In the recent CommOps meeting, we discussed a few updates around some initatives to make available tooling for community members to interact with Fedora’s data. To understand the health of our community, we want to leverage data from the Fedora Message Bus to look in aggregate into how people are interacting with Fedora.
In our meeting yesterday, we’ve hit a question that I want to open up to the community. What is our community’s thoughts on usage of our data from the Message Bus and other Fedora systems for analytics, and what is things we would like to not do with it?
Our current agreed norm as a team has been to stay away from looking at induviduals and looking at the whole when doing our work. The best way I could codify the values I’ve said during the last year or so would be:
- Make data avaialble equitably, meaning that you don’t have to download an entire copy of the Message Bus to do something.
- Focus on the big picture, not the induvidual. We don’t want to be looking at what specific people are doing, but understand what the bigger picture of our community is.
- Move from insight to action by focusing on storytelling. We don’t want to obsess over data and not do something about it.
To better help you understand the type of data we’re looking at, here are some example items of things I look at in the Message Bus today:
- Datagrepper (my update today to my FAS Account)
- Datagrepper (a GitHub webhook)
- Datagrepper (a build being signed)
An example we could use to guide this conversation is to examine is the CHAOSS Community’s Data Policy which outlines how data in their project will be leveraged. Down the road and based on your feedback, we could draft and make a recommendation to the Project of something similar to codify this as well.
And a final thought for you to consider is feedback about AI Models and their usage in our community might be better discussed under this existing Council ticket.