Ivan Krstić: OLPC Doesn't Have a XO Laptop Deployment Plan


Wow. Reading Ivan Krstić's 4,400 word manifesto on OLPC, Sic Transit Gloria Laptopi, you can tell that, by his own admission, he is angry.

bitfrost ivan Krstić
Ivan Krstić in happier days

He calls out both the Free and Open Source Software community and One Laptop Per Child for their dueling thoughts around software that is distraction from the overall goal - educating children. Then, he confirms what I concluded long ago about Nicholas Negroponte's view of OLPC's mission:

In fact, I quit when Nicholas told me - and not just me - that learning was never part of the mission. The mission was, in his mind, always getting as many laptops as possible out there; to say anything about learning would be presumptuous, and so he doesn't want OLPC to have a software team, a hardware team, or a deployment team going forward.
But its when Ivan talks about deployment that I really get scared. His description of the OLPC implementation plans (or lack thereof) was the exact reason I started OLPC News so long ago - my great fear that this very scenario would come to pass:
Other than the incredible Carla Gomez-Monroy who worked on setting up the pilots, there was no one hired to work on deployment while I was at OLPC, with Uruguay's and Peru's combined 360,000 laptop rollout in progress. I was parachuted in as the sole OLPC person to deal with Uruguay, and sent to Peru at the last minute. And I'm really good at thinking on my feet, but what the shit do I know about deployment?

Right around that time, Walter was demoted and theoretically made the "director of deployment," a position where he directed his expansive team of - himself. Then he left, and get this: now the company has half a million laptops in the wild, with no one even pretending to be officially in charge of deployment. "I quit," Walter told me on the phone after leaving, "because I can't continue to work on a lie."

Best of all is Ivan's money quote for all of us who love the idea of technology as a catalyst for change, love the clock-stopping hot technology, but yet also worry about the impact such a high-profile project has on the entire technology for development movement:
The real laptop challenge
That OLPC was never serious about solving deployment, and that it seems to no longer be interested in even trying, is criminal. Left uncorrected, it will turn the project into a historical fuckup unparalleled in scale.
That is exactly the fear that drove me to get medieval on OLPC's ass for the first year of this humble site. I've mellowed in time, not due to any great strides by OLPC in its abilities to lead, follow, or even get out of the way of others, but by the amazing success of independent efforts like el proyecto Ceibal, OLE Nepal, Teaching Matters, Waveplace Foundation and despite OLPC's best attempts to muck it up, Give One Get One, which for the record, is still the largest deployment of XO laptops to end users.

Ivan concludes by touching on a thought that a few of us are thinking hard about these days. With the coming plethora of 4P Computing options, platforms that focus on performance, power, portability, and price factors favorable for educational deployments in the developing world, there is no need to be monolithically focused on any single platform.

The real need is in educational software and content for those platforms, and the deceptively tricky act of deploying them at scale. It's just too bad that Ivan doesn't think OLPC is even going to attempt either. That in itself is a historical fuck-up paralleled all too often by lesser organizations.

Related Entries


Who knew what, and when?

Or the news is, what took it so long to be made public?

"when Nicholas told me - and not just me - that learning was never part of the mission." I guess the problem here is in defining learning. I agree with Nicholas' assessment, even though wonderful work has been done as to software for exploring and creating. But let us not despair. As long as he provides the machines, we can _and_should_ provide the learning components, which I know some people are intently working on, even though we do not have the Intel billion.

> With the coming plethora of 4P Computing options, platforms that focus on performance, power, portability, and price factors favorable for educational deployments in the developing world, there is no need to be monolithically focused on any single platform.

This is all very true. But don't go too far and dismiss the XO as the wrong tool for learning because you feel cheated by Negroponte. The XO is still pretty much the best tool for the job that I can imagine.

Not that anybody said anything else. Just trying to get people back onto the ground before this gets too emotional.

Otherwise: Pretty much exactly my thoughts, very well presented by Ivan.

Don't get me wrong, Heng, I too think the XO is still the best tool for a certain job - educating primary school children in the developing world when a country can afford a 1:1 program. Yet the XO does start to loose some luster when those variables change.

And if we take Negroponte's approach - its all about laptop sales - then the hardware becomes just one component in a full solution set, of which OLPC has little in comparison against others.

Very interesting writing by Ivan Krstic...however, as the cliche goes, "it raises more questions than it answers".

In trying to comment on Krstic's half-truthful statements, I find myself in the highly unusual position of having to defend Negroponte. I'll try to be brief.

1. Unless Krstic is mentally retarded, it's very hard to believe that he didn't notice the obvious absence of an implementation plan or an educational strategy until...6 months ago!

2. Krstic very conveniently glosses over the fact that him, along with Walter Bender are a very important part of the team that delivered a deeply dysfunctional mess called "Sugar" and that a long time will be needed - with or without Negroponte's involvement - to fix all the current issues.

The truth is that all the internal warfare at OLPC began six months ago when MaryLou Jepsen got the seet licensing deal to make money off the work paid for by OLPC. Now, as it becomes increasingly clear that Negroponte's project is dead, the usual suspects have no incentive to continue being part of Negroponte's "big lie", as Bender and Krstic call it.

Pretty sleazy stuff from everybody, to be honest. Yes, Negroponte is a compulsive liar and an unethical person. I have said it for 3 years and I have been called all sorts of names for telling the truth (but that's ok - it feels good to be proven right :-) ). But it is highly hypocritical of Krstic, Bender and others to pretend that they were not willing participants in Negroponte's scam.

Krstic correctly points out that if the enabling technologies were separated from the UI and ported to other platforms we'd have a very Good Thing in terms of a cross-platform environment embodying tools for developing truly useful educational software.

What he forgets is that the community of developers which he hopes will provide this software using the Presence and DataStore API *has no idea how to do this because the API documentation doesn't exist*!

At least Microsoft publishes its API. I'd rather write code in C++ for Microsoft .NET than write in Python while trying to *guess* how the Sugar API is supposed to be used!!!

On a parallel subject, how not to prepare for things, the G1G1 problems encountered and how the support team was overwhelmed by having to deal with between 1-10% of the G1G1 problems and what those problems were, Kim Quirk has posted a note with graphs on this, soon to be on a wiki page?

"... it is really important for us all to see the metrics from the RT
system to answer the kinds of questions that everyone has: How many
questions are on getting started, on wifi, on donor services (where is my laptop?), etc.

I have attached some pictures of metrics... but it might be more informative to put that info on a public wiki page so we can get trends over time, etc.

A few things about these stats:
1 - these stats are almost entirely from the Give One Get One donors.
2 - these don't represent all questions as there was a partner taking 99% of the phone calls and about 5-8 times as many email questions as our RT system.
3 - we did these stats WITH the donor services questions (where is my
laptop?) and without those questions, which represents 'tech support'

About the need of educational software, the cost of porting an application from one platform to another one is not negligible. Until now, it looks to me Squeak/Smalltalk is a very interesting solution in respect to portability, quality of the language and IDE. For those deducating free time to develop free educational software, it is a very valuable bonus.

This is exactly what I found when porting DrGeo (http://wiki.laptop.org/go/DrGeo) from C++/GTK+ to Squeak/Smalltalk. Thanks to the Squeak/Smalltalk portability, even if OLPC or Sugar collapse, I can port it to many platform at a minimum developer cost.
I will be dead meat, if I decided to port to Python+Sugar. (Well, I would have allocated free time for a re-port effort...)

Really a big bonus...

Ol' Ivan certainly tees off on a variety of subjects but the one that jumped out at me was his pocket review of the Dakar experiment - it flopped - and what that flop implies.

The explanation given is that there were personality conflicts and mismanagement but that explanation hardly satisfies. If constructionism is such a worthwhile means of engineering education why wasn't another attempt made or rather, why weren't multiple attempts made each failure acting as a guidepost for the next attempt?

