4 ms·

One Laptop per Child

25 points by amrrs 5y ago

ddtaylor 5y ago
Does anyone know where the video of the crank failing is?
- codewiz 5y ago
  iirc, the "prototype" that was unveiled to Kofi Annan was just a polyester model: https://www.youtube.com/watch?v=hii3ih_XtTw https://www.youtube.com/watch?v=hii3ih_XtTw
  - codewiz 5y ago
    I meant to say polystyrene.
rektide 5y ago
XO-PC is the attempt everyone seems to love to hate. It's a magnet for "look how bad a job was done." It's tragic that disbelief & belittelement sells so easily, so readily. We are so hungry to hear these tales, & they are so much more easy to lay out at our feat, for they're spoken in terms of failures we can understand, that are evident. What's much harder to do is to illuminate the absolutely magical fantastic incredible feats that this project delivered. It's hard to stress just how much got done, how radical a revolution in OS design XOPC was. The ambitions of the project made it exceedingly far, creating in the process sophisticated, cutting edge open-source contributions that we still have yet to begin to reproduce adequately. It's so much more nuanced & complex to tell the tale of what was really at stake here; it's so much easier to list some failing & flailings, to instigate a point & laugh response from the knows-no-better audience. The device was magical. There's never been an attempt, ever, to build as ambitious a physical system. A screen that worked inside, & outside. A mesh network that should let two kids collaborate on a hilltop, a classroom collaborate together with or without access points and/or internet, or, as mentioned, the not so working attempt to relay along a single internet access point through a mesh. A crank power generator. Semi-ruggedized. In 2007. That they were trying to find a way to make for under $150 or $100. What makes the litany of faults posts/books/&c so transparent to me is that they fixate on hardware concerns. First, I feel it was remarkably close. Had it achieved any scale, they probably could have worked out more of the issues. Second, and most chiefly, this is judging the book by it's cover. Of the complains & whining & shit-talking about this device, 98% are about hardware. Almost no one has an appreciation for what a masterwork of software was brought together. They don't speak of what was attempted inside the book. This was the most ambitious pro-child, pro-teen, pro-learning computing environment ever created, and it has a huge number of absolutely unmatched, totally novel, remarkably useful & functional inventions that were poured into it, to take a standand-alone desktop of programs, to create a exoskeleton of computing to narritivize & make cross-app visible what was going on, & to then stitch together multiple folks's experience into something wider & classroom scale. SugarOS was built upon the best-of-the-day available open source technologies, technologies we still have nothing even remotely as capable of or as powerful as. Each laptop ran a local XMPP server/agent, which connected with other laptop's XMPP servers/agents, over mesh, using link local ("serverless" xmpp, implementation: Salut) or internet addressing (classic federated xmpp, implementation: Gabble). This was thanks to the Telepathy communication framework, from 2005 by Robert McQueen, which was a pluggable DBus powered communication framework, described in Architectures of Open Source[1]. Telepathy links, in turn, was then used to tunnel DBus connections over the network, called Tubes[2]: something still incredibly powerful, unmatched today. This powered SugarOS's Shared Acvitities[3][4] (and Presence), which were a range of interesting collaborative learning apps that a classroom could collaborate on. DBus, the common well-accepted Desktop Bus of Linux, was being used as an internetworking transport layer to build apps that could coordinate & work together. App designers didn't need to think about complex networking models like they do today: DBus let the local objects of a system speak for themselves, and tubes let other far away participants interact with those objects. The Activities Model was an powerful idea that apps should be generating a log of what they're up to, and writing to the system. That way a student could go back and see what they've done over time, and revisit these activities. It was, in my view, a more powerful, interesting system than Android Intents, which would ship in 2008, and as well as being more powerful, Activities were persistent/enduring, and user facing, whereas in Android they only serve to momentarily stitch together the different pieces of the OS. That act of Activities being visible & persistent underlies many of the choices of the SugarOS. Sugar is supposed to be a learning OS, one that shows you about computing & computers. I've already mentioned that it used best of breed available open source technologies, not inventing it's own paradigms but making use of FreeDesktop standards & the best protocols available at the time. It tried to manifest the ideas of Constructivists[5] (such as Seymore Papert) in teaching, who assert that we do not transfer knowledge to each other, but reconstitute experience on our own terms, and it did this by making what was happening visible, alterable, by keeping the learner in control, as much as it could. All these hit pieces on XOPC: none of them can address 1/10th of the ambition & hope & potential that went here. They all dissuade, convince you not to