Tag Archives: low bandwidth

Follow on to ‘IT Next to the Beneficiaries’

This response is a follow-on to Jeff’s post which can be found below as well as some responses that were received to a cross post.  This is my take and hopefully Jeff (and others) will comment.  Thanks to Jeff for letting us know that once again we are still way behind in the solutions game.

What Jeff experienced is not new and is readily experienced by almost every aid worker that is deployed to the field today as well as those individuals we work with.  If you are in Nairobi or sequestered in a compound you are likely to reap the benefits of an over-sold but none the less fat pipe VSAT.  However, more often than not you are going to end up doing exactly what Jeff did in the Congo, what we did in Indo in 2007, and what our associates have done in every location around the world for the last 10-15 years – squeeze water from a rock.

This is not a new problem and yet it always manages to avoid a fix.  It is not so much an issue of how much bandwidth you have but rather how well you use that bandwidth.  These days many people believe that everyone needs enough bandwidth to be able to stream TV, watch endless YouTube videos, and download files in 2 seconds rather than 20.  Yet most of the world gets by just fine on a trickle of bandwidth, web mail and a chat client.  Given that fact (my fact) we should still focus on increasing bandwidth to emerging markets but also not ignore bandwidth conservation and the incredible impact it could have on accelerating the adoption of technology in the developing world.


Mobile networks will soon become the primary provider of Internet connectivity for most of the developing world.  Google’s investment in o3B (sort of) bears this out granted such a backhaul can also be used for WiFi/WiMAX, etc.  I still think that most people will be getting their mail on their phone or doing exactly what Jeff’s associate resorted to which was to plug his desktop into his phone and stick the antenna out the window.  Even if you do not have the connection you can still run a sneaker net just as they did.  And there is still good old dial-up.  Again, the connection is not the key issue here, the software is.


My guess is that most people in emerging markets are running bootleg OS’s which, while functional, are total junk.  You can walk into any bootleg DVD store and while you are picking-up the latest movie that was poorly filmed by some kid with a camcorder you can also get a cracked version of just about any software out there.  The problem is aside from bootlegs being illegal in certain parts of the world it is also hazardous material.  Bootleg anti-virus packages often come pre-loaded with viruses as do most cracked OS offerings.  It is just not worth it and no it is not the only solution.  Personally, I think Ubuntu is the solution.

Ok, ok, I can already hear the moans but who the hell is going to buy a real OS?  No one.  Who really wants to use a junk cracked OS?  No one.  Who does more than check their web mail, chat with friends and some minor word processing and spreadsheets?  Almost no one.  Really, why not use a FREE and readily available OS that is almost devoid of viruses.  Right there you can solve most of the problems.  Most of Macedonia has adopted Ubuntu so why can’t the rest of the world?


AVG FREE Personal Edition.  Enough said.  Don’t even waste your time on anything else.  This will take care of 90% of your problems.


I don’t care which browser you use as long as it does one thing: block adds.  Advertisements are the bane of the developing world.  Does someone in Nairobi really want to see a Flash animation ad or some gaudy banner ad when all they really want to do is check their mail?  No.  One of the best tools I think you can use is Adblock Plus which comes as a Firefox Add-On.  Therefore, use Firefox.  It is FREE and dramatically cuts back on a tremendous amount of bandwidth clogging imagery.

In the US, with our ridiculously cheap bandwidth (think oil until about 2008), we can afford to drool over insurance ads or truck ads but out there such a luxury is not an option.  Seriously, the mindset these days is that gobbling massive amounts of bandwidth is perfectly acceptable.  Does anyone realize how much it costs to power the backbone?

Web mail

Lastly, just go get a FREE Gmail account and stop struggling with all the rest of the services.  Use the Thunderbird client if you want to POP your mail but by stripping out the web junk you’ve already solved most of the issues.  However, if the line regularly drops out better to POP it.  Thunderbird integrates seamlessly with Gmail.  You can do a lot with Gmail that you cannot do with the other services.


How do you get the software to the beneficiaries?  Use the same delivery system that is already in place.  Rather than selling bootleg software why not provide the real thing?  Once they download one copy of FREE and available software they could either give away or sell a legit copy of something useful.  (I am not endorsing this option for every application as I am sure it violates a handful of EULA’s) .  But is it against the EULA to charge a ‘service fee’ to cover the vendors cost of downloading and copying the software but not for the actual software?  If folks out there can get a computer they can get an OS and in that case they could have a perfectly stable and efficient system for close to nothing.  What ever the case I am sure that all the vendors mentioned above would love to get their product into the hands of new customers.


