Usage Statistics: Difference between revisions

Latest revision as of 01:55, 21 April 2018

There are a lot of Wine users and this page attempts to estimate how many there are. It isn't just an academic question because who our users are should shape our priorities. How many are gamers? How many use Wine for a single application? How many care more about improving performance on apps that work already, and how many want more apps to work?

Now that we directly package Wine Downloads for several popular distributions, we can count statistics ourselves (not sure if we do though), but even then, there are many ways to estimate Wine's popularity.

Popcons

Several distributions offer popularity contest (popcon) packages, which report package install statistics for users who have checked the box to "send anonymous package usage data." Besides the raw figures for users with Wine installed, proportions can be estimated by taking a ratio of Wine installations to the installations of a core package for the distro. However, we suspect that popcons miscount actual Wine users for various reasons, such as self-selection bias (are popcon users more or less likely to also use Wine?)

Wine popcon pages for various distributions:

Surveys

Maybe we could setup our own survey through SurveyMonkey to get some new data?

At the end of 2007, DesktopLinux.com did a survey about the Linux market. In their results, 31.5% of 38,500 reported using Wine to run Windows applications.

Every Linux Unix Computer User is a Potential Wine User

Work in progress: This page is currently undergoing extensive revision. External links to specific parts of this page may now be broken due to the section having been edited, moved, merged with another section, or removed altogether. Consult the table of contents to find the section you are looking for. There may be related discussion on the talk page.

* Work backwards from estimates of global computer and OS users (gstatcounter, Garnett, et. al)

Website Traffic as a Proxy Statistic

Work in progress: This page is currently undergoing extensive revision. External links to specific parts of this page may now be broken due to the section having been edited, moved, merged with another section, or removed altogether. Consult the table of contents to find the section you are looking for. There may be related discussion on the talk page.

While it introduces a layer of indirection, we can treat things like web traffic to WineHQ as a rough estimate of overall interest in the project.

* Use Google Trends

@@ Line 1: / Line 1: @@
-There are a lot of Wine users and this page attempts to estimate how many there are, as well as suggest the growing importance of Wine to developers.
+There are a lot of Wine users and this page attempts to estimate how many there are. It isn't just an academic question because who our users are should shape our priorities. How many are gamers?  How many use Wine for a single application?  How many care more about improving performance on apps that work already, and how many want more apps to work?
-We really do need to start thinking about just who our users are.  How many are gamers?  How many use Wine for a single application?  How many care more about improving performance than getting more apps working? Such questions may help us prioritize development.
+Now that we directly package Wine [[Download]]s for several popular distributions, we can count statistics ourselves (not sure if we do though), but even then, there are many ways to estimate Wine's popularity.
-== Why statistics are hard to get ==
+== Popcons ==
-Many, if not most, people don't get Wine by downloading it from one of our file mirrors: dl.winehq.org, ibiblio, and budgetdedicated.com.  Instead, they use the (old) Wine package that comes default with their distribution.  It is unknown what percentage of Wine users are using an older version of Wine, however it is likely greater than 90%.
+Several distributions offer popularity contest (popcon) packages, which report package install statistics for users who have checked the box to "send anonymous package usage data." Besides the raw figures for users with Wine installed, proportions can be estimated by taking a ratio of Wine installations to the installations of a core package for the distro. However, we suspect that popcons miscount actual Wine users for various reasons, such as self-selection bias (are popcon users more or less likely to also use Wine?)
-== Ubuntu Popcon ==
+Wine popcon pages for various distributions:
-The Ubuntu popularity contest (popcon) package reports package install statistics for users who have checked the box to "send anonymous package usage data."  http://popcon.ubuntu.com/  As of September 2009, 495283 participants had installed Wine and 32660 had run it within the past week.  Comparing this with a package that everyone has and you find there's a total of 1226340 survey participants.  This means that Wine is installed on an estimated 40% of Ubuntu PCs, although only about 6.7% of them have used it within the past week.
+* [https://popcon.ubuntu.com/by_inst Ubuntu popcon results (sorted by install count)]
+* [https://qa.debian.org/popcon.php?package=wine Debian popcon results for Wine-based packages]
-There is reason to suspect that Popcon substantially undercounts Wine users, however, because survey participation is not offered to users who install Ubuntu directly from Windows using Wubi.  Since Windows users are more likely to need Windows applications and thus Wine, this will result in systemic bias against Wine.
+== Surveys ==
+{{Info}} Maybe we could setup our own survey through [https://www.surveymonkey.com/ SurveyMonkey] to get some new data?
-== Total Wine users ==
-*Estimated total users of Wine (all versions of packages, across all distributions): unknown (I've heard ~1.5 million)
-*Estimated total users of Wine, Crossover, or other Wine-containing products (ie Google Picasa): unknown, but much larger
-*Estimated total users of anything containing Wine source code (say, Parallels): unknown, but much much larger
+* At the end of 2007, DesktopLinux.com did a survey about the Linux market. In their results, 31.5% of 38,500 reported using Wine to run Windows applications.
-== Surveys ==
+== Every <s>Linux</s> <s>Unix</s> Computer User is a Potential Wine User ==
-End of 2007 [http://www.desktoplinux.com DesktopLinux.com] did a survey about the [http://www.desktoplinux.com/cgi-bin/survey/survey.cgi?view=archive&id=0813200712407 Linux Market].
+{{Wip}}
-.5% of 38,500 reported using Wine to run Windows applications.
-== Every Linux user is a potential Wine user ==
+  * Work backwards from estimates of global computer and OS users (gstatcounter, Garnett, et. al)
-Around October 2007, there were an estimated 6 to 12 million Ubuntu users.  By May 2010, Fedora claimed 24 million users based on update server statistics, which are unknown for Ubuntu. However if Ubuntu is believed to be twice as popular as Fedora, then there could be an estimate of around 60 million+ combined Ubuntu/Fedora users.  Most of these are desktop users.  When other distributions are added in, there are even more potential users.
-If we combine the 60 million+ usage estimate with a somewhat conservative 5% wine-usage estimate based on Ubuntu's Popcon above (which may or may not be valid for Fedora), then we arrive at approximately 3 million active Wine users.
+== Website Traffic as a Proxy Statistic ==
+{{Wip}}
-== Future Growth ==
+* While it introduces a layer of indirection, we can treat things like [https://www.winehq.org/webalizer/index.html web traffic to WineHQ] as a rough estimate of overall interest in the project.
-Newer distributions are making installation of Wine even easier, such as prompting to install Wine as soon as the user attempts to open an executable file (see UserExperience for more).
+ * Use Google Trends
-Desktop Linux adoption is also growing.  Desktop Linux usage doubled between 2006 and 2007.  Based on web browsing data, Desktop Linux is currently more popular than iPhone OS and other notable platforms.
 ----
-[[Category:About]]
+[[Category:About]] [[Category:Data]]