LWN: Comments on "The sods must be crazy: OLPC to drop tablets from helicopters to isolated villages (ars technica)"

The sods must be crazy: OLPC to drop tablets from helicopters to isolated villages (ars technica)

vomlehn — 2011-11-28T23:16:09+00:00

Bless Mr. Negroponte's heart, but if he is serious, this is silly. Technology requires a context. If I give you a pencil, and you don't know what writing is, you are not going to know how to write in a year. I'm all in favor of OLPC but he is commiting the same mistake we engineers are prone to: forgetting that technology is made by people, for people, and has no relevance outside that. (When technology gets relevance outside of people, the discussion will get a lot hairier)

The sods must be crazy: OLPC to drop tablets from helicopters to isolated villages (ars technica)

vomlehn — 2011-11-28T23:02:30+00:00

"Been to poor countries?" I've *lived* in poor countries. Oh, and I've also spoken enough of their language to have serious conversations and listened to them tell me what their lives are like. Education is irrelevant without food, just like food is irrelevant without water and water is irrelevant without air.

Piaget sort of falls apart at the higher level of the needs levels, but the bottom ones are spot on. When you have food and water and shelter and physical safety, *then* you get benefit from education. And the benefit at *that* point is huge.

SFnal observation

NRArnot — 2011-11-10T12:35:10+00:00

Vernor Vinge was way ahead of the game yet again.

Much the same idea is a major part of the plot in "A Fire Upon the Deep" (published 1992).

Back to the future

jg — 2011-11-09T03:30:47+00:00

Knowing Nicholas, it probably is somewhat a stunt: but it's based on a very interesting project: The Hole in the Wall experiment in India.

But I know for a fact (having had the conversation with Oscar Becerra, who is in charge of the OLPC project for the MOE of Peru myself personally) that some of the schools that the MOE in Peru wants most to serve are really that inaccessible, and that there is no prayer of getting a helicopter to the ground. They are 3 weeks walk from the nearest supply point. The teachers walk for a month to get to the schools that are there (such as they are), teach for a month, and walk for a month to get home again: the kids get one month out of three of instruction.

When I understood this, my jaw hit the floor. I had had no idea there were still places on the planet this remote. There are... Crazy as it sounds, the experiment is probably worth a try (and Nicholas is a showman).
- Jim

Back to the future

emkamau — 2011-11-08T16:57:44+00:00

My comment was both disgusted and sarcastic. I doubt that any place they would be sending computers would be too remote to land a chopper. As you point out, these are places where they would need a power supply plus also evaluate progress with the project. That means its a place they can actually get into.

This is just a stunt.

emk

Back to the future

jg — 2011-11-08T16:18:06+00:00

There may not be a large enough opening in a jungle for the helicopter to even land...

Some places are *really*, *really* remote, such as the upper Amazon basin in Peru. It takes 3-4 weeks on foot to reach some of these locations, and there is unbroken rain-forest canopy.

The more interesting question is where the power is coming from; if the XO-3 can get low enough power consumption, then human power as originally envisioned becomes feasible. If not, then you have another hard problem to solve (such as solar panels on the top of the canopy, and getting the power down the trees).

Different worlds indeed

pboddie — 2011-11-07T17:17:54+00:00

Well, it's the accelerated graphics on the Broadcom chipset that's the primary issue, I think. One could argue that the close relationships between the initiative-takers and Broadcom, possibly a result of a fair number of Cambridge-based, ex-Acorn people (hence the BBC Micro vibe) being pulled into positions in Broadcom subsidiaries, may have led to a degree of complacency about such issues, but time will tell.

(There was a remark that a port of the proprietary RISC OS to the Raspberry Pi would probably work out because one of the people involved also works at Broadcom.)

Different worlds indeed

man_ls — 2011-11-07T16:39:25+00:00

Thanks for the insightful comment. I did not know that the raspberry pi was so closed; let us hope that it is at least completely hackable with free software. Binary blobs would be very disappointing.