read further, not to investigate: they tell of folly & failure. To me, these authors seem uneducated & foolish, unable to understand the absolutely pure unbridled magic that was trying to be released. XOPC is the most advanced, most interesting OS on the planet, to this day. It's tried to jump the gap, tried to empower & make powerful the user, & to help them work with other users, in ways far surpassing what any OS has ever attempted, and it bloody did it. It build the tools for collaborative apps, it built a journal to help one see the course they've taken through computing, across a wide range of apps & experiences. And it did all this while shipping an interesting capable piece of hardware (that was not without flaws) that does more than many computers we do today: worked indoors & out, successfully connected with our peers, had some ruggedness. Hats off to SugarOS & XOPC: the most incredible, revelatory, inspiring OS ever made. May one day we aim to bring great open source more manifestly into the world, more visible, more powerfully-interactably, learnably/constructivistly again, more collaborative & cross-user. Today's future is boring, uninspired, a recycled washed out repetition of the past, endless application shoeboxes, whereas SugarOS succeeded admirably in building a better future, where we were connected with our computers more broadly, able to see & visualize the system, able to work not just the tools but computing itself, and able to stitch together our computing interesting across systems, across users, across kids, across people. SugarOS is the most inspired powerful vision of computing anyone has ever stated, and technically is mostly did work. Communicating what SugarOS is & how radically it succeeded is much harder, and much more important than once again harping upon XOPC's unsuccesses (and it acts as chaff that cover's today's uninspired furtherances of the past). [1] https://www.aosabook.org/en/telepathy.html https://www.aosabook.org/en/telepathy.html [2] http://wiki.laptop.org/go/Telepathy http://wiki.laptop.org/go/Telepathy [3] http://wiki.laptop.org/go/Activity_sharing http://wiki.laptop.org/go/Activity_sharing [4] http://wiki.laptop.org/go/Collaboration_Central#Collaborative_Activities http://wiki.laptop.org/go/Collaboration_Central#Collaborativ... [5] https://en.wikipedia.org/wiki/Constructionism_(learning_theory) https://en.wikipedia.org/wiki/Constructionism_(learning_theo...
- rektide 5y ago
  Following on just a little here.. much of the latter part of this Twitter thread is about trying to take sure that education is the primary objective, that tech does not lead. I continue to think that there's little sign educators have any chance to develop & support more healthy "creator" centric tech in the current environments (which is the sole positive/pro- tech thing this Twitter thread encourages). We're not going to get good at using computers to teach math or langauge or science, so long as computers themselves are a prescriptivist tool, so long as we don't empower people to be competent & capable & comprehending of technology. XOPC tried valiantly to set a powerful new relationship up between the student & their computer. It tried to make tech not just a tool for teaching, but to make tech open terrain a pupil could wander through, understand, learn about, on their way to their other tasks. This to me is so very much more interesting & ennabling than what I think of as the purpose-minded "how can we make test scores go up" learning that I see as the focus of this twitter thread. This author is probably right about what current tech startups need to do. But he's wrong about what the future really needs.
- aaron695 5y ago
  SugarOS is pure fucking evil. The rich, the Hiltons or the Buffett's or Trumps or any normal Western person would not waste their child's future on SugarOS. They will train them on what matters. The fact we did this to the poor natives in some sort of experiment says a lot about us. I'm 100% for testing on poor people to help the rich... with reasonable pay. But there was no need here. Microsoft was willing to come to the table but was rebuffed. Much of what you mention simply does not work in a classroom or the real world. Perhaps we didn't know at the time what works. But OLPC refusal to try and design a system the first world could also use or gave dignity to the poor meant they failed slowly and alone without a chance to pivot to working ideas.
- rasz 5y ago
  > Almost no one has an appreciation for what a masterwork of software You mean the deal between Negroponte and Gates to ship XP on it? Most likely facilitated by no other than Epstein.
poulpy123 5y ago
I remember the time thinking it was a solution in search of a problem
aufhebung 5y ago
>Enthusiastic Teachers & care-givers who gave support & encouragement to the children to use their laptops for creation & not just for consumption. I worry about the prevalence of iPads in (first world) schools. iPads are great for consumption, but pretty terrible for creation outside of a few areas like drawing.
protomyth 5y ago
In my mind, the greatest failings of OLPC was their reliance on foreign governments and NGOs, and their contempt for poor students in the US. The whole buy 2 get 1 crud was just insulting if you wanted to buy one for any poor students in the US. Their utter lack of programs for children in the US or ignoring people wanted to buy them for poor children in the US (oh, say a reservation) was just foolish.

