Greetings, This is my first post to this site. I am presently a business development manager in the employ of a Japanese Software Development Company in Tokyo. We are exploring the idea of building an online virtual world similar to the likes of There.com and Second Life. This is an idea still very much at a conceptual level, however I would appreciate any insight the pros here could lend on the following: 1. If building from the ground up, what game/physics engine(s) are most condusive for use in this type of application?I'm certain many of you have tried There and SL, but a few of the many 'must have' capabilities are: A. Ability for the avatars to move about freely and 'teleport' to different locations on the grid. B. Variable camera angle and object zoom. C. Ability for users (avatars) to create scripts and scripted objects. D. Integrated voice client and non-intrusive chat method. E. Ability for users to upload content designed offline. F. Ability for users to stream audio and video. Note- AI is not a consideration as this will not operate as a game. 2. General idea of the number of programmers and 3D graphics designers required to get a product launched within 6 months. 3. General idea of what kind of network architecture would be required to sustain the 'world'. I've read that SL has over 1400 Linux blade servers divided into nodes supporting the various 'regions' within SL. I have no idea how There.com is structured. Personally I find the There application to have a much more immersive quality primarily due to the elegant way that chat bubbles are generated above avatars head, and the seamless voice client. The physics engine in use (proprietary?) also works much more realistically than SL's. However, the graphics look dated, and it has a cartoony, childish feel to it. SL on the other hand is graphically appealing, but full of lag and user designed clutter. The physics engine (havoc) is awful. overall experience is frustrating. The positive attributes include the ability to design and customize just about anything you want. Sorry for the long post. I am not a technical expert in game design, but I am excited about the potential of this idea, and look forward to any feed back. Cheers, Timmy

I can't comment on all your questions, but regarding question number 2, I can tell you that to get this Designed, Prototyped, Beta'd, and Released within six months would take about half the population of China to create the graphic assets, the other hald to create the sound assets, and all of India to do the programming.

In other words, something like this cannot be done in six months. Most virtual worlds take 3 to 5 years to create, a team of around 50 creating sound and art assets and another team of 50 to program and average about 10 million USD to produce. So to get this done about six to ten times faster would requires about 10 to 100 times more money and 1000 to 10000 more people....just can't be done in that timeframe, not on the scale of SL or THERE.

You could (perhaps) whip up a prototype in six months with a group of around 20 people and about 500,000 to cover the engines and payroll. Just figures off the top of my head...I'm sure others here can give you more accurate figures.

It's funny you bring this up, because Gamasutra featured an article by a couple of the team members of Second Life. It's an interesting read.

Quote: Original post by fastlane69
In other words, something like this cannot be done in six months. Most virtual worlds take 3 to 5 years to create, a team of around 50 creating sound and art assets and another team of 50 to program and average about 10 million USD to produce. So to get this done about six to ten times faster would requires about 10 to 100 times more money and 1000 to 10000 more people....just can't be done in that timeframe, not on the scale of SL or THERE.

Although I agree that 6 months is not feasible with such a large project, remember that in second life much of the content is created by the users themselves. Of course, you don't want the first users to log in to an empty world, so there needs to be a good amount to start with. I've never created any software to this scale, so I'll decline making any sort of an estimate.

Thanks for the quick responses.

Fastlane, I agree that 6 months would be an unrealistic target. I know the overall task could be vastly shorter (time wise) if one were to take an existing application and add the necessary customizations. This part I believe could be done within 6 months. This now begs the question of whether there are any half-built or abandoned games for sale :^)

SMR, thanks for the link. Incidentally, the two Linden folks mentioned in that 2003 article are the two gents now running the company.

Cheers,
Timmy

The problem there is that there are no such applications that can reduce the time it takes to make an online world by a factor of 6. The only company that I think has any chance of fulfilling this request would be Big World technologies (www.bigworldgames.com), but even they don't provide you with the artists and programmers necessary to customize the game.

Consider that to create the design document for an online world takes at least two months (working 40 hour weeks, I wouldn't trust any software that had less than 300+ hours of requirements and design analysis put into it) and to Beta a online world you need a MINIMUM of 3 months to work out all the bugs (programming and functional) and you are basically left with 1 month in which to hire programmers (for the customization) and artists (even bigworld doesn't offer an artist pipeline for it's customers) and actually get things done.

Trust me: 6 months in the present market is impossible without spending a sick amount of money to speed the process up...even then, it is highly doubtful that you can pull it off considering the scheduling of tasks. Give yourself a year to get this beta and you'll have a fighting chance.

Quick Question: Do you have anyone in your organization (or your partners) with formal project management experience? I urge you to go to them or hire a firm to do an analysis of what you want to do. ONce a PM does a WBS of your project and then links all the dependencies and such, I think you'll find that the critical path alone will take more than 6 months and more than 500K USD.

Check out Big Worlds though. They are your best bet at a pre-built commercial online world.

Fastlane,

Thanks again. I was not aware of bigworlds, and looking at their videos I see that they

We are actually in contact with a local firm similar to bigworlds that has a product which we could develop upon. However, I needed an outside opinion from a seasoned veteran such as yourself, to make a better determination on how to move forward.

I would be really intersted in hearing any opinions you may have on this whole 'Metaverse' or Non-genre MMORPG as some call it. Some people herald the coming Metavers as and end to the browser as a way to navigate the internet. Personally I don't think people will ever stray from 2d content entirely, but a lot of potential exists in this are for sure.

We have some unique ideas for business models that will capitalize on the now nearly ubiquitous levels of broadband penetration in Japan. Additionally, Japan is seeing a significant surge in PC sales that will peak at 94 million by 2010. I am limited in what details I can share at this point, but I can say that we would be able to bring a much more aggressive business model to bear in this type of environment.

Cheers,
Timmy

There is no doubt in my mind that browsers will fall by the wayside in another 10 years or so. By then, Browsers will be to the Metaverse as Dial-up is to Broadband...still supported but on their way out.

I also disagree that people won't stray from 2D for the simple reason that we don't live in 2D and thus any experience that brings us closer to reality will generally be accepted. In this Metaverse concept, it the difference between shopping online for a camera and reading the specs or going online and seeing the camera from all angle, perhaps even holding it to see how it looks.

I'm personally exploring education in these online worlds. I have a 3-year grant from the US gov't to explore this idea. Even though I've been working on this alone for the last three years, it will still take year one to properly prototype and alpha it, year two to do a private beta, and year three for a open beta, all leading up to public release around 2009. So I'm a firm believer in these environments but I'm also a firm believer in taking your time.

The best piece of advice is that you work on the four documents you need to guarantee success: your Business Plan, your Game Design Document, your Technical Design Document, and your Project Plan.

Each addresses a different but synergistic aspect of creating an online world. For example, your business plan will show you how to fund it's developement and how to fund the post-release support; it's also critical to get investors on board. Your GDD will lay out how the game (doesn't have to be a game, but it's the best word right now) actually works...what people can do, why they would enter your world, etc. The TDD lays out the actual nuts-and-bolts of your software and hardware which in a online world, not surprisingly, is a huge deal. Finally and based on all previous documents, the PP outlines the time, cost, and quality that your game will acheive and lays out the exact path you are going to take to make it happen.

If you don't have these already, any plans to create and deploy this world will fail spectacularly. Personally, I would spend a few months generating a prototype of these documents and then use them to guide you along your path. You won't be able to create these documents all at once since there is alot of cross-polination in these documents (your BP will affect your PP...your TDD is based on the platform you choose, which could be a factor that comes from the BP and the requirements of the GDD), so you should work on getting a v1.0 ready asap and then constantly iterate it as decisions are made down the line.

Only when these documents are "locked in" should you even consider developing IMO.

Fastlane,

Thanks again. Actually we knew about the GDD and others documents required as we have been in discussion with VC at this point.

Your comments have got me thinking about how to handle the technical development side. After reading this recent post on Terra Nova: http://terranova.blogs.com/terra_nova/2005/12/make_your_own_v.html#comments , and poking around the Multiverse site (http://www.multiverse.net/index.jsp?cid=0&scid=0) it seems that this type of solution could greatly reduce the time to market of a virtual world type application. I understand that developing the in-world content alone could be an enormous and time consuming endeavor in itself, but I am very interested in your professional opinion of the Multiverse offering.

If you have the time, I would appreciate the chance to talk to you offline at some point. I would like to reveal a bit more about what we have in mind with this business idea.

Given that what you are doing is very similar to some of my own projects, it's probably not a good idea that you reveal too much to me. I will be as helpful as I can with my advice on these forums, but any offline discussion could seriously undermine both our business propositions.

The problem with all these solutions (multiverse included) is that they are untried and untested. Thus, you are making a huge time and money risk on the chance that these products deliver. One thing that disturbs me about all these offerings is that they give absolutely no metrics to the public...we have no idea on client load, resource use, average bandwidth, etc. I think that the toolset that will truly dominate this segment will be the one that can come to the public and say "look we have tested this throughly, it workds, here are the numbers, now come and play"...to me, Big Worlds is the ONLY company that comes close to this.