Friday, May 01, 2009

Google opens up Mapmaker Data for Kenya

This is exciting enough to rouse me from my blogging lull.

Google announced in this blog post that they are making the map maker data for Kenya available under a pseudo creative commons license for non-commercial use. Governments, non-profits and individuals are allowed to use the data in any non-commercial way with attribution.

If you remember from my previous blog post, my reservation about map maker was that the data was not owned by the public. All though this announcement does not put the data into the public domain it goes a long way to eventually making it happen.

Another positive is that this opens the possibility of data portability with Open Street maps.

This is exciting news and I hope this trial is successful so that more countries can get access to their mapping data.

Technorati Technorati Tags:

, , , , ,

Friday, November 09, 2007

Mapping the "Dark Continent"

Mapping services like mapquest, yahoo maps, google maps and map-based gps units do not work in most African countries. This is because there is no GIS data for these places. This is a problem that needs a tropically tolerant solution and I am glad to say that Google might have the answer.
Google is using crowdsourcing to solve this problem in India. What they are dong is providing software that allows local people to draw/overlay GIS data on top of satellite imagery. This project came to light at the Cambridge Conference in July. Michael Jones, CTO of Google Earth talked about this in his presentation at the conference. Phil Bridges has an audio recording of the talk in this post and the relevant part of the talk has been transcribed by Dan Karran in this other post.

I think getting the data is one part of the solution. The second part of it, is how the data is used. In Ghana, most of the streets do not have street signs. So giving someone directions based on street names and house numbers will probably not be a tropically tolerant use of the data. The question is will Google go the extra mile and provide directions that make sense in the appropriate cultural context ?

I am also curious about who ends up owning the data ? It will be great if Google allows the community to own all of the data or a portion of licensing fees, etc ?

Technorati Technorati Tags:
, , , ,

Thursday, August 30, 2007


In a comment to anti-social-bookmarking, Joitske Hulsebosch pointed out webmark - an online bookmark manager by Henry Addo.

Webmark tackles the bookmarking problem presented in anti-social-bookmarking without the secondary features of social-networking and folksonomy. The feature set is complete for managing and keeping bookmarks online. To use webmark, all i had to do was create a free account and then define some basic categories/tags.  I could then use a webform to submit urls. The FAQ provides a bookmarklet that can be used instead of the webform. In addition, webmark also allows users to post the same URL to This however requires the user to trust webmark with their credentials. I find this problematic especially since webmark makes no statements about the limits within which it will use the credentials.

Webmark, is a simple straight forward online bookmarking product. There is still some improvements to be made. It would be nice to have free-tagging. This means that a user does not need to define a category before saving a url under that particular category. Also, being able to save under multiple categories will be a good additional feature. There is also no clear definition of who owns the bookmarks. did a very good job of stating that bookmarks belonged to the users and provided a way for users to download/export the url data. I think webmark should implement this feature to make the product more attractive.

In conclusion, I think webmark is a good example of a tropically tolerant solution to bookmarking and I look forward to seeing it grow. I also hope this is the first of many more tropically tolerant solutions from Henry Addo.

TechnoratiTechnorati Tags:
, , , , , ,

Wednesday, September 27, 2006

Idea for a payment network

In my last entry, I suggested that it was time for the average Ghanaian developer with an Internet connection to participate in the new web economy. I concluded that the main barrier to entry is a distribution network for the money being earned.
Today, I present to you an idea for a tropically tolerant money distribution network.

The problem statement is this:

  • Ghanaian banks do not make interbank transfers easy.
  • Its really expensive to transfer money from outside the country.
  • A good percentage of the population don't use banks

Whiles it is true that most Ghanaians do not use banks or have bank accounts, most Ghanaians have mobile phones. There is a joke that even the beggar at the stop light has at least two cell phones. The pervasiveness of mobile phone networks can be used as a tropical solution to the money distribution network.

Mobile phone operators have already established large cross-country distribution and payment systems. In order to establish a globally integrated payment network - there is a need to implement a system that connects say Paypal to the local mobile phone network. The implementer would have a Paypal account for incoming/outgoing international systems. On the ground in the tropics - there would be a need to implement and deploy a system that interacts with the mobile phone operators payment system and the Paypal API. The design / implementation should be minimally invasive to the already established mobile phone operators payment system. This will make it easier to create buy-in from the mobile phone operators.

To better illustrate my idea here is a use case:
The Characters:
Kofi Babone - a Ghanaian web designed based in Takoradi, Ghana.
Abrotsire Fashions - a Zambia owned clothing retailer in Canada
Shikatse - a Ghanaian company that implemented the system i am proposing.
Spacefon - a mobile phone operator in Ghana

Kofi Babone gets a job to design a website for Abrotsiri Fashions in Canada. Upon complention, payment is made to Shikatse's paypal account with a unique id for Kofi. Shikatse then credits Kofi's Spacefon account in Ghana with the money minus the paypal's commission, shikatse's commission and spacefon's commission. Shikatse's system notifies Kofi via an SMS text message that his money is available. Included with the text is an encrypted password. Kofi uses his personal password/key to decrypt this password and goes to the nearest Spacefon payment center. He presents the unencrypted password and he is given his money. He walks out with a smile and heads to the nearest akpeteshie bar.

Issues and Assumptions:

  • Creating buy-in from the mobile phone operators
  • Assumption that all payment centers are part of a wide-area-network(WAN).
  • I haven't looked at Paypal's licensing policy
  • Laws - there might be some laws that could prevent this
  • Ease of integration with payment system

Thanks to Profit for the initial seed in this idea and to Paypal Mobile for the proof of concept.

TechnoratiTags: , ,

Thursday, August 24, 2006

Earning dollars in a cedis economy or Participating in the new web economy

About two months ago, during a fienipa conversation with Paa Kwesi, we started talking about search revenue. This inadvertently led to a discussion about firefox and flock. The conversation about flock is quoted below:

  • me: flock has a similar idea

  • Paa: ah, like the search box
    i know one of the flock guys

  • me: who? the yale MBA guy ?

  • Paa: yeah, hehe,
    kwee, this guy has become popular damn

  • Paa: well, he's a whiz
    another discouraging thing when i think ghana wants to compete

  • me: compete ?
    yeah - we can compete

  • Paa: not in ghana

  • Paa: maybe out here where i sit together with "flock guy" in class and brainstorm together

Edited for grammar and brevity

Yes, we cannot compete in Ghana but we can compete in the global village. Just to be clear, I am not saying Ghanaians cannot compete but rather the web economy in Ghana is virtually non-existent. There is however no shortage in skills and a desire to earn dollars.
How should the people with these skills generate dollars ? Simple - build an outward facing product.

By outward facing I mean a product geared towards an audience in developed countries. Lets take the dollar theme a step further. Ghanaians should build a web product for an American audience. How ?

The following are required to build a web product

  1. Developers to develop the web application

  2. Hardware to host the site

  3. Revenue generating model

  4. A product

Developers is a non-issue.

Hardware to host the site is also not a very big deal especially in the beginning stages when traffic is low. Once the product achieves viral status, scaling of the hardware becomes an issue. However, services like Amazon's S3 and EC2 look very promising. They are currently not very cost effective but the hope is that it is just a matter of time.

Revenue generating model - this is a no-brainer: ADS. Why spend the time trying to figure out complicated pricing models when you can subscribe to the ad-network of your choice and boom revenue is generated so far as you get enough site traffic. However, there is a catch. How do you get access to your hard earned dollars ?

The Ghanaian developer in Ghana has no easy way to get the ad network in America to transfer the money to him. Paa Kwesi elaborates further on this challenge. The challenge of an easy payment system that is globally integrated is a key problem that needs to be solved in order to participate in the new web economy.

Coming up with a product geared for an American audience is difficult but not impossible. There are two possible tracks to follow.

