Hey,
I've implemented server side clustering (using quadtree and distance based approach), instead of creating this process in-memory i do it via batch.
Why do i need client-side clustering , you ask...good question:
real-time changes - because i don't do the actions on the server's memory, i must deal with delete / insert / update of markers. That means to update existing clusters in different zoom levels.
Solution:
Using the existing cluster from the server via client-side clustering, a.k.a, markercluster plus (for v3).
So after this intro, how can i tell markerclusterer plus:
"here are my existing clusters: a,b,c,d,e..."
and then
"take these new markers (real-time deltas from the batch creation till now) and cluster them"
and
"remove these markers from clusters.... " (will decrease the cluster no of markers, and other parameters that aren't unique)
is this possible? your thoughts on it, anything is welcome
tnx