What we see here is not a technology problem, it is an problem with education and advertising.  There are plenty of newspapers and billboards that are read/seen on a daily basis throughout the developing world and if Ubuntu can buy billboards in Silicon Valley along US 101 they certainly can cover the cost in select African locations like Ghana, Ethiopia, Uganda, Kenya, and South Africa which is their own backyard.  (Perhaps they already have an ad campaign in Africa – I have no idea.)  I am really not sure that we are the best choice for teaching people in the developing world how to conserve resources but bandwidth is our baby and if we invest a little up front and show them how to more efficiently use what we’re pushing their uptake will be greater than if we let them suffer with our clutter.  Ultimately, the ads will flow and the market will develop but we’re just not there yet.  Better to have smart users than none at all.

IT Next to the Beneficiaries

My friend Jeff has rescued me from my flu induced apathy by sending an excellent reality check highlighting the limits of IT in the field.  Here it is in full:

Jon writes here often about how cutting edge apps might be adapted for use by aid workers. As a tech guy myself, I love to read and dream about this kind of thing. But as a field guy, I’m unfortunately here to today to bring the dreams crashing down to reality.

The thing is, the IT life I have observed in the field, near to the beneficiaries, is not all shiny and high bandwidth. And “near the beneficiaries” is what matters. Coordination meetings with Google Earth maps projected on the wall, integrating UNOSAT imagery and placemark databases showing wells and refugee camps is the dream. But we must never forget why we are there: to work with communities to make the temporary and lasting changes necessary to let them support themselves. So that’s why the IT landscape near to the beneficiaries matters to me more than the IT context in the capital or headquarters.

Here’s a beneficiary for you: The man I’m thinking about is the data manager for a Congolese health district. He’s lucky: he lives on the eastern border of the country, close enough that the strong Ugandan economy supports his own struggling one. He’s got 30 kbit/sec Internet implemented via a serial cable plugged into a desktop mobile phone with an antenna out the window pointing across the border at Uganda. He works in a district office, so there are several computers, one of which is on his desktop, all donated by various vertical programs (the “AIDS computer”, the “Malaria computer”, etc). The WHO has supported the DR Congo strongly over the last decade, so he’s had the chance to travel 8 hours by motorcycle to the regional capital to attend classes in Microsoft Excel. He has a Yahoo email account.

During a meningitis epidemic, it’s his job to enter all the paper forms arriving by bicycle and moto from around the region into a spreadsheet setup by the WHO. It’s hard for him to get his job done… which is how I came to know him. My boss sent me over to him to try to find out why he could never give us the data we needed on time in order to make good decisions about our meningitis intervention. He explained that he can only work when the generator is on. Then he explained that sending files with Yahoo mail is slow, and doesn’t always work. Then when we looked at his computer to see why, we found evidence of several viruses making it impossible for him to open any document with “virus” in the filename. (I’d already noticed the evidence: each time our USB key came back from his office, it had a file named h.vbs on it.)

Across town in the MSF compound (a rented hotel), I was working with my own constraints. We also were dependent on the generator, since there is no electrical grid in eastern DR Congo. We had three laptops for five people, approximately the same ratio as the health office (2 desktops, 4 people putting significant time into meningitis). But because we had laptops, we could work when the generator was off. We had internet via BGAN satellite at 128 kbit/sec. It is expensive, at USD 5 per megabyte, meaning that we had a filter on our mail preventing us from receiving attachments larger than 2 megs.

The attachment limit turned out to be another problem. On the off chance that the generator, the phone, and Yahoo mail worked, and he managed to send the week’s data out to us via e-mail, we couldn’t receive it, because it was too big. We solved that problem by getting in the Landcruiser and driving across town with a USB key. Which, if you’ll recall, is how I knew that I should go over and remove his viruses.

How do we remove viruses? It’s easy, right? Update your virus scanner and scan. But an update is 12 megs, or USD 60 on satellite, and completely impossible to download on a 30 kbit/sec link from Uganda. So no update. MSF solves this problem by sending virus update CDs to the field monthly. But the viruses you catch in Africa tend to be locally written (this one was from a technical university in Nairobi). If they ever end up in the virus databases at all, they arrive late. Certainly later than my most recent update CD from Geneva. I got lucky and found instructions online for how to edit the registry and remove the virus. But I’m a programmer… who would have fixed it without me?

Keeping computers working next to the beneficiaries gets a lot harder as you pile on the constraints, doesn’t it? But it’s probably worth it, because of the benefits they can bring.

Or is it?

In Myanmar, my girlfriend worked 6 days without a computer doing mobile clinics, then came back to base on Sunday for a rest and to enter stats into the MSF spreadsheets. By necessity, she finished her stats and weekly report in an hour of work Sunday afternoon, then got on with the real business of resting up for the next week of mobile clinics. She said it was liberating to not use a computer in the field… which might be a lesson for us as well.

