@callmebackdraft said in So whats with the hit detection?:
1 up to 6 ships per server
Their only 5 max
So whats with the hit detection?
@xultanis-dragon Now you're just being too broad - vehicles don't just apply to what moves on the ground. It applies to what moves in the air and in the water too. In essence, you've gone too far!!
Too far!
Too far.
Too far...
Um. Does anybody else hear an echo?
@galactic-geek said in So whats with the hit detection?:
@xultanis-dragon Now you're just being too broad - vehicles don't just apply to what moves on the ground. It applies to what moves in the air and in the water too. In essence, you've gone too far!!
Too far!
Too far.
Too far...
Um. Does anybody else hear an echo?
I see you have played knifey spoony before.
@xultanis-dragon said in So whats with the hit detection?:
@galactic-geek said in So whats with the hit detection?:
@xultanis-dragon Now you're just being too broad - vehicles don't just apply to what moves on the ground. It applies to what moves in the air and in the water too. In essence, you've gone too far!!
Too far!
Too far.
Too far...
Um. Does anybody else hear an echo?
I see you have played knifey spoony before.
Yes? 🤥
Lol, cloud based vs dedicated servers is not a thing.
There are loads of cloud services that are are fully run on dedicated hardware.
Scalability also has nothing to do with this, i can make a scalable system just as easy to implement, scale and deploy on a multitude of dedicated servers as i can on a multitude of virtual servers.
Because guess what, the communication to said systems works exactly the same.
Nowadays these things dont really are as “simple” as vps vs dedicated.
You work with microservices, containerisation, deployment groups yadda yadda yadda.
I am not a spokesperson for azure, nor am i for aws or google.
What these services (any of the ones above, provide is a software overlay to their datacenters that give easy of setting up services, networks, pipelines etc.
For anything you need, be it dedicated hardware, vps’, services, kubernetes clusters or whatever.
Thats it and with their vast amount of money pumped into their platforms they have the ability to have enormous fiber optic backbones backbones to their data centers and thats not all.
They also have the ability by using their platform to create networking tunnels to and from their locations, be it inside a single data center but over the split of buildings or to and from other sites globally.
Like i said, azure, aws, google whatever is nothing more than a layer of management software that makes it easy to work with a global cluster of sytems be it virtual or dedicated.
Again, please do some deeper research, you are suffering from the dunning kruger effect.
And thus you are making statements that are simply untrue.
Do you not grasp the concept between virtual and physical in the technical terms??
They also have the ability by using their platform to create networking tunnels to and from their locations, be it inside a single data center but over the split of buildings or to and from other sites globally.
Am I to understand it that you just believe what is going on is that there are a ton of servers in multiple locations and what people are doing is just acclimating ONE of those servers for use of business and now that server is the dedicated server for that service or business??
Broski I really hope not and if anyone is suffering from Dunn Kruger syndrome its you because it seems you actually don't understand the concept between virtual and physical when it comes to networking.
I don't know if you actually understand networking but you need to research what virtualization is and understand that the cloud is using that.
They also have the ability by using their platform to create networking tunnels to and from their locations, be it inside a single data center but over the split of buildings or to and from other sites globally.
You think this is a special feature?? Networking tunnels are NOT something that cloud does. That is just networking. Thats the routers and the switches doing that NOT the servers.
Honestly man, I don't think you fully understand whats going on here. I think you are mixed up on what is ACTUALLY going on versus what you are picturing inside your head.
If anyone needs to do a deeper research its your because you are talking as if a server that you are using in a remote location is the same as "cloud" and it is NOT.
Just because its remote and in a different location does not make it Cloud based.
Cloud based or Cloud computing is the implementation of virtual system.
"The main enabling technology for cloud computing is virtualization. Virtualization software separates a physical computing device into one or more "virtual" devices, each of which can be easily used and managed to perform computing tasks. With operating system–level virtualization essentially creating a scalable system of multiple independent computing devices, idle computing resources can be allocated and used more efficiently. Virtualization provides the agility required to speed up IT operations and reduces cost by increasing infrastructure utilization. Autonomic computing automates the process through which the user can provision resources on-demand. By minimizing user involvement, automation speeds up the process, reduces labor costs and reduces the possibility of human errors."
You should really understand the wiki quote you posted and actually READ what cloud is and understand the concepts.
Cloud is Virtual, not just a server in a remote location, understand that concept first please.
Yes i do know all those concepts.
