| Pages: [1] :: one page |
| Author |
Thread Statistics | Show CCP posts - 0 post(s) |
Jehovah Cooper
H A V O C
The Requiem
   |
Posted - 2008.07.24 05:40:00 -
[1]
Maybe this fits in another topic; I'm sure its something the devs have considered but its a lot of work of course, and I don't want to argue about the overall topic (lag) but just put this idea out there and hopefully hear if anything like this is in the works.
Basically the idea is to have graceful degradation in game functionality as node capacity is reached. Turn off non-essential features could be one thing (rather than just making them so slow you want to die).
But more importantly have tiered combat grid approaches. So for example when it starts to get "too laggy" the server stops bother to track missiles or drones and simply lets them do some average damage to targets...it stops bothering with calculating transversal and other things that require lots computations that grow exponentially as the number of objects increase. It stops sending the client any information not on the brackets. It dumbs down combat and makes it less real progressively as the server reaches each new threshhold limit. At some point maybe its 1000 v 1000 its basically turned into more of a turn-based strategy game with little or no real-time elements and the clients all make this obvious to the end-user . While its not as satisifying as a miraculous and infinite amount of processing power would be it at least would not be infinitely frustrating and soul-crushing the way horrible lag and node crashes are.
Of course smaller engagements all work as they do today - its about taking the engagements that simply don't work at all today and letting them work a little bit. Like I said I know this would be very time consuming but could really lead to increased enjoyment and is a strategy that would not be odds with continuing to improve performance at the hardware & software levels b/c as things improve we just have to do less and less of the "dumbed down" combat but you will always reach a threshold no matter how good it gets so I think something like this is needed so that the breaking point isn't so ugly and frustrating. Anyway enough rambling, I don't want to argue about any of this w/ anyone just I want to know that a dev has thought of it and its maybe something being considered or talked about.
|
Sir Substance
Minmatar
Sunspot Requisitions
Worlds End Consortium
|
Posted - 2008.07.24 08:09:00 -
[2]
idea needs development. but as a solution to fleet lag, it shows promise.
 Originally by: Ekrid
because trying to fit medium beam lasers is like trying to shove a cow in your mailbox. doable, but it gets messy.
|
Astria Tiphareth
Caldari
24th Imperial Crusade
|
Posted - 2008.07.24 11:27:00 -
[3]
Actually this sounds both quite good and an incredible pain. Degradation of key things like transversal and collision detection would discourage blobbing, because quite a few key areas of EVE combat rely on these. Bumping would become impossible. Speed-tanking would cease to be effective. Drones doing average damage would negate a lot of the specific faction-fit stuff on carriers etc.
Probably needs work but yes, not bad at all as a notion rather than just dying due to lag.
___
My views may not represent those of my corporation or alliance, which is why I never get invited to those diplomatic parties... |
Sir Substance
Minmatar
Sunspot Requisitions
Worlds End Consortium
|
Posted - 2008.07.25 01:59:00 -
[4]
im gonna go ahead and bump this back to the top page.
Originally by: Ekrid
because trying to fit medium beam lasers is like trying to shove a cow in your mailbox. doable, but it gets messy.
|
Jehovah Cooper
H A V O C
The Requiem
|
Posted - 2008.07.25 02:46:00 -
[5]
Thanks for the feedback and comments.
I think you raise an interesting point, that the changes might favor certain types of ships and setups and it could steer people towards that. We already have that in the sense that large fleets look for different compositions than small ones, but this could take it to a new level and it would be a bit unpredictable what all the outcomes might be. But you could only game it so much because node stress and lag are somewhat unpredictable too and you couldn't be sure what the conditions would be like for your fight. The ideal would be to keep ships reasonably balance through each tier of simplification but that wouldn't be entirely possible. Still I think that any sort of working game - even rock-paper-scissors - beats a lagfest.
|
Sir Substance
Minmatar
Sunspot Requisitions
Worlds End Consortium
|
Posted - 2008.07.25 03:10:00 -
[6]
i agree. the reality is that when you have 100 ships a side on the grid, you are in for a broken experience anyway. better to make this predictably broken rather then random. certainly it will unbalance stuff in the process, but thats part of the brokenness, and i feel inbalance is better then unplayability.
Originally by: Ekrid
because trying to fit medium beam lasers is like trying to shove a cow in your mailbox. doable, but it gets messy.
|
Kitoba
Minmatar
Legion of Dynamic Discord
|
Posted - 2008.07.25 11:13:00 -
[7]
The most important term in the OP's post is "turn based strategy", ie. taking the randomness out of lag and giving everyone an equal chance to give commands. /me as a dev would definitely consider it, as long as you can approximate combat mechanics to the game designer's content it solves a hell a lot of problems with one strike.
|
Jehovah Cooper
H A V O C
The Requiem
|
Posted - 2008.07.25 22:34:00 -
[8]
Yeah we've all heard the term "slide-show" in use to describe a really laggy fight. Why not make this obvious, user-friendly, managed and fair.
|
NanDe YaNen
The Funkalistic
|
Posted - 2008.07.25 23:31:00 -
[9]
Turret amalgamation. Most people use all turrets on one ship at a time and activate all of them together. Turns eight activations into one. Eight calculations into one. Use average values for tracking, damage, instead of individual modifiers. i.e. "Your seven blah blah blah hit Skizzlefist for blah blah blah"
Reduced updates to locations for slow ships...wtf can you possibly care about targets 200km away when you move 500m/s with MWD. Who cares whether a BS is... changing.... directions... Give fast ships, who need to know these things, the tip of the hat when it comes to updating info. Let people see fast ships more clearly.
Amalgamate drone AI. Is it even possible to assign drones to individual targets? Not like I care. It's usually going to be the same as turrets. Fire. On that target. All of you.
Make the drone updates more like a fat client. I don't care where the server tells me the drone is most of the time. I just want to orbit that drone and shoot that drone. More likely, I really only care about which drone is shooting at what or headed for what. The details I could care less about. Even more likely, I wish drones were about as detailed as turret effects. I really don't need any of the other information.
Missiles, more like a fat client. I don't really care where the missile is to the server. Just tell my client to draw it flying from one thing to the thing it's headed for.
Given how many transactions per second servers can support nowadays, I suspect the code implementation is already so bad as to deserve open criticism. It's not like these calculations are that hard. Performance should scale linearly per ship. Remove anything that is n-body. Client-side and server-side. If you don't know what n-body is, hire me.
|
Ruze
Amarr
No Applicable Corporation
|
Posted - 2008.07.26 00:20:00 -
[10]
This is a very good idea. I really like the concept of a systematic degradation of a fleets 'effectiveness', the larger it gets. Maybe throw one more out there, you tell me if you think it'll help anything:
After a system has 100 players, local goes into 'delayed' mode, until the system is brought back under 100 players.
Genesis Project |
Sir Substance
Minmatar
Sunspot Requisitions
Worlds End Consortium
|
Posted - 2008.07.26 01:52:00 -
[11]
Originally by: Ruze
This is a very good idea. I really like the concept of a systematic degradation of a fleets 'effectiveness', the larger it gets.
i think possibly you might have missed the point. this is an anti-lag strategy, not an anti-blob strategy.
Originally by: Ekrid
because trying to fit medium beam lasers is like trying to shove a cow in your mailbox. doable, but it gets messy.
|
Ruze
Amarr
No Applicable Corporation
|
Posted - 2008.07.26 01:55:00 -
[12]
I think the two go hand-in-hand in a lot of circumstances. The largest issues of lag in EvE are related to large-scale fleet warfare. And if the effectiveness of large scale fleet warfare is reduced, you will be encouraging those who wish to compete at their full potential to fight in smaller groups.
I think the ops ideas are good ones, better than anything I have read or my little brain can come up with. And, I think it addresses a two-fold problem.
Genesis Project |
Astria Tiphareth
Caldari
24th Imperial Crusade
|
Posted - 2008.07.30 17:29:00 -
[13]
I'm going to bump this as well, because I think it's a fascinating notion, and even if never implemented it might prompt some sort of eureka from someone.
It also occurred to me that the notion of degrading experience has already been in games for years, in the form of level of detail. Why not take it a stage further?
___
My views may not represent those of my corporation or alliance, which is why I never get invited to those diplomatic parties... |
Sir Substance
Minmatar
Sunspot Requisitions
Worlds End Consortium
|
Posted - 2008.07.31 08:00:00 -
[14]
Originally by: Ruze
I think the two go hand-in-hand in a lot of circumstances. The largest issues of lag in EvE are related to large-scale fleet warfare. And if the effectiveness of large scale fleet warfare is reduced, you will be encouraging those who wish to compete at their full potential to fight in smaller groups.
I think the ops ideas are good ones, better than anything I have read or my little brain can come up with. And, I think it addresses a two-fold problem.
well, thats true. however, i feel if we proceed with this idea on the bases of "stop blobs" instead of "minimise lag" we will all suffer for it.
