Molecular models in SL

A friend asked about molecular models in SL so rather than putting it on a notecard, here is a summary:

Proteins (see also the secondlifemolecules wiki)

• Hiro Sheridan's sculpted prim system based on Hex and now extended to include textures for surface amino acids, polarity and charge.
  
• Erich Bremer's Monolith (release imminent).
  
• My Protein Rezzing Toolkit based on Troy McLuhan's scripts and extending them by making the components interactive and highlightable. A similar approach has been used by Asorel Todriya, Rez Tone and Archivist Llewellyn (updated 2nd July).
  
• Use of Polyview and 2D animated gif conversion as described previously.
• 2D cartoons created in UCSF Chimera.
  

Small molecules

• Hiro's classic molecule rezzer which is notecard-based (xstreetsl [see orbital models as well under Hiro's merchant id] and ACS Island [includes specimen molecules)]).
  
• Orac which is addressed using the SMILES etc code for the molecule and then conducts its own search. Contact Hiro for availability.

More on Hiro's projects.

There is also a SL group on this topic, viz Molecular Structure.

Helen Keller Day

I was fortunate to be peripherally involved in a resident-inspired and LL-endorsed event to mark Helen Keller Day. I was asked by Jeremy Braver to provide some measure of inworld visual appeal using StoryMachine as a front to a web-based effort to develop a resource wiki to support hearing- and sight-impaired users of virtual worlds. I took the opportunity to add the option that StoryMachine chat its operation such that text2voice applets can register changes for sight-impaired users.

Inevitably traffic on the sim concentrated on the talks (>100 avatars at various times despite, sadly, some griefing) but it was fun to chat to various of the event staff and some old friends. Needless to say, I didn't actually get to any of the talks but highlights were contributions by Keller Johnson, a relative of Helen, and public announcement of Max, the very cool VHH Guidedog of which more anon.

Inworld presentation for University Learning & Teaching Day

My SL presentation at the University Learning & Teaching Day seemed to go OK. I took up the challenge of doing a continuous demo inworld. However, I mixed this with slides not displayed in PowerPoint or an inworld viewer but instead on a simple HUD that I could flick out of the way using Shift-Alt-H. The slide selection was by dialogue rather than Next button. This proved beneficial as I started late and ran out of time so had to improvise the end of the talk. Running the session in this way was not especially high risk as I also had the PowerPoint slides to fall back on. The other gadgets that helped a lot was a TKD navigation HUD that supported drag-n-drop landmarks and my trusty wifi dongle.

I successfully demonstrated that content in SL changes, as a visit to the Theatron sim revealed a different (floating!) theatre from that on the slide made the day before (no surprise as they can be rezzed to order). Trying (and failing) to get in one door gave me the opportunity to demonstrate the classic "camming through wall" technique. The audience guessed that the theatre was probably Roman.

Other observations on this approach were that HUDs fail on sims where you can't run scripts (e.g. Virtual Cell on Genome) and that any special builds need to be sited in an appropriate location and sequence that plays nicely with any improv. I did two custom builds, one on Edgar Dale's (much abused) Cone of Experience and the other charting student feedback (which was mainly positive) and ended up using both before I actually described what we had been attempting to do. Plans for next year sadly had to remain unexplored.

Plenty of questions so it seemed as though pre-prandial hunger had kept the audience awake. Personally, I would happily use this approach again though clearly it would be a little more awkward (and fun) if I had an audience inworld as well.

Rocking the Metaverse, Part 4

This is just to convince myself that I was there. Suffering from woeful bandwidth, I spent ages downloading patches and failing to teleport to the Brandenburg Gate in Twinity where the final gig was held. By all accounts it was wonderful (e.g. image) but sadly for me the stage never rezzed (I think it's the blank wall on the right), the music stuttered badly and I was logged out. Moreover, the other avatars took so long to rez that I must have blundered through them a few times. Apologies. I hope the bruises heal quickly.

