Hard to believe after all these year that you still can’t handle a reasonable sized battle without the game near grinding to a halt.
1 Like
What other games have 3k+ player fights? How is this a reasonable sized battle?
2 Likes
Are you serious? Look at what needs to be rendered? With clustering, cloud tech they should be spooling up additional resources based on performance alerts. There’s no excuse at this stage of the games life cycle.
Again, what other game deals with these kind of numbers?
2 Likes
How you can defend this performance is beyond me. I’m currently working on mid range kubernetes clusters handling billions of transactions. I’m sure this is backend issue. No way this is render issue. I’d love to see the transaction queue on a database and or infrastructure hosting it.. But at this stage in the games life cycle these issues should have been addressed by now.
1 Like
Who cares what other games do or have? CCP markets these battles as aspirational goals, as the pinnacle of entertainment and gameplay. Yet, they do nothing to make them feasible and enjoyable. This is disgraceful and has been a frustrating failure for years from CCP.
We are currently on a reinforced node with only 3.5k chars in system, yet the game performs like there were double that number on an unreinforced node. Nothing activates, nothing locks, every intentional or unintentional camera movement runs the risk of the game crashing. There is no way to feasibly do something else other than being in a fleet ship (like having a T3D sniper squad taking care of hostile ewar). Docking takes ages (I’ve been trying to dock for the last 20 minutes and just now entered the black screen docking phase, probably for another 20 minutes), trying to get back after reshipping takes ages, getting disconnected means at least an hour on client startup and blackscreen.
It is completely irrelevant that there are no other games with battles at this scale. If you market yourself based on such battles and actively encourage them (by virtue of game mechanics necessitating them), as a company you should do much more to make this desired gameplay experience usable and feasible.
GPU Pipeline was supposed to improve performance, but even in Potato Mode, I notice nothing of that sort. In fact, I think the performance has gotten even worse thanks to GPU pipeline. This is a disgrace.
4 Likes
The
of EVE online:
- One of the twins stays at home playing EVE participating in a giant battle with severe time dilation, taking 12 hours to play 30 minutes of battle.
- The other twin goes to a party and gets laid.
The first twin had zero fun, while the other twin had a lot of fun.
CCP and Nullbloc leaders: Being an F1 monkey and participating in TiDi battles is the pinnacle of EVE Online accomplishments.
3 Likes
Well since you’ve avoided answering the question, I’ll assume you can’t answer it. From my knowledge, there is NO other game that can come close to the scales that EVE does. CCP is proud that they can have these large battles. Yes, we want there to be less TiDi, better performance, etc. But the tech just isn’t there. Other mmos have hard limits to what is possible. In EVE there is only one “instance”. EVE does reinforce systems in which large battles occur, it’s why you could get that many players in.
What are you proposing as an alternative? Limiting systems to only 500 players? Then hostiles can’t invade and then you’ll actually have a blue donut. I bring up other games as that’s their competition.
Yes things break down when you reach the limits, this isn’t a failure if no other game can do it better.
GPU pipeline never said it could make large battles have more server performance. Just more performance for your PC. The limit here isn’t your PC, it’s the server itself.
2 Likes
There is no reason to answer a question that is irrelevant to the point.
Yes, they are proud to misrepresent these battles in their promotional materials as something great, and that has literally less than 0 to do with reality ingame.
Distribute the load to more systems. If a large battle for a structure is about to happen, the main target is not relevant to the fight any longer. Instead, it spawns nodes in several systems that you need to hold as owner for the timer to get the structure process to go through, or hold as the hostile to prevent the timer from ticking up. We already have this kind of mechanic with opposing timer progression from Insurgencies, FW influence, even Sov Node Warfare.
No more single point big battles in one system but spread out battles with several fleets in several systems all being useful and actually able to do something. If the tech is not there for big battles, prevent the big battles altogether, without preventing the big fights. Spread it out on several nodes instead. That tech is already there, it has already proven itself successful in distributing unreasonable server loads, and it should be applied to anything that goes beyond practical system loads the server can reasonably handle.
Another 15 minutes gone since I got into the Black Screen docking phase, by the way. I just arrived in the hangar.
3 Likes
So if no one has been to mars, it’s a failure of everyone trying to make it possible so they shouldn’t even try?
They are not misrepresenting anything though, they are very clear that those battles took 20+ hours. If you don’t like it, don’t do it.
And what about wormholes? Or smaller corps. That battle was over a Fort, not a Keepstar.
Welcome to TiDi, at least you didn’t crash. I don’t like large fights because of TiDi either, which is why I stay in low sec. But without it, those large fights would just cause server crashes and an invading force can’t perform objectives. Even if CCP were to spread things out, the null blocks would just overload everything anyways as objectives got completed and they converge to where the most resistance is. Y’all aren’t even bringing 10% in most cases as it is.
