if you want a physics dynamics engine, you want ODE. if you want a small amount of code (a high bang-per-buck ratio for the students i.e. they get results fast), use python. ODE has python bindings, so you win both ways. there are plenty of example recipes for python-pyode with OpenGL and pygame, which you don't entirely have to "understand", just copy cut/paste just like any other programmer would, and it gets the job done.
now, if you're looking for web-based, i cannot recommend "pure javascript". it's too much like hell on earth: you'll be teaching the kids the wrong lessons (namely: stay the hell away from web development).
i _can_ however recommend pyjamas. we have 3 or 4 example games already done: one uses SVG Canvas, it's "asteroids". then there's minesweeper, hangman, lightout and i also wrote a sudoku game a year ago. not one single one of those examples is over 1,000 lines of code. i think the asteroids one is about 600. http://pyjs.org/examples
the nice thing about pyjamas is that you don't have to "mess about" setting up the libraries. you just go straight in. thus, the bang-per-buck ratio (lines of code required to achieve a particular goal) is very very much higher than average.
http://en.wikipedia.org/wiki/The_Difference_Engine - by Bruce Sterling and William Gibson is a fascinating and complex exploration of exactly this concept: namely that Babbage succeeded. The key historical difference - the premise of the book - is that England's backing of the American Civil War succeeded, due to cryptography in part. Towards the end of the book it's made clear that the continued war between France and England has turned "cold" and thus much effort is dedicated to sneaking obfuscated "divide by zero" algorithms into the opposing side's Difference Engines. this book is one of the only sci-fi books (out of over 500 that i've read) that i actually found it hard to understand even 50% of what was going on. still made a damn good story, though.
By branding Bitcoin a "money transfer service" in the eyes of the public, powerful banking interests will then be able to begin loading Bitcoin down with regulations.
precisely. just like in "real life" with "real" money. you know - the stuff that is printed on-demand with no link to reality? (read senator ron paul's book "end the fed")
Regular wallets can not be stolen with a computer program. Stealing money from a bank account is possible for a program, but there are ways to recover it after the fact, unlike with bitcoin.
err, no. the banks are willing to *refund* the "electronic" money that is stolen, by replacing it with another lot of "electronic" money. and neither the police nor the bank of england (or whatever) will print you another bank note with the same serial number, will they?
so all that's required to "emulate" the present situation is for someone to set up an online bitcoin bank, and to offer the same "insurance" against theft as regular banks. of course, that means that they will need to charge for the service, and will need to be happy with the massive fluctuation in the currency. actually what would happen is that that bank would bitch like hell that it was your fault and that you didn't have the right anti-virus updates, and would delay the insurance claim just like any other bank.
yes. a friend of mine pointed out that bitcoin would not be considered a "serious currency" until you could buy drugs and pay for sex with it. then he found out that there is a web site that accepts bitcoin payments and will mail you some drugs, anywhere in the world. all that's left now is to find an intelligent and enterprising prostitute willing to understand and accept bitcoin, and we'll know that bitcoin has come of age, hurrah!
somebody in the u.s. hasn't been reading the geneva convention. if the U.S. is hell-bent on linking the words "cyber" and "warfare", then the U.S. had better be ready for the consequences. the consequences of "declaring war" on another country are very very simple: under the Geneva Convention, a declaration of war legitimises and grants the right for any citizen of the country being attacked to immediately take offensive action, no matter where they are, against citizens and against all soil of the aggressors.
in other words, should the United States respond with physical force against another country's citizens just because a computer which was wide open to the world (with 3 letter passwords), that is an "act of war", and the citizens of the country being attacked are automatically granted the right to take immediate offensive violent action against any United States Citizens or against any United States "property" and soil.
in other words, this is an incredibly stupid thing for the United States Government to be doing. especially given that many people in the United States Military have absolutely no idea what constitutes a cyber attack, and they certainly don't understand that 3 letter passwords are an invitation to go "cooeeee! i 0wn youuu!"
