Apr 092013
 

I’ve been keeping an eye on the stats of the distribution of my Xen and Kernel-Xen packages now for around 12 months. I find it interesting that they just keep getting more and more popular. First, some stats:

Month
Est New Installs
November 2012
1,213
December 2012
1,256
Janurary 2013
1,292
February 2013
1,152
March 2013
1,318

How did I get these stats? Thats a little more complex. This is the count of requests for the file kernel-xen-release-6-4.noarch.rpm from a single mirror. This package is only installed once – and not redownloaded after initial installation. The mirror is referenced in most of the install guides that are found around the internet. What it doesn’t take into account is the other five mirror sites, local copies etc etc. I think this figure may well be accurate, however there is a good possibility that it is greatly underestimating the real number of servers out there referencing my repos.

Package
Number of releases
kernel-xen
45
xen
10 (v4.2), 2 (v4.1)

This consists of approx 22 security vulnerabilities of Xen across the now non-maintained v4.1 branch as well as the current 4.2 branch. I have not counted kernel vulnerabilities as these are fixed normally within the standard release cycle – and not as seperate patches.

So how does this turn into finances?

Well, I’ve asked for donations for a while to help offset the costs in producing and developing these packages. Since starting in 2011, I have received exactly $80AUD in donations. This means that given the number of hours working on these packages, I effectively work for about $0.12c per hour. This also excludes any expenses in hosting, equipment or server expenses.

How does this affect the projects?

This is easy. If a server dies, I can’t replace it. If I suffer a hardware failure, I can’t replace it. If hosting costs increase, I can’t pay them.

So – this makes me wonder… How do people survive doing open source development? Sure there are companies like Citrix, RedHat etc that employ people to contribute to FOSS – however what about everyone else? Comments welcome below!

  One Response to “The economics of Open Source development”

  1. I get the feeling that business contributes to open source where licensing is permissive (eg: BSD, Apache, etc). As for projects licensed under the GPL, can anyone cite one that is in heavy use in production environments (think Apache or PostgreSQLSo) aside from the Linux kernel? Unless the license allows such use, such projects never gain critical mass.

    I’d even go as far as saying that copy-left software projects are plagued by newbie evangelist programmers who make changes with reckless abandon, giving rise to streaming piles of “version 3 will be a complete rewrite, coming soon!” abandonware.

    So, back to the economics. Some people are of the opinion that “you get what you pay for”. Those people will always pay for something, and rightly they will enjoy better service in return. Let them pay for service, and make sure it’s top notch – that’s how you make money in open source. Release your software using a permissive license and the companies who can afford your service fees who find your product and see value in it will bolster your downloads, and hopefully fill your pockets. 😉

 Leave a Reply

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>

(required)

(required)