104 réponses à ce sujet

[ALTBOULI]

Spin-Orbit - I can't see it as any other thing. Otherwise I would be effected as well since [b]if it was the code only, then everyone would be hit equally[b] which isn't what we are seeing.
 
It may be a combination of course (and likely is).
 
Frostbite enforces sync of animations across players, and this sync performs very poorly with high pings, as seen on the Battlefield 3 and 4 games. (Most decent servers enforce a 150ms max ping policy using ProCon).
 
If we assume (and this is always dangerous to do ) that the same applies to DAIMP, then 150ms ping would let me connect with the US East cost (I get 110-120ms ping times there), but typically not the West coast or latin America, meanwhile my ping times to the baltic area is 15-25ms, and those are typically the Group of people I get matched with the most.

Not necessarily it all depends on the matchamking logic and how it priortizes matching players. In a densly populated area the matchmaking may appear to be working fine but I didnt encounter anyone outside of Europe (In most cases it was rare to play with a non-UK player). I dont encounter American/Canadian players unless I get an invite from a friend. And from what I inderstand its the same for them with European players

[TormDK]

I don't see one (drinking) excluding the other (doing a packet capture analysis)

 

In fact, it might help get some attention to the problem if we had some hard facts to go on. Speculating is all well and good, but rarely gets us anywhere.

 

But if we go with a star topology, then it's back to looking at people's ISPs/routing (other than hosting center node load at the time of capture, which we likely will never see any details on). Looking on the net, I haven't been able to find any details on where EA has their hosting centers, other than references to Equinix having helped them build them.

[Jeremiah12LGeek]

Spin-Orbit - I can't see it as any other thing. Otherwise I would be effected as well since if it was the code only, then everyone would be hit equally which isn't what we are seeing.

 

I have to point out that this is the opposite of how it would actually work.

 

If the netcode is an issue, it would not effect everyone equally. It would have a smaller impact on players with good local networks, a large playerbase within the region-lock (or whatever it is) and at least decent, consistent speed.

 

As players begin to lose one of those advantages, the problem would begin to get worse. As they lost more of those advantages, the problem would begin to become severe.

 

A player in a region with poor networks, a small playerbase, and low speed would be exposed to nearly every problem with the netcode.

 

In other words, the evidence supports the idea that the netcode has problems that need to be addressed. If the problem was with individual connections/networks, then it would present similarly across multiple MP games, rather than expressing specific, reproducible symptoms only when DA:I is played.

[ALTBOULI]

I have to point out that this the opposite of how it would actually work.
 
If the netcode is an issue, it would not effect everyone equally, at all. It would have a smaller impact on players with good local networks, a large playerbase within the region-lock (or whatever it is) and at least decent, consistent speed.
 
As players begin to lose one of those advantages, the problem would begin to get worse. As they lost more of those advantages, the problem would begin to become severe.
 
A player in a region with poor networks, a small playerbase, and low speed would be exposed to nearly every problem with the netcode.
 
In other words, the evidence supports the fact that the netcode has problems that need to be addressed. If the problem was individual connections/networks, then the it would present similarly across multiple MP games, rather than expressing specific, reproducible symptoms only when DA:I is played.

Well said!

[Spin-Orbit]

I have to point out that this the opposite of how it would actually work.
 
If the netcode is an issue, it would not effect everyone equally, at all. It would have a smaller impact on players with good local networks, a large playerbase within the region-lock (or whatever it is) and at least decent, consistent speed.
 
As players begin to lose one of those advantages, the problem would begin to get worse. As they lost more of those advantages, the problem would begin to become severe.
 
A player in a region with poor networks, a small playerbase, and low speed would be exposed to nearly every problem with the netcode.
 
In other words, the evidence supports the fact that the netcode has problems that need to be addressed. If the problem was individual connections/networks, then the it would present similarly across multiple MP games, rather than expressing specific, reproducible symptoms only when DA:I is played.

actually this isn't what the empirical evidence always agrees to. I am in east coast US a high density, large player base (even with the PS3) with robust internet connection. I often go into empty lobbies and often connect with small group of consistent users. Now there seems a preferential matchmaking algorithm perhaps trying to match fast connection with each other? Once in a strange blue moon I get matched with people from South America and Europe. Additionally the throughput issue doesn't guarantee smoother performance once in the game as well.

For the actual runs

The star topology makes sense but we don't know all the elements are actually originating from the host machine.

