ARM Powered OLPC XO-1.75 Laptop Is Faster Than X86

Charbax writes "Not only is power consumption halved to less than two Watts and price of the motherboard reduced, the performance of the next generation OLPC Laptop is actually better for running full Fedora Linux compared to x86. Here's a video interviewing OLPC's CTO, Edward J. McNierney, where he explains how and why OLPC's world class engineers are making this change of CPU architecture. If OLPC XO-1 threatened Intel enough to start the netbook market and has reached two million poor kids in third-world countries thus far, XO-1.75 may help start the ARM-powered Linux laptop market. Do you think Fedora/Sugar will do, or should OLPC attract Chrome OS and Android solutions for education to get faster help from the big boys of Silicon Valley in bringing Linux software successfully to the next billion PC/laptop users?"

Comments

Funny

[Anonymous Coward]

It's the intel laptops that cost an ARM and a leg.

Re:Funny

[TaoPhoenix]

You had it and missed Anonymous!

They cost an ARM AMD a leg!

Re:Funny

[ichthyoboy]

You had it and MIPSed Anonymous!

FTFY

Re:Funny

[Culture20]

You had it and MIPSed Anonymous!

CPU ADD I.T. AMD MIPSed Anonymous!
Fini. No more replacement puns. Just stop.

Android for the masses

[drinkypoo]

If they plan to sell the machine widely so as to produce as many units as possible then ideally it would run Android. If they're only selling it for educational use then it doesn't much matter what is on it so long as it isn't (only?) Windows.

Re:Android for the masses

[aheath]

I am not sure that Android would be any better than Sugar. I participated in the buy two get one program so that I could look at the original OLPC XO 1.0 laptop. I was not impressed by Sugar. I would prefer to see OLPC provide a path from the XO to a full blown Linux distribution that does not require children to learn a new UI. OLPC should stick to developing affordable hardware and ask Canonical to provide optimized versions of Edubuntu and Ubuntu for the OLPC XO-1.75 laptop.

Re:Android for the masses

[takowl]

... a full blown Linux distribution that does not require children to learn a new UI...

You know, I think any computer UI is likely to be a new one for many of the children they're targetting. They've got a rare chance to design an interface for people who don't already have expectations of how to use a computer. I know I'd take that opportunity to see if I could work out a better model.

Re:Android for the masses

[Anonymous Coward]

The kids in most of the countries that got one of the originals were not learning a new UI. They were just learning a UI.

Re:Android for the masses

[Charbax]

I think the OLPC laptops by default all come with a full fedora linux desktop, as you can see in the video, which the kids can easily dual-boot into if they want "advanced mode", with full Gnome desktop.

Re:Android for the masses

[asnelt]

I don't think Android would form the right basis. It was basically designed to be an entertainment consumption OS. It is not designed for IT productivity which would be an important part for kids in third-world countries (not the only one but an essential part). The goal should not be to create countries of consumption drones.

Re:Android for the masses

[jonbryce]

If it ends up being anything like this
http://www.reghardware.com/2010/11/03/review_netbook_toshiba_ac100/
then no.

Re:Android for the masses

[Charbax]

That hardware is basically awesome, it's now just a question of software to make it fully Intel Atom netbook killer. It's nearly half the weight, potentially 50% of the price, runs 3x longer on a battery, may run even twice as much or more with a reflective Pixel Qi screen.

Re:Android for the masses

[jedidiah]

No. It's YOU that can't be serious. First you start with an obviously false premise and then expect us to take for granted something that no one really knows for sure.

Tablets just aren't that widespread. Most of what people know about them is taken on "faith".

Besides. If one "unix" with a tweaked shell can be successful then so can another.

Re:Android for the masses

[DrXym]

I am not sure that Android would be any better than Sugar.

Well it would be if they ever intend to produce a tablet version. Come to that, it might be better for a desktop too since it would allow apps to be written in Java and developed on Windows, Linux or Mac. This could help enormously to popularize the project and might even allow some crossover with apps being a downloadable for other Android devices.

Re:Android for the masses

[Anonymous Coward]

You should probably watch the video because the inverviewer asked about android.

Re:Android for the masses

[Anonymous Coward]

True that, but we all know slashdotters rarely read the fine articles.

Re:Android for the masses

[c0lo]

I would prefer to see OLPC provide a path from the XO to a full blown Linux distribution that does not require children to learn a new UI.

Since when being in a position to learn new things is bad for a kid?

Note that it is not the knowledge that's important, but rather to "flex that muscle" involved in learning and make learning (and, if possible, critical thinking) a constant through the life. Something that the westernalized "civilizations", so blinded by efficiency/cost-reduction, have lost the focus long ago - I'd venture to say for as long as 1950-ies. No wonder the "taming" process now called "education" is seen by the kids like a burden and also as a "cost" by the society in general.

No wonder a constructivist like Negroponte, in addition to a very low price, took another radical step: to make the OLPC not feel like a pure laptop but as a tool for leaning. A disputable choice, as there are many other choices leading to the same result, but at least the mission is very well defined:

To this end, we have designed hardware, content and software for collaborative, joyful, and self-empowered learning. With access to this type of tool, children are engaged in their own education, and learn, share, and create together. They become connected to each other, to the world and to a brighter future.

Also, some other quotes from Negroponte's personal vision:

It's an education project, not a laptop project.

Laptops are both a window and a tool a window into the world and a tool with which to think They are a wonderful way for all children through "learn learning"

.

Re:Android for the masses

[elgaard]

That is a good point. I have been playing with the AC-100.

The Android is not impressive on the AC-100.

But you can install Ubuntu on it; which is slow.

But it you use LXDE and trim Ubuntu a bit, it is actually good, much faster than Android.

I do like how the Android browsers make excellent use of the screen.

The Ubuntu/Debian package system is so much better than the silly app markets. More usefull applications and much easier to install.
(i didn't even manage to install Emacs on Android).

Re:Android for the masses

[LingNoi]

The goal is to use the laptops to further education and creativity, not to learn linux or the gnome desktop. I've seen a lot of open source advocates simply not get it and think the whole thing is about pushing linux. It's not.

Re:Android for the masses

[Anonymous Coward]

Because there's no Java for UNIX systems now?

And I'm not sure how an operating system that's deliberately built to make a general-purpose computer function smoothly as an appliance (and in the process sacrificing its computing capabilities) fits with the OLPC project's gola of helping kids learn something about computers.

Re:Android for the masses

[Anonymous Coward]

I am not sure that Android would be any better than Sugar. I participated in the buy two get one program so that I could look at the original OLPC XO 1.0 laptop. I was not impressed by Sugar. I would prefer to see OLPC provide a path from the XO to a full blown Linux distribution that does not require children to learn a new UI. OLPC should stick to developing affordable hardware and ask Canonical to provide optimized versions of Edubuntu and Ubuntu for the OLPC XO-1.75 laptop.

My you do have low standards. Canonical produces one of the worse Linux distributions out there.
Sugar is a very nice UI, suitable for education and very suitable for a small screen. It is designed so that the features are discoverable.
Under the hood is a full linux distribution. If you had really participated in G1G1 you would of know that.

By the way I took part in both G1G1 events.

Re:Android for the masses

[IB4Student]

Sugar Android full blown distro

Re:Android for the masses

[naz404]

Hi. Just reminding everyone that the current default OLPC OS now allows dual-booting to Fedora+Gnome aside from initially booting into the kid-oriented Fedora+Sugar desktop environment, making it suitable for more traditional uses by older users as well as being capable dev machines.

http://wiki.laptop.org/go/Release_notes/10.1.3 Will be posting this multiple times here (please don't mod as redundant) as Slashdotters really need to be made aware of this fact.

Re:Android for the masses

[Anonymous Coward]

MIcrosoft can do that,piece of cake

Re:Android for the masses

[DrXym]

Because there's no Java for UNIX systems now?

You're not getting my point. Sugar is a C++ app running on a purpose built version Linux. A prerequisite to doing anything is Linux and a knowledge of C++ development on Linux. That immediately limits people who will bother to contribute.

If OLPC were using Android then one immediate benefit is that any developer regardless of OS could write apps. Google provide an SDK which includes pretty much all they need to get going and it works through Eclipse on Windows, Mac or Linux.

Since the tools are available to more developers and they're friendlier too, OLPC will get more volunteers.

I'm note trivializing the work that would have to be done to move to Android. After all, it's merely practically a total rewrite of the entire UI. But it would have long term benefits for contributors, platform portability, and possibly even making Sugar a GUI that could be installed on other Android tablets. I imagine that if Sugar existed for tablets it would be enormously popular for parents, especially if its the same shell found in educational tablets.

Re:Android for the masses

[tehcyder]

If they plan to sell the machine widely so as to produce as many units as possible then ideally it would run Android. If they're only selling it for educational use then it doesn't much matter what is on it so long as it isn't (only?) Windows.

What's so good about Android as a general computer OS? It's fine for mobile phones, but I wouldn't want to use it instead of a proper Linux distro.

Re:Android for the masses

[Anonymous Coward]

*buntu on this would be a waste. *buntu has never been optimized for low resource usage, and low resource (specially power) usage is the most fundamental idea of OLPC. Yes, you say to get Canonical to optimize the OS instead of just putting *buntu there, but you know it's too heavyweight. A basic, thin Linux layer + Sugar is the best option.