Although it's sad to end on a duff note, the Metaverse Tour has been good fun and taken me to virtual worlds I would possibly never otherwise have explored. While the SL experience was the smoothest (I'm not qualified to judge Twinity obviously), ReactionGrid was the most intimate and Metaplace the most fun so each venue had its high notes. And thanks, of course, to the musicians (Dizzy, Grace, Slimmie and DD) for their forebearance, great music and good humour.

Meeting experience: Elluminate, SL and Twitter

ALT kindly hosted a free Elluminate session on Serious Games (SGs) today. The talks were very interesting if possibly a little rushed in parts (not a problem I ever have *cough*). Sara de Freitas described work underway at the Serious Games Institute and I was particularly interested to hear of the underlying rationale, evaluation and development strategies. This was followed by Chris Brannigan from Caspian Learning who described some of the promising new tools, including one of Caspian's, that are lowering the technical barriers and costs of SG development.

I haven't used Elluminate for a while so it was interesting to compare it with the usual (and sometimes choppy) SL meeting experience. Firstly one had to have Java installed (a potential download) and test one's audio (no change there then; see Torley's new vidtut for the SL equivalent).

My guess is that a maximum of 35 out of the anticipated 140 (is that right?) turned up. That would fit in a sim OK. Most people gave their full names but a few didn't and there were no profiles to read. Perhaps as a consequence there was negligible chat going on in advance and no "friendships" to take away. Or tee-shirts.

The moderator was very good at putting the speakers at their ease, explaining the system and keeping the show on the road. It was clear that at least one of the speakers had some technical concerns at the outset so no change there. Most speakers asked for confirmation that their voice was coming over OK and the audience had fun (?) responding with their limited range of emotes (smiley, perplexed, applause and thumbs down).

The Elluminate interface is pretty easy to manage though I didn't figure out how to resize/position the three principal components (participants, chat, slides) so that on my old monitor there wasn't really a decent sized chat box and the slides were a little on the small size. This may be a "feature" of the system or it might be down to my failing to read the joining instructions. No change there either.

Both speakers made use of application sharing, in one case showing a video of an SG, in the other an application used for development. In both cases the result for the viewer was slow to render but just about OK. Synchronous video in SL can be dodgy in my experience though they may be working off dedicated servers here so not necessarily comparable.

The real-time polls were put to good use to determine, for example, the gaming experience of the audience and, at the end, the overall impression of the future of SGs (an audience-inspired question that gave a jolly if, to my mind, potentially over-optimistic result).

The chair kept a good rein on time and ably facilitated a nice range of text questions which were given a fair hearing by both speakers. No attempt was made to use audience voice; I got the impression that this had caused issues in the past.

If there was one thing missing, it was chat on a backchannel (and, of course no distracting IMs), a common feature of SL presentations and increasingly RL ones too, courtesy of Twitter. As mentioned previously, I suspect the audience for the most part did not know one another so that might have been a barrier anyway. However, using chat to line up questions also inhibited more general conversation.

Others have bemoaned the absence of a custom Elluminate-type setting for SL meetings though I suspect it would not take much to make one. Training users in the general vagaries of SL for the purpose of a one-off session is, however, another issue, albeit one worth considering if greater participation from the audience were part of the objective.

Meanwhile Twitter provided a backchannel for some at the Science Learning & Teaching Conference at Heriot-Watt. The fact that there were multiple parallel sessions and relatively few tweeting may have diminished the remote experience but I felt I gained little and could contribute even less by following the hashtag #sltc09. No tee-shirts there either.

For a more tee-shirt-friendly talk, albeit not inworld, see Andy Powell's latest at TERENA.

Occupational therapy in SL

You should go to Eduisland II. There's currently a marvellous display of occupational health exhibits created by graduate students of Thomas Jefferson University. The subjects covered include healthy ageing, stroke, carpal tunnel, school backpacks and dementia. Lots of innovative approaches to presentation, including holodeck rooms that show how layout and content can be modified to suit particular conditions.

I was lucky enough to catch a presentation by one of the students and an excellent job she did too, fluently managing slides, chat and the backchannel. I had a chance to IM with the tutor Zsuzsa Thomsen and she told me that this was their second year of teaching in SL and that the project-based approach they'd taken this year had yielded real dividends. Some of their content had already been displayed on HealthInfo Island with more likely to follow. She was already formulating exciting plans for next year's teaching.

The current exhibits are likely to stay in place until September but I'll certainly be going back before then.

How many UK universities have a presence in Second Life?

Dusan Writer had a mental blip recently and mistakenly inferred from two rather dated lists of universities and colleges in SL that the number inworld was trivial. John Kirriemuir has done a great job of logging UK activity in SL and has suggested that as many as 90% of higher and further education institutions were active in some form. Given the high number of early adopters one might expect in HE, this is not especially surprising. I, however, was interested in "tangible" evidence in the form of a sim or parcel that could reasonably be linked to one of the institutions that report to HESA, i.e. UK/NI and non-Further Education.

I don't claim my study was exhaustive and I didn't consult any email lists. Rather I just checked out John's site, island-hopped a bit, looked at Simteach and googled a little (there are 160+ institutions on HESA's list and I most certainly didn't try to hunt them all down). My educated guess is that around 60 universities and similar own land that is readily located, ranging from multi-sim sites like Edinburgh down to single-student parcels. About 10 of those institutions operate behind closed doors. This is most likely an underestimate and certainly doesn't reflect teaching *activity* as you don't need land for that. Nature's new developments, for example, are available for occasional use by teachers and EduNation also used to act in that fashion and may still. Mind you, a few parcels looked distinctly quiet as though people had recently moved on.

I'll wait for John Kirriemuir to produce the definitive guide to who is doing what and whether it's working. In the meantime I've started to make a simple map-driven inter-sim teleporter based on 47 of the slurls I collected (you can list the entire set in chat by searching on channel 1 for "all"). In a separate initiative, Neosome Anatine has started to map some of the sims to 2D RL maps as discussed previously.

Return to Elucian Islands

Joanna Wombat was the first person to give me building rights on a piece of land in SL just behind what I used to call, erm, the Drexel dungeon on Second Nature (almost two years ago). I think I provided her with a list of outlandish ideas, most of which remain dear to me even today. It was a particular pleasure therefore to attend the opening of the new SkyLabs on Nature's Elucian Islands given that these are spaces for science-related SL newcomers to build on.


There were some great talks by Max Chatnoir, Hiro Sheridan and Eightball Magic though, sadly, I had to leave at that stage. I should like to say I am pleased to note that I took the opportunity to establish a new record for the number of avatars simultaneously present in Max's virtual cell. However, there is also the possibility I should keep such thoughts to myself in future.

Either way, an enjoyable event and hopefully the start of a valuable new initiative. Joanna blogs that much of the space has indeed been allocated so I shall look with interest for future developments.

Further vague ramblings about text-to-speech and back again

I have been playing around with StoryMachine. Firstly I modified a little tourbot (original script from the ICT Library) so that it would visit 12 stations around the main disk representing the hours of the (half) day. Then I added a sensor so that arriving at one of these stations would list the nodes nearby. At the moment the avatar is constrained to visiting each station in turn rather than being able to jump around. There is also no way of reading any node notecards or discerning the nature of any links between them.

Carol kindly came up with the idea of using Vocaroo to store short voice clips. However, both of us have had problems in getting the Flash player to start properly (in my case using Firefox) and, of course, it won't display in the SL client web browser due to the Flash requirement. Vocaroo is good, however, in allowing users to leave short voice messages without having to sign up. The .wav format is also compatible with the parcel media stream though the requirement for Flash would make it hard to automate storing of the info, even if it might be possible to playback relatively painlessly.