Ivan links to an article about the Hennigan school but that article's over ten years old. That gives us two data points. What's happened to Hennigan school in the intervening twelve years?

Greatschools.com suggests that however good constructionism is, it isn't good enough to vault a school out of the middle of the pack. Not the stuff of revolutions.

If aerodynamics had followed the same course of development we'd still be arguing about which wax works best to stick feathers to our arms.

Ivan also refers to constructionism as a theory but that's hardly an appropriate use of the word. A theory's the result of the formulation of a hypothesis, the testing of the hypothesis and agreement between the hypothesis and the expermental/observational results. So where are the experiments/observations that confirm the validity of the hypothesis of constructionism? There aren't any so how come constructionism is referred too as a theory? A more appropriate term would be "school of thought" but that doesn't have quite the scientific cachet of "theory".

This observation about constructionism provides a nice explanation for the reticence on the part of the OLPC organization towards pilot programs. Adherents to a school of thought aren't troubled by failure because they're always due to personality conflicts, mismanagment, lack of commitment or the direction of the prevailing winds. But adherents aren't unaware of the way non-adherents view failure; they're quite likely to view it as a failure of the school of thought and that's not acceptable.

So calls for pilot programs/experiments are met with insults, obfuscation and circumlocutions. When that's no longer good enough you start to prepare for the inevitable failure by creating distance between the experiment and the adherents. Hence, the Dakar experiment didn't collapse because there's no "there" there with regard to constructionism but because of mismanagement and personality conflict. The school of thought and it's espousers are free of the taint of failure and when the dust has settled it's time to start peddling the idea again.

When the rubes start to catch wise you shut down the con and slip out of town against the day the con can be run again.

The singular accomplishment of the OLPC may turn out to be the establishment of the low-priced laptop as an viable commercial market segment.

Manufacturers are drawn to the higher-end of the price spectrum because it's easier to make a profit selling a $2,000 laptop then a $200 laptop but competition cares nothing for the preferences of the manufacturers. The successful $2,000 laptop begets the succesful $1,900 competitive laptop. That natural progression may stall for various reasons and various lengths of time but over time the direction is clear and the movement inexorable.

The OLPC kicked off a scramble to populate the low-price end of the market and that may be its lasting legacy.

Intriguing post. It could be that Ivan is calling it a theory in the colloquial although it doesn't hurt to remind people that it (theory) has a well defined and precise meaning. Note that there doesn't seem to be much in the way of counter experiments either (this reminds me a lot of anarchism, what few attempts have been made at it have been seriously hampered by outside forces - indicating that some additional thought is required to deploy ...:-)

It may be that just stirring the educational software pot has helped as well. Squeak is certain to get more exposure because of this, the Epaati activity may make an effective model for other local content authors - the fat lady is nowhere near singing in this opera. We've yet to hear from MLJ and WB is still quite active on the OLPC development lists.

This is what I was saying for a long time -- here and elsewhere:

"Choosing to reinvent the desktop UI paradigm means we are spending our extremely overconstrained resources fighting graphical interfaces, not developing better tools for learning. … It is most important to recognize that the graphical paradigm changes are inessential both to our core mission and to the Sugar core ideas."

Everyone under-estimates the difficulty of making a desktop environment. It is a huge ego-trip that has killed more then one company.

They should have just used one of the many light & free desktops, like xfce.


> A Windows-compatible Sugar would bring its rich learning vision
> to potentially tens or hundreds of millions of children all over
> the world whose parents already own a Windows computer, be it
> laptop or desktop. To suggest this is a bad course of action
> because it’s philosophically impure is downright evil.

Well those people that suggest "Sugar compatibillity on Windows is wrong" are exactly the people Nicholas Negroponte calls "Open-Source fundamentalists". So to me, this just looks like Nicholas Negroponte and Ivan Kristic agree.

> while Ivan Kristic is unequivocally enthusiastic about Sugar
> being ported to every OS out there, he is absolutely opposed
> to Windows as the single OS that OLPC offers for the XO.