Re:Android for the masses

[Junior J. Junior III]

You know, I think any computer UI is likely to be a new one for many of the children they're targetting. They've got a rare chance to design an interface for people who don't already have expectations of how to use a computer. I know I'd take that opportunity to see if I could work out a better model.

And thereby screw all those children by training them on a UI paradigm that they'll never see anywhere else.

Best case, you'll design a better UI and they'll hate being forced to use inferior, but entrenched, UI paradigms in the real world. Worst case, you'll design a worse UI and they'll still hate having to re-learn the mainstream UI paradigm even if it's nicer.

Re:Android for the masses

[SiChemist]

Sugar is NOT C++; It's written in Python. The person in the video mentions that in order to develop for android, you need a different system that doesn't run android. In other words, you couldn't develop for the machine ON the machine. It's a very important point since one of the goals of OLPC is to encourage tinkering.

Android just isn't a good fit for OLPC.

Re:Android for the masses

[Culture20]

They've got a rare chance to design an interface for people who don't already have expectations of how to use a computer.

So they have a rare chance to engender expectations that run counter to the rest of the computing world's expectations, to purposefully create a knowledge rift?

Re:Android for the masses

[Culture20]

My you do have low standards. Canonical produces one of the worse Linux distributions out there.

They take Debian and make it up to date and usable. Except for Gentoo (one of the truly worst in my opinion; fun for a year, but I'd rather not have a core or two constantly compiling), nothing else as visible is as bleeding edge and willing to use non-free elements for the sake of usability.

Re:Android for the masses

[jaydonnell]

I would prefer to see OLPC provide a path from the XO to a full blown Linux distribution

Did you watch the video? He said that you can easily switch to gnome.

I think you may be over stating things...

[Anonymous Coward]

I don't think the XO laptop kick started the netbook market for Intel. They were already on that path when XO came to the attention of the public. That and XO was not covered very heavily in the mainstream media.

Re:I think you may be over stating things...

[Charbax]

OLPC was founded in 2005 with the "$100 Laptop" idea which Bill Gates, Intel, everyone imediately poo-pooed. Intel's Asus Eee PC Atom platform was a direct reaction to OLPC's hype, Eee was promoted as $199 Laptop miod 2007 but introduced by the end of 2007 as a $399 netbook.

Re:I think you may be over stating things...

[Frosty Piss]

OLPC was founded in 2005 with the "$100 Laptop" idea which Bill Gates, Intel, everyone imediately poo-pooed.

And they have yet to come close to delivering a $100 anything. As well, they can talk about the specs of this latest version, but they have not yet actually built any.

Re:I think you may be over stating things...

[Charbax]

The $100 has always been target that can only be reached once more than 6 million units are mass manufactured, that was always the original idea. Intel tactics though, it has been proven in official state letters, successfully blocked OLPC from reaching countries like Nigeria, China, India, etc. But even though they "only" sold 2 million laptops to children in some of the poorest places in the world, you can find plenty videos online http://olpc.tv/ , see how the kids and teachers are using those daily, it's a huge success. I mean comon, OLPC may have deeply changed the lives of 2 million families in more or less very poor third world countries. Sure enough, it'd be better they reached 2 billion kids by now, by they I mean OLPC or anyone else in the industry. It's all about lowering cost and lowering power consumption of laptops and also bringing internet everywhere.

Re:I think you may be over stating things...

[Nadaka]

An arm tablet/netbook can probably have a per unit production cost of under $100 if they opt for a relatively small screen and battery. It is definitely a lot cheaper than intel for similar power and endurance.

Re:I think you may be over stating things...

[Frosty Piss]

The $100 has always been target that can only be reached once more than 6 million units are mass manufactured, that was always the original idea.

Than why bandy it about? It's an unrealistic number. They might as well say the can offer it for $50 if the make 20 million of them... It's a number pulled out of the sky that has no relationship to what will realistically happen.

By the way, I bought 4 of the original version (2 for me, 2 for Africa...)

Re:I think you may be over stating things...

[theillien]

And they have yet to come close to delivering a $100 anything. As well, they can talk about the specs of this latest version, but they have not yet actually built any.

That isn't the point. Charbax was responding to the ACs claim that the XO didn't kickstart the netbook market. Whether OLPC has delivered on its promises is another discussion entirely.

Re:I think you may be over stating things...

[Charbax]

2 billion kids are waiting. US spends billions dollars every day on useless wars in Afghanistan and Iraq. Why is it so unrealistic to have a vision where one actually puts money in educating the kids as soon as possible, before they grow old and miss their opportunity of getting inspiration to do big things in their future.

Re:I think you may be over stating things...

[Frosty Piss]

US spends billions dollars every day on useless wars in Afghanistan and Iraq.

This isn't about the US war in where ever. And the US ins't the only country neglecting "2 billion kids".

Re:I think you may be over stating things...

[Charbax]

It's a US project made by the good people at the US MIT, without US initiative this project might not exist yet. Sure thing, I wish India, China, Europe, Saudi Arabia, all join together and make sure every child on this planet get a fair chance at education now. It's politics that decide the priorities and where to put the money, tax who and sponsor what. The idea is sure enough we need to build a few million more/better schools, and bring Internet to all. But, even though those things have to happen, better food, better health and security, we might as well give the children a school in a box which a Laptop has the potential to be.

Re:I think you may be over stating things...

[H0p313ss]

2 billion kids are waiting. US spends billions dollars every day on useless wars in Afghanistan and Iraq. Why is it so unrealistic to have a vision where one actually puts money in educating the kids as soon as possible, before they grow old and miss their opportunity of getting inspiration to do big things in their future.

It has been commented over and over that the best way to make the world a safer, saner and more friendly place is to improve education, heath care (and health care access) and develop healthy economies. The OLPC project is one among many that can actually make the world a better place by providing affordable computing.

Re:I think you may be over stating things...

[Frosty Piss]

It's a US project made by the good people at the US MIT...

Again, this has what if anything to do with US wars?

Re:I think you may be over stating things...

[Charbax]

It's useful to compare things, analyse money transactions for understanding the value of things. For the price of 1 day of the US war in Afghanistan, all the kids of the whole country of Afghanistan could get a laptop. Shouldn't the head of military strategists consider that it may be worth it, that giving all the children in the whole country this type of tool could calm down some of the suicide bombers, might convert some of the extremists? Most of the girls in Afghanistan don't go to school at all, because they are scared or just not allowed to, why not give them this tool so they can at least try to learn themselves at home? You can't say this is not about politics.

Re:I think you may be over stating things...

[Nadaka]

In the case of things useful to compare, you should note that the actual daily cost of the wars is a lot closer to 200 million (US), not several billion.

Re:I think you may be over stating things...

[Frosty Piss]

It's useful to compare things, analyse money transactions for understanding the value of things.

No, there is no relationship at all between the was being carried out by the US and a social program being carried out by some former MIT folks. None.

Re:I think you may be over stating things...

[Charbax]

The relation is Governments and Money. It would absolutely be a military strategy to invest in education, every $1 spent on education could be worth 100x more towards bringing peace than every $1 spent on artillery or tanks.

Re:I think you may be over stating things...

[Charbax]

Total US Military Spending 2010: $685.1 billion That's about $2 Billion per day.

Re:I think you may be over stating things...

[mikael]

Just get the users to assemble the computers themselves - worked for Clive Sinclair and the ZX series of home computers ($149.95 assembled, $99.95 in components). Manuals and assembly instructions are provided at no extra cost.

Re:I think you may be over stating things...

[killkillkill]

Those children might have a bit of trouble using that laptop if the Taliban or similar Islamist groups again control the country. I'm not a big fan of ends-based ethics, so I don't want to make a case supporting the war. It is a bit unrealistic, however, that the lack of the US fighting the war would mean that the lives of these children would be substantially better and that even if that money could be made available for a program like this any of the children would actually see or use the computer.

Re:I think you may be over stating things...

[Anonymous Coward]

You know, it's shitholes like you that keep society back. If shitholes like you got their way, we'd all be living in Cuba or Venezuela or North Korea under some dictator living in a mansion while doling out "sage advice" to us peons.

The fact is, you're some cunt who lives off of your girlfriend, or perhaps you're a "trustifarian". At any rate, you don't add any value to the economy, you're nothing but some sponge shaped like a turd.

Re:I think you may be over stating things...

[Anonymous Coward]

I started programming with a $99 computer the ZX81 when I was 8, pile of crap but you could program for it and it was simple to use. Not sure I'd want to be an 8 year old trying to program for linux

Re:I think you may be over stating things...

[inglorion_on_the_net]

... even though they "only" sold 2 million laptops to children in some of the poorest places in the world, you can find plenty videos online http://olpc.tv/ , see how the kids and teachers are using those daily, it's a huge success. I mean comon, OLPC may have deeply changed the lives of 2 million families in more or less very poor third world countries.

To me, it's pure win. I mean, imagine working for a high-tech company and in so doing having improved the lives of 2 million families. Doesn't that sound GREAT?

Re:I think you may be over stating things...

[Carewolf]

No, actually the XO laptop started the ENTIRE netbook market.

Re:I think you may be over stating things...

[gl4ss]

http://www.dealextreme.com/details.dx/sku.42015

if you really want, here's an 88 bucks arm laptop(the windows ce on it might be pirated, tho), sure it's not super fast but olpc never was supposed to be. what sucks about olpc, is that they haven't been very good at lowering costs, instead spending time on subsidised plans and mesh research while the 'real' industry has been driving costs down like crazy(and really, moving a linux desktop to arm isn't such a bid deal that you should be hyping up your world class engineers).

and taking credit for the netbook market? they must be hiiiiiiigh.

Re:I think you may be over stating things...

[Nadaka]

A: Total US military spending does not go towards the wars. The cost for the wars directly is a less than $200 million a day. Even in a time of relative peace, we would still need provide funding for military readiness.

Re:I think you may be over stating things...

[tehcyder]

Please clarify exactly how GP's notion of spending money on education instead of pointless wars is holding society back.

Thanks.

Re:I think you may be over stating things...

[sznupi]

OLPC is largely an inexpensive Toughbook, with communities around it to boot. What you linked to...you might be lucky if it's not DOA.

They don't take credit "for the netbook market", they rightly take credit for initiating this category.

I want ARM power!

[Cyberax]

I want to buy a powerful ARM laptop, with the fastest CPU, most cores and the biggest screen (15" is preferable).

Is there anything like this on the market?

Re:I want ARM power!

[Anonymous Coward]

I want ARM power!

Curls, reverse curls and seated presses are what you need to do.

Re:I want ARM power!

[Anonymous Coward]

What about cheese curls? I do those all the time!

Re:I want ARM power!

[rwa2]

I want to buy a powerful ARM laptop, with the fastest CPU, most cores and the biggest screen (15" is preferable).

Is there anything like this on the market?

No, not really. Dual-core 2Ghz ARM chips are supposed to come out this year.

I just bought a Tegra 2 tablet to play around with (got the Viewsonic G-Tablet for cheaper than it would have cost to upgrade my midrange Android phone). It's all right. But the performance system you're looking for is still a ways off.

Re:I want ARM power!

[Charbax]

I also filmed a 14" 2Ghz ARM Cortex-A9 laptop at CES, see here: http://armdevices.net/2011/01/07/nufront-arm-powered-laptops/ In Europe Toshiba has released the best looking ARM Cortex-A9 Tegra2 Powered 10.1" Laptop, it's available for 160 euros for new (sub $200 retail price, consider Europeans pay approx 25% taxes). The only problem with that Toshiba AC-100 is current lack of decent laptop-oriented software, the Android that's loaded on it is not mature enough and Toshiba is very secretive about software update status. That Toshiba AC-100 has been rooted and impressive hackers have loaded Ubuntu on it but it's buggy for now, sound doesn't work yet for example, and it's risky to install, some people have bricked their units doing it. Shuttleworth said at recent Ubuntu conference that the Toshiba AC-100 is his favorite device. Much more may be coming soon in ARM Powered laptop segment. You can follow my site if you want news, or even post your news on it if you find something.

Re:I want ARM power!

[SimonTheSoundMan]

I don't think there are any today.

As for the article, I remember back in the mid and late 1990's when Acorn Computers had ARM powered desktops that were faster than x86 at the time, faster than what both IBM and Intel had to offer.

Intel 386DX vs. ARM 6 66MHz or Pentium II 300MHz vs. StrongARM 600MHz. It's a shame Acorn Computers are no longer.

Re:I want ARM power!

[Anonymous Coward]

I don't think there are any today.

As for the article, I remember back in the mid and late 1990's when Acorn Computers had ARM powered desktops that were faster than x86 at the time, faster than what both IBM and Intel had to offer.

Intel 386DX vs. ARM 6 66MHz or Pentium II 300MHz vs. StrongARM 600MHz. It's a shame Acorn Computers are no longer.

You do realize that ARM is a spinoff of Acorn Computers

Re:I want ARM power!

[inglorion_on_the_net]

That Toshiba AC-100 has been rooted and impressive hackers have loaded Ubuntu on it but it's buggy for now, sound doesn't work yet for example, and it's risky to install, some people have bricked their units doing it.

You would think that these hardware manufacturers have something to gain by selling as many units as they can, and that it helps there if they let people run whatever OS people happen to want to run. Apparently, that's not the way it works - not all hardware manufacturers give you a choice, and those who do usually give you a choice among two operating systems. Also, it seems they take a lot of hard work of getting the operating systems they do offer to run on the devices themselves, even though I am sure other people would be more than happy to do that work for them. What gives?

Re:I want ARM power!

[Anonymous Coward]

You should wait for the ARM Cortex A15, which will be on late 2012, yes, that happens to be the date when Windows 8 ARM version releases.

Cortex A15 has Quad-Core 2.5GHz version and has comparable performance to mainstream x86 cpus.

Re:I want ARM power!

[KingMotley]

You might think so, but a lot of hardware devices are actually sold for cost, or at a loss. They do that so that they can establish a market for the device, and eventually make money selling updates and/or software for them.

Unfortunately, allowing people to change to OS defeats both of those, and then the manufacturer just loses money on the hardware sale with no way of recouping the loss, and possibly make a profit.

Re:I want ARM power!

[owlstead]

When I was working for Gateway way back, I think it was about 2 telephone calls to support and the device was sold at a loss. Fortunately, most people never call in. Imagine what happens when people were allowed to run linux distro X *and* be eligible for support. I'm a staunch supporter of Linux, trying new distro's all the time and I've been running Ubuntu for 2 years now (on both my laptop & my PC), but even I cannot see how you could ever support such systems.

The problem with O/S is that you never know what you are dealing with. Managers just hate uncertainty. It's all fine and dandy, but you cannot start or support a project that may never finish, even if it does not cost loads of money. If you start a product you make commitments to clients/upper management. Managers won't make a commitment unless they are sure they can deliver.

Now I do think that linux should have much broader support from hardware companies, and I do think it will deliver if they do (if only to allow the price of Windows to be lowered in the long term). But there are quite a few snags on the road.

Good, back to hacking my video/monitor setup, and trying if I can get TRIM to work, I'll have to convert to EXT4 first, (I think). Oh, and my power safe features have failed *AGAIN* - no suspend and I cannot upgrade my SSD firmware for some reason, and that locks my machine in suspend mode now and then - the times that suspend works that is. Ugh.

Hand crank charge

[valnar]

"OLPC has learned from experience that finding electricity is the main problem in getting the XO laptops to kids in developing countries. Many areas lack electricity, making it hard to recharge laptops. The group has come up with a number of novel ways to recharge the laptop's batteries.

McNierney recharged the XO-1.75 with a hand crank. It takes 1 hour and 47 minutes to fully recharge the battery by hand, he said."

I'll be expecting to see this on a remake of Gilligan's Island in the future.

Re:Hand crank charge

[amiga3D]

The Professor did better than this. They pedaled a bicycle to recharge batteries on the island.

Re:Hand crank charge

[HouseOfMisterE]

They stirred salt water in coconut shells once to charge the batteries from their portable radio. That Professor was a smart guy!

Re:Hand crank charge

[zippthorne]

Too bad it never occurred to him to turn the VFO into an HF transmitter to send an SOS and get off the island, though...

Re:Hand crank charge

[Anonymous Coward]

He'd already run afoul of the FCC for doing that in the past. He was worried they'd triangulate his position before a passing ship could rescue them. He'd rather take the odds of not getting rescued over pissing off the FCC. :)

Re:Hand crank charge

[Gilmoure]

He was stranded on an island with two young, pretty women and just imbeciles as competition for their attentions. He was in no hurry to leave.

Re:Hand crank charge

[valnar]

Yes, that is a classic episode.

Re:Hand crank charge

[mikeabbott420]

He never lost hope that he could get both Ginger and Mary Ann if the competition was old, fat and retarded.
At night he would go out and do what he could to obscure any traces of occupancy that might be visible from an airplane.
A devious and obsessed man.

Re:Hand crank charge

[wonkavader]

As was said by Bob Denver, playing someone who look a lot like Gilligan, in "Back To The Beach": "[The professor] could build a nuclear
reactor out of a pineapple and a couple of coconuts, but he couldn't build a _boat_!"

Re:Hand crank charge

[amiga3D]

I can see that. Ginger was okay but the brunette was HOT!

Re:Hand crank charge

[rubycodez]

Except he was gay and modeled after Alan Turing. Poor Gilligan, having to be the bitch for Skipper, Mr. Howell (in his "sham marriage") and the Professor. Hope his man-gay riders used lots of coconut butter....

Re:Hand crank charge

[Gilmoure]

Was pissed off to find that Dawn Wells had a restaurant near my hometown (Clearwater, Fl) but I only found out after she'd moved away.

D'oh!

Re:Hand crank charge

[Grishnakh]

No, he didn't want to build a boat. The show never showed what was going on at nighttime. The prof was the only good-looking and not-retarded unmarried man on the island, with two hot young babes. Why would he want to leave?

Re:Hand crank charge

[Grishnakh]

No way, he was a buffoon. He thought he was the Alpha, but it was the Professor who was the true Alpha. He just didn't need to be loud about it.

Re:Hand crank charge

[zippthorne]

Every character Bob Denver ever played looked quite a bit like Gilligan....

Re:Hand crank charge

[wonkavader]

Crap. You're RIGHT!

The problem here is that I watched Gilligan's Island when I was less than 10 years old. I just didn't look at it the way ANY adult would. Duh. Of course.

I'm sorry I said anything. The professor was indeed a genius.

Build them and an app store.