Different worlds indeed

pboddie — 2011-11-07T16:03:50+00:00

The OLPC set out to change education worldwide, and failed. Asus put out a new retail product, and they not only sold tens of millions but created a new retail category; thus they succeeded.

I think that with less shouting and more listening there might be more agreement on display than either of you realise.

Asus certainly deserve the bulk of the credit for the retail netbook segment, but it was arguably an act of opportunism on their part: establishing a brand reputation for retail products while shipping low-end parts (small screen, recently-introduced CPU line) that other laptop producers probably weren't interested in. But had the OLPC project not stimulated or articulated demand for such machines, small form factor laptops might have remained a premium item, like the Palm Foleo or the current craze for ultrabooks (although the latter will probably also have 17" screens and the usual nonsense over time, if not already).

As for the Raspberry Pi, the project wisely avoids many of the areas that you keep criticising about the XO-1. There's no screen being shipped (a significant innovation in the XO-1 that probably won't be adopted by many other products for reasons that deserve further investigation), and the more radical networking aspects of the XO-1 are replaced by more pedestrian options, but the project has still fallen into the trap of relying on highly proprietary essential chipsets - even more so, one might argue - and this may cause a lot of problems in future.

Note also that the Raspberry Pi is not exactly blazing a trail: there's a glut of embedded solutions at around that price point, which is why the project seems to be able to hitch a ride with a vendor for the solution being offered. Some of the buzz around the project seems like nostalgia for the days of the BBC Micro. Although I was a huge fan of the BBC and Acorn computers in general, let us hope that the mistakes of that era - a proprietary standard from a single vendor - are not also reproduced alongside the benefits of the endeavour.

Different worlds indeed

man_ls — 2011-11-07T09:57:57+00:00

Thanks for defining my position so clearly; unluckily it is a bit more complicated than you portrait. You are also mischaracterizing the Raspberry Pi's intentions.

First, the raspberry pi is targeting general availability, not just selected education deployments. "The minimum order quantity will be one unit" counts as "retail" to me. So, even if only retail counted, it should not be discounted.

Second, goal achievement should be measured according to what those goals were. The OLPC set out to change education worldwide, and failed. Asus put out a new retail product, and they not only sold tens of millions but created a new retail category; thus they succeeded.

Last but not least, sorry that you will not be reading my comments, but I am not sure I will be missing your answers. It is sad that we are often so close-minded to dissonant opinions.

It can still happen

man_ls — 2011-11-07T09:54:36+00:00

Nice! The BeagleBone at 69 looks even more attractive.

It can still happen

BlueLightning — 2011-11-07T09:46:47+00:00

Drop the RS-232? That would not be good for kernel development (or system development, occasionally.) In any case, the Beagleboard XM *does* have an ethernet port.

Rather in the "no facts so my opinion is as valid as yours"

man_ls — 2011-11-07T09:41:23+00:00

Bending the facts a little now, are we? According to the last Wikipedia link, Asus demonstrated the Classmate PC and the Eee PC at the same time:

At the Intel Developer Forum 2007, Asus demonstrated the Classmate PC and the Eee PC

What you link here is not a demonstration but an announcement. But OK: it was a joint development between OLPC, Intel and Asus, and the cookie should be divided in three portions. You think that OLPC and Classmate PC (education-oriented machines) were precursors to the Eee PC, fine. In my view innovation is not just thinking about something but making it happen; if it was the former then the cookie should be mine too because I dreamed about such a machine in 2005. All happy now?

Sorry, but we live in different worlds

khim — 2011-11-07T09:33:24+00:00

The raspberry pi can at least happen because it is in the future.

Not in your world. You've outlines you position pretty cleanly: only retail counts. Raspberry pi does not target retail thus it'll never happen in your world. Raspberry pi may happen in my world, but never in yours.

If they don't ship, and another manufacturer comes out with a $40 real thingie which computes, then the latter should get the credit for their work.

Got that. If that's indeed your position then it's pointless to continue discussion with you. Thank god for comment filtering. Happy trolling.