Well from what I have heard, Nicholas Negroponte only talks about Windows XP booting from an SD card as a dual-boot OS. That requires of course that the SD card reader is just as fast as the internal memory so that the performance would be just as good booting from the SD card.

Of course nobody, not even Nicholas Negroponte can stop Microsoft or states that puchase the XO-1 laptop from choosing to install Windows XP exclusively (it could be through money under the table by Microsoft or justifying it by any other choice such as saving $4 on the 2GB SD card with Windows XP to dual-boot), then I don't see why that scenario should stop OLPC from anyways supporting that scenario of the Windows-only XO laptop.

Of course OLPC I think will always prefer to pre-load the fully open-source and fully optimized and constantly evolving Linux Sugar OS by default on the internal memory of the laptop. Sugar Linux runs pretty well already, and still a bunch of optimizations such as better power management, standby/resume support, autonomous WiFi mesh packet forwarding, improved ebook interfaces, new and optimized educational activities.. Of course Sugar Linux I think will most probably remain the DEFAULT OS of the OLPC project.

All the work by 40 Microsoft employees over the past two years to provide an optimized Windows XP for the laptop is GREAT, this makes it possible for all the kids to experience the kind of OS we have been using for the past 15 years in the developped countries. And Sugar Linux being there on the laptop by default, makes it sure that Microsoft will be pulled in the right direction to do all the right choices to bring cheaper and better laptops to all the people of the world.

Without OLPC there wouldn't have been the Asus Eee, without OLPC there would be no Intel Atom processor project, without OLPC there would be no $3 Windows XP licence to the third world, without OLPC there would be no industry wide recognition (Asus, Acer, Gigabyte, Quanta, HP, Dell, VIA...) of this new ULCPC product segment. So already OLPC is in my opinion the biggest success ever in this laptop industry. but there is still a huge way to go, now the laptops need to be put in the hands of the 5 billion people who don't have access to computers and the Internet. In my opinion, OLPC is the best at pushing the whole industry in the right direction, and I expect to see OLPC accellerate it even faster with the cheaper ARM-based XO-2 coming up.

Consider another scenario, Nicholas Negroponte constantly travels all around the world to meet all the heads of state and the computer and Internet industry CEOs. Consider that Nicholas Negroponte accepts a secret invitation by the Microsoft CEOs. Consider that the guy who is taking over as CTO after Bill Gates actually is _not_ completely evil.

Consider Microsoft decides to learn from Google and decides to "Do no evil" as well.

Consider now that Microsoft for now does not plan to open the source code of "Windows XP Lite", that Windows XP version that will continue to be available exclusively to ULCPC's after next month (Microsoft will exclusively sell Windows Vista after June), consider Microsoft tells this to Nicholas Negroponte:

"We at Microsoft plan to open the source code of Windows XP within a couple of years, for now we licence it for $26 for ULCPC manufacturers, but soon within two years, we plan to reduce the licencing fee to $0."

"We at Microsoft were only bluffing recently when we talked about taking over Yahoo for close to $50-billion. We have now decided actually that we instead will invest half that amount of money supporting the expantion of computers and the Internet to the emerging markets, we want every of the remaining 5 billion people to get on the Internet within the next 10 years, and the other half of the $50-billion will be invested with or without Yahoo into creating Web 2.0 technologies around an ultimately open-source and free Silverlight platform."

What should Nicholas Negroponte say about that? Fact is, Microsoft has not in the past, but could choose in the future to "Not be evil", and this could mean to put Microsoft's whole efforts into good causes like cloud computing and the OLPC project. How should OLPC's current 15-employee Sugar Linux crew compete with that? And isn't it what all the open source fundamentalists ultimately want, for Microsoft to decide to give away an open-source version of Windows XP and even invest in Linux development?

> As far as I know, there is no real study anywhere that demonstrates
> constructionism works at scale. There is no documented moderate-scale
> constructionist learning pilot that has been convincingly successful

In my opinion, you don't need to look very far to see a huge convincing working success of a constructionist approach to learning, just look everywhere in society:

- The computer and the Internet has proven itself.

