Monday, October 26, 2009

Reflections On Usher For EDUCAUSE

cryptbuilding
Location: House of Usher

Colleague Mirt Tenk asked me to reflect on the House of Usher build, for her talk on SL case studies at EDUCAUSE. I'm a head-case, so this will be easy. I need to take stock anyway of what we have done, or not, at the House of Usher.

Mirt asked:
"I'd like to ask if you'd be willing to share information about your House of Usher build. Specifically, I'd like to ask WHAT you would have done differently, and WHY, and what you learned from doing it the way you did it (if not covered by the first two)."
I learned a lot from helping with the Usher build. Here's what I would avoid or repeat:
  • Megaprims: We should have started with them. I stupidly did not do enough research on the issue at Linden's Labyrinth...I mean wiki [read the entry on lag and megaprims]. It turns out that one may use megaprims on a private island, but not on the mainland. Replacing hundreds of 10x10 prims has been tedious in the extreme, but the results can be dramatic: a fivefold reduction in prims for some walls. The House began as terribly laggy and over-primmed. It still is.
  • Student Skills: Our I.T. folks picked a talented student with no SL experience. She became a good builder (the crypt entrance she did is a masterpiece), but I should have done more early on to educate her about lag and over-primming. This is solidly my fault.
  • Sketchup=Messup: Partly we ended up this way because we started with Google Sketchup file and imported the files to SL using a tweak first discussed in New World Notes [have a look at one implementation from Eightbar, a UK site]. I would instead do a rendering on paper or online and then build IN SL from the start, with prims.
  • Tedium? Why Not?: I'm comfortable with it. I love crawling through bibliographies, for instance. Students are impatient, including our builder. I think the Sketchup shortcut comes partly from student discomfort with the painstaking steps of building in SL. Our initial builder had to go back and recheck hundreds of prims so they were properly set to "share with group" and I could texture them. Most of these prims later got replaced by megaprims.
  • Test Build: The House is huge. I wish the team had built something smaller together first. We might have learned the steps to avoid sharing issues, include megaprims, and rethink design by starting smaller, perhaps on an Edgar Allan Poe Visitor Center at ground-level.
  • Up Up and Away: Here our Estate Manager, Kevin Galbraith, had a stroke of genius. We wanted to separate the content from our island, so we put it on a platform far in the sky. It's now inside a big bubble textured with stars and clouds. It keeps the always-midnight feeling of a Poe tale alive, whatever the island's day/night settings.
  • Bring on the Gloom: Textures became my best friend here to keep things dark. I used dull textures and made them even duller by tinting them in grays or browns. Even at midday in SL, the interior is appropriately gloomy. The candles and lamps actually do their work. We used copper sheathing in the crypt, as in Poe's tale, and in the story is gave off its own luminance, so the ambient SL lighting helps retain the effect without the walls having to "glow" (an effect that's easy in SL).
  • Faculty Involvement: Though I will beta-test the House with my first-years, I am still concerned. The first teacher to use the House after me lacks the building skills to see what we needed to change. I wish I'd coerced him to take a building class, so he'd have a sense of the pedagogical necessities of certain features. I could have used a second faculty builder on the site as well. I spent at least 100 hours on this project, time I could have instead spent on writing another juried article or SF story for publication.
First Steps
All that said, I am glad to have gone through the process because it was invaluable to me.

I'm a long eagle, sometimes to the point of arrogance about my skills. Had I infinite time, I would have done it all myself, the way I build model aircraft. Then, however, I'd have made different mistakes. Working with a team that included a faculty colleague, instructional designers, and a student builder showed me the potentials and limitations of a major build in SL.

I can now build with a sense of conserving server resources and allowing for how a team should start its work. We had our eyes on the summer deadlines and on the House itself, and though we did include pedagogical considerations all along, we may have become too enamored of our product, during the process.

A final thought: I have long wanted to write an entry here about why online games bore me.

This post could, however, suffice. Slaying orcs or zombies and solving riddles posed in a game pale, for me, next to the challenges of slaying lag and solving riddles in a maze of my own making.

No comments: