Something to think about...

[email]hughperkins@yahoo.com[/email]: Oh, I want it OpenSource
[email]hughperkins@yahoo.com[/email]: The whole point of OSMP is that Linden are strapped for dev resources whilst being surrounded by dev resources
[email]hughperkins@yahoo.com[/email]: It's like dieing of thirst in a swimming pool.
eggy: LOL
eggy: I will so quote you on that!

im lost.

Which part did you not understand?
Hugh was just commenting on how ironic it is that LL is strapped for developers when they are surrounded by them. It's funny. I dont see SL being open sourced any time soon, but I do think it would be fun to poke around the code instead of whining about how much the inventory sucks.

aghh i get it now im dead brained i think my mom dropped me on my head when i was a baby and stomped my head.

Here here!

This is ABSOLUTELY the point.

Everyone should email philip and tell him to check out this thread immediately.

This can NOT be ignored.

There are probably 100s of very very able developers that are probably just dying to do things for Linden Labs for free but they are not letting them.

Why not just open source a part of the engine? They gotta leverage the community much better.

In fact, it's so good I'm going to use it as my signature for awhile.

I agree with this, to a point. I believe LL should open-source the LSL bytecode structure as well as little things like input interfaces, etc. Cory Linden has said he would outright give the LSL bytecode structure to someone if he felt they were competent enough to write a better language than LSL.

Hell, I would request the bytecode structure myself, but the bulk of my experience is in enterprise applications, databases, and architecture rather than low-level stuff like writing a compiler.

From a business and money perspective, there is way too much risk associated with going open-source with much more than that, so don't expect it anytime soon. Even with NDA's and all that junk, the code is likely to get leaked to the public and that is not a chance they can take.

I think a restrictive lisence would be work out fairly well. Make it so that commercial or for-profit use of the code is against the terms, and nearly anything else they think they might have a problem with.

Great new things would pop all the time. llCreateNotecard would be nearly instantaneous. A linux port would be underway. Syntax highliting in-game would be available. Alternative languages would have been started. Outside modelling made possible, alternatives to poser made available. Who knows what would be there. Open sourcing the game would be a gigantic leap in the right direction.

There would be problems with security (people finding ways to exploit the system more easily, and people even adding backdoors of their own to it), copyright infringement (people using parts of SL's code in other projects, or using code from other projects in SL, people running their own SL servers for free), and then you would have the occasional developer who later changes their mind and throws a fit because they wants their work removed from the game (even though they can't do anything legally, it's still bad P.R.).

I can see several problems from making SL open source.

Devs stealing code from this game, or using code from other projects in this game.

Devs finding bugs to exploit for their benefit, or even implementing a backdoor of their own.

People complaining about how they shouldn't have to pay for a game when people are working on it for free. (Many people don't understand that it costs money to maintain servers.)

Devs who later change their mind and throw a fit because they want their code removed.

People running their own SecondLife servers instead of paying Linden Labs.

People could compile clients that would remove some of the limitations to protet other users. (For example, a client that could give the plaer the key for any texture or even save the texture as a tga)

Would take a lot of time but parts of SL could stay closed source, and stay in a compiled library, and parts could be left totally open for dev's to search for bugs, or add features. Like Quake

Certain parts would lend extremely well to open source, such as an external development tool that doesn't require the bandwidth / fps overhead.

Yes, open-source griefing would be quite a pox upon the land. I'd really not want to see that, thanks. Imagine what some moderately-knowledgable goon could do with the source, if their intent was malicious.

Bad idea.

And before someone chimes in with the open-source +10 shield of 'but we can code defenses' cry, just remember that it would be a continually escalating war - is that really what you want?

Well then, maybe the solution should be something in the middle.

Some parts should be moderatly safe to open source. How about LSL for example?

(Required Forum Disclaimer: If you dont like this example, dont stick to it, call it silly, and use it as a judgement to call the entire idea "stupid" and "unuseable". Try to find another example that is more to your liking that may work in this frame. This is all about a concept.)

Well as we all know, the wisest way to introduce something of this magnitude would be gradually. I would first like to see LL move to a more modular plugin-based architecture that would allow us to write, say, our own LSL editor or replace whatever handler they use with a simple call to our favorite editor along with a trap to pass on the code to the compiler.
Writing our own interface would be spiffy. It would enable more complex ways of interaction. We can currently write whatever application we want and have it communicate with LSL scripts over email and XML RPC, but it wouldnt be very fast or integrated with SL.
I'm thinking of something more like being able to trap mouselook mode and overlay the SL UI with a HUD.
If LL ever starts opening their source, it would probably start with the "hair". I would love to work on that horrible inventory. I really would.

Actually, Maxx, it would probably not be as bad as you fear. SL works on a thin client architecture. The client is supposedly just a viewer of content, like internet explorer. You are very well capable of coding your own browser, but that does not mean that you can deface websites with a custom browser.
I am obviously not aware of every little implementation detail, but IMHO they could probably open source the whole client without a hitch. That's what thin client architectures are all about, right?

Problem is eggy, we won't know how 'pure' that scenario is until the source is open, and once it is - if there is some data that *is* trusted from the client, it could mean a few unpleasant things.

If not, then yay, have at it.

Personaly I agree with Eggy - the only possible trouble eminating from opening the whole this is copyright infrigment - mostly from having access to textures and other media that you are usualy unable to save.

It's not like you can't already do it
Stealing textures has always been possible in SL. There is no way to prevent it.
If you can see it, you can copy it.

What about clothing?

If you get someone to fix the lag, nobody will bitch.

The thing is LL does hire from the ranks of players. I'm suprised nobody has relized this and more over who they are. While I would love a job at LL i relize i'm still under the URU NDA (which i can't lay my hands on to double check) which i'm pretty sure has a 1 year exlusionary statement from working in the gaming industry (why couldn't i have found the SL beta instead? ).

I was in the Uru Beta, too. It may very well be that since Ubisoft discontinued the Uru multiplayer stuff that you may be free to pursue LL employment since Uru no longer is competition in the MMOG market.

~_~ not exactly true...
Until URU

Clothing is even easier to steal than regular textures. But I wont bore you with the details. Not publically, at least