Hold up, let me open word to take notes. I sorta know my way but I'm still somewhat a beginner.
And yet such things happens in real games
Yea, but always thought it was due to time constraints or not making a reliable template at the beginning. It was one of the first things taught to me: make templates for different events and triggers to save time, then use and update them as needed. For professionals to not do that...
Did we know it was turn-based?
In one of the panels shows the game had a combat scene very reminiscent of FF, combat menu and all (page 14, 2nd panel); so I can guess it was.
Either way, plenty of those still have attacks in the overworld, sometimes as a "first strike" triggering mechanic (ie. pmttyd) or to solve puzzles (pmttyd), or sometimes simply pure outright flavour.
I guess, but that's like flipping a coin if it has it or not, and either way, that kind of events doesn't always include movement in 3/4 perspectives in classic ERPGs.
But if we are allowing really shoddy programming, it could honestly be as simple as an input-check that is bugged such that if you mash a you get a frame where the collision code doesn't run at all (though this would let you clip anything anywhere), or at least does so if you mash a on a trigger that does nothing (1 frame where player moves, but instead of collision it runs the trigger, but trigger just returns and does nothing).
![huh :huh: :huh:](data:image/gif;base64,R0lGODlhAQABAIAAAAAAAP///yH5BAEAAAAALAAAAAABAAEAAAIBRAA7)
ha... ha... Yea, that would be awful...
flashbacks to events unaccounted for in own code
But back to more reasonable stuff, maybe it's not that it moves, but that the trigger pauses all physics (like you said, most games stop physics) but when physics return there is a single frame where collision is skipped (because of lack of deltatime maybe? plenty of games where similar stuff has happened for various reasons, some intentional, some just a bug).
Fuck, didn't thought of that. Makes sense. Still isn't that dependent on the load on the hardware? If the load wasn't that big to trigger the bug elsewhere, why did the load suddenly rise? Overflow? Non-crash infinite loop? Too big&full scene?
This would make the bug work on any trigger next to a wall though, but maybe the trigger needs to be somewhat inside the wall for you to not get pushed back out when you leave it (so you can always clip slightly into such walls, but most don't let you through entirely)?
That's the thing, apparently it only happens there so any other triggers are fine.
![thonk :thonk: :thonk:](data:image/gif;base64,R0lGODlhAQABAIAAAAAAAP///yH5BAEAAAAALAAAAAABAAEAAAIBRAA7)
Unless...
Or all triggers returns a position and angle for player to start in when the trigger finishes (to enable cutscenes), and that one is simply a bugged early version without that so it interacts weirdly with players actual position and velocity?
I though of a possibility: that all triggers of similar kind or use the same template are bugged in similar ways, but depending of what they do or where they're used cause different kinds of havoc. For example, if two event uses A or B of the trigger and both A and B are bugged, even if A causes problems, the B bug wouldn't affect anything save occupying resources.
Overall, there are plenty of games that for various reasons has had clipping occur, even only in very specific spots and angles, for weird reasons. You have speedrunners manipulating animations, states (botw loves this, "skew" and stuff), pause menu's/features, and pretty much anything.
I always though the vas majority of that happens fue to hitbox dissonance or not loading the hitboxes in time, not what I would find in old 2D games, but I guess it could happen. And don't get me started on speedrunning using those, feels like they're cheating or something.
![fml :fml: :fml:](data:image/gif;base64,R0lGODlhAQABAIAAAAAAAP///yH5BAEAAAAALAAAAAABAAEAAAIBRAA7)