would it be possible to add a window behavior to snap edges together rather than to automatically and effortlessly glide over inactive windows? for one, it’d make stack and tile easier to actually perform, and for another it’d allow for some static tiling alongside our sweet, wonderful, amazing dynamic tiling. as maybe an option that could be toggled on/off it’d give some extra bite.
As some coder at work used to say, anything that can be logically described can be programmed. 
I admit that how you describe your idea, I don’t know exactly what your idea is…
To automatically tile a window to the one below when their edges touch? That could become messy quickly. What do you mean with static tiling? Tiled windows that cannot be-untiled again?
Maybe, if you can describe your idea more detailed and clear, someone feels motivated to tweak our S&T. OTOH, remember that it was developed with the help of actual studies of user experience by the Auckland University. Extending S&T beyond their findings is not to be done lightly…
I raised this a few years ago, but I would love to see S&T scriptable. maybe in R2 …
it is logically described, we’re just not using a shared language.
“snap” means within the bounds of a certain threshold, an object will sit in one spot during a drag operation. outside that threshold, the object will move normally. it’s a feature common in computer aided design, graphics and video editing applications – anything requiring precise placement of objects in relation to each other. on/off toggle means this behavior can be switched off.
the opposite. they don’t move together, they don’t resize together, they’re just two windows butted up against each other. as opposed to the more dynamic tiling in haiku, where windows match size along their mated dimension and move around together as a cluster. this wouldn’t change any behavior, it’d just look like something. meaning, you could still use the key combination for stacking and tiling. stack & tile would get easier, in fact, since currently the threshold for doing either is small and windows will continue movement past that threshold at regular speed, so there’s a lot of moving back and forth to land in the right spot, a lot of misses before one can actually pull it off. snap would encourage stacking and tiling.
it’s true, they did this and it’s amazing. it can be better. we can do more to bring it front and center.
this helped, thanks.
i remember that thread. that one seems more involved (but i’d love to see that, also – it’d make working with a/v plugins a lot more manageable than the usual scattered windows)