  1. Come up with a unique idea from scratch

  2. Take an existing idea and add a unique twist

It seems like most of the new start-ups are going for option 2. For example there are a number of social book-marking services available each with its own quirk. This week alone two web-based instant messaging start-ups launched. But I digress ...

The point is there is money to be made. Ghanaian developers only need to come up with a globally integrated payment network, an outward facing product and voila: dollars.

TechnoratiTags: , , , , ,

Monday, June 19, 2006

Internet connection not required

In this previous post, I mentioned that one of the ways to handle the connectivity problem of a tropical region is to cache data locally and sync with a central server when a connection is available. Well, guess what ?

A company called Webaroo has done just that. Webaroo recently released a free beta of their web unplugged service. This service allows users to search and browse portions of the web anytime, anywhere and unconnected. It accomplishes this by crawling the web, analyzing the pages and creating topic specific web packs for download. The downloads are stored locally on a users computer and served via the Webaroo server/client over local http. In addition it allows a user to cache and index specific websites for offline browsing. The locally stored data is updated when the user goes online.

The Webaroo server/client is a free download with the following system requirement:
Laptop/Desktop System Requirements

  • Windows XP SP1+ or Windows 2000 SP4
  • 1 GHz+ Pentium class processor or equivalent
  • 500 MB RAM (Recommended 1GB for larger Web Packs)
  • Internet Explorer 6.0+ or Mozilla Firefox 1.5+
  • Microsoft .Net 1.1 (Downloaded during installation if required)
  • Broadband Internet connection while downloading and updating content
Sadly, the system requirements are not tropically friendly. In addition, on my DSL line(384Kbps - 1.5Mbps), it took about ~15-30 minutes to download the world soccer web pack. On the "plus" side, a search for Ghana, whiles offline, did not fail to remind me of Ghana's 2-0 loss in the FIFA world cup game against Italy.

For those of us who are unable to meet the system requirements, a poor mans version can be implemented using a browser. Most browsers allow a user to save specific web pages for offline browsing. A more challenging method, would be to write/modify a Firefox extension to provide similar functionality.

Alternatively, the combination of web feeds and news aggregators present interesting tropically tolerant ways to access web content. I will cover this in a later post.

There is an argument to be made that instead of developing technologies to cache and serve data we should focus on expanding broadband availability. My response would be, there is nothing wrong with a multi-tiered approach or a stopgap solution.

In a world where internet connectivity is intermittent, cache-and-browse is the way to go.

Sunday, May 07, 2006

Anti-social bookmarking

I call them bookmarks, others prefer the name favourites, and a minute number of people use the term hotlists. According to wikipedia, "Bookmarks are pointers - primarily to URLs - built into the various internet web browsers".

Bookmarking is a way for an internet user to find his/her way back to a web-page on the world wide web. Without them many-a-page would fall through the cracks in our memory, never to be found again.

Traditionally, whenever an internet user discovered a web-page worthy of keeping, they used their browser to store the URL locally on their computer. Whenever they needed to go back to the page, they went to their browsers and with minimum fuss found their way back. This methodology works great for users in "developed countries" but it is not necessarily a good fit for sub-saharan africa.

Millenium Indicator 48 for Ghana: 38% of the population have computers and 78% of the population use the internet(Source: UN Statistics Division).
Lets ruminate on that for a minute.

Okay, time's up.

I don't know about you, but for me the fact that 40% of internet users in Ghana use public computers means that the traditional method of bookmarking is not tropically tolerant (tt). So, what are we going to do about it ?

Nothing, its already been done for us.

The solution: storing the bookmarks server-side, can currently be found in social bookmarking websites. These services exist mostly for social-networking and folksonomy purposes. However, looking at them from a tropical mindset, social-networking and folksonomy are secondary features. Social bookmarking sites are simply, free, online storage for bookmarks. What this means is that an internet user in Ghana can have his own "private" list of bookmarks that is available whenever he is online without having to own a computer.

Fellow tropical netizens, what we have here is tropically tolerant bookmarking.


Technorati technorati tags: , , ,,