These forums have been archived and are now read-only.

The new forums are live and can be found at https://forums.eveonline.com/

EVE General Discussion

 
  • Topic is locked indefinitely.
 

Lag, TiDi, 6-VT and you...

First post
Author
Tippia
Sunshine and Lollipops
#221 - 2013-07-31 15:59:06 UTC  |  Edited by: Tippia
Murk Paradox wrote:
And I already spoke to saying I do not share that viewpoint. It was after it was already answered you kicked in.
…with the follow-up question of asking for more details.
Also, just because you accidentally half-answered it doesn't change the fact that you rather massively misunderstood whose question it was and to whom it was posed — instead, you assumed that people had said things they didn't say.

Quote:
And you should take your own advice. It isn't your discussion. It isn't your question, and it sure as hell isn't your answer.
I already do. It is indeed my discussion, and while neither the question or the answer are mine, I would like to hear more of yours because it's pretty hollow as currently stated.

Quote:
That's what lag is. A delay.
No, lag is a bit more than that. And even if lag were just “a delay”, that wouldn't mean that all delays are lag. TiDi, for instance, is not lag, in spite of the fact that it delays the completion of a server tick.
Murk Paradox
Ministry of War
Amarr Empire
#222 - 2013-07-31 16:05:52 UTC
"Sorry, I like to write about computer science-y stuff. Where this is going is explaining the main way in which tasklets behave under heavy load. When the server is overloaded, it can take upwards of five seconds or more to get another turn at doing stuff after you've yielded. The upshot of all of this is that tasks that are either very nice or have lots of communication requirements can end up taking a whole hell of a long time to complete as they spend most of their time waiting for a turn to execute. Ship death is a great example of this - it takes many round trips to the database, so yields very frequently, causing the event to stretch over many minutes before finally completing. Similarly, modules can get very far behind in processing because the tasklet that manages them yields after 100 milliseconds of execution.

(http://community.eveonline.com/news/dev-blogs/introducing-time-dilation-tidi/)

"Lag is a common word meaning to fail to keep up or to fall behind.[1] In real-time applications, the term is used when the application fails to respond in a timely fashion to inputs.[2][3]

Lag is also often used in reference to video games to describe the delay (or latency) between an action by a player and the reaction of the game.[4]

(http://en.wikipedia.org/wiki/Lag)

I am still not understanding why people are trying to say TiDi is not lag.

TiDi by design, increases the time the application processes the input on server side. It's the design of it.

"it's not lag, it just slows things down" is the most ******** argument I have ever seen.

This post has been signed by Murk Paradox and no other accounts, alternate or otherwise. Any other post claiming to be this holder's is subject to being banned at the discretion of the GM Team as it would violate the TOS in regards to impersonation. Signed, Murk Paradox. In triplicate.

Alavaria Fera
GoonWaffe
#223 - 2013-07-31 16:05:58 UTC
Maximillian German wrote:
No. You obviously don't know how the system works. People have already explained what TiDi is and what it does multiple times. Yet you still ignore their explanations and then pull incorrect math out of your ass in an attempt to prove a point that you don't even have. When asked a question, you fail to give even a straight answer. You are literally too stupid to hold constructive argument. This is the last time I will reply to your idiotic posts.

You know something.

I think everyone here's fallen for an EVEO General Discussion troll. Tech 2.

Triggered by: Wars of Sovless Agression, Bending the Knee, Twisting the Knife, Eating Sov Wheaties, Bombless Bombers, Fizzlesov, Interceptor Fleets, Running Away, GhostTime Vuln, Renters, Bombs, Bubbles ?

Murk Paradox
Ministry of War
Amarr Empire
#224 - 2013-07-31 16:07:51 UTC  |  Edited by: Murk Paradox
Tippia wrote:
Murk Paradox wrote:
And I already spoke to saying I do not share that viewpoint. It was after it was already answered you kicked in.
…with the follow-up question of asking for more details.
Also, just because you accidentally half-answered it doesn't change the fact that you rather massively misunderstood whose question it was and to whom it was posed — instead, you assumed that people had said things they didn't say.

Quote:
And you should take your own advice. It isn't your discussion. It isn't your question, and it sure as hell isn't your answer.
I already do. It is indeed my discussion, and while neither the question or the answer are mine, I would like to hear more of yours because it's pretty hollow as currently stated.