-
PvPers always say "GB2WoW". the message is that EVE is hard, and people just need to deal with it. wasn't it funny how when nano's started making it hard for *them*, that all went out the window? |
Tyson Gallane
Caldari
Political Warfare Executive
|
Posted - 2008.07.31 08:39:00 -
[15]
Say, if after a fleet reaches a certain size then my skills cease to effect play, my super BS becomes "Stock BS with X STD guns/launchers". If I want my skills to be useful then I will try to find smaller battles, thus there will be some slight effect on blobbing.
If the fleet needs me there for numbers though, then I will go along and be a (simplified) BS.
One danger is that people will start to fit very basic T1 ships, knowing their fittings are unimportant in a fleet fight.
T.
|
Jehovah Cooper
H A V O C
The Requiem
|
Posted - 2008.07.31 23:33:00 -
[16]
Originally by: Tyson Gallane
One danger is that people will start to fit very basic T1 ships, knowing their fittings are unimportant in a fleet fight.
T.
Certainly you could game it like that if you absolutely knew you'd be facing a lag-bomb. But then that ship wouldn't hold up very well in a 40 vs. 40 engagement and are you really going to have stacks of ships in your hangar based on the fleet size you expect? Maybe some would but it wouldn't be that big of a problem. Also there would still be some way to incorporate the value of your skills and fittings without doing all the math in real time - it could pre-calculate some sort of effectiveness level when you undock in the ship based on all your skills and the fittings you have on the ship. So that my T2-fitted megathron pilot with 10m SP in gunnery gets a much higher multiplier to his effectiveness than some nub that can barely undock with a t1 fit.
Great ideas people, keep them coming.
|
Xindi Kraid
|
Posted - 2008.07.31 23:58:00 -
[17]
I like the idea for this.
One thing to note is that alot of the lag people is client side rendering lag where you get frame dropping since you video card can't render ll the objects on screen fast enough. Because of this the server simplifecation should also come with coding to gradually reduce the visual detail as more stuff happens on screen in order to avoid graphics lag.
That way when theres only a few ships you see them in thei shiney shiney majesty and when theres a few hudred firing at each other with drones out its reduced to just genral shapes but you can still tell whats going on
--Bird of Prey: Forum God
Caveat Emptor
Caveat Venditor
CAVEAT |
Sir Substance
Minmatar
Sunspot Requisitions
Worlds End Consortium
|
Posted - 2008.08.01 03:55:00 -
[18]
Originally by: Xindi Kraid
I like the idea for this.
One thing to note is that alot of the lag people is client side rendering lag where you get frame dropping since you video card can't render ll the objects on screen fast enough. Because of this the server simplifecation should also come with coding to gradually reduce the visual detail as more stuff happens on screen in order to avoid graphics lag.
That way when theres only a few ships you see them in thei shiney shiney majesty and when theres a few hudred firing at each other with drones out its reduced to just genral shapes but you can still tell whats going on
that could be a tricky one. the textures and meshes are loaded into memory, it would generally require a restart to down or up grade them. but, i would approve of simply removing the textures at higher lag levels. everyone's ship would probably look grey at that point, but hey.
-
PvPers always say "GB2WoW". the message is that EVE is hard, and people just need to deal with it. wasn't it funny how when nano's started making it hard for *them*, that all went out the window? |
Spurty
Caldari
Trader's Academy
Blue Sky Consortium
|
Posted - 2008.08.01 04:18:00 -
[19]
I like this idea although it will make combat a bit unpredictable. For example one of the easiest ways to lag out a fleet jumping into your camp is to release drones.
Making people's drones return to drone bay and be unlaunchable would be a good plan for reducing lag, but make combat a little unfair for the old kings of pvp, gallente
--
Two prostitutes standing on a street corner. One says to the other, "Have you ever been picked up by the fuzz?" The other replies, "No, but I've been swung around by the ****!" |
Jehovah Cooper
H A V O C
The Requiem
|
Posted - 2008.08.01 12:10:00 -
[20]
Client lag isn't a very big problem and there are already lots of optimizations you can do to improve it; the topic here is specifically server-lag and node death. When you click a module and it doesn't activate for 5 minutes that is server-lag. When you crash and then can't login to a system for 45 minutes its again server lag. When the node crashes and all the POS you spent 10 hours kiting are now back to full health that is again server-side issues.
|
Jehovah Cooper
H A V O C
The Requiem
|
Posted - 2008.08.01 12:12:00 -
[21]
Originally by: Spurty
Making people's drones return to drone bay and be unlaunchable would be a good plan for reducing lag, but make combat a little unfair for the old kings of pvp, gallente
The idea is more, stop doing all the things with drones that cause lag. Let them still have some effectiveness but don't calculate it all in real-time. Just taking away drones is a huge nerf to certain ships.
|
| |
|
| Pages: [1] :: one page |
| First page | Previous page | Next page | Last page |