the 100ms delay doesn't mean that latencies higher than 100ms are hopeless. Don't know how you gleaned that from it. The only time that would hold true is if the server operator had sv_maxunlag set to 0.1, but 1.0 is the default. Players with a lower ping/latency still have an advantage in some respects, but this 'unlag' feature takes a good bite out of that latency sandwich for you.
On Thu, 14 Jul 2005 23:55:48, Dan Sorenson <[EMAIL PROTECTED]> wrote: > Heck, I'm suprised I came so close on some of these > assumptions I made, and thanks for the pointer. I was basing > most of my logic on the old classic engine, and it's been > about 4 years since I've investigated its behavior. > > However, I don't see where gaming the prediction and > interpolation by cutting client-sent updates wouldn't throw > off the results. Whether the engine would react by going more > or less in favor of one client in this case is not mentioned, nor > would I expect fine details of the server's logic. > > What has me more curious is where it states the Source > engine does entity interpolation with a 100ms delay. I don't think > I can get to my upstream tier 2 provider in under 100ms, let > alone to a server, so right away interpolation is suspect. > That explains a lot of the oddities I see in-game, actually. > It's not the overall rate that's a problem -- I rarely see more > than 64Kbit/s inbound from most servers, it's the latency. > I'll have to experiment some, as we've a number of international > players on our servers with pings in the 120-200ms range > trying to take on 30ms players. Improving things such that > interpolation remains accurate would be a good thing for them. -- Clayton Macleod _______________________________________________ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds_linux
