Here's the first step in breaking down the performace issues. I chose the Balogna Pony because most people are familiar with that map. I also chose this map because it has many of the elements we want to test.

The first map is just the way dash created it. Simple amd sweet, it contains 3 textures.
Most of the variables are in the sky section. Here's what we have. They are the same beteewn the two maps as I want to only thest the effects of adding textures. I will also make changes to the other elements as we progress.
Here's the skinny.

New Sky(Sky) {
position = "336 136 0";
rotation = "1 0 0 0";
scale = "1 1 1";
visibleDistance = "1600";
useSkyTextures = "1";
renderBottomTexture = "1";
SkySolidColor = "0.640000 0.148000 0.215000 0.000000";
fogDistance = "1000";
fogColor = "0.150000 0.010000 0.320000 0.000000";
fogVolume1 = "1 1 1";
fogVolume2 = "1 1 1";
fogVolume3 = "1 1 127";
materialList = "game/data/skies/skyblue.dml";
windVelocity = "1 1 0";
windEffectPrecipitation = "0";
noRenderBans = "0";
fogVolumeColor1 = "1.000000 1.000000 1.000000 1.000000";
fogVolumeColor2 = "1.000000 1.000000 1.000000 1.000000";
fogVolumeColor3 = "1.000000 1.000000 1.000000 1.000000";
cloudSpeed3 = "0.003";
cloudSpeed1 = "0.002";
cloudHeightPer0 = "7";
locked = "true";
cloudHeightPer2 = "0.199973";
cloudHeightPer1 = "0.3";
cloudSpeed2 = "0.001";
Balogna Pony 3 textures. (DB3.mis)
game/data/terrains/FTerrain/FTerrain01
game/data/terrains/FTerrain/FTerrain02
game/data/terrains/FTerrain/FTerrain03

Bologna Pony 8 textures (DB3(8).mis)
game/data/terrains/FTerrain/FTerrain01
game/data/terrains/FTerrain/FTerrain03
game/data/terrains/FTerrain/FTerrain02
game/data/terrains/minions/marble04
game/data/terrains/minions/drygrass4
game/data/terrains/minions/uw2
game/data/terrains/minions/MossyRock01
game/data/terrains/RedTerrain/RedTerrain04

To truley test this we need to have a single server running on a machine. The host must be willing to run onlt these to maps for a while. Additional servers can alter the performance and skew the results of any testing performed.

Who should host this test?
I need help packing this server with 10 players so we have league like conditions.

Last edited: Monday, March 20, 2006 at 11:47:43 AM

If you need me to host do or anything I am at your disposal.

I'll be testing a different location for my server....... Work XD

If the results are to my satisfaction, I'd be more than happy to host for night testing.

B*D

Last edited: Monday, March 20, 2006 at 1:40:17 PM

Put it up I will visit.

Will we only use the scale how it feels for lag, ie players one observation?

Or will you email out the pass so that we can verify pings?

Or is there other software that is available for MACS that we can all use and send you the results?

I'm loking for an existing hoster that we all have used. Like dash, nuts, or BC. I can host the thing myself if we're looking for non-server class hardware.

Dash would probably be the best suited and NUTS & BC's MI server would be next in my opinion.

I'll host it. Send it

We played a full season of PSL on the Pony. Lag was minimal if noticeable at all for any East Coast player. We never had any complaints at all, unless they remained silent.

Excellent to see this going.

Here is an idea. A scoring sheet where we can note the conditions at the time and indicate some performance ratings.

Here's a crack at it.
-------
Server Testing Scoring Sheet
Server name:
Host:
Map:
My location:
My ping:
Date:
Time of day:
# Players in server:
Anyone with ping>200?:

Shot delay:
Shooting efficiency:
Collision accuracy:
Comments:

-----

I am sure I didn't think of everything. Would be nice to get many reports and compare notes.

Per JJ's suggestion I'm moving my performance discussion here.

Even if Napalm can iron out the maps for smoother play, centrally located data center servers might be of value for predictability. I've found http://www.colomart.com in Dallas with killer prices and http://www.riptidehosting.com in Denver with good carriers(Internap, Savvis, and Level3). Neither of them has a setup fee or more than a one month minimum contract, so it'll be low risk to try them.

Last edited: Tuesday, March 21, 2006 at 6:06:41 PM

Good work here, nappy. Looking forward to seeing what yous come up with.