You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
We'd like to start tracking aggregates and statistics for Bridgy Fed usage. We do a bit of that now, eg total users, broken down per network, current activity rate, etc. But we'd like more, notably historically over time, eg MAU/DAU, activity rate, etc. Sadly our data volume is already high enough - our current datastore is ~1T - that we don't keep all the raw data we'd need to calculate those kinds of things, either historically or going forward.
These kinds of stats would be useful for a number of things, notably conversations with potential funders.
I don't expect to store all of BF's raw data long term, in any form, but I can estimate some stats historically, and more importantly I can start tracking them going forward. So, what do we want? Assume these are both totals and broken down by source and destination network:
active users per day, month, ...?
counts of activities/interactions across the bridge, broken down by type
account migrations in, out, and across (eventually)
custom usernames/handles
remote "instances" seen, interacted with
explicitly opted out users, instances?
total domains seen/interacted with...?
total interactions seen/handled, whether or not they get bridged?
costs, both total and per user, per active user, per interaction
We'd like to start tracking aggregates and statistics for Bridgy Fed usage. We do a bit of that now, eg total users, broken down per network, current activity rate, etc. But we'd like more, notably historically over time, eg MAU/DAU, activity rate, etc. Sadly our data volume is already high enough - our current datastore is ~1T - that we don't keep all the raw data we'd need to calculate those kinds of things, either historically or going forward.
These kinds of stats would be useful for a number of things, notably conversations with potential funders.
(As a counterexample, Bridgy classic is small enough that I have kept all the historical raw data, mostly in cold storage, and I've regularly crunched and reported statistics for a long time: https://snarfed.org/2023-06-02_bridgy-stats-update-8 , https://snarfed.org/?s=bridgy%20stats . It's fun! Mostly 😁)
I don't expect to store all of BF's raw data long term, in any form, but I can estimate some stats historically, and more importantly I can start tracking them going forward. So, what do we want? Assume these are both totals and broken down by source and destination network:
cc @anujahooja
The text was updated successfully, but these errors were encountered: