FPS on default client

[NOOBY93]

mass PvP.

nobody would actually come

It would only work in Valhalla where nobody is afraid of getting knocked out or humiliated, so people come, so it becomes mass pvp

[loftar]

Loftar apparently doesn't believe that we guys are lagging due to the default client fault.

I'm not sure where you got that idea. All I've said is that I can't reproduce it in an environment where I can test and analyze it.

Would help if Loftar could push a slight update to the client that feeds him timestamped statistics, then he could fascist log Valhalla and correlate that with the metrics he gets from the clients.

Perhaps, but before I could do such a thing, I'd need to have any kind of idea of what to even log to begin with. :)

[Granger]

Loftar apparently doesn't believe that we guys are lagging due to the default client fault.

I'm not sure where you got that idea. All I've said is that I can't reproduce it in an environment where I can test and analyze it.

Would help if Loftar could push a slight update to the client that feeds him timestamped statistics, then he could fascist log Valhalla and correlate that with the metrics he gets from the clients.

Perhaps, but before I could do such a thing, I'd need to have any kind of idea of what to even log to begin with.

I would suggest: make a way to dump the complete data send toward a client into a file, timestamped, so you can replay it by simply feeding it back into a client on demand to see how that specific data is processed in it (and be able to profile it).

Give the clients in valhalla a feedback channel where they report:
- on connection system metrics (like a call to dxdiag for windows and pipe it to the server, as an example) so you know with what kind of system you deal with
- last message from the server before rendering current frame (I guess you timestamp them somehow) so you can correlate what the client reports to what it got, or list of messages received (which would contain their ordering) before beginning to render (which iirc halts parsing of new messages)
- time needed to render last frame, or more detailed depending on what currently is available
- other metrics that you can obtain from the client that might be interesting to you (amount of stuff currently loaded, current CPU load, current CPU frequencs free RAM, JVM status, GC runs, whatever)
In case the server resources permit it, apply this to all connections to valhalla for testing.

This should give you replay files and metric feedback to correlate what happened in the client when things went south. Replaying such a file (according to the timestamps, so the replay will be somewhat similar) should either show you, in your client, the problem -or- it won't and then we can search further for what might be happening to give the effects your users see.

I would also suggest to delay your connection to the server when testing at times, strange things can happen in network stacks should the ping get higher (or jitter happens) which won't be noticeable in a normal testing environment as then you'll be near the server...

[DPblH]

I'm not sure where you got that idea. All I've said is that I can't reproduce it...

Just play your game. Its simple. Also hittest is still laggy and takes alot of time. Best way to test - play in a big village, each click drops your FPS by 20 units.

[LadyGoo]

I'm not sure where you got that idea. All I've said is that I can't reproduce it...

Just play your game. Its simple. Also hittest is still laggy and takes alot of time. Best way to test - play in a big village, each click drops your FPS by 20 units.

Join my plot, Loftar

[sMartins]

Join my plot, Loftar

Maybe he is already there watch your back

[iamahh]

best test inside a town with farms, animals and buildings

[linkfanpc]

I play at one-two frames at all times.

Fix your freaking game dudes.

I'm not kidding, the FPS counter says 1-3 FPS.

[Gigglepuddy]

i get some lag with amber client when i have around 400+ pieces of dirt lying on the ground. native client, it takes close to 2000. and by some lag i mean it drops from 60fps on amber,by like 8 fps every additional piece or dirt once the slowdown starts, while on native it can take a few hundred to drop a few extra frames.