And yes virtualisation is the main driving power behind cloud systems.
However that doesn’t mean that all cloud systems need to be virtual.
I know network tunneling is not a just one of the features that these services offer, however they can do it on a different level and in that case yes its a special feature and its not just “the routers and switches”
Anyway, i am done with this conversation, if you wish to spread misinformation go right ahead, the misnomers you have used all over your post are really telling and secondly your sweeping generalisations might be even more so.
If you wish to be kept mostly in the dark of the options, methods etc.
Cloud doesnt equal virtual as much as a vehicle is not equal to car.
Cloud systems can be virtual just as much as vehicles can be cars....
Anyway like i said not my job to educate you, theres schools for that.
@pithyrumble ah yes support rare when they break their own game and haven't been able to fix it for the past 3 years while filling up the emporium with new stuff every month. This isn't only about hit reg this is about rare caring about their game breaking issues. If people just say oh rare its fine that the hit detection doesn't work well, then what if other game breaking things happen? why should they fix those if the community is fine with a broken game?
@uzugijin how are we supposed to tell rare to fix their servers, or code? its not like I have access to all their files. All we can do is tell them its broken and submit videos.
@theeggoplant said in So whats with the hit detection?:
@uzugijin how are we supposed to tell rare to fix their servers, or code? its not like I have access to all their files. All we can do is tell them its broken and submit videos.
you don't have to they know its broken
and they've said it wont be a quick fix
@callmebackdraft said in So whats with the hit detection?:
Yes i do know all those concepts.
And yes virtualisation is the main driving power behind cloud systems.
However that doesn’t mean that all cloud systems need to be virtual.
Did you just forget to hit reply or were you intentionally trying to get the last word??
It pretty much is. Thats whole concept behind cloud. You are using resources from multiple sources to run your program. If you were a business and you wanted to run a program through the cloud, then the cloud would virtually run that program. That program isn't designated to ANY one server or ANY one device. They could lose half their servers and you would still be able to access that program because the program is hosted virtually and not on a physical desktop.
Cloud services are becoming mainstream for hosting gaming and its causing problems everywhere as more and more companies switch over to it.
I've already asked you to prove to me that there is a game that requires player twitch reactions that is NOT having problems right now.
Before a game missing a hit reg was a random occurrence that was almost ALWAYS associated with lag.
Now you have games all over the place have hit registration issues.
Virtually hosting a server is not as fast or responses as a dedicated one. It might work well for something like Minecraft or even WoW but not something like a FPS game.
I know network tunneling is not a just one of the features that these services offer, however they can do it on a different level and in that case yes its a special feature and its not just “the routers and switches”
On this one I misspoke. For some reason I thought you were trying to suggest that Cloud acted as its own layer 3 device. You were speaking about something like VPN's and the like.
Anyway, i am done with this conversation, if you wish to spread misinformation go right ahead, the misnomers you have used all over your post are really telling and secondly your sweeping generalisations might be even more so.
Its not misinformation if I'm right. Dedicated servers are faster at responding. Cloud servers are great inovation but are garbage when it comes to hosting high responsive games and that goes ESPECIALLY for Azure servers.
If you wish to be kept mostly in the dark of the options, methods etc.
Cloud doesnt equal virtual as much as a vehicle is not equal to car.
Cloud systems can be virtual just as much as vehicles can be cars....
Yeah for some reason I don't think you actually grasp the concept of virtualization.
Its using software to pretend you are using hardware to run software. Like I said, you are using resources from the servers to CREATE a virtual PC that you tell to HOST stuff. You are NOT picking out a server in a rack somewhere to physically host the game.
Anyway like i said not my job to educate you, theres schools for that.
Maybe you should take your own advice and actually try to figure out the difference between virtual and dedicated, because I'm pretty certain now that you think they are both the same thing.
@jollyolsteamed if we go quiet than we approve of this being an issue. Hit reg has been bad for years. Most of this games lifetime this has been an issue. We are coming up on year 3 of sea of thieves and it is not fixed. I understand that this isn't a quick fix but this has gone on for years and a lot of players are really frustrated.
@xultanis-dragon Virtualization can mean that it streamlines the interface by making it appear as though it is running on one system, but the applications themselves are still run natively on separate servers.
So if I have App1, App2, and App3 running I might see them as a single list of running apps, but App1 and App2 and running on Server1 while App3 is running on Server2.
It's an abstraction layer, because you don't need to know that information a lot of the time. The game servers themselves are likely not running in a virtual environment.
