EVE Information Portal

 
^ Back to top

Topic is locked indefinitely.

4 Pages123Next pageLast page
 

New dev blog: Time Dilation – How’s That Going?

Jump to first DEV post
Author
C C P Alliance
#1 Posted: 2012.02.09 17:13  |  Edited by: CCP Guard
Time Dilation has been doing some pretty amazing things for big fleet fights...and those aren't just our words.

CCP Veritas of Team Gridlock, has written up a status report blog on Time Dilation. It has pretty graphs and is sprinkled with lovely pilot testimony. Check it out right here!
CCP Guard | EVE Community Developer | @ccp_guard
#2 Posted: 2012.02.09 17:15
Those are some sexy graphs.
No Holes Barred
#3 Posted: 2012.02.09 17:15
CCP Veritas = best Veritas!
CSM 7 Secretary
CSM 6 Alternate Delegate
@two_step_eve on Twitter
My Blog
Vote Steve Ronuken for CSM
#4 Posted: 2012.02.09 17:25
Third Big smile

But more seriously, looking good. Yes, tidi-ing the node is better than not dilating anything. Sure, it's a trifle annoying for other systems, but that's life. Good Job

Upholders
#5 Posted: 2012.02.09 17:27
I hear there have been some issues.

One day our wormhole was time dilated, only our wormhole and with only 5 people in it at that time.

Care to shed some light on the possible reason as to why that would happen?
Welcome to EVE online, here's your rubix cube, go F*** yourself ♥
C C P Alliance
#6 Posted: 2012.02.09 17:32
Mr Bigwinky wrote:
I hear there have been some issues.

One day our wormhole was time dilated, only our wormhole and with only 5 people in it at that time.

Care to shed some light on the possible reason as to why that would happen?

Wormholes, being nice semi-population-capped things, group up very nicely. Your average wormhole-housing node is going to have a few hundred wormhole systems on it.
CCP Veritas - Technical Director - EVE Online
Adhocracy
#7 Posted: 2012.02.09 17:33  |  Edited by: jonnykefka
Mr Bigwinky wrote:
I hear there have been some issues.

One day our wormhole was time dilated, only our wormhole and with only 5 people in it at that time.

Care to shed some light on the possible reason as to why that would happen?


Easy: It was on the same node as a different W-space system (or k-space, but I think all w-space systems are on w-space only nodes) that was under heavy load. TiDi applies to ALL systems on the node, regardless of how busy any individual system is. I've ended up in a nullsec system being the only one in local and the system at 50% TiDi.

Remember, nodes are huge clusters of systems. I've heard rumors that W-space may run on region-sized nodes because of the relatively low population, meaning that only one system in your entire region has to be loaded enough to trigger TiDi for you to experience it in your system.

EDIT: Beaten by Veritas. Curses!
Amarr Empire
#8 Posted: 2012.02.09 17:37
TiDi works wonderfully and is possibly one of the greatest improvements to large fleet fights in the past couple years. Add in client side optimizations in the coming months and null sec will be happier than it ever has been. Keep up the great work CCP Veritas and Team Gridlock! We love you long time for improving our overall game experience.
Dirt Nap Squad.
#9 Posted: 2012.02.09 17:37  |  Edited by: Trebor Daehdoow
+1!

Um, actually, based on average time-dilation, that should be +0.96875!
Retiring CSM Chairman • My picks for CSM9: http://treborofthecsm.blogspot.com/2014/04/how-to-vote.html
Minmatar Republic
#10 Posted: 2012.02.09 17:37
Congrats on TiDi guys - that and the compressed local were enough to make me resub for a year. Hopefully the coming year sees continued UI and performance improvements.

I think the major pain point right now is around session changes - not just the timer, but delays incurred due to jumping, docking, and logging in. TiDi really exposes how bad this is, since it makes the rest of the game much less painful. I realize it's probably a pretty fundamental problem, but it would be well worth the time invested if you guys could work out some kind of precache strategy or something to make session changes less costly.

Also, I think TiDi magnifies the effect of session changes during jumping. Before TiDi, traffic control would have a tendency to spread a large fleet out among several systems by keeping people from jumping. TiDi encourages a clumping effect, though - the main blob of the fleet slows down whatever node they're on. As people leave a system it speeds up, so they warp faster into the overloaded system where everyone is warping under time dilation. Traffic Control kicks in hardcore and annoys people, but it can't actually help the problem.

One possible solution would slow down the rate at which dilation reduces when people leave a system. The idea would be to keep people in the outbound system slowed long enough for people to pass through the inbound system, thus preventing a big traffic jam.
#11 Posted: 2012.02.09 17:45
I suppose the trick now is to use TiDi to be able to drop a system to 0% for a few seconds while it changes to a reinforced node without losing any players. Also: instead of disconnecting everyone over DT, the universe drops to 0% for the 15 minutes or so, then ramps back up when everything's done.
C C P Alliance
#12 Posted: 2012.02.09 17:49
The soon-to-be-released client performance optimizations from Team Gridlock got a minor mention, but imho they are majorly full of win.
- EVE Online Quality Assurance -
#13 Posted: 2012.02.09 17:56
Coool. Cool

And be noted that I've never been to nullsec, so that's how close I have been to experience the (in)famous lag.

But Cool is Cool.
The Greater Fool Bar is now open for business, 24/7. Come and have drinks and fun somewhere between RL and New Eden! Ingame chat channel: The Greater Fool Bar
No Holes Barred
#14 Posted: 2012.02.09 18:01
CCP Masheen wrote:
The soon-to-be-released client performance optimizations from Team Gridlock got a minor mention, but imho they are majorly full of win.


Does this mean you're fixing bracket lag?!?
Gallente Federation
#15 Posted: 2012.02.09 18:08
CCP you did a wonderful work there! congratulations!
Please read these!
> New POS system
> New SOV system
C C P Alliance
#16 Posted: 2012.02.09 18:10
BeanBagKing wrote:
Does this mean you're fixing bracket lag?!?

Work has been done along those lines. I can't say if any individual will call it "fixed" with what's been done, but it's possible.
CCP Veritas - Technical Director - EVE Online
#17 Posted: 2012.02.09 18:24
TiDi is some wickedly awesome black magic CCP. Thanks. Cool
WE FORM VOLTRON
#18 Posted: 2012.02.09 18:28  |  Edited by: Femaref
I remember your talk about it at last years fanfest Veritas, and how much I enjoyed the technical aspect of it. Really love to see it ingame and working, you, sir, really know your stuff.
#19 Posted: 2012.02.09 18:32  |  Edited by: Vincent Athena
BeanBagKing wrote:
CCP Masheen wrote:
The soon-to-be-released client performance optimizations from Team Gridlock got a minor mention, but imho they are majorly full of win.


Does this mean you're fixing bracket lag?!?


I did the last Sisi test. Bracket lag was much improved. I was able to "show all brackets" and still be able to do stuff. Also I was able to click back and forth between overview tabs without the client becoming a hopeless lagged out mess. Thanks CCP Veritas!

Now about collateral TiDi (that is TiDi in systems which have few people in them, due to a big event elsewhere). One way to reduce this is to just have more nodes. Any options for just getting more hardware? After all if you had 8000 supernodes the issue goes away. Or how about having multiple "virtual" nodes on one real node?
#20 Posted: 2012.02.09 18:57
Reppers go “wommm”, just fyi.
“If you're not willing to fight for what you have in ≡v≡ you don't deserve it, and you will lose it.”

Get a good start: Newbie skill plan 2.1.
4 Pages123Next pageLast page
Forum Jump