Skip to main content
Jump to: navigation, search

UDC/Filtered Usage Data

Revision as of 00:58, 20 February 2011 by (Talk | contribs) (Example)

Filtered usage data is made available for general consumption in CSV format. You can obtain the data from the archive server.


The data is provided in multiple files. Each file represents the data gathered in a specific year and month. For example, the file filtered_usagedata_200902.csv.gz (1,231MB), contains the data collected in February of 2009. Note that each file contains the information that was uploaded in the indicated period; that data may have been generated in some previous timeframe.

The CSV file is generated with the following implied header:

userId, what, kind, bundleId, bundleVersion, description, time

These fields represent:

  • userId - A generated ID for a specific workspace. This value is generated by our database using information provided during the upload. You can use this ID to connect records that come from a common source.
  • what - What happened (i.e. the event), e.g. "activated", "started", "executed", ...
  • kind - The kind of thing that caused the event (i.e. the source), e.g. "view", "editor", "command", ...
  • bundleId - The name of the bundle that "owns" the thing that caused the event, e.g. "org.eclipse.ui.editors"
  • bundleVersion - The version of the bundle, e.g. "3.6.0.v20100520-0800"
  • description - Event-specific information, e.g. "org.eclipse.pde.ui.PDEPerspective"
  • time - Time of the event (millisecond clock value from the workstation)

More detailed description of the data that is captured can found on the What Gets Captured page.



The first line tells us that, in the workstation assigned ID 336262, the perspective with "org.eclipse.jst.j2ee.J2EEPerspective" from the bundle ID "org.eclipse.jst.j2ee.ui" (undetermined version) was activated at 2008-10-16 12:59:38 (1224176378264).

Back to the top