Well.. he's back, with friends.

It's a good video to watch when you worry about things like war, unfair trade practices, poor foreign policy, dictatorship, and so on -- it reminds you that people are globally friendly, silly, happy folk if given a chance. Which is always true, but not always easy to remember.

It was a fascinating discussion. I'm a strong supporter of sustainability (otherwise, why bother?) and Al Hammond gives a passionate and convincing argument for the central role of business in creating sustainable solutions. Talking with him beforehand, he mentioned (paraphrasing heavily) comparing the measurable benefits of the past five decades of foreign aid versus the last decade of private sector mobile phone rollouts -- the long-term benefits greatly favor the mobile phones.

I can only imagine that once mBanking really gets rolling, all doubt will be erased that the cell phone has helped the Next Four Billion more than 2.3 trillion dollars in aid. My mind quickly out-paces itself when I begin to ponder mBanking benefits for everyone from rural artisania workers able to take and receive payments for commissioned artwork to p2p payment systems to direct-to-market agricultural benefits...

Now, I have a few outstanding doubts about some parts of these two plans - some scalability and malicious-user problems with the Vietnam model, and some privacy and franchise-enforcement questions with the healthcare idea. Now, I have fewer doubts on both of these concepts, combined, than, say, the OLPC Project (though I strongly believe that a base-of-the-pyramid approach to the OLPC could work well). The huge difference between a BoP, market-driven approach and traditional development is that investors bear the brunt of failed projects, a pleasant change from the recipient country being in deeper debt regardless of the outcomes of debt-financed aid projects. I think traditional development will forever have a role in humanitarian and post-conflict aid, but in infrastructure and service creation, the BoP, private-sector approach will prove long-term much stronger than pure-play foreign aid programs, for the simple reason that it applies reasonable risk management to development projects. What a concept!

Without further ado, my full meeting notes after the jump...

Low-cost Rural Connectivity Model: Vietnam pilot

NextBillion blog post on rural connectivity , NextBillion on mBanking

• Approach to ict: think about scale; therefore focus on mobile companies, only a market driven approaches will scale
• Pilot in three rural communes in Vietnam using wifi and voip to extend network to rural area; pilot then invite mobile companies backed by intel
  • wifi backhaul, but wimax and vsat approaches possible
• Sin Mua commune: pop 2203 mountainous, fiber in 12km and decent power
  • ptp antenna with omni distribution 7km hop
  • meshed wifi APs ; local calls stay local via switch; EVN telecom
• nghia tho commune: ppo 939; further to fiber, simple install for regular electrician not engineer required
  • $1000link can go 65km ptp backhaul
• tinho tho third cmmune, more compex net mesh
• performance
  • wifi phone to wifi phone 400 meters range (so 800m total between phones with AP in the middle)
  • wifi ph to mobile network 500 m
  • laptop internet access 1km
  • invisible voip system, just a phone
• economics
  • $22k/community with labor and nondiscount/mass equip, likely lower number
  • capex 14.47/house. 3/21person
  • 20% of gsm network cost
  • low power, don't need generators for network devices
• Value prop to customers and mobile providors
  • 1/2 of calls stay local to a community => free calls in town b/c never hits network => increased penetration and long-term increased revenue b/c of increased penetration and access to non-local calls via switched net
  • 3% teledensity in rural vietnam
  • mobile companies can make money off these models
    • 30% household penetration generating $4 rev/month, networked infrastructure payback is 1.5yrs
  • addresses fears of mobile companies against voip as competition
• Addon benefits
  • remote management of switches, auto interference detection and channel switching, auto-rollover to other wifi networks
  • shut down virus and p2p traffic
  • broadband net allows voice, software flexibility
  • easy training of local engineers to build out network, sustainable
  • 4G network?? wimax doesn't win out in rural areas in tech and loses w/r/t cost; follow on gsm tech in future?