I would think the host
1. handles actual run
2. syncing timers
3. download from server the run information (loot i.e. gold or item, enemy type and count)
4. handle the spawning
5. runs combat calculations since this has to be synced
6. Uploads the information to the server

Users
1. communicates with the server to load the basic run info
2. retains certain individual run information such as gold count
3. downloads after the run item type (drop)
4. experience/run count/etc. sync to server

For the actual matchmaking

1. PC/Console sends a packet to the server requesting a session (to specific region?)
2. this is the mystery
a. Server maintain a database of PC/Consoles requests (from specific region?)
b. just broadcast the request (to specific region?) acting as a simple man in the middle
c. not be involved at all and the initial step 1 is broadcasted
3. Once a host is found (some PC/console that couldn't find a match) it has a on a flag to accept incoming requests, sends acknowledgement what parameters they used for this is in the netcode
4. Host and user starts predetermine steps for connection, latency, PC/Console type etc.
5. Once the checks are accepted between host and user the joining process begins (at which point the netcode actually would be used and would be the culprit for the hanging in endless joining or typival 3 timeouts and disconnect that explains the endless loop when not connecting during a failed joining process)
6. There should be two timout timers, one for the actual IP process that can cause the disconnect second should be in the netcode should there be incomplete joining of the user to the host (this may not be implemented)
7. Once connected there should be syncing timer parameters (causes of rubberbanding) however this seem to be host to each user individually evident from some users experiencing rubberbanding and not others at a given time.
8. If the host itself experience serious slowdown all users see issues in lag such as the frame rate issues.

ME3MP why it works smoother maybe be due to more server based runs (another wild speculation here), that would provide a database of users, matches and be able to provide a non-regional matchmaking. DAIMP they may have went to much ore of peer model because they though it would be more efficient, less costly since server requirements are reduced, quicker matchmaking, etc. this is obviously a issue


Keep in mind that all of the above are my speculations as I don't really have any information to suggest anything I wrote might be accurate, off-base or etc. This includes how they would go about addressing the issue.

tldr

[Jeremiah12LGeek]

actually this isn't what the empirical evidence always agrees to.

 

To be honest, I don't know what this means.

 

I read through the rest of your post, and I did not see you present any evidence that contradicted anything that I said.

[Spin-Orbit]

I don't see one (drinking) excluding the other (doing a packet capture analysis)
 
In fact, it might help get some attention to the problem if we had some hard facts to go on. Speculating is all well and good, but rarely gets us anywhere.
 
But if we go with a star topology, then it's back to looking at people's ISPs/routing (other than hosting center node load at the time of capture, which we likely will never see any details on). Looking on the net, I haven't been able to find any details on where EA has their hosting centers, other than references to Equinix having helped them build them.

As I thought about it longer, I realize any packet capture will give me the IP information, assuming it is TCP I will see the header, destination information but the application payload should be encrypted (never know with BW) so I will see if it is specific server or IP address and would need same done from couple users to see if there is a North America, European etc. spread or if it is a broadcast address. It wouldn't get me into the interworking of the actual process itself.

[Spin-Orbit]

To be honest, I don't know what this means.
 
I read through the rest of your post, and I did not see you present any evidence that contradicted anything that I said.

I basically said it isn't necessary that the netcode really at bottom of this as much as the architecture

[Jeremiah12LGeek]

I basically said it isn't necessary that the netcode really at bottom of this as much as the architecture

 

I never said that the netcode was solely responsible for the issues. I said that the evidence supports the idea that the netcode has issues that need to be addressed.

 

And while speculation is fine, speculation is not empirical evidence, as you described it.

[Spin-Orbit]

I never said that the netcode was solely responsible for the issues. I said that the evidence supports the idea that the netcode has issues that need to be addressed.
 
And while speculation is fine, speculation is not empirical evidence, as you described it.

empirical evidence is contradictory to what you suggested on the players with good networks, large playerbase etc.

If the netcode is an issue, it would not effect everyone equally, at all. It would have a smaller impact on players with good local networks, a large playerbase within the region-lock (or whatever it is) and at least decent, consistent speed.
 
As players begin to lose one of those advantages, the problem would begin to get worse. As they lost more of those advantages, the problem would begin to become severe.
 
A player in a region with poor networks, a small playerbase, and low speed would be exposed to nearly every problem with the netcode.

[Jeremiah12LGeek]

empirical evidence is contradictory to what you suggested on the players with good networks, large playerbase etc.
 

 

You didn't provide any empirical evidence that contradicts what I said. I'm not sure if you're misreading what I wrote, or if you don't know what "empirical evidence" means.

[Spin-Orbit]

You didn't provide any empirical evidence that contradicts what I said. I'm not sure if you're misreading what I wrote, or if you don't know what "empirical evidence" means.

actually this isn't what the empirical evidence always agrees to. I am in east coast US a high density, large player base (even with the PS3) with robust internet connection. I often go into empty lobbies and often connect with small group of consistent users.

your into loops as well I see, you know 8-track was one big tape loop

[Jeremiah12LGeek]

your into loops as well I see, you know 8-track was one big tape loop

 

You're making less and less sense with each response to me. What exactly are you accusing me of?

[Spin-Orbit]

You're making less and less sense with each response to me. What exactly are you accusing me of?

your into groundhog day or into The looper? I personally think groundhog day is funnier

[Jeremiah12LGeek]

your into groundhog day or into The looper? I personally think groundhog day is funnier

 

Is this supposed to be empirical evidence, too? 

 

I asked you what it is that you're accusing me of. Why would you be unwilling to answer that quesion?

[Spin-Orbit]

Is this supposed to be empirical evidence, too? 
 
