Jump to content

All of the problems with PvP right now


Shaman

Recommended Posts

  1. Damage multipliers between construct sizes are too high
  2. There is no reason to fight (ofc this is being worked on)
  3. Fighting is unprofitable as any loot you get will most likely have lives taken out of it/ be worthless
  4. The investment to get an L core PvP ship is too low, L cores are supposed to be the large org battleships which need a large investment to build and run, but are too easy to get atm. (XLs should be the equivalent to Titans, if not more)
  5. The hit probablility for small constructs are too forgiving to the offender, even super tiny pvp ships with a <2m^2 cross section get destroyed easily. It should be a lot harder if not impossible for large ships to hit constructs like these
  6. pvp is too long range (this one is more bias, but if constructs had a lot less range like 500-1000m most of these kind of problems would disappear, as well as being a lot more cool looking. you might want a way to catch up with ships as well if this is implemented)
  7. players can just enter third person under layers and layers of voxels, which means that it is hard to kill players in most PvP ships. Making seats more like the currently useless cockpit would make more sense.
  8. PvP is boring to watch + buggy
  9. most meta ships are just glorified cubes atm
  10. players can just warp out of danger
  11. nothing is stopping you from building a ship the same size as an XS/S/M core with a L core instead to get the benefits that the larger guns give, and the benefits of using smaller cores are currently negligible
  12. most PvP battles end in a retreat as fuel, scrap and ammo requirements are too high for any outcome to happen
Link to comment
Share on other sites

1 hour ago, Shaman said:

 

  1. most PvP battles end in a retreat as fuel, scrap and ammo requirements are too high for any outcome to happen


More like ships have a bit too much durability and there are no penetration mechanics so killing a ship is like peeling an ? 

 

1 hour ago, Shaman said:

 

  1. players can just enter third person under layers and layers of voxels, which means that it is hard to kill players in most PvP ships. Making seats more like the currently useless cockpit would make more sense.

AVA When?

1 hour ago, Shaman said:
  1. The investment to get an L core PvP ship is too low, L cores are supposed to be the large org battleships which need a large investment to build and run, but are too easy to get atm. (XLs should be the equivalent to Titans, if not more

L core PVP ships are not cheap believe me. However if new ore gaining methods come this may ring true.

1 hour ago, Shaman said:
  1. players can just warp out of danger

Players can obtain any grade resources and use warp to completely skip danger..

 

 

13: PvP exploits, server sync issues and bugs even bad pvp talents

Link to comment
Share on other sites

NQ really needs to understand that battles being boring to watch is not going to work in today's age of streaming. It will deter players from trying the game instead of being a vital well of marketing. If battles aren't interesting to watch, most players won't find them that interesting to participate in either...especially as a gunner. 

 

I also agree that combat is too long-ranged. What's the real design reason driving this idea that combatants should only ever be represented as radar dots? Realism? Not a good enough design justification. Also a very likely driver of combat lag (due to high velocities)

 

I think they really need to redesign most of their combat systems...keeping in mind that this isn't 2003 where this style of hardcore combat might have worked in an MMO. 

 

 

Link to comment
Share on other sites

On 4/29/2021 at 5:52 PM, ShippyLongstalking said:

NQ really needs to understand that battles being boring to watch is not going to work in today's age of streaming. It will deter players from trying the game instead of being a vital well of marketing. If battles aren't interesting to watch, most players won't find them that interesting to participate in either...especially as a gunner. 

 

I also agree that combat is too long-ranged. What's the real design reason driving this idea that combatants should only ever be represented as radar dots? Realism? Not a good enough design justification. Also a very likely driver of combat lag (due to high velocities)

 

I think they really need to redesign most of their combat systems...keeping in mind that this isn't 2003 where this style of hardcore combat might have worked in an MMO. 

 

 

 

The way that most space dogfighting game make combat fun is one or more of these things, but usually a combination of:

 

  1.  low max speed cap (say 1000km/h for DU)
    This is necessary to force combat into close quarters as well as held deal with netcode issues better if it's aim / projectile based fps combat
  2.  warp drives with variable speeds that can warp wherever you point your ship.
  3. weapon projectile distance caps to again force close quarter combat
  4. g-forces and loss of consciousness due to g-forces moving blood... this adds limits to maneuverability and forces the pilot to consider how to maneuver the ship properly in order not to pass out while fighting.
  5. ship design is 6dof focused, meaning that maneuverability of a ship on all axis is very important in order to dodge projectiles and outmanevuer opponents
  6. projectile speed is capped, bullets are projectile based rather than instant hit (hit scan). Speed of projectile is capped so that if a pilot is good enough they can outmaneuver bullets or at least avoid majority of damage. Combat is not about trying to build tanks to soak up bullets, it also allows for outmaneuvering them

 

The combination of all these things leads to very fun and challenging combat. Part of the reasons for this that combat relies on pure player skill and not on talents or rpg skills. It means that if you suck then it's mostly your fault. Which leads me to another list of what these mechanics lead to:

 

