Loading...
  OR  Zero-K Name:    Password:   

Performance issue testing

8 posts, 702 views
Post comment
Filter:    Player:  
sort
I've changed the engine version from 104.0.1-358-gb58f0bb ->

I've applied Spring 104.0.1-286-gd47727e to the server to get feedback on the performance issue that some have reported (@Anarchid FIrankAdminAlcur and others). We were previously using 104.0.1-358-gb58f0bb. I have noticed the performance drop sometimes but it is hard to pin down in a benchmark. The likely candidate is 104.0.1-309-ga8b5ffc with its change to garbage collection.

This change is possibly a regression for some rare crashes, but it is the kind of tradeoff we need to make. Play some massive FFA and get back to me.
+3 / -0
6 years ago
Initial testing says performace seems the same, maybe slighty better. How old is the reverted engine?
+0 / -0


6 years ago
11th of March.
+0 / -0
6 years ago
I had performace issues way before 11th of March. This is the battle I first had the "ping"->"catching up"->complete lagout thing. https://zero-k.info/Battles/Detail/488213
+1 / -0
6 years ago
Played some games. Performace definitely hasn't been improved with the revert.
+1 / -0
DErank1a
6 years ago
I wanted to benchmake but was given broken links :(
Is testing only done through vague anecdotal postings of players like psanic?
+0 / -0


6 years ago
There was a benchmarker and I don't know whether I should control what dead links people post. Technically it can be used but it is difficult. It also didn't tell me much when I tried so I need to think up a better one.
+0 / -0

6 years ago
Inbuilt bench marker available through options could be very good idea. And unique feature. I never saw game with inbuilt bench marker.
+0 / -0