my site's a pyjamas application. it is therefore 1,000 lines of python.... or, when compiled (and therefore including the pyjs runtime which does stuff like dict, list, exceptions etc. all emulated in javascript, as well as including the library of widgets that are used on the page) it's 1.3mb of really obtuse but functionally correct javascript.
do you know what? i think this is *precisely* why companies have rejected so-called "embedded" processors for mainstream computing. they look at the pricing, and go "wtf?? the CPU's _how_ much less than the cost of the LCD panel??" freak out and refuse to design it into a product.
a modular design makes no such assumptions or restrictions. if people want to buy a $15 CPU card and put it into a $400 17in laptop chassis with a 1920x1080 LCD and a 3.0TB Hard Drive, they're mooore than welcome to do so.
also you have to bear in mind that $15 FOB (free on-board) is *not* $15 retail....:)
the CPU card planned to use the allwinner a10 is pretty functional, and the reference platform recommends something called the AXP209 which can handle USB-OTG power, 3.8v lithium battery packs, as well as straight 5.0v input. so we *can* do a complete stand-alone computer, which then can either plug straight into another machine and draw power, or plug in to a USB-OTG-powered hub and basicallly be completely independent from then on (it's got HDMI out, headphones etc.).
with the _added_ advantage that yes, you then plug it into other systems when they're available, including an 8-way rack-mounted server and including a 24-in all-in-one LCD TV / Internet / Computer "thing". lots and lots of options, here.
it's NDA confidential, but you could likely get one online if you looked hard enough:) the plan is to work information into a wiki, based on the source code and based on my access to the datasheet. everything that's needed _will_ be available so that people _can_ use these CPUs. if it gets really challenging and/or time-consuming i'll put in a request to the Board of Directors at Allwinner, ok?
(tftp, apologies, slashdot screwed up, i had to rescue this reply, without formatting)
JAE makes these connectors [jae-connectors.com]. I'm afraid you aren't trying hard enough. The job that you are about to undertake requires a lot of effort. You can't just take the path of least resistance.
ah! i got all excited, then i saw the number of mating cycles: maximum of 50. we're looking for something that will be reliable when removed and re-inserted by the average person, potentially several times per day
COM express - saw that one. it's not user-removable - factory only.
I'm unsure what you mean by that. I have COM Express boards and connectors right here, and I assure you they are perfectly removable. There are many docking station connectors that are of similar size and pitch.
you're not a mass-volume end-user. the ultimate target is not the occasional 1k or even 10k units: the ultimate target is multiple orders potentially in excess of 500k per month (inside china, where we "lowly" rest-of-the-world people happen to get the crumbs swept off their table in the form of lower pricing).
in that context, a COM Express board card simply won't do.
remember also: we're aiming for *mass-volume*.
You don't have the story to manufacture anything in mass volume. In essence, who are your customers, outside of Slashdot?
sorry, that's confidential. there is one abbsolutely massive one in china, some retail stores in the UK and one deal in turkey. every one of them depends on us having samples paid-for entirely *by us*, because they are such large customers they _expect_ to be served up with "free samples". hence the "bootstrap" process.
PCMCIA's 68-pin connector was designed for repeated insertion/removal.
The most popular RF connector, called SMA, is usually rated for about 100 mating cycles. The message here is that you need to realistically assess how many insertion/removal cycles are really needed. Yes, PCMCIA is more rugged, but it comes at a price. Reliability is one of components of that price.
i'll take it. (see above).
we can even re-use pre-existing casework
But is it worth it to lose a business just to reuse a couple of simple fixtures?
yes. we're running this as a CIC not a Ltd Company. profit maximisation is not the be-all and end-all.
