has LL ever generated investment dollars from its residents

has there been a time when LL has solicited investment dollars from its own residents? I constantly read about freezing and lag etc.. would it not be smart to look towards the very residents of your creation to further advance the technology of your virtual platform. Money can be requested in phases, depending on what is needed. If this world were mine (luther vandross) i would give the very people who had confidence in my idea, the right to invest (even if it is short term) in the world that they themselves help to create. Example: 1 miliion residents+ten USD a piece=10 million dollars to give everyone a better 3-D world to enjoy. "your world, your imagination".

They have enough investors already. Plus there are regulatory hurdles they would need to clear before taking our money. I would certainly be interested in such an investment though.

Not a bad Idea, really.

I suspect though the number of actual humans behing all these user accounts is quite a bit smaller than 1 million.

Still - =) could add up.

Maybe even a performance charge lol "Do you wish to pay $1 more a month to go directly to Second Life performance improvements?"

Kind of like your taxes and the President re-election fund.

It won't help much.

There's a reason why tens of millions invested haven't already solved the freezing and lag and whatnot.

The problem isn't one of the platform, but rather the wild variety of ever-changing client hardware, and the drawbacks of the internet itself.

Two things need to happen.

One, a 'maturing' of client hardware, much as what happened in the automobile industry. There were a zillion kinds of engine once, but now, for good or ill, the 4 cycle 4/6/8 cyl engine is 'the' engine.

The computer industry will someday standardise a lot more (mature as a market), but until that happens, there will be a huge number of constantly moving targets to build a client for. As such, it's not a money-solvable problem for long.

To illustrate the point, consider all the peripheral manufacturers every time Microsoft releases a new operating system. Even if you are Hewlett Packard, there is no way to rapidly respond. Microsoft will rush its product to market more rapidly than you can design around the problems. Over and over again.


Second thing: most people really don't know what we lost, when the 90's economic bubble popped.

Had it hung on another couple of years, we would be running the gigabit internet, the one upon which Second Life really needs to come into its own.

History dictated otherwise, cool heads prevailed in 2001 not 2003, the irrational exuberance ended early. So we are each left with anachronistic, low megabit, data-over-copper datastreams in the last mile to our computers. Which basically define the 'tech problems' of the grid.

Can't get the data we need fast enough, so we spend thousands of hours designing around the problem. Our grid is nearly defined by this problem, in fact. Sim borders, textures, resource limits...

So I think of it as kind of Victorian Internet... the real one stillborn in 2001, and us mostly still using a lot of copper line - the 21st century equivalent of brass gears.

Well said, Desmond. You're way ahead of the curve, even for a steampunk

That's really interesting Desmond and well explained - please go on, it's fascinating. IS it really true that most of our probs come from that short bit of copper from the exchange to us?

Cheers,
Duder

I certainly can't speak for them, but in my view LL does not want that type of investment. If they wanted something like that, all they would have to do is eliminate the free accounts or limit them to a certain amount of time trial basis. They are unwilling to do that as the HUGE amount of [seemingly] account holders fit their business plan, whatever that is.

My perception is, that at this time, two non paid accounts who cause multitudes of trouble to other residents are much more prized than one minding his or her own business paid account. We're basically nobodies. Right now they are after numbers. That seems to be all.

Would I invest in LL? Possibly...I don't know if they are publicly traded on the stockmarket. I would not at this time however. I DO think they have latched onto the wave of the future here but I'm not sure if they will continue to ride it. Non existant or cavalier customer service, myriad technical problems, pursuing unneeded and generally unwanted "bells and whistles" [e.g. voice] instead of fixing existing problems and generally ignoring the desires of paying customers would NOT make me want to invest at the present.

I would, however, gladly pay a larger monthly fee IF they phased out the free accounts. I do not believe that will happen, unless a viable competitor enters the market place without the above listed problems.

I must admit at times to feeling a bit like an investor already. As an island owner, I am paying Linden Lab nearly US$4000 a year (just ask Desmond what his yearly cost must be!). Now certainly I am receiving a lot of value for my $4000. But it's also clear that Linden Lab is generating most of it's revenues from island owners, and some portion of that revenue must be used for continued development and infrastructure work. I do not feel bad about this at all. I love this place, and I am honored and happy to contribute to it's continued growth. I wish more people felt that way sometimes.

I have invested more money here already than with any other internet company, with the exception of Amazon.com.

It is a question of priorities. I understand they blew my entire personal donation on cocktail weenies for the Coldwell Banker welcome party.

Just fix the damn performance issues already!*




*the ones that are within you control

Money from porn is what funded the early internet. Now we can all get our porn for free.

Help technology! Give porn sites your credit card!

I respectfully disagree on almost all of your points.

The client hardware is not the bottleneck or even the design limitation. SL is coded to OpenGL which is why it migrates from Windows to Mac to Linux with not monumental effort. SL does indeed push OpenGL harder than any other game does which does unconver driver deficiencies like nothing else. But that is not what causes the majority of the big issues such as inability to teleport, inventory failures, asset failures, etc. Those are all server side issues and are the result of design limitations.