@d3adst1ck said in So whats with the hit detection?:
@xultanis-dragon Virtualization can mean that it streamlines the interface by making it appear as though it is running on one system, but the applications themselves are still run natively on separate servers.
So if I have App1, App2, and App3 running I might see them as a single list of running apps, but App1 and App2 and running on Server1 while App3 is running on Server2.
It's an abstraction layer, because you don't need to know that information a lot of the time. The game servers themselves are likely not running in a virtual environment.
That is very true, however if server 1 and server 2 go down, then server 3 and 4 take over. The apps are held together by the infrastructure not by the machine.
Sea of Thieves uses PlayFab. PlayFab is virtual.
Servers per machine - Required. The maximum number of game servers that should be operated on a single virtual machine. A typical configuration may have four servers on a Standard_D2s_v3, providing each server effectively 50% of a core.
Deploying PlayFab multiplayer server builds - PlayFab
Describes how to deploy a PlayFab multiplayer server.
Basics of a PlayFab game server - PlayFab
Describes how game servers are packaged and integrated with PlayFab systems.
So regardless of whether Azure servers can be good for gaming or not, the current server configuration is very much NOT good. A tickrate of about 30 is unacceptable for a first person game.
@theeggoplant said in So whats with the hit detection?:
@jollyolsteamed if we go quiet than we approve of this being an issue. Hit reg has been bad for years. Most of this games lifetime this has been an issue. We are coming up on year 3 of sea of thieves and it is not fixed. I understand that this isn't a quick fix but this has gone on for years and a lot of players are really frustrated.
Why would that mean that we approve of it?These posts wont make hitreg fix itself any faster
All this serves as is a reminder to rare that hitreg is a problem which rare is already aware of
at this point these posts are just spam
@callmebackdraft said in So whats with the hit detection?:
@xassflamex this used to be 6 and indeed i said 6
its 5, and its never been 6.
@xassflamex said in So whats with the hit detection?:
@callmebackdraft said in So whats with the hit detection?:
@xassflamex this used to be 6 and indeed i said 6
its 5, and its never been 6.
It used to be 6. There are plenty of videos on youtube confirming this.
@xultanis-dragon I don't know much about playfab deployment, but it's possible since they provide the base package with os details for development, that when you publish it is pushed to a native environment that you manage as a virtual server (a cluster of servers running the base package as the OS). This would allow you to increase storage, cpu, ram, etc.. without having to worry about hardware. I could be wrong though.
Even if it was running under a full vm, I don't think this would affect performance in regards to hit registration. That is likely more dependent on server executable optimization. If the tick rate was higher it would help, but it also has to be stable. Fluctuating tick rate is probably worse than a lower, stable one.
I'm pretty sure them capping ships on a server to 5 down from 6 has also helped. They've put effort and honestly I think they should removed like 1/3 of the roar to help also. It looks odd having an extension of the map and having 18 islands cut down to 12 would surely help in an area that most people avoid anyways.
We could easily cut 1/2 sea post, 2/5 large island and 3/9 small islands.
EDIT: I realize that voyaging in this area would be easier due to less islands and smaller space but the reward of hit reg and less lag would make it worth, not to mention the area would be frequented more often by voyagers and people looking for pvp.
@xassflamex like @D3ADST1CK said it used to be 6 but they turned it down to 5 to try and mitigate some of the server issues they have been having
@awsmstaccntname said in So whats with the hit detection?:
I'm pretty sure them capping ships on a server to 5 down from 6 has also helped. They've put effort and honestly I think they should removed like 1/3 of the roar to help also. It looks odd having an extension of the map and having 18 islands cut down to 12 would surely help in an area that most people avoid anyways.
We could easily cut 1/2 sea post, 2/5 large island and 3/9 small islands.
EDIT: I realize that voyaging in this area would be easier due to less islands and smaller space but the reward of hit reg and less lag would make it worth, not to mention the area would be frequented more often by voyagers and people looking for pvp.
Cutting or increasing map size won't change anything because areas are only active around player ships, otherwise they are dormant.
So problem with hitreg is that people already had figured out how to use it as part of the game.
I played against somebody who implied that they were using hitreg, and to prove it that player jumped on the galleon and walk through tornado of swords. Then pulled blunder and wipe out galleon.
I figured it was only matter of time.
And Yes, supposedly works the same way in adventure.