voip questions

• The wifi is essentially open wifi for laptops; also tourist selling point for free wifi?
• Along with wifi and VOIP rollout, pilot places 3 telecenters in each commune, one at the commune headquarters, one at the main healthcare location, one at the school, with hopes that individuals could set up home-based cybercafes
• negotiated with local radio station to use their tower, gave them an omni AP, wanted internet even though they didn't have computers
  • so did they get PCs? pilot started 3 wks back, equipment failure from mistake by other cell provider (corporate sabotage?) and weather problems since, so no reports yet.
• telecom ministry hearing demand from other communes
• SMS over voip network?? no, no sms in pilot; a real dual mode wifi+gsm phone in plan for rollout
  • $20 phone + $5 wifi chip + $5 biometric reader = $30 sale price at 1M units phone
  • sms capability and biometric reader for full support of mbanking with security of transactions
  • connectivity platform
    • mbanking about to explode to benefit of rural
    • then eduction, entertainment = > huge mkts making basic connectivity look small
    • get paid for crop via mobile phone?

Scaling Healthcare and ICT

NextBillion Blog post on this

• building infrastructure including healthcare
• 150+B market in developing world currently for healthcare, 1/3-1/2 on pills
• problems with doctor/hospital avoidance, informal providers, fake drugs, inflated pricing
• last mile healthcare - mi farmicita; franchised, tightly managed distribution platform with high tech info / feedback loops
  • franchise distro system of pharm via secret shoppers, audits, etc - to weed out anyone not playing by the rules; provide a private sector discipline that doesn't exist - one strike and you're out
  • franchise pharmacies as scalable model, low startup cost, generic localy produced drugs, new techs and new distribution channel
• services: drugs, then diagnostics, minor remote surgery (req. broadband), **grows from pharma to accessible medical kn and diagnostics
  • future build out of new techs
    • diagnostic tech for matching patient and drug, dna analysis unit for $50 ea, 5 minute analysis, 50 cents/test, don't need dr to read it, tests for specific diseases
    • remote practice tech, remote reading of sonogram, etc
• Information usage
  • walmart logistics, detailed patient records, remote
  • 2 way info - collect patient data for health min, marketers, consumer data, biometric data
• incubation, 3rd party insurance to subsidize access to very low income,
• model assumes internet (which can be sold on the side)
• economics
  • pharmacy net - 5M investment, 1-2 yr value is 25M

healthcare Q&A

• How scale at right speed to avoid same problems as doc/hospital avoidance? lots of culture/trust issues aswith any medicine/new tech?
  • focus on selling pills, build trust and branding thru good product and fair prices; 1 yr to build trust in kenya, build trust gradually and also intro svcs gradually, remote treatment requiers good records
• selling of healthcare info?
  • local decisions, incentive to use data for revenue, run tests for health ministry for a fee, sell to marketers, insurance; balance with value prop

• • hope towards regulatory rules, system built to pass HIPAA standards
• Other models, like similares in Mexico (untested meds)
  • Regulatory and trust problems
• Maintenance costs of the remote practice equipment?
  • upkeep of diagnostic - consumables all on sample chip of 5 cent, one tool one test, so low upkeep, designed for bop use, low expertise and cheap
• govt role?
  • public sector provision problematic due to high red tape, low trust. so lower hanging fruit is private sector; Philippines law allowing remote pharmacist approval to pill dispensing
  • Kenya Mexico, Philippines and India as rollout countries
  • high variability of prescription standards across countries (doc, pharmacist, anyone sells) and enforcement and behavior varies from that. education of regulators during scale, but value prop is a massive improvement over current
• telemedicine via fee for service private sector care, but maybe a nonprofit insurance system also?
• Current examples of remote practice include Nighthawk, which has radiology expert across the world for late night xray analysis

Some more questions...

