First attempt at a Biozone OAR

As I am temporarily resident at Dreamland Metaverse, I took the opportunity to start to compile an OAR. In this case I think it is more awful than oarsome but I can't see my having the time to do much more in the immediate future and I don't want to come to the end of my stay without doing something (my flight from SL was chaos incarnate in retrospect).

It comes with more imperfections than explanations so please don't redistribute it at this stage. Readers of this blog will have a vague idea of what it's meant to be about even if it's riddled with bugs and errors.

Some small elements expect a MOAP- and mesh-enabled viewer. I haven't tried it with Sim-on-a stick and the file size is 17.6 Mb. I've left it uncompressed so you can upload directly if you or your provider prefer (Dreamland works that way). No warranty of any kind, thoughts on licence to follow (some form of CC).

Please play nice.

Nu Athens

This is part of the atmospheric but alas unfinished Nu Athens RP build by Lordfly Digeridoo who is best known by educators for much of the Clever Zebra architecture in SL. I mention it here (a) because it's a great build and available as an OAR (well, more properly an xml2 file, iirc), and (b) because I have a temporary standalone rental with Dreamland Metaverse.

I need the space in the coming term as well as OpenSim 0.7.3 for mesh and OARs for flexibility. Kudos to Snoopy who runs DM for a very smooth and rapid registration and deployment system. I and my students can still hypergrid to NWG without problems and DM have a great web interface for support of OAR/IAR archiving so I can play with great builds such as this.

Mixing mesh with MOAP to faux texture molecules

I now have a number of potential strategies for colour-coding the mesh molecules:

• use 3Dcoat, BlackSmith3D or Blender to generate UV textures
  
• split the molecules into structure-based components in Chimera, import, add colour conventionally inworld and reassemble (harder than it sounds as the parts seem to get distorted)
  
• import the protein into Blender, split it there, make textures for individual parts and then reassemble inworld (it works for Austin Tait)
  
• use prims to highlight the structural elements
  
• use MOAP to live paint the mesh
  

The movie shows the latter two approaches. In the background with transparent prims added is the entire cholera toxin reduced to 21000 faces in MeshLab after export from Chimera in .wrl format.

In the foreground are the beta strands of chain A coloured using MOAP. The web app I am using is Twiddla and you can see two problems. Firstly, I have no menu and hence can't change colour and secondly I have a problem with brush strokes being replicated, either due to there being multiple repeats or too large a pen size. The balance is certainly not right yet.

Ultimately it may be possible to use a grid to direct texturing as is often done with sculpted prims and hence to have a mix of inworld and offworld texturing and perhaps more than one person active at a time. One approach would be to paint one part of the molecule in red, change the red colour offworld and then paint the next part inworld in red.

Whether a static texture could be imported subsequently to replace the MOAP remains to be seen. The twin redeeming features of this approach is that perfection is not required and the tools and conception are relatively accessible.

Mesh proteins: the quest continues

You may recall that I made some mesh proteins and other biological structures in SL in the very early days following the release of mesh. I'm now trying to improve the process but it is far from simple.

The ribbon-like structures you can see above represent the A chain of cholera toxin [PDB: 1xtc] in cartoon format as exported from UCSF Chimera. The more bulky structure is a surface model for the entire toxin. You will note that some of the ribbon structures appear latticed. These were exported from Chimera in X3D format. By contrast, exporting in stl format preserves the cartoon structure intact (an improvement). In both cases the structures were further processed in MeshLab (using Quadric Edge Collapse Decimation) to reduce the face count to 20000 and convert the structure to dae (COLLADA) format. So far so good. (updated 7 Jan 2012 for latest Chimera version).

The problem comes when you try to generate a UV map to colour the three different structure types. I've tried a couple of approaches. Firstly, I broke the molecule into parts in Chimera and exported the three parts separately before colouring them inworld and attempting to put them back together (structure in background). Notionally that is very simple but in practice it had me going goggle-eyed, even with the white template structure to work from.

The second approach involved "painting" the molecule using one or more packages designed for the purpose. Again, using the obj format file seemed best in terms of compatibility. Thus far the packages that look most hopeful are 3DCoat (trial version) and the free version of BlackSmith3D. While the former appears much easier to use, the latter works OK too and I was able to generate and export a painted UV map from it. The problem I am now faced with, however, is aligning the texture on the mesh. The structure at the front right should be mostly red with a few green splodges (it's just a test); it's mainly splodges.

It may sound odd given the distressed image above (and do click through for detail) but I'm actually quite pleased with this progress, albeit that there's every chance I'm missing more obvious solutions (do tell!). At the moment I run OpenSim 0.7.2 as Sim-on-a-Stick, upload the mesh using SL Viewer 3 and use Dolphin 3 for viewing.

Don't yawn: yet another hypergrid group teleporter

There has been interest lately in standardizing on a hypergrid teleporter design. Gates are much loved by sci-fi fans but there are other options such as the searchable hgurl boards. Personally, I think it too early to settle for one design, not least because the gizmos are not always in evidence and one design rarely fits all. That said, I do like the idea of a common icon to identify such devices.

Gates have the advantage that they are readily identified and groups of people can be channelled through them relatively easily. English people will also appreciate the opportunity to form a queue although this can be challenging for n00bs. Many variations on the gate theme have been devised. I like the in-sim teleport gates on the NWG Physics region as you can see what your destination looks like.