I then had a play with the MS Speech Recognition kit that comes with Vista (apparently installable for XP as well). There is some Python code for using the toolkit in the form of the pySpeech module. First attempts suggest that this works quite well. There is also another project called Dragonfly that optionally uses DNS for voice recognition and has an accessory library of control macros for Windows.

There is a complementary text-to-speech library called pyTTS which is based on the MS SAPI toolkit. This not only converts text-to-speech but can also save it in .wav format and optionally play it back. It becomes feasible therefore to record text as speech, upload it to a website and then recall it via the parcel media URL.

However, it isn't clear to me how well the two modules play together so my current thinking is to use pySpeech to (optionally) generate a text snippet from voice input. Once stored in a database, this can then be reconstituted as speech using pyTTS.

Problems include: (i) MS-dependency, (ii) overlap with existing software for TTS (and SL voice), (iii) the fact that not everyone wants to control their PC via speech, (iv) establishing a sensible balance between single- and multi-user experiences, (v) keeping things simple. As intimated previously, these concerns may be sufficient to negate this approach but I'm keeping a reasonably open mind for the moment.

The Seven Pillars

Yesterday my avatar visited Infolit iSchool for one of Sheila Yoshikawa's meetings. The theme was the Seven Pillars of Information Literacy with H1N1 (swine) influenza as an example. I was a little late arriving as I'd forgotten about the new Release Candidate viewer (which seems a little less stable than the previous one to me).

The Seven Pillars provide a useful framework for researching a particular question as well as an opportunity to consider the skills and resources required to successfully carry-out the investigation. We discussed possible flu-related questions for such a study and then had a tour around the existing build.

One of the questions that occurred to me was "Do face masks help prevent transmission of flu?" and Sheila immediately produced some for us to try on (with varying degrees of success). I can't help thinking we all look rather sinister.

Joining up the dots a little, I also suggested that the Theatron3 HUD would be a useful adjunct when it emerges from beta testing as (I think) it could deliver pillar-specific per-avatar audio and video, and also that builds might be used for multiple questions if a holo-emitter were used. This was probably muddying the water a little but I have a tendency to do that.

Show some (HUD-based) emotion

I went to Theatron Island yesterday to see the beta launch of three HUDS developed for Eduserv-funded eponymous project which is recreating theatres of historical interest in SL. This is a long way outside my normal ambit except, of course, I am interested in history and I like to go to the theatre.

Only two of the three HUDs were explored in depth, an audience HUD which supports a limited range of gender-typed emoting as well as permitting the director camera control at performances after the manner of Camsync. The HUD was nicely implemented though at one stage I managed to get locked into a protracted belly-laugh when I was aiming for a healthy chortle. For some reason sound had died so I was slow to recognise the ongoing problem (which was readily cured by teleporting).

Willow Shenlin has done interesting work in generating emotes for teaching use and demonstrating the positive effect they have on student perception of SL. This is something I need to follow-up though I am concerned about ceding screen space to large numbers of HUDs. The Theatron HUDs support both partial transparency and show/reveal in order to limit this encroachment.

A second MS Natal-based video (above), this time from Lionhead, shows the use of the 3D camera to interface a human with a computer-generated avatar (Milo) with mutual recognition of facial cues as well as partial inworld immersion of the human. The former is an area that the Lindens hope to exploit with their webcam-based mapping system.

I do know that some students (and quite possibly business people too) perceive this as very important and Willow's work confirms this. I suspect, however, that, as with voice, I personally would use it only episodically as it is fairly intrusive.

The other two HUDs included one intended to support navigation and location-specific dissemination of information via sound, video and web. This is notecard-based and probably has very general application on educational builds.

The final HUD, which we only scratched the surface of, provides support to actors during plays, coordinating chat and recorded oral delivery of lines as well as supplying animations and stage directions (apparently there is a director's HUD as well).

As a consequence I was able to play the part of Lucifer in Cornish in Ordinalia, a medieval story based on Genesis. While I was probably rubbish, it was good fun and I was able to spot the odd line in the story with the assistance of the very patient Theatron folk (there is an alternative version with a subtitled translation). All in all, a very interesting way of interacting with a build and an interesting contrast to the PIVOTE and Citrus Virtual approaches.