[LWATCDR]

Build the hardware and sell it at cost or maybe less then create an app store to make more money.
WIth that money develop new version and or subsidize the sale of the hardware.
If you want to put a GPL app in the store it is free if you want to put in a none GPL it costs x and if the app isn't free as in beer you take y% of the price.

Not only are you getting the device into the hands of people that really could use them but you are opening up development and ways of making a live to people that may not have the opportunity otherwise.

Re:Build them and an app store.

[VortexCortex]

Build the hardware and sell it at cost or maybe less then create an app store to make more money.

Huh? So, only the "rich" poor people can afford the "cool" apps?

Besides, it already has a free "app store" (AKA activity repository).

openSUSE has packaged about 50 activities in total for Sugar, with more activities available for installation from the sugarlabs.org activities repository. Activities that haven't been packaged can be downloaded directly from http://activities.sugarlabs.org/ and installed by the user through the browse interface (the repository is similar to firefox addons.)

Re:Build them and an app store.

[91degrees]

Why not let people charge for GPL (and other free) apps as well? It's permitted by the GPL. The FSF is quite clear that you can do this with their blessing.

Re:Build them and an app store.

[Charbax]

Definitely they should add Android apps support somehow, I think the kids in Peru are going to enjoy Angry Birds.

Re:Build them and an app store.

[Anonymous Coward]

Yeah, sure, get them entrenched in the false economy of IP right away, so that they don't have to be bombed into submission later.

Re:Build them and an app store.

[Pharmboy]

Build the hardware and sell it at cost or maybe less then create an app store to make more money.

Yeah, brilliant plan. Surely no one will figure a way around that. All they have to do is install DRM on Linux for that to work, which will surely be bulletproof. Ask Sony or Apple or anyone else trying to have a monopoly for software on hardware they built but no longer own.

Re:Build them and an app store.

[LWATCDR]

I never said they couldn't sell GPL apps. The question is should those be hosted for free and just a percentage taken

Re:Build them and an app store.

[LWATCDR]

Well except for one detail to start with.
They where planning on sell the hardware at just about cost anyway as a not for profit and use the sale of hardware in some nations to subsides the same of hardware in others.
This way the profits from the app store could just add to the subsides. Plus you don't have to have a monopoly. Most android devices now will allow you to use other app stores on the device and or even just side load without jail breaking.
DRM is a different problem. I think that DRM is usless for the most part. It only hurts the honest consumer. A good example is me with Microsoft Flight Simulator. I always buy FS but the older versions made me put in the disk to run it. So I downloaded a crack for it to break the DRM. Now IMHO I have done nothing wrong since I own a copy of the program.
The music industry has found that DRM is pretty much useless. Now we need to get the Movie and book industries to get a clue and we will be in good shape.
So I would like to see it DRM free and move forward. It may actually give none mainstream movie, TV, and book creators a better shot at the market that way.
So yes I do think it is a brilliant plan since the OLPC was never supposed to make a profit to start with as far as I know.

Re:Build them and an app store.

[LWATCDR]

Who could afford what wild be up to the authors. Odds are that like most FOSS a lot of the really cool apps would be free just like on Linux and in the none FOSS world on Android and iPhone.

Arm powered ARM powered computer?

[schlachter]

Anyone else notice that they are building an Arm powered ARM powered computer? Now requiring only half as many cranks.

Re:Arm powered ARM powered computer?

[Duradin]

The OLPC hasn't had the integral crank since the design phase.

Re:Arm powered ARM powered computer?

[Anonymous Coward]

Now it only needs an OS built by cranks to operate the crank-powered computer. But GNU/Hurd keeps getting nowhere. /I keed because I love

Re:Arm powered ARM powered computer?

[macraig]

My family and friends insist that I'm seriously cranky, so I volunteer. Half - 1.

Re:Arm powered ARM powered computer?

[Caerdwyn]

If cranks can power an OLPC, Slashdot should be able to light up Las Vegas.

Re:Arm powered ARM powered computer?

[anexanhume]

And you don't need to take out an ARM on your house to afford one.

Re:Arm powered ARM powered computer?

[Arthur Grumbine]

Relative to the average income in a third-world country, it does cost an ARM and a leg.

Re:Arm powered ARM powered computer?

[anexanhume]

Last I checked, they don't offer ARMs on shacks. Jesus, this thread just got depressing. Let's kill a baby seal to liven things up.

Re:Arm powered ARM powered computer?

[Hognoxious]

Yeah, you need to keep away from all those minefields.

Re:Arm powered ARM powered computer?

[Anonymous Coward]

You're forgetting to count all the cranks who came up with the OLPC idea in the first place.

Re:Arm powered ARM powered computer?

[hedwards]

Dumb move. I kind of wanted one, but if they've nixed the crank, then what's the point. I know that they get really good battery life, but I thought a part of the point of it was not needing to plug them in.

Re:Arm powered ARM powered computer?

[killkillkill]

Well, if your arm is already worn out what what could you possibly need a computer for?

Re:Arm powered ARM powered computer?

[internettoughguy]

Dumb move. I kind of wanted one, but if they've nixed the crank, then what's the point. I know that they get really good battery life, but I thought a part of the point of it was not needing to plug them in.

I reckon a foot pedal would be more suitable, you have your hands free that way.

Re:Arm powered ARM powered computer?

[EnsilZah]

Yo dawg, we made an arm powered ARM powered computer so you can crank while you crank.

Re:Arm powered ARM powered computer?

[sznupi]

Now imagine how a more "standard" machine would look... (and in case you don't know: high tech toys tend to have large premium when compared to our prices, in absolute numbers)

Um, faster than...an 8 year old x86

[AcidPenguin9873]

The original XO-1 uses an AMD Geode LX 800, which was released in 2002/2003 or thereabouts. This latest XO-1.75 uses a Marvell Armada 610, and the marketing material I'm looking at from Marvell has a copyright of 2010 on it. The CPU in there is a Marvell Sheeva which the earliest reference I can find is from 2008, but that's not even a fair date because that's when they announced it, not shipped it.

So yes, this processor is faster than an 8-year-old AMD Geode. I would like to see power/performance tradeoffs vs. today's Atom and AMD Fusion stuff before everyone goes nuts about how ARM is faster than x86 for half the power.

Re:Um, faster than...an 8 year old x86

[pak9rabid]

That's what I was waiting for. I didn't believe for a second this ARM-powered box was anywhere near as fast as the current x86 offerings.

Re:Um, faster than...an 8 year old x86

[dirtyhippie]

Yeah, it's a misleading title, to say the least. Should have said "... faster than X86 version" or so.

Re:Um, faster than...an 8 year old x86

[h4rr4r]

Atom is a power hungry son of a bitch compared to Arm gear. The lowest power PineView based one is at 6.5 Watts and that is the CPU alone, the Arm stuff is all SoC. The whole SoC power budget is going to be less than that.

Once you start to value power consumption above all else Arm really starts to make sense. When you can plug in your laptop every couple hours well less so.

Re:Um, faster than...an 8 year old x86

[Ossifer]

Definitely faster than an Intel 8086.

Re:Um, faster than...an 8 year old x86

[Noughmad]

Well, it could mean it's just faster than some x86. I would say it outperforms a 386.

Re:Um, faster than...an 8 year old x86

[Charbax]

You won't find a sub-5Watt system (including the screen) running Intel Atom or AMD Fusion, so those cannot be compared. Why AMD or Intel don't go lower power, go ask them. the VIA Nano allows for a 4Watt system, which is a bit lower than AMD Geode.

Re:Um, faster than...an 8 year old x86

[h4rr4r]

It does not need to be, it only needs to be faster/watt.

Re:Um, faster than...an 8 year old x86

[AcidPenguin9873]

So you're saying an ARM SoC is lower powered than an x86 system? I agree! That's not what the title of the article says though.

Re:Um, faster than...an 8 year old x86

[jedidiah]

Yes. Because the measure of a computer is how much power it consumes.

No. If you need to do something computational then you need to do something computational. There's just no getting around it really. There are some limited edge cases where you can tailor hardware to a select few computational use cases but sooner or later you will probably want to wander off into unusual territory. THAT after all is what a general purpose machine is for.

Call a spade a spade and cut out the fraud and nonsense.

Re:Um, faster than...an 8 year old x86

[h4rr4r]

The title refers to the x86 version of the OLPC, clearly. I believe that. I also believe that these ARM SoCs are faster than any x86 system constrained to the same power budget.

Re:Um, faster than...an 8 year old x86

[jedidiah]

Atom is a power hungry son of a bitch compared to Arm gear.

It also performs a whole lot better. In x86 terms, the raw CPU power of an ARM is like a Pentium 300Mhz.

Intel might even be able to throw together some recycled 90s technology with comparable performance and end up with something ARMs equal in power.

That was the model for mobile computing for a time (adapt outdated desktop tech).

Re:Um, faster than...an 8 year old x86

[KiloByte]

Or, alternatively, put a chip with four of those instead of one. Being faster with one means you'd suddenly be more than four times faster than x86 at the same power.

Re:Um, faster than...an 8 year old x86

[Jecel Assumpcao Jr]

I suppose that by "x86 OLPC" they mean the current XO 1.5 which is powered by a 1GHz VIA chip.

http://wiki.laptop.org/go/Hardware_specification_1.5

Re:Um, faster than...an 8 year old x86

[Tanktalus]

You're calling this a false comparison because the title/summary are vague, and you have an expectation of comparing modern to modern. But that's far from the only way of comparison. Often you'll see products compare model-to-model. "We've managed to squeeze 13% more mpg out of this year's vehicle" or "we've added feature X to the new version" or "we've fixed 100 more bugs this release than we did last release."