  1. 1) very high skill ceiling. This is a double edged sword because this combat system relies on pure player skill, which requires proper player training, while some people actually prefer rpg skill mechanics that increase their effectiveness in battle. Very high skill ceiling introduces merit based combat, in other words, the best fighter wins, not the best voxelmaster or ship builder, nor even the best ship design. Very high skill ceiling usually means "its about the driver, not the car", while rpg mechanics and talents systems focus more on the car than the driver. People will complain about high pvp skill ceiling because actual player skill will become barrier to entry into pvp rather than talents / quanta / rpg skills.
  2. these mechanics require 6dof (6 degrees of freedom flight model). DU actually has it, except it is useless in a fight because you can't "dodge" and combat is so long distance that there is no practical purpose to strafing / rolling, etc. 6dof combat can be very very difficult to grasp due to maneuver complexity and very insane freedom of movement in close quarters.
  3. In DU context this model is improbable to achieve as the whole codebase is pretty bugged and it looks like DU is just a minimum viable product in terms of pvp. Starbase is doing it with their hybrid p2p system though, so it can be done. Furthermore I personally know developers who have achieved very effective ways of mitigating issues due to high ping across continents (I've played such a game against a person across the world and the projectile prediction was surprisingly on point, everything was smooth). It's definitely possible but it will require NQ to rewrite... everything.
  4. this whole thing is a pipe dream for DU tbh, devs have said they are sticking to lock and fire. Idk if they even know of ways to mitigate high latency seamlessly... I just think the current DU pvp model is a very unappealing mess.
Link to comment
Share on other sites

7 hours ago, IvanGrozniy said:

1) very high skill ceiling.
2. these mechanics require 6dof (6 degrees of freedom flight model). 
3. In DU context this model is improbable to achieve
4. this whole thing is a pipe dream for DU tbh

 

These are interesting ideas!

 

They definitely are hardcore concepts, which I have mixed feelings about. 

 

High skill ceiling can be fun and extremely rewarding, but IMO it works best when players have ample opportunity to practice -- that means either fighting against NPCs to refine your skill and/or making it easier to reset after any death. If combat is both very high skill and very punishing...that's not going to work in 2021, not as an MMO. 

 

The most popular skill-driven competitive games understand that playing against especially high skill people can feel absolutely crappy, which is why they match players based on a rating system. Can't do that in an open-world game.

 

Fair is a very relative concept and is more about player perception than reality. Is it fair to get ganked? It never feels it. Is it fair to fight against someone very skilled as a new player (or simply as someone without great reflexes)? It won't feel like it. 

 

By definition, most people are average at combat; not everyone can "just get good" because that's not how skill-based competition works. 100% skill-based competitions rely on matching people of similar skill to feel fair. Even those with 1000s of hours of practice will never be as good as top-tier skilled players, and it won't even feel like a contest. 

 

I think high player skill should be a huge factor, but not the only one. IMO, the model you describe here would work better in an instanced competitive game with a rating system more than an MMO in today's industry (I'd play that game). 

Link to comment
Share on other sites

41 minutes ago, ShippyLongstalking said:

I think high player skill should be a huge factor, but not the only one. IMO, the model you describe here would work better in an instanced competitive game with a rating system more than an MMO in today's industry (I'd play that game). 

 

It's a double edged sword though. Because tbe problem you'll have with focusing on "the car rather than the driver" is balance. Once you get into the business of balancing you have to commit to balancing everywhere... basically being an overbearing mother that constantly gets into the way of kids playing in a sandbox. And doing that is frustrating for both players and developers. More so for developers because now they have to balance everything. The devs basically create a sandbox mmo and then nurf the sandbox into the ground by fighting the emergent mechanics of a sandbox to absurdium.

 

In a more "its about the driver, not the car" kind of game, you do get a merit based system. But yes not everyone will just practice practice practice to get better. But some will. Nothing wrong with that. It's "fair" depending on your definition of fair. Everyone can increase their pvp skills if they choose to practice them. But yes you'll get super good players that would beat anyone in a 1v1, 1v2 or 1v3... but it's "fair" because they put in a lot of time into practice. Again... depends on how you define and look at these things. Matching up elo ratings in an mmo is practically impossible, this is the nature of an mmo.

 

"Its about the car, not the driver" typically favors the players who have been longer in the game because they accrued more gear / talent points, etc... if something feels unfair people typically blame the game balance and game mechanics (if they were logical about it lol). "It's about the driver, not the car" favors naturally talented players more and their ability to learn and improve. In this scenario people typically blame each other or devs for allowing "exploits" because some players can't comprehend how much more skillfull their opponent is.

 

Also the reason why I divide these poles so strongly is because we hardly ever see a good blend of both simply because of complexity and sunk costs. It's easier to focus on just one type of system otherwise if you do both, you're working on systems that cancel each other out and that become useless. For example, you can have gforces in the game but also a talent system that helps you deal with it. Older players will max that talent and render gforce mechanics essentially useless because they are canceled out by talents. And this can happen for multiple systems across the board such that an older player has way too many mitigating talents/skills that prevent them from experiencing the limitations of the game in the same way that a fresh player would experience it... and there again we come to questions of fairness :). But essentially the question would be: why design the game such that most skills barriers and limitations are negated by rpg mechanics simply because of a player having an older account? You'll get 30-40-50% of the mechanics of the game nurfed or not be used at all due to them being canceled out. So why even try to sink time and money into such a system...

 

It's frankly useless to talk about fairness in an mmo. It's more useful to create game design where a player can improve as a player over time. I think that should be more emphasized.

