I started this series with:
- Second Life: Educational Community Update 2015 March
- Second Life: Next Generation Platform Update 2015.
- Second Life: Next Generation Platform Update-II 2015
- Second Life: Next Generation Platform Update-III 2015
- Second Life: Next Generation Platform Update-IV 2015
This sixth article in the series is about what I gleaned about the types of users the Lindens see using their new product from the speech at VWBRE 2015 that touched on the coming Next Generation Platform (NGP).
Who They Will Support
Ebbe is thinking they’re supporting three categories of people.
One category of users is professional high-end artists. Now that could be artsy artists (think Bryn Oh) or 3D modeling artists (think Slink). This is a group of people that are likely to have high end modeling tools and possibly work in the game design field.
Another category of users is the hobbyists (think most of the people selling in the marketplace). This group of people are likely using Blender.
A third group is what he calls customizers. That last one was new to me.
He thinks of a customizer as being a person that decorates with furniture rather than making the furniture. These are the people that place things to furnish a home or building and wear things. For virtual world merchants these are the customers. This category makes up the majority of users currently in Second Life. They do not use a modeling program.
This customizers category of people needs layout or placement tools built into the viewer. Ebbe tells they are looking to improve on the tools we have in viewers now. Expect really good layout tools to be built into the viewer. Or… at least hope for them.
One of the tools that users have wanted for some time we call alignment tools. We have alignment tools in the Build Panel of most third party viewers. However, we have never gotten them in the standard SL Viewer. The reason the Lab has given is the tools behave inconsistently with some objects. The Lindens believe that would be too confusing to include in the SL Viewer.
Presumably the NGP viewer and server side build system will have objects with an improved bounding box that will allow more consistent alignment of objects inside the virtual world. It has been a problem in Second Life.
Since the alignment problem touches so many of the virtual worlds users, the Lab will find a solution.
From other things said by Ebbe, I have the impression professionals and hobbyists will get more ‘design’ consideration than customizers in the area of creation. I expect they will have beeter support in NGP than they do in SL.
I also think customizers will get more user face and building layout tools design attention. Ebbe’s goal Is to make all things better.
Because the lab has decided to start out with proprietary viewers we won’t see third-party add-ons for some time. That means we’ll have to start out in New World with whatever tools lab thinks are most important. And for now, as far as the viewer is concerned, that appears to be import and layout tools. I assume alignment is going to be a big part of layout.
In SL we are used to third party developers being able to push the Lab by adding features to their viewers and showing how popular the features are. Often the Lindens just can’t recognize the value of a feature or gauge it’s popularity.
In NGP that won’t be possible. The NGP viewer is closed source, proprietary. I expect it to remain that way well into 2016. Ebbe definitely gives the impression that the viewer will at some point become open source or at least have the ability for third parties to add plug-ins. With Oz Linden being a Lab employee I think chances of the viewer going open source are pretty good. But again, this is a point I’m not betting on.
Thinking about plug-ins, I have doubts as to whether they will take that route as a primary path for user added features. They planned for users to add plug-ins to the existing viewer in the area of Media On A Prim (MOAP). It didn’t happen. I can’t say what the Lindens may have learned from that experience or even how they view it. But, it was not successful. From other experiences with user acceptance we know the Lab has give significant weight to our behavior over our words. So, I don’t look for plug-ins to be a big priority for NGP Viewer development.
I am not putting money on whether they will push toward full open source or go for plug-ins. But, I am confident we will get one or the other. And I think open source has a bit of an edge.
Summary
The Lab is not big enough to do everything at once. So, while they want to support the three target categories they have to start somewhere. I think it obvious the priority is professionals and then hobbyists. Following that will be the customizers. I think there will be considerable overlap in all the categories. So, the priorities may not even be apparent to the users.
Alpha and beta users are likely to be professionals and hobbyists. The customizers tools could be developed during alpha and customizers would find then there when they come in.
My point is this prioritizing of categories isn’t favoritism but design and implementation practicality.