- People with the computer and the Internet are better off then people without it. This digital divide and the effects of it is well documented within developped countries and between developped and developping countries.

Basically that is I think the basic philosphy of the OLPC project, in following order of priority:

1. Get those laptops in the hands of children

2. Try to connect as many of those laptops to the Internet using Mesh networking to save costs (the scenario where 1000 laptops can share one access point, thus an $0.20 monthly Internet connection per laptop)

3. Add to this a bit of rethinking in the way the interfaces and software is put together, not only to empower local governments to customize the software and the content because languages and cultures are different, but also because a bit of rethinking of the way the OS, collaboration, security, software all that can use a bit of rethinking.

These 3 main points are in my opinion those trojan horses that Nicholas Negroponte has been talking about. Give the children the tools, connect them to the Internet and those kids will know what to do with it.

Sure a whole lot can be achieved in the point 3, and this can be a very important part, of course software and content is very important, sure a good support infrastructure can be important for the success of each deployment, but those are not the most important parts of the OLPC revolution. Support and deployment can be mostly taken care by the local governments, they put together local task forces that translate software, upgrade firmwares, upgrade and customize software, digitize content and all those things that are of course needed. But point 1 and 2 are the first and second priority trojan horses.

Of course having great support and deployment strategies is very important, but keeping all the manufacturers interested in investing their time and money in continuing to push forward the revolution in cheap laptop manufacturing is more important. The patience of Quanta, AMD, Marvell, CMO and others might be kind of limited. All those very generous partners of the OLPC project on one hand have donated limited amounts on funding the OLPC foundation, but on the other hand in certain cases they have also invested perhaps billions in building the new infrastructure that is needed to make those cheap laptops a reality. Those companies do have financial responsabillities towards their investors to show that the OLPC project is moving towards if not profitabillity, at least break even. Nicholas Negroponte probably needs to reassure Quanta, CMO and the others that soon there will be millions of laptops shipped and not just hundreds of thousands.

So commercialising the XO patents, managing OLPC more like Microsoft, selling more laptops to more countries, invovling the developped countries in figuring out all these support and deployment questions. That is the way to go. Let Australia, Birmingham Alabama, Italy, UK and the EU, India, South-Africa, Mexico, involve all those people to bring their software developpers and their educators into the project to help figure out how to best implement this.

> As far as I know, there is no real study anywhere that demonstrates
> constructionism works at scale. There is no documented moderate-scale
> constructionist learning pilot that has been convincingly successful

Oft praised example is the Reggio Emilia, Italy.

http://en.wikipedia.org/wiki/Reggio_Emilia_approach (Hmm, this wikipedia article doesn't capture the real richness, but it is a start.)

I visited their pre-school a few years ago, and it was surely a rich, rich environment.

The city's budget is about 120 millio Euro per year, and they spend 25 million Euro or such on their 3 month through 6 years old children education. (In Italy, the school system beyond 6 years is taken care of the state and the city cannot make decisions.)

Yes, there are people serious about education. Reggio Emilia people pays attention to the kids (including budget spending).

"Posted by: Charbax on May 15, 2008"

Who the Hell are you, in which parallel Universe did you live up to this day, and how can I get there from here?

If the XO was commercially available, as is, and there was an interest to harness a bigger community that supported and maintained this device as well as Sugar, then maybe Negroponte wouldn't be able to change the whole direction of the project, and other problems would be fixed in the process, better documentation would have to be released. Not to mention that OLPC could deploy more laptops, and therefore become more relevant if it opened to commercial distribution. If for every five laptops with a premium price, for example, one laptop was delivered to a child, then that possibly translate in thousands of more units being delivered free of cost to children.

The best of luck to all of those involved.