Link to comment
Share on other sites

On 4/30/2021 at 11:13 AM, IvanGrozniy said:

 

It's a double edged sword though. Because tbe problem you'll have with focusing on "the car rather than the driver" is balance. Once you get into the business of balancing you have to commit to balancing everywhere... basically being an overbearing mother that constantly gets into the way of kids playing in a sandbox. And doing that is frustrating for both players and developers. More so for developers because now they have to balance everything. The devs basically create a sandbox mmo and then nurf the sandbox into the ground by fighting the emergent mechanics of a sandbox to absurdium.

 

In a more "its about the driver, not the car" kind of game, you do get a merit based system. But yes not everyone will just practice practice practice to get better. But some will. Nothing wrong with that. It's "fair" depending on your definition of fair. Everyone can increase their pvp skills if they choose to practice them. But yes you'll get super good players that would beat anyone in a 1v1, 1v2 or 1v3... but it's "fair" because they put in a lot of time into practice. Again... depends on how you define and look at these things. Matching up elo ratings in an mmo is practically impossible, this is the nature of an mmo.

 

"Its about the car, not the driver" typically favors the players who have been longer in the game because they accrued more gear / talent points, etc... if something feels unfair people typically blame the game balance and game mechanics (if they were logical about it lol). "It's about the driver, not the car" favors naturally talented players more and their ability to learn and improve. In this scenario people typically blame each other or devs for allowing "exploits" because some players can't comprehend how much more skillfull their opponent is.

 

Also the reason why I divide these poles so strongly is because we hardly ever see a good blend of both simply because of complexity and sunk costs. It's easier to focus on just one type of system otherwise if you do both, you're working on systems that cancel each other out and that become useless. For example, you can have gforces in the game but also a talent system that helps you deal with it. Older players will max that talent and render gforce mechanics essentially useless because they are canceled out by talents. And this can happen for multiple systems across the board such that an older player has way too many mitigating talents/skills that prevent them from experiencing the limitations of the game in the same way that a fresh player would experience it... and there again we come to questions of fairness :). But essentially the question would be: why design the game such that most skills barriers and limitations are negated by rpg mechanics simply because of a player having an older account? You'll get 30-40-50% of the mechanics of the game nurfed or not be used at all due to them being canceled out. So why even try to sink time and money into such a system...

 

It's frankly useless to talk about fairness in an mmo. It's more useful to create game design where a player can improve as a player over time. I think that should be more emphasized.

It is interesting. The MMO industry standard is definitely to favor the long term player over the newcomer and I think you ignore that at your peril, but I do have some sympathy for the natural skill angle (even though I will never be that person). In my opinion you would ideally make the skills compensate certain things to simulate an 'experience' advantage but not to the extent that it would completely negate natural skill. Things like the g-force would be a 'trained' skill but then if the system can cope with point and shoot then that maybe just stays pure skill based.

 

The hard part of the technology as you scale up though is certainly things like dynamic projectile calculations. If you get a big battle that could be a problem. I suspect that is partly the reason why they have what they have. I would also say, not even including space, large ships tend to not be line of sight when fighting - just look at battleships in WW2 never mind what happens today. This is supposed to be advanced tech. Pretty much any Sci-fi book would have you fighting at thousands if not tens of thousands of miles (at least) apart!!

 

Still interesting discussion - hard trade offs to be sure.

Link to comment
Share on other sites

6 hours ago, Cheith said:

The MMO industry standard is definitely to favor the long term player over the newcomer and I think you ignore that at your peril


I know what you mean and I don't ignore it, I simply don't want to add caveat after caveat, too much rambling on) Yes that is the mmo standard. Time is the ultimate pace halting mechanic in mmos. I did say that the "it's about the car, not the player" approach favors players who have been playing longer.

 

6 hours ago, Cheith said:

 

The hard part of the technology as you scale up though is certainly things like dynamic projectile calculations. If you get a big battle that could be a problem. I suspect that is partly the reason why they have what they have.

 

For sure that is NQ's reason for not having fps combat. That doesn't mean it is not doable. Starbase will have fps combat mechanics for both ships and robots despite the mmo nature of their game. I know of at least 1  indie space combat project that has some amazing netcoding that allows me to fight someone from across the world in ship fps combat with almost no disagreeable latency for projectiles (some fancy prediction is used there and it works without a hitch). NQ is a new company, they have never made a game, let alone an mmo, which, in terms of netcode, is an entirely different beast. This is why we have been having some insane netcode issues in DU, for example massive amounts of data being downloaded during pvp where the mesh requests have absolutely insane rates. FPS combat can be done in an mmo with some tradeoffs, and in the hands of talent it can be done very very well. NQ does not have that kind of talent, even the current lock and fire system is very much scotchtaped together, it simply does not scale. You can check out an open alpha (free) called Hunternet for multiplayer space fps combat. You can also check out Infinity Battlescape for relatively ok netcode and massive space battles (they do have a bunch of bots but they have them replaced by players when they join servers).

 

To be honest I would not trust NQ to make a good fps combat experience in space. It takes someone who actually understands how these things work in game context. and as far as I'm aware, they don't have such a person. Even Star Wars Squadrons failed in that regard.

