Hidden away on the Second Life Wiki, well not hidden very well, you’ll find the pathfinding cookbook. Described as a ready made solutions for some critters, it has some interesting concepts, it’s also incomplete or written with invisible ink.
In terms of interesting concepts we have such potential solutions as:
Sentient Treasure Chest
Treasure chest with legs. Sleeps, but runs and hides if anyone approaches.
Which makes me wonder if a wizard called Rincewind is having some input over at Linden Lab! There are sections for dogs, cats, rats, snakes, vultures which all sound like feasible critters people may want to engage with and then we get:
Hopping Scarecrow
Wanders about hopping and approaches random people, lingers and moves on.
Hmmmm!
I like the concept of pathfinding but like the pathfinding cookbook, it feels incomplete. The lack of a rigged skeleton for non player characters being a rather large missing piece of the jigsaw, the tier being too damn high being another because people aren’t likely to create too many creative pathfinding sims whilst tier is so high, on the other hand you can make a pathfinding solution on a small parcel.
What the whole pathfinding project is crying out for is the highlighting of some good examples of pathfinding in use. I’ve seen a few comments in pathfinding discussions along the lines of “What can you do with it?” now as we know in Second Life, people’s imaginations can run wild and come up with some chaotic solutions as to what can be done with things, it’s the beauty and beast of Second Life and user generated content. For example you’ll see something and think “Wow good idea” and then you’ll see another use case and think “Oh come on, you didn’t have to use if for that!“, although in the latter case some people will think “Wow good idea” too.
The whole pathfinding thing seems to have gone a little quiet, maybe the arrival of a sentient treasure chest could get the project moving again.
Pathfinding seems weirdly too far ahead of its time to be useful. It is pretty much pure magic compared to the rest of LSL, and it feels like a script which can actually take advantage of it would itself need to be horribly complicated. Animation springs to mind, and something like mesh bones or morphs, or even just hierarchical linking of normal prims would work wonders for such projects.
There probably are some good, easy cases, but even just handling all the errors/events raised by the running pathfinding and interpreting them requires a fair bit of situational awareness and logic.
(I don’t know if LL was playing the *really* long game with pathfinding, or just playing around with the components available, but I sort of have inappropriate dreams about a mix of pathfinding and Versu, and I want to have its babies).
I think you’re bang on the money about pathfinding being too far ahead of its time, it seems like a great concept and then you start piecing it together, I’ve had patrolling prim cubes running around my sim and played with making a building pathfinding friendly but getting anything that would be a good use case going has not happened, although I do like those patrolling prim cubes!
Versu is still coming soon, we should know by now that soon in LL time usually isn’t the same as soon to the rest of the world’s population but a pathfinding version of Versu would be absurdly appealing!