My deep condolences to all those who receive lame April Fool's Day jokes today. It's an unfortunate side effect of data being delivered at the speed of light.
When at regular speed, I found the internet connections in Mali to be generally quite good. They are after all linked up via Senegal and thus they are drinking from the top of the data hose for the connections that wrap around West Africa and go down to South Africa. But while the speed and general latency were quite good, getting online was quite hit and miss.
It was typically the case that when arriving at a hotel in the various places I stayed and finishing filling out my passport information, I would ask about wifi and what the password to it was. Then, in some kind of Mission:Impossible fashion, I would be slipped a piece of paper that had something like the following written on it:
121kja98889akka237asaaak9887aasm2
Sure, I'm big on security and having strong passwords, but this is a bit ridiculous, especially given that it's a basic WEP (and thus easily hackable) connection. At first I thought this was a one-off situation, but found it to be the case time and again anywhere I stayed that had an Orange internet connection, which was everywhere.
It also happened that I would try to connect and when the router was assigning an IP address to my machine, it would sit and think, give and address, reject it, and then assign it again. From my days as an IT manager, it looked like the DHCP had run out of leases or that there was no way for a new computer to sit on the network at that hotel, which was weird given that are 255 available addresses in any basic network setup.
Near the end of my trip, I stayed at a hotel where the owner was a web developer. I asked him what the problem was to which he told me one simple word: "Livebox". The way things are set up is that in Bamako (and I assume in most of Mali), they're now running all the new phone lines and internet over WiMax. The terrain of Bamako lends itself to this extremely well as the town somewhat sits in a bowl around which antennas can be placed for excellent line of site from homes.
That's the good news. The bad news is that the Livebox is the device people get in-home to gain access to all of their data goodness. The Livebox has a multitude of problems, the first being that the passwords for wifi are ludicrous. The second being that the PPPoE system built in to it is worthless. The third is that the DHCP is by default set up to only hand out three addresses. Yes, only three local addresses are available, which is asinine, and why I was getting bounced as it does this for both wifi and LAN connections. Even 10 would be nuts, but still doable. The fourth big problem is that they don't give users access to any of the configuration settings for the Livebox so people are stuck with these dumb passwords and lousy DHCP setups.
I am now going to let you in on a secret. If you want to access these configuration settings which should be at 192.168.0.1 (or .1.1 as I can't remember), the username is your account username and the password is that same username. That's it. Type that in there and go wild or set things up to bypass most of the Livebox altogether. I have no idea what Orange was thinking when they decided to deploy devices in this manner as I'm sure it must give them more service calls than just relaxing the restrictions.
The BBC has put together a nice, animated map showing how internet penetration has spread over the world up through 2008 (not sure why 2009 is the ugly stepchild year in all of this.) As is usual, the big blank spot on there is Africa, with the exception of Morocco, who seem to be doing pretty well, most likely due to the proximity to Spain which allows cable to be deployed quite rapidly.
Nothing on this map should come as a surprise to anyone following African technology and it's all part of the BBC's large SuperPower report they're doing on how the net is changing lives around the world. But, that said, it's not changing the lives of too many in DR Congo with the 0.45% penetration rate there that they even make special note of, although I assume that will change a good deal once their inland cable reaches Kinshasa and potentially, 10 million new users. Also on that note, not including 2009 really skews things a just a wee bit as the East African cables have changed the game a great deal on that side of the continent.
Anyways, a map worth a look is you enjoy all thingies animated.
So, I've been settling in to my new home well. Admittedly, as I saw, all the action is down in Abidjan and Abengourou is a bit slower pace. It's relaxing though and as there is ready internet here, I have no complaints. There really aren't too many choices in landline internet and as far as I know, Aviso is the only one in these parts.
At $45 USD a month for 1.2Mb down and 384Kb up, as internet goes in West Africa, it's pretty reasonable and so far, pretty reliable. Yes, there are times when it is slower than others, but that just makes me have more of a purpose in online activities. The farting around aspect in being online gets largely reduced.