I clearly am not in on your business discussions, and perhaps I'm missing something very obvious to you. I just don't see who would be your customer, and I haven't heard anyone in the last decade (until you showed up) arguing for the PCMCIA connector. If you build the card it will have to be plugged into similarly proprietary backplane.
it's been over two years of research, negotiations, discussions, and conversations and phone calls with my associates for at least 1 hour a day, often longer. it's a long _long_ story:)
There is one thought, though, that you might find useful. There are thousands of ARM boards on the market. Yours would be yet another one, with no significant difference. However there are very few cheap x86 boards. The lowest cost you can find is still two or three hundred dollars after you add the m/b, the CPU, the heatsink and whatever else might be needed. Can you fix that?
ah... no. i looked at it. actually i'd forgotten that i'd evaluated all of the x86 options available, with the exception of the transmeta (is that still dead?). we even looked at the new MIPS Loongson Leemote (upcoming) which will have 200+ hardware-based "acceleration" instructions suited to running x86 assembly code with only about a 30% performance penalty. as the new MIPS leemote is planned to i think it's a quad-core version initially, 1ghz, with the single-core version later, and even the vector processing unit being 512-bit wide on the single-core version will be capable of 1080p video decode at 100fps, i'm really excited by the possibilities on that... *if* they can get the bloody power down.
tftp, thanks for the heads-up. insight: i'm using openscad and mm3d: solidworks had better have the source code available before i touch it. that may give you a clue:) i'm a software (libre) developer, so this initiative is targetted at software (libre) developers. hence the quite deliberate decision to use ikiwiki as an open web site. i did actually add a CSS file, that's quite... that's quite something, that is, in software (libre) terms:)
specs are on the page; interfaces are on the elinux.org wiki; power requirements i've discussed on arm-netbooks (it's a PCMCIA card) it's max 5.0 watts; thermal considerations likewise (it's a PCMCIA card) max is probably about 3.5 watts, possibly pushing 4; mechanical drawings: it's a PCMCIA card, period - the mechanical non-interoperability requiremetns are on the elinux.org wiki; software is to be developed by software (libre) developers.
so the information's there - just not all in one neat place.
so, thank you for pointing out that it's critical to find good connectors!
we didn't pick PCMCIA because "it's cool", we picked it because it's still a mass-volume part (Conditional Access Modules) but is legacy as far as portable computers are concerned. and because it's user-removable. and because, quite simply, there isn't anything better. yes i looked at MiniPCI, i found _one_ image of a removable MiniPCI with ejector assembly: could i find who made it? could i hell.
COM express - saw that one. it's not user-removable - factory only. that defeats the object.
remember also: we're aiming for *mass-volume*. "enthusiasts" happen to be a critical link in the chain to bootstrap our way _up_ to mass-volume.
in other words, by the time we get _to_ mass-volume, we can always go and either raise some funds or spend existing funds raised by that point to get a proper impedance-matched PCMCIA connector specially made. or, say "what the heck" and get something else made.
regarding 0.1" headers: PCMCIA's 68-pin connector was designed for repeated insertion/removal. 0.1" headers are not. yes i looked around, tried to find a low-cost connector that would go onto the end of a PCMCIA card (in place of the 68-pin one). couldn't find one. if in your experience you know of something that would work, here please do tell me!
the other thing to take into account is that by sticking to PCMCIA 68-pin connectors we can even re-use pre-existing casework.
*sigh* absolute absolute last resort: both the SATA _and_ USB2 can be ramped down in top speed. SATA i think you can do as low as 150mbits/sec, something like that. i did look this up as part of the evaluation process. didn't quite do as experienced a job as you, though, so thank you for the heads-up.
i think the difference is illustrated by the following: i actually did seriously evaluate getting a beagleboard into a low-cost laptop (15in, 1920x1080 LCD). it was impossible. we had to find a massive "gaming" laptop with a 1in case that would normally have a huge fan in it, to give the height clearance sufficient to fit the beagleboard, comfortably. in the end we gave up with that one.
