The finale is here and my last post for the wish list will be the last of the year. And it will be the bestest most amazingest post EVAH! !. Are you ready?
oh and sorry for the huge amount of word spam you are about to be subject to which I thought would be wise to contain withing the bowels of the blog.. you know so the crazy obsessiveness isn’t all out and visible to the world
Server Mainframes and Client Coding
or in basic terms I’m talking about culling and the what the fuck just killed me phenomenon.
This is a very serious issue for me but I am very much bias towards it as most of my play time for GW2 is spent in WvW. I am not going to pretend it is the most important and biggest issue EVA fix or I’ll quit type thing because it’s not. Not to most people anyway as they are either happy enough with what we have or have other pressing issues that align with their own interests.
The thing with it is while not entirely game breaking (although it is to an extent) it kind of ruins my enjoyment often enough that it is a problem. It breaks the feel and how you embrace combat. I and many others were after that epic feel of large scale warfare, masses of people and pure carnage but that just isn’t happening right now. In a big enough fight it only ends up being a bubble of combat surrounding you and the rest of the world need not exist. That bubble is relatively small as well meaning I can interfere with things I can’t see and others I can’t see can interfere with me. One of the key principles in combat is to know your enemy so as to make decisions, to enact strategies, and to adapt to a situation and right now this is being greatly hindered .
They recently said in the WvW dev video that they wanted to focus on the mechanics above all else and get the base game to where they want to be. And then they go and focus on progression systems when there are far more pressing issues. Culling is the pressing issue here but there are many factors in play, many which I don’t even understand or even know about that are adding up to the issue
Right now the belief is that the server determines which entities your client can see, and sends you information on what those entities are doing and where they are. It is a flow of information that continues to perform many calculations with changes be constantly updated. Most of everything is done on the Arena net servers with the rendering and such being up to the user. Now culling may not make much sense but there are a few reasons for it.
- To limit the amount of data being sent to and from the servers and to and from the client. Bandwidth can be expensive for both parties.
- More efficient by not sending inconsequential information
- Not as much stress placed on the client in terms of rendering. Slideshow FPS anyone?
Now in terms of the amount of data being sent the actual numbers and size of the packets is rather contested. It is understand that the amount of data that needs to be sent is multiplicative based on the amount of people in the immediate area which could add up to quite a bit and if too much could place require a rather large bandwidth to deal with, however most seem to believe that the amount of data contained needed to display characters and their actions is relatively miniscule. However even if it was a considerable load on the servers the argument isn’t really apt, one comment on the forums explained it quite well
if you have an high end computer, you can currently see players even beyond both the targeting limit, and the LOD Draw limit (where 3D models change to ‘art’). The client already receives data from beyond the farthest draw distance the options allow.
If this is the case then most of the needed information is about everyone in the area is already being sent and received by the client, anything else needed would just be collision mechanics and other trivial things which leaves the load time being mainly based on rendering
If it is a rendering issue then there are a few things that may be able to be done to allow the Arenanet to lessen the restrictions. Right now there is a lot of detail going into loading each individual character in terms of game assets, character customisation options are all included which involves the race, gender, facial and body options, and armour which seems like a waste to me. While it is a nice touch to show off the true appearance of the enemy and allies it takes extra processing power to do when in all honesty I would rather see more people but less detailed. Having basic models for the 3 different teams maybe being based only on the basics of race and gender with a generic armor model would be a better option.
The restriction also seems based on Arenanets wish to have the game as compatible with as many users as possible and with very cheap minimal specifications. While a noble gesture it has made a few issues for those with better systems and with the general state of the game. While they have optimised it for the lower end not as much work has gone into improving performance for those with better machines, there is only so much you can do do with the limitations of the lower end without impacting others. You can see this during gameplay when you lower the performance options as you really don’t gain that much of a boost to performance and funnily enough it still seems to use only slightly less CPU and GPU resources. There is pretty much a hardcoded bottleneck placed on the client, an obvious change is probably just giving the user the options needed so they can take better control of how many people they want to render and how much stress they want to put on their computer.
But I don’t believe it is entirely a restriction placed on solely based on the users hardware either as there are a few niggling parts. Firstly the engine seems to take a large amount of time to fully render those in the characters immediate area regardless of culling restrictions, even when it is trying only to render a limited number the amount of people in the area it takes a certain amount of time with this getting exponentially worse with the amount of people. Most games I’ve played had some sort of default model that loaded while client side information was being processed or while more information is being sent which lets people to react, models here don’t load until the client has received all the relevant information and has been completely processed which is far too long during combat.
The bad part though is that this is even happening when there are very little people in the vicinity when server and client side resources needed is at a minimum. The Thief clipping issue is a perfect example of this although the issue is very visble during any sort of stealthing action as the amount of time it takes to reload is seconds which can happen regardless of the situation.
The issue of excessive loading of character and NPC models is not just an issue during WvW with lots of players but during small scale battles, throughout the PvE world during events and just general roaming, and even during SPvP matches. It seems then that it is a fundamental problem with either the client coding, the server mainframes or both. Guild Wars 2 uses a redesigned version of the GW1 engine which really had a very different design approach in mind. The approach they had there was to separate players into much smaller groups through instancing.. it was an engine designed without the need to code and render large quantities of players. And this engine base is what we have now.
