| <!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN" |
| "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd"> |
| |
| <html xmlns="http://www.w3.org/1999/xhtml"> |
| <head> |
| <meta http-equiv="content-type" content="text/html; charset=utf-8" /> |
| |
| <title>FAQ - Recording downloads via MirrorBrain</title> |
| <meta name="keywords" content="Marketing, OpenOffice, OpenOffice.org" /> |
| <link rel="stylesheet" type="text/css" href="marketing.css" /> |
| </head> |
| |
| <body> |
| <div id="MyContent"> |
| <h1>FAQ - Recording downloads via MirrorBrain</h1> |
| |
| <h2>Why do we measure downloads?</h2> |
| |
| <p>One of the <a href="http://marketing.openoffice.org/">Marketing |
| Project's</a> key statistics is the number of downloads from the |
| OpenOffice.org site. This gives us immediate feedback on the success of |
| promotional activities, such as <a href= |
| "http://marketing.openoffice.org/press_releases.html">press releases</a>, |
| as well as trend information over time.</p> |
| |
| <h2>What is MirrorBrain?</h2> |
| |
| <p>When people request a download from the <a href= |
| "http://downloads.openoffice.org/">OpenOffice.org download page</a>, they |
| are directed to <a href="http://mirrorbrain.org/">MirrorBrain</a>. MirrorBrain redirects them to their nearest |
| mirror site for the download they require, and records the request in a |
| log file.</p> |
| |
| <h2>Does MirrorBrain provide the full story?</h2> |
| |
| <p><strong>No!</strong> - it only records downloads started a single |
| point, the <a href="http://downloads.openoffice.org/">OpenOffice.org |
| download page</a>. It omits:</p> |
| |
| <ul> |
| <li>downloads which people make directly from mirrors</li> |
| |
| <li>downloads via other mechanisms, such as <a href= |
| "http://distribution.openoffice.org/p2p/">peer-to-peer |
| networks</a></li> |
| |
| <li>downloads from other third-party repositories - including GNU/Linux |
| distributions (see next question)</li> |
| </ul> |
| |
| <p>Note also that MirrorBrain logs when it successfully redirects someone |
| to a download site - if the user chooses not to download, or cancels the |
| download, then MirrorBrain will not be aware.</p> |
| |
| <h2>Why do MirrorBrain figures look low for GNU/Linux?</h2> |
| |
| <p>The OpenOffice.org Community provides OpenOffice.org software in the |
| two most common GNU/Linux package formats (rpm and deb). However, the |
| majority of GNU/Linux users receive software direct from their |
| distributor's respositories. This is a significant omission and explains |
| the relatively low number of GNU/Linux downloads in our figures. How big |
| is the gap? When we released OpenOffice.org 3.0, Ubuntu - possibly the |
| largest desktop GNU/Linux variant - <a href= |
| "http://www.theregister.co.uk/2008/10/27/shuttleworth_ubuntu_commitment/"> |
| estimated</a> they had around eight million users.</p> |
| |
| <h2>Are MirrorBrain downloads the same as the number of users ?</h2> |
| |
| <p>No. Nothing like it.</p> |
| |
| <ul> |
| <li>MirrorBrain records only a percentage of total downloads (see |
| above)</li> |
| |
| <li>Downloads are only one distribution method - people buy CD-ROMs, or |
| get one from a magazine cover, or borrow one from friends...; people |
| buy PCs with OpenOffice.org pre-installed... etc.</li> |
| |
| <li>One download or CD-ROM can be installed on thousands of computers - |
| or none at all</li> |
| </ul> |
| |
| <p>Note: the <a href="http://distribution.openoffice.org/">Distribution |
| Project</a> exists to make OpenOffice.org available through as many |
| channels as possible.</p> |
| |
| <h2>Do other people provide statistics?</h2> |
| |
| <p>Several of the <a href= |
| "http://projects.openoffice.org/native-lang.html">Native-Language |
| Projects</a> record download information for their languages.</p> |
| </div> |
| </body> |
| </html> |