A short introduction (10-15min?) to https://wikimedia.biterg.io/ for anyone interested in statistics about our technical community who use our technical infrastructure (Wikimedia Git; Gerrit; Phabricator; IRC; Mediawiki.org).
Followed by questions, (hopefully) answers, discussion.
This is a rather spontaneous session and I'm probably only going to showcase some very basics.
I'm looking forward to comments, feedback, and some questions.
Upstream project: https://grimoirelab.github.io/
Upstream repositories and bug reports: https://github.com/chaoss/
Downstream documentation: https://www.mediawiki.org/wiki/Community_metrics
Downstream bugs and feature requests: https://phabricator.wikimedia.org/tag/analytics-tech-community-metrics/?order=priority
List of database indices (to use in the advanced search field on the website): https://wikimedia.biterg.io/app/kibana#/discover
Notes/script used by @Aklapper during the session:
- Goals of this session:
- 1. Create awareness that this exists
- 2. Show some basics what you can do (from a user point of view, not admin stuff); about 15 minutes
- 3. Get feedback, is this useful or not, etc
- 4. This session will not answer your developer questions about Kibana or Elasticsearch code.
- Grimoirelab platform/suite, under the CHAOSS umbrella of the Linux Foundation, written in Python and all FOSS, ran by Bitergia (company spin-off from a research group at the University of Madrid), also providing services for Mozilla and other organizations
- Gathering activity data from different sources (see the left pane of https://wikimedia.biterg.io ) like Git, Gerrit, Phabricator, IRC, Mailing Lists, mediawiki.org -> indexing -> Database. Does not mean all and any data from those sources is indexed, though (e.g. not GitHub, see T186736).
- UI: if you are used to Kibana, no news for you:
- Overview, click items like names, repositories, organizations -> filtered view applied. Enable/disable them below text field.
- Timeframe in upper corner.
- Exporting short URLs possible via "Share"
- Textfield for advanced search but need to know exact index names; see "Discover" link in docs. Standard Lucene search syntax, example: `author_bot:false AND author_name:"Foo"
- Questions and answers; discussion