Link to comment
Share on other sites

9 hours ago, IvanGrozniy said:

For sure that is NQ's reason for not having fps combat. That doesn't mean it is not doable. Starbase will have fps combat mechanics for both ships and robots despite the mmo nature of their game. I know of at least 1  indie space combat project that has some amazing netcoding that allows me to fight someone from across the world in ship fps combat with almost no disagreeable latency for projectiles (some fancy prediction is used there and it works without a hitch). NQ is a new company, they have never made a game, let alone an mmo, which, in terms of netcode, is an entirely different beast. This is why we have been having some insane netcode issues in DU, for example massive amounts of data being downloaded during pvp where the mesh requests have absolutely insane rates. FPS combat can be done in an mmo with some tradeoffs, and in the hands of talent it can be done very very well. NQ does not have that kind of talent, even the current lock and fire system is very much scotchtaped together, it simply does not scale. You can check out an open alpha (free) called Hunternet for multiplayer space fps combat. You can also check out Infinity Battlescape for relatively ok netcode and massive space battles (they do have a bunch of bots but they have them replaced by players when they join servers).

 

I guess my question is what your acceptable level of scale is - is it EVE scale with 1000 vs 1000? Or is it a couple of dozen? This does get important as you even try and build it and whether or not you are calculating trajectories for potentially 100,000s of objects if there is travel time for the objects (which in the skill version there would be). I understand the predictive side, but that would not work for intersections of ship/projectile if you get your skill based option - the last minute twitch would never work - but as you say it is a trade-off. I am also not sure, from what is said, whether or not any of this predictive stuff is client side - I would hope not as that is where the cheating comes in and it always done. Client side code is the easy way to shed a lot of calculation effort but sadly all players cannot be trusted to play fairly and there really is no good way of stopping it (just making it more difficult). Also client side anything is a great desync route.

 

Link to comment
Share on other sites

3 hours ago, Cheith said:

I guess my question is what your acceptable level of scale is - is it EVE scale with 1000 vs 1000? 

ATM game becomes a shitshow even with 5v5, different people see same ship at different locations and there are always the "lucky" ones that get all their shots as "miss" due to server being unable to calculate whatever it needs and just dumps the shots as "miss". You talk 1000 vs 1000 ?

 

Skill based ?

Did you try following another construct at least at like 10,000kmh ? That thing jumps like a frog, even "lock and shoot" is not sure where it should shoot at.

To make it remotely skill based they have to replace the engine, in other words, make another game.

 

Link to comment
Share on other sites

This is one of best discussions of game mechanics I've seen on the forums in some time. Ill do my best not to derail it. 

 

Speed limits: The first thing I've gathered is that combat speeds need to be slowed way down. DU cannot and most likely will never be able to support combat at such wide relative speeds. 

 

I like the idea of keeping combat under 1000kph. I will also suggest that the different size cores have different base max speeds and that ships be able to use afterburners (for lack of a better term) at the cost of greatly increased fuel consumption and potential hull damage. 

 

Combat speed limits will make cheaper warp or some kind of sub warp engines necessary.  Sub warp could require either far less warp fuel or a lesser version.   Perhaps it could still burn space fuels to keep the number of fuels required down?  

 

But everyone will be able to avoid PVP with warp!  When coming out of warp into safe space, yes.  I don't think ganking haulers as they come in to planets is good gameplay anyways.  There should be reasons for people to choose to risk warping their ships into PVP space.  To prevent prey from simply warping away, Eve style scramblers are needed. 

 

Unprofitable fighting: I've said from the start, the current wear and tear system isn't going to work.  I've seen some good ideas that could fix it but the most simple solution would be to have, at best, a 50% chance that a module damaged in PVP can be repaired. Half of any cargo in a destroyed container is lost. Done. 

Link to comment
Share on other sites

1 hour ago, XKentX said:

ATM game becomes a shitshow even with 5v5, different people see same ship at different locations and there are always the "lucky" ones that get all their shots as "miss" due to server being unable to calculate whatever it needs and just dumps the shots as "miss". You talk 1000 vs 1000 ?

 

Skill based ?

Did you try following another construct at least at like 10,000kmh ? That thing jumps like a frog, even "lock and shoot" is not sure where it should shoot at.

To make it remotely skill based they have to replace the engine, in other words, make another game.

 

It was a question - what is the end goal. I get they are not there now but what the end point is makes a huge difference to how you design it. If the goal is small scale combat then you can design it for some level of player skill based activity and probably active projectiles when it is finished. If your goal is EVE like epic battles then likely not a chance of any player skills involved other than trained skills - lock and fire would likely be it . 

Link to comment
Share on other sites

9 hours ago, Cheith said:

I guess my question is what your acceptable level of scale is - is it EVE scale with 1000 vs 1000?

 

In both my examples of games you can achieve 500v500 I think... and Infinity Battlescape can do 1000v1000 I believe.. it's been a while so I may be wrong on exact numbers. Nonetheless both of the games I talked about achieve "massive" battles with projectile based, fps combat. I do think NQ is in for a ride with their voxel system though, the lack of optimization is staggering in pvp with all the crazy voxel update requests and such. It's a huge problem. Like @XKentX said, even lock and fire 5v5 right now is impossible in DU, and what's worse is when it starts lagging it also affects players from across the system with pending ops... I mean.. there's loads and loads of problems, you can crash the server by warping 20 or so docked constructs at the same time... or was it 16?.... been a while :).

I believe @XKentX is right in saying that NQ devs will need to rewrite the whole system in order to even support good sized pvp battles with lock and fire... NQ does advertise massive pvp battles, which is false... not only because we don't have enough players to actually "enjoy" this pvp lock and fire system, but also because well... servers can't even handle a 5v5.

 

We can also talk about lock and fire and how to improve this system instead on focusing on something NQ will be very unlikely to even consider, as seen here:
723937932_unknown(7).thumb.png.ea8675e73a59bf3a8196a4f7bf7c0812.png

 

They will VERY LIKELY stick to lock and fire due to server constraints (talent constraints imho :) ). They will very likely stick to long distance combat, otherwise they will have to solve the slow server ticks at distances where you can see the opposing construct. It's just hard to even talk about lock and fire in a positive way when there are so many massive problems with it right now.

Link to comment
Share on other sites

58 minutes ago, IvanGrozniy said:

 

In both my examples of games you can achieve 500v500 I think... and Infinity Battlescape can do 1000v1000 I believe.. it's been a while so I may be wrong on exact numbers. Nonetheless both of the games I talked about achieve "massive" battles with projectile based, fps combat. I do think NQ is in for a ride with their voxel system though, the lack of optimization is staggering in pvp with all the crazy voxel update requests and such. It's a huge problem. Like @XKentX said, even lock and fire 5v5 right now is impossible in DU, and what's worse is when it starts lagging it also affects players from across the system with pending ops... I mean.. there's loads and loads of problems, you can crash the server by warping 20 or so docked constructs at the same time... or was it 16?.... been a while :).

I believe @XKentX is right in saying that NQ devs will need to rewrite the whole system in order to even support good sized pvp battles with lock and fire... NQ does advertise massive pvp battles, which is false... not only because we don't have enough players to actually "enjoy" this pvp lock and fire system, but also because well... servers can't even handle a 5v5.

 

We can also talk about lock and fire and how to improve this system instead on focusing on something NQ will be very unlikely to even consider, as seen here:
723937932_unknown(7).thumb.png.ea8675e73a59bf3a8196a4f7bf7c0812.png

 

They will VERY LIKELY stick to lock and fire due to server constraints (talent constraints imho :) ). They will very likely stick to long distance combat, otherwise they will have to solve the slow server ticks at distances where you can see the opposing construct. It's just hard to even talk about lock and fire in a positive way when there are so many massive problems with it right now.

Interesting - well I'm not surprised. Voxels are problematic and if you are going to treat each voxel individually in a battle then you have what you have. I suspect none of the other games have that particular issue. If all you are doing is broad surface calculations it is a lot easier and the rendering is a lot easier too so both the server and client load is significantly less.

 

Glad I am not trying to solve this - it is certainly not going to be easy and, to be honest, I suspect you would have to ditch the whole voxel aspect of it to get acceptable performance in a large battle unless it was all done by radar. Most computers would struggle to render a large voxel battle, especially if it was the closer range battles people are asking for - assuming you could even get the data there fast enough and never mind the updates.

 

I am wondering if the voxels actually need to be 'welded' into groups for this kind of activity so that they are no longer individual small entities but much larger single entities. Then you could do the more normal surface kind of damage calculations. Basically normalize the voxel surface. Anyway totally speculating. Been interesting and we'll certainly see what comes out of this.

Link to comment
Share on other sites

1 hour ago, Cheith said:

Glad I am not trying to solve this - it is certainly not going to be easy and, to be honest, I suspect you would have to ditch the whole voxel aspect of it to get acceptable performance in a large battle unless it was all done by radar. Most computers would struggle to render a large voxel battle, especially if it was the closer range battles people are asking for - assuming you could even get the data there fast enough and never mind the updates.

 

This... although there are ways to offload the whole thing to the gpu, not really sure how they do it under the hood but you can do some powerful magic on a gpu... but there are plenty of reports of people maxing out their gpus with no one around and no constructs... and in space. Sounds like a lot more problems happening than just voxels getting in the way of better performance. The other issue is the shear amount of data being requested to load all those voxels in. SB also has these problems so they deal with it by staggering viewing distance. it's not working great at the moment but they're working on it.

Lol, I'd be happy to do away with voxel systems entirely and just "build" with prefab components that are designed for sci-fi immersion instead of trying to accouunt for every single voxel.... say, you need a corridor in your ship, well, there can be prefab designs you can place that look really good... but that requires lots of artwork on NQ's part...speaking of which... they did have a fancy corridor thing in early alpha videos but it just disappeared...

 

corridor.jpg.0833b1344e35376d13e5ebc668a800d7.jpg

Link to comment
Share on other sites

1 hour ago, IvanGrozniy said:

 

This... although there are ways to offload the whole thing to the gpu, not really sure how they do it under the hood but you can do some powerful magic on a gpu... but there are plenty of reports of people maxing out their gpus with no one around and no constructs... and in space. Sounds like a lot more problems happening than just voxels getting in the way of better performance. The other issue is the shear amount of data being requested to load all those voxels in. SB also has these problems so they deal with it by staggering viewing distance. it's not working great at the moment but they're working on it.

Lol, I'd be happy to do away with voxel systems entirely and just "build" with prefab components that are designed for sci-fi immersion instead of trying to accouunt for every single voxel.... say, you need a corridor in your ship, well, there can be prefab designs you can place that look really good... but that requires lots of artwork on NQ's part...speaking of which... they did have a fancy corridor thing in early alpha videos but it just disappeared...

 

corridor.jpg.0833b1344e35376d13e5ebc668a800d7.jpg

Nice - but the issue there is (as you probably know) that you then need to hire the graphics folks to work on that - and to do a good selection of assets is non-trivial. One of the reasons (imo) people go for player made content is that it is cheaper to build from a graphics perspective and in some ways you actually get a better variety of outcomes. Players also have different expectations for the look of ships they have made themselves as opposed to the supplied variety. I might just be overly cynical here, but I suspect not.

 

You are probably also right on the 'other issues' - as I have said ad-nauseum building this type of game is definitely non-trivial on the server side of things. I work on moderately high throughput distributed server systems (5-10,000 events/second or so) so I have some idea of what they are dealing with but I think this, in real time, is just another challenge. Figuring out what the acceptable trade-offs are must be a nightmare in itself and then there are the whole host of network unknowns to deal with - especially for lag. Is the server bottlenecking, is it too much data for the client, did one if the ISPs get slow, is Level 3 having another bad day? Too many variables!!!

 

 

Link to comment
Share on other sites

I don’t think the economics are there for PvP.

 

You can log into EvE and find fun builds that take an hour or two of play time to set up (hell frigate fights are probably the most fun and diverse PvP meta and those can be 5-20 mins of game time to build).  then if you need to go higher the civilization side of the game has ship replacement programs, along with “income multipliers” for participants of the defense outside of PvP (ranging from intel watch channels, to PvP areas more “secure” than high Sec, and defense response fleets).

 

something like Mortal Online or Albion Online have gear set up so “fun” gear can be made in under an hour, along with the skill system set up so people “spam” gear for leveling/training.  This means that often times gear is “cheaper” to buy, than the time value is to make.

 

————

Frankly this game utterly fails on making gear cheap, and only somewhat grasps the concept of “civilization multipliers”.

 

even if PvP gets rammed in most of the community will just sad out from PvP pregrind, and the rest of the community is likely to follow.   PvP is also lame as all hell in this game you have EvE style targeting, with fewer guns/utility modules, and don’t get the ability to move yourself.   Or you get to move yourself but get maybe a single module to play with.  (Lame to play, lame to stream, but great for getting drunk with some buds I guess).

 

The worst part is that the devs are literally fighting in the wrong direction for fixing the game, by increasing grinds and similar.

Link to comment
Share on other sites

4 hours ago, HangerHangar said:

I don’t think the economics are there for PvP.

 

You can log into EvE and find fun builds that take an hour or two of play time to set up (hell frigate fights are probably the most fun and diverse PvP meta and those can be 5-20 mins of game time to build).  then if you need to go higher the civilization side of the game has ship replacement programs, along with “income multipliers” for participants of the defense outside of PvP (ranging from intel watch channels, to PvP areas more “secure” than high Sec, and defense response fleets).

 

something like Mortal Online or Albion Online have gear set up so “fun” gear can be made in under an hour, along with the skill system set up so people “spam” gear for leveling/training.  This means that often times gear is “cheaper” to buy, than the time value is to make.

 

————

Frankly this game utterly fails on making gear cheap, and only somewhat grasps the concept of “civilization multipliers”.

 

even if PvP gets rammed in most of the community will just sad out from PvP pregrind, and the rest of the community is likely to follow.   PvP is also lame as all hell in this game you have EvE style targeting, with fewer guns/utility modules, and don’t get the ability to move yourself.   Or you get to move yourself but get maybe a single module to play with.  (Lame to play, lame to stream, but great for getting drunk with some buds I guess).

 

The worst part is that the devs are literally fighting in the wrong direction for fixing the game, by increasing grinds and similar.

EVE also has 'rats' to fund your PvP with - and in nullsec the rats can be pretty tasty.

Link to comment
Share on other sites

19 hours ago, Cheith said:

Interesting - well I'm not surprised. Voxels are problematic and if you are going to treat each voxel individually in a battle then you have what you have. I suspect none of the other games have that particular issue. If all you are doing is broad surface calculations it is a lot easier and the rendering is a lot easier too so both the server and client load is significantly less.

 

Glad I am not trying to solve this - it is certainly not going to be easy and, to be honest, I suspect you would have to ditch the whole voxel aspect of it to get acceptable performance in a large battle unless it was all done by radar.

 

DU doesn't render constructs as voxels. 

 

Rather, it converts voxels to meshes server-side. That's one of the most fundamental and complex parts of their tech stack. 

 

That's one reason why constructs pop-in at the last second instead of loading progressively like a voxel structure could -- it has to jam the whole mesh into the GPU before it renders. 

 

So...the performance issues we see with ship battles are not likely voxel-related, at least not directly. I'm sure that their voxel-to-mesh algorithm isn't exactly the most effective at generating efficient meshes and obviously their LOD system needs a lot of work. 

Link to comment
Share on other sites

7 minutes ago, ShippyLongstalking said:

 

DU doesn't render constructs as voxels. 

 

Rather, it converts voxels to meshes server-side. That's one of the most fundamental and complex parts of their tech stack. 

 

That's one reason why constructs pop-in at the last second instead of loading progressively like a voxel structure could -- it has to jam the whole mesh into the GPU before it renders. 

 

So...the performance issues we see with ship battles are not likely voxel-related, at least not directly. I'm sure that their voxel-to-mesh algorithm isn't exactly the most effective at generating efficient meshes and obviously their LOD system needs a lot of work. 

Interesting - do they handle damage calculations the same way? If not then it is still a lot more work and then you have multiple layers of voxels. Obviously as one can't see the code only guessing!!

Link to comment
Share on other sites

  • 2 weeks later...

Some other stupid stuff that I think affect performance:

 

1) Constant "block" recalculation.

Did you have that your engines are OK and then, midflight, the games decided they are "blocked"... WTF. It seems it periodically recalculates the shape of your ship even that it never changes and the algorithm is not deterministic or something.... Why don't we jet get a button like "Prepare dynamics" that you should click 1 time after changing anything in build mode that will do the calculation and keep that until you change anything in your ship. Then you need to press it again before flying.

Getting your stuff blocked midflight + recalculating it again and again when nothing changes is bad both performance wise and gameplay wise.

 

2) Damage and voxel banding

It's cool that you can see a big circle and voxels have an actual hole but that what prevents any PVP from actually happening.

If 5 people, 6 lasers each shoot on a target (target is M-core silicon cube with 1 person on board) the guy on the target cuble will download at least 100mbit/second of the "changes" to his ship.... We call it "download your damage" lol. When you are shot at in DU you are literally DOSed. 

This is bizzare. It can't work. NQ should compromise here, make the voxels not "bend" etc but just the damaged part like change color or texture becoming less visible so it's clear that it can be shot thru.... Or maybe just part of voxels disappear without changing shape of the others(not sure if possible). It can be banded together with that "calculate dynamics" button, once you click it, your voxels are "frozen" in shape and damage will only alter colors and stuff not the shape. (As shape changes is what I think requires lots of complex stuff).

No one wants all his ship voxes bent just because of 1 shot anyway it pain in the back to repair manually and no. I don't put my L core to 24h repair because of 1-2 shots.

 

Don't know why I still post here, I don't play since like 2 months ago but don't know. Maybe they will improve.

Link to comment
Share on other sites

I really want to like PvP in DU but at this point I dont think anything they do will really ever make it fun enough to want to throw away millions of credits to even enguage in it or that there is no ground game or even hand to hand combat to be had that would ever make this fun in any capacity.

 

!: Scripting

The first and biggest gripe I have with it in general is that NQ needs to either disable PvP scripting as its basically devolved into aimbot hacks to where you can target people from much further than your radar should be able to see the target, tell if its a friendly org member or foe, pull a full health/diagnostic report of the ship, match speed and course correction, and just autofire faster than people should be able to rail things in space.

 

PvP scripts should either be disabled or be required to be submitted for review and put into a repository that everyone can buy with credits in-game to even begin to make the PvP in this game fun. Hell NQ can even charge for submissions regardless of approval and if they are good enough attach them to actual parts rather than having the game within a game mechanic of secret personal guarded scripts that will ruin any sort of progress the developers make to try to improve DU PvP.

 

2: Target Lock

The second biggest gripe I have is with how target lock is handled in general and the overall tab target RNG feel it has which requires no skill or longevity to combat that this system provides. It is akin to modern day fighter jets using radar and other systems to spot enemy jets long before yu ever see them and launching missles and destroying them before any sort of real combat can be had or any chance that you can turn the tables after somebody tries to jump you because there are probably already multiple vollies coming your way. In DU its is pretty much full auto scripting with no chance on the recieving end. Its like the Aegis drones to where they can aquire targets before taking off, go to the location, destroy the enemy, fly bakc, land, and park without ever needing a human to do anything since it does everything itself with perfect precision.

 

There is a reason nobody likes PvP except for those who have these scripts and using the target lock to their advantage since again they are pretty much hacking the code through scripts and even if they arent target lock still destroys the fun of PvP since it requires no real skill to use or any form of line of sight or actually maintaining the lock to keep the missles going towards their target. There is also a reason EPs like Star Wars have target lock but it still maintains a sense of WW1/2 dog fighting nostalgia so it appears that the heroes or villians have both skill and a chance of survival unless you have ships behind you and still they rarely use missles/torpedos instead of just using the lazer cannons since it would be stupid if the epire rolled in with and destoyed the rebels with the tools commonly used in DU with target locks and scripting.

 

The only way to fix this is to make the nature of weapons actually require skill like everything else in this game requires and disabling the tab target RNG part lock system and require the attacker to be in visual range to actually both fire the weapons but also to maintain a visual lock for missles and moving to a system that relies on more lasers or physical ammunition to actually fight with instead of the garbage going on and needing to lead shots with lazers or to maintain your aim on a target for it to go to where you are aiming and have the possibility to miss if you lack skill.

