From StatsJam

[edit] Data Analyses

At present, we have the following data analyses available. Feel free to add to them, comment on them, point out bugs and flaws in reasoning (yes, queries can most definitely have bugs). This is a community site for everyone to do data analysis (see Introduction below).

Note! Many of these analyses take significant amounts of time (some may take a minute or two). See Notes on Performance below.

• Basic ingimp Statistics
• Defining Significant Users
• Describing Sessions
• Command Counts in Sessions
• Per Command Statistics
• Describing Images
• Activity Tags
• Notes on Data Analysis

[edit] Introduction

Welcome to ingimp's Stats Jam, an open forum for analyzing and discussing data collected from the ingimp project. ingimp and Stats Jam are projects led by Michael Terry at the University of Waterloo.

[edit] What Are ingimp and Stats Jam? A Quick Intro

ingimp and Stats Jam are described in more detail below. Briefly, ingimp collects data about how GIMP is used in the real world, while Stats Jam is a MediaWiki extension that allows one to embed SQL queries within wiki text. Stats Jam allows one to embed live queries against the database within the wiki page itself. On this site, Stats Jam provides the front-end to the ingimp data set. See Background for more information on these projects and their goals.

This project represents the first effort in openly collecting and analyzing data describing how open source software is used in practice. The need for speculation and gut instinct is lessened -- now one can analyze the data of hundreds of users to paint a picture of how the software is used day-to-day, by real people around the world.

[edit] Useful Resources

• Background -- More information on ingimp and Stats Jam
• General wiki help: Help:Editing
• Stats Jam SQL help
• ingimp DB schema
• ingimp DB views
• Notes on Data Analysis
• Sandbox

[edit] Getting Started

• Before editing any page, we recommend you create an account
• If you're new to Stats Jam, but are familiar with wikis:
  • See Stats Jam SQL help
• If you need a quick introduction to wikis:
  • Help:Editing
• You can also try things out in the Sandbox.
• To be able to write queries, you need to understand the DB schema. See:
  • ingimp DB schema for a list of the DB's tables
  • ingimp DB views for a set of views

[edit] Notes on Style in this Wiki

Stats Jam's ability to embed live DB queries in wiki text opens up a host of new possibilities for public, collaborative analysis and discussion of data. We don't quite know what the proper conventions are for having these discussions -- we expect conventions will arise naturally through use of the site.

That said, we have a page, Notes on Data Analysis, to host discussion about ways to perform the data analysis for this data set, and in this medium.

[edit] Notes on Security

You might be saying to yourself, "Wait, anyone can embed executable SQL queries within wiki text? What about security?!?!". We've thought about that. We have very strict permissions set on the database. Furthermore, we have an SQL validator which validates SQL before executing it. But there could still be security issues we are not aware of. If you feel you absolutely must test out the security to find weaknesses, please do so in the Sandbox. Then tell us when you find a suspected hole.

[edit] Notes on Performance

For security and usability reasons, the response time of Stats Jam will be very slow at times. Like, you may want to get some dinner while some pages load. As use of this site increases, these performance problems will, of course, get worse. We know these problems exist and will fix them as time permits. Of course, you are free to hack the code to help us improve performance. See http://www.statsjam.org for pointers to the source code.