The only problem I'm having (and this is where the request for help comes in) is that this package came with this clunky USB modem to connect with. There are two and at times more of us who need to use this connection, so this simply won't work as it is a one to one setup.
From the US I brought two ADSL modems (a Zxyel and a Brightport) as well as an Apple Airport Express. I figured that at least one of the modems would work and that the Airport could be set up with a PPPoE connection to bypass this whole USB modem garbage. It turns out that this doesn't work and I'm stumped. In theory, everything should be happy. The Airport is set up and everything gives the appearance of connecting. The modem has a solid DSL light and seems to be happy. But, when it comes to actual browsing; nothing. Pages don't resolve, nor do attempts to ping outside addresses.
At this point, I'm kicking myself for bringing the Airport because, while compact, easy to carry, and hapy to switch between 120/240v, it is also without a great many options and I can't configure the WAN side of things as I would like to. So, is it the case that I need to spend another $100 of Aviso blood money to get their router/hub (which is the upgrade to this USB modem) or is there simply some setting somewhere that I'm not aware of?
Any help would be greatly appreciated and earn a Flag or Castel (the big one, not the small one mind you) upon sucessful wifi-ification of this connection. Thank you!
A couple months ago, I wrote about the Central African Backbone which was in the process of being planned. More news has come along on this front and it runs a bit counter to what seemed to be the plan in my previous article. The good news in all of this is that, according to this article, the World Bank has indeed fully committed to the development loan:
The Board of the World Bank has approved total project funding of US$215 million, of which US$26.2 million will go towards the first phase of developing each country’s national backbones to give them access to the international landing station in Cameroon.
The only thing is that instead of the line running from Algeria to the south, apparently the line is going to run from the Cameroonian coast, inland to the east, in order to connect Chad and CAR. The goal being to obviously favor countries starting with the letter C as C stands for 'connectivity' or something like which I've just made up... The genius of this plan is that they expect to run the cable along an oil pipeline that terminates at the coastal town of Kribi in Cameroon. Sounds good to me if it means that people will get proper terrestrial connections in the very near future.
This recent article makes a brief mention of redundant connections, which may be where the connection up through Algeria comes in to play at some later date. The ambiguity goes a long way to explain why the CAB hasn't made as much news as it should because in reality, the coastal cables are incredibly easy in comparison to connecting up those who are further inland. Hopefully we'll get more coverage as the deployment proceeds and solidifies.
I think that it's been for the last 10 years or so that people in the US have been hearing about the amazing possibilities of running data via power lines or PLC. If PLC doesn't sound familiar, it's because Americans were mostly familiar with just one aspect of it which is BPL or Broadband over Power Lines. To date, it's sounded a great deal like snake oil and to a large degree has fallen by the wayside. I hadn't really thought much of it as it seemed to be dead in the water.
It turns out that it actually isn't and in fact, PLC is very much alive and has the potential of being deployed in Africa on a large scale. I got this out of a session I attended at the BarCamp Africa UK which was coordinated by Andrew Boye and Kofi Abban-Sackey of Cactel Communications. In all truth, they were there to pitch the work that their company was doing, but at the same time, in addition to an introduction to their firm, I think that everyone in attendance was able to take away a great deal from the session about PLC in general. Given the questions asked be all of us, I think the potential possibility of this technology was largely unknown.
The biggest point that stuck out at me was the figure of technology penetration, in that power lines have a 55% penetration rate, mobile phone a 30% rate and landlines a 5% rate. Look at that. There is such a big deal being made about data connectivity via mobile phones in Africa when power lines already have nearly twice that rate! This is an untapped resource which could readily be taken advantage of, today.
Naturally, living most of the time in the US and seeing a presentation on power line connectivity, I had to ask, why is it that this would be viable in Africa? Yes, the ability to reach a great many more people is there, but we haven't seen it implemented in the US and like I said, it seems to have just died out. I would never ascribe to the belief that something has to first work in Africa to then work in the US (actually it's quite the opposite), but at the same time, the last group of people in the US to get access to broadband would be readily served by access via power lines as they are quite rural and not served by any connectivity options other than dialup or satellite. This hasn't happened. Why? And how is this able to be mitigated in an African setting?
