will SL succeed?

AW can have as little as 30m up to 200m visibility range for objects (terrain is up to 1200m which is also the max far clipping plane's distance). Object count varies but there are cell data limits that restrict how many objects can be in a 10x10m cell. However, objects download and render pretty fast--about the same, if not a bit faster actually, as SL objects (but that could be because SL's objects fade in--but not out!--and AW's objects don't).

Refer to AW's help for more info.

The problem with SL's full-screen mode (which looks EXACTLY like it does windowed, by the way) is that it takes a few seconds to restore after switching to another app (at least on Windows anyway), so now I always run SL in "windowed" mode even though it's maximized to full-screen (and doesn't look any different). AW's interface isn't as tight as SL's but SL's gets cluttered up too easily (which I've proposed a redesign of).

Regardless, there's no reason one couldn't use either the integrated modeller or an external modeller to create objects.

It's Hole but check the link for more info (and it's not my world, but Mauz's).

javascript.

They're already using XML in building the user interface. They might want to consider borrowing more than Gecko from Firefox and integrate something like XUL into the client. Just don't let scripts or webpages auto-install it or request its installation... make the installation a deliberate action (like, opening a file on the desktop) on the user's part to avoid them getting caught by the "reflex approval" problem.

The combination of XML and Javascript for user interfaces is getting a lot of traction: Firefox, Konfabulator/Yahoo Widgets, Dashboard, and of course AJAX.

Personally, I agree with Daniel James when he says that SL needs to be open source in order to succeed in the long term. LL does not yet have a consensus on whether or not to really make a transition to OS, however I would guess that a vote today would see the pro-OS camp win.

In any case, I think an OS move today would be bad (as in "too early" -- the whole thing probably would go the way of Netscape, not the corporate overmismanagement part, but the need for a Mozilla-like rewrite. The main problem IMHO is the network protocol between the SL client and the SL servers. If we were to make that more stable, robust, and flexible then I'd promote immediate open sourcing of the SL client (well, once the few proprietary libraries had been ripped out). 1.9 will make some changes underneath that may make it a little bit easier to overhaul the network protocol, so it might actually happen someday.

Yeah, that sounds like a neat idea.

Though I'm hoping that SL eventually will evolve into something that can used as the desktop environment on my computer, where I could have all my applications running on something similar to a sim that would be my desktop, that would not be connected to anything other than my local computer except for by one-way portals out to other places. A bit like a mix of OpenCroquet, Sun's Project Looking Glass and Second Life.

This seems like a funny thing for me to say as a long time Open Source advocate, but...

How would you handle the digital rights in an open-source Second Life? I can see ways to use certificates that would allow someone to create their own SL client that could interact securely with their server, so you could set up your own "second life" grid, but you wouldn't have any non-full-rights content from the original grid, and you wouldn't get to connect your client to the "real" grid.

Do you guys really have a scheme to solve the open-source versus DRM problem?

Yes. I'm setting myself up for flame, but stop thinking money money money. I would like a 3D VW to sell my RL stuff as well as a world to host my digital creations. The only reason i sell stuff in SL is because it's the ONLY way I can afford SL. I would go tomorrow on a server with no DRM and rely on donation ware if I had to. If I can host a server it wouldn't cost me much more to run that my current web provider charges me.

I honestly think this current rush of being able to make silly money from selling virtual goods is probably a temporary thing. When the boom is over it will settle down to something more realistic. Maybe not, there is a case for creating DRM for virtual goods, look at Intels plans.

Open source VWs will happen eventually it's inevitable. People want it so it'll happen.
Maybe some closed banking system could be created, maybe that's what LL could provide !

As an example of what I'm talking about. Look at what's happened to the music industry. People now create and distribute music for free, they make money by selling touchy feely real stuff, spin off products from the music. A minority of artists find it harder to survive in the new world because they've been molly-coddled by their labels for so long. The majortiy of artists now have a way to be heard that never existed before.

I'm not. Linden Labs is. That's why I was surprised to hear that kind of comment from a Linden... because I don't see how the current SL model could continue to work with an open source client (let alone an open source server) and I'd really love to hear why Andrew thinks I'm wrong.
By no means am I putting that down. I would happily forgo the whole Linden Economy... but I don't think Linden Labs would want to.

And certainly long term there is going to be an open source virtual reality that's at least as good as Second Life. But... Andrew said "The main problem IMHO is the network protocol between the SL client and the SL servers. If we were to make that more stable, robust, and flexible then I'd promote immediate open sourcing of the SL client"... and I really found that quite surprising.