Quote:
That's what lag is. A delay.
No, lag is a bit more than that. And even if lag were just “a delay”, that wouldn't mean that all delays are lag. TiDi, for instance, is not lag, in spite of the fact that it delays the completion of a server tick.



I do not think razzle dazzle is a solution to tidi. I do not think other improvements to the game are anything other than other improvements to the game.

There's my answer. (I don't think the dev team is single threaded =P).

This post has been signed by Murk Paradox and no other accounts, alternate or otherwise. Any other post claiming to be this holder's is subject to being banned at the discretion of the GM Team as it would violate the TOS in regards to impersonation. Signed, Murk Paradox. In triplicate.

Tippia
Sunshine and Lollipops
#225 - 2013-07-31 16:12:31 UTC  |  Edited by: Tippia
Murk Paradox wrote:
I am still not understanding why people are trying to say TiDi is not lag.
Because the entire purpose and design of TiDi is to ensure that the server can keep up, and that the application keeps responding in a timely fashion to inputs — you know, kind of the exact opposite of the meaning of the quote you used.

To be precise, and to use a quote from the blog, directly after the part you used (which merely describes the problem, not the solution): “The EVE Online server should degrade in such a way that the queue of tasklets waiting for execution is minimized - ideally zero. That's precisely what Time Dilation aims to do.”

In other words, what TiDi does is remove the execution delays by ensuring that the server never reaches that overloaded state where some tasks get bumped into bajillionth place in the queue (which is what we would call lag).
Maximillian German
Task Force Coalition
#226 - 2013-07-31 16:17:02 UTC  |  Edited by: Maximillian German
Murk Paradox wrote:
Tippia wrote:
Murk Paradox wrote:
And I already spoke to saying I do not share that viewpoint. It was after it was already answered you kicked in.
…with the follow-up question of asking for more details.
Also, just because you accidentally half-answered it doesn't change the fact that you rather massively misunderstood whose question it was and to whom it was posed — instead, you assumed that people had said things they didn't say.

Quote:
And you should take your own advice. It isn't your discussion. It isn't your question, and it sure as hell isn't your answer.
I already do. It is indeed my discussion, and while neither the question or the answer are mine, I would like to hear more of yours because it's pretty hollow as currently stated.

Quote:
That's what lag is. A delay.
No, lag is a bit more than that. And even if lag were just “a delay”, that wouldn't mean that all delays are lag. TiDi, for instance, is not lag, in spite of the fact that it delays the completion of a server tick.



I do not think razzle dazzle is a solution to tidi. I do not think other improvements to the game are anything other than other improvements to the game.

There's my answer. (I don't think the dev team is single threaded =P).


ok, can't help replying. My resistance is weak. Anyway, the OP didn't say that razzle dazzle(really shiny stuff) is a solution, but rather a distraction by CCP from the problem. Maybe it'll make more sense like this:

players: "CCP, why aren't you working on getting rid of lag?"

CCP: "well you see kids--oh look a new scanning system!"

players: "COOL. WE'VE TOTALLY FORGOTTEN TO PESTER CCP ABOUT LAG!"

The OP is asking why we let ourselves be tricked, more or less. Also, a follow up TiDi dev post:

http://community.eveonline.com/news/dev-blogs/time-dilation-hows-that-going/
Maximillian German
Task Force Coalition
#227 - 2013-07-31 16:20:01 UTC
Tippia wrote:
Murk Paradox wrote:
I am still not understanding why people are trying to say TiDi is not lag.
Because the entire purpose and design of TiDi is to ensure that the server can keep up, and that the application keeps responding in a timely fashion to inputs — you know, kind of the exact opposite of the meaning of the quote you used.

To be precise, and to use a quote from the blog, directly after the part you used (which merely describes the problem, not the solution): “The EVE Online server should degrade in such a way that the queue of tasklets waiting for execution is minimized - ideally zero. That's precisely what Time Dilation aims to do.”

