Open Source Second Life - Prop: 1483

The InnerLife team would be interested. We're trying to get a good data input model for biometric data to control events in SL, and XML-RPC just has too high latency for what we're doing. Being able to modify the client and server (which we would then have to host) would really help us. Perhaps if we did the development work, someday our changes might get folded back into the main source tree and even supported on the LL servers (if they continue to exist).

That being said, it would pretty much wreck the content creation economy in SL. More than it is already, I mean. And making the servers available, unless there's a way to travel from one to another, would fragment the SL world irreparably.

Maybe that's inevitable.

neko

The Second Life community is not really based around anything except "Oh, jee we're in second life"

People of like interests always find ways connect, and I think communities would spring up like IRC channels or usenet boards.. A large collection of servers would be able to host even larger and diverse communities..

That's what secondlife has now. It's pretty large and has a good ammount of diversity. Niches could be filled in around the larger networks like Linden Lab's own.

that looks like a distorted version of the Debian definition of open source.

I don't think it's at all definitive, simply because it's got a huge inaccuracy.
Software linked with GPL libraries inherits the GPL when it's distributed, unless the GPL library is merely one implementation of the API it exposes, and the program doesn't depend on any extensions unique to that library. It doesn't have to be distributed with the GPLed library, because it incorporates GPLed interfaces and header files in its build.

The derivation clause is worrisome to me, because it's basically a copyright on the API, not on the actual code. I'm surprised they were willing to stomach this, given that so many of the founders of the FSF (including Stallman, author of the GPL) were also instrumental in creating the LPF... and the whole point of the LPF is to oppose software patents and interface copyrights.

Anyway, it sounds like whoever wrote this document is mixing up the aggregation clause with the library clause. They're separate exceptions to the derivation clause, and operate independently.

I personally fully support this idea - at least the server part. One of the biggest features I would like to see is the ability to build offworld (yes I know there is a blender project that allows this - but personall I prefer the simplicity of the inworld tools). Being able to run my own server to allow me to build stuff without the hassle of running out of prims would save me a lot of time and effort. I could even have several server setups to work on different projects. It would quite simply be a massive bonus to builders. Yes there would be problems with assets - probably texturing your builds would be a problem unless you could dump the textures you have to disk and load them into your off-line server.

I heard this being discussed - by a LL representative in a secondcast (or townhall) presentation and he was asked about this. In particular he was asked if providing the tools for people to be able to run their own servers would be a threat to LL, and the response was no. SL has built up a culture/set of cultures over the time it has been running, someone starting up another server or set of servers would have a hard time competing with SL - but the benefits to content creators would be good.

-
Timmins

Thanks for your support on this proposal ;0

Some related threads from General:

The future of SL: Why not help make it happen!

Should Linden Lab Open Source Second Life?


I've been trying to keep discussion going on these topics, but interest wanes with all the drama of current events.

As a professional developer, I feel compelled to put in my 0.02 by saying that while I believe that Open Source (in general, not just FOSS) is absolutely fantastic and I fully support it (even participating in a few projects as time permits), I strongly feel that there should not be pressure (by community or otherwise) on *any* company to release their code to the public. I have myself been the recipient of a moderate amount of criticism for not releasing my source, which just mystifies me to no end.

So far I think that SL is a great game, notwithstanding the warts, and am perfectly content with closed source. I would, however, love to hear more stories "from the trenches" and would like to see more videos on Google TechTalks. It was the Google TechTalk video on Mono microthreading that got me into SL in the first place Any Lindens listening?

P.S.: I would also love to see a licensed/approved API mechanism where third parties could provide "enhancements", similar to how you have to get approved to create an eBay-licensed application or an Authorize.NET payment solution.

I support the "licensed/approved" API mechanism. We need as many barriers to entry as possible in creating add-ons to protect my profits- *AHEM* everyone's security.

I would be ok with the CLIENT being open source. That would allow people to change the GUI a bit. And SPECIAL releases that have extra features. BUT anything else would be a BAD idea. Allowing anyone to host a server and change it's coding would give us the following problems.

1.) More power hungry people who think their gods and ban people just to be a total asses.
2.) Lack of stability on servers hosting sims. Atleast with LL i can trust that my sim will be here till i can no longer pay for it, but if Joe Pentium-1 hosts it on his mommys emachine i douht i'd be staying with him...
3.) The Linden Dollars will become worthless... Meaning everything i'm working for in SL will be worthless aswel. I definatly would not put any pride into my work after that...
4.) Nobody would know which technical suport team to rely on, since everyone will be tweaking it in their own way.
5.) Griefing would take a new level. People too average to know how to program, will rely on friends to compile and just use as is. Which meens NO development. I would probably have to have several different clients to access other grids. I would rather be on a server by people who put Security and Privacy measures in their work, limiting abilities to grief.

Probably not.. Any self hosted servers would likely not be connected to the grid and I suspect would be more like personal sandboxes. I agree that allowing people to host their own sims and connecting them to the grid would be a bad idea.




Same as point 1....



Sorry I dont get this one at all........ How would an OS server cause this?



Same answer as 1 and 2 - if the server was not public - or at least not connected to the main grid this would not be an issue.




Not sure how a few people not developing something will mean stagnation.... Most of the people who use Linux for example are not developers and use it "AS IS"... Does this mean that there is no development? No!