25 points by amrrs 5y ago

ddtaylor 5y ago

Does anyone know where the video of the crank failing is?

codewiz 5y ago
iirc, the "prototype" that was unveiled to Kofi Annan was just a polyester model: https://www.youtube.com/watch?v=hii3ih_XtTw https://www.youtube.com/watch?v=hii3ih_XtTw
- codewiz 5y ago
  I meant to say polystyrene.

rektide 5y ago

XO-PC is the attempt everyone seems to love to hate. It's a magnet for "look how bad a job was done." It's tragic that disbelief & belittelement sells so easily, so readily. We are so hungry to hear these tales, & they are so much more easy to lay out at our feat, for they're spoken in terms of failures we can understand, that are evident. What's much harder to do is to illuminate the absolutely magical fantastic incredible feats that this project delivered. It's hard to stress just how much got done, how radical a revolution in OS design XOPC was. The ambitions of the project made it exceedingly far, creating in the process sophisticated, cutting edge open-source contributions that we still have yet to begin to reproduce adequately. It's so much more nuanced & complex to tell the tale of what was really at stake here; it's so much easier to list some failing & flailings, to instigate a point & laugh response from the knows-no-better audience. The device was magical. There's never been an attempt, ever, to build as ambitious a physical system. A screen that worked inside, & outside. A mesh network that should let two kids collaborate on a hilltop, a classroom collaborate together with or without access points and/or internet, or, as mentioned, the not so working attempt to relay along a single internet access point through a mesh. A crank power generator. Semi-ruggedized. In 2007. That they were trying to find a way to make for under $150 or $100. What makes the litany of faults posts/books/&c so transparent to me is that they fixate on hardware concerns. First, I feel it was remarkably close. Had it achieved any scale, they probably could have worked out more of the issues. Second, and most chiefly, this is judging the book by it's cover. Of the complains & whining & shit-talking about this device, 98% are about hardware. Almost no one has an appreciation for what a masterwork of software was brought together. They don't speak of what was attempted inside the book. This was the most ambitious pro-child, pro-teen, pro-learning computing environment ever created, and it has a huge number of absolutely unmatched, totally novel, remarkably useful & functional inventions that were poured into it, to take a standand-alone desktop of programs, to create a exoskeleton of computing to narritivize & make cross-app visible what was going on, & to then stitch together multiple folks's experience into something wider & classroom scale. SugarOS was built upon the best-of-the-day available open source technologies, technologies we still have nothing even remotely as capable of or as powerful as. Each laptop ran a local XMPP server/agent, which connected with other laptop's XMPP servers/agents, over mesh, using link local ("serverless" xmpp, implementation: Salut) or internet addressing (classic federated xmpp, implementation: Gabble). This was thanks to the Telepathy communication framework, from 2005 by Robert McQueen, which was a pluggable DBus powered communication framework, described in Architectures of Open Source[1]. Telepathy links, in turn, was then used to tunnel DBus connections over the network, called Tubes[2]: something still incredibly powerful, unmatched today. This powered SugarOS's Shared Acvitities[3][4] (and Presence), which were a range of interesting collaborative learning apps that a classroom could collaborate on. DBus, the common well-accepted Desktop Bus of Linux, was being used as an internetworking transport layer to build apps that could coordinate & work together. App designers didn't need to think about complex networking models like they do today: DBus let the local objects of a system speak for themselves, and tubes let other far away participants interact with those objects. The Activities Model was an powerful idea that apps should be generating a log of what they're up to, and writing to the system. That way a student could go back and see what they've done over time, and revisit these activities. It was, in my view, a more powerful, interesting system than Android Intents, which would ship in 2008, and as well as being more powerful, Activities were persistent/enduring, and user facing, whereas in Android they only serve to momentarily stitch together the different pieces of the OS. That act of Activities being visible & persistent underlies many of the choices of the SugarOS. Sugar is supposed to be a learning OS, one that shows you about computing & computers. I've already mentioned that it used best of breed available open source technologies, not inventing it's own paradigms but making use of FreeDesktop standards & the best protocols available at the time. It tried to manifest the ideas of Constructivists[5] (such as Seymore Papert) in teaching, who assert that we do not transfer knowledge to each other, but reconstitute experience on our own terms, and it did this by making what was happening visible, alterable, by keeping the learner in control, as much as it could. All these hit pieces on XOPC: none of them can address 1/10th of the ambition & hope & potential that went here. They all dissuade, convince you not to read further, not to investigate: they tell of folly & failure. To me, these authors seem uneducated & foolish, unable to understand the absolutely pure unbridled magic that was trying to be released. XOPC is the most advanced, most interesting OS on the planet, to this day. It's tried to jump the gap, tried to empower & make powerful the user, & to help them work with other users, in ways far surpassing what any OS has ever attempted, and it bloody did it. It build the tools for collaborative apps, it built a journal to help one see the course they've taken through computing, across a wide range of apps & experiences. And it did all this while shipping an interesting capable piece of hardware (that was not without flaws) that does more than many computers we do today: worked indoors & out, successfully connected with our peers, had some ruggedness. Hats off to SugarOS & XOPC: the most incredible, revelatory, inspiring OS ever made. May one day we aim to bring great open source more manifestly into the world, more visible, more powerfully-interactably, learnably/constructivistly again, more collaborative & cross-user. Today's future is boring, uninspired, a recycled washed out repetition of the past, endless application shoeboxes, whereas SugarOS succeeded admirably in building a better future, where we were connected with our computers more broadly, able to see & visualize the system, able to work not just the tools but computing itself, and able to stitch together our computing interesting across systems, across users, across kids, across people. SugarOS is the most inspired powerful vision of computing anyone has ever stated, and technically is mostly did work. Communicating what SugarOS is & how radically it succeeded is much harder, and much more important than once again harping upon XOPC's unsuccesses (and it acts as chaff that cover's today's uninspired furtherances of the past). [1] https://www.aosabook.org/en/telepathy.html https://www.aosabook.org/en/telepathy.html [2] http://wiki.laptop.org/go/Telepathy http://wiki.laptop.org/go/Telepathy [3] http://wiki.laptop.org/go/Activity_sharing http://wiki.laptop.org/go/Activity_sharing [4] http://wiki.laptop.org/go/Collaboration_Central#Collaborative_Activities http://wiki.laptop.org/go/Collaboration_Central#Collaborativ... [5] https://en.wikipedia.org/wiki/Constructionism_(learning_theory) https://en.wikipedia.org/wiki/Constructionism_(learning_theo...

