Second Life: Next Generation Platform Update-IV 2015

I started this series with:

This fifth article in the series is what I gleaned about the possibility of internal content creation tools from the speech at VWBRE that touched on the coming Next Generation Platform (NGP).

In Motion

In Motion by Daimaju Clowes [always taking clients], on Flickr

Internal Tools

in Second Life we have no internal mesh creation or editing tools. We only have tools for working with the parametric primitives. Of course the layout tools in the SL Viewer will handle either prims or mesh. But those layout tools are showing a little bit of age and the main SL Viewer still doesn’t have prim alignment tools. 

Numerous people ask from time to time if Linden lab is going to build in 3D modeling (mesh modeling) tools. The question applies to both NGP and Second Life. These questions are usually asked with the speaker thinking of some internal tool being like Blender or 3D Max. The idea being such a tool would allow us to create or edit mesh items from within the viewer.

I suspect they are hoping for something way less expensive than 3DMax or Maya and easier to learn than Blender. Unfortunately, that is pretty much an impossibility. 3D modeling is a complex task and the tools are sophisticated and come with high learning curves. It is way doubtful that the Lab vould produce a better tool on an initial try than cmpanies that have been the business for decades.

Whatever the case, Ebbe has given us a definitive answer about whether the Lab build a mesh modeling tool into the viewer, that’s not going to happen. What they plan to do is provide good import support for 3-D modeling tools. My previous article covers that support.

Ebbe explains they are using Maya for internal reasons. Maya easily provides them with the most complex content and they use that stuff to test the NGP software their developing. They’re trying for efficiency and performance with the new NGP. They need to quickly make complex content that stresses the system to find content rendering issues that slow the rendering process and optimize those bottlenecks.

Maya is a huge program. An entire company develops and supports it. The same is true of Blender, 3-D Max, AutoCAD and numerous other programs. Adding the equivalent of one of those programs to the NGP viewer would make for a huge workload. It just isn’t reasonable or practical build an editing tool into the NGP viewer that’s equivalent to Blender. Especially when several exist. The Lab can save that time and and spend it on something else, like NGP

If you’re wondering why the lab is using Maya and what those internal  reasons might be, it is all about revealing the places where they need to optimize and polish their work on NGP. It is all about what is easy and quick for the Lab. It says nothing about which tools they are designing for support-wise. The only hint it gives us about tools that will be available for creative users is that Maya definitely be one of them.

Summary

We have three major modeling tools most people work in; 3-D Max, Blender, and Maya. I think the majority of SL hobbyists use Blender.

Building another 3-D modeling tool and building it into the viewer would delay NGP development. Plus a new and unique mesh modeling tool would add to the learning curve of all users. I’m sure those are things that just are not going to happen.

By using existing external modeling tools people will be able to use knowledge gained elsewhere to work inside NGP. That’s a good thing.

8 thoughts on “Second Life: Next Generation Platform Update-IV 2015

  1. I hope that they are considering a very important third option for building tools – a middle ground solution, if you will – for developing content. The ability for anyone to build in world was critical to the success of SL, at least prior to scuplted prims and mesh. The loss of that ability is considered by some a significant contributor to the decline that SL has experienced over time, with outside development becoming more and more necessary to remain competetive and/or produce high quality content.

    That \middle ground\ would be a set of tools which, while not nearly as robust or feature-rich as a full production suite or 3d application, would still elevate building capabilities for the average user well above what the existing primitive geometry tools provid. A likely solution would be the use of voxel building tools, which are relatively new yet extremely capable, flexible and refined already as well as being compatible with the OpenGL standards already adopted. They are also readily convertible to other formats, as well, such as verious polygonal mesh types. This would allow made items to become assets no different when complete than content imported by the user from other applications, so it can then be managed, distributed, sold or used the same way. Yet voxel creation tools have already shown to be considerably more advanced than our beloved, proprietary prim. Initially rough and inarticulate in games like minecraft, the voxel tools today are much more advanced and can already provide environment and rendering capabilities beyond what SL is capable of, even with mesh.

    I really miss the days of building in world. Losing that ability, as I joined the sculpty and mesh wagon train, remains a personal grievance of mine. It really does make a difference. Having the ability to import content from outside SL is much different than it being the only solution. Leaving out that important ability, to make within the world (even to make things together there), would be a terrible mistake and easily avoided by making the right technology choices.

    • What ability to build has SL lost?