It turns out (and I was very unaware of this, although it makes sense) that profitability for a power line data connection node is the same as it is for cable, phone, or anything else in that if you have very few people on that node, you will have a hard time making any money from it. This is the reason that it's never taken off in the US. While it would work great in a setting like New York City or San Francisco, there already is a great many options in these places and so installing it doesn't make sense. Stick it in the middle of the prairie in Kansas and you'll lose money. Where it would make sense in Africa, is a place like Accra or many other locations. The power system is already there and all you need to do is plunk down a node for about $200 USD, configure it and off you go. This is in sharp contrast to having to lay new fiber which could take months if not years.
Andrew and Kofi nailed their talk and there simply wasn't enough time to get in to everything that they presented. We ended at their showing a map of a demo project they set up on the University of Ghana campus. There are obviously a great many fine details that were not gone through completely, but I came away from the session with a renewed interest in PLC and feeling pretty good that this could very well be a solidly viable connection option for over half the people in Africa.
I've recently come across the site Online Africa which really picked up publishing steam in this month of October. The site has great loads of promise, working to cover connectivity issues across the African continent. No small feat obviously as there are a great many countries and a great many issues.
One of the things that the site is trying to do that makes it relatively unique in what it's covering is to break down data in to individual countries. There is a great deal of data out there and it's good to see someone else trying to make use of it and present in an easy to digest form.
Of course, there are others writing about these topics as well, so people may ask as to what is the point of another blog? I would counter that the point is to have more of a conversation and to have as many people covering what happens in regards to the connectivity for the billion people living on the African continent. When Apple has the smallest of bowel movements, there are literally thousands of blogs echoing news about said bowel movement. When a cable lights up internet for millions in Africa, are there even 100 blogs that cover the event?
So, I'm really happy to see Online Africa start up. I wish the site a great deal of luck and hope it will grow and cover more over time.
A great deal of articles have been written about all the East African cables being deployed. Rightly so, given that connectivity is paltry and slow at the moment and about to get a good deal faster (hopefully.) But while this is all good, when you read about how fast the cable is laid out in the ocean (10+km a day), you realize that it's really the inland part that's tricky and there's a lot of inland land in Africa. Sure, you don't need a boat to make it happen and the ever-present media-spawned threat of pirates is less, but the issue of cutting across fields, farms, and most importantly, international borders on land is pretty daunting.
I suppose it's because they haven't broken ground on the project yet, or probably more to the point that the majority of coverage has been in French, but the Central African Backbone is starting to gain a bit of momentum. (Please add to the Wikipedia link if you know more as I had to create it when writing this article.)
A good deal of what I know came from this article on ZDNet in French which covers the basic layout of the cable and the fact that it will most likely start in Algeria and connect to Europe to the north and Sub-Saharan Africa to the south. While there was a little bit mentioned about this last February, it's the fact that Algeria has decided to really set forth and start laying the cable that has garnered more worthwhile attention lately. Their Information Minister is pushing it in the name of getting rid of the satellite connections and helping Algerians better connect to the internet. But in reality, anyone smart knows that this is going to be a major cash cow for Algeria once they punch through to the other countries including Chad, Cameroon, and CAR. All of that comprises Phase 1 of the project and I'm not exactly sure how they're planning to connect Chad with Algeria and not be connecting Niger which lies between the two. Apparently there is a bit of a "and then a miracle occurs" aspect to the planning currently.
Phase 2 of CAB is even more in the Wild West portion of planning as a great number of countries have been tossed around in the mix to connect. While Nigeria might be there, the most probable candidates are Gabon, Equatorial Guinea, São Tomé, and Congo-Brazzaville. Congo-Kinshasa is also on the list to possibly be connected. Of course, Kinshasa could be very well have its own link very soon, so it's not quite as crucial for that area. What would be more impressive is if Northern Congo-Kin could get in on the link coming in to CAR assuming that they would naturally run any fiber link to Bangui, the capital, which sits against the border with Congo. If only Mobutu would have lived to see the day when Mad Men could be live streamed at Gbadolite...