• How resistant is the wifi mesh to malicious usage - i.e. someone who sets up their own AP in the mesh to capture data (e.g. mbanking info, healthcare info?) as it goes through, or to reroute calls?
• Healthcare privacy issues will become paramount since the entire model is built on creating consumer trust; one person who's HIV status is leaked kills the project. It's great that the system is built to conform to HIPAA standards, but I think this decision has to be part of the franchise's strictly managed rules, only sell info in aggregate or at least with patient identification info stripped and in sufficient quantities that small communities don't reveal data by accident.
• The franchise model is great, but problematic with free-wheeling copyright/trademark infringement will produce copycats, will require massive marketing system especially when you have to de-authorize a pharmacy.

XO laptop owners need more jabber servers to mesh network on. Every time I look at my empty neighborhood view I am sad. Yet I am not geek enough to run a jabber server solo. I need the help of a jabber expert to set one up for DC.

The super nice folks at Obscure.org set up a jabber server for the DC area, and there are many other community jabber servers listed at Laptop.org wiki list of community jabber servers. Any one server can only handle up to ~150 users at a time, so please choose your community's server (if your community doesn't have one, try setting one up, or contacting Harper Reed of XOChat.org for an XOChat regional server)

Find friends and shared activites

It takes a few seconds to set up via the sugar control panel, and works amazingly well. The process is documented over in the OLPCNews forums, and I've included the short-and-sweet instructions in the full article (click below).

Jabber servers can get overwhelmed by too many users, so make sure you connect to one that's near to you (or set up your own Jabber server for your local OLPC users group!) . If you're in DC however, this is a call to action to sign on to dc.olpc.obscure.org, share chat activities and create some online meshy XO community love! Setup instructions, server information for the rest of the world, and some tips and tricks after the jump.

1. Load up a terminal session
2. type sugar-control-panel -g jabber
   (this shows you the current setting)
3. type sugar-control-panel -s jabber jabberserver.example.com
   (this sets it to your new jabber server. DC's server is dc.olpc.obscure.org . If you're not in DC, check out the Laptop.org wiki list of community jabber servers or the XOChat list
4. Go back to the "mesh" view by pressing the ring button
5. Other XO icons should pop up if anyone's online!

Don't forget to fix the chat bug if you've upgraded since the last time you fixed it:

su
cd /usr/share/activities/Chat.activity/
mv chat.py chat.py.old
wget http://dev.laptop.org/~morgan/chat.py

Where next?

• 5-25 words: $ 12.50
• 26-50 words: $ 17.50
• 51-100 words: $ 25.00
• 101-250 words: $ 50.00
• 251 words and up: $ 100.00

the AP had a long history of quoting more than 100 words from bloggers -- and not even linking back to the original blog. Now, in a bit of ultimate irony, the AP's own article about this brouhaha quoted (without linking) twenty-two words from TechCrunch. That's 18 words more than the supposed four word "limit" the AP has suggested. With an ironic chance that wide, TechCrunch's Michael Arrington couldn't resist, and asked his lawyer to send a DMCA takedown notice to the Associated Press, along with a bill for $12.50
(via TechDirt)

You can read TechCrunch's hilarious take on this as well, but I'll refrain from quoting them based on their feisty legal department.

The general sense is that everyone wants portability, but is unwilling to sacrifice anything to get it. I say bollocks -- you can keep your sore shoulders, I'll make a few minor sacrifices, adapt my lifestyle a bit, and carry on. After the jump is my full response.

