[Edu-sig] OLPC G1G1 sales start today

[Paul D. Fernhout]

Michael Tobis wrote:
> I think this "limited time only" idea defeats the purpose of the whole
> software environment. Why should I buy a community building tool that
> will not be available to potential members of my local community?

A very good point. And you can find lots of people who agree the OLPC
distribution model is problematical for various reasons, for one example:
http://www.olpcnews.com/sales_talk/countries/negroponte_change_olpc_distribution_.html

But as I see it, this is not about the OLPC XO-1 as an innovation by itself.
For me, this is about experimenting with doing education with lightweight
portable "dynabooks"
  http://en.wikipedia.org/wiki/Dynabook
running GNU/Linux supporting dynamic languages (like Python) with wireless
connectivity, which don't break when you drop them or use them outdoors
around a little bit of water or heat or dust. That's the big picture, long
term.

I've bought several portable devices (with hopes of running Python software
on some of them) -- none have had the (projected) robustness of the OLPC.
The robust alternatives I could buy still seem to be in the thousands of
dollars, for example:
  http://en.wikipedia.org/wiki/Toughbook
Still, it is possible the XO-1 won't live up to the hype of toughness. And
if it doesn't, in my eyes it will have been a failure, because I think that
is the most important aspect of it -- reasonably performing GNU/Linux and
Python in a tough but cheap enclosure. The software can be fixed up, and I
expect will improve over time like it did for the Zaurus SL-5500 or Nokia
N800. If the hardware has problems, then you're stuck.

Here is a related article with a video clip of earlier (prototype) OLPC's
with problems:
http://www.youtube.com/watch?v=Fm9bTrpjTZ0
http://www.olpcnews.com/countries/nigeria/olpnc_galadima_school_interview.html
>From there: "On walking around school I saw kids that had laptops with
spoiled screens, torn off "ears", and defect batteries ­ one kid even told
me that his laptop had been stolen. There are teachers for whom the idea of
using the laptops in learning is not yet natural, and there are challenges
regarding how to digitalize the curriculum the kids must go through during
primary school."
I found that clip a little heartbreaking as at the end the child puts his
hand across the apparently defective laptop screen, me thinking -- is he
wonderingly if perhaps there goes his chance at a hopefully better
education? Probably need at least 1.2 laptops per child to account for
spares. :-(

Still, even if only a few people can only get a few now, the general
principles related to "dynabooks" will become more and more ubiquitous as
years go by. And they are already happening -- one can get a Nokia N800 for
a similar price (although it isn't quite as robust, and it is harder on the
eyes). This has been an ongoing trend for many years. With a couple decades
this trend might even culminate in devices equivalent to the OLPC XO-1 given
away for free on the back of cereal boxes. :-)
  "Electronics goes organic"
  http://www.my-esm.com/digest/showArticle.jhtml?articleID=199700530

But even if that seems to hard to believe, it's not much of a stretch to
imagine in ten years buying the equivalent of the OLPC XO-1 used on ebay is
only going to cost in the tens of dollars. At that price, almost everyone in
the world can afford one (or someone can afford to buy them one if they
can't themselves). We are seeing the continued emergence of Pierre Teilhard
de Chardin's "noosphere".
  http://en.wikipedia.org/wiki/Noosphere
Whether that will be seen as good or bad for most remains to be seen.

So, I see the XO-1 as mostly a funky vehicle for this kind of "dynabook"
experimentation. But I expect in two or three years this OLPC XO-1 will be
obsolete (if it isn't already. :-) Though on the other hand the Newton
MP2100 has held its value for almost a decade, and probably still exceeds
the OLPC XO-1 in good power management and ease of use -- although that is
mostly software related, for example the Newton may sleep 100 times during
the time it take you to draw a line across its screen.

I could write a lot on OLPC failings based on the Newton's successes (OLPC
didn't use an ARM processor like the Newton, color maybe wasn't essential
and if it was, it could have been done the usual way,
  http://wiki.laptop.org/go/Image:Blurry-develop.png
the Newton OS and GUI really did get out of your way most of the time and
worked well in a mobile setting). I could even point out at length how the
OLPC project duplicates some of the Newton's mistakes (making it hard for
developers to use existing code or develop using existing desktop equipment,
as OLPC could have made Sugar optional add-ons to Gnome instead of
essentially replacing Gnome, etc.),
 http://www.osnews.com/story.php/16582/The-OLPC-Sugar-Interface-Dont-Do-it/
Still, I think there are also a lot of good points about the general OLPC
concept anyway (and even about Sugar's aspirations). I also admire the way
it was done as a non-profit using free and open source methods. One can
quibble about how organized they are, etc., or lots of other things they may
have had difficulties with, or with the distribution models, or the hype
machine, but as I see it the OLPC is the biggest game in town working on
hardware and software this way -- and they have chosen Python to pull
everything together.

