Tag Archives: spread firefox

The Firefox download map: Remixed

So with the help of my very cool friend Jeremy V. we’ve remixed the Firefox download map.

The official download map shows us the absolute number of downloads per country. From a marketing perspective this was a great idea. However, since writing the post in which I mashed up the Pentagon’s New Map with the Firefox pledge map I’ve been wondering if their is a correlation between the number of pledges/downloads of Firefox and how “connected” a country is.  To find out, we’ve recreated the Mozilla download map but changed the variables. Using population census and internet user stats from Wikipedia.

Current Firefox Download Map
Firefox download map

Remix 1: % of population that downloaded firefox

You need to upgrade your Flash Player

//

Remix 2: % of internet users that downloaded firefox

You need to upgrade your Flash Player

//

Lots to dive into here. Most notable is how India and China’s strong showing in the initial download map is obviously nothing more than a function of their sheer size. In the later two maps they do not fare as well. The fact that other browsers dominate the marketplace in these countries may also play a role. This is certainly the case in China (though I’m less certain about India). Either way though Firefox’s presence in China and India is a lot weaker than the download map suggests. What, if anything, can be done about this is worth debating.

The biggest surprise has to be Iran. It would appear that Iran is anything but the insular isolated country the western media likes to portray it as. Indeed, outside of the western world Iran has one of the highest download rates per capita. This would suggest Iran is quite well connected (and, I suspect, deeply mistrustful of MicroSoft). This should pose a challenge to Barnett’s thesis, in which connectedness should make a country safer to the international system. John Harris who wrote an excellent critique of Barnett’s thesis, would argue that Iran isn’t perceived as a threat because it is disconnected (it clearly is connected). Instead, what makes it a threat is that a ruling elite strives (however unsuccessfully) to keep it as disconnected from the rest of the world as possible. This creates tensions and challenges that exasperate the complex challenge that is Iran.

One of the most interesting impacts of open source software is that it may help countries – especially those transitioning into modern economies – increase their connectiveness. These maps seem to confirm this hypothesis. A scan of the % of the population that downloaded Firefox shows a robust number of people in emerging economies – especially in Latin America, Eastern Europe and to a lesser degree, South East Asia. Indeed, the government’s of many Central and Eastern Europe countries have mandated that their public services use open source solutions where ever possible. With many countries in the region counting download rates at over 2% of the population(!!!!) Eastern and Central Europe really is emerging as a dominant market place for Open Source solutions.

Ultimately, what is notable is that Core countries do by and large well, Gap countries do poorly and countries that sit around much of the perimeter of Barnett’s line separating the “gap from “the core” score in between. For example, the deep pits of disconnectedness are focused around the Stans and Africa, broken up only by Iran and Israel. Consequently, in my opinion the the Firefox download map gives us data that validate Barnett’s thesis. The bigger question is what can open source generally, and Firefox specifically, do to help change the map?

Two notes:
1. Since being slashdotted I’ve received many comments and emails both encouraging and critical. Please keep sending them and adding to the analysis here. Many heads are better than one and so please do contribute
2. I apologize that data for some countries (like Iceland) is not available on this map. Recreating the map wasn’t easy, and it was all in our (and really Jeremy’s) spare time so we did the best we could.

Firefox pledge map – pledges as a % of population

(sorry this post was initially messy – when you get off a transatlantic red-eye and have fifteen minutes to copy and paste in your post, things are bound to get ugly. It should be all cleaned up now)

Mozilla kindly sent me a preliminary data set of download pledges. This data is about 2 days old but should still be more or less instructive. I hope to get more up to date data soon and will update accordingly.

According to the raw data the 20 countries where the most people – as a % of the population – pledged to download Firefox. Population figures are based on those available on Wikipedia.