I've been using my "G1G1" OLPC XO (http://aptop.org) as my travel machine. It does RDP and VNC (and ssh, natch) if I need to connect remotely to a machine, has basic browsing (and Opera works just fine on it with a little effort), document editing, and so on. It can deal with Google Apps (for spreadsheets) and is super tiny, weighs about as much as a full coffee mug, rugged, and I can read the screen in full sunlight. Bonus: it can charge from those annoying cellphone-only charging stations and a full battery lasts 4-5 hours (wireless turned off) which is perfect for watching a movie or two^H^H^H^H^H^H getting work done on a long plane flight. It's not a powerful processor and has only 1G onboard storage (with an SD slot; I have an additional 4G on that), a few USB drives, and nothing else. You'd be surprised how functional that can be as long as you remember to copy whatever you want to work on to a thumbdrive before heading out the door. If not, it has some of the best wifi reception I've ever seen, and the first round of G1G1 systems came with a year of free Starbucks/tMobile wifi, which most airports have somewhere. If you're willing to accept the limitations of the system, it's hard to beat as a travel system. There's rumors of another round of G1G1 coming soon. If anyone is curious about the XO, there are monthly XO meetups (posted at olpclearningclub.org) and other news and events posted at OLPCNews.com (Note: I write for OLPCNews)

Finally, is there a good TCO study on a mac laptop vs. a hardware-similar PC laptop that includes security software, general hassles, and so on? Having recently blown a few weekends working with bluescreen-level driver problems on my boss's brand new Vista machine, I have newfound respect for a system that Just Works.

This is one of the smartest uses of Twitter for public outreach I’ve ever seen — giving folks a sense of a personal connection to this high-tech mission to find water (and signs of life) on Mars. (Some members of the Phoenix team are also blogging.) I especially like that Mars Phoenix is replying to questions sent in by its Twitter friends (like me).

Makes it all seem so much less… alien!

Hubris is a longstanding problem in development work, as William Easterly (among many others) has been writing about in sordid detail for years. If you haven't read The Elusive Quest for Growth, go to your library or local bookstore now and grab a copy. It's fascinating, disturbing, and clearly written.

The OLPC project has sadly failed to learn from the many, many missteps in large scale, top-down development projects, as we've been writing about over at OLPCNews.com for years now. Without careful implementation working with in-country experts, the project will never come close to fulfilling the original vision, as Peru is revealing:

Even with these results, the Unified Union of Education Workers of Peru, representing some 320,000 public school teachers, is skeptical. "These laptops aren't part of a comprehensive educational, pedagogical project, and their usefulness is debatable," says Luís Muñoz Alvarado, the union's general secretary. Muñoz never had a chance to explore the laptops, though. In what seems an easily avoidable blunder, the Education Ministry has not explained the program to the union.

So in a haphazardly, too-little-too-late fashion, 1CC is piecing together an implementation plan as they go, which makes about as much since as building an airplane in mid-flight:

Recognizing the need to integrate the laptops into communities, OLPC is scrambling to develop guidelines for deployment based on the experiences in Uruguay and Peru, the two countries with the largest distribution so far. The group is also bringing in consultants to advise countries on how to integrate the PCs. One, Edith Ackermann, a visiting scientist at MIT, says OLPC should have involved more educational experts in creating and testing the applications. Instead, she says, "The hackers took over." The result is some programs are too complex for many children to use. "Now we have to deal with this. I don't know if it's too late," says Ackermann.

While some critics have called on OLPC to hire aggressively so it can provide on-the-ground support for dozens of countries at a time, Negroponte and Kane plan instead to rely even more on outsiders. They'll forge alliances with local tech companies and nongovernment organizations that will provide deployment support.

That is one move that OLPC is making correctly -- presuming that those alliances are contractually required to help build a shared and open knowledge base that can create a community of practice and body of knowledge that future implementations of similar projects can take advantage of. OLPC hiring internally to do this work can never be as useful as partnering with existing, local institutions who will continue to be around after the OLPC paratroopers move on to the next implementation. It's almost a sustainable plan, and it looks like there's some work to create a set of best practices:

Although each country has a different situation, they can learn from common experiences. OLPC plans on using Haiti, the poorest country in the Western Hemisphere, to test ideas about how to best integrate the computers with society and to create a template for other countries.

The (lack of) implementation plan is only part of the OLPC hubris on this project. Underlying that is the more insidious educational theory level of the program.

