Moving on

On Monday I gave a tour of the sim to avatars from the annual JISC Online Learning Conference. After some 30 months, it was the last event on the island and I'm glad to say it went OK -- even SL voice for the most part behaved itself. The picture shows my avatar in Molecule Park extolling the virtues of SL mesh.

Anyway, I've since cleared the sim and I'll submit a Region Cancellation Request tomorrow. Nothing profound to say. Still renting a small parcel in SL but mostly looking to OpenSim now.

Fiddling...

The urgent need to generate some sculpted prims has finally driven me to invest in a high-end inworld editor, namely Sculpt Studio (I bought it inworld on the Tech Expo sim). This, together with Sculptris for mesh, is putting off the evil day when I have to learn a "grown-up" 3D modelling package. I shall resist the temptation to ask for guesses as to what I'm attempting to model. Suffice it to say that, yes, I could have done this with SculptyPaint but I do think that having an inworld interface provides an opportunity to augment the interface with, for example, 2D images of the intended product.

As you will probably have gathered, Sculpt Studio works in terms of slices. Ordinarily you edit the shapes represented by the slices using point prims connected by particles. However, in this case it sufficed to simply move a few prims around to compress one end and elongate the other. After a little smoothing, I sent the data to a website offworld that generated the sculpt texture for uploading. It also provided a .obj file for UV texturing and text so I could regenerate the sculptie from a notecard.

Having applied the sculpt texture (to the dark prim above the slices), I had to stretch the sculpty a little and still it's not perfect. However, it's close to being adequate which is more than I've typically managed previously. Furthermore, I think I could explain to a student what I've done so, expense apart, it's not necessarily a technical deadend. Moreover, the experience was sufficiently painless that I can imagine using the tool regularly. We shall see.

It's desperately trivial and, of course, I should be clearing the sim...

Using Sculptris for SL mesh

The gaming news this week has been all about the Kinect (formerly Natal) controller for the Xbox 360. While some early development work was done for SL, it isn't clear that a reimplementation will be facile, nor that it would be generally useful to educators outside of an individual use, small group or front-of-class setting (... OK, that's not bad).

Meanwhile Jon Himoff of Rezzable has been comparing OpenSim (+ Imprudence -- or Kokua as it is now called) with Unity3D. Having championed the former, he now seems to be shifting towards the latter, most notably due to ease of implementation via the web and on mobile devices. The downside for me is the loss of simple prim-based building, not just for me but for my students. I can, however, appreciate that from a commercial perspective Unity has a lot going for it. Moreover, it may be that a simpler platform with pre-packaged solutions will appeal to the average teacher, as opposed to the technology maven, in a way that SL for various reasons hasn't.

However, the notion that my students will build in Blender for Unity3D (or SL for that matter) seems untenable. On the other hand, digital sculpting software such as Sculptris (presently an alpha version free download but Windows-only) appear to be a much better match for generating "organic"-type SL meshes. The user is presented with either a flat or spherical surface and generates the shape by grabbing, inflating, flattening and smoothing it -- imagine students being able to use something like this inworld via a haptic interface!

The image above shows an SL mesh for a large protein whose overall structure has only been determined at low resolution using electron microscopy. It's only a rough first attempt on my part, generated in Sculptris and ported into SL via MeshLab (Sculptris exports in .obj format) but the use of the intuitive "digital clay" environment was both engaging and informative. True, I could have used real clay but I suspect it's simpler and cleaner to keep such creations on a computer!

Best of both worlds?

The renewal on our island falls due at the end of November and, as blogged previously, I am not pursuing the lock-in option. The Lab's CEO makes it clear in a recent open letter that he sees educators as "important" but basically on a par with any other SL community (apart from the lock-in, presumably) and somewhat less important discount-wise than the land barons. There is some hard-nosed commercial logic in there, albeit tinged with a fairly callous implementation.

While the future points to OpenSim, I'm keeping a base in SL by renting a small parcel on an education sim. The accompanying picture shows a villa temporarily rezzed on the new parcel. I had notions at one time of using it as a multi-purpose centre for the old sim but in the new context it has a ludicrous prim count as well as looking out of place. Nevertheless, it underlines one of the benefits of SL, namely the ready availability of content.

Moreover, if something doesn't exist, there are often handy tools for its construction. The good news, however, is that there are strategies that mean you can derive many of the benefits in OpenSim without breaking the SL Terms of Service, e.g.:

• If you use Puppeteer to animate prims (instead of simpler portable open source animation scripts), you can shoot machinima in SL and use that.
• The particle organ is another tool I am reluctant to abandon. I can't transfer its prim-based output directly but I can use it to generate scripts that should port.
  
• I use PipeMaker for making, um, pipes. I can do that in SL and port the product to other grids. That said, PipeMaker is a freebie and its licence means the script could be ported to OpenSim.
• I can still use tools like ShapeGen, Lathe and Polygon Maker to generate symmetrical shapes in SL and port the products, having swapped my own base prim for the one provided so I am named as creator.
  
• Likewise, I can use Prim Oven and Cord Maker to make sculpted prims I can port to OpenSim. These are used inworld but the trend may be towards standalone applications so that content can be ported to any grid. This is already possible with tools such as SculptyPaint but will presumably become the norm for meshes.
  

Yes, I could use all these in a sandbox but the intention is to have a base for small groups of students, both as a backup to OpenSim and as an orientation centre when visiting edu projects in SL. While some significant science-related sims are closing (e.g. CDC, ACS), there are still interesting places to visit. Hopefully in due course Troy McLuhan's list will be augmented by a complementary one drawn from non-SL grids. While it may not be possible to move seamlessly between SL and OpenSim grids, the HyperGrid affords that opportunity already for OpenSim users (where enabled).

Yeast chromosomes in SL

A blog by Lars Juhl Jensen drew my attention to the haploid yeast genome structure recently published in Nature. The pdb file loads perfectly into UCSF Chimera but in my hands only displays in atom mode (as wire, for example). While this can be saved in x3d and pdb format, neither seems to load in any useful way into MeshLab, the source of the COLLADA dae file needed by SL mesh (caveat: I haven't tried Pymol, Blender or, for that matter, BioBlender). Accordingly, I loaded the pdb file into MeshLab directly and, as you might imagine, it is seriously big and the chromosomes are not colour-coded. For the moment I have had to resort to deleting large chunks in MeshLab as best I can, the idea being to reconstitute the genome subsequently. Anyway, the structure on the right represents the protruding part of chromosome XII (as ever, click through for a larger image). The structure on the left is a smaller chromosome (not to scale, in other words) saved from Chimera in .x3d format, uploaded into Vivaty Studio (the company recently acquired by Microsoft) and saved from there in .obj format for final processing in MeshLab prior to uploading to SL. The extra random messing about generated a marginally better result but, as things stand, neither product is especially pretty nor useful. Just intriguing.

[In case you wondered why the sim is so quiet, it's set to 5 min auto-return and is thus not very attractive to serious mesh developers]