As far as groups are concerned, another option is a script that collects avatar touches and force teleports the "touchees" at intervals. This is particularly good for larger groups whose simultaneous arrival might otherwise overwhelm the receiving region. However, it does suffer from being somewhat mundane. The example used by the HGAC was based on a New York cab. While that's a nice touch, it is not very immersive -- you don't actually get in the taxi. Incidentally, the HGAC is back in session now that jokaydiaGrid has moved to OpenSim 0.7.2.

The design here involves the owner setting a destination based on a notecard-driven dialog (a search option would be nice but at least the teacher can specify/limit the regions available). A particle-based image of the destination then appears although this preseently suffers from persistence issues with the previous image. In a perfect world a map or a sculpted terrain prim might also appear, plus data on current occupancy. It would also be good to extend usage rights to named avatars on a notecard, e.g. other teachers.

The teacher then selects the Rez option to make the teleport sphere appear. The sphere senses nearby avatars and visits them one-by-one based on proximity to the teleporter (nearest go first). Avatars receive an alert by IM when their turn comes. If the avatar touches the sphere, it is teleported; otherwise (as with the first avatar in the movie), the sphere waits a short while before moving onto the next avatar in the list. It self-deletes when the list is exhausted.

The net effect is in my opinion pleasingly kinetic (although the sphere movement is physics-based so your avatar may get swiped), teleporting is phased (remember, good for groups) and avatars can readily see what is happening.

The downside is that avatars need to stay where they were sensed and there will be some lag generated by the script though that might be mitigated a little by using an OSSL function rather than llSensor. For some reason the physics movement also fails to run on at least one other grid I've tried (but that could be the sandbox settings). I think there's also a limit on the number of avatars that can be sensed though it would be easy enough to setup a second teleporter nearby. The teleport prim is current temprez but that is easily fixed.

With the proviso that I haven't tried this out for real, it looks like an interesting variation on the group teleporter theme. The main issue, of course, is finding good places to visit and there teleporters are of limited use. This is an issue I will return to in a later blog.

What's missing from this picture?

Spacecraft? Prim dinosuars? A comical caption? None of these but rather the the green waves that emanate from the white blob/aerial that hovers above avatars when they are using voice to chat (the blobs and waves don't appear in viewer snapshots by default).

John Rogate has Vivox voice working on a couple of the Educator Enclave OSgrid regions including the one above. While I cannily demonstrated my lack of facility with voice in Imprudence, I did make a slight contribution by demonstrating that voice can be used by avatars arriving from other grids by hypergrid teleport (NWG in my case). The voice is also fully spatial; when you choose to attach it to the avatar camera, the volume changes as you move the camera around, just like SL.

If I was going for a quasi-comical caption it would probably be along the lines of "Dr Livingstone, I presume". However, the caption also underpins a resolution to get out and see more of the grid with an emphasis, of course, on education. Formerly this was done under the excellent auspices of the Hypergrid Adventurers Club but this hasn't met in its regular slot for months now and, rather than wait any longer, I think it's time for me to move on.

In which I pose some questions and return to Romenna

I think this is where I'm supposed to make predictions but instead I'll ask some questions about 2012 without pretending to know the answers:

1. Will Kokua, the successor to the Imprudence viewer, be released and will it be SL MOAP- and mesh-compatible?
2. Will a useful hypergrid directory be compiled?
   
3. Will the HyperGrid Adventurers Club (HGAC) start meeting regularly again?
4. Will Kitely launch its subscription model and help educators and their students gain access?
   
5. Will Enclave Harbor open its doors?
6. Will New World Grid upgrade to OpenSim 0.7.2?
7. Will OpenSim move into beta?

If you want some good news, Vivox (the voice engine used by SL) is being made available gratis to non-profit OpenSim grids and many commercial grids are adopting it as well. That should make a big difference to language teachers and their students.

And just as importantly, I can report that the mythical city of Romenna has survived the transition to OpenSim 0.7.2. One of the earliest (24 Jan 2011) and most impressive destinations for the HGAC, Romenna was the creation of Nick Lassard and runs on the very low-power PMgrid (don't bring a friend, don't even cough). PMgrid is centred at 7000, 7000 and now runs OpenSim 0.7.2. Accordingly it is easily accessed via NWG and FrancoGrid using the address secondlife://www.pmgrid.org:8002/. Once on PMgrid, do a map search for Rom and the list of sims will appear. Select your sim and then be incredibly patient and expect to use your alt-zoom camera skills rather than move your avatar.

Then-and-now screenshots suggest that not much has changed since the HGAC visit apart from the move to 0.7.2 and it would not surprise me if the sim never saw its intended RP use. Tangentially, there is a conference in Cardiff in January on the subject of digital preservation of video games and virtual worlds. I have seen many excellent builds in OpenSim lately but have to say that Romenna would be a shoo-in for me even though it has some rough edges.

The bonus image below shows where I ended up when I fell through a gap and ended up in the sewers. I am happy to report that there was indeed light at the end of the tunnel.

Happy New Year!

Merry Whatever!!

I had all sorts of great plans for the image above but apparently "Done is better than perfect". Whatever it is that you celebrate at this time of year, I hope it's a good one and that next year is kind to you too.

Clothes and spex courtesy of New World Grid. No idea where I got the skin/shape from (ReactionGrid?). Shoes (possibly). Windlight setting (it's a mystery). Single prim sculpty deck from The Fairy Godmother a.k.a. Linda Kellie, ditto icicles and tree.