Actually I was thinking 'unsustainable' rather than revolutionary.
Having lived in/through/with the birth/blossoming/commercialization of both the microcomputer and the non-factory robotics businesses, I find the breathless ecstasy of 'getting into production' a wonderful thing, but its a lot like falling in love, there are a lot of things that happen next that you're going to have to live through.
Given that they are taking nearly zero margin on these devices (at least according to this post), they have many pot holes to look out for:
1) Their Chinese factory turns out to be some guy's brother-in-law who got them a cheap price because he's using counterfeit parts. Or worse the guy that bid for the job expects to build the factory to build them with the money your going to pay him up front.
2) Their board vendor tech sneezed and accidentally added a direct short to the films they are using to make the boards.
3) The part they ordered for the USB connector is discontinued by the vendor who gave them a price of $0.25 and its now only available from some other vendor for $2.25 each.
4) The pick and place robot put a set of boards on 90 degrees out of kilter, so the entire batch has all the parts soldered into the wrong places.
5) The customs guys open them up, and not seeing the requisite safety symbol/OfCom/FCC/ISO/CET whatever certificate denies them entry into the country
The list goes on and on. One of the challenges of manufacturing is knowing not only what the parts cost but what costs the system is going to impose on you. You need to build in a margin that covers both otherwise you'll underestimate what it will cost you to build a product.
So new manufacturing experience aside (and its going to be a great learning experience if they can keep reminding themselves that it is a learning experience), it sounds like they have a really great way of bringing home to the politicians how their policies affect jobs locally. It almost seems like a reasonably responsive government should sponsor some group to manufacture a new product domestically and then carefully track the barriers to doing so and eliminate those that are amenable to a political fix. Tariffs on imported semi-conductors (protecting the local fabs?) and electronic components when such components cannot be locally sourced seems like a good starting point.
I certainly hope they will be able bring that message clearly to Parliament. It would be refreshing to have really hard to object to facts in the debates going on.
David Braben has been in the computer games field since the 1980s and has made a fair amount of money from it, Eben Upton works for Broadcom (the chip supplier for the Pi), and the rest are all pretty well known in their fields too.
So overall, I think they should by now know what they're doing, and with Braben on board should have some money to make it happen.
Have you guys really been reading about the raspberry pi? The founder is an SoC architect at Broadcom, I'm sure they'll find a reputable and cost effective manufacturing partner with their industry contacts.
Yes, I've been reading their stuff. And I'm hoping they are wildly successful! But I had to chuckle at your comment that their founder is an SoC architect because it reminded me of a really funny moment at this company I helped start back in the mid 90's. We had this new PCI card that we were getting manufactured for our systems and the guy who designed it was complaining about the delays in getting cards to put into test systems. Our director of manufacturing was really pissed off at the emails he was dealing with asking why things weren't ready and his response was along the lines of
"... some architect designs something and thinks that I can just bend over and pull a few thousand out of my ass. Doesn't he get that someone has to build the fucking thing and then somehow we have get them here?"
It was my first exposure to the more sordid details of getting a piece of hardware from prototype into production. For years up to that point I could easily get 10 or 100 of something made, but I hadn't ever tried to get 10,000 made, or had contracts in place to have them made on a regular basis.
Manufacturing acuity is not in the standard make up of senior technical people, its a process puzzle but with people rather than logic, with regulations rather than resource constraints, and with cash rather than electricity. The art of creating streams of parts which arrive at facility to take part in a process of assembly (and perhaps sub assembly) test and delivery, is a very different skill than designing an SoC.
That being said, its an amazing sight to behold when its all working right, sort of an eternal Rube Goldberg machine. And some folks make it look effortless.
Indeed, I think they really did set themselves in by setting a price point. We have seen so many projects, like the OLPC $100, I mean $200 computer, the OpenPandora and a number of others go down this same route and learning the hard way.
Fortunately there's a massive amount of buzz around this project, so even if the initial batches are unreliable and the units get to market at double the initial intended price, people are willing to put up with it. Perhaps the initial buzz is not a bad strategy, since it has a similar effect to Microsoft's strategy in the 90s. Many people hold off on buying other things waiting for this product, and even if it's not delivering as promised they are now psychologically invested in it.
As you say, it's also a good way for many people to learn about the barriers UK manufacturers have to face. I hope this will help bring about some change.
The first thing I am targetting is semi lame, but useful for me - a batman style lamp. Its for my parents and me to birectionally coordinate on when to come on skype. Right now I first have to call them and then ask them to come online and same the other way around. This way either party can speculatively just push the button and if the other one acknowledges the computers can be switched on.
If you are looking for a standalone Linux box just to turn on a lamp I think you are over-thinking things. Check out X10 automation kits: here's one for ~$24.
You should be able to control that with a small script (python, perl, etc. anything that can write to a serial port) that you keep running on any computer in your parents' or your own home. X10's advertising is notoriously lowbrow, but the hardware itself is decent and a little googling will find you a lot of hackers that have done crazy stuff with it.
I cannot persuade my parents to keep their computer on 24/7. That is the issue that I am trying to sidestep.
However thanks for the X10 link, it looks very interesting.
Wouldn't it be easier to build a standalone voip phone with a rasperry pi and avoid turning on a computer for skype altogether ? I'm not sure if you could do bidirectional video with the pi, but voice should be no problem.
I believe a standalone video phone should be quite easy. My understanding is that the chip can support 1080P encode or decode (so 720P bi-directional video should be fine) and that it will have OpenMAX libraries to allow you to drive the hardware codecs.
OpenMAX code is quite fragile in my experience (from trying to use such libraries on Android phones) so I was wondering about writing a Python wrapper (based on the ctypes module) for the video encode/decode to make it easier to use. (Unless someone has done this already?)
I would also be interested in how this could work. I've tried setting up VLC as a video phone, but the lag time was horrendous and that was when calling localhost.
Nice idea! You could hook it up to a mini projector or something that projects your face/an image of your choice onto the wall?
You could go one step further with that and project a "growl" style wall notification system. When I'm in the kitchen I often miss people trying to get into contact with me via IM or whatever, it would be great to see notifications on a surface or something to say "X has just messaged you!" or whatever
If volume commercial purchasing is possible at prices slightly above what they charge schools and students, we could see a revolution in low-volume consumer electronics. If 3D printing or other low-volume enclosure manufacturing is cheap enough, that is.
Oh man, we would see democratization of manufacturing! How durable the stuff printed from such printer? How long would it last (doesn't deteriorate/vanish/melt)?
I forgot to mention FCC and similar testing as well -- that would need to be made a lot more accessible and affordable. We need either an exemption for low-volume electronics (maybe such an exemption exists?), or a streamlined process for modular electronics so that each component (e.g the Pi) needs to be tested only once, and anyone using previously-tested components can just pass on their certificates, kind of like RoHS.