Country % of the pop that pledged to download
Anguilla 5.78
Falkland Islands 2.00
Montserrat 0.66
Andorra 0.60
Faeroe Islands 0.40
Iceland 0.35
Liechtenstein 0.27
Slovenia 0.26
Greenland 0.25
Estonia 0.25
Poland 0.23
Palau 0.20
Aruba 0.19
Netherlands 0.17
Maldives 0.17
Malta 0.17
Antigua And Barbuda 0.16
Latvia 0.15
Turks And Caicos Islands 0.15
Lithuania 0.15

On the one hand it is no surprise that many of these – about half – are small island countries with populations under 100,000 people. Moreover, 14 have populations under 500,000. Consequently, for many of these countries even a small number of absolute pledges translates into a high total percentage of the population. But I also think this says something about the economics of open source and the internet. It enables small isolated and sometimes neglected parts of the world economy to get effective, top of the line software for free, software that in turn enables them to participate in the global economy.

But things get more interesting if we strip out the countries with populations under 500,000. This picture eliminates the small countries (which are, in some ways, outliers) and allows us to focus on the rest of the world (167 countries made this list). Here are the results for the top and bottom 20 countries.

Top 20 (with population greater than 500,000
Country Core or Gap % of the pop that pledged to download
Slovenia Gap 0.263
Estonia Core 0.248
Poland Core 0.235
Netherlands Core 0.170
Latvia Core 0.154
Lithuania Core 0.146
Finland Core 0.144
Norway Core 0.143
Belgium Core 0.139
Portugal Core 0.131
Albania Gap 0.130
Denmark Core 0.126
Singapore Gap 0.125
Croatia Gap 0.124
Hungary Core 0.123
Canada Core 0.110
Chile Core 0.107
New Zealand Core 0.109
France Core 0.106
Australia Core 0.105
Bottom 20 (with population greater than 500,000
Country Core or Gap % of the pop that pledged to download
Ghana Gap 0.001
Burundi Gap 0.001
Kenya Gap 0.0006
Chad Gap 0.0006
Uganda Gap 0.0005
Burkina Faso Gap 0.0005
Mali Gap 0.0005
Rwanda Gap 0.0005
Cameroon Gap 0.0005
Guinea Gap 0.0004
Mozambique Gap 0.0004
Bangladesh Gap 0.0004
Malawi Gap 0.0004
Niger Gap 0.0004
Myanmar Gap 0.0003
Sudan Gap 0.0003
Nigeria Gap 0.0002
Tanzania Gap 0.0002
Dem. Republic Of The Congo Gap 0.0002
Ethiopia Gap 0.0001

I admit that when I wrote the post on Monday about the correlation between the pentagon’s new map and the firefox pledge download map I thought that once the per capita data was analyzed it would sharply change the outcome. The reality is, it doesn’t. Core countries are far and away dominant on the list. In the bottom half of the list (84 of the 167 countries with populations over 500,000) only 4 countries are in the core: India, China, Mongolia and South Africa. (of course as a % of Function Core, or even the worlds’ population, this is a lot of people!).

Eastern Europe is clearly an emerging open source powerhouse. Of the top 20 countries as a percentage pf population who pledged the top 3 are Eastern Europe and a total of 8 make the list. Only 4 of the countries are “non-integrated gap” countries all of which are transitioning (or arguably have transitioned, into “New Core” countries. Indeed, there is an argument that open source software allows new core countries to integrate into the core more rapidly by not only making some of the key tools that facilitate this transition more readily and cheaply, available but also by enabling the population to participate in their development thus building world class skills without the requisite FDI or multinational corporate investment.

The more grim news is at the bottom of the list. Perhaps unsurprisingly, but still another sad reminder, virtually every country on the bottom 20 is African (Bangladesh and Myanmar are the exceptions). In short, the countries most in need of this software, software that is freely available, still are least likely to have the capacity and infrastructure to download it.

Other notable placements were Venezuela (62) and Iran (77), much lower down the list than I initially suspected they would be.