Now we are in the "ignore the facts if they don't support my theory" territory...

khim — 2011-11-07T09:26:18+00:00

I read the link, I was alive at the time and remember the story.

Selective memory? OLPC creation: 2005. Classmate PC demonstration: 2006. Eee PC demo: 2007.

Meanwhile (probably not as a consequence as the wikipedia article seems to imply) Asus came up with a product that had decent specs and their own operating system.

Oh yeah. First ASUS was given the Classmate PC sample and then, year later, it "independently" created Eee PC with identical specs. LCD, CPU, all the hardware is the same - the only difference is the case (less rugged but more stylish) and amount of memory (256MB/2GB vs 512MB/4GB). Sorry, but I can not call soldering of bigger chips "huge innovation".

Asus did not come out with the Classmate PC and afterwards decided to build a "civil" version, the dates don't match.

You mean: year is not enough to build clone? Come on: ASUS engineers are pretty good.

Ignorance is bliss?

man_ls — 2011-11-07T04:38:36+00:00

ASUS quite explicitly created Classmate PC for Intel as response to XO-1 (when Intel was kicked out of OLPC) and it decided to sell "civil" version on the side.

I read the link, I was alive at the time and remember the story. Intel created the Classmate PC as a response to OLPC. Meanwhile (probably not as a consequence as the wikipedia article seems to imply) Asus came up with a product that had decent specs and their own operating system. Even if the Eee PC was inspired in the Classmate PC which was inspired on the OLPC, and something more than a wikipedia article should be required to prove it, it was Asus who thought it might make a viable computing platform and started the category of netbooks. My personal opinion is that it was a case of concurrent rather than causal development: Asus did not come out with the Classmate PC and afterwards decided to build a "civil" version, the dates don't match.

[the heavy delirium about hordes of third-world open source developers being unmet "technical goal" skipped]

OK, but you conveniently oversee the other failed technical goals: dual e-ink displays, the bloody crank, mesh communications, the failed touchscreen concept and the $100 price point. So no, they did not meet all their technical goals.

Not an order of magnitude

dlang — 2011-11-07T04:37:36+00:00

the OLPC got down to $200 each in quantity purchases. the $400 price purcased 2 macines, one that you got and one donated elsewhere (and yes, those donated machines really were delivered)

there are a lot of devices that are not available to the general public in small quantities, that doesn't mean they don't exist

It can still happen

man_ls — 2011-11-07T04:14:23+00:00

The raspberry pi can at least happen because it is in the future. The project can achieve their goals because they are not a pipe dream: "promote the study of computer science and related topics, especially at school level, and to put the fun back into learning computing." If they don't ship, and another manufacturer comes out with a $40 real thingie which computes, then the latter should get the credit for their work. Just as the BeagleBoard has its own merits, although someone should have told them to drop the RS-232 and put an Ethernet connection in there.

Not an order of magnitude

man_ls — 2011-11-07T04:05:10+00:00

I bought my Eee PC 701 2G for $250 on Easter 2008. And that is counting with retailer and manufacturer profit. Meanwhile, the XO-1 never got below $400 retail, and the theoretical $200 gave you a machine with worse specs than the 2G. That was probably because the 2G used standard parts, but that is the whole idea of netbooks: use yesterday's parts to sell wicked machines.

I'm not saying raspberry pi is a failure...

khim — 2011-11-07T00:57:20+00:00

the raspberry pi team has only slipped their shipping target a month, it's far too soon to call them a failure.

I'm not saying that they are failure. I'm saying they will not achieve more then OLPC. But since man_ls's definition of "success" is strictly "you can buy something in store" raspberry pi does not exist yet and probably never will.

in both cases, the ideas being introduced are what will matter far more in the long run than any particular hardware.

Sure, but to be consistent man_ls should attach ideas from raspberry pi to ASUS, Acer, HTC or anyone else who'll first manage to sell them in stores.

Either, or, please...

dlang — 2011-11-07T00:48:26+00:00

the raspberry pi team has only slipped their shipping target a month, it's far too soon to call them a failure.