Here’s some tech lessons I’ve learned in the field:

  • Use POP, not webmail. Slow, unreliable, expensive connections are the enemy of web apps.
  • Don’t expect to be able to use web apps, even the ones your headquarters says you should be using.
  • Have lots of USB keys around to move documents
  • Expect viruses and struggle on regardless, there seems to be no workable answer to this. (At one Congolese internet cafe, the operator scanned USB sticks before allowing us to use them. I gave him my USB stick clean, and I got a virus back from him. Great.)
  • The simplest thing possible might work. Anything more complicated won’t.

The number one lesson: The work is worth doing, and to do it right you have to be out there, at the edge, in the mud, with the beneficiaries, in their communities. If the IT can’t keep up, leave it at home.

Opera Beta Offers Email “Low Bandwidth Mode”

Lifehacker has a story about Opera’s new browser featuring a “Low Bandwidth Mode”.  This could come in mighty handy for those of you stuck in the bush.  When is the rest of the world going to figure out that most of the world fights the low bandwidth battle every day.  From the post:

A new beta of the Opera web browser is now available for download by willing testers and adds a “low bandwidth mode” email feature for when you’re on skinny pipes (or just trying to save a few bytes), along with enhanced browser data sync between remote computers, email client improvements, and a better RSS feed preview. Let’s take a look.

Read on…

‘Low Bandwidth Networking to maximize informations per dollar.’

Renaud Gaudin of Geekcorps-Mali has an excellent summary of ways to optimize your network when you are stuck with an over-priced and often over-sold VSAT connection.  These low bandwidth/high latency networks are the bane of the humanitarian aid industry and suck up millions of dollars every year that would normally be used for purchasing medicine, food, supplies, etc.  Renaud writes:

To use the Internet more effisciently and get the max of those tiny 200KB, we have used the following strategies:

  1. Cache web pages on the client computer at the radio station.
  2. Force the client computer to communicate only with a central server.
  3. Have the server strip images and ads from web pages.
  4. Have the server email news summaries to the client.
  5. Transfer email between client and server just once per day.
  6. Strip attachments from email messages.
  7. Compress all communication between client and server.
  8. Disable web access when the daily quota is exceeded.
  9. Provide continuous feedback of daily network usage to help the user learn to efficiently use the Internet.

He goes on to link to include a link to a Low Bandwidth Networking technical document.  This is a fantastic document outlining the use of a Squid Proxy, Loband Filter, and even Ubuntu for running a lean and mean system when faced with major connectivity hurdles.

Defeating the scourge of the dirty, low bandwidth/high latency over-priced networks is the last great battle that has yet to be fought.  Humanitarian organizations dump more money into bad connections than they chose to think about.  All the philanthropic assets of the IT industry should be trained right now on solving this priority issue.

Chris Wilson and the folks over at Aptivate.org have been wrestling with this 400lb. gorilla for years now and they created Loband.org, wrote ‘How To Accelerate Your Internet’, and contributed to host of projects in an effort to defeat it.  I started my own NGO, Humanlink, also in an effort to develop a solution to pressing humanitarian telecommincations and technology problems.  We ran a successful mission in Indonesia last year where we rolled out Clark Connect Community Edition boxes at no cost to the organization we were assisting.  Mercy Corps has been using Clark Connect for years due to it’s web caching (and other) abilities.

Shrink Pic – Simple photo compression for the field


I should have written about this long ago when I first heard about it but unfortunately this blog did not exist back then.  A fine Dutch gentleman in Indonesia mentioned that I might try it as he found it worked splendidly over low bandwidth networks and indeed it does.  What is it?  Shrink Pic is a simple little app that floats around in the back of your machine bothering no one and only steps up when you email a photo to a friend or send it via IM.  Shrink Pic automatically compresses the image before the transfer and what normally takes minutes now takes seconds.  You don’t see it, it doesn’t bother you and, best of all, it is free.

I love those folks who make our lives easier for free.  Download it here.

The beauty of forgotten networks


Forgotten networks are beautiful things. They are the unheralded heroes of the rest of the world. They are the pathways through the rice patties along which the rickshaws of the internet carry all the world’s staples. Unacknowledged by the networking gods and shunned by their flashier, sexier counterparts they move the data that none of see but that all of us consume. Made of millions of always failing crappy connections infested with viruses and junk they somehow seem to function in spite of it all. Nobody wants them but everyone needs them.

I just spent a very pleasant evening with Yahel Ben-David of airJaldi.com at the Intel Research Lab in Berkeley discussing these forgotten networks. We discussed how we could detoxify them and breathe new life into them as he is doing in Dharmsala and as we have done in Indonesia. Yahel is developing some amazing technologies out of simple packages that have been around for years but that have proven their worth time and again. We found value in these same packages when we rolled them out in Indonesia in the form of Clark Connect.

Forgotten networks are veins of pure gold that may once again draw prospectors. They are the overlooked and undervalued information highways that, with a little maintenance and TLC, can bridge the gap between the ‘haves’ and and ‘have nots’. To bridge the Digital Divide look no further. All that is needed now is that initial first step, that leap of faith.