I don't really care if the OLPC Foundation continues to exist. But I'd like to think that, because of the G1G1, one child who otherwise would not have a laptop, has one. And perhaps it will inspire him or her to dream and reach further in life. Because of this thought, I won't bitch about being lied to by the OLPC concerning their mission. Otherwise though, adios OLPC. It's not been a pleasure.

P.S. I hope Sugarlabs junks the Journal. I dislike it greatly, and would prefer a good file manager any day of the week.

I'm with teapot on this. What's the deal Charbax? Have you gone over to the dark side?

Up until this last you've been warning about how Intel wouldn't ever allow cheap computers to get into the hands of the poor so as to maintain their bloated profits, etc. Now you're a pal of Bill Gates? What's going on?

I went back to Ivan's article (I had browsed through it in a hurry before) and now I find his logic as bad as his Latin. Laptopii, please, genitive plural from the singular second declension laptopus, or lap-t-opus, penguin-tablet worn on the lap, as used by Roman schoolchildren with a stylus (we don't have that yet).

Too bad, I once saw him as one of the best guys around. I still think he's clever enough, but I would have wanted more of this sharing with the community when he was inside and was able to do something about things.


My Ubuntu upgraded to Hardy, and Firefox to FF3B last night. Guess what, the list of visited links on FF3B looks now _just_ like Sugar Journal. I don't know if to laugh or cry... It very cleverly will put on top your most visited links, which is cute, BUT is not what I expect when I am looking for the most recent site. Oh well, comes with the territory.

"Who the Hell are you, in which parallel Universe did you live up to this day, and how can I get there from here?" - teapot

Charbax wrote over two pages and you dismiss it all with one line. Don't be Irvin.

"now I find his logic as bad as his Latin. Laptopii, please, genitive plural from the singular second declension laptopus" - Yama

Um, forget about his Latin, what is it about his logic that you find so bad?

Aw, David, c'mon. He's a brilliant man. We shouldn't care less whether what he writes makes complete sense. I can fill in some of the blanks. Sorry for the distraction.

What I care for, though, is that he took his sweet while to tell us the king is naked.

I will be happy to see him and anybody support this project further, but to all and sundry, let me tell ya, I'm tired of denials, defensive lines, partial truths and eye candy. I'm tired we don't dare tell people that certain "essential" features don't work, especially to the teachers who would deserve at least a bit of respect because they were told so much was marvelous, so they would get to like this project, and now they are finding it out the hard way that there ain't such a thing as a Saint Nick, and as long as we don't come forward in supporting them in what does work we are all part of the problem.

I've written elsewhere about the need for fighting corruption. We need to start at home, by being straight in telling the truth, even if it's not as pretty as we wish.

Believe it or not, people, the right people, the good people, the little people, they do like the truth. It's the corrupt ones that love to hear lies, to hear the XO can run off a yo-yo and solve all your problems. Ivan has come out, and we appreciate. We wish those that stay in not to wait to stand up and share now what needs fixing, so we can be more together in getting things working, in cooperation, in community.

"25 million Euro or such on their 3 month through 6 years old children education"
Good to know. We're cooked. We are now budgeting about Eu 150 per kid. No way our success can approach Reggio Emilia's.

Hmm, you haven't described any "bad logic" so far. Anyway, it would have been hard for him to write this rant while still working for OLPC, wouldn't it? That rant is the sort of thing that would get anybody fired. Perhaps he could have written something more muted in the past without resigning, but I don't hold it against him. The truth is important, but as long as he was working for OLPC, perhaps he thought he could do some good, privately. I certainly think improving the situation is more important than exposing it; but if it can't be fixed, then it ought to be exposed and that's what Ivan did.

No deployment plan. Incredible.

My best friend figures that often the founder of a movement is not competent to run it. He described a nonprofit that he worked for, in which the board of directors voted the founder out of power but gave him a figurehead position. In that situation, the founder knew he was in over his head and handed over the reigns willingly. But here I have the impression (perhaps wrongly?) that Nick Negroponte thinks he can do no wrong.

I'm also puzzled why NN seems to have changed his mission. It used to be educating children, now it's just selling laptops. How did that happen?