Sounds to me like the article (which, of course, I've not read, and don't plan on) is comparing the next-generation of the OLPC to the previous-generation of the OLPC. That is a fair comparison. They're not trying to get into a design debate with geeks like us. They're telling everyone that their design is different, and one of the benefits they got was a faster system.

You're right that it's not as big of a deal as the title sounds when considering the choices they could have made staying on x86. But they had other goals, met (most of) them and still managed to come out ahead of the previous offering in raw speed. As a package, it sounds like a pretty decent thing, though I don't think any of it was OLPC's design or hard work. They just took components, stuck 'em together, and measured what they ended up with. The amazing thing to me is how fast ARM is coming along speed-wise without giving up on the power consumption. Which, of course, has no bearing on OLPC, and doesn't help them sell product. So if they want to sell stuff, they gotta make it sound like it's their uniqueness in the market that is making progress rather than just being another computer store selling other companies' parts simply smooshed together.

Re:Um, faster than...an 8 year old x86

[Hognoxious]

Yes. Because the measure of a computer is how much power it consumes.

Do you know what the word "per" means?

It's perfectly valid to measure efficiency in the form "stuff you get out per stuff you put in". Miles per gallon, for example.

Re:Um, faster than...an 8 year old x86

[hedwards]

You can do a lot if you're changing the instruction set that drastically. The AMD and Intel processors in use today include a lot of kludge from previous revisions which are necessary for backwards compatibility. I don't think that you need an emulator for the processor bit until you get at least to the 386 and probably not till earlier revisions. You can still run DOS on modern hardware if you wish, you mostly just don't get all the features or the RAM.

Re:Um, faster than...an 8 year old x86

[hitmark]

Benchmarks?

Re:Um, faster than...an 8 year old x86

[TrancePhreak]

According to this, Silverthorne maxes out at 2W. Dimondville is 2.5W. We don't need to put the highest end Atom chip in the machine. http://en.wikipedia.org/wiki/Intel_Atom

Re:Um, faster than...an 8 year old x86

[mikael]

I was told the ARM instruction set combined multiply and addition instructions into one, much like DSP chips do. As multiplication and division instructions are traditionally the most energy-hungry operations due to all the transistors required to implement them in hardware, it's more energy efficient to emulate them.

Even more optimisation include logical shift instructions are implement using a barrel shifter, and multiplication is implemented using Booth's algorithm on 8-bit blocks. Thus, instead of one big block of transistors to implement multiplication, it's a microcode routine. Early SPARC chips emulated floating-point instructions for this reason.

Re:Um, faster than...an 8 year old x86

[AcidPenguin9873]

The title refers to the x86 version of the OLPC, clearly.

That was not at all clear to me.

I also believe that these ARM SoCs are faster than any x86 system constrained to the same power budget.

I completely disagree. If you allow both CPUs 20W, there is not an ARM microarchitecture available today that will outperform a Core i3 in the 18W TDP, and I doubt that the Cortex-A9 (coming soon) would either. An A-15 is in the ballpark but that's a couple years away right now.

If you allow both systems 500mW, there isn't an x86 available to do a comparison. Designing a CPU for 500mW power budget with the expected (relatively) low performance is an entirely different game than designing a CPU for 20W power budget and the expected (relatively) high performance.

Re:Um, faster than...an 8 year old x86

[h4rr4r]

No Core i3 system will make it on 18 watts. Only the CPU will, which is great if you don't want to actually use it. Even at 20 watts, no x86 system can compete.

By the same power budget I meant the one the OLPC operates on. I should I have more clear.

Re:Um, faster than...an 8 year old x86

[h4rr4r]

You need to look at more recent numbers. Intel could make a CPU maybe with similar power, but we are talking about SoCs here. So you need everything in that low a power budget.

Re:Um, faster than...an 8 year old x86

[h4rr4r]

Cool now add in the FSB and the South bridge, and the video card, etc. All that stuff is on the ARM SoCs.

Re:Um, faster than...an 8 year old x86

[froggymana]

I'm not too familiar with the performance of the VIA chips , but some of their Nano processors claim to have a 100mW idle power usage

Re:Um, faster than...an 8 year old x86

[froggymana]

Oh, and this claims that it has a maximum power usage of 5W http://benchmarkreviews.com/index.php?option=com_content&task=view&id=180&Itemid=63&limit=1&limitstart=1

Re:Um, faster than...an 8 year old x86

[jimicus]

It also performs a whole lot better. In x86 terms, the raw CPU power of an ARM is like a Pentium 300Mhz.

Intel might even be able to throw together some recycled 90s technology with comparable performance and end up with something ARMs equal in power.

That was the model for mobile computing for a time (adapt outdated desktop tech).

What sort of ARM are we talking about? A 1987 ARM 2 (which ran at all of 8MHz with no cache whatsoever), a StrongARM, an ARM 610, a Cortex A9?

Re:Um, faster than...an 8 year old x86

[epine]

It also performs a whole lot better. In x86 terms, the raw CPU power of an ARM is like a Pentium 300Mhz.

For the new ARM devices, maybe a Pentium III 300, which wasn't all that slow unless bloated down by a fat OS. Note that we're talking a Pentium III 300 with a gigabyte of (faster) memory to spread its wings, not the one you remember owning with 256MB that specialized in PATA cable supervision.

Intel might even be able to throw together some recycled 90s technology with comparable performance and end up with something ARMs equal in power.

Not a chance in hell, unless you give Intel a huge process advantage. This is a myth that predates your four digit user ID.

The myth was that RISC would outperform CISC. Never happened. It was true that a small design team implementing some beautiful clean RISC would outperform a similar small design team implementing some clunky CISC. When's the last time a small design team implemented x86 at the high end?

The ISA is only half the battle. Both team have to work equally hard on the memory subsystem, bus protocol, cache coherency, and glue logic. The complexity of a branch predictor has little to do with the ISA (unless your architecture is so weird it enforces different coding practices, witness Itanium or Cell, neither of which engendered fondness).

The myth should have been that RISC would outperform CISC with a constrained power budget, but the reality distortion engines of the day were unable to burn such a low octane fuel. Doesn't show up much in your pet Photoshop filter.

Fundamental power advantage was always true of ARM vs x86. Decoding an irregular instruction set takes more power. Scoreboarding a huge in-flight register set sucks power (which x86 needs to so, because the named register set was too small). Tracking partial flag register updates sucks power. All these sucky things in x86 can be hidden, because with cleverness (aka well resourced design team) all these tricky units can do their magic in parallel on every cycle. Performance achieved is fine, heat is not.

x86 actually gets a bit of a boost from all this trickiness, since you can build some optimization into it (you already paid for the extra pipeline stages). Bonus from the huge in-flight register set is that it's very good at hiding latency. x86 always benched well on Paris-Dakar, but was less impressive on groomed track compared to the RISC thoroughbreds.

When you back out the clever furnace rooms from an x86 implementation you end up with a WinChip. Remember those? Those were essentially a 200Mhz 486 with no latency hiding and still a lot hotter than an ARM chip with comparable performance.

The performance constraints with ARM doesn't matter much these days. Performance critical functions like video transcoding or compositing can be handled by dedicated silicon. A 200 watt CPU plays chess at a 3000+ chess rating these days. For most people, a 1 watt CPU plays well enough. ARM is still underpowered for self-hosting. Are there crazy kids out there these days that can't figure out how to use a wall outlet, or don't know which end is up on a full sized keyboard?

If you think Intel can just grab some old technology from the 1990s, maybe you think you could take a Corvette engine block from the 1970s, disable a couple of cylinders, refit the carb. and get 50 mpg.

Or you could start with a cylinder block designed for the purpose.

Re:Um, faster than...an 8 year old x86

[KingMotley]

The myth was that RISC would outperform CISC. Never happened.

That depends on how you define RISC and CISC. When that was the debate the industry was much different. In many cases, it could be said that RISC actually did win, and is currently the best performer. Today's x86 chips in many ways are more like a RISC engine (micro-ops) with a translation layer put on top of it for compatibility. Now that may not meet your definition of RISC, but just about everyone will admit that many of the RISC principles are alive and well in todays microprocessors, being either a RISC with a CISC translation layer, or at the very least a RISC/CISC hybrid.

Re:Um, faster than...an 8 year old x86

[h4rr4r]

Yeah, for the CPu. Now add in the FSB, the south bridge the graphics card, etc. These ARM SoCs contain all that.

Re:Um, faster than...an 8 year old x86

[owlstead]

Funny you would say that, none of my use cases involve doing too much computational intensive stuff. Applications that really require CPU power are things like voice recognition (if this is still the case) and image processing (well, video processing and high end photographic editing at least). I don't think that the math the students have to learn will tax the CPU beyond its limits.

The number of times that I've taxed both cores of my CPU at work to the max are countable on one hand, although having idiotic settings of McAfee installed by the IT dept certainly helps. The 4 cores that came with my home system? Never! My processor is running at 27 degrees Celsius at the moment (unfortunately my fan-less dedicated GPU is running at 60 degrees Celcius all the time).

Re:Um, faster than...an 8 year old x86

[sznupi]

On the bright side, general consumers seem to catch onto the concept of "good enough" lately, even if not some slashdotters...