Unlike the Globacom cable, this initiative is indeed being financed be foreigners, namely the World Bank. You can view a brief overview of the loan, here. They quote Phase 1 as being $30 million USD. The ZDNet article quotes it as being €17 million, so I don't really know which figure is for certain. Phase 2 is set to cost in the neighborhood of $160 million, so obviously due to the amount and the number of countries involved, it can be understood as to why it's so undecided at this point.
However the pieces fall, once the links start getting put in to place from Algeria, there will be a world of change for the interior of Africa which usually relies solely on VSAT connections that are obviously better than nothing, but still suck. As to speed for CAB, I couldn't find any hard figures. It is said that Cameroon will have 12 optical fibers and Chad will have six, which doesn't make a lot of sense, nor does it really say much about speed. I guess once things actually start happening over the next year, we'll get a better sense of all this.
I came across this good read yesterday about the closure of Senegalese internet cafes. For those reading this blog right now, it might be a tough go as it's in French, but Google Translate can be a good friend if you want to look it over and get a tasty overview of the internet situation in Senegal currently.
Basically, it comes down to the fact that the monopoly is a huge problem in Senegal. Sonatel (who apparently offer the iPhone) is the current government-owned monopoly on the main internet connection to the outside world. Naturally, since they have no competition, they don't really care about offering competitive prices. No matter if it's Orange, Tigo, or relative newcomer, Expresso, it doesn't matter that they offer data plans as they all have to buy bandwidth from the Sonatel connection and costs work out to generally the same. This is a common problem throughout most of Africa unfortunately.
In Senegal, the situation was exacerbated by the fact that internet usage took off and people got in to getting online. This in turn made for more internet cafes. More internet cafes made for more competition. The prices for users dropped from about $3 USD an hour to $0.65 per hour. This of course made for even more people getting on the internet, but only once those price points were reached.
These prices it turned out, were not tenable for business. Despite a massive demand, internet cafes have been closing left and right simply because they can't afford to pay the connection fees and offer service at the prices people are able to afford. Naturally, the connection fees didn't drop with more usage, which has been the inverse of this situation in the US, Europe, and Asia, with a number of Asian countries making out the best on price vs. speed (Japan is spoiled rotten with 100Mbs per second at $25 USD.)
Tossed in to this is also the case that a growing number of people have been able to connect at home through the growing ADSL network, especially in towns such as Dakar. It's all been downhill for those on ground trying to provide cafe connections for the masses.
While the circumstances are a bit different, one does have to wonder how much life an internet cafe has left anywhere in the world? Outside of tourists locations, they seem to be drying up everywhere to some degree as more and more of us travel with laptops or at the very least, wifi/highspeed data enabled phones that can do simple browsing anywhere we go. And this is a trend you see in Africa as well. Like I found in Cape Coast, Ghana, there are a number of growing cafes that have ceased to be the typical cafe of the past. They offer more of a place to plug in and work for the day, aka, the coworking space.
Obviously when your outbound connections are controlled by one entity, any change in the market is treacherous as the free market is anything but. Even still, I'm not sure that only connection costs are to blame for less internet cafes in a country like Senegal. User habits as to how people get online are changing the world over and I ask if the public internet access point soon be a relic? As I watch people wondering the streets, checking Facebook on the iPhones, I say yes, although it's a yes I don't necessarily like as the 21st coffeehouse was something I was enjoying.
So 2009 appears to be going down in the books as the Year of the Cable in Sub-Saharan Africa what with all the cable lighting up in East Africa and now the Glo-1 landing in Lagos over the weekend which will provide additional bandwidth to West African countries. Friends in Ghana were excitedly talking about this when I was there last month as it will open up a second route out of the country above and beyond the single line that they currently have which is tied in to the old SAT-3 line.
