The poll came as a response to a similar question in two surveys: the Eclipse Community Survey and one from Dr. Dobbs Developer Technographics. According to this article on CNet the question in those surveys was about the deployment operating system and our question was about the development operating system but interestingly many people added comments in which they make the distinctions.
| Linux Ubuntu | 214 | 21% |
| Linux Debian | 162 | 16% |
| Mac OSX | 160 | 16% |
| Windows XP | 72 | 7% |
| Linux RHEL | 53 | 5% |
| Linux Fedora | 49 | 5% |
| Linux - any distro | 37 | 4% |
| FreeBSD | 36 | 4% |
| Linux Centos | 33 | 3% |
| Windows 7 | 32 | 3% |
| Solaris | 29 | 3% |
| Linux SUSE | 28 | 3% |
| Linux - Other | 23 | 2% |
| Linux Gentoo | 23 | 2% |
| Other answer... | 18 | 2% |
| Windows Vista | 15 | 1% |
| Linux Slackware | 13 | 1% |
| OpenBSD | 6 | 1% |
| Windows 2003 | 6 | 1% |
| HP-UX | 3 | 0% |
| NetBSD | 3 | 0% |
| Windows - Other | 3 | 0% |
| Linux Mandriva | 3 | 0% |
| AIX | 3 | 0% |
The poll ran between 15-22 April 2010 and apparently it attracted exactly 1024 responses during that time. It was promoted via Blogs.Perl.org, in the Perl Mongers group on LinkedIN the Twitter account of Gabor Szabo, on the IRC channel of Padre and on some of the Perl Monger mailing lists.
This time only a single answer was allowed for each person which means the number of answers is also the number of people who answered. While we cannot be sure it seems the number of respondents is comparable to the previous two polls.
The potential answers were listed relatively well and I managed to add the missing values relatively quickly so the "Other" field was used only 18 times (1.76%). On the other hand the idea to give "Windows - Other" (3 votes) and "Linux - Other" (23 votes) was probably not a good idea as we don't really know where those answers belong. The "Linux - any distro" got quite a lot of reponses (37 votes 3.61%) after it was added but that only shows the (lack of) preference of the developer and not so much what actually is being used.
On both Blogs.Perl.org, and on the Perl Mongers group on LinkedIN many comments were received. In many cases people indicated a difference between development and deployment operating system and a difference between the OS used at work and at home. Further surveys might want to make that distinction. There was also a comment indicating that some people run their development tools on one machine while they are actully running the code on another machine. Indeed many IDEs support a way to open a remote files via ssh or ftp.
Raw data
answers.csv and other.csv
Previous polls
We ran two earlier polls, one about Which editor(s) or IDE(s) are you using for Perl development? and the other one about What other technologies, languages, templating systems are you using besides Perl?.