("unfortunately" / it's not like it's overheating, isn't it?)

Re:Um, faster than...an 8 year old x86

[sznupi]

Did you care to look at any benchmark numbers at all? Here's one not too bad comparison. With ARM chips which are already faster than Atom.

Re:Um, faster than...an 8 year old x86

[sznupi]

Certainly faster than such P3, also than Atoms (which seem to show up in some servers...)

Re:Um, faster than...an 8 year old x86

[metasj]

I suppose that by "x86 OLPC" they mean the current XO 1.5 which is powered by a 1GHz VIA chip.

http://wiki.laptop.org/go/Hardware_specification_1.5

That's correct.

Faster then any x86 based computer!

[Anonymous Coward]

Take that Intel!

Chrome, Android = Google, right?

[countertrolling]

I have a bad feeling about this.

Re:Chrome, Android = Google, right?

[Charbax]

How much is Google spending on non-profit projects per year? Billions of dollars? Wouldn't it be good if the craze that goes on around Android and Chrome could be fully utilized by a global education project instead of these things happening separately? If OLPC can find a way to make it central to the whole Mobile Computing trend to also support full productivity, full educational purposes, real usefulness and not only Angry Birds. That could be very positive couldn't it?

Don't bother with Android

[rickb928]

It doesn't solve a problem that XO has. Linux fits very well.

Windows on ARM doesn't solve any problem XO has either, and potentially causes some, like licensing and lock-in. between you and me, if we're gonna start kids off with computers in the Third World, Linux makes WAY more sense than Windows. Even more than Android. Crome is not ready, and the cloud may not be Third-World-Friendly for a long time. try not to rely on resources that are either not available, cost more than food, or can be taken away by other nations, or even their own.

If ever there was a project that leverages the maximum potential for freedom via the Internet, this is it. Really, give the kids someething they can work with and watch out. Somethings wonderful will happen.

Re:Don't bother with Android

[AHuxley]

Licensing, ad tracking and lock-in is all an Apple, Microsoft or Google has. Think of the shareholders, trusts and sovereign wealth funds that need that flow of new consumers.
The world faced this with oil and mineral rights after decolonisation.
Past generations sold out to Washington or Moscow, expect the same with MS and Google.
Linux is the perfect fit, but will it be seen as a hardware base for a night of the long install by a MS or Google as a national 'gift'.

Re:Don't bother with Android

[Anonymous Coward]

Linux kernel is the operating system because it is a monolithic kernel and not a microkernel (as GNU people likes to call it to be).

Android use Linux. Android is Linux operating system.
MeeGo use Linux. Meego is Linux operating system.
Google Chrome OS use Linux. Chrome OS is Linux operating system.
WebOS use Linux. WebOS is Linux operating system

XO computer use Linux with Sugar GUI. It has same way a Linux OS as Android, MeeGo, Chrome OS, WebOS and so on.

The Linux operating system has proofed that monolithic operating systems has not died like almost everyone told to happen in early 90's and the Server-Client architecture would be only correct OS architecture to be used.

Linux has conquerored the whole world, but not in front of the people, but behind the marketing forces. We have TV's, DVR's, ADSL/Cable modems, cellphones, smartphones, desktop computers, laptops, servers, supercomputers, rest of the embedded systems etc etc controlled by Linux OS.

Linux is something what FSF never got work with HURD operating system. So they needed to start a black campaining against Linux. MS fears Linux so it has already learn it is better to be a quiet than even trying to fight against it.

Re:Don't bother with Android

[BitZtream]

if we're gonna start kids off with computers in the Third World, Linux makes WAY more sense than Windows

Not really. It only make sense from the incredibly short sighted financial aspect only.

From a 'which is more likely to get someone a job', Windows is far more useful in most places. And lets face it, it may be 'one laptop per CHILD' ... but mommy and daddy are going to take advantage of it too if thats the only computer in the house.

Freedom is great, right up until you starve. Please stop trying to use to throw 'free'dom into it, its as retarded as those 'freecreditreport.com' commercials ... with enrollment in triple advantage for 12 months.

Re:Don't bother with Android

[naz404]

Hi. Just reminding everyone that the current default OLPC OS now allows dual-booting to Fedora+Gnome aside from initially booting into the kid-oriented Fedora+Sugar desktop environment, making it suitable for more traditional uses by older users as well as being capable dev machines.

http://wiki.laptop.org/go/Release_notes/10.1.3 Will be posting this multiple times here (please don't mod as redundant) as Slashdotters really need to be made aware of this fact.

Re:Don't bother with Android

[rickb928]

I don't see the XO as an option for regions where 'get someone a job' is the most likely and desireable outcome. Mind you, I have NO idea where all the XOs go, but it was at least intended that they go where there is no electricity, no infrastructure, not much at all. That is the power of seeding computers into places where there is so little.

So in that light I don't see 'get a job' as so important as 'get an education', or 'encourage ideas'. Yes, many of these will result in getting a job, and that's good, but it's not the only good outcome, nor is it perhaps the best one, depending on circumstances.

But if getting a job is a primary goal, then how does a Windows lock-in do better than a Linux lock-in? Both offer excellent development environments, both have large markets, substantial need for applications, etc. Windows will force an XO user to repidly confront licensing and costs to expand much beyond the intial XO experience, while Linux can be done virtually for free, though distribution is the most likely cost of acquiring new software, IDEs, etc.

And if you can program in or leverage Linux, you'll do just fine with a Windows machine and a fairly short orientation period.

The XO is not intended for 'most places', so far as I knew, but if its appeal is broadening the market, great. Sounds like we may want more than one sort of XO. This is good.

ps- The Zulu will figure out how to transition from Linux to Windows. So will many people. I figured out how to transition from Windows to Linux, which seems much more difficult, even in hindsight. Now I can straddle both worlds. This would be good for nearly every XO audience.

Same chipset as the Marvell Moby reference design

[davide marney]

Gigahertz-class CPU, integrated full HD 1080p encode and decode, 6MP image captures, integrated audio processing engine, advanced 3D graphics. Renders 45Million triangles-per-second. Includes 802.11n wireless, Bluetooth 3.0, HDMI, USB 2.0, 3G Baseband, SD/MMC card, and camera. It is powerful enough to simultaneously decode 4 1080p video streams at a time. Some videos of an early reference design here: http://www.youtube.com/watch?v=1s17KwfzTFY

So poor kids are using it...

[Targon]

Poor children and those in third world countries generally are not customers who would be spending money. This is a key point to this whole issue, where the idea that just because there may be thousands or even hundreds of thousands of users does not mean that there is a lot of money that can be gained from that market. Two billion of these machines will still not end up as profitable for software developers as two million regular PCs running MacOS or Windows for that reason.

Re:So poor kids are using it...

[Fri13]

Actually they will but not in next quarter. Companies need to wait that children grows up and they start searching software or other products for their own computers.

Companies should do same thing what tobacco companies did what was to make a toy cigarettes and similar things for them to play. Later when children was a teen and wanted to try smoking, they toke the familiar brand.

The tobacco companies understanded the long run marketing and branding what will give big profits in years to come.

But computer world lives only in quarters and they do not understand that the brand needs to be made so strong that child remembers it when they are a teens.

Re:So poor kids are using it...

[Cwix]

Sweet candy computers!!

Why android over standard Linux?

[chipwich]

As I've watched Android dominate the tablet market, I'm bothered by the fact that these devices do not give root access without "jailbreaking". Isn't Android a major step toward the very scary world of "Trusted Computing"? That is, the hardware manufacturer, government, or whoever else has power can deny the ability for a user to run a program (or all programs!) at whim. Right out of the box, the user is denied permission to use their hardware in the way that they see fit.

I feel much more comfortable with a full Linux distro that empowers its users, rather than makes them comfortable with someone else holding the keys to their machine. Besided, android hardly seems compatible with the "open" goals of OLPC. A full distro would take advantage of a real JVM and a much richer software eco-structure instead.

Re:Why android over standard Linux?

[Charbax]

Archos officially allows you to root your Archos Android Tablet using a "Special Developer Edition" firmware. Thing is, for official root support on such compact embedded devices you have to loose the warranty. At least for now. For example, you could write code that constantly over-clocks the processor or abuses something else in there and that can harm the hardware, so manufacturer is not able to support that.

Re:Why android over standard Linux?

[mattack2]

Android dominate the tablet market? Where are the sales figures that back that up?

Re:Why android over standard Linux?

[Fri13]

He meant by number of models, not by number of sales. As sales iOS has over 95% market share.

Re:Why android over standard Linux?

[mattack2]

That's what I thought... So I think 'dominate' was incorrect to use in that situation.

Re:Why android over standard Linux?

[owlstead]

Android *by itself* does not restrict you to anything. The thing may be that Android is *too* open, as it gives a lot of power to the manufacturer and network operators. I can very easily install proprietary firmware on my HTC Hero phone and I can use it as a modem out of the box. As a tech savvy user I asked my network operator before buying though, this is not something your average consumer is going to do.

MeeGo?

[Kryptonut]

What about MeeGo? Already runs on the Nokia N900's ARM processor.

You aren't the target market

[jeff4747]

OLPC is targeted at...shockingly enough...children in the third world. Where you don't have power outlets scattered around the house every 12'. As such, low-power is a critical requirement.

If they were rich enough for a power grid, they wouldn't need the aid.

