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
Currently, determining the appropriate view distance is done by incrementing/decrementing the current view distance by 1.
Using a search algorithm similar to a binary search would be very helpful, as it would have a time complexity of O(log n) instead of the current O(n), which can be quite noticeable when starting up a server.
With a minimum view distance of 4 and a maximum of 24, it took around 7 minutes to reach the full render distance of 24, when the MSPT was at 3.85 when the render distance was set to 5, and when increased to 24, the mspt was only 4.05.
Had it done a binary search, it could have much more efficiently determined the point at which the mspt approaches 50 (the limit I have set), and then from there gone up incrementally, as the tps can fluctuate a bit naturally (so a perfect binary search wouldn't work)
The text was updated successfully, but these errors were encountered:
Currently, determining the appropriate view distance is done by incrementing/decrementing the current view distance by 1.
Using a search algorithm similar to a binary search would be very helpful, as it would have a time complexity of
O(log n)
instead of the currentO(n)
, which can be quite noticeable when starting up a server.With a minimum view distance of 4 and a maximum of 24, it took around 7 minutes to reach the full render distance of 24, when the MSPT was at
3.85
when the render distance was set to 5, and when increased to 24, the mspt was only4.05
.Had it done a binary search, it could have much more efficiently determined the point at which the mspt approaches 50 (the limit I have set), and then from there gone up incrementally, as the tps can fluctuate a bit naturally (so a perfect binary search wouldn't work)
The text was updated successfully, but these errors were encountered: