Serving MOUL Peer To Peer

[JWPlatt]

You're just running the server code on your local machine. The physical location of the server code changes nothing. It would be a huge maintenance burden to have two different versions of the server.

So basically we are back to the idea of developing a server and making the code public so that anyone could start a shard.

Peer to peer games usually require all participants to be at the latest version. End of maintenance burden.

It is incorrect to think in terms of a shard to which everyone connects. Peer to peer is defined as a different topology. Keep to the definition.

Come on folks. I think the Guild of Writers can do far better than think in naysaying terms. That's why I asked for how it could be made to work - not excuses for giving up. It is off topic to give up. Please restrict contributions to the possibilities and what can be done to overcome the challenges. Anything else is a waste of time.

[teedyo]

Hate to be a naysayer but, peer to peer just doesn't scale well. The information coming in to a client will remain about the same as a client <-> server setup but the data being sent out is an order of n-1 larger where n is the number of peers. ie. if you have 10 peers; you'll be sending out 9x the data you would with a client <-> server. In effect, you'd be putting the server's load on the client which still has to do all the rendering, avatar tracking and etc.

[Trylon]

However, combining the two infrastructres would be interesting.
For example: A number of servers run together in a peer-to-peer fashion, while clients connect to the best-available server.

[JWPlatt]

Teedyo, you are correct on both counts. The load is why I put into my specifications that it be intended for small groups, not serve shard-like numbers. And, yes, you were naysaying and off topic. So stop it.

That's the spirit, Trylon!

I mean it. Don't bother with saying it's too hard or too this or too that. I was very impressed at the speed of which the abiltity for offline MOUL was released. That's the kind of thing the GoW is capable of. I registered and made the home for this thread here instead of MOUL because I expect more from the kind of people who would be part of the GoW; to be the kind of talented and gung-ho crowd who would rather jump up and ask how it can be done rather than say how it cannot.

[Lontahv]

I think we should start fresh with a lot of stuff, we should not try to hack poor plasma to death and rather work to make something legal.

If we do make some kind of fan-run Uru-Style MMO, I would push for some kind of dynamic-P2P, a hybrid. The way it would work is, if someone has a fast upload speed then they serve their ages(avatar movements/emotes/chat/etc.) the public areas would always be served by the main servers. Just think of the server being a super-client, rather than a server. It would be beefed up in the server parts and there would be no GUI.


I imagine all the file serving to be P2P(much like bit-torrent). And with the official server being just a super-client, it would download the age and then serve it, just like everyone else.

There seemed to be lots of people who could run UU servers from their homes. I don't have enough bandwidth to do that, but those who have lots of upload speed--they would be the ones that could run an age, if not a few ages.

The idea I have for age creation would go like this(after age is approved it will be added to the library, in which case, the official server downloads and serves the age):

You link to your new age then, you want to have some friends come(the only instance of the files is on your computer). So, you give them invites. They link in and get the age served to the regardless of what your upload speed is. As they download from you, they are added to a list on the server, something like this:

<agename><ownerID>
<ownerID><download speed/upload speed>
<PlayerID><download speed/upload speed>
<PlayerID><download speed/upload speed>
</agename></ownerID>

Now, when someone comes to the age again, it checks upload speeds and then looks up the player with the fastest upload speed. If this player's is offline it goes to the next best speed and so on.

Now, the worry about P2P is that this could tun into 2ndLife, not because 2ndLife is P2P(it's not even close to P2P) but because with P2P comes a lack of buffers, you are actually contacting another person direct(EEEEP ). There would be no laws about age's content until it got to the main-stream(nexus-thing, which means to the main server), this is bad unless the laws are very strict and the Maintainer's tools powerful.

By powerful I mean POWERFUL.
Such as...

Being able to download and play an age without invite or warning.
P2P lockdown if age has bad stuff in it(no one would be able to use an invite to an age that's locked down).


Now, the problem with P2P-ish stuff(like MOUL) is that open-source is not really an option.
Well, as long as you don't try to connect with your modded client you'll be ok.

