The game default tickrate is 20. If you use a newnet it needs a minimum of 60. Most players prefer newnet and if you have a fast single core speed on your server's cpu and not tons of players newnet at 100 tickrate will be a very nice experience.
Check this out and especially the screenshots comparing tickrates:
https://wiki.unrealadmin.org/Netspeed_Tutorial_(UT)
For my test server I usually set it to 60(normally it's just vanilia with relics), did mess around with 120 tick rate when I had a better pc as it was just me messing around with bots testing stuff for node utstats.
Also stop whining about ACE being closed source every 5 posts. 
snowguy
Thanks for replying.
drunken_fool
besides the anti cheat the newest ace has better crosshair support
the_cowboy
why would you open source anti cheat? wouldnt the cheaters just have a better understanding of whats going on? maybe im not following but doesnt seem like a good idea
friendofafriend Security of the system has nothing to do with closed source nature. If your abstractions are principled enough, you don't need to hide your implementations. I am not sure how versed you really are with coding, you can take example of linux, an open source software being used on servers of google and Microsoft and one of the chief reasons is that Linux is more secure.
You are doing a great disservice to the community if you hide your software implementation under the pretext of security because guess what, even uscripts are able to bypass ace make it a laughing stock personally.
Cowboy does make a good point the cheat anti-cheat communities do seem to have it backward. If I was the ACE of ACE I'd want the number 1 server in all of UTville showcasing my mad coding skills but then again SnowGuy probably doesn't wanna run test versions on his servers anyway.
How much would it cost to run a good 100 tick rate server for a year? Maybe we can all donate for a full year?
London 2 and Atlanta combogib are already at tickrate 80. I could put either one at 100 if a few players started playing there regularly and wanted it moved up. Yesterday, there were 3 or 4 players on London 2 while the main combo was full and stuck on niven-caos 
What i meant was, to put london 1 on 100 TR and most modern Newnet while keeping the 18 players max
I'm pretty sure that lots of people (Myself included) wouldn't mind donating to get something like that
London 1 is the most populated server and it would be nice to play on a modern server settings with all these people
There is no way you are getting 18 players and 100 tick rate with newnet, look at the netstats(F6) on the zp server with 18 players, then look at how much higher it is on a 100 tick rate server with newnet with 5 players.
Pug servers with 100tr have 10 players max, the public server with 12 players couldn't handle newnet 100 tick rate, had to be put down to 65 or 85.
Ooper Pug servers with 100tr have 10 players max, the public server with 12 players couldn't handle newnet 100 tick rate, had to be put down to 65 or 85.
Hmm, just wondering if native C++ code can help in this area. I mean to scale 100 TR to 20+ players.
I am complete noob in this type of calculation and one way I can improve my knowledge is to actually make native newnet type of mod given that'd indeed be useful i.e better scalable.
Ooper It doesn't really have to be 100. 65 or 85 is also fine.
Anything is better than ZP and 20 TR
The choices that are made are usually just compromises between the various hardware and software limitations with the hope that the result will be the best for the majority of players.
The London 1 host does not offer faster CPUs in their London data center. Redirecting all players to the London 2 server would be the solution to that but I don't want to do that until more players are happy with the set up there. Additionally, I actually prefer the hosting company that provides the server for London 1. Maybe before long that host will have newer CPUs on offer that the server can be moved to.
We shouldn't get held up on the idea of a 100 tick rate anyway. Newnet at 60/65 tick rate can still offer a very good game for lots of simultaneous players. I only moved the Atlanta and London 2 servers up to 80 TR because it made the grapple feel a little less jerky on the most updated newnet available. The recommendation for newnet at a 100 tick rate was specifically for X-Cell because he operates a server with a very fast single core speed CPU for friends and family, most likely always less than 10 players.
Increasing the tick rate on zeroping might not actually improve things anyway. There is a bug with zeroping that cause some unregistered hits from time to time and increasing the tick rate may actually exacerbate that issue. The improvements in the feeling between the London 1 and London 2 servers are more likely to be the differences between newnet and zeroping than they are between the different tick rates.
There are many other considerations in all of this but it would take a long time to detail every choice that is made. Ultimately, the plan is to always improve the experience for players and one day the main combo server (wherever it ends up being) will most likely run a newnet at a higher tickrate.
Edit: London 1 is ZP at 40 TR by the way 
the_cowboy
A native implementation of newnet could improve performance but without some way to profile things properly we couldn't be certain of where the best place would be for your optimization efforts. I mean the newnet code might not be the bottleneck when trying to run the game loop at 100 times per second
the_cowboy I will admit i'm not well versed at all in the matter, but the guy does seminars n stuff about cheat protection i'm sure he has a good reason to keep it closed..
have you ever tried talking to him? maybe he would let you help him out with some anti cheat stuff? when ace first came out it was awesome it did catch a bunch of notorious trigger botters.
regarding newnet vs zp i always preferred zp at least for instagib, but moving around on high ping feels awful compare to newnet.
friendofafriend Ye I have already talked to him and he seems to be of the mindset "security by obscurity" which is not good in my books.
For instance, he claims that HwID estimated by ACE is part of security feature and therefore should be closed source.
To counter that precisely, I have written MrEHasher and have shown that electronics hashes can be obtained via code that can be made public.
P.S. I'd have loved if he had posted the video of that talk!
Just leaving this thesis here
https://liu.diva-portal.org/smash/get/diva2:1605200/FULLTEXT01.pdf
I'd be glad to discuss and debate the lag compensation and zero ping relevant stuff.
Now we got cr37 another cheat. It's amazing lol.
Can you check if that's lolz , ! . etc... couldn't see his flag last night I was bugged.
The thing with this guys is they have no shame, they literally set their aimbot on nightmare mode. At leats sbt is little more discreet lol.
I know you're under a lot of pressure snowguy and not much you can do really.
I just wish one day all this lamers get caught. I wouldn't even banned them , I would just let them play with no hacks and let them get some pay back.
drunken_fool Now we got cr37 another cheat. It's amazing lol.
Yes, I've checked him already and you are probably right about the aliases you've named. I don't want to reveal player aliases as it allows you to almost play anonymously, if you choose or have reason to do so. I will only say that this player has been discussed here before. Hopefully, that doesn't give it away completely.
I am not sure how you can detect "assisted aimbotters" spectating doesn't show all mouse movement accurate from what it seems, I did spectate them both and they do miss shots, sbt does however have an insane reaction time to flickshot me around a corner on a pretty long range 8/10 times