In other words, what TiDi does is remove the execution delays by ensuring that the server never reaches that overloaded state where some tasks get bumped into bajillionth place in the queue (which is what we would call lag).


thus, we see a slowdown, but the execution queue does not. I guess you could kinda say we're getting rid of server lag, but creating client lag, assuming that we are taking lag to mean slowdown in this instance.
Murk Paradox
Ministry of War
Amarr Empire
#228 - 2013-07-31 16:29:15 UTC
Tippia wrote:
Murk Paradox wrote:
I am still not understanding why people are trying to say TiDi is not lag.
Because the entire purpose and design of TiDi is to ensure that the server can keep up, and that the application keeps responding in a timely fashion to inputs — you know, kind of the exact opposite of the meaning of the quote you used.

To be precise, and to use a quote from the blog, directly after the part you used (which merely describes the problem, not the solution): “The EVE Online server should degrade in such a way that the queue of tasklets waiting for execution is minimized - ideally zero. That's precisely what Time Dilation aims to do.”

In other words, what TiDi does is remove the execution delays by ensuring that the server never reaches that overloaded state where some tasks get bumped into bajillionth place in the queue (which is what we would call lag).



Increasing the time is not "in a timely fashion".

This post has been signed by Murk Paradox and no other accounts, alternate or otherwise. Any other post claiming to be this holder's is subject to being banned at the discretion of the GM Team as it would violate the TOS in regards to impersonation. Signed, Murk Paradox. In triplicate.

Tippia
Sunshine and Lollipops
#229 - 2013-07-31 16:33:36 UTC  |  Edited by: Tippia
Murk Paradox wrote:
Increasing the time is not "in a timely fashion".
It is when you know when (and that) it will happen, as opposed to having no idea if or when your commands will ever be processed.

Once a tick is in a timely fashion, even if that tick happens to be a bit longer than usual.
Maybe in a tick (or six hundred (perhaps (if you're lucky (this time, but don't expect it to work the same the next time)))) is not in a timely fashion.

TiDi ensures the former as far as possible; lag is when we have the latter.
Murk Paradox
Ministry of War
Amarr Empire
#230 - 2013-07-31 16:33:49 UTC
Tippia wrote:


To be precise, and to use a quote from the blog, directly after the part you used (which merely describes the problem, not the solution): “The EVE Online server should degrade in such a way that the queue of tasklets waiting for execution is minimized - ideally zero. That's precisely what Time Dilation aims to do.”

In other words, what TiDi does is remove the execution delays by ensuring that the server never reaches that overloaded state where some tasks get bumped into bajillionth place in the queue (which is what we would call lag).



Aims to do, but does not accomplish. Because TiDi stays throttled (most likely due to events since the time of the blog and by growth or lack of).

"Thankfully we do have a means to throttle load that buys us a lot while leaving the design intact: make time run slower. A large majority of the load in large engagements is tied to the clock - modules, physics, travel, warpouts, all of these things happen over a time period, so spacing out time will lower their load impact proportionally. So, the idea here is to slow down the game clock enough to maintain a very small queue of waiting tasklets, then when the load clears, raise time back up to normal as we can handle it. This will be done dynamically and in very fine increments; there's no reason we can't run at 98% time if we're just slightly overloaded."

which is much further down the area you quoted.

That is how TiDi "creates" lag; by extending the time. The blog, as a read, shows how they want TiDi to handle spikes of server load and then compensate.. but it doesn't. Whether it be from pilots or server load of actions and extra elements (ammo, drones, etc) the TiDi plateaus constantly.

That's why it becomes lag, and not simply a "collision" of time that balances itself out which is what you are implying it does (it should, but doesn't).

This post has been signed by Murk Paradox and no other accounts, alternate or otherwise. Any other post claiming to be this holder's is subject to being banned at the discretion of the GM Team as it would violate the TOS in regards to impersonation. Signed, Murk Paradox. In triplicate.

Maximillian German
Task Force Coalition
#231 - 2013-07-31 16:33:56 UTC  |  Edited by: Maximillian German
Murk Paradox wrote:
Tippia wrote:
Murk Paradox wrote:
I am still not understanding why people are trying to say TiDi is not lag.
Because the entire purpose and design of TiDi is to ensure that the server can keep up, and that the application keeps responding in a timely fashion to inputs — you know, kind of the exact opposite of the meaning of the quote you used.

