Several days after my article was published, one or more
anonymous posters put up vague messages suggesting criticism
of one part of the FLOSS survey. The effect of the messages
was not to explain any disagreements or contribute a better
understanding of the survey, but just to besmirch the image
of the survey, and by extension my article.
I have talked to some participants in the debate in order to
clear the air, and have established the following key
1. Out of the six parts of the survey, complaints were
lodged concerning one part: the developer survey.
2. A few people criticized the survey because it did not use
the kind of scientifically representative sample that
depends on identifying the boundaries of the population,
etc. An explanation by the authors shows why such a
sample could not be taken.
3. No one has suggested how the survey could have been
conducted differently. The critics say that the results should
be taken as only indications rather than statistically
meaningful results. The authors of the survey believe
that the kind of sample they took is a random sample that
is valid for drawing statistically strong conclusions. See the URL above for details.