Re:You aren't the target market

[sznupi]

I would wish them to not repeat some of our mistakes / power usage efficiency is a good thing (alas, a form of post-colonial mentality - valuing wastefulness enabled by prosperity, seeing it not as outcome but as one of the reasons - is too prevalent...)

YEAH !! RIGHT !! WHAT A TARGET MARKET !!

[Anonymous Coward]

Just like linux, the third-world and dirt-poor. Just think, I lose $32 a unit but I make it up in volume !! SHeez, no wonder Linux users are STUPID users !! Mod this if you agree !! Moooohahahahahha !!

Dump Lin-sux and Lamedroid: ask Apple for iOS!

[Anonymous Coward]

One upon a time, Apple offered the OLPC people *FREE* copies of OS X and like the MORONS that the OLPC folks are, are they turned Apple down (and just look at how Lin-sux KILLED the OLPC project). Well here's an opportunity to dump the crappy quality of "open sores" software and go with something that RULES THE DAY. Just be sure to ask very very nicely because you IDIOTS turned Apple down once, but now you are desperate.

Re:Dump Lin-sux and Lamedroid: ask Apple for iOS!

[Charbax]

Apple can install Mac OSX on any OLPC XO laptop today and have always been able to, just to prove that they can make it run. OLPC laptops are 100% fully dual-bootable, just put any other compatible OS on the SD card and reboot. To this day, I don't believe Apple has provided any compatible version of OSX for these laptops. OSX is closed source, only Apple can come and provide a compatible "Light" version of their software to run on the machines. Apple prefers to keep their money at doing other things.

Re:Dump Lin-sux and Lamedroid: ask Apple for iOS!

[Cwix]

Lol obvious troll is obvious.

Re:Dump Lin-sux and Lamedroid: ask Apple for iOS!

[Anonymous Coward]

Sure I'm "trolling" with the tone, but you KNOW that everthing I said is true. First, that Linux KILLED the OLPC project. Second that putting iOS on the OLPC would make it a FAR FAR FAR more successful product, bringing UNIX-standard computing to millions of more kids who will otherwise go without due to the projects bizarre obsession with terrible "UNIX-alike" software that is Linux.

ARM Powered OLPC XO-1.75 Laptop Is Faster Than X8

[Anonymous Coward]

And here I thought x86 referred to an instruction set or cpu architecture or some such. Can this amazing laptop also leap tall buildings in a single bound? I think the OLPC project is great, wish I could say the same for slashdot summaries...

Negroponte, please

[Gothmolly]

Why again should the developed world hand out laptops to the rest of the world?

Re:Negroponte, please

[Ignacio]

Because it can raise their level of civilization, which in turn will allow the developed world access to their resources and will allow them to buy products from the developed world.

And solve world hunger, cure cancer, and all of that, but no one cares about that bit.

Wrong Target.

[crhylove]

I don't want Windows. I don't want Sugar. I don't want Fedora. I don't want Ubuntu. I don't want Anroid with their crappy market.

I want Linux Mint. It's faster, more stable, and more feature filled than any of those OSes out of the box. Dead simple, my mom was even a convert, and it is rock solid. I put Mint on a machine, and never get a tech support call back, which is exactly what I want.

Mint and Forget. And in this case I mean forget the other operating systems. Linux year of the desktop should be 2011, and it should be Mint version 10 which is incredible.

Don't flame me or troll me until you've installed it on 3 or 4 machines. It will shock you. I literally haven't hunted for a driver since the new mint came out. Not one. On about 20 different machines.

The only post format chore I have to do in Mint is make video files default to VLC, change the shortcuts a bit in the start menu, and install audacious and delete rhythmbox. It already has Firefox, Open Office Write, Brasero, Pidgin, and almost every other program an end user needs. Oh, except for Skype. I have to install that often as well.

Re:Wrong Target.

[Charbax]

You are welcome to get a XO development machine and show that Mint works great on it.

Re:Wrong Target.

[Anonymous Coward]

By the time I've bought and installed 3 or 4 machines, 12 years have past. That's not very 2011.

Re:Wrong Target.

[Anonymous Coward]

There is no Linux Mint for netbook! You'll have to install Ubuntu Netbook Edition.

Re:Wrong Target.

[Anonymous Coward]

You are welcome to get a XO development machine and show that Mint works great on it.

He won't be able to as it is just another crappy Ubuntu distro variant

Re:Wrong Target.

[riyad.parvez]

probably you forgot linux mint is a derivative of ubuntu, and doesn't add any extra feature or stability except adding some custom packages and proprietary codecs

Re:Wrong Target.

[crhylove]

I know precisely the difference. Clearly, you did not take my advice and try 3 or 4 installs before commenting. You should Mint is a much, much superior experience to Ubuntu. When you switch you won't go back. Not only that, the latest versions have a Debian base that is superseding Ubuntu altogether. For all the obvious reasons.

Re:Wrong Target.

[Anonymous Coward]

Poulsbo chipset.
'nuff said.

Re:Wrong Target.

[BitZtream]

I want Linux Mint. It's faster, more stable,

So in order for your statement to be true, Mint must be using some completely different kernel than all the rest that is mysteriously better.

I some how find that hard to believe, which makes you either completely ignorant of the OS your fanboying about, or a really retarded/bad troll.

Second, some douche in his moms basement who already has four or five machines in his house isn't the target of the ONE LAPTOP PER CHILD project now is it?

OLPC isn't a design contest so you can have something to jerk off with, you are not the target. They don't have the wrong target, your just too self centered and ignorant to realize the world doesn't revolve around you and what you want, which is pretty typical on slashdot so don't take it personally.

Re:Wrong Target.

[Anonymous Coward]

Mint is Ubuntu with few more packages. If you don't believe me check your /etc/apt/sources.list

Re:Wrong Target.

[crhylove]

I regularly install linux mint on netbooks. It by default has a smaller foot print, less screen real estate, and runs excellent. In fact, that is the primary platform I install mint on.

Re:Wrong Target.

[crhylove]

Your assumptions about me were all wrong. I do IT for a living. Mom's basement is on another continent. LOL And most importantly, kids can use Mint. It looks pretty much like Windows, and I know tons of kids that can operate windows. Making them learn a foreign interface with no bearing on actual computer work is doing them a disservice.

OLPC should broaden it's target. Impoverished people of all ages the world over need a cheap reliable machine. Linux Mint and a good piece of hardware kit would nicely fill the bill.

You haven't tried Mint at all yet, have you? It's not the kernel that is superior, it's every other part of the OS. You'd know that if you had tried it out.

YOU are the troll. And here I'm feeding you.

Re:Wrong Target.

[crhylove]

A few DIFFERENT packages. All of them a better choice for almost everybody. Plus the Mintmenu and default layout are vastly superior to every version of Ubuntu I've tried.

Re:Wrong Target.

[Anonymous Coward]

Nobody cares what you want. What matters is what's best for children's education.

Re:Wrong Target.

[Anonymous Coward]

Agreed!!! :-)

Re:Wrong Target.

[Anonymous Coward]

doesn't sound terribly different from my 6 years of ubuntu, except for a larger community to support the real nerd goods.

Re:Wrong Target.

[FranTaylor]

It's faster

Various articles I've read indicate that the kernels in the various distributions are very very close in performance, so you must be speaking of user land.

Well then your problem is that one can strip away as much as you want with many distributions. For example with Fedora you can do a minimal text install that boots up almost instantly. Or you can install as much or as little as you want and your startup time will increase accordingly.

more stable

When you have a situation like OLPC, the developers are tweaking the kernel hard for the particular hardware and it is going to be rock solid. I would expect OLPC to be as robust as any approved RHEL or SLES installation.

The only post format chore I have to do

Again with a distribution tailored to the hardware, you don't have to do any of this kind of stuff at all.

Re:Wrong Target.

[SwedishPenguin]

But what does it matter when the software comes preinstalled? It's not like they're going to install and configure each and every laptop individually when they can install and configure one laptop as they want and then dump the image onto all the others.. I'm not saying Mint is not better out of the box than Ubuntu, but when it comes preinstalled and preconfigured, the distribution does not matter much.

Re:Wrong Target.

[crhylove]

It certainly matters. Having a real OS and real tools will make it a real computer. Having some weird experimental toy software (Sugar) that even *I* can't figure out (>20 years professional IT) makes it a toy. If this was a Linux Mint laptop, kids, adults, anybody could get online, surf the web, write some apps, whatever they wanted to do.

Re:Wrong Target.

[SwedishPenguin]

I agree that you would want root access and proper tools and such, but what I'm saying is that it doesn't matter much which distribution is used when the computer is preinstalled and preconfigured with that distribution, and by distribution I mean the likes of Ubuntu and Arch, not Android which has a completely different userspace from a standard GNU/Linux distribution.

Not an Intel initiative

[Troll-Under-D'Bridge]

Intel's Asus Eee PC Atom platform was a direct reaction to OLPC's hype...

It might not be your intention. But by prefixing EeePC with the possessive (Intel's), you make it sound as if Intel was the primary mover behind the introduction of the EeePC. Intel had its own take on the OLPC. This was the Classmate PC, described thus by Wikipedia as "Intel's entry into the market for low-cost personal computers for children in the developing world". The Asus EeePC was, well, Asus's attempt to cash in on what Asus saw as a trend toward smaller form factor computers (which continuing with the increasing popularity of tablet computers, CPU/GPU "fusion" chips, and mini-ITX motherboards). The EeePC originally had a Linux-based OS produced by one of the lesser known distributions, the semi-free Xandros.