As for the different clients for different grids - this is possibly true... But if you onle want to connect to the real SL grid then there would not be a problem.

Yeh, I'd LOVE to be able to host an isolated sim in my house.

Open sourcing the client would give us far more than just little gui changes. Well, let us just say that there would probaly be a total GUI overhaul. And like you say special features.. Hopefully not as entire new client releases, but through a plugin interface.. command line clients.. text based clients... 2d clients.. Imagine that ;0 a 2d client that uses your profile picture as your in world representation.. Truth is though, things like that are already starting to happen through libsecondlife...


Why does it matter if they own the server? If they are not the kind of people you want to associate with, why worry about what they want to do with their land/servers?

Also, not an issue. I could host my website with the some guy running a webserver off a coke machine, but why would I do that? There are services that guarentee 99% uptime and have quality tech support already.. Dedicated servers cost like $100 a month and up. Imagine if a respected hosting company set up a grid of shared server hosting where they sell sims by the parcel.. Just like Linden Lab ;0 Now imagine if they signed a contract with LL to get their servers connected to the main grid.. or at least as an island.. Before that happens, you will probably have people hosting sims off their breadbox... You're not going to live there, I'm not going to live there, but I doubt anyone running a server on their breadbox is doing it to sell real estate.


What exactly is the L$ worth and where does it come from?

Your L$ balance is worth what some other guy who also uses Second Life is willing to pay for it. It's really just a note in some database at Linden Lab that says. "Joe has L$500 / Joe paid Jack L$100 / Jack Has L$100 / Joe Has L$400." The fact that Jack gave Joe 48 Yen for that L$ is irelevant.

Maybe Joe bought a Furry av for his L$100, and then Jack sold his L$100 to Jill for virtual sex. And Jill sold her L$50,000 in virtual whoring profits to Clark for $250US.

Clark used his newly aquired L$50,000 to rent land from Penny's virtual estates.

Penny pays WeHost4U $150 a month for the use of a server which is connected to the grid. Penny collects L$50,000 a month from Clark and sells it back into the market to pay for her value added land rental business.

It's possible that Clark did not even need L$ though Penny accepts them just like she accepts credit card or paypal... whatever works.


If the user has an issue with his or her land. They contact their provider. I hope would know who you purcased your land from...

If Penny owns the land your rent or purchased.. Penny is your support contact.

If you own a server, contact your hosting provider.

If you purcahsed land from Linden Lab, contact them...

If you don't know who you bought it from maybe you shoudn't have purcahsed it.. None the less support of some kind should be available to contact in the land menus if that were the case.

If your client screws up contact whoever made it, or ask in some forum......

Edit: This is just like using any website service or browser...



I don't see how this equals griefing... Sure as the system is now... the client has far too much trust, but if the project were to be open sourced, all the vectors for such grief and exploitation would be in the purblic view.. And when your bugs are public, you better fix them fast right?

The fact that John Public doesn't program or compile his own client is not even an issue either... There is no reason why he should have to compile anything at all. Any project worth John Public's time is going to offer binaries for his operating system that install the client automaticly.

John will open his client and rez into the world(or rez in invisable mode) at whatever predefined location is set by that client.. Perhaps he has an account at Yahoo.com and he wanted to be there in stead.. He just pops open his handy dandy adresss bar and types in "yahoo.com" the client knows to prepend secondlife:// to his URL because he's using a SL client which teleports him directly to the public yahoo world.

From there he is presented with yahoo's destination menu/hud. With the yahoo hud he may search SL or the Web, read his email, link his client to yahoo's inventory system and retrive his objects and bookmarks... It all works because it's part of an open standard. John Public is happy; he just downloaded the latest version of SLIExplorer from Microsoft, and while his computer may now be acting as a spam zombie for some evil internet marketing firm, that's nothing new to John.

This proposition just broke 200 votes and Page 3 of the total votes list ;0 By far my most successful proposition yet ;0 It could use a little Linden recognition though!

I'm wondering if there is any more interest in this proposition?

I'm surprised that, on the 3 subjects posted about that, no LL staff have answered seriously..nor on the 3 voting proposals about that. I'm curious of what they think about that...
Is this a taboo on LL or they are just unaware??? oO

I think their aware, they just don't wana talk about it.

Needless to say my opinions on open sourcing have changed since my last post in this thread. In a related thread, my mind was changed. I'm definatly eager to have someone else work on it now. Linden Lab has proven to many of us that they don't seem to view SL as serous as they used to.

It would be best to just give it away to someone who can build upon it, then let them continue to hack it to death like they been doing. Either they should sell it to a company that can do better without ruining how it currently is any more, or open source it to let the residents fix the problems.

I for one do not know C+. I know how to script in LSL, Frucadia's dragonspeak, and a few oldschool programming languages like BASIC. However i would be willing to give it a shot, if that chance arises. I don't like to critisize, but Linden lab has serously f*cked SL over lately and it's time to fix it.

I've had several serious discussion with Linden Lab about open source.. It's true they have not replied to my posts about it, but they do support the cause. Just look at libsecondlife if you do not believe.. We have a functional library that can perform many of the same functions the official client has, and several Linden developers have expressed interest in using libsecondlife for various projects.

Most recently at SLCC, libsecondlife spoke with Linden Lab about our project and there was talk of open sourcing parts of the SL codebase. Already they have been helpful to our project with several hints and other documentation on the Second Life message system.