So yeah hitreg needs immediate fix at this point because people will be walking out with Athena chests right out of the boat.@d3adst1ck I get that but Rare also said that they wouldn't be adding new areas of the map for this exact reason though which makes me think that some memory is used still. Loot, etc, not sure that's handled back end if it's spawned upon player entry of zone and remains there if not picked up for the next crew to see and grab or despawns if no other ship is in the area. It'd take some memory to still keep that loot available.
@awsmstaccntname the reason they don't want to expand the map is because it will spread players out to much, not that the game can't handle the extra space. If they're able to add more ships to an instance without sacrificing performance, they'd probably be willing to expand the map also since they have an ideal encounter rate between player ships.
@D3ADST1CK when will you realise your wrong and have been beat lol. Your so wrong it actually hurts to read your reply. Also, @Xultanis-Dragon happens to be a computer IT guy so your really arguing with the wrong person and look like a fool.
Anywho,, hitreg in SoT is not host side problem, its problem with user side game portion. Thats why it can be solved on user side, or that's why affects some people more than others because of the hardware difference.
Cloud service that runs Host code has nothing to do with it.
Everybody who are in the same game session are talking to the same code. If it was host side than everybody would be affected at the same time, but they are not.@jadescissors32 said in So whats with the hit detection?:
Anywho,, hitreg in SoT is not host side problem, its problem with user side game portion. Thats why it can be solved on user side, or that's why affects some people more than others because of the hardware difference.
Cloud service that runs Host code has nothing to do with it.
Everybody who are in the same game session are talking to the same code. If it was host side than everybody would be affected at the same time, but they are not.Negative. Even players with the new series X have issues as well. Its not just host side or client side. The issue is that the majority of the problem is host side.
Its not that the Cloud is running host code. Its how the cloud initiates the server and then how it acclimates the resources needed to run it. The OS, the RAM, the CPU, etc etc.
Client side has never been a severe issue in the past and has only been an issue since gaming companies have moved their hosting services to cloud infrastructures.
Don't mind creed. He got you mixed up with someone else. Even still, I am mad at him at the moment. Most days I'm mad at him though. Everytime I get online its always "HEY DRAGON WE JUST HAD A HUGE FIGHT WITH 80000000 SHIPS, IT WAS AWESOME!!!" - I get online and its just dead silence on the seas. He gets online and its fight after fight after fight. Some seriously envy and salt I got going on but yeah dont mind creed.
@xultanis-dragon I still think that under a full VM, where the game server exe is running in an emulated OS, that the Azure servers are more than capable of providing the compute power and throughput to service players properly.
If you want a comparable, you can look at Rainbow 6 Siege which is also hosted on Azure/Playfab. A battlefield youtuber who started doing network analysis and branched out to other games did a few videos on R6Siege also.
Here's a video from 5 years ago when it was running under 50Hz and mentions a few problems that seem similar to what you can experience in SoT. At 5:08 he goes into Lag Compensation on the server end which seems to be very similar to the 'noodle sword arm' and 'backtrack' on guns behaves suggesting that SoT doesn't consider player latency when rewinding game state to determine hits.
https://www.youtube.com/watch?v=vBZj4c_j9Qs
This one details changes made to the netcode in one of the patches from 4 years ago. They didn't improve lag compensation, but did make the ping kick a bit more restrictive. I know that the player latency in SoT is pretty lenient because I have had a pretty crazy ping while my local network was congested for several seconds and didn't lose connection.
https://www.youtube.com/watch?v=_FlWQ2gRT3k
Here's one from when they updated to run at 60Hz which seems to address a lot of the problems with lag compensation, but doesn't fix all of them.
https://www.youtube.com/watch?v=KvIyUxaF5pE
It's possible for games running on Azure, whether VMd or not, to run at higher tick rates so I don't think it's a server hardware or VM issue. The issue has always been the game server software. Either it is not optimized enough to run at a stable higher tick rate, or it just isn't possible with the amount of calculations that it needs to run to simulate the game world, ships and players. It's also possible that Rare has some kind of instance count target that they want to hit to run a certain number of instances per region/group and/or for cost reasons.
Of course high tick rate won't solve everything. You can find players complaining about hit reg in higher tick rate games like Overwatch, TF2, CSGO and Valorant, and some of these might be due to calculation bugs or simulation inaccuracy, but the higher tick rate helps keep the game state closer to reality as seen by the players' client simulation.
Of course we're both theorizing from the outside without knowing all the details, so who knows.
79 von 156