While this philosophy is essential to the mission of OLPC, it's also a source of tension. Current educational leaders in Peru embrace Constructionism, but most countries base their education systems on the idea that teachers pass their knowledge to receptive students. That was a problem for OLPC in China as well as India. India's education department, for instance, calls the idea of giving each child a laptop "pedagogically suspect," and, when asked about it recently, Education Secretary Arun Kumar Rath barked: "Our primary-school children need reading and writing habits, not expensive laptops."

Now, you can argue until the cows come home about pedagogical theories, but at the end of the day you must respect a country's sovereignty and right to choose its own educational track. Despite India and China's different approaches, it would be hard to accuse either country of not achieving some impressive educational outcomes and economic growth by following their current path. To close out with a quote by Easterly regarding the OLPC pedagogy, "It's arrogant of them. You can't just stampede into a country's education system and say, Here's the way to do it."

Almost universally, they’ve [OEMs] asked for standard Ubuntu packages and updates, with an app launcher that’s more suited to new users and has the feeling of a “device” more than a PC.

The implementation is, overall, quite ingenious in many ways, but there are still places where it feels a bit clunky. The project is clearly early in its development and we will likely see the rough spots even out as it evolves.

Beyond just a more device-like application launcher and a tabbed window structure; Mark also mentions "two companies that want more radical user interface innovation":

Canonical is participating directly in the design and implementation of one of those UI’s, and we’re integrating someone else’s UI on an Ubuntu base for the second project. I haven’t seen either of those UI’s, for confidentiality reasons, but I’m told that the teams working on them think they have great ideas that will elevate, in different ways, the state of the art.

Who is Canonical working with?

Mark concludes;

All in all it will be exciting to see how the netbook era stimulates innovation in the Linux user experience, because there are a lot of companies wanting to build differentiated UI’s on a standard Linux base. And directly or indirectly Canonical will help to bring that innovation to KDE and GNOME and hence to the wider Linux ecosystem.

With any luck, the 4PC market that the OLPC has helped to create will also spawn a new round of UI considerations which traditional software companies (Microsoft and Apple) will be interested in designing for as well, creating functional but light-weight versions of their OS (WinCE hardly counts, Apple's iPhone OS might be a sleeper candidate however).

When I saw this reappear after years of being out of production over at ThinkGeek it brought on a wave of nostalgia:

Embedded Video

Again, trying not to sound snooty here, but the lessons weren't very "new" either, but the way they phrased them were -- instead of speaking about crowdsourcing, peer-production and open source/sharing, the presenters framed the same general concepts in communications and strategy language like message control (it's dead), reader-focused theories of change, stakeholders/champions, voice and vision, and so on. This gives me more relevant vocabulary to use to champion the full-sharing concepts when I speak with nonprofits.

Read on for my run-down and links to even more event notes!

• Use the keyword in the ad title text (to capture the user's eyeball, since that's what they're looking for and just searched for)
• Second line (first of "text") should be clear short engaging description of what you'll go to
• *Third line is call to action
• 4th is the link

• Google Sites could be a great way to engage constituency or as an intranet with easy webpage/section setup by users
• Google Presentations can now effectively be a powerpoint-broadcast with live text chat; combine with a conference call number and it's a free webconference!
• Google Docs/Spreadsheets are really well set up for collaboration
• Google Spreadsheets can now create live widget-izable charts based on spreadsheet data!! (Another reason to use GoogleApps more!)