Also interesting, and perhaps a possible challenge for Barnett (and the world) is that the 3 Core countries with fewest number of pledges were (in order from fewest to most) China (123), India (116) and South Africa (89)

Also, just in case you were wondering, I believe the distribution of the countries rank by percentage of population makes a power law graph :)

Powerlaw pledge graph

 

Shocker.

Firefox 3 pledge map vs. the Pentagon’s new map

What are the geopolitics of open source? To find out I thought it would be interested to see how Thomas Barnett’s map meshed with the Spread Mozilla Firefox 3 download pledge map.

Some brief background for those not familiar with “The Pentagon’s New Map.” It is a map that sits at the heart of a book of the same title written a few years ago by Barnett. It is a compelling take on what America’s grand strategy should be for the 21st century and how it is, and more importantly isn’t, ready to execute on it. Better yet, it is engaging, thought provoking, interesting, and written so anyone can read and understand it.

The core of the book’s thesis (remixed from Wikipedia and very high level) is as follows:

  1. International systems of rules reduce the likelihood of violent conflict (e.g., the WTO Dispute Settlement Understanding)
  2. The world is divided between the Functioning Core and the Non-Integrated Gap.
    Function Core = economic interdependence, incented to abide by rules
    Non-Integrated Gap = unstable leadership and absence of international trade, weaker incentives
  3. Integration of the Gap into the global economy provides opportunities for individuals to improve their lives, presenting a desirable alternative to violence and terrorism
  4. US grand strategy for the 21st century… help countries migrate from the Non-Integrated Gap into the Functioning Core

According to Barnett’s thesis, countries in the Non-integrated Gap, because they are less connected, should probably have fewer computer users, fewer people downloading software and fewer people participating in Open-Source projects. Mashing up his map with the Firefox pledge map might give us some a clue to how well open-source conforms to his thesis.

(Note, I’ve remixed Barnett’s map to make it is easier to read on a computer with the Function Core countries in green and the Non-Integrated Gap countries in red . You can find the original map here.

PNM remixed

Below is the spread Firefox pledge map, which tracks how many people around the world have pledged to download Firefox on its release day (June 17th). I’ve overlaid the Non-Integrated Gap/Function Core border over it.

firefox PNM mash up

Some comments/thoughts:

  • Interesting correlation between low pledge totals and Non-Integrated Gap countries
  • All but two Non-Integrated Gap countries (Colombia & Turkey) have 10,000 download pledges or fewer. (I also think it is interesting that Barnett doesn’t include Turkey in the Functioning Core…)
  • Most countries within the Functioning Core have 10,000 pledges or greater (South Africa, Nordic Countries and the Baltic States are notable exceptions)
  • Non-Integrated Gap countries with the most pledges are Iran, Turkey, Venezuela, Peru, and Indonesia – interesting list. Seems to suggest that many of the countries the US tries to isolate are actually the most connected.
  • According to my Mozilla friends Poland (yes, Poland) was the first to hit the 100K pledge mark. Many new Core countries are adopting Open Source en mass to avoid paying for expensive Microsoft software. Open source may be offering them a cheap way to increase connectivity and integrate with the core faster, and on their terms. Fantastic outcome.
  • This map DOES NOT account for population variation – would be fascinating to see a map based on per capita pledges (I’ve contacted my friends at Mozilla and they’ve passed the raw data along to me so I will follow up with that analysis ASAP)
  • I will try to update the map with the final data on download day (June 17th) when all the pledges have been tallied
  • Note: Firefox pledge map copied on June 15th, 2008, 8:30 pm PST

Lots more thoughts and analysis to be done on this. I hope to blog more on this shortly. If Barnett responds in any way I promise to update – would love to hear his thoughts/reflections on this.

(One final aside, if you get the chance to see Barnett present, do so. He’s up there with Lessig in his delivery. I remember seeing him at a conference. He went long by 10 minutes. The US ambassador to Canada was in the next room waiting to give the next presentation but if any of the organizers had tried to intervene and hurry Barnett up, they would have been lynched. FYI, You can see his TED talk here.)