but yes: bari (one of the subscribers to arm-netbooks) has volunteered to convert the beaglebone, or any other board with open HW schematics such as the IMX53QSB, to EOMA-PCMCIA-compliant format, *if* there is enough community interest. personally i really like the AM3357, it's about the only low-cost decent CPU that would do for FSF Hardware-Endorsement. http://lists.phcomp.co.uk/pipermail/arm-netbook/2011-December/001155.html
yeahh i have a friend who went across the border into... i think it was the phillipines. there's something INSANE like a 140% tax on luxury goods. regarding the NGO efforts: you should read professor yunus's book "creating a world without poverty", he says that in comparison to CICs, NGOs have a disincentive to success. very very interesting book. thanks for the support, clarious.
pete - i get it, i really do. the thing is, i've been waiting around for companies to stop doing GPL violations, and they're just not getting it. in this article, i haven't described how much of a god-awful mess the situation in china is (see the other article comments a few days back, link at top of page).
there are plenty of people left in the world who still find what you've done (and i too) really exciting. they _like_ getting their hands mucky:)
we'll get to where you want to be, i promise. i'm gonna be soo pissed if doesn't happen:)
CRFF? you mean CompactFlash? i'd not thought of that one - damn good idea. let me think... it would be ideal... except it's only 44 pins (it's IDE, basically - see hwtools.net they have converters). 44 pins is not enough. thinking about it, we miiight be able to get two CF slots side-by-side, but who's going to make a double-sized case? naah, 68 pins by a jammy coincidence is just enough, _and_ we can still use the 3.5mm stainless steel off-the-shelf cases. jammy, huh?:)
Slashdot Mirror
hey fuck you too, google+! http://lkcl.net/SANY0051.JPG
if you want a physics dynamics engine, you want ODE. if you want a small amount of code (a high bang-per-buck ratio for the students i.e. they get results fast), use python. ODE has python bindings, so you win both ways. there are plenty of example recipes for python-pyode with OpenGL and pygame, which you don't entirely have to "understand", just copy cut/paste just like any other programmer would, and it gets the job done.
now, if you're looking for web-based, i cannot recommend "pure javascript". it's too much like hell on earth: you'll be teaching the kids the wrong lessons (namely: stay the hell away from web development).
i _can_ however recommend pyjamas. we have 3 or 4 example games already done: one uses SVG Canvas, it's "asteroids". then there's minesweeper, hangman, lightout and i also wrote a sudoku game a year ago. not one single one of those examples is over 1,000 lines of code. i think the asteroids one is about 600. http://pyjs.org/examples
the nice thing about pyjamas is that you don't have to "mess about" setting up the libraries. you just go straight in. thus, the bang-per-buck ratio (lines of code required to achieve a particular goal) is very very much higher than average.
http://en.wikipedia.org/wiki/The_Difference_Engine - by Bruce Sterling and William Gibson is a fascinating and complex exploration of exactly this concept: namely that Babbage succeeded. The key historical difference - the premise of the book - is that England's backing of the American Civil War succeeded, due to cryptography in part. Towards the end of the book it's made clear that the continued war between France and England has turned "cold" and thus much effort is dedicated to sneaking obfuscated "divide by zero" algorithms into the opposing side's Difference Engines. this book is one of the only sci-fi books (out of over 500 that i've read) that i actually found it hard to understand even 50% of what was going on. still made a damn good story, though.
By branding Bitcoin a "money transfer service" in the eyes of the public, powerful banking interests will then be able to begin loading Bitcoin down with regulations.
um... how?
precisely. just like in "real life" with "real" money. you know - the stuff that is printed on-demand with no link to reality? (read senator ron paul's book "end the fed")
yes of course they will "fall for it", because people are running automatic programs which offer exchange of money for bitcoins and vice-versa.