To be precise, and to use a quote from the blog, directly after the part you used (which merely describes the problem, not the solution): “The EVE Online server should degrade in such a way that the queue of tasklets waiting for execution is minimized - ideally zero. That's precisely what Time Dilation aims to do.”

In other words, what TiDi does is remove the execution delays by ensuring that the server never reaches that overloaded state where some tasks get bumped into bajillionth place in the queue (which is what we would call lag).



Increasing the time is not "in a timely fashion".


But TiDi does not increase server response time, it simply slows down the in-game clock. While modules may take 2x as long to run their full cycle, they will still start that cycle as soon as you hit 'em. The problem before TiDi was waiting minutes for modules to start the cycle, having modules never stop cycling, etc.
Murk Paradox
Ministry of War
Amarr Empire
#232 - 2013-07-31 16:34:32 UTC
Tippia wrote:
Murk Paradox wrote:
Increasing the time is not "in a timely fashion".
It is when you know when (and that) it will happen, as opposed to having no idea if or when your commands will ever be processed.

Once a tick is in a timely fashion, even if that tick happens to be a bit longer than usual.
Maybe in a tick (or six hundred (perhaps (if you're lucky (this time, but don't expect it to work the same the next time)))) is not in a timely fashion.

TiDi ensures the former as far as possible; lag is when we have the latter.



no, it isn't. When I click a module and it takes 3 seconds, then a reclick takes 10 seconds, then 2 minutes, then 13.... that is not a timely fashion.

This post has been signed by Murk Paradox and no other accounts, alternate or otherwise. Any other post claiming to be this holder's is subject to being banned at the discretion of the GM Team as it would violate the TOS in regards to impersonation. Signed, Murk Paradox. In triplicate.

Murk Paradox
Ministry of War
Amarr Empire
#233 - 2013-07-31 16:36:40 UTC
Maximillian German wrote:



I do not think razzle dazzle is a solution to tidi. I do not think other improvements to the game are anything other than other improvements to the game.

There's my answer. (I don't think the dev team is single threaded =P).


ok, can't help replying. My resistance is weak. Anyway, the OP didn't say that razzle dazzle(really shiny stuff) is a solution, but rather a distraction by CCP from the problem. Maybe it'll make more sense like this:

players: "CCP, why aren't you working on getting rid of lag?"

CCP: "well you see kids--oh look a new scanning system!"

players: "COOL. WE'VE TOTALLY FORGOTTEN TO PESTER CCP ABOUT LAG!"

The OP is asking why we let ourselves be tricked, more or less. Also, a follow up TiDi dev post:

http://community.eveonline.com/news/dev-blogs/time-dilation-hows-that-going/[/quote]

I don't see it as being tricked, because like I said, I do not think the game revolves Time Dilation. I think CCP has many many things on their plate and can improve multiple things at various times.

This post has been signed by Murk Paradox and no other accounts, alternate or otherwise. Any other post claiming to be this holder's is subject to being banned at the discretion of the GM Team as it would violate the TOS in regards to impersonation. Signed, Murk Paradox. In triplicate.

Maximillian German
Task Force Coalition
#234 - 2013-07-31 16:40:51 UTC  |  Edited by: Maximillian German
Murk Paradox wrote:
Maximillian German wrote:



I do not think razzle dazzle is a solution to tidi. I do not think other improvements to the game are anything other than other improvements to the game.

There's my answer. (I don't think the dev team is single threaded =P).


ok, can't help replying. My resistance is weak. Anyway, the OP didn't say that razzle dazzle(really shiny stuff) is a solution, but rather a distraction by CCP from the problem. Maybe it'll make more sense like this:

players: "CCP, why aren't you working on getting rid of lag?"

CCP: "well you see kids--oh look a new scanning system!"

players: "COOL. WE'VE TOTALLY FORGOTTEN TO PESTER CCP ABOUT LAG!"