OLPC had a far larger impact than the number of shipped laptops alone. It is fair to call it a failure in terms of the number of laptops shipped, but only compared to the incredibly insane numbers that were being talked about initially.

raspberry pi is not aiming nearly as high, and if they 'only' ship the 10K devices that they expect to have available starting in a month or so, I think they will be a success.

in both cases, the ideas being introduced are what will matter far more in the long run than any particular hardware.

Either, or, please...

khim — 2011-11-07T00:42:58+00:00

Now, the raspberry pi board...

Does not exist. Because you can not buy it in stores.

Please be consistent: either OLPC created netbook concept or raspberry pi does not exist yet. Please choose: either the first tradesmen gets credit or artist gets credit, you can not have it one way in one case and another way in other case.

I doubt Raspberry Pi Foundation will actually achieve more then OLPC did.

Ignorance is bliss?

khim — 2011-11-07T00:34:47+00:00

The Eee PC was the first usable computer to feature: solid state disks, desktop (or rather, laptop) Linux, below 2kg, for about $250.

Have you actually looked you your link? It quite explicitly says:

The Eee series is one response to the XO-1 notebook from the One Laptop per Child initiative. At the Intel Developer Forum 2007, Asus demonstrated the Classmate PC and the Eee PC, and listed specifications for four models of the Eee PC.

ASUS quite explicitly created Classmate PC for Intel as response to XO-1 (when Intel was kicked out of OLPC) and it decided to sell "civil" version on the side.

[the heavy delirium about hordes of third-world open source developers being unmet "technical goal" skipped]

Asus said: "computers for regular people don't have to be so expensive".

More like ASUS said "we have this cheap educational laptop - why not give it an interesting name and sell in stores". Good idea, but hardly a thing worth worshiping.

Meanwhile, Negroponte said "kids should get a bizarre education with a computer and no training".

Well, sure. Yet his educational ideas failed but technical approach revolutionized world.

I think there is a big difference.

About as much as difference between Xerox Alto and Mac: first was revolutionary development which spawned the whole "GUI revolution", second one was just one of copy-cats with delusion of grandeur.

I doubt it; the ideas were floating all around and everyone was having them, just that Asus was the first to deliver.

Actually ideas were not "floating around" but were concentrated on laptop.org website instead. ASUS only needed to reimplement them.

As Jobs said, real artists ship.

Jobs was master of deception and while I admire his ability to call white black and convince audience that he's right... white is still white, sorry. Real artists rarely ship - because this is different skillset. But it does not mean that salesperson which grabbed the work of artist and run with it can be called "real artist".

Don't get me wrong: Jobs was an artist and he did ship - but this was rare exception, not rule.

Not an order of magnitude

dlang — 2011-11-06T23:57:22+00:00

before the OLPC nobody was interested in making low-capability, cheap laptops. everyone was focusing on the side of moors law that said that you could get more for the same amount of money, so everyone was keeping prices of low-end devices about the same, they just got more capable over time.

the OLPC was the first project to really use the flip side of the law that said that for the same capability the price drops.

no, they didn't quite get to their $100 target, but part of that is that they had to increase the specs of the machine in order to run current software, and a larger part is that they never got the real quantities they were aiming for (which would have driven prices down), that said, they did get things to <$200 per machine, which was unheard of at the time they started

Asus and others realized from the OLPC project that there was a market here, and created devices in the $400 range.

the raspberry pi is another stab in this direction, it is skipping quite a few things that the OPLC included: keyboard, monitor, power supply, case, networking (wireless at that), once you add in all those things I don't think it is any cheaper. It's advantage is that in 1st world countries, many of these things will be available to be borrowed.

Not an order of magnitude

man_ls — 2011-11-06T22:48:16+00:00

Agreed. But the OLPC did not make that order-of-magnitude change, not even the promise: they just rode the wave. OLPC promised a $100 laptop when laptops could be had for about $600, and then failed to deliver. They were not a game changer in any sense of the word, but just used the round figure (100) to get attention.

