World Finder

This little app grew out of some internal work on how to advise clients as to which virtual world platform to choose and why, but we thought we'd share it with the community. The worlds in the finder are currently defined by the world's we'd be happy to work in with the client, and which we think can offer the client a good solution. We might extend the list to include others - but don't expect them to rank highly!

Which brings us to the ratings allocated to each world characteristics. These aren't scientific. These are our personal view, BUT reflecting over 10 years of virtual world expertise and the delivery of over 100 virtual world projects to paying commercial clients. We've deliberately limited the parameters to 12 (we couldn't manage 10 and could easily have had 50), but have tried to chose those which we think are the most critical in virtual world selection and which force users down one route or another (eg graphic quality varies only by degrees but in-world building is binary). More than happy to debate if we've chosen the right 12 (but only if you say which you'd drop), but we WON'T enter into debate about the ratings, they are just our view.

We have assumed that the use case is around virtual worlds being used for training & education, built environment visualisation, data visualisation or collaboration.

---

Essential - The platform MUST support this feature, it it doesn't it's excluded
Highly Desirable - Would really like this feature, but don't exclude the platform if it doesn't
Desirable - Would like this feature, but don't exclude the platform if it doesn't
Not Bothered - Ignore this feature in any ranking
Not Required - Do not need this feature

---

Use Case

Can build and customise in-world:	Essential Highly Desirable Desirable Not bothered Not required	(rather than use an external 3D design tool or ask us to change everything)
Highest possible quality graphics:	Essential Highly Desirable Desirable Not bothered Not required	(yes this is a use case not technical requirement, and note that not virtual world will currently match the best computer game)
Need to model a large area (> 1 sq km):	Essential Highly Desirable Desirable Not bothered Not required
Custom client:	Essential Highly Desirable Desirable Not bothered Not required	(e.g. no vendor branding, only functions needed by user etc)
Engagement mode:	Single User, No avatar Single User, With avatar Groups, No avatars Groups, With avatars	(This is about users per activity - it is assumed that there will always be multiple concurrent users)

User Access

Runs in browser (with plugin):	Essential Highly Desirable Desirable Not bothered Not required	(no 3D virtual world runs yet in a browser without a plugin)
Runs on mobile:	Essential Highly Desirable Desirable Not bothered Not required

Technical/Security Considerations

Deploy on own servers:	Essential Highly Desirable Desirable Not bothered Not required	(i.e. behind firewall to keep data secure)
Robustness/security of code:	Essential Highly Desirable Desirable Not bothered Not required	(i.e. minimal chance of code failures or exploitation)
Import of existing 3D models:	Essential Highly Desirable Desirable Not bothered Not required	(e.g. from Collada, 3D Studio max etc)

Commercial Considerations

Lowest Cost Possible:	Essential Highly Desirable Desirable Not bothered Not required
Security of Supply:	Essential Highly Desirable Desirable Not bothered Not required	(combination of size of user base and likelihood of vendor still being in business in 5 years)

---

---

---

Assume every world has: text chat, voice chat (but costs except sl/vastpark), web access, web services access and will be used by multiple users simultaneously.

Assume every world will need: decent (but not outstanding) computers, and decent bandwidth

If you must have a Mac solution forget Vastpark. If you must have a Linux solution forget VastPark and Unity.

If you just need something for a single user, desktop based (or otherwise) then it's probably Unity.

Other considerations (tend to favour SL/Opensim, but could be implemented in others but with significant extra cost): hypergridding, in-world currency, in-world social media functions (profiles, groups, friends etc)

Yes we know that: Mesh is in beta in SL and in OpenSim 0.7.1. and Chris Collins is doing work to combine OpenSim and Unity.

Worlds/engines considered but not yet listed (and/or considered by us to be less attractive across the board than those we've actually listed): Blue Mars, CryEngine, Olive, Open Qwaq, Open Wonderland, Protosphere, Teleplace, Unreal, Jibe, WebAlive

Near misses for selection criteria: Ease of deploying in-world voice chat. Being based around open standards (and open source code, but that is less imprtant than standards).