The Fediobserver has daily and monthly stats for active users. I assume daily indeed follow daily variations. You also saw those huge bumps when LW updated and started counting votes as activity
I think the observer shows daily and monthly stats for the active users per month and active users per half year so the active users per month wouldn’t change as fast I think.
Also about it being a botfarm I do think that is a possibility. Actually there is more evidence for it when you see extend the graph to 120 days and see a huge uptick in users and servers at the same time. Edit: 2024-7-29
I think the observer shows daily and monthly stats for the active users per month and active users per half year so the active users per month wouldn’t change as fast I think.
Indeed, I was remembering it wrong, you might be right!
Also about it being a botfarm I do think that is a possibility. Actually there is much clearer evidence for it when you see extend the graph to 120 days and see a huge uptick in users and servers at the same time. Edit: 2024-7-29
would this not happen if a whole instance got shut down?
Either that or blocked the crawler on their Instance.
I have a feeling it was caused by something like that, likely a failure of some kind. The drop is way too steep and abruptly for it to be natural.
An active instance would have an impact on active users
Not immediately though right? since the active users are a month or half-year. Or does it automatically update that too?
Anti Commercial-AI license (CC BY-NC-SA 4.0)
Active users are daily too (you have day, week and month counters in the sidebar of all communities)
I was talking about on the fediverse observer. It wouldn’t show up immediately there.
Anti Commercial-AI license (CC BY-NC-SA 4.0)
The Fediobserver has daily and monthly stats for active users. I assume daily indeed follow daily variations. You also saw those huge bumps when LW updated and started counting votes as activity
I think the observer shows daily and monthly stats for the active users per month and active users per half year so the active users per month wouldn’t change as fast I think.
Also about it being a botfarm I do think that is a possibility. Actually there is more evidence for it when you see extend the graph to 120 days and see a huge uptick in users and servers at the same time. Edit: 2024-7-29
Edit: wording
Anti Commercial-AI license (CC BY-NC-SA 4.0)
Indeed, I was remembering it wrong, you might be right!
Definitely
The source code is open source so I believe someone can tell us if need be.