Regular wallets can not be stolen with a computer program. Stealing money from a bank account is possible for a program, but there are ways to recover it after the fact, unlike with bitcoin.
err, no. the banks are willing to *refund* the "electronic" money that is stolen, by replacing it with another lot of "electronic" money. and neither the police nor the bank of england (or whatever) will print you another bank note with the same serial number, will they?
so all that's required to "emulate" the present situation is for someone to set up an online bitcoin bank, and to offer the same "insurance" against theft as regular banks. of course, that means that they will need to charge for the service, and will need to be happy with the massive fluctuation in the currency. actually what would happen is that that bank would bitch like hell that it was your fault and that you didn't have the right anti-virus updates, and would delay the insurance claim just like any other bank.
yes. a friend of mine pointed out that bitcoin would not be considered a "serious currency" until you could buy drugs and pay for sex with it. then he found out that there is a web site that accepts bitcoin payments and will mail you some drugs, anywhere in the world. all that's left now is to find an intelligent and enterprising prostitute willing to understand and accept bitcoin, and we'll know that bitcoin has come of age, hurrah!
somebody in the u.s. hasn't been reading the geneva convention. if the U.S. is hell-bent on linking the words "cyber" and "warfare", then the U.S. had better be ready for the consequences. the consequences of "declaring war" on another country are very very simple: under the Geneva Convention, a declaration of war legitimises and grants the right for any citizen of the country being attacked to immediately take offensive action, no matter where they are, against citizens and against all soil of the aggressors.
in other words, should the United States respond with physical force against another country's citizens just because a computer which was wide open to the world (with 3 letter passwords), that is an "act of war", and the citizens of the country being attacked are automatically granted the right to take immediate offensive violent action against any United States Citizens or against any United States "property" and soil.
in other words, this is an incredibly stupid thing for the United States Government to be doing. especially given that many people in the United States Military have absolutely no idea what constitutes a cyber attack, and they certainly don't understand that 3 letter passwords are an invitation to go "cooeeee! i 0wn youuu!"
madness. absolute madness.
my site's a pyjamas application. it is therefore 1,000 lines of python.... or, when compiled (and therefore including the pyjs runtime which does stuff like dict, list, exceptions etc. all emulated in javascript, as well as including the library of widgets that are used on the page) it's 1.3mb of really obtuse but functionally correct javascript.
@BEGIN PGP SIGNED
... facebook happened.
@END PGP SIGNED
do you know what? i think this is *precisely* why companies have rejected so-called "embedded" processors for mainstream computing. they look at the pricing, and go "wtf?? the CPU's _how_ much less than the cost of the LCD panel??" freak out and refuse to design it into a product.
a modular design makes no such assumptions or restrictions. if people want to buy a $15 CPU card and put it into a $400 17in laptop chassis with a 1920x1080 LCD and a 3.0TB Hard Drive, they're mooore than welcome to do so.
also you have to bear in mind that $15 FOB (free on-board) is *not* $15 retail.... :)
hattig, hi,
the CPU card planned to use the allwinner a10 is pretty functional, and the reference platform recommends something called the AXP209 which can handle USB-OTG power, 3.8v lithium battery packs, as well as straight 5.0v input. so we *can* do a complete stand-alone computer, which then can either plug straight into another machine and draw power, or plug in to a USB-OTG-powered hub and basicallly be completely independent from then on (it's got HDMI out, headphones etc.).
with the _added_ advantage that yes, you then plug it into other systems when they're available, including an 8-way rack-mounted server and including a 24-in all-in-one LCD TV / Internet / Computer "thing". lots and lots of options, here.
hi fredan,
it's NDA confidential, but you could likely get one online if you looked hard enough :) the plan is to work information into a wiki, based on the source code and based on my access to the datasheet. everything that's needed _will_ be available so that people _can_ use these CPUs. if it gets really challenging and/or time-consuming i'll put in a request to the Board of Directors at Allwinner, ok?