I wouldn’t hold my breath on NGP to become opensource… For once, I think going OS has done more harm than benefit to SL as a whole. Opening the viewer source has ultimately cannibalized SL enabling competition in certain areas through OpenSim, it has spawned a range of hacked copybot viewers, and which game-breaking new features have ever been introduced by TPV developers, aside of changes in the UI? Has it ever helped opening SL to new applications? Has it helped to attract more users? I’d argue without the viewer being OS, SL would still be about the same as it is today… minus a few handy little features…
Also, I think a modern platform will rely on even more proprietary components. We’ve seen that happen with Kakadu and Havok in SL and I expect an even tighter integration of the physics engine into the client, since that would help a lot to relax the sim from workload and improve performance…
Plugins are far more likely to happen, and it would democratize viewer-customization, by allowing a wide range of developers to build extensions, rather than limiting it to just a few teams who have the capacities to develop and maintain a full, alternative viewer. I think somewhere in his speech Ebbe mentions they’re looking on how to make the UI customizable, even…
As for MOAP: I’d love to use it, and I think it would allow for a far more intuitive and rich interaction with the world than we have today, with clumsy prim HUDs and the blue menus. Unfortunately, the feature is fundamentally broken tho:
https://jira.secondlife.com/browse/VWR-17719
Always requiring a click on the MOAP surface first, before we’re actually able to interact with the content is annoying and effectively renders it useless for anything but playing a Youtube video… The fact that it’s been unfixed for almost 5 years now indicates it might be a more elaborate issue with the WebKit framework they’re currently using, and Oz has mentioned they’re planning to switch to a different HTML framework, which will hopefully solve that issue and maybe even allow us to fully overlay the 3D view with semi-transparent HTML5 user interfaces…
I agree holding ones breath waiting for open source is a really bad idea.
Disagree about the problems from open source (OS). OS has not created the problems. It has made it easier for anti-social people to build viewers that steal content, copybot. But, one does not need one of those viewers to ripoff content. The OpenGL debugger provides all the tools one needs. A couple of years ago I played with it to see just how extensive its abilities are. Basically if it renders on my screen I can capture it.
The copybot viewers I’ve seen are never open source and are often Trojans sold or distributed for the hidden purpose of stealing passwords.
Open source has given us materials and a ton of features the Lab would never have had time to implement. It frees the Lab to work on the projects that require viewer and server changes to implement.
I agree with you that an NGP viewer is likely to have tighter integration with Kakadu, Havok, Vivox, and other utilities provided by third parties. You also make a good case for the plug-in path. I wouldn’t bet against the Lab taking that path.
There are a number of viewers made by a single person. Also, teams tend to be 2 to 6 people. Not my idea of large. Even in those teams whole sections of a development are handled by one person. Only Firestorm and the SL Viewer are developed by what I think of as large teams. So, I don’t see viewer development as a problem or requiring teams. The Kakadu has an open source replacement. Other parts, like Havok, have Linden Libraries that can be used. I expect that to happen with NGP too. So, while I make take a bet against there being open source, I would have to get really good odds and a limited time frame
MOAP has possibilities. I expect it to go toward CEF in SL and have some counterpart in NGP. As to the HUD’s, the SL Viewer already has HTML dialog panels, just no one uses them and there is almost no support info in the Wiki for them. That lack of use may signal to the Lindens to do something different with NGP.
The Lab never intended to extend MOAP. They stated they expected third parties to write plug-ins to support it. So, that too may suggest to the Lindens a different approach is needed. CEF and HTML5 is very likely to be part of NGP and to some extent the use of those is coming to SL.
@Wolkenreiter, OS is the only way to go if you want to create a rich, fast evolving environment that attracts developers and creates a market. Even behemoths like Microsoft have finally understood that, as they are opening their entire .NET framework source code.
As for SL, LL never opened its server-side code. OpenSim was created because devs reversed-engineered the SL protocol for client to server communication. As for viewers, there is no way for LL to block the development of alternative viewers. As someone said time ago (can’t remember who), if a viewer communicate with a server you have a door to create an alternative viewer.