Following https://tarte.nuage-libre.fr/c/fediverse/p/194717/we-need-more-users I decided to explore data a little bit more. I'm not the biggest fan of growth-as-as-target so I wanted to see how much the people were participating in the discussion.

The data

I took the data from the API explorer in https://api.fediverse.observer/ with this query:

query {  
  monthlystats {  
    date_checked  
    softwarename  
    total_posts  
    total_users  
    total_comments  
  }  
}

Then parsed the json with this https://jqlang.org/ filter:

jq '.data.monthlystats | map(select(.total_users > 0 and (.softwarename == "lemmy" or .softwarename == "mbin" or .softwarename == "kbin" or .softwarename == "piefed"))) | group_by(.date_checked) | map( {date_checked: .[0].date_checked, total_users: ([.[] | .total_users] | add), total_posts: ([.[] | .total_posts] | add), total_comments: ([.[] | .total_comments] | add)}) | map({date_checked, posts: .total_posts/.total_users, comments: .total_comments/.total_users}) | sort_by(.date_checked) | map([.date_checked, (.posts | tostring), (.comments | tostring)]) | .[] | @csv'

(As you see I filtered for the threadiverse. I also did the same with all software, I'll put the graph for that in comments)

Then did a good old' chart

What to think of it

I don't know. Users' activity is on the rise and I find it nice

you are viewing a single comment's thread
view the rest of the comments

[–] rako@tarte.nuage-libre.fr 2 points 10 hours ago

Yeah if you want to do the same community it's going to be harder, but if you want to make your own community with your own content and views it's different.

Also, the history of the internet contradicts your point, communities have moved servers since the beginning, there never was a unique central point for everything. Lemmy is a bit inferior here because it only allows you to see communities one by one, but piefed can group communities into feeds that you can directly follow. By not placrng focus on a single one piefed can push for much more diversity