(tftp, apologies, slashdot screwed up, i had to rescue this reply, without formatting)
JAE makes these connectors [jae-connectors.com]. I'm afraid you aren't trying hard enough. The job that you are about to undertake requires a lot of effort. You can't just take the path of least resistance.
ah! i got all excited, then i saw the number of mating cycles: maximum of 50. we're looking for something that will be reliable when removed and re-inserted by the average person, potentially several times per day
COM express - saw that one. it's not user-removable - factory only.
I'm unsure what you mean by that. I have COM Express boards and connectors right here, and I assure you they are perfectly removable. There are many docking station connectors that are of similar size and pitch.
you're not a mass-volume end-user. the ultimate target is not the occasional 1k or even 10k units: the ultimate target is multiple orders potentially in excess of 500k per month (inside china, where we "lowly" rest-of-the-world people happen to get the crumbs swept off their table in the form of lower pricing).
in that context, a COM Express board card simply won't do.
remember also: we're aiming for *mass-volume*.
You don't have the story to manufacture anything in mass volume. In essence, who are your customers, outside of Slashdot?
sorry, that's confidential. there is one abbsolutely massive one in china, some retail stores in the UK and one deal in turkey. every one of them depends on us having samples paid-for entirely *by us*, because they are such large customers they _expect_ to be served up with "free samples". hence the "bootstrap" process.
PCMCIA's 68-pin connector was designed for repeated insertion/removal.
The most popular RF connector, called SMA, is usually rated for about 100 mating cycles. The message here is that you need to realistically assess how many insertion/removal cycles are really needed. Yes, PCMCIA is more rugged, but it comes at a price. Reliability is one of components of that price.
i'll take it. (see above).
we can even re-use pre-existing casework
But is it worth it to lose a business just to reuse a couple of simple fixtures?
yes. we're running this as a CIC not a Ltd Company. profit maximisation is not the be-all and end-all.
I clearly am not in on your business discussions, and perhaps I'm missing something very obvious to you. I just don't see who would be your customer, and I haven't heard anyone in the last decade (until you showed up) arguing for the PCMCIA connector. If you build the card it will have to be plugged into similarly proprietary backplane.
it's been over two years of research, negotiations, discussions, and conversations and phone calls with my associates for at least 1 hour a day, often longer. it's a long _long_ story :)
There is one thought, though, that you might find useful. There are thousands of ARM boards on the market. Yours would be yet another one, with no significant difference. However there are very few cheap x86 boards. The lowest cost you can find is still two or three hundred dollars after you add the m/b, the CPU, the heatsink and whatever else might be needed. Can you fix that?
ah... no. i looked at it. actually i'd forgotten that i'd evaluated all of the x86 options available, with the exception of the transmeta (is that still dead?). we even looked at the new MIPS Loongson Leemote (upcoming) which will have 200+ hardware-based "acceleration" instructions suited to running x86 assembly code with only about a 30% performance penalty. as the new MIPS leemote is planned to i think it's a quad-core version initially, 1ghz, with the single-core version later, and even the vector processing unit being 512-bit wide on the single-core version will be capable of 1080p video decode at 100fps, i'm really excited by the possibilities on that... *if* they can get the bloody power down.
anyway: i updated the faq
i'm the boss: i rebelled at the request [from myself] :)
tftp, thanks for the heads-up. insight: i'm using openscad and mm3d: solidworks had better have the source code available before i touch it. that may give you a clue :) i'm a software (libre) developer, so this initiative is targetted at software (libre) developers. hence the quite deliberate decision to use ikiwiki as an open web site. i did actually add a CSS file, that's quite... that's quite something, that is, in software (libre) terms :)
specs are on the page; interfaces are on the elinux.org wiki; power requirements i've discussed on arm-netbooks (it's a PCMCIA card) it's max 5.0 watts; thermal considerations likewise (it's a PCMCIA card) max is probably about 3.5 watts, possibly pushing 4; mechanical drawings: it's a PCMCIA card, period - the mechanical non-interoperability requiremetns are on the elinux.org wiki; software is to be developed by software (libre) developers.
so the information's there - just not all in one neat place.
