WvW culling suggestion: Detailless figures

WvW culling suggestion: Detailless figures

in Suggestions

Posted by: Greiger.7092

Greiger.7092

I can’t really think of a title that quite describes what I’m talking about quite right.

Problem: The servers can’t send all the data needed to render all the members of a zerg all the time.

Proposed solution: When the server needs to send more data than it can send, Send only the alliance, location, and direction of a player, and have the client simply render a simple team color stylized humanoid silhouette until it gets more information.

-----
From what I understand much of the problem with the culling is that the servers can’t send all the needed information to all players quickly enough for the full detail renders. For a player to be rendered at this point the server needs to have sent us race, class, all armor, all weapons, boons, conditions, hairstyle, haircolor, skincolor, skin pattern, height, buildtype, eyecolor, all the facial feature variables, animations, location, heading... Basically every single character creation option has a numeric value that must be sent from the server when the game wants to see a player.

That’s a hell of a lot of data to get one player rendering. Might not be much with current high speed internet, but when a server needs to send all that to hundreds of players at the same time, it adds up which seems to be giving us the culling issue.

That’s why I propose an intermediate level between rendering fully and not rendering at all. I think if the servers sent the location and heading of a player first, which the client only renders a quickly painted looking simple humanoid figure for, and then sent the rest of the data when it gets the chance the culling issue would be greatly reduced if not removed completely.

Location and heading is only a fraction of the data that needs to be sent currently. I for one would much rather have 2 people fully rendered with an army of faceless figures behind them that slowly fills out, than 4 people fully rendered and an empty field behind them that slowly renders a portion of the army.

WvW culling suggestion: Detailless figures

in Suggestions

Posted by: Kutai.4971

Kutai.4971

Sadly, whilst I don’t know the back-end of GW2, I can tell you this wouldn’t help. The “information” the servers are struggling to send/your machine is struggling to interpret are simply the position, direction and current actions of the other players.

Even if they were low detail actors, shadows or something similar, it probably wouldn’t work better than it does now. I mean right now, enemy players appear as generic “invader” models for this very reason.

WvW culling suggestion: Detailless figures

in Suggestions

Posted by: Greiger.7092

Greiger.7092

I mean right now, enemy players appear as generic "invader" models for this very reason.

I’m pretty sure this isn’t the case, I took a screenshot when I ended up in a bad situation and looked through it. To see if I could see generic invader models.

Instead I see 2 warriors with 2 different armor sets, 2 asura with different hairstyles and coloration, the hammer warrior near the front has a backpack while the others do not as well as a different armor than the Charr near him... Some are wearing helmets, others are not...

Even if I managed to catch all the differences with such a small group that are possible, that is still a lot of unnecessary information that could just not be sent.

Hell, I wouldn’t be surprised if those individual players could identify themselves in the screenshot.

Attachments:

(edited by Greiger.7092)

WvW culling suggestion: Detailless figures

in Suggestions

Posted by: Edragor.9164

Edragor.9164

Like this idea and proposed it in some other thread , too.
And have to disagree on the complaints some have.

You CAN hit culled players and even anticipate their movement, which makes it evident that culling CANNOT be related much to network or server lag, but must be some client-engine related performance issue.

Better to see only 5 animated chars in your vicnity and 20 generic-non-animated placeholders to indicate the other players…
…instead of seeing 10 chars in flawless animation, but having to be clairvoyant about the other 15 players..

The former might be more pretty & immersive, but the later is more of tactical & in-fight use…

(edited by Edragor.9164)