The second day was a lot more theory/what you can and should be doing with Social Media

Holly Ross of NTEN began the day
Her plenary reinforced the change in information and the role that nonprofits have played as librarians/translators, and how that must change to more of a facilitative role now. NPOs are now a bottleneck between their stakeholders and the information they're looking for; and NPOs need to find more and easier ways to provide this information and enable the users to then spread the message to their own networks. People don't need as much "basic information" like best practices and tip sheets -- your website is not a destination for that information anymore, google is. Your role is to be a better information broker/conduit.
Holly then laid out a compelling web strategy:

• Does your website help you LISTEN?
• Does it start conversations?
• Does it let users share?
• Does everything fit together (branding/interlinked, as well as technically)

She strongly encouraged the use of blogs by people at the organization talking transparently and honestly about the organization, as proving grounds, places to hone the message based on comment feedback and finding a valid and specific voice. She further encouraged people to blog/twitter completely separately from the organization, but as professionals, as a way to add a human connection/dimension and more places for stakeholders to comfortably interact.

Tucker McLean presented on SalesForce, which provides 10 free user-licenses to nonprofits. SalesForce is also integrated with GoogleApps, and can be used in a variety of interesting and flexible ways. We might even look at it to see if it can be re-aligned to be a GMS. Red Cross used it to track volunteers and partners in their response to Katrina, and the Center for Employment Opportunities uses it as a case management tool. SF has a 1/1/1 "strategy" for their CSR; 1% of time, 1% of product and 1% of equity goes to nonprofit causes, in 2007 alone SF employees did 77,000 hours of service in the form of consulting, CrossCultural Solutions service trips in Peru, and Habitat for Humanity work.

Social Media and Messaging was a topic I ducked into after the SalesForce talk; the big takeaway was "message control is dead" -- stakeholders are champions, not servants, and if they have a personal stake in your org's success, they will spread the message for you -- the message might vary, but it will be more powerful because of the network/personal connection.

MoveOn gave a talk on their view on email marketing/communication, focusing on two strategies, the moment and the movement:
Moment emails consist of three parts:

• crisitunity; crisis==opportunity; hooks for message/action
• rftc reader focused theory of change ; call to action with an acheivable goal that engages user
• ask: actual action that leads the rftc; show your support/raise your voice is not an ask; not a direct action; donate or write or etc. for a specific cause/event
• 3 parts; crisis/opportunity, enable/show community / ask : 2-3pargraphs before the link, goals and timeframes can help if don't make too long
• strength test: make sure top text is strong; add strengtheners under the fold to bolster any weak spots; credibility, show reasons why

More focus on the importance of letting go and empowering the users -- case in point; UNICEF had been uninsterested and even against engaging Facebook causes, until a 14 year old in Washington decided to create a Cause to donate to UNICEF; she built the group to 100,000 people and UNICEF started paying attention when they goy a check for $10k; but it was entirely user-driven with no management/communication oversight from UNICEF.

That being said, ongoing work in social media requires a staff commitment, both from current staff and from a dedicated "online community organizer"

The end-of-the-day closing comment was
"pick one thing and lose control of it"

Even More notes:

My full notes are at http://joncamfield.com/wiki4dev/index.php/2.0_NonProfits

ForumOne folks blogged more extensively about the conference

In a direct contraction to Nicholas Negroponte guidance at the November, 2005 IADB meeting, where he told Ministers of Education that "To do a pilot project is ridiculous!," the IADB is not only piloting OLPC, they're also going to have objective testing on the efficacy of a one to one education model

It's probably not an accident that the World Bank's InfoDev just published a Handbook on Monitoring and Evaluation of ICT in Education Projects...

I'm interested in the hard costs of the project, which weighs in at a total of $5,100,000USD. You can trace the evolution of the project back to the original November 2007 project summary and watch the numbers dance around as they move towards the final plan of operations, signed in March 2008. Between November and March they realize the need for school servers (at over $1,000USD each), energy and security solutions, as well as increase their setup, implementation, contingency, and measurement costs, (contingency spending alone goes from just under $16k to almost $260k).

Since the 5.1 million total doesn't shift, these costs are taken out of other parts of their implementation plans. IADB's laptop purchase goes from 7,000 to 3,700 (OLPC is providing another 10,000 XOs presumably from G1G1), training and content gets cut down by $200,000, and maintenance gets $30k shaved off.