As ever, I forgot to grab images of the Theatron event so my apologies for that.

Rocking the Metaverse, Part 3

Just in case I don't make part 4 (in Twinity next Tuesday), I should briefly blog the Metaplace event yesterday. Roland Legrand has already given the general low-down but I thought I would reflect briefly.

Firstly there was a delay in my avatar to appear and some friends arriving late reported severe lag towards the end even though those already present had no problem once their avatar had resolved from the gas cloud. The maximum concurrency was 85 that I heard which is pretty good. The sound was the best yet. As ever, there were occasional technical glitches but nothing serious.

As per Roland, I found the lack of Alt-Zoom camera shifting disconcerting in the 2.5D environment and it took me a while to figure out the private IMs (needs other avatars around for practice). It even took me a while to work out how to animate my avatar (doh! just click on the avatar and choose from the menu) which underlines the disconnect that even "experienced" people can have, e.g. gamers going into SL.

Towards the end of the show, things got a little, erm, playful (or silly as some might say), as folk began to explore the other animations, most notably the ability to throw snowballs and tomatoes. This, rather than exhaustion or consumption, is the explanation for the prostrate avatars in the image above. From an educators perspective, it does underline the fact that if, as in SL, you give someone, say, a melon gun or an appearance slider, you must expect them to use it.

Cathy Cheal has recently published a short paper on student perceptions of Second Life in the context of a course on virtual worlds that included scripting and building, i.e. advanced techniques. This has subsequently spun off into the media and edublogosphere. She attempted to make the course more engaging by incorporating some gaming-related features. The paper has been highlighted for showing that students do not necessarily take automatically to this type of environment (not a surprise to me). It does highlight, however, the extent to which the students became engaged (there were complaints from students about "spam", for example). Moreover, "oral evidence throughout the course indicated that most students enjoyed what they learned during class sessions and were engaged with their projects".

While others have commented on the small numbers involved, the fact that the students were being asked to shift out of their comfort zone, as well as possible design or implementation issues, the one thing that doesn't seem to have been addressed widely is that the paper specifies that the students came from a non-gaming background and chose to do the course on the basis that it might be more enjoyable than a maths-based alternative. Given the pervasiveness of gaming, it is hardly likely that the students would be well-disposed in that direction at the outset. Moreover, the view persisted that this was a game rather than a learning environment and that having fun was somehow contrary to the academic ethos (the author indicates that the fun angle will not be emphasised in future).

Although the paper has attracted some negative comment, I think it is in fact a valuable example of using feedback constructively and one that I shall consider closely when I re-model my own teaching in the light of feedback.

That said, the student comments on the complexity of the SL user interface for building is familiar territory. I haven't explored Metaplace with students but Daniel Livingstone reports a favourable reaction from his students ("fun simple"). However, these students were, I suspect, pro-gaming.

Microsoft Project Natal

Following on from my "Sitting down considered harmful?" blog, Microsoft have been kind enough to preview Project Natal, an Xbox addon that provides motion capture and voice recognition. Something similar was previously promised by then LL chair Mitch Kapor for SL but somehow never quite arrived. Mind you, Natal apparently won't ship in 2009 so there is still a window of opportunity.

At a more prosaic level, a new freebie motion screen capture tool just caught my eye, viz BBFlashBack Express. It captures screen, audio and webcam and exports to swf, flv and avi.

Sitting down considered harmful?

Noted SL edublogger Iggy O waxes lyrical over the virtues of getting away from the keyboard, something that SLers are signally bad at doing if LL stats are to be believed.

It transpires that he may have a point as a recent paper (also evaluated by F1000 but registration required) indicates that time spent sitting is an index of subsequent mortality risk whether you take recommended levels of exercise or not.

(The usual caveats apply: I am not a doctor and I haven't read the actual paper but it does make one wonder, no?)