#change11: third time lucky?

This is my third attempt at an introductory post. The first was too boring (this is my third MOOC, hope I last a bit longer this time), the second too whacky (a meta-story based on discovery of a strange calendar device in a space colony mysteriously abandoned by its inhabitants).

I wil, however, use OpenSim (an open source immersive environment closely based on Second Life) and, provided time allows, create artefacts that model my understanding of the course even if I abandon the concept of a meta-story. If anyone is interested in OpenSim, do let me know.

Most likely I will dip in for sessions rather than aim for blanket attendance. Clark Aldrich and Diana Laurillard are standout contributors for me though many of the others familiar and deserving of equal attention. Even if I falter, I will scan the daily newsletter and follow a few links to keep pseudo-current.

Anyway, good luck to the spaceship #change11. Prepare for launch!

In which I attempt to keep track of the future

According to what I could gather from the iffy audio of John Naughton's talk at ALT-C (better version here), it's impossible to predict the future. The best we can do is recognise:

• The future is already here — it's just not very evenly distributed (obligatory Gibson quote; I suspect the timescale here is about 20 years)
  
• What's "here" gets mashed together in ineffable ways to generate stuff that gets adoipted or not
  

Not that this stops people guessing (or my blogging it interminably so I don't forget).

As far as the Gartner Hype Cycle is concerned, virtual worlds, for example, are still in the Trough of Disillusionment, at least five years from mainstream adoption. However, the criteria for emergence from the Trough include evolution of second generation products and this is happening at Linden Lab and more generally.

One approach to second-guessing the future is to get some wise people together and get them to prognosticate. The JISC has done this recently in a joint project with the New Media Consortium. Their "ones to watch" in the near term:

Time to adoption: One year or less
Cloud computing, mobiles, open content, tablets

Two to Three Years
Game-Based Learning, Learning Analytics, Semantic Applications, New Scholarship

Four to Five Years
Augmented Reality, Collective Intelligence, Smart Objects, Telepresence

Many of these have some overlap with virtual worlds. We are seeing more open content, we want to be able to use virtual worlds on tablets and mobiles, they have potential for game-based learning and AR was one of the components of the original Metaverse Roadmap. Telepresence is interesting as it seeks to provide a more authentic shared presence, albeit at significantly greater cost, than virtual worlds.

JISC CETIS also does an informal horizon scan.

Some times it helps to see where we've been. There's a great presentation on the history and future of learning technology by Nick Shackleton-Jones from BP (via Seb Schmoller). While virtual worlds get short shrift around 2003, Nick can't quite kick the futuristic habit as (spoiler alert) we come back to the post-tablet sci-fi glasses in 2020 (or virtual retina devices if you prefer the Rainbows End version).

In terms of what is accessible now, Keele is doing interesting work with CAVE approaches at the molecular, anatomical and simulated ward levels (via Kali Pizzaro). A combination of the virtual and face-to-face has much to offer as Rich White has demonstrated with OpenSim.

Finally, if you want to see a Second Life-oriented perspective on the future, Patty Rangel's SLCC talk based on her graduate studies at Singularity U is interesting. While she talks about VR contact lenses, not to mention mind-control, she also references the heliodisplay technology which does away with the glasses/CAVE altogether (though I guess they could be used to overlay the real-world overlay). Patty's vision is a billion users in 10 years so some way to go yet.

So nothing new really but that's what we expected. Right?

Hypergrid Educators Map and Meetup this Friday

There was supposed to be an informal meeting this Friday for OpenSim/Hypergrid educators. Unfortunately a variety of factors have conspired to make this difficult so it is going to be more of a dropin than a meetup. I'll be on BioZone on New World Grid from 14-15:00 BST (06:00 SLT) on Friday 9th September so do come and say hello. New World Grid is now on OpenSim 0.7.1.1 so you won't be able to jump from ReactionGrid or JokaydiaGrid but you can register a NWG avatar and jump to the OSGrid Educator Enclave and visit John Rogate's freebie centres, including the new full-sim OAR from Linda Kellie (but remember to delete any inappropriate couples animations before use in teaching).

I'm not planning any formal talks now (maybe we can schedule a session another time) but if you want to catch Justin Clark-Casey, he is speaking in SL TODAY (7th Sept). Further info here.

If you come inworld at NWG Welcome, you can either do a map search for BioZone or walk into the adjacent building. There's a scripted teleport in the gallery at the end on the right. Just touch the image and that should get you to the sim. Once there, touch the big blue arrow and you should arrive at a sky-platform as per the picture. As ever, arrive a little early if possible. If you're using a series 2 viewer, don't be surprised if you freeze and have to close the viewer and login a second time. This seemed to be less of a problem last week (I suspect caching is part of it) but latterly has reappeared. Regrettably I've not had time to delag the sim so you will have to take us as you find us.

