popularity contest data for economic theory testing
Nagy Viktor - Akasha
viktor.nagy at gmail.com
Wed Sep 9 08:58:26 UTC 2009
Dear Ubuntu Devs/Webmasters,
I would like to know whether it is possible to receive the popularity
contest data (without the MY_HOSTID identifier) from the very past until
The reason why I ask is to run an empirical study on open source
development, and software development in general. In economics, a special
market is a two-sided markets. It is characterised by three players, 2 user
groups and a platform. The distinguishing feature of these markets compared
to other markets is that members of let's say group 1 are more likely to
join the platform if more people of group 2 join it, and vica versa.
Examples include dating clubs, where men and women would like to meet, or
programming projects where developers are more likely to develop for the
iPhone because of its large user base, then for Android, and this assures
that iPhone has nicer apps, and thus more users. Of course, this can be said
about other IT projects as well.
Even though software development is one of the most noted examples in
theoretical articles, no study have looked at the effect empirically.
The idea about popularity contest data is to extend that dataset (probably
for a limited set of projects) with mailing-list, Trac and version control
log files in order to have measures for development intensity, and test how
strong the two-sided effect might be in open source software development.
Moreover, such a study might come up with insights on the critical times in
a project's life, that might help Canonical and other open source software
companies to better organise their resources.
I hope you will find the above idea interesting, and will help me by
providing these data.
Thanks for your time and effort,
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the Ubuntu-devel-discuss