Stuck at 15 FPS or "We need Halo Style Performance, Game Play, and graphics"

Not only do I believe this is possible, I believe it is necessary.

While I do not believe the metaverse is meant for games, I do believe that entertainment can provide a foundation of sorts for our world. Movies are being overtaken by games, and SL should be there when it does.

How can this be done? We need three things:

First, we need the client to become a SIM with a cache as asset server. Everything is there - physics, prims, textures, animations, scripts (in bytecode format), sounds, avatars .. everything. Yes, in theory, people could steal the various data - but they already can. I can do screenshots, I can record everything hitting my sound card. True, animations are a bit tough to steal, but they'd be tough to steal if they were in my cache as well. Same with bytecode.

In this situation, the SIM simply becomes a replicator.

And yes, people will be able to cheat. But people can cheat with Halo and it is still very very popular.

Next, we need pre-rendering tools.

Pre rendering tools lets us go through all our prims and textures and alter them to have light and other effects added. Lighting needs to be done beforehand. It adds a very important element to the experience and can not be done after the fact,

Finally, we need performance optimization tools. We need better insight into what's lagging our collective experience.

This is what the customers want.

There is a selective group of people who don't want this, but they are very selective and not particularly willing to pay enough to support an entire company.

We all understansd what you're trying to do, LL, but someone is going to hook up halo style engine to a world where you can dynamically update the content - and you're going to be immediately out of business.

Nobody wants to be stuck at 15 FPS when they don't have to be.

It could also be that we are 'stuck at 15fps' because SL uses a radically different 'level style' than say Halo.

Halo gets BSP and other similar optimisations; SL cant prerender anything; the new 2.0 renderer is pretty much all innovation - LL's creating new technology & research into the area of optimising dynamic enviroments. Stuff that just hasnt been done before on this scale.

By SL's very nature - we cant get these types of features. However, I do happen to agree we need greater ability to find what causes lag in SL.

There also needs to be some form of encouragement for people to do things like use properly sized textures. (Prehaps having different upload fees for different resolutions?), and also encourage texture re-use, etc. Prehaps on a similar vein, some mild feature improvements may also help us combat lag, like the ability to get Self-lit, instead of using up valuable lighting computation when it's not needed.

-Adam

That's one of the best ideas I've heard in a long time

Strongly support!

Isn't there a new renderer in the works..?

Agreed. A little communication would help so much. There are people reformatting their machines, buying new hardware in hopes it will fix the lag and then are surprised to see that nothing changes. When you are more then 3 generations ahead of hardware ahead of the system requirements you'd think there would be a performance increase. Top of the line systems (talking over 3GHZ and Nvidia 6800 or ATI x800 with 1GIG+ of RAM) would be capable of over 15 FPS in most situations. Most people upgrade their hardware to improve the games/programs they play and this seems to be the only program that doesn't benefit from a hardware upgrade.

Regarding the localized cache, Valve pre-released Half-Life 2 for weeks prior to the game being released and to date, this cache has not been compromised. Adding encription to the cache would never hurt performance as bad as downloading the same things over and over. I would imagine that the cache is not on our drives because the space requirements to keep a decent utilized cache would require too much disk space but that is just speculation on my part.

Adam:

Take:

http://www.planeshift.it/

And then add the ability to incrementally stream new content into its cache.


Voila - dynamic world with halo like performance and graphics.

yeah, so builders have to build content offline. Hell, I do most of my work with pen an pencil already anyways.

Planeshift has it's content prerendered; it's not suitable for comparison. Working SL content 'offline' pretty much goes against what SL's core design is. If you want something like that - then your best working with Blaxxun/VRML/ActiveWorlds which focus on offline editing of the actual structures.

-Adam

ActiveWorlds has very poor graphics and very poor physics. VRML? You're trying to be funny, right? As for Blaxxun, I just tried it out. But, again, you're kidding, right?

Anything less than modern MMORPG style graphics won't cut it. Secondlife and There.com barely fit in that category.

Planetshift definitely does.

Dynamic building, while neat, will reach a very tiny majority of people at the moment. I do see it as the future, but everything is about timing.