That's not to say that I haven't been doing at least some of what I'd envisaged. The picture above shows a model of the lower left quadrant of the hypergrid centred on the plywood structure on which my avatar stands (click through for more detail) which is at 10000, 10000. This corresponds to Wright Plaza on OSGrid. OSGrid now extends to 20000, 20000 but I'm not aware of any educators beyond the area shown though I haven't investigated ScienceSim in sufficient detail (this has the Newton region at 10000, 10000 -- remember that grids are essentially separate or, if you prefer, overlap as shown here).

Looking into the distance, the nearest blue platform represents the central areas of New World Grid with some green markers for regions of particular educator interest. Below it is the OSGrid Educator Enclave showing as mainly green. Note that these are magnified 100x compared to the region they represent. The maps were made on a best endeavours basis using Microsoft ICE to meld together screen captures and ImageMagick to cut them into suitable sizes for upload.

In the distance you can see JokaydiaGrid and ReactionGrid, the latter being the lower of the two.

The blue and pink quadrats show distances of 4096 regions in either direction, the limits for grid jumps. I've used different colours as the two lower grids are currently on different versions of the Hypergrid so there is no possibility of direct movement between them and NWG. Hopefully that situation will change soon but even then you will need to use an intermediate Hypergrid Gateway (there's a green marker for one on the platform in the overlap area). The maps themselves are possibly dated and in most cases small fragments of much larger ones. Other grids with edu content are missing, most notably closed institution-run grids and non-grid based entities such as Kitely. Of course, presence on the map does not necessarily reflect access but I thought it worth marking the fact that Drexel and UC Irvine have a presence on OSGrid, for example.

Most of the map-based markers are generated on-the-fly from notecards. The data from these are stored in a Google spreadsheet. A separate doc contains a list of educators whose id is used as contact in the region database (well, the list is just me at the moment). Clearly it would be better at some stage if these migrated to a database but I like the idea that anyone can edit, at least at this early stage. I want the markers themselves to be interactive, e.g. to offer a teleport and display metadata visually, but that is for the future. I'll post the URLs for the spreadsheets nearer the time.

Hope to see some of you on Friday!

(hat-tip to Jeff Kelley for the prototype of the map display)

A new day

Thanks to Olivier, New World Grid is now running on OpenSim version 0.7.1.1. There are a few things that are broken, most notably landmarks and, until yesterday, llHttpRequest (now fixed -- thanks, Olivier!), but it is very early days and the crucial aspect is that regions and inventory are functional. The llHttpRequest script function is used by a number of my gadgets and presumably also by the HGURL boards that I often use for hypergrid jumps as these are also temporarily broken (now working again after a fashion).

Although mesh is supported in this version of OpenSim, it is not compatible with current viewers and the Linden Lab version rolled out recently. You can upload mesh with Kirstens S21r7 viewer (an old release). However, when you view with SL Viewer 3, only the mesh bounding box is present (the avatar bumps into an invisible prim) and nothing else renders, not even the markers normally present in viewers that do not support mesh. If you want to use mesh (and I've only played briefly), you have to use Kirstens S21r7 to view as well and even then the bounding box does not resize when you change the size of the mesh. The bounding box also seems to persist after the object is deleted so that a region restart is required to remove it. In short, I think mesh is best left alone for the present.

While many of the problems I reported previously have been resolved (my avatar now rezzes in Series 1, 2 and 3 viewers, I can access my main sim from those viewers and shared media works), the hypergrid remains an issue. That said, osTeleportAgent appears to work so I can now use a scripted teleport to get to Wright Plaza on OSGrid as OpenSim versions 0.7.1.1 and 0.7.2 both use the same Hypergrid version. FrancoGrid has also made the transition and I briefly visited the Orion sim on that grid. The viewer map is also non-functional in the viewers tried thus far though map searches do work.

I've been using SL Viewer 3, Kirstens S21r9 and Dolphin 2. At the moment I am leaning towards Dolphin 2 which is simpler than some and feels less laggy too though slightly more prone to crash during scripting. It handles shared media and has a grid manager I can configure. It lacks mesh but that's no longer a major prerequisite. One slight annoyance when building is the absence of support for prim dimensions greater than 10 m.

Another option would be Firestorm which is apparently due out with a grid manager and mesh in the next two weeks but I really have to make a decision before that.

Although some things that were broken now work and, to a lesser extent, vice versa, my general feeling is one of definite progress and things coming together, both for me personally and for the wider OpenSim community.

Given that mesh is a no-show for molecular structures, I decided to play a little with those I already have. I used Pipemaker to build a pathway round the multi-sculpted LysinB model. The model is phantom and the pipe is hollow so an experienced, sensible-sized avatar can walk and fly it. It's probably not the last word in such things as it is actually very hard to navigate, the problem being one of disorientation and steep climbs with twists that require jumps or flight. It's quite easy to double back on yourself so I set a feature that colours a prim. Anyway, a good test of camera and avatar control, a bit of fun and maybe I can build some science into it too.