The OP is asking why we let ourselves be tricked, more or less. Also, a follow up TiDi dev post:

http://community.eveonline.com/news/dev-blogs/time-dilation-hows-that-going/


I don't see it as being tricked, because like I said, I do not think the game revolves Time Dilation. I think CCP has many many things on their plate and can improve multiple things at various times.
[/quote]

There we go, now you've answered the OP's question. Onwards to the lag argument.

edit: also, it looks like the quote button is misbehaving
Tippia
Sunshine and Lollipops
#235 - 2013-07-31 16:43:57 UTC  |  Edited by: Tippia
Murk Paradox wrote:
no, it isn't. When I click a module and it takes 3 seconds, then a reclick takes 10 seconds, then 2 minutes, then 13.... that is not a timely fashion.
It is when you compare with the alternatives, which are (in order) blackscreen, blackscreen, 20 minutes + stuck module, CTD. After all, in your cases, the module commands are being processed as opposed to… you know… not.

Quote:
That is how TiDi "creates" lag; by extending the time. The blog, as a read, shows how they want TiDi to handle spikes of server load and then compensate.. but it doesn't. Whether it be from pilots or server load of actions and extra elements (ammo, drones, etc) the TiDi plateaus constantly.
That's because the source of the load is not a spike, but a constant demand. Of course TiDi will plateau if the load does not go away. If there is but a spike, then TiDi compensates well enough, but those aren't the kind of situations that have created this whining.

And no matter how many times you repeat it, TiDi still does not create lag — processing overload does. TiDi reduces that overload.
Murk Paradox
Ministry of War
Amarr Empire
#236 - 2013-07-31 16:51:06 UTC
Maximillian German wrote:
Tippia wrote:
Murk Paradox wrote:
I am still not understanding why people are trying to say TiDi is not lag.
Because the entire purpose and design of TiDi is to ensure that the server can keep up, and that the application keeps responding in a timely fashion to inputs — you know, kind of the exact opposite of the meaning of the quote you used.

To be precise, and to use a quote from the blog, directly after the part you used (which merely describes the problem, not the solution): “The EVE Online server should degrade in such a way that the queue of tasklets waiting for execution is minimized - ideally zero. That's precisely what Time Dilation aims to do.”

In other words, what TiDi does is remove the execution delays by ensuring that the server never reaches that overloaded state where some tasks get bumped into bajillionth place in the queue (which is what we would call lag).


thus, we see a slowdown, but the execution queue does not. I guess you could kinda say we're getting rid of server lag, but creating client lag, assuming that we are taking lag to mean slowdown in this instance.



Well, if lag is to be a slowdown or delay between execution and input of commands (highschool glossary term?) then yes.

This post has been signed by Murk Paradox and no other accounts, alternate or otherwise. Any other post claiming to be this holder's is subject to being banned at the discretion of the GM Team as it would violate the TOS in regards to impersonation. Signed, Murk Paradox. In triplicate.

Murk Paradox
Ministry of War
Amarr Empire
#237 - 2013-07-31 16:54:47 UTC
Tippia wrote:
Murk Paradox wrote:
no, it isn't. When I click a module and it takes 3 seconds, then a reclick takes 10 seconds, then 2 minutes, then 13.... that is not a timely fashion.
It is when you compare with the alternatives, which are (in order) blackscreen, blackscreen, 20 minutes + stuck module, CTD. After all, in your cases, the module commands are being processed as opposed to… you know… not.

Quote:
That is how TiDi "creates" lag; by extending the time. The blog, as a read, shows how they want TiDi to handle spikes of server load and then compensate.. but it doesn't. Whether it be from pilots or server load of actions and extra elements (ammo, drones, etc) the TiDi plateaus constantly.
That's because the source of the load is not a spike, but a constant demand. Of course TiDi will plateau if the load does not go away. If there is but a spike, then TiDi compensates well enough, but those aren't the kind of situations that have created this whining.

And no matter how many times you repeat it, TiDi still does not create lag — processing overload does. TiDi reduces that overload.



Compare to the alternatives.... but we aren't comparing. We are defining.

"Dark brown is a light color" is a false statement. It clearly says dark and we know by looking it it isn't light.

"Dark brown is a light color when compared to black" would be a true statement.

I'm tired of splitting hairs with you. You simply keep adding addendums to prove a point, which deviates from the topic.

If you want to consider that defeat or victory, so be it. You will simply forgo a stance in order to win a literal argument regardless of the impact it has on players. I care more about the impact.

This post has been signed by Murk Paradox and no other accounts, alternate or otherwise. Any other post claiming to be this holder's is subject to being banned at the discretion of the GM Team as it would violate the TOS in regards to impersonation. Signed, Murk Paradox. In triplicate.

Frostys Virpio
State War Academy
Caldari State
#238 - 2013-07-31 16:54:58 UTC
Murk Paradox wrote:
Maximillian German wrote:



I do not think razzle dazzle is a solution to tidi. I do not think other improvements to the game are anything other than other improvements to the game.

There's my answer. (I don't think the dev team is single threaded =P).


ok, can't help replying. My resistance is weak. Anyway, the OP didn't say that razzle dazzle(really shiny stuff) is a solution, but rather a distraction by CCP from the problem. Maybe it'll make more sense like this:

players: "CCP, why aren't you working on getting rid of lag?"

CCP: "well you see kids--oh look a new scanning system!"

players: "COOL. WE'VE TOTALLY FORGOTTEN TO PESTER CCP ABOUT LAG!"

The OP is asking why we let ourselves be tricked, more or less. Also, a follow up TiDi dev post:

http://community.eveonline.com/news/dev-blogs/time-dilation-hows-that-going/


I don't see it as being tricked, because like I said, I do not think the game revolves Time Dilation. I think CCP has many many things on their plate and can improve multiple things at various times.
[/quote]

Do you have any idea how much code would need to be redone to get away from the current limitation? What will happen after they rebuilt it to sustain 6k players with no lag at all on the new reinforced node with better coding when the CFC decide to drop 5k pilot VS a fleet of 4k pilots in the same system?

Ah **** we're back at square one because it's the nature of the game. Anything even remotely possible is bound to happen. Remember Asakai? Before it happened, if you asked people around, do you think people would of belived a single missclick could of started a supercap escalation battle over a tower bash? Do you think people would of belived before 6VDT that a single system siege would amount to a 4k vessel fight over a station shield timer? What about Luminaire?

The more space you put in the system, the more pilots will go there to be part of the next big thing. Asking CCP to halt all other developpement until a way to handle super fights is silly at best. It's still on the table but not top priority.
Frostys Virpio
State War Academy
Caldari State
#239 - 2013-07-31 16:57:04 UTC
Murk Paradox wrote:
Tippia wrote:
Murk Paradox wrote:
no, it isn't. When I click a module and it takes 3 seconds, then a reclick takes 10 seconds, then 2 minutes, then 13.... that is not a timely fashion.
It is when you compare with the alternatives, which are (in order) blackscreen, blackscreen, 20 minutes + stuck module, CTD. After all, in your cases, the module commands are being processed as opposed to… you know… not.

Quote:
That is how TiDi "creates" lag; by extending the time. The blog, as a read, shows how they want TiDi to handle spikes of server load and then compensate.. but it doesn't. Whether it be from pilots or server load of actions and extra elements (ammo, drones, etc) the TiDi plateaus constantly.
That's because the source of the load is not a spike, but a constant demand. Of course TiDi will plateau if the load does not go away. If there is but a spike, then TiDi compensates well enough, but those aren't the kind of situations that have created this whining.

And no matter how many times you repeat it, TiDi still does not create lag — processing overload does. TiDi reduces that overload.



Compare to the alternatives.... but we aren't comparing. We are defining.

"Dark brown is a light color" is a false statement. It clearly says dark and we know by looking it it isn't light.

"Dark brown is a light color when compared to black" would be a true statement.

I'm tired of splitting hairs with you. You simply keep adding addendums to prove a point, which deviates from the topic.

If you want to consider that defeat or victory, so be it. You will simply forgo a stance in order to win a literal argument regardless of the impact it has on players. I care more about the impact.


You try to prove yourself right while supporting the assumption that blackscrren and client crash were not worse than slow but ongoing fight. Tippia won this argument many pages ago.
Tippia
Sunshine and Lollipops
#240 - 2013-07-31 17:00:10 UTC  |  Edited by: Tippia
Murk Paradox wrote:
Compare to the alternatives.... but we aren't comparing. We are defining.
No, we're using a highly relative term. “A timely manner” depends on the context and what you compare it against.

In this case, the commands are being handled in a timely manner (that time being within the server tick) rather than being forgotten or dropped or constantly being pushed further down the queue.

Quote:
I'm tired of splitting hairs with you.
Then stop making silly claims that rely on hair-splitting.