      What voxel editing tool goes beyond mesh? I looked through 2 Google Pages of voxel editors and everything I found was Minecraft primitive…

      The growth or decline of SL is hard to factually define. Ebbe mentioned that at the height of SL there were 1.1± million logging in. Today that is down to 900k. That isn’t a lot of shrinkage since 2007-8. Rod spoke once about the participation in SL changing because of mesh. I agree users do spend SL time outside the world building in Blender type tools. So, it is possible to surmise that we have almost as many daily users but those users spend less hours in SL. I track concurrent users. I know we have dropped from 80k peak to 60k peak. But, we can’t know if people spend fewer total hours or more because there is no way to track hours people work on SL that are not spent using the viewer (or otherwise logged in).

      Ebbe mentioned using voxels but little more than just ‘using’. I assume there will be some type of prim building and the voxel comment suggests prims will be better.

  2. I think the main difference between the pre-sculptie/mesh SL and post-sculptie/mesh SL is that most anyone could make decent money building and selling in SL in the early days. Now the standards have risen, more is expected and people want to pay less. The high-end mesh artists are turned off because they don’t want to sell their work for pennies. The old school prim builder’s work doesn’t sell any more at all. Because those who do sell their mesh items have to accept such low prices and land is so expensive they can’t afford to have an in-world store so many just have a marketplace presence which likely isn’t seen by new users who like to dress up their avis and go shopping.

    The NGP has a better chance of succeeding – but if the Linden fees on sales is too high and the prices of their products has to be too low to compete it will still turn off the high-end mesh artists. They should be able to afford an in-world store though.

    • The elimination of the MagicBoxes decreased the need and demand for land. That decreased the $ value of land. It may have changed the entire financial model for SL. Ebbe’s idea to lower land cost and implement more transaction tax could be the result.

      Land area has been decreasing as mainland occupancy and private regions are given up. A shrinking SL may be more a measure of the move to Marketplace then a measure of people leaving.

      I’m not sure the cost of goods has really changed. How much is a dress? How much did the price change? I can only find empirical and anecdotal information. Some system clothes are still expensive. Some well made fitted mesh is expensive. But, in general dresses sell for L$200 to L$1,000 and a few even more. That is about the same as when I came into SL in 2008.

      Fitted mesh making has not yet stabilized nor become ubiquitous. Work flows are still improving, which will reduce the creative cost. But SL appears to be on par with prices in the Renderosity Store. So, value of goods is not a problem unique to SL. SL and renderosity.com have about the same number of unique visitors per day. So, they seem to be relatively equal markets with similar prices…

  3. https://www.youtube.com/watch?v=Bj1UQ7yL4mw
    https://www.youtube.com/watch?v=loJ8Ds0FPik

    Actually, I can see two options for inworld tools:
    -Voxels
    -Mesh componenent building, along with the means to deform them through inworld tools… Think putting up a row of mesh wall segments, then deform them along a curve to fit them into the landscape. Mesh components bear the advantage, that LL could leave the complex technical details, like uv mapping, LOD or physics to the more experienced mesh builders and just account them after their impact on render performance.

    I don’t expect we’re going to see prims in NGP, for they’re too limited and too inefficient to render, with a lot of uneccessary geometry and draw calls, due to their number of texture faces…

    LL is obviously aiming to make NGP a VR platform, and VR is too hardware demanding, they simply cannot allow the use of inefficient content (a convincing VR experience will require stereo UHD rendering for super-sample-antialiasing, running at 90FPS). The current professional high-quality VR demos squeeze out the last bit of performance out of the most-capable hardware you can buy today…

    This is probably the reason they’re holding back with information on how they plan on inworld building to work. It will require more careful engineering than just developing a system that looks nice on paper and hope for the technical reality to eventually adapt.

    I think we’ll have to get accustomed to the fact that there will be much stricter limits on user creations and we won’t enjoy the same level of creative freedom we have in SL, today….

    • I agree and disagree. On restrictions I totally agree. We will see something like the SL Land Impact cost in NGP. But, probably not so tied to land. Starting from scratch the Lab will not be held back by prior costing plans. In SL texture use and prime counts are killing us from poorly designed clothes.

      Creative freedom, I cannot see them restricting freedom. The Lab has a history of providing creative freedom. I expect they will try to impose a high cost for poor design, with poor design being based on high render cost being bad and thus expensive. Otherwise we will be free to create whatever we can imagine.

  4. To clarify: I did not mean that SL has lost that ability to build in world. What I meant was that in order to develop higher quality content it became necessary to work more and more outside SL using other tools.

    As for voxels, take a look at the progress being made with Voxel Farm: http://procworld.blogspot.com/

    There really is a lot more to voxels than simplke blocks. You can store all kinds of data in them. Procedural terrain is awesome.

Leave a Reply

Your email address will not be published.