The end result is focusing on 13,200 students plus 500 teachers (the November plan was for 19,000 students, but didn't have laptops for all of them or any teachers embedded in the budget) in forty communities. The student:teacher ratio would be a respectable, if not quite believable 26.4:1 . I'm creating a spreadsheet to hold all of this information with some formulas to sort it out that you can look at yourself.

What interests me is to see how my cost estimates from November 2006 stack up. I calculated for a 5 year total implementation process, and was using some slightly different numbers (the laptop cost $148 then, for example), which in some ways overlaps with the IADB budget. I tried to honestly extract a "first year" budget from my old numbers and re-arrange the IADB budget to group it into the same categories. the November costs are from the original November 2007 project summary, with the per-laptop costs just dividing by the ran numbers of laptops in the plan (not by the projected numbers of students and teacher recipients, which was a larger number). The March numbers are from the final plan of operations, signed in March 2008, with the per-laptop costs being the total divided by the number of laptops (they fixed the number glitch), and dividing the server costs among the laptops. The 5year projection is not just multiplying that by five, as most of the costs are designed to be one-time implementation costs, but tweaking where necessary to cover maintenance and so forth. The Jon5yr cost is my original guestimate, and the Jon1yr extracts the setup costs and divides the rest of the recurring costs by 5:

You can view the entire spreadsheet to see what went where in detail at this Google Spreadsheet of the OLPC/IADB costs with the formulae visible to see how I constructed my numbers (and the linked PDF files have the original arrangement of the costs if you're so interested).

Now, IADB doesn't break down connectivity very well, so it's unclear what part of that is installation and what part (if any?) is service provision, and it's this figure where my estimates appear way off. IADB calculates $3,000 USD/school for "connectivity" . According to a February Community News email:

Michail had a conference call with SES-Americom. They agreed to provide the (C-Band) space segment and internet termination (via their Maryland teleport) for our upcoming deployment in Haiti.

So I'm going to guess that that's $3k for a VSAT installation with donated bandwidth; with no maintenance budget or money for a paid subscription if the bandwidth donation ever dries up -- does the community have any insight on this?

As for the other "line items", I was $10 over per laptop on total implementation costs, and $2 over on maintenance costs, not bad. If I presume that an Internet connection is free of cost, my original number 972 less five years of Internet access (coming in at $541 using the UN dial-up costs discussed here), I end up $11.94 cents higher than a 5-year projection of the IADB's numbers, (not accounting for the change in the laptop price from $148 to $188).

So if anyone can provide some on-the-ground information about bandwidth costs in Haiti (all the ISP websites I found were broken in one way or another); we can see what the ongoing costs of this project will be.

XP on the XO

Update Unsurprisingly, it's not true. From James' blog, emphasis added:

As I have posted earlier, we had to write multiple custom drivers and a BIOS to get Windows to boot from an SD card in order to do the Windows port to the XO. This is the initial implementation customers will be able purchase when the product RTMs and will be a "Windows only" XO that Nicholas Negroponte himself has described as running "really fast." Customers can also choose to buy the existing Linux/Sugar XO. Longer term, the OLPC plans to write a new BIOS and increase the amount of flash storage on the XO to support a "Dual Boot" option that would enable children to use either Linux or Windows on the same machine. This is fine with us as long there continues to be an excellent Windows experience on the XO.

Recording audio

Recording video

"Sharing"

Power

Wifi

Security ?

Conclusions

XO is built from free and open-source software. Our commitment to software freedom gives children the opportunity to use their laptop computers on their own terms. While we do not expect every child to become a programmer, we do not want any ceiling imposed on those children who choose to modify their machines. We are using open-document formats for much the same reason: transparency is empowering. The children—and their teachers—will have the freedom to reshape, reinvent, and reapply their software, hardware, and content.

And I certainly don't see how he can with any straight face maintain that "The mission statement of OLPC has not changed in three years"