Second Life To Open Source Server Code
mrspin writes "Having already taken the timid steps of open-sourcing the code for its client software, Linden Lab has confirmed that they'll be going the whole way, and will soon be opening up the server code for Second Life. This furthers Second Life's ambitions to be a fully distributed 3D network — built on interoperability and not owned by one company — a bit like the Internet itself. ZDNet's The Social Web asks: 'who will be the first to offer Second Life hosting or use the server code for their own internal purposes? IBM would be an obvious candidate, perhaps offering corporate Second Life services. And for the rest of us? GoogleLife, free virtual land — ad supported of course. It's certainly a possibility.'"
Once its all open, guess who will be in the line to download the code and get programming? Yep, the pr0n industry!
Support NYCountryLawyer RIAA vs People
I always wondered what the point of Second Life was. I don't get it at all. Going to a virtual world and playing a character? I understood that was the venue of Everquest and World of Warcraft.
But, if you're suggesting there's untaxable income to be made, then perhaps I've been looking at this thing completely wrong.
Still, what in the world are corporations doing playing games? I won't understand that... unless it's the same money thing.
Oh, say does that Star-Spangled Banner entwine / The myrtle of Venus with Bacchus's vine?
Distributed between two data centers, that they control, Linden Labs can't manage better than about 95%-98% uptime. Inventory items and sometimes even portions of entire sims regularly go into the bit bucket when the data centers have connectivity issues.
And to this mix we will add a heterogeneous server base, geographically dispersed, with network connections of unknown reliability?
Get ready for a Second Life experience akin to IRC in the 90s.
Does anyone else think that this could be the beginning of "The Metaverse" as envisioned by Stephenson? (see Snow Crash)
Could someone build at least one world in which you purchase "land" based on the power/CPU requirements of the land, rather than its (virtual) area.
The "necessity" of getting a return on your per-square-meter fees causes SL to be overtaken by casinos and brothels. Make the fee dependent on something of actual economic value.
Just thinking aloud, don't have time to do it myself
My turnips listen for the soft cry of your love
This sounds like another key step to making the web how some had originally envisioned it. Back in the day when VRML was born there was the idea of creating virtual worlds where we walk to a clothing store like we would in 'first life', of course the technology wasn't quite there yet... Now with Second Life we're a hair closer but as long as proprietaryness is in the way that's just one more silly road block. Personally I want a Google Earth version of second life so I can travel the world and see a decent recreation of it made with actual photos and 3d satellite imagery, I also want to recreate my college campus and attend class virtually...
define "first time they saw one."
Did your simple minded idiot see their first website in 2000 or 1990?
Progression is always logical in retrospect.
I'm not so sure. I think it's a last ditch effort to boost their users. I mean they have been spamming the hell out of SL for the last couple months/years and I don't think it has worked. They just never got critical mass for their system.
The thing is, while many people have checked it out, not many have stayed. Like Vonage they must have a huge turnover rate because eventually people realize that compared to alternatives it sucks. SL seemed pretty damn buggy (both in terms of performance and plain bugs) the times I tried it. The development environment sucked too. It doesn't have value as a game so you're left with "Why am I wasting time with this?"
Maybe eventually we will see something like Second Life take off but I hope it isn't SL unless they do a major rewrite.
I'm interested in the prospects of a distributed MMOG. The world is split up into regions. Instead of an authoritative server doing the processing tasks and stating what is and what isn't, specific clients are granted authority about the regions. Multiple clients, that is. Everyone listens to all of the authoritative clients running a region, and decides what's true based on a simple majority vote. The key is that clients don't get to pick what regions they're authoritative for; it would be distributed by something like a Kademlia network. The only way to take majority control of a region would be a massive DOS, kicking off a large percentage of the network.
Bandwidth requirements are certainly notably higher (due to the fact that there's not one authoritative server per region, but several), but on the other hand, it's everyone's bandwidth being used; no one company has to pay for it.
It's actually more complicated to this, since the loads for a given region will vary greatly. You'd likely need realtime tesselation and merging of regions to keep the loads reasonable for a given client -- either that, or very small regions, with each server running a large number of them (when the load gets too high, a server starts offloading some of its regions). Still, the basic premise seems feasible.
Present day. Present time.
Err... I'm afraid not. Take a look at Croquet's design. It's an old fashion P2P protocol in which each user forwards only their inputs (e.g. keypresses) over the network to other users. Every user must run the full simulation locally, making total network traffic and resource usage O(n^2) with the number of users.
This cannot scale to more than a handful of users. Croquet's design is fundamentally incapable of being "massively multiplayer". I would say that that makes it not "a better choice than Second Life" in quite a few cases.
(Never mind the fact that Second Life is a huge, proven, production system with hundreds of thousands of users whereas Croquet is an academic experiment.)
One of my professors is toying with the idea of working with SL for some lectures. The lectures are still in thr real world, but the assignments revolve around building stuff in SL. For example, one assignment might revolve around designing an automated "assembly line" that reacts to certain events Probably the biggest gripe he has with SL so far is that not everything is possible - he's currently trying to get a Petri network simulator going.
Having access to the SL source code would enable him to set up his own server at the university; that way we'd have much less (network-induced) lag. Also, we wouldn't have to worry about being interrupted by walking penises.
USE HOT GRITS WITH STATUE OF NATALIE PORTMAN (NAKED AND PETRIFIED)
That's not true at all. There are two full time employees at Linden Labs that are responsible for maintaining the GPL licensed releases and they are kept in sync with the production releases within a few days. Right now, April 19th 2007 @ 1:40PM, the GPL releases are in sync with both the production grid and beta preview grid.
OpenSim is a great project, I work with those guys frequently, but contrary to popular belief it is not a child project of libsecondlife. It is an independent project that happens to use libsecondlife. It's also an important project because the simulator code will likely be released as GPL, while OpenSim is a BSD licensed project. I'm not a big Croquet fan, but I'll have to check out Ogoglio.com and Verse.
- John Hurliman, libsecondlife lead developer
I want it to work like this: I buy a small house in Second Life, and anyone who comes through my "door" ends up on my server, and the inside of my house is hosted exclusively on that server, and I can control who comes in and out. And it can be HUGE on the inside, a la the Tardis.