Firefox/Projects/StrandedUsers: Difference between revisions

Revision as of 17:50, 28 September 2009

We strand some hundreds of thousands of users on each point release - people who don't come along with subsequent security updates.

We should understand why that is.

We need more data. We need to figure out what data that is.

Are these users significantly different than our overall user base?
- All mac users?
- All europeans?
- All occasional users?
Why do certain versions have half, or twice, the overall "strand rate"?
- Is it a function of release lifetime?
Is the lack of update a deliberate choice or a technological limitation?
- Compare AUS and Blocklist to gauge "intentionality"
How much can be attributed to third party distributors (e.g. Linux distros) that have their own release cycle?

Talk with Metrics
Figure out what telemetry we can send to help understand, without harming user privacy in any way.

@@ Line 8: / Line 8: @@
 We need more data. We need to figure out what data that is.
+=== Open Questions ===
+* Are these users significantly different than our overall user base?
+** All mac users?
+** All europeans?
+** All occasional users?
+* Why do certain versions have half, or twice, the overall "strand rate"?
+** Is it a function of release lifetime?
+* Is the lack of update a deliberate choice or a technological limitation?
+** Compare AUS and Blocklist to gauge "intentionality"
+* How much can be attributed to third party distributors (e.g. Linux distros) that have their own release cycle?
 == Next Steps ==