Portals to interiors (post 2.0 feature)

Feedback on past, current, and future development.
User avatar
Ravenwing
Posts: 322
Joined: 02 Jan 2016, 02:51

Re: Portals to interiors (post 2.0 feature)

Post by Ravenwing » 07 Aug 2019, 21:02

This feature would be fantastic. I have always hoped for realtime windows that are essentially the same thing, minus being able to move through the portal. I'm sure the computational cost is the reason it's never a thing, but one can dream. Maybe once we have realtime ray-tracing ;)
mjmax wrote:
06 Aug 2019, 13:49
That's a really great idea. If the interiors don't perfectly align with exteriors, you need some sort of non-Euclidean space to make a seamless 3D world work without redesigning everything, and portals are probably the most reasonable solