I think there is a world market for maybe five computers...

[raspberrypiguy1]

'I think there is a world market for maybe five computers. -- Thomas J. Watson, founder of IBM (1874 - 1956)'

A tad out wasn't he?

The Raspberry Pi Guy

[OtherCrashOverride]

He is in good company:

"640K ought to be enough for anybody. " - (allegedly) Bill Gates, 1981

and

"10,000 of them should be enough" - (allegedly) Eben Upton, 2013 when submitting the RPi camera order.

[raspberrypiguy1]

Hahaha! So true! I mean can't everyone function on 640K?!

The Raspberry Pi Guy

[Burngate]

From wikipedia "Although Watson is well known for his alleged 1943 statement, "I think there is a world market for maybe five computers", there is scant evidence he made it"

Never-the-less, 5 in 1943 to 10,000 Pis in 2011 is quite a large inflation rate.

[mikerr]

Whether the quote is true or not, there are similar statements around that time

In context, in 1943, the "computers" referred to took up the space of a dedicated room, and cost the equivalent of several million dollars.

That's the investment level of governments, not companies

Now a $25 computer really is a modern wonder

[Gert van Loo]

I once read that the whole miniaturization and going to devices like the 7400 series was possible because of the need
(and willingness to pay) of the Apollo program for such devices.
Curious in how far THAT is true.

By the way i was there when Eben decide on a first batch of 10.000 cameras and he KNEW it was not enough.
It was just a convenient amount for a first batch to get things started.

[toxibunny]

"You may think me cocky sir, but I can see the day when there will be a telephone in every town!"

[jackokring]

I predict a computer in the phone in every town by the turn of the next decade. One must be bold with prediction, lest you underestimate popularity and look foolish.

[Ravenous]

I predict that by 2014 there will be a computer in every computer. Discuss.

[mikerr]

..but where are the flying cars ?

[Ravenous]

I once read that the whole miniaturization and going to devices like the 7400 series was possible because of the need (and willingness to pay) of the Apollo program for such devices. Curious in how far THAT is true.

Maybe - I read something like that, the Apollo computer was one of the first applications of integrated circuits. (I think there were two gates per chip/module.)

Here's a crazy man who made one recently:

http://klabs.org/history/build_agc/

(Look at his "Part 1" pdf for photos of his copy and a basic summary. Silicon Heaven!)

[raspberrypiguy1]

I predict that by 2014 there will be a computer in every computer. Discuss.

Mind = blown!

The Raspberry Pi Guy

[riffraff]

I predict that by 2014 there will be a computer in every computer. Discuss.

Already there. 7 adults in this house, and my wife and I just upgraded our Androids to newer Samsungs with UMA. Imagine my shock when upon returning home from work one day last week my router issued me the IP 192.168.2.27!
I've just spent the last 1/2 hour mapping out all of these devices with Fing. I've got to reallocate some of this wireless crap cuz nobody's getting a decent connection. To make matters worse, every house around us has one or more wireless routers and mine keeps channel hopping every time a vehicle passes.

[Bakul Shah]

I once read that the whole miniaturization and going to devices like the 7400 series was possible because of the need
(and willingness to pay) of the Apollo program for such devices.
Curious in how far THAT is true.

You read that right! From PBS's absolutely fascinating Silicon Valley program:

Narrator: It took two years, but in March 1961, Fairchild introduced its first commercial integrated circuit, called Micrologic.

At $120, Micrologic was far out of reach for the average private company. But when President John F. Kennedy announced an ambitious new space program in May, it was clear that the federal government would be willing to pay for it.
....
Narrator: At a time when the average computer was a room-sized machine containing mile upon tangled mile of wires, Fairchild's integrated circuit -- or microchip -- made it possible to put a computer right on board a spacecraft and send it all the way to the moon.

Jack Yelverton, Fairchild Semiconductor: This was one of those really important disruptive technologies, that this was something that was going to change everything in electronics.

Leslie Berlin, Historian: Brilliant people with brilliant ideas exist all the time; it's just a question of being a brilliant person with a brilliant idea, in the right place at the right time, where people want what you've come up with.

Narrator: Fairchild landed the contract to supply chips for NASA's Apollo Guidance Computer.

Hopefully this program can be seen outside the US.

[Jim Manley]

Fairchild landed the contract to supply chips for NASA's Apollo Guidance Computer.

We just happen to have one of the Apollo Guidance Computers as an artifact at the Computer History Museum here in Silicon Valley (I believe it was one of the test units that remained back at Mission Control or Cape Canaveral used to help troubleshoot problems during missions). Despite pioneering microelectronics, they also contained "rope" modules which were, of all things, core memory, but stretched out to be linear in configuration rather than planar to save space (think of pulling on the diagonally-opposite corners of a standard core memory plane's grid wires and twisting them until it resembled "rope" in shape). The Apollo 11 mission was fast-tracked for launch to the point where the Lunar Excursion Module (LEM) was never completely system-tested with all of the electronics in their final configuration. When Buzz Aldrin activated the radar altimeter as he and Neil Armstrong were approaching the lunar surface, the computer suddenly reset.

Fortunately, because of the core memory, minimal size of the code, lack of an operating system's overhead, and real-time nature of the system that it only needed to process current flight data (i.e., no history was saved), it could resume processing immediately after reset. The computer was suffering from inadequate shielding from the kilowatts of radar-frequency energy emanating from the altimeter and it was reset several more times during the descent. If you listen to the audio from the landing, you can hear Aldrin reading off the altitude, range to landing site, and vertical and horizontal velocities (based on radar doppler).

Armstrong had to maintain a mental picture of their position and movement toward the surface, but they had done hundreds of simulated landings, so he had become very proficient. Due to the high reflectivity of the lunar surface and high sun angle, they couldn't really tell how large the features on the surface were until they were within about 100 feet in altitude. It was then that they realized that the rocks they were seeing in the landing area selected from Surveyor 5 and Ranger 8 images (obtained just before their impact nearby up to two years earlier) were actually boulders the size of houses, and they were close to the rim of a wide, shallow crater not seen in the images. Armstrong leveled off the LEM and maneuvered it another 3.5 miles beyond the intended landing site where the surface was a mix of fine dust and rocks up to 2.5 feet across. They landed with less than 15 seconds of fuel left.

[Ravenous]

For more about the first landing and the computer problems, there's a great (and extremely detailed) write up from one of the programmers here:

http://www.doneyles.com/LM/Tales.html (Note figure 8 which shows how loaded the CPU would get on a normal run)

Neil & Buzz's tricky day at the office is described in their own words here:

http://www.hq.nasa.gov/alsj/a11/a11.landing.html (search for "1202" or "16/68" for example)

Proper engineers!

[Jim Manley]

Thanks for the links Ravenous - I need to make sure we have at least the documents by Don Eyles, et al, in the museum's collection associated with the AGC. We have pretty much all of the technical documents, including the six-inch thick code listings on 11 x 15 inch green-shaded fanfold. I almost fell out of my chair when I read that the software engineer responsible for the portion of the software where most of the troubles were experienced (not due to bad software, but bad data during the landing) was named ... HAL!!!

Don did discuss some things that better explain some of the events experienced by Neil and Buzz during their Excellent Adventure. While they did see some 1201 and 1202 warnings during simulator runs, they didn't see anywhere near the number, and especially at the times they did, during the actual landing. As a pilot of both helicopters and airplanes, I can definitely identify with their consternation at winding up with both their "heads in the cockpit" dealing with nattering alarms and blinking lights that should not have actually been occurring, instead of Neil remaining looking out the window, which delayed his surveying the prospective landing spots.

I knew that there was a problematic interaction between the landing radar altimeter and the computer (and had forgotten there was a separate rendezvous radar) and I had understood there to be an EMI issue between them (I need to go find the source that led me to believe that). However, the real problem was the overloading of the computer with bogus data from the landing radar altimeter. I found it quite unbelievable that it was possible for them to encounter so many unforeseen problems, such as the premature low-fuel warning due to fuel sloshing in the tank - why wasn't the tank based on a piston-style or collapsible design where the volume of the tank was always the same as the fluid contained inside? Perhaps it wasn't possible to seal combustible fluids within a piston-style tank within the weight constraints. Apollo 13 certainly pointed up the dangers of having to move fluid around in tanks with volumes larger than the contained combustible fluids.

I loved Neil's question to Buzz effectively asking whether they should ditch the landing radar altimeter - typical test pilot cockpit banter by guys wondering whether they were headed for a smoking hole. I also found illuminating his comment about the various scenarios leading to aborting the landing being the focus of the thousands of hours of training, but they looked for any excuse they could to make the landing happen. That was even to the point of potentially crunching onto the surface from as high as 40 feet when the fuel ran out, while the engineers had designed the landing gear to be able to withstand a powerless landing from only 10 feet.

I had misidentified the remaining fuel upon landing - it was 15 seconds to Bingo (a term used by military pilots as the point of no return if you don't head for a landing or to go get gas via air-to-air refueling), not 15 seconds total fuel remaining. They would have had no more than 20 seconds past Bingo before fuel exhaustion, so, they actually had 35 seconds of fuel left at engine shutdown.

There's a comment in the transcript about how, in 15 years or so, people would be relearning these hard-earned lessons when future landings on bodies in the solar system were attempted. Little did they know they would be off by at least 50 years if the first Mars landing occurs even by 2035, which isn't even considered a firmly achievable goal.

This put a bug in my brain about how to encapsulate the lessons of the landing as a scenario in the STEM educational game system I'm working on, Pi-finity! I'm thinking about not only reproducing the Moon landing, but also what a Mars landing would entail.

[Ravenous]

Interesting that the earlier Apollo 10 was partly meant as a rehearsal for the landing, probably hoping to find possible problems (hardware, software and procedural). However in Apollo 11 the actual things that caused trouble were things that weren't covered by the test! There's a moral there (that most of us already know about).

Back onto the computer: I just read the wikipedia page on the apollo guidance computer, and it's interesting that it was re-used for early fly by wire experiments. It links the following file (4MB):

http://www.klabs.org/history/history_do ... omayko.pdf

I'll be off reading this for several days!

[OtherCrashOverride]

Its 2013 and we can't put a man on the the moon! You expect me to believe we did it in the 60's !?!?!!?!

[Lob0426]

We, the U.S. no longer have the ability to focus on any project, no less returning to the moon. We have not even passed a budget in years, can't fix immigration, or figure out how to put people to work.

Returning to the moon would give us a leg up in a planned trip to mars.

We do not even have our own launch platform to put astronauts up to the international space station. That would cut too big of chunk from the corporate bailouts, Unemployment and Welfare budgets!

In all our government has abdicated it's lead in space technology to other countries and to private companies.

So do any of you think we, the U.S. will make it to mars in 2035?

[Gert van Loo]

So do any of you think we, the U.S. will make it to mars in 2035?

Of course!
I bet Coca cola are dying to put some white text on the red planet!

[W. H. Heydt]

'I think there is a world market for maybe five computers. -- Thomas J. Watson, founder of IBM (1874 - 1956)'

The unnerving thing about that for me is that Watson was born less than 10 years before any of my grandparents (all 4 were born in the range of 1878 to 1892).

As for Apollo 11...I got a day off work because of that. I was working for the Federal Summer Employment Program (tab operations in the San Diego Naval Supply Center). Nixon gave all Federal employees and day off in honor of the Apollo 11 landing.

Oh...and in 1961, not all computers 'filled a room'. Go find pictures of the IBM 1620.

[Lob0426]

@ Gert van Loo
Coca cola did not get a bail out so they cannot afford a space program either
Besides Mars is not that trade mark red color that they need. So a lot of extra cost for dye.

[Jim Manley]

Oh, come on, everyone knows we could put a man on the Moon or Mars next year ... the other half of President Kennedy's statement, " ... and return him safely to Earth ... ", well, not so much! Picky, picky, picky! However, if that man were to be Vladimir Putin, Ayman Al Zawahiri, or Kim Jong Un, I'm sure we could find the funding tomorrow.

There's actually a Dutch (Gert???) private commercial effort to send people one-way to Mars, and 100,000 losers ... I mean, pioneers ... have already volunteered, paying $38 each. They're going to select 40 "lucky" stiffs (sooner or later, most probably sooner) by the end of 2013. I wonder if the "stiffs" get free mobile devices filled with e-books so they can catch up on their e-reading during the 7 - 9 month trip to Mars, or would they get "lifetime" Netflix via NASA's Tracking and Data Relay Satellites (TDRS)?