Bot Paths A question about draggables & sliders

@D.S.Masters - Short answer: it depends on the location of the slider in respect to its placement during the design of the level.

From upcoming documentation:

“No Step Slider”

Normally a bot (both B and R) is able to step on a horizontal slider. However if you tap on the slider at its default (starting) position, an X would appear, showing that he actually can't step onto it. But if you move the slider and try again, the bot would now happily step onto it.

• Final Guard, @Akira Wong, 1 August 2016

In other words, its a "feature." Does that answer your question?

 

It's possible that I have it wrong. Can you please point out some conflicting results? It would be good to get to the bottom of this.

Thanks for sharing!

 

I might need an example to understand. But from what you described, B technically could get to the "adjacent surface" before: when the slider was in the position en route to this "adjacent surface." Before falling in the place of the slider, the slider (it seems) would have made a "proper path" to the "adjacent surface." No?

 

Thank you for explaining further. So what is going on in this example is part of the "autopilot" routine. Just like if you send B to a spot and then drag a draggable of no more than a single-block height to the spot before he gets there, B will climb up onto the draggable.

In this case you simply made him fall instead of climb.

As far as the software is concerned, B is where he is supposed to be - because you sent him to a legitimate spot. Hence B should be able to continue onto any further legitimate spot.

Again, from upcoming documentation:

Spoiler: Autopilot

Autopilot (AKA "Path Completion”) Exploits

So we know that if a spot that B can reach in a level is tapped, B will immediately walk to that spot. But what happens if the path changes during his traverse? It all depends, and how it depends has spawned many new exploits.

Essentially what is going on is that when a spot is tapped, the game runs a calculation. First it calculates if the spot is reachable, and if not, it displays a red X on the spot. If it is reachable, not only does the game display a bubble on the spot, it calculates how many steps and in which direction B will need to make in order to reach the spot. In other words, B is assigned a path and then set on autopilot to complete the path. Here are some variants of how this knowledge has been exploited:

In other words, the game software does not continuously re-calculate all the rules of acceptable locations/situations or not. It simply makes one calculation after you tap a spot. If it is against the rules then an 'X' will appear. Otherwise B will go as instructed and the physics engine will adapt B's movement as the environment changes. Once complete, the process can be repeated.

It's actually a good thing that B can continue on, even if he ends up in an otherwise illegal spot. If not, play would have to be restarted each time.