I feel a little about it like I feel about GNU/Linux, which I resisted for
years because it was obvious UNIX was an antiquated technology compared to
all the other promised OS's (Plan 9, Microkernels, Inferno) or even just
plain old pre-DOS QNX or the Mac OS. But in the end, the GNU & Linux
communities have kept chugging a long and become the biggest game in town in
that free OS arena.

I don't know if OLPC will have that much momentum, of course. I doubt it --
not because it isn't a ultimately a good idea to make robust cheap dynabooks
for kids (or adults :-), but just because there will be so much hardware
competition with dropping prices that in a few years one may have many
choices with a variety of OS's or software environments (e.g. the Java
powered PepperPad http://www.pepper.com/ to name just one competitor). And
some of this is happening precisely because the OLPC project paved the way
and stirred up the commercial competition.

But all that is good news ultimately for anyone who wants to use computers
in education -- better, faster, cheaper, lighter, more robust. It's all part
of an ongoing revolution, perhaps leading up to Vernor Vinge's
Hugo-award-winning short science fiction story "Fast Times at Fairmont High"
  http://www.kurzweilai.net/articles/art0518.html?printable=1
where he talks about the equivalent of sprinkling smart dust (called
"localizers") around schools. From the review: "One thing that was
believable is that it seemed that a lot of the kids cheated, and it was
almost impossible for the adults to catch them. With universal network
connectivity it would be hard to make sure kids are doing their work on
their own. I got the impression the school sort of looked the other way, the
idea being that as long as the kids solved their problems, even if they got
help via the net, that was itself a useful skill that they would be relying
on all their lives."

One can even question the whole idea of whether kids in developing nations
are better off without today's conceptions of computing. Certainly many of
these developing nations already rank as "happier" than much of the
developed world.
  http://news.bbc.co.uk/1/hi/world/africa/3157570.stm
  http://thehappinessshow.com/HappiestCountries.htm
Are we really sure we want to tinker with that happiness? But it is
happening anyway -- so at least hopefully they can get the benefit of the
good part of it while perhaps using their own unique cultural strengths to
work around the bad parts.

The OLPC XO-1 is a first of a kind. It's almost certain it will fail as a
specific product because it is a very ambitious undertaking. Or, as "Fake
Steve Jobs" sarcastically wrote:
 http://www.olpcnews.com/commentary/press/one_fake_steve_jobs_sarcasm.html
"Frankly I'm shocked to see these guys having problems. I mean, a brand new
hardware design, a new screen technology, a customized Linux operating
system, a one-off user interface, and the customers are the poorest nations
in the world, …and the whole project will be run by woolly academics who
have never even worked in a real company let alone run one. What could
possibly go wrong?"

But, as an idea, I think it has a lot of momentum, even if others in the end
may carry the idea forward in better ways. And it is an idea -- the Dynabook
-- with roots back in the 1960s or even earlier. This is just a step on the
way. I think the biggest thing to take away from this is that the dynabook
hardware is pretty much there. It is the software and content that is the
problem. And that's where edu-siggers etc. can play a vital role. Again, from:
  http://en.wikipedia.org/wiki/Dynabook
"Though the hardware required to create a Dynabook is here today, Alan Kay
still thinks the Dynabook hasn't been invented yet. The key software and
educational curriculum are the missing pieces."

> What are the prospects of a commercial product running the XO software system?

Probably zero. I would even think the OLPC project would think this
probability should ideally be zero, to minimize the risk of theft. Having
said that, I know there was interest by groups who do outdoors type
applications -- like surveying or inspection.

> I agree, I am tempted to buy four and get two rather than buy two and
> get one. That starts to be enough money that I hesitate though.

I agree. I actually really wanted three. Even just two seems like it doesn't
fully get at the idea of, say, a third person joining an existing networked
group activity. I was hoping maybe I could substitute one emulator session
on a desktop for that third laptop.

> I disagree about Java, though. I think that moves in the wrong direction.

True, Java has its own issues as well. Sometimes it seems it can't make up
its mind if it is an OS or a software library or a VM or a language. And
dealing with a proliferation of versions and configurations is its own set
of potential headaches, and a major weakness in the Java concept. Nothing is
perfect, in part because one person's priorities may differ from another's
-- in my case, easy Mac installation is important for me (while still being
cross-platform) and I like Swing better than, say, wxWidgets.

--Paul Fernhout