This is why 2ndLife has to do this:

"Your client must ask the server if it's ok to breath."
-Lontahv

Yes, it's true. That is how it can be open-source.

If you can mod your client's code, you can send viruses to others and, hack the servers.

I think that the code should be open so people can work on it(and test it in another grid/server/thing/universe).
But, every time you try to connect is does this:

You connect to auth
Auth sends you an executable and it checks your files and client.
The executable sends back an MD5-checksum to the server and checks it to clients past and present, or at least the main files.

You can't have P2P and have it be super open, it just doesn't work. Ever wonder why Skype has no files, just a few plugins and an executable?


Sorry for the long post.

~Lontahv

[BAD]

Anything is possible, however that doesn't make it practical or able to be done easily.

I'll ask that we refrain from telling other how to post. People are entitled to their opinions and on this board they are fully allowed to nay say, or disagree with others opinions as long as they remain civil and respectful.

No matter how passionate you feel about an idea, blinding yourself to derision of your idea won't make it any easier to do.

We could spend time creating a peer to peer network for Uru. It would only require writing completely new netcode to run along side the client. I don't know if their is netserver code out there that is peer to peer, available for free, and compatible with Uru's code, but I wouldn't bet about it.

So my suggestion is that instead of asking generally for a how to, that you seek people individually who know how to create netcode and alike and recruiting them to make it happen. You can use the GOW as a focal point and we offer space for any ideas that we see as interesting and helpful to the cause of age writing.

I just also want to point out that this is a working forum. We did not set up this group just to ponder about how to make ages or software for uru, but to do it. So if you want something done, you have to work at it to make it happen here. We entertain a lot of suggestions on this board, but we are a loose group, not a machine. People will work on what they want, when they want to. Your job (as a person with ideas) is to find the people who can do what you want, and convince them to help you out.

[JWPlatt]

BAD, if you believe in naysaying, then have my account deleted. I'll get the hint and I won't be back. I want to be part of something constructive. I will not have tolerance for those who only post to beat down an idea. It's a waste of time because things won't happen without support anyway. If something doesn't fly, at least we know it wasn't for lack of trying or because someone refused to allow the light to shine. As far as I know, it is common practice to disallow off topic posts, usually in accordance to the original post and the original poster who can act as facilitator.

If you believe, however, in criticism and review which then offers progressive solutions and alternatives to an idea, such as Lontahv did with a completely constructive post, even though I am prepared to cordially disagree to some extent and support my own concept, then I'll be back later to do exactly what you said since it should be obvious to anyone that I posted for exactly the reasons you stated. That means saying "I don't like that, but what if we did it this way? It might work even better." I think that fits your views of allowing dissent, and that's what I want.

If I seem short, it's because my time is short. The absence of Uru is giving me time to do more important things in real life. But this is something I am willing do, or organize, or lead, if that is made possible by a positive attitude and people willing to go on an expedition. I'll be back with more, or I won't, depending on whether this Guild, its forum, and its representatives choose to respect my wishes for paying no heed to naysaying. I have nothing to lose at the moment, but I would certainly like any effort to count for something as time goes on.

[BAD]

Who is beating down your idea? Why are you getting defensive?

The ultimate decision about if you stay or leave is in your hands. I nor any admin or moderator will delete a persons account just because they may disagree with our philosophies.

If your looking for an argument go elsewhere. This is not the place for it. You can PM me if you like for an in depth explanation of my beliefs on posting and forum management.

Now if you want, we can get your thread back on track.

Does anyone have experience with writing server software and would like to give this a try?

[Paradox]

It's probably possible that the Alcugs server code could be updated, but it would likely be easiest to start fresh and fully implement the server using the new data that we have as opposed to the (lacking) information we had in 2004.

[Tsar Hoikas]

I do have a few ideas on how this could be accomplished, but it's a really involved process, and I'll need to do some research into somethings in Plasma that I've never used before so that I can determine if my ideas are plausible.