h a l f b a k e r yExpensive, difficult, slightly dangerous, not particularly effective... I'm on a roll.
add, search, annotate, link, view, overview, recent, by name, random
news, help, about, links, report a problem
browse anonymously,
or get an account
and write.
register,
|
|
|
This is a bit of a ramble, but I have a lot of the ideas build up over time and I'm listing several here.
The idea is basically to take a client-server architecture and spread it among all connected computers. This would be used in a 3D online world. Unlike traditional distributed solutions, each
computer would be a dedicated host to a portion of the world. The world would contain "link zones" to a list of other servers. These link zones would consist of coordinate boundaries and a list of links. The links would be evaluated in an order specified by the link zone (random, in order, etc.) and it iterates through the list if a server is unavailable until it finds a good connection or it times out. On timeout it will only display the local zone as if there was no link zone in place at all. (Or optionally a placeholder object/texture)
To the user, instead of displaying the current server in the link zone - the remote server would be displayed. This would allow you to walk from one server to another.
This would allow for many small-scale servers to be connected into one giant world. A central server could be used for authentication purposes, global scripts/commands/etc. and possibly server-server transitions.
Content would be provided by both a common content package as well as dynamically downloaded server (or namespace for several grouped servers) specific content.
People would be able to define permissions on their servers to limit the number of people connected and who can enter it.
===
Example: A simple 2 connection
Person logs on and enters their locally hosted area (home area). The area outside their home is marked as a link to server 2, so out the windows/doors they see server 2. When they walk out the door they are seamlessly transferred to server 2 and server 2 contains a link to their home area which they see as the house. Could also have the inside of the house only as a separate server so it's bigger inside than outside.
===
Note: Server-Only would also be able to be run for remote server purposes and the idea would be open source to allow modifications to zones (no grav here, etc.) and different clients to be written.
Technical limitations: There would be a limit on the number of adjacent zones that could be displayed at a time
[link]
|
| |
Yes, but Stephen Hawking won't let you turn off gravity in
your neighbour's field. |
|
| |
why have any "home" software at all? Time's a commin when having your own hard drive is going to be for cabin crouching Luddites and your data is going to live "elsewhere". |
|
| |
It's so you could have your own customized space (like an evolved version of the web... each server is a website), and there would be no need to run a centralized server at all - so no cost to run it (other than devel costs and possibly the authentication server). |
|
| |
As for data elsewhere, until they get speeds fast enough that there's virtually no difference in local vs remote there will be hard drives. Now when that happens, not only will you have no hard drive - you won't have to do any computing either and will only be running a display of some sort. |
|
| |
Welcome to the concept of dumb terminals, sagron! |
|
| |
Isn't this just the internet? |
|
| |
When I switch on my pc, I'm on my 'home' machine, if I run a web server on my machine, then it becomes a node that other people can visit - I'm not sure, but I bet there are settings that allow me to limit the number of concurrent connections that my machine will accept. |
|
| |
Then, if I want to look at the contents of another machine, I can open my browser, and type the address of that other machine, and go and see what's on there. If I want to go back to my home machine, I can. |
|
| |
If I'm looking at a website, and its author likes the content hosted on another machine, they can provide a link to that other machine, which, if I click on it, will take me to that machine. |
|
| |
And so on - how is this idea any different to what we're using to have this discussion? |
|
| |
Oh, and 'cloud' computing is bollocks. Data has to be stored somewhere, by somebody - the fact that people can pretend that they don't care where doesn't really seem that paradigm shattering. It's just a pricing model. |
|
| |
It is very similar to the internet, something like an extension to it... Like you said, each server is equivalent to a web site and each adjacent zone is equivalent to a hyperlink. The main differences
a) it's a 3d environment (obviously)
b) it's seamless between them (servers) |
|
| |
And yes, I've programmed with the predictive streaming capabilities but the biggest differences between that and my idea is that my method allows for areas to be bigger than the surrounding area (inside of house bigger than outside idea) and that the servers can be hosted by anyone and in fact can be dynamically changed during runtime |
|
| |