance being standardized: A Victorian lady in a department store could make purchases on bank credit, with the store simply sending that data directly to the bank's engines via telegraph, which then calculates the changes to her (or her husband's) account.

Taxes are also something that could be made much easier to calculate and collect here. If someone invents witholding, you could see effective and efficient income taxes in Britain by the 20th century. Other taxes are possible to envision as well--If tax liability is linked to a bank account, a corporation or individual could have excises, sales taxes, or VAT style taxes calculated for them based on data they report in on a regular basis, with this reporting being further telegraphed to tax collectors for reporting and automatic payment from that same bank.

This is infrastructure-intensive until someone discovers how to run multiple signals down a single telephone/telegraph line, but when that occurs you could see an explosion of innovation on the idea.
 
Could it have some application in cryptography? I think that you really need to find useful military uses for it to receive the necessary funding.
 
I can't remember where I heard about this, (it might have been here), but one possible path I've heard proposed for a TL with a more successful Babbage and Lovelace is a Victorian proto-internet. Imagine those early machines connected by telegraph, able to communicate via a set standard. You could have finance and industry driven by the rapid tabulation of prices and delivery schedules, increasing efficiency. You could see consumer finance being standardized: A Victorian lady in a department store could make purchases on bank credit, with the store simply sending that data directly to the bank's engines via telegraph, which then calculates the changes to her (or her husband's) account.
you read the novel the difference engine?

https://en.wikipedia.org/wiki/The_Difference_Engine
 
I can't remember where I heard about this, (it might have been here), but one possible path I've heard proposed for a TL with a more successful Babbage and Lovelace is a Victorian proto-internet. Imagine those early machines connected by telegraph, able to communicate via a set standard. You could have finance and industry driven by the rapid tabulation of prices and delivery schedules, increasing efficiency. You could see consumer finance being standardized: A Victorian lady in a department store could make purchases on bank credit, with the store simply sending that data directly to the bank's engines via telegraph, which then calculates the changes to her (or her husband's) account.

Taxes are also something that could be made much easier to calculate and collect here. If someone invents witholding, you could see effective and efficient income taxes in Britain by the 20th century. Other taxes are possible to envision as well--If tax liability is linked to a bank account, a corporation or individual could have excises, sales taxes, or VAT style taxes calculated for them based on data they report in on a regular basis, with this reporting being further telegraphed to tax collectors for reporting and automatic payment from that same bank.

This is infrastructure-intensive until someone discovers how to run multiple signals down a single telephone/telegraph line, but when that occurs you could see an explosion of innovation on the idea.

I'm not convinced that it would happen (Then again, 30 years ago would you be convinced of the Internet to the point that internet is my career!), I'm certainly fascinated by the idea. So you'd somehow be connecting the analytical engine to a telegraph somehow? Telegraph-Punchcard Signals! Perversely, if we run with the idea, if someone could essentially telegraph in a request/command, then we could see something not unlike cloud-computing or early workstations. I.e. I want to do something, you do it for me, and I'll pick up the paperwork later / you send my printer the instructions of what to print. At that point you've basically had to convert mechanical memory into electrical digital memory, and back to mechanical. That could realistically lead to someone trying to maintain an electrical memory.

(I haven't read the Difference engine mentioned by someone else, but I could dig that, if someone explained how it all happened).
 
Top