Why does it matter if some players make the game easier for themselves?

[Aewarin Jameson]

question, did you ever play any Online game, to know what the level of creativity of people is to find exploits and use them to their advantage?

Oh yes, plenty thanks. Agreed there are some very creative players but for the years I admin'd a UT server the number of cheaters compared to actual players was low .. now that admittedly was in 2000 and times certainly have changed. However, even playing WoW PvP / BF2 / BF3 the actual number of cheaters compared to players is still low .. can't say for sure recently as I don't play that much any more but I like to think the majority of players still don't.

I consider solo/private group switching a potential exploit and a method of griefing (e.g. steal stuff from under the Allgroups nose).

I understand you point of view on this

 

[jabokai]

double-post... oops.

 

[jabokai]

Not according to the DDF Archives.

Any and all of these seem improvements over "tough luck". I accept that if I lose connection in combat then it's tough luck, but outside of combat I don't expect to just have my ship sit somewhere in space until I log back in.

I hadn't seen that - before my time in the DDF!

Was there ever a poll created that the thread refers to? Was there a final proposal? I can't see it (with a cursory glance through the DDF archive). Those all seem to be player suggestions with no actual decision? :S

 

[DigitalDuck]

Was there ever a poll created that the thread refers to? Was there a final proposal?

Dunno - I'm not in the DDF, so I have no idea what actually transpires there unless it's in the archives, and I can't see anything either.

 

[Kerrash]

If I remember, if you log off out of combat you ship would appear to enter hyper space and exit agin when you log back in.

 

[Cathy]

@ Erimus. Something that does concern me a little is that a player who pulls the plug, as you put it, might [edit]affect the hosting of your session, which would be unlucky and annoying. But if that happened with that player a lot I would give them the benefit of the doubt and conclude their internet connection is too unstable - and put them on ignore so my chances of being matched with them even in the all-players become very slim. Then even if they are doing it deliberately, problem solved

 

[ianw]

@ Erimus. Something that does concern me a little is that a player who pulls the plug, as you put it, might be the one who his hosting your session, which would be unlucky and annoying. <snip>

Okay - being a bit dense here possibly - but is that how this P2P thing works? One of the PCs is hosting the session for everyone (in that session)?

 

[Kerrash]

@ Erimus. Something that does concern me a little is that a player who pulls the plug, as you put it, might be the one who his hosting your session, which would be unlucky and annoying. But if that happened with that player a lot I would give them the benefit of the doubt and conclude their internet connection is too unstable - and put them on ignore so my chances of being matched with them even in the all-players become very slim. Then even if they are doing it deliberately, problem solved

Connectivity between players has been confirmed to be P2P, so that shouldn't happen

 

[Cathy]

Connectivity between players has been confirmed to be P2P, so that shouldn't happen

I hope so. Many little irons in the fire should keep it going - but I just have this little contraindications concern with stuff I don't understand in detail

 

[Cathy]

Okay - being a bit dense here possibly - but is that how this P2P thing works? One of the PCs is hosting the session for everyone (in that session)?

No no... sorry, I have put things badly. What I'm trying to get at is, if one player d/cs a lot, might that affect the session somehow (frame-rate, etc.) as others in that session keep playing?

I wish I was more technical

 

[ianw]

No no... sorry, I have put things badly. What I'm trying to get at is, if one player d/cs a lot, might that affect the session somehow (frame-rate, etc.) as others in that session keep playing?

I wish I was more technical

No prob - I've yet to see it put somewhere concisely how it will work. For instance does the experience for all in the session default to the capabilities of the poorest quality broadband circuit in it?

I expect it will become clearer when the alpha is out...

 

[jabokai]

No prob - I've yet to see it put somewhere concisely how it will work. For instance does the experience for all in the session default to the capabilities of the poorest quality broadband circuit in it?

I expect it will become clearer when the alpha is out...