rektide 5y ago
Following on just a little here.. much of the latter part of this Twitter thread is about trying to take sure that education is the primary objective, that tech does not lead. I continue to think that there's little sign educators have any chance to develop & support more healthy "creator" centric tech in the current environments (which is the sole positive/pro- tech thing this Twitter thread encourages). We're not going to get good at using computers to teach math or langauge or science, so long as computers themselves are a prescriptivist tool, so long as we don't empower people to be competent & capable & comprehending of technology. XOPC tried valiantly to set a powerful new relationship up between the student & their computer. It tried to make tech not just a tool for teaching, but to make tech open terrain a pupil could wander through, understand, learn about, on their way to their other tasks. This to me is so very much more interesting & ennabling than what I think of as the purpose-minded "how can we make test scores go up" learning that I see as the focus of this twitter thread. This author is probably right about what current tech startups need to do. But he's wrong about what the future really needs.
aaron695 5y ago
SugarOS is pure fucking evil. The rich, the Hiltons or the Buffett's or Trumps or any normal Western person would not waste their child's future on SugarOS. They will train them on what matters. The fact we did this to the poor natives in some sort of experiment says a lot about us. I'm 100% for testing on poor people to help the rich... with reasonable pay. But there was no need here. Microsoft was willing to come to the table but was rebuffed. Much of what you mention simply does not work in a classroom or the real world. Perhaps we didn't know at the time what works. But OLPC refusal to try and design a system the first world could also use or gave dignity to the poor meant they failed slowly and alone without a chance to pivot to working ideas.
rasz 5y ago
> Almost no one has an appreciation for what a masterwork of software You mean the deal between Negroponte and Gates to ship XP on it? Most likely facilitated by no other than Epstein.

poulpy123 5y ago

I remember the time thinking it was a solution in search of a problem

aufhebung 5y ago

>Enthusiastic Teachers & care-givers who gave support & encouragement to the children to use their laptops for creation & not just for consumption. I worry about the prevalence of iPads in (first world) schools. iPads are great for consumption, but pretty terrible for creation outside of a few areas like drawing.

protomyth 5y ago

In my mind, the greatest failings of OLPC was their reliance on foreign governments and NGOs, and their contempt for poor students in the US. The whole buy 2 get 1 crud was just insulting if you wanted to buy one for any poor students in the US. Their utter lack of programs for children in the US or ignoring people wanted to buy them for poor children in the US (oh, say a reservation) was just foolish.