Re:Not an Intel initiative

[Charbax]

Sure mr troll under the bridge. Intel has no relations with OEMs like Asus whatsoever, and Intel also manufactures the Classmate all by themselves, not by any chance in the same Pegatron factories that make Asus Eee laptops. Asus Eee was a staged marketing stunt orchestrated by Intel for Asus to block OLPC sales, go look back.

XO By The Numbers

[westlake]

If OLPC XO-1 threatened Intel enough to start the netbook market and has reached two million poor kids in third-world countries thus far, XO-1.75 may help start the ARM-powered Linux laptop market.

Deployment of XO laptops

Global: 1.8 million

Latin America: 1.5 million

Peru: 870,000
Uruguay: 460,000
Columbia 65,000
Argentina 60,000
Mexico 50,000

Africa: 135,000

Rwanda: 120,000

Asia: 24,200

Oceania: 10,000

Australia: 5,000

The geek has some explaining to do when his allegedly potent combination of durable, cheap, laptop hardware, FOSS software and constructivist philosophy of education finds almost no acceptance beyond a single language, region and culture.

As for resuscitating Linux-on-the-Netbook, it isn't easy to make to make the case that OLPC has been a significant force for a broader adoption of Linux anywhere on the planet.

Mobile vs Desktop - South America
  Top 5 Operating Systems - South America

Re:XO By The Numbers

[Charbax]

This is essentially 100% of the developing world kids to have gotten any laptops at all.

Touch screen

[kentsin]

Anyone make a kit for touch screen for XO 1.75?

Forget portable computing

[Trogre]

I want one of these as my home 24/7 web/mail/project server. My power bill would be grateful.

Re:Forget portable computing

[gl4ss]

http://www.arm-board.com/

go nuts.

Re:OLPC?

[gnola14]

Fucking troll

Comment removed

[account_deleted]

Comment removed based on user account deletion

Fedora vs Android

[Anonymous Coward]

I have not used Android, but have been a staunch Fedora user for the past 5 years. Fedora is stable, even though it is a developement Linux, It is well written, almost legacy in architecture. Android is brand new, and while it may be fine for cell phones, we dont know the following:

a) How good is it for laptops and desktop systems

b) How bug free is it. Fedora has more than 5 years of solid codebase behind it, Android is still teething. And with regards to security Fedora is quite far ahead. Other thoughts too, less fear of some unknown patent infringement penalties, especially if you live in the USA.

equally, a chance to f up

[r00t]

They've got a rare chance to design an interface for people who don't already have expectations of how to use a computer.

There is the risk that the new UI can't ever beat the standard one, the risk that it won't because actually implementing it well is astronomically difficult (how many people and years have gone into each of the normal UI implementations?), the certainty that apps will be badly ported or wrappered, and the cruelty of wasting people's time on a UI that not even its own developers will tolerate.

When the developers and some unrelated non-developers start using a new UI exclusively, then we can rightly begin to consider shipping it to other people.

Re:equally, a chance to f up

[naz404]

Hi. Just reminding everyone that the current default OLPC OS now allows dual-booting to Fedora+Gnome aside from initially booting into the kid-oriented Fedora+Sugar desktop environment, making it suitable for more traditional uses by older users as well as being capable dev machines.

http://wiki.laptop.org/go/Release_notes/10.1.3 Will be posting this multiple times here (please don't mod as redundant) as Slashdotters really need to be made aware of this fact.

Re:equally, a chance to f up

[r00t]

Children deserve better. Now we need:

1. fix GNOME to be less of a confusing mess (for example, the double taskbar/menu things are 100% design by committee)

2. ditch the initial boot into Sugar to improve speed and reduce confusion

3. save space by making Sugar an optional download from some obscure site

no, we get it

[r00t]

If you want us to code for free though, you have to push Linux.

Don't want to push Linux? OK, fine, we cost $50,000 to $200,000 per year generally. That's not counting employer contributions to health insurance, 401K (matching), life insurance, disability insurance, unemployment insurance, social security taxes, dental insurance, etc.

IMHO, tolerating our Linux zeal is extremely cost-effective.

Re:no, we get it

[LingNoi]

You don't speak for the whole open source community.

Re:no, we get it

[r00t]

OK, some would prefer OpenBSD or Hurd.

An "open source community" or an "open source advocate" is, no surprise, all about the open source. Only a tiny subset of these people will greatly care about some other random worthy cause.

You may only care to "further education and creativity", but that won't get the software written. Random ordinary teachers and creative people won't get the software written. You need software developers. You can pay them, or you can offer them the chance to push the software that they are passionate about. Your choice.

You're being offered an amazing deal. It's like getting -- very roughly -- a hundred million dollars of engineering effort for "free". All you have to do to get it is promote free software.

Re:no, we get it

[LingNoi]

Random ordinary teachers and creative people won't get the software written.

Well lets see now.. There is the:
One laptop per child project
http://hfoss.org/
Arduino was started by a bunch of teachers and lecturers.

but I guess these people don't qualify in your opinion because it would go against your whole argument. woops.

Re:no, we get it

[tehcyder]

So you seriously believe that someone who is earning $200K+ a year will give that up and work for free in a good cause?

Re:no, we get it

[r00t]

I was going to say you'd found the "tiny subset" that I mentioned, but then I followed your link. I see...

"computing faculty and open source proponents at Trinity College, Wesleyan University, and Connecticut College."

OK, these aren't ordinary teachers and creative people. They are computer science professors. In other words, they are software developers.

"using Free and Open Source Software (FOSS)"

Oh look, they do care about software freedom. They won't be happy with Windows.

"We are funded by the Directorate for Computing & Information Science & Engineering (CISE) of The National Science Foundation (NSF)"

Heh, they aren't even doing the work for free!!!

Re:no, we get it

[r00t]

A few have actually done that.

Many have given up a portion of their time, myself included. (not working overtime that was available) I do this for Linux and maybe for kids in the USA.

Sometimes people take a few months leave.

you forget that software is like a gas

[r00t]

Software expands to fill the hardware, and then some more.

I'm thinking they could start a full-screen XO-1 emulator at boot. Naturally the emulator would be written in a scripting language with duck typing. I'm sure this would be educational.

laptop==IED, so the kids lose out

[r00t]

The adults will take the laptops. Most likely the laptops will be intercepted before they reach the kids, with laptop activation being subverted by corruption and violence.

The laptops are perfect for triggering bombs. They have cameras, microphones, DC input, and mesh networking.

Countries like Afghanistan have problems that run deep. It's not like a poor and uneducated version of a 1st-world country.

No chrome OS

[Neil Boekend]

Chrome OS would be quite useless because large areas of 3rd world countries do not have a reliable internet connection.

This is something people seem to forget

[Sycraft-fu]

ARM has not discovered some magic loophole in the space-time continuum ala Star Trek to be able to produce chips that are much faster than anyone else, and then for some reason refuse to use that to produce high end chips for the markets that want them (like supercomptuers).

What ARM is good at is making low power chips in both senses of "power." They use little electricity to do what they do, and they do not perform all that many calculations per second. ARM's high end is around Intel's low end. That's great, there's a market for all kinds of stuff. Some devices need small, efficient chips, other need big powerful ones. I wouldn't want a Sandy Bridge in my cellphone and more than I'd want a Cortex based chip in my desktop. There's room in the world, and indeed a need, for both.

I think people on /. are a little quick to get on the ARM worship. They see the low power numbers the ARM CPUs get and think that means that a big ARM CPU could be just as powerful as Intel's stuff and extremely low power. No, probably not.

Re:This is something people seem to forget

[gl4ss]

well, they don't remember the same thing happening in the '90s.

what would be an interesting test, sort of, would be to measure the amount of electricity used and the speed which they can raytrace a scene or encode a video.

Poor summary.

[seeker_1us]

What the video says is that the ARM chip is more powerful than the old x86 chips that OLPC used. Considering it's a SOC, and much newer technology (the original processor was 130 nm node, and the Armada is 65 nm node) it really should be.

Wrong customer

[spage]

You don't appear to be an educational facility contemplating a deployment of "one laptop per child" or a child in a developing country, so I sincerely hope OLPC doesn't give a damn about what you want.

People have ported Debian and Ubuntu to the current XO hardware, see http://wiki.laptop.org/go/Category:Linux_distributions

Android would be a very wrong choice.

[JustNiz]

Android: commercial product tied to one company. Includes marketing-based artificial dependencies already built-in like App-store, DRM etc. I know Google/Android are 'cool' at the moment, but at the end of the day they are a business and are ultimately in it for the money. Superficial business tactics aside, they are just another Microsoft or Apple.

Linux: non-commercial product, independent, altruistic, good free community support via the internet, no artificial functional limitations or dependencies baked-in by marketing types. Many more professional-quality apps available for free download than Android. Even has licenses to protect its 'freeness'.

You tell me which would be a better choice for poor kids and their family budgets.

What is the goal of OLPC

[jbolden]

I think this comes down to asking what the goal of OLPC is.

1) To build cheap hardware usable in 3rd world
2) To build cheap hardware usable in the 4th world (i.e. not electricity).
3) To figure out a way to do logistics for semi valuable property in the worst of the 4th world
4) To create an optimal interface for children in 3rd world countries where a computer infrastructure doesn't already exist.

These goals unfortunately conflict.