Throughput is going to start out at 640 Gbs and eventually be cranked up to 2.5 Tbs. There's been a bit of coverage on it which you can read at 27 Months, TechMasai, Vanguard, ITNewsAfrica, and This Day. I'm probably missing a lot of others as cable landings are a pretty big deal and this one is made even more so as it's being deployed primarily by Globacom Limited who are a Nigerian company.
The only thing that should probably be mentioned in all of this is that from Ghana to Senegal, the cable leapfrogs six countries (yes, I am indeed counting The Gambia in there). I'm not sure if this is because these other countries didn't want in on the connection (which colleagues in Côte d'Ivoire tell me is often the case) or if the countries were simply left out. It's a shame about this as a lot of Information Ministers are going to hobble their countries in the future if they don't have enough connectivity coming in. They'll have to run connections to neighboring countries who do have it and then pay a premium for something they should have had directly in the first place. So it goes apparently, but bandwidth is decidedly becoming a hard currency around the world and countries need to get in on it when they can.
Given the focus of what I'm doing in the coding of Maneno, I am very concerned as the state of bandwidth and availability of internet in African countries. While traveling through through Accra, Cape Coast, and Kumasi, I got a decent taste of the speed and reliability. I have to say that overall, it is quite good. In the internet cafes, the speed worked out to somewhere around 256 up and 1.5 down. The down was faster at various times, but the up was never much more than 256 as far as I could note.
The only noticeable problem was what I call the "bog effect". I'm not sure if it was the line or the number of users online at a certain time, but everything would be chirping along fine and it would be like a bucket of cold water was dumped on the connection as everything would suddenly time out. Thirty or so seconds later, everything would come back up and be fine. It's more of annoyance than it is an actual impediment and I even have this happen every so often on my home connection in the US.
As stated earlier, there is much more of a latency problem than there is a bandwidth problem. Even when you go farther in to the interior in a place like Kumasi, the speeds are the same as they are on the coast where the cable touches down. Of course, outside of the major towns and cities, connectivity drops off massively. However when inside the towns, the only real barrier to widespread internet access is cost. While paying $0.65 USD an hour to use an internet cafe is quite cheap for me, someone who earns maybe $20 USD a day is going to have a hard time being online. The only free connection I've encountered was at Smoothy's in Accra (where the Ghana blogging meetup is) but you'll need to have your own laptop to make that happen.
It needs to be noted that when there is a direct correlation between distance from the Atlantic, speed, and cost. For instance, I was told that in a town like Tamale, which is about 2/3 of the way up north, speeds are massively slower and massively more expensive for that slow speed. The issue being that "lovely" Vodaphone bought out the original company phone system of Ghana Telecom and has a monopoly. The backbone of cable running inland gets more and more narrow the further north you go and they have no interest in expanding it.
But in contrast, a town like Takoradi, which is near the western edge of the country on the coast, there are some of the best connections in the country. Why? Oil was discovered there this year and so countless foreign companies are setting up shop to tap in to it. They want to have network access and so bandwidth is getting a huge bump there currently. I heard this from some of the "gang" I met up with in Accra who often find themselves zipping out there for a day to work on support issues and the amount of trips they are making is only increasing.
Home connections are a different issue altogether. Very, very few people have them. The cost is exorbitant and with Vodaphone having bought out the state telecom a short while ago, there has been little care as to whether people can actually afford this or not. The other problem is that the bandwidth is nowhere near as good as for business connections. In fact, from everyone I've talked to who has it, it's insanely slow. It only picks up speed between 02:00 and 05:30. Those who really want to do all their hardcore internet work often stay up until then to do it. Needless to say, the geeks of Ghana are a young group.
In general though, Ghana is in pretty good shape for connectivity and bandwidth even though apparently 70% of it is hogged up by video sharing sited like YouTube (according to some fellows I met in the Busy Internet building.) The only big problems I've encountered were ones like last Saturday where apparently the entire link for the country went down and while connectivity within Ghana was fine, anything outside was impossible. That's obviously going to be an issue when you only have one link with the rest of the world. Just another argument for developers to someday host their local sites locally.