I do remember seeing something from a dev but cannot remember where. My recollection is that the strongest connection (bandwidth/ping/recorded dropouts?) handles the session but there's always another one ready to take over.

Apologies if I misremembered, maybe it'll jar someone else's memory and make the quote findable!

 

[Aewarin Jameson]

I do remember seeing something from a dev but cannot remember where. My recollection is that the strongest connection (bandwidth/ping/recorded dropouts?) handles the session but there's always another one ready to take over.

Apologies if I misremembered, maybe it'll jar someone else's memory and make the quote findable!

From my understanding of it (and I don't have anything to back this up) when you enter a region of space with other players you would open P2P links with everyone creating a full mesh. (Perhaps that's why there is a limit of 32 ?). As you do anything you update everyone and as they do something they update you .. and so on. (Going to be fun talking to me then out in the Styx )

If this is true then you wouldn't need a master / slave model as you're all peers of each other. I do admit this model would be rather inefficient as the NW overhead is greater than if you had a master / slaves

 

[jabokai]

From my understanding of it (and I don't have anything to back this up)

Pfft - that's just guesswork and conjecture then, whereas my post was based on actual* recollection!


* It may have been a dream.

 

[Aewarin Jameson]

Pfft - that's just guesswork and conjecture then, whereas my post was based on actual* recollection!


* It may have been a dream.

Guess I am ...

 

[DigitalDuck]

From my understanding of it (and I don't have anything to back this up) when you enter a region of space with other players you would open P2P links with everyone creating a full mesh. (Perhaps that's why there is a limit of 32 ?).

[...]

If this is true then you wouldn't need a master / slave model as you're all peers of each other. I do admit this model would be rather inefficient as the NW overhead is greater than if you had a master / slaves

That's what peer-to-peer means...

It's no more or less inefficient than the server/client model, as you're sending the same amount of data in total in both cases anyway. If we assume a full instance of 32 players:

- In a P2P model, each player receives 31 players' information and sends their own information to 31 players (total communications sent per update: (31 + 31) * 32 = 1984)

- In a server/client model, the server (which in this case is one of the players) receives 31 players' information and sends everyone's information to 31 players, while the clients each send their own information to the server and receives 31 players' information (total communications sent per update: 31 + 31 * 31 + (1 + 31) * 31 = 1984)

However, in a server/client model the clients are not connected to each other, so if the server disconnects the entire session is lost.

From what I understand, though, the game will choose one player to act as an arbiter to settle mismatches, and it's this that changes when players disconnect/change instances.

 

[jabokai]

Found what I remembered, the post being by one of our own and then confirmed to be basically correct by Mike - http://forums.frontier.co.uk/showpost.php?p=100538&postcount=326

I was sure there was more... perhaps it was another thread, or perhaps it really was a dream!

 

[Aewarin Jameson]

It's no more or less inefficient than the server/client model, as you're sending the same amount of data in total in both cases anyway. If we assume a full instance of 32 players:

The data amount is the same but the distribution is different.

A full mesh relies upon all nodes having the same amount of BW available to them which, depending upon the data_size could swamp some pilots internet. A client / server model relies upon the server having more than enough BW to support everyone and the clients only need to speak to 1 person, not all.

• More efficient in terms of number of packets being sent around the network from each pilot (except the server)
• More efficient in terms of BW required by the clients.
• Less efficient in terms of update times as everything has to go via the server and not direct.

<Shrug> Many ways to skin a cat.

From what I understand, though, the game will choose one player to act as an arbiter to settle mismatches, and it's this that changes when players disconnect/change instances.

That's the interesting part .. suggests a Server/Client model right and not full mesh ?

 

[jabokai]

.. suggests a Server/Client model right and not full mesh ?

See the post linked to in my post above - you may have missed it as you were typing at the time!

 

[Aewarin Jameson]

See the post linked to in my post above - you may have missed it as you were typing at the time!

Spot on - good find sir!


(how topics sway from OP to now is funny )