http://www.usb.org/developers/usbfaq#sig6
http://www.interfacebus.com/Design_Connector_Serial_ATA.html
the other link i found, on fciconnect, the impedance stated "100ohm minimum".
so, thank you for pointing out that it's critical to find good connectors!
we didn't pick PCMCIA because "it's cool", we picked it because it's still a mass-volume part (Conditional Access Modules) but is legacy as far as portable computers are concerned. and because it's user-removable. and because, quite simply, there isn't anything better. yes i looked at MiniPCI, i found _one_ image of a removable MiniPCI with ejector assembly: could i find who made it? could i hell.
COM express - saw that one. it's not user-removable - factory only. that defeats the object.
remember also: we're aiming for *mass-volume*. "enthusiasts" happen to be a critical link in the chain to bootstrap our way _up_ to mass-volume.
in other words, by the time we get _to_ mass-volume, we can always go and either raise some funds or spend existing funds raised by that point to get a proper impedance-matched PCMCIA connector specially made. or, say "what the heck" and get something else made.
regarding 0.1" headers: PCMCIA's 68-pin connector was designed for repeated insertion/removal. 0.1" headers are not. yes i looked around, tried to find a low-cost connector that would go onto the end of a PCMCIA card (in place of the 68-pin one). couldn't find one. if in your experience you know of something that would work, here please do tell me!
the other thing to take into account is that by sticking to PCMCIA 68-pin connectors we can even re-use pre-existing casework.
*sigh* absolute absolute last resort: both the SATA _and_ USB2 can be ramped down in top speed. SATA i think you can do as low as 150mbits/sec, something like that. i did look this up as part of the evaluation process. didn't quite do as experienced a job as you, though, so thank you for the heads-up.
i think the difference is illustrated by the following: i actually did seriously evaluate getting a beagleboard into a low-cost laptop (15in, 1920x1080 LCD). it was impossible. we had to find a massive "gaming" laptop with a 1in case that would normally have a huge fan in it, to give the height clearance sufficient to fit the beagleboard, comfortably. in the end we gave up with that one.
but yes: bari (one of the subscribers to arm-netbooks) has volunteered to convert the beaglebone, or any other board with open HW schematics such as the IMX53QSB, to EOMA-PCMCIA-compliant format, *if* there is enough community interest. personally i really like the AM3357, it's about the only low-cost decent CPU that would do for FSF Hardware-Endorsement. http://lists.phcomp.co.uk/pipermail/arm-netbook/2011-December/001155.html
yeahh i have a friend who went across the border into... i think it was the phillipines. there's something INSANE like a 140% tax on luxury goods. regarding the NGO efforts: you should read professor yunus's book "creating a world without poverty", he says that in comparison to CICs, NGOs have a disincentive to success. very very interesting book. thanks for the support, clarious.
pete - i get it, i really do. the thing is, i've been waiting around for companies to stop doing GPL violations, and they're just not getting it. in this article, i haven't described how much of a god-awful mess the situation in china is (see the other article comments a few days back, link at top of page).
there are plenty of people left in the world who still find what you've done (and i too) really exciting. they _like_ getting their hands mucky :)
we'll get to where you want to be, i promise. i'm gonna be soo pissed if doesn't happen :)
nonono, nooo :) that's _just_ for the tiny micro-engineering-board which is an *extra* board! sorry if that wasn't clear on the elinux.org wiki page.
thanks symbolset.
ok thanks, i finally got it, and added "max budget" field.
CRFF? you mean CompactFlash? i'd not thought of that one - damn good idea. let me think... it would be ideal... except it's only 44 pins (it's IDE, basically - see hwtools.net they have converters). 44 pins is not enough. thinking about it, we miiight be able to get two CF slots side-by-side, but who's going to make a double-sized case? naah, 68 pins by a jammy coincidence is just enough, _and_ we can still use the 3.5mm stainless steel off-the-shelf cases. jammy, huh? :)