The design flaws arise from early decisions that were fine for a proptotype of a thousand concurrent users but are failing miserably under current load. Most of these concern the shared resources: the asset server, the presence server, the tiny size of both the development group and support group.

There were other protocol decisions - for example sending your client everything in your draw distance and letting the GPU figure out which small fraction of them need to be rendered - that increase the bandwidth demands on the servers at least fourfold. Again, this was fine for a prototype but if the servers were smarter, they could send out a fraction of the data that they do now. Multiply that across 40,000 concurrent users and the aggregate data rate - and therefore the networking requirements - becomes much smaller.

The reason that money hasn't solved these problems is that they are neither big enough nor interested enough to step back and do the enormous re-design and re-write that these changes would need. It has been almost axiomatic in systems design since at least 1975 that a prototype shows you where your initial design decisions were wrong. Thus Fred Brooks' point:I don't think I could better describe how SL appears to the customers than this passage does.

That is the saddest, yet best post I have ever read. I'm downgrading to basic...

Weren't there lifetime accounts at some early point in development? I would call something like that a resident investment.

In any case, regarding the throw-one-away philosophy: It most certainly has to happen at some point if SL is to mature. A version 2 needs to happen at some point, and reasonably soon I suspect, or some other upstart who has learned from LL mistakes will make a new product that's better, and overtake SL. Let's face is, the current version of SL is clearly a beta - the real question is whether it graduates from that stage or not.

Malachi has a *very* substantial point just above, but we are on different levels of discussion I think.

Let's start with Malachi's bottom-up redesign. Say you get the *best* most optimal Second Life grid as we know it. It's *still* hard-limited by network performance.

Jopsy describes the problem very well here; it's exponential:

To quote Jopsy: (source: /206/f5/172570/2.html#post1446642 )




As for issues with teleporting, asset failures and such - all solvable. These are growing pains from .5 million to 5 million accounts created in mere months, and unsurprising, really.

So even with all that 100% solved, there is only so much better the grid can be; only so many more avatars in view of themselves unless we 'can' the experience like a standard online video game and dramatically limit what we can do.


The Company is working on the server side even now, changing out quite a bit under the hood for efficiency. Yes it's not a total redesign, but they are changing a substantial, relevant chunk. I consider (as I hope they do!) that the grid as we know it *is* the 'pilot plant.'


As for gigabit net:

http://www.informationweek.com/news/showArticle.jhtml?articleID=15800331

http://www.networkworld.com/news/2005/032105specialfocus.html?prl

http://www.wired.com/wired/archive/11.08/pipedream.html

http://fibers.org/articles/fs/8/3/3/1

We were soooo close. So close. Another couple of years, and Second Life would have been a relatively light network demand on 10 gig lines - hitting hard limits again of course on finite bandwidth, but at a far more realistic place. But not today, not with the way history unfolded.

Malachi, Desmond, Jopsy, thank you for the illuminating posts.

I had a strong feeling that LL was merely shirking their duties. If they need money, and human resources LL has access to both right here in-world. This platform is too good of an idea to subject it to shotty patch work, and poor customer service.
Rosedale has admitted that everything created in SL causes lag, and it seems he is comfortable with that being a persistent problem. When do you hire enough developers? When do you properly staff and train a customer service department? These are basic needs for a large IT company, are they not? I paid my $72.00 the first day i read about LL. When i encounter problems in-world, i assume my money is going to the betterment of SL, so i tell myself that things will get better as time goes on. But the more time that passes, the more it appears that LL is flying by the seat of their pants.

Jamil

Thats a good point not all investment is technology related.

Investment into making SL a better/ more responisve company has merits on its own.

You are quite right that any multi-way environment becomes exponentially more expensive, but there are factors that mitigate even that.

Pardon me if I'm being pedantic, but computer science types like to use Big O notation to describe the complexity of a problem. Read the link if you dare, but for our purposes here it is enough to say that multi-way communication has a cost of O(n^2) or exponential based on the number of participants (n) as Jopsy describes.

However, when translating to real behavior of a system there is a constant factor that O is multiplied by - that is, the base cost for one operation. Take the example of IRC which is a multi-way communication IM-type service. IRC has exponential cost but can support hundreds of simultaneous users because the constant factor - the amount of data that needs to be retransmitted - is very small. So even with 100 users, 100^2 "interactions" can be feasibly handled. Thus, if SL was smarter about what it sent to the client, that constant factor becomes smaller.


How could it be smarter? First of all, contrary to Jopsy's example above, certainly not all players are chatting at once nor would you want to try to make sense of it if they were. In fact in large crowds, most avatars are usually not doing anything at all except dancing. We know that avatars are the single greatest expense in SL traffic, and a room of dancers requires a huge amount of data to be sent to all present. But the animations are relatively small, the problem is that the animation is run on the server and the results of that are sent to each client in view distance. But it isn't just the clients inside the club that get sent that data, it is every client within view distance even if the club walls are opaque. Start chiseling away at these sort of "data being sent only to be ignored" interactions and that constant factor continues to go down.

Why was SL written like this in the first place? Because no one expected the scale we now see and it was quicker and easier to send unnecessary data than to do more server side work to cull it. There is plenty that can be done to reduce the O(n^2) problem. The question is: are they planning to?