I get the impression from the above that one possible use of Splinter would be to have a rapid ability to switch between desktop environments. So instead of having everything you frequently use in one place, all the time, you could define a desktop for, I dunno, gaming, another for work, another for surfing... allow for all the crossover that happens between those things too. If I'm reading this stuff right, then an important marker of difference between the standard Windows desktop and Splinter could be that a "Splinter Desktop" isn't just a wallpaper, it's a wallpaper with embedded functionality. So if you can switch to a new desktop by choosing it off a pane of available desktops you can also be changing your working environment entirely, not just the way it looks.
Or am I still misunderstanding what this is supposed to be about?
-oblivion
You're definitely on a good track there, I'll reverse it a little and say that it is a "dynamic desktop" so for example while I don't have the energy to do it right now/never, You could program the splinter with my current layout of icons/(Triggers?) and be able to play "Jack in the box" with it all, for days it's too much noise, "put it all away". Wallpaper not even required.
However another new idea occurred to me this morning. Dmd it struck me that a current downside is that it all seems to be coded by hand. What if there were a master-scripting logic that resides in a master file, that the program reads/imports and then it spins out all these linked nodes?
Going on from there, I think it would be fun if it made a desktop chess game viewer. How? Because on one side chess notation is standard, maybe a middleware connector to slightly expand the assumptions behind the notation, that's relative cake for you hotshot dev types, then Splinter starts from a stock layout of pieces and reads in the notation and moves the pieces by using the notation moves as triggers! It's like BattleChess for 2012!
