Debian contributors, a tool to measure and credit a diverse community

Debian contributors is a new software for gathering, measuring, and crediting the active participants in the Debian Project, for any kind of contribution, technical or otherwise.

It is a web application with a client reporting tool supporting a wide range of data mining options. It offers comprehensive identity management which allows to credit a person's work across multiple emails, logins or gpg keys.

Results for several data sources and proof-of-concepts are visible in https://contributors.debian.org/ currently retrieving 16 types of contributions (only one of them was previously tracked). Source code is AGPL licensed, and quickly evolving in the public Git repository of the Debian project.

Besides being a technical tool, it is a social experiment: the intent is to change the way the Debian community perceives itself, by making a diverse range of contributions visible, and showing them with the same level of importance.

More info here: https://wiki.debian.org/DebianContributors/DcDocs

Transcript

1. A quick tour around Debian Contributors
   Laura Arjona Reina
   2014-07-20
   Laura Arjona Reina A quick tour around Debian Contributors 2014-07-20 1 / 14
   

   View Slide

2. What is Debian Contributors?
   A website with a list of people contributing to the Debian
   Project.
   A software infrastructure to gather data from dierent sources.
   A project to make non-packaging contributions more visible in
   Debian.
   A new hat: you get/loose it depending not on your adquired
   reputation. . .
   . . . but on your actual contributions.
   A tool to help the Debian New Member process, and teams.
   Laura Arjona Reina A quick tour around Debian Contributors 2014-07-20 2 / 14
   

   View Slide

3. Quick info
   Website: contributors.debian.org
   Source code: http://anonscm.debian.org/gitweb/?p=nm/dc.git
   Authors: Enrico Zini, New Members Front Desk team.
   Technologies used: Django, JSON.
   More info: https://wiki.debian.org/DebianContributors/DcDocs
   Laura Arjona Reina A quick tour around Debian Contributors 2014-07-20 3 / 14
   

   View Slide

4. Data sources
   JSON structure with array of people IDs, and contributions
   (type, date)
   Dened in contributors.debian.org website
   Data mined and pushed using dc-tool (e.g. with a cron job)
   Laura Arjona Reina A quick tour around Debian Contributors 2014-07-20 4 / 14
   

   View Slide

5. The dc-tool: data mining (I)
   Lots of data mining methods
   Files, gitdirs, svndirs
   Gitlogs, mailfrom, postgres (SQL query)
   Template to build an URL pointing to contribution details
   Some other options (select dirs, blacklist/whitelist addresses. . . )
   The sender only needs to dene data source and mining method
   No coding skills required for many data sources
   Some coding skills required to adapt other sources
   Laura Arjona Reina A quick tour around Debian Contributors 2014-07-20 5 / 14
   

   View Slide

6. Results: contributors.debian.org
   List of contributors for current year, and whole history
   Laura Arjona Reina A quick tour around Debian Contributors 2014-07-20 6 / 14
   

   View Slide

7. Results: Data Sources (I)
   List of teams/data sources
   Laura Arjona Reina A quick tour around Debian Contributors 2014-07-20 7 / 14
   

   View Slide

8. Results: Data Sources (II)
   16 data sources already handled
   Debian System Administration, Security Tracker, BTS
   Bazaar, Subversion and Git committers
   Collaborative Maintenance of packages, Package uploaders
   Mentors and New Members Frondesk
   Blog authors, Video subtitling, Publicity, Spam review
   Wiki and Website editors.
   Laura Arjona Reina A quick tour around Debian Contributors 2014-07-20 8 / 14
   

   View Slide

9. Results: Proposed Data Sources
   8 data sources propsed to implement
   Debconf participation, Debian Perl team, Debian blends
   Translations of packages and website, of manuals and
   documentation, of package descriptions, of debian installer
   Voters
   Some sources need a bit of tweaking, others just adopt them
   using dc-tool
   Laura Arjona Reina A quick tour around Debian Contributors 2014-07-20 9 / 14
   

   View Slide

10. Results: Contributor Dashboard (I)
    The system matches emails, login, PGP keys, UIDs
    Coming feature: identity claim
    Laura Arjona Reina A quick tour around Debian Contributors 2014-07-20 10 / 14
    

    View Slide

11. Results: Contributor Dashboard (II)
    Settings to control which info is shown/hidden
    Laura Arjona Reina A quick tour around Debian Contributors 2014-07-20 11 / 14
    

    View Slide

12. Further work
    Include people not having @debian nor Alioth account
    Identity claim, opt-in system
    Promotion inside the project
    More data sources
    Teams involvement
    Publicity outside the project
    Laura Arjona Reina A quick tour around Debian Contributors 2014-07-20 12 / 14
    

    View Slide

13. Why not using it in other projects?
    Adopt the goal: making diversity skills more visible
    Use the methodology and the code, if you like it
    Any kind of feedback is really appreciated
    Laura Arjona Reina A quick tour around Debian Contributors 2014-07-20 13 / 14
    

    View Slide

14. Take away
    Debian contributors improves the perception of the Debian
    project
    Diverse skills more visible
    More accurate info, automated tools, distributed maintenance
    Website: contributors.debian.org
    Documentation: wiki.debian.org/DebianContributors/DcDocs
    Contact: Laura Arjona, Enrico Zini
    Thanks!
    This work is licensed under CC-By-SA 4.0 International
    Laura Arjona Reina A quick tour around Debian Contributors 2014-07-20 14 / 14
    

    View Slide