Now, the raspberry pi board...

Real artists ship

man_ls — 2011-11-06T22:43:41+00:00

The Eee PC was the first usable computer to feature: solid state disks, desktop (or rather, laptop) Linux, below 2kg, for about $250. While the OLPC XO-1 was a toy which did not meet its goals: neither sub-$100 laptop nor a boon for third world education. (And by the way, mass production for under $100 was a technical goal.) Other technical goals not met by the OLPC project: the weird touchscreen, dual e-ink displays, the bloody crank, mesh communications, hordes of third-world open source developers after magically viewing source code. Not to speak about self-sustaining education materials or other organizational goals.

Of course, being sold in shops was a major market shift and part of the real revolution. Believe me, I was in the market for a similar thing for a few years, but there was nothing similar. Asus said: "computers for regular people don't have to be so expensive". Meanwhile, Negroponte said "kids should get a bizarre education with a computer and no training". I think there is a big difference.

By the way, the Eee PC 2G shipped before the XO-1 (either in "Give 1 Get 1" programs or to developing nations). Did the 2G use some of the ideas of XO-1? I doubt it; the ideas were floating all around and everyone was having them, just that Asus was the first to deliver. As Jobs said, real artists ship.

orders of magnitude aren't trivial

quanstro — 2011-11-06T15:21:00+00:00

in computer science, it's usually a bad idea to call orders of magnitude
trivial. orders of magnitude are were mere quantitative change
intersects qualitative change.

What is needed is one IT classroom par school, not one laptop per child.

Alterego — 2011-11-06T12:37:28+00:00

Sorry to disagree with you.

- You did not learnt your mother tongue on your own, it was teached my your parents, and it is hardwired in your brain to learn langage.
- You did not learnt to read on your own.

I agree that later when you are *mastering* reading, you can learn on your own with their (witings) help, book are only an intermediate between the knowledgeable one and the learning one, they are a transmission tool, they does not suppress the teacher.

Sorry, but no.

khim — 2011-11-06T09:11:27+00:00

No, the olpc project should not get any credit for netbooks: Asus should.

Both OLPC and ASUS were integral to the success of netbooks. First ASUS netbook was basically "OLPC for mere mortals" so it's hard to say that netbooks have nothing to do with OLPC, but later models switched to Windows, dropped still more OLPC ideas and that is what made them mainstream.

The eeepc was far more revolutionary than the olpc in my view.

Wow. Name one capability which Eee PC 701 which was neither in XO-1 nor in normal laptop. Eee PC was quite obviously made by using some revolutionary ideas from XO-1 and keeping other, so how can it be "more revolutionary"? Just because it was sold in regular shops? Is it such a revolutionary idea?

The $100 laptop was just a trivial generalization of the sub-$1000 desktop of the late 90s, and a goal that the olpc project failed to meet.

Actually they met all the technical goals they shoot for. Political goals (sales measured in millions, etc)... not so much. Some (not all) ideas from XO-1 were reused in Classmate PC and, later, in Eee PC. To say that XO-1 is "just a trivial generalization" but Eee PC is "far more revolutionary" is to say that facts don't matter.

This is the end, my friend.

man_ls — 2011-11-06T01:41:30+00:00

No, the olpc project should not get any credit for netbooks: Asus should. The eeepc was far more revolutionary than the olpc in my view. The $100 laptop was just a trivial generalization of the sub-$1000 desktop of the late 90s, and a goal that the olpc project failed to meet.

The sods must be crazy: OLPC to drop tablets from helicopters to isolated villages (ars technica)

drag — 2011-11-05T23:53:29+00:00

Prosperity is the state of flourishing.

That is a prosperous society is one that individuals have the ability to garner wealth. The higher level of prosperity the easier the ability for people to garner wealth.

Wealth is defined by goods/services that make a person's life better by filling needs and desires. It is impossible to put a overall number of wealth or quantify wealth because it is a extremely subjective.