I'm gobsmacked and very happy to be proved wrong! Good news to hear it's still on the table

Active Worlds has its own lacking features, you can't customize your appearance, only choose from a premade list of avatars which are defined by the world, and you can't carry around objects or script them to follow you around everywhere or talk to you. Essentially you are playing by the rules of the world owner, you cant build stuff out of any pieces except those that are sitting in the object yard, you cant buy and sell stuff, sure there is capability for some of those things like scripting but it just doesnt happen as much in AW, the place is dominated by lego-like structrues and everybody wearing the same things. Importing 3d assumes that you own your own world, sure it is easier to host in AW, but in SL you can play with objects while in other people's worlds. Each world you go in you have to pick from their avs and objects instead of being free to customize, unless you are firends with the people in control of the place then sometimes they will use your custom stuff.

I dunno, I feel like SL is a releif from the limits of AW, importing 3d created externally would be the killer feature for me though. I have this folder full of VRML files just waiting to find its way into SL. Integration with existing structures like email and the web would also be nice to have as well, but SL sucks up all my RAM as it is, no need to bloat this program any further.

Uh, so get more RAM; it's cheap.

Anyway, AW world owners can create avatars (or you can give them yours for them to put in their world) but, you're right, SL is definitely better than AW in avatar customization for the most part. AW avs can still be smaller/larger of ANYTHING (or just an invisible empty clump as I used to be) and not be limited to size constraints and having to cover/fold up a severely deformed avatar skeleton like SL's avs are.

Yes, AW's action commands pale in comparison to LSL but at least AW's rotate command can be synced and AW accepts URLs for pictures, textures, and sounds--unlike SL! AW's lighting is also more advanced: type ("point" or "spot" compared to SL's "point" only), radius independent of object size, various effects (blink, fade in/out, fire, flicker, flash, pulse). SL can be scripted to do most of the effects but it takes a bit of tweaking to get them right--and even then SL can't update the lighting fast enough because the server tries to do it all instead of letting the client handle it for each individual user. No, the effect might not be syncronized but for most lights--especially ones on static objects--it just doesn't matter. AW can also handle more lights than SL--in OpenGL mode where it renders the first 7 lights in hardware and the rest in software. Whereas, in SL, it, apparently (though I'm still doubtful) renders all lights in software, which is stupid.

More on modelling, object surfaces (ambient and diffuse) can be adjusted to reflect light differently, unlike SL's fixed object surface reflectivity.

Overall, SL is much better in MOST respects--but not all...

Basically, AW (when I messed with it) is themed around making a 3D website. People own a world and make it to their liking.

SL is themed around doing stuff. Limits exist, but some of these are there to ensure your stuff plays nicely with other stuff, and still fail at times, at that.

AW offers total control; SL offers collaboration.

No, that whole 3D website thing was only introduced a couple years ago (and never succeeded, along with anything else AWI tried to promote AW to be--with the possible exception of virtual education, but I doubt even that's been very popular in ANY 3D platform).

AW also offers collaboration: I built Cubed world with 2 others for a virtual education contest years ago. Many worlds have been built by multiple people. SL has more things to do, gamewise, but AW has its fair share: bingo, cards, chess, even an RPG like Dark Life--only in a MUCH larger world!

No, we don't have a scheme for solving that problem yet, but that problem is not set loose by an open source client. Were an open source client to arrive on the scene then an open source server would be inevitable and that problem would have to be tackled. It may be that there will be a niche for 'trusted' servers on a big grid that also runs wild open source servers -- content and services that anyone wanted to protect under DRM might never be able to leave their little trusted pocket of the world. An intrepid explorer could wander into the wild grid, but might not have access to all objects in his/her inventory. Meanwhile two trustworthy continents could agree to trust each other and thereby each would be able to provide DRM'd goods and services that the other provides.

If I have the source to the client then what's to stop me from saving any animation, texture, sound, linkset, or combination thereof to disk in a format that could be uploaded with my name on it? The only asset that I couldn't copy would be scripts.

I actually came to SL from Puzzle Pirates, because fame and being a political powerhouse began to be too much for me. Naturally, it put a large smile on my face to see Cleaver (James) commenting on SL. In regards to the subject however, i'm not sure what to say that hasn't already been said. Linden Labs really has two options. Take the smart, time-tested route and OS it when they feel it's right or continue to do what i feel they have done and change stereotypes. Break barriers.

The latter will take a whole new definition of "cult-like MMO community", and really could kill SL. The flipside would be significant.

This entry on Cory Linden's blog talks about those issues. Fundamentally stuff like that cannot be protected using only DRM technology, even under a proprietary client because the client can be hacked.

So I shouldn't worry about LL sticking a snooper in the client that'll freak out if I use an OpenGL sniffer to poke around the mesh?

No, I don't think you need to worry about that.