1 Like
Wrong analogy and weird argument in any case. This is not about going to Mars. This is trying to go to Mars with a system that is not capable of going there feasibly. If you watched For All Mankind, what we currently experience in EVE is akin to the North Koreans attempt to get to Mars. They got there but in tatters, with human sacrifices, and in no shape to do a proper operation on the planet.
These battles are not like a decision whether to go or not, they are about whether to use a reasonable and practical system to get there comfortably and safely or with a system that is not meant to do that journey.
None of their promotional materials about any of these big fights mention anything about the utter rubbish performance level, the issues, the wasted time. They gloss over all of this just so that they can talk about how much RL money was destroyed in a battle.
In the past, I limited my suggestion to only Keepstars but since big battles can errupt over anything, the server should just by default spawn nodes for anything involving big null block structures. As for wormhole space, their systems are also part of constellations and have connections to other systems via wormholes. The server can use that to determine where to spawn the nodes. This would only occur if a big battle is likely to happen that the server cannot reasonably handle in one system. The nodes would theoretically not even have to be in different systems. Already not being on grid with the main battle reduces tidi stress considerably. With all their crappy distributed server node level tech stack stuff, CCP should also be able to set different parts of a system (planets for instance) on different nodes and put them on RF nodes individually.
I do not see how smaller groups would be impacted in any way by that. They cannot spawn 3k chars that could trigger this mode. And even if, that’s a detail that can be ironed out during the proper feature development.
I rather prefer a BSOD client than this crap, to be quite honest. AT least you know that this stuff is done and gone and you can just do something else.
No, that is not possible. As stated in past instances of this proposal, you need to stay on those nodes and control them for the entirety of the timer. This is not meant to be like running a node to completion and then another node spawns. There are only X nodes in total and you need to control the majority for the timer to progress in your favor. This way, you can stack a huge fleet number on one node, but then other nodes can be controlled by hostiles and you still lose the timer. This is meant to discourage the stacking of N+1 fleets on one node.
Right now, both sides have 6 fleets each at least. Let’s say, there are 5 nodes. Each side has to invest at least 1 fleet per node. That leaves 1 fleet for each side to reinforce 1 node fleet. Other reinforcements can then be brought into other systems to help or attack the 2 fleets in those systems and get node control for their side, while the 4 fleets are active in another system, wasting their time and ships there.
1 Like
So screw small alliances? At minimum you need 5x256 pilots to anchor a fort? Or do anything else?
1 Like
Why would they be screwed? Are they likely to spawn a big fight with thousands of chars involved? Have you heard of anyone creating a huge tantrum over a random small group’s Fort in an unremarkable place? I have not.
As said, this would only trigger if big battles are likely to happen. Some random scrub alliance anchoring a Fort in Providence, Cloud Ring, Heimatar or Aridia is not going to do that. For smaller groups, nothing would change unless they manage to draw the ire of some big block for whatever arcane reason. In that case, their Fort is dead, with or without distributed fights.
People bitching about gaming tech they know nothing about. This game has by far the biggest battles of any game and each ship is controlled by one of thousands of other players all from different parts of the planet and they all want to know why there is latency issues!
Glenny, stop eating the puff balls. They are not popcorn, but opium drops. 
There are hardly any latency issues. These are processing power issues.
1 Like
30 years in the IT industry on the infrastructure architecture side. I know what I’m talking about, This is a legacy code issue. They haven’t bothered to upgrade the back end. I believe it was on MySQL at one stage. No excuse at this stage given the games age,
1 Like
I think Star Citizen has gotten up to 2k.
I am pretty sure that is the shard (server) cap, not the instance cap
you need to rephrase to ‘any more’ - 5 years ago the servers acted weird with 10k players and ate capital ships with no killmail, however, fights with 5k players went pretty well.
It looks like somthing changed in between that adds extra server load.
5min to process a single command is slow, too slow
1 Like
I agree with @Wookster33 . This is not an argument. Perhaps no other game has battles with so many players, but neither does EVE. When 3k players is in the same system, and perhaps lower ammounts can trigger it as well (can’t prove, I only saw 3k+ recently), the game is absolutely unplayable and not fun at all for any of the side.
I accidentally entered Otsasai two weeks ago when there was such battle with 3000± players in local. Just exiting the system took me 15 minutes.
What is point of allowing such battles if the server can’t handle them? It just results into bad experience for all involved.
Naturally, players are at fault as well. AFAIK they intentionally bring so many players (and alts) to trigger the TIDI which is then more beneficial for defenders.
EVE might allow battles with 3k+ players unlike other games, but its is extremely unfun to be in such battle as the game is extremely slow, laggy and unresponsive.
2 Likes