I don’t think the Arenanet team have had much practice working with creating larger virtual worlds either and were not as confident in their abilities here, their use of overflows and such is part of this as it was not just a way to improve the dynamic event experience but to limit the pressure placed on the servers, engine and client that is caused by large groups of players. They were terrible unprepared, WvW is one part but we can see how blinded they were to the issues when the karka event came around and they funneled the playerbase into Lion’s Arch… huge problems and they even had to change how they now plan events based on the outcry. But they can’t stop groups from coming together in Wuv Wuv which is now is kind of a baptism by fire, and if it is ever going to be anything more then some casual fun and achieve there original dream then they must come to face with the restrictions and limitations they have set and are entrenched in the system.
In this day and age what we have just isn’t good enough, there have been dramatic strides in technology with improved server and computer hardware as well as internet networks. Now I don’t expect a linear correlation between hardware/ bandwidth and client performance, that is unfounded but I still expect more then we are getting. Technology is booming ahead and it seems companies are only trying to use the newer tech instead of mastering the old.
Older games had better systems for displaying many models then we have today and were able to make larger scale combat, since then we haven’t really seen much improvement in the field except in small circumstances.. in fact in many cases like in Guild wars 2 it is actually worse. It doesn’t make much sense. Now some people believe it is the pressure placed on the server by the combat system which involves projectiles and collision mechanics but action games have been doing this for a long time now. Guild Wars is also rather detailed but then you have games like Planetside 2 which have gorgeous worlds, collisions and projectile mechanics and have huge scale warfare including many individuals as well as tanks and planes being constantly visible in the immediate and surrounding space. When comparing what we have to games around us the implementation we have doesn’t make any sense.
Arenanet really needs to step it up here and start competing in a field that right now is doing many things better then them, maybe they were scared by the Ilum failure and didn’t want to see the fallout in their own game but this fix is just as bad. I understand they have different teams for certain updates and such but culling I believe could be and should be a group effort in order to fix this sooner. Better character models with less polygons and such that can render faster, easier and produce less stress on bandwidth, design basic models that pre-load when needed, better coding of client information and packet compression, bug squashing, engine optimisation, toning down aoe, and better server mainframes are all options that may improve the situation.. there is lots that can be done but considering the amount of effort required to make it work I don’t see a fix in the near future.
Cost Benefit Analysis
One part holding this back is I don’t think WvW makes much money compared to other areas of the game, server maintenance and bandwidth is expensive so the fault may also be from the new payment model. GW2 was mostly designed as pick up and play with a relatively small to moderate continuous player base and burst play surrounding events. WvW however is a constant drain on resources and while people will come and go the player base is more consistent in their play. WvW has a constant cost but no real consumerist way to support it, while it is as connected to the shop as the PvE side it isn’t as ingrained in that playstyle to purchase and use items.
If this is the case though it is hard for a company to justify spending money on areas that are a loss which may be why we have seen less updates on the area. It is a huge draw for people to the game but maybe enough people aren’t transferring to other parts of the game and spending to support it. In terms of the payment model as well for those that focus on WvW there really isn’t much reason to buy expansions either which is another big part of supporting the game and the company. So throwing all your resources into a part of a game that isn’t going to translate into a worthy return is just not good business which is why we get a small group working on a huge issue that encompasses the games engine, the clients abilities, and the servers capabilities. And why we, after many months now are still stuck with a pressing issue and no fix in sight except for bandaid updates (that get reverted).
Now most of what I have said here is subjective rambling, (especially this end part) and really isn’t based on any objective facts so take it with less then a grain of salt., a pixelated grain of salt To be honest I have no idea what goes into running an mmo, how a server performs, and how there servers interplay with our own hardware… it is a mystery to me and that’s kind of the way I like it. None of this stops the thought though that something is terrible wrong with the game and it’s ability to properly render characters that severely impacts gameplay. The amount of times I have come across large groups I never saw is numerous and that needs to change. I want to see the people around me, I want to witness the large scale combat we were promised and I want that epic feeling of two large forces crashing into each other and watching the carnage unfold.
Now these aren’t the only things I would like, there are many parts of the game I would love to change around based on my interests but this is the main one I want and feel that is needed NOAW, NOAW, NAOW. I didn’t really come across this list very easily as there is so much I would like in GW2 to make it the game of my dreams, or close enough that I would be happy spending exorbitant amounts of time in during this brand New Year. I came up with quite a few little points, some others though that are worth mentioning :
- housing – personal housing or large guild buildings that have a certain amount of customisability and lots of collectables
- group Functioning - Better commander tools, Ability to find groups easier without having to resort to third party sites, and maybe custom chat channels
- Dueling Arena – the Arena in the Charr area would be absolutely perfect for this. Picture a gladiator arena were matches go on at the bottom with 1v1, guild battles, and even last man standing battle royale’s. It would be glorious and those above can even watch the matches unfold and even bet on them.
- A Real Story – because the one we got is terrible, some races are better then others but still…
- SPvP crossover – Giving people Exp on their characters and having gear transferable between game types
- Meaningful Crafting and econmy – but that is so beyond the bounds of the game now
- WvW Progression
- Balance – whatever that means anyway
And what would your top 5 changes, fixes and new mechanics be for Guild Wars 2?
Well that’s all folks.. for now anyway. Thank to all those who have stopped by to read my unintelligent rambles and see you in the New Year