I asked you what it is that you're accusing me of. Why would you be unwilling to answer that quesion?

LOL I changed my mind your more like a slinky, instead of going around in circles with same argument, you argue with slight changes continuously until it is completely off in a different direction. If you can deduce that from your arguments, then I can't have you see it.

[Jeremiah12LGeek]

LOL I changed my mind your more like a slinky, instead of going around in circles with same argument, you argue with slight changes continuously until it is completely off in a different direction. If you can deduce that from your arguments, then I can't have you see it.

 

I'm having a very difficult time following you, and I honestly don't believe the issue lies with me. If you are accusing me of "moving the goal posts" when you say "argument changes continuously," could you be more specific? Because the original quote was:

 

the evidence supports the fact that the netcode has problems that need to be addressed

 

While I should have used the word "idea" rather than "fact," it was a small misstep in grammar (which for reasons I won't get into you ought not to be criticizing me for.)

 

When you misquoted me, I referred to it again:

 

I never said that the netcode was solely responsible for the issues. I said that the evidence supports the idea that the netcode has issues that need to be addressed.

 

In this instance, I used the correct word "idea" rather than "fact." If that is the sole reason you are claiming that I changed my argument, I have to point something out. You are taking issue with a small grammatical error in what I said, and claiming that it is the equivalent of "moving the goalposts."

[Ispan]

The issue seems to be that anecdotal evidence is being mislabeled empirical data and presented as accurate and reliable information rather than the relatively limited observation that it is. Then there's this weird looping accusation that comes up when anyone disagrees

[Jeremiah12LGeek]

The issue seems to be that anecdotal evidence is being mislabeled empirical data and presented as accurate and reliable information rather than the relatively limited observation that it is. Then there's this weird looping accusation that comes up when anyone disagrees

 

I think it may well be both of my guesses:

 

You didn't provide any empirical evidence that contradicts what I said. I'm not sure if you're misreading what I wrote, or if you don't know what "empirical evidence" means.

 

The only part that bothers me is that I'm apparently being accused of a couple of things that are some kind of secret that I'm not allowed to know. I really would prefer that he just tell me what these accusations are supposed to be, instead of these weird sideways insults that only he seems to understand.

[Ispan]

I think it may well be both of my guesses:


The only part that bothers me is that I'm apparently being accused of a couple of things that are some kind of secret that I'm not allowed to know. I really would just prefer that he tell me what these accusations are supposed to be instead of these weird sideways insults that only he seems to understand.

Honestly I went afk for a moment in the middle of writing and missed several of your posts that I would've just 'liked' instead of writing my own had I seen them. +1 for being patient and levelheaded, you are a saint

[Spin-Orbit]

The issue seems to be that anecdotal evidence is being mislabeled empirical data and presented as accurate and reliable information rather than the relatively limited observation that it is. Then there's this weird looping accusation that comes up when anyone disagrees

Empirical is observation based evidence, my observation is presented, limited yes. However that can be disputed, refuted or confirmed by others. That would be how you build the limited to a wider base of observations.

I think it may well be both of my guesses:
 
 
The only part that bothers me is that I'm apparently being accused of a couple of things that are some kind of secret that I'm not allowed to know. I really would prefer that he just tell me what these accusations are supposed to be instead of these weird sideways insults that only he seems to understand.

My accusation is simple you start with one argument and move sideways to another and seems to me that is what your accusing me of here. I presented metaphors and if you look at a circle vs slinky it is self evident. If your confused then I have said it above.

[ALTBOULI]

Stop trolling him Jeremiah!

[Spin-Orbit]

Honestly I went afk for a moment in the middle of writing and missed several of your posts that I would've just 'liked' instead of writing my own had I seen them. +1 for being patient and levelheaded, you are a saint

Really...seem like a this for that to me...interesting observation (funny thing is censored)

[Jeremiah12LGeek]

+1 for being patient and levelheaded, you are a saint

 

Don't be fooled. That glowing halo around my head is just caused by all the alcohol.

[Jeremiah12LGeek]

My accusation is simple

 

Oh, here's hoping that turns out to be true...

 

you start with one argument and move sideways to another

 

Okay. Could you tell me what this change in argument was, or provide a quote from me where I do this?

 

seems to me that is what your accusing me of here.

 

Okay. Could you provide a quote from me where I accuse you of this?

 

I presented metaphors and if you look at a circle vs slinky it is self evident.

 

Your accusation has officially abandoned any notion of simplicity. That didn't last long.

 

So... I'm a circle vs slinky. Care to explain how that reference makes any sense? Or can you even do that?

 

If your confused then I have said it above.

 

I am definitely confused, although I'm not sure I can concur that the confusion technically belongs to me. It's more of its own confusion.* You said what above? Rather than making partial sentences that refer to a mysterious thing that you never identify, could you please tell me, specifically, and using examples, what it is that you are accusing me of?

 

 

 

*In my defense, I held out on making a grammar joke until now. I'm not made of stone. I wasn't going to last forever.