A example:

Say I have a tuna sandwich and you have a glass of milk. I don't really want to have a sandwich because I am not hungry, but I am thirsty. You have all the milk you need, but would actually prefer a sandwich. So we exchange it. We perform a voluntary transaction by exchanging goods. As a result we are wealthier. In this transaction no wealth was stolen, but exchanged. So we are both better off with our needs and desires better met. And, in a small way, the net wealth of society is increased. Even after the food is consumed we are still better off because that we preferred to consume it to satisfy our appetites.

People often confuse wealth with money. It's a very easy mistake to make since on a individual level it is almost always true that having more money can make you more wealthy. But this is not ipso facto true that a society with more money is wealthier then a society with less money. Money exists to facilitate trade. It allows people to easily gauge the relative costs of different goods and is the key to the ability to do accurate accounting of profits and losses.

Confusion between wealth and money is just about one of the two most common economic fallacy there is. It's source of all sorts of really bad thinking and mistakes by individuals and governments. Even the majority of economists are not aware of the fundamental mistakes they can make with associating wealth with money. Very destructive.

What is needed is one IT classroom par school, not one laptop per child.

ovitters — 2011-11-05T21:25:52+00:00

Agreed. I learned my mother tongue on my own. I learned IT on my own. Think people underestimate what you can learn if you're interested. Teachers are not something magical. I've given various trainings, if people are not interested, a teacher is not going to change that.

The sods must be crazy: OLPC to drop tablets from helicopters to isolated villages (ars technica)

ovitters — 2011-11-05T21:23:08+00:00

Pretty sure they've thought of it.

This is the end, my friend.

ovitters — 2011-11-05T21:21:28+00:00

Finally, dropping them from helicopters most likely means there is no customer for these units. No customer means no payment.

What a strange argumentation. I don't get it.. at all. You mean whenever help is a provided, e.g. by means of dropping food, you assume that it is bad because they're not a consumer or something? There are difficulties to overcome. Not 100% of the devices will serve a useful purpose. This applies to pretty much everything.

The sods must be crazy: OLPC to drop tablets from helicopters to isolated villages (ars technica)

Pawlerson — 2011-11-05T13:25:35+00:00

Just don't use Damn Vulnerable Linux for this operation. Something like Unbreakable Linux would be better.

The sods must be crazy: OLPC to drop tablets from helicopters to isolated villages (ars technica)

blujay — 2011-11-05T09:32:07+00:00

That begs the question, what is prosperity?

The sods must be crazy: OLPC to drop tablets from helicopters to isolated villages (ars technica)

csigler — 2011-11-04T22:05:21+00:00

As an American, please allow me to correct you.

You're thinking of a UAV, a.k.a. drone.

What is needed is one IT classroom par school, not one laptop per child.

Alterego — 2011-11-04T21:06:38+00:00

You are right, %computer per people is a much better x-axis.
As one can choose the axis on the graph (i guess you find it) one can draw the said graph , litteracy vs computer per 100 people.
http://www.gapminder.org/world/#$majorMode=chart$is;shi=t;...

For me there is no corelation between both: africa has very low literacy rate compared to asian countries with the same computer % (and 50 years ago asia was poorer than africa, with a

It seems that country which managed to get 5 computer per 100 people have a very high litteracy rate (more than 90%), no need to get one laptop per child in countries where half the population is under 25 yo... but one IT room per school would be a good start ;-)

The sods must be crazy: OLPC to drop tablets from helicopters to isolated villages (ars technica)

Cyberax — 2011-11-04T19:43:27+00:00

You are wrong.

I've been to poor countries and education is THE BEST way to help them. That's the best way to create prosperity.

Besides, you can't help poor countries by giving them food (that's about the worst possible thing you could do, in fact) and providing security is not really feasible either.

The sods must be crazy: OLPC to drop tablets from helicopters to isolated villages (ars technica)

endecotp — 2011-11-04T18:06:42+00:00

If an American helicopter drops something on your village, the safest bet is to assume it is a bomb.