The way to build the metaverse is to get the eyeballs and then to migrate them into the metaverse. Trying to build the metaverse first, while fun, won't be so fun as you watch an incremental style planetshift run past you very quickly.

what I'm trying to say, simply, is that llRezObject could be added to planetshift. And soon. When that happens - watch out.

Oh, so everyone gets drastically low FPS like I do? =O

I thought it was just my Radeon 9000...

Actually no it cant. For small objects yes - that might be possible; but for big ones, like scenery, buildings, etc, no it absolutely cant. At least not without hitting the performance issues we currently have.

SL's 2.0 renderer is going to be fairly revolutionary here, since it's going to be the first properly researched 3D renderer built from the ground up for dynamic scenery, while maintaining a good framerate.

-Adam

Will SL 2.0 increase the system req.?

I'd doubt it; By the look of things, it's just going to make things faster; but I'd ask a Linden to comment if you want an official answer to that question. It should also be noted the 2.0 renderer is a long way off. We will still have at least 6-9 months before 2.0 is launched (maybe longer)

-Adam

Yeah, let's just remove real-time collaboration and building. It's only the core idea that SL is built around. Who needs it. We should be all about the mass market appeal!

Have you tried playing Halo in order to get Halo-like play? Personally, I wish SL was more like PowerPoint

I wouldn't have SL any other way! It's the most INNOVATIVE form of entertainment I've ever seen!!!

It should be more popular: Needs more advertisement!

Wouldn't construction zone sims be possible under this scenario? That is, a place for collaboration to occur for group builds until it is ready for prime time? At a cost of course; think money sink.

I would have guessed that the major portion of creating is done in singular mode. This method would also provide security, privacy, and eliminate the social distractions oft complained about (im's, drop ins,...) for the creator off making magic happen in his own little world.

I don't intend to structure my individual plan to be all about mass market appeal but I do think it is a goal for LL. As such, I consider it to be a significant factor when evaluating any suggestion.

Advertising doesn't make a product like SL, word of mouth does.

The future of the metaverse will be the teenagers of today.

Most people older than 20 or 21 are not as immersed in the world of computers that teenagers are. The key to creating the metaverse will be to grab their attention and then grow as they hit university / college / etc.

All these streaming hacks are going to be irrelevant as bandwidth improves.

I think being able to see the world being constructed around you in real time is a major part of the draw. SL doesn't have the tools to create content comparable to mass market MMORPG's or single player games so it shouldn't try to compete on that playing field. LL needs to keep SL focused on what sets it apart.



Now you're on to something Hardware and bandwidth will catch up to SL in time and as long as LL stays focused on their core mission (which I don't think is to compete with WoW or Halo) they'll be years ahead of anyone else when a fully streaming world CAN look and perform at a comparable level. Patience, grasshopper.

I don't get those kind of fps unless I visit the mainland. Since I spend 90% of my time in single-owner sims where the owners/group are fanatical about controlling textures and scripts, usually my fps is very nice.

I'd like to see some more tools for finding lag. The ability to identify textures really helped. How about some script critiquing tools.

IMHO you are on crack. The majority of my friends are 30+ and we are all immersed. Raised on the net doesn't automatically mean you will be knowledgable, intelligent, nor aggressive in your pursuit of new technology. I know plenty of teenagers who are plug-and-play passive in their use.

The kids certainly lead the way by utilizing and introducing the technology but the mindset will be adapted by those who can benefit most from it. Someone older and established will spend the bucks to have the latest and greatest and will embrace if it is useful. Technology makes cocooning easy.

Kids want to hang out at the mall. Drive to the mall with the kids, baby in a stroller, pick out some clothes and head to a fitting room/closet. Or, sit down at the computer, right click to see how things look on an avatar which is accurate representation of your physical shape. Swing the camera around to see what your butt looks like in that swimsuit in the privacy of your own home. Click to order your selection. Find a block of time and arrange for a sitter in order to spend a night out with friends. Or meet them immersed in a 3d world for a bit when you've got a little time to squeeze out of the schedule.

Most people owning computers, cell phones, and home theaters today had no idea they wanted them even 10 years ago. An old dog will learn any trick you want to teach him if it's worth knowing.