> recent survey of
> SourceForge projects confirmed it and found that the median was 1.
> According to this dogma, we will rule at least half the project.

Offtopic, but that survey wasn't worth a damn. 
After reading that article, I wanted to do something similar for statistics
project.
So I took a close look at his results. He screened the projects for high
maturity level,
so I did the same. I then sorted the results by activity, and looked at the
first 100 results (smaller sample than his, IIRC).

First thing you notice is that above 80% of the projects there are small web
front ends.
Things that I do in about two days at work. This gives a horrible bias to
his results.
Those projects mature fast, since they take two days to develop, and you
don't need more
than one programmer to do it.

Then you notice that the range is very large. The largest project there is
either Python
or TCL, with 40+ developers. You really can't compare those to web front
ends.

So a sample of mature, active projects contain only two kinds of projects:
Tiny, one 
person, 3 days projects, and huge projects that take 50 people and 8 years. 

IMO those web front ends should have been removed from his sample, 
since they can't be defined as a project at all. This will change the
picture considerably.

If you consider only projects written in C/C++, you see a very different
picture.







----------------------------------------------------------------------------
To unsubscribe, send a message to [EMAIL PROTECTED]
Archives available at http://www.mail-archive.com/[email protected]/

Reply via email to