 

3: Sliding

It does not matter if you are in space or in the atmosphere there is a slide when you change direction. In the atmosphere the stabilizers are meant to keep your ship from sliding and is meant to act like the fletchings of an arrow in order to maintain a true trajectory. In DU however no matter what you do if you dont start off with a perfect course you will inevitably slide off course and unless you come to a complete stop you can never truely fly in a true straight direction. Since there is no wind in DU its like always sliding on ice and it is no different in space even with the space adjustors or atmo adjustors which should stop this. It is even worse in space because you cannot ever relly correct the trajectory without coming to a full stop or you will slide many SU off course.

 

This ruins any chance for combat in space since you either are forced to use an autopilot script or the PvP variant and it turns it into more of a drag race than allowing any kind of space combat dog fighting from occuring ever until this either is taken out oof the game or somehow lessened to allow for some slide more like a rally car would handle on snow or drifting but still allowing people to maneuver in atmo/space to open up the door for evasive maneuvers based on the mass of the ship.

 

4: Speed

Speed is very important if this drag race type system is to stay. Its like your ship could go much faster but there is a governor placed on your speed for no real reason but allowing gankers to catch up even if you leave the bubbles at max speed you still cant go faster than 29999 which gives other place the ability to jump and with the rediculous range scripts or target lock has and the equally rediculous range on missles which are not bound to the speed cap just blow you out of the sky.

 

The first thing that needs to change is the cap and you should be able to achieve warp speeds if you are consistently using the space engines as well as even if you let off the gas you should still maintain your desired speed without slowing down unless you hit the space brakes to slow down even if yur ship is blown to bits it should still slide and keep drifting through space and require the enemy player to have to match your speed and board your ship.

 

But also that there shuld also be different warp tiers which go faster than the initial warp spped and rangin from Warp 1-5 depending on if it is T1-5 or even more with basic-rare. It should not also be an instant o-shit button either as much as just a faster way to move and if space engines could be used on top of it to gain more speed if the cap was abolished it should allow players or other PvPers to catch up to you based on the masses of both ships to make warp speed battles a thing rather than just ripping people out of warp which keeps getting proposed and rather requiring the attacker to actually desroty the warp drive to disable their warp.

 

The other thing that needs to happen is the ability through talents or part variants to have atmo burn resistance in the atmosphere to make it so that you can go faster than 1200kmph and open up mach speeds based on aerodynamics rather than some artifical hard cap of planetary speed. Its not like planes burst into flames doing this in RL and it seems pretty stupid that you cant take full advantage of speeds in DU if there were parts that added heat resistance through parts/talents.

 

5: Turrets/AI based Turrets

The lack of any kind of turrets that players or AI can cantrol which arent fixed is appauling. Even AA type turrets akin to the AI based ones you encounter around the large TUs or the giant planet teraforming Tus that actually do something or that you as a player can hop into like the turrets in WW2 bombers is rediculous or the fact that even if you put them on a ground based construct NQ has to know that you will never get someone to man it 24/7 and some kind of AI is required to make it possible to defend yourself when you are offline.

 

This goes doubly so for PvP in atmo/space in a dynamic core based ship in the fact that haulers/freighters need heavy defenses even if they are AI driver auto turrets that arent as accurate as a player.

 

6: Power/Overheating

The lack of power to run heavy weapons without a downside to just constantly being able to fire with no charge time or power consumption to do more damage or that if your power core or overheated other systems go offline or require different systems to be turned off to trade off on attack/defense/radar/detection/stealth meta and just all system go all the time needs to change. There should be some form of balance between Attack, balance, or defense capabilities and what you can reliably take onboard or power or trade off lazer or energy weapons with physical machineguns, cannons, or missles and having things like shields or warp drives rely heavily on power or overheating based on usage or duration of usage to where you need energy based weapons to creack the power keeping a shield up before physical weapons can be used and resistances added for various weapons to parts which might require multiple weapons to take out a ship or that there could be different types of propulsion which again require different weapons to disable before destroying that do not require fossil fuels.

 

7: Nature of Weapons vs Voxels

The weapons themselves seem way overkill for the sizes they can destroy in terms of the voxel damage they can cause or the splash damage they do. Something like a simple laser cannon or even a physical machine gun or cannon should start out with XS size being 1vx in size with L sizes doing 5vx worth of damage to whatever it hits with no splash damage. Beam weapons on the other hand or rail guns should go through multiple targets with beams acting like metal cutting lazers in RL to where it needs a sustained direct stead beam to cut through metal vs a rail gun which should do some DD damage but go through everything in the ship and do less and less damage with everything it goes through with the 1-5vx size from XS-L weapons and again requiring tremendous amounts of power to opperate and requireing either an equal dynamic core size or clever engineering and placement of parts to accomplish the right to use it to the mass/power/thrust/lift to make it work. Where as machineguns/cannons/missle should actu in the same way as well to where machineguns or gatling guns should put out a lot of fire but do little damage to pepper a ship doing 1vx size with XS-L putting out more shots but it ships away doing basically DOT based damage with sustained fire and missles doing heacy DD damage on the initial hit and then with the splash damage doing much much less damage to surrounding parts in a cone rather than a full AoE.

 

 

 

If NQ keeps doing more of the same without meaningful changes before TU warfare comes in without ground weapons, hand held weapons, or any kind of reason to fight in the first place it will never be worth the development time. If it is allowed to be fully automated via scripting no matter what they do its still DOA. Its kinda a lost cause at this point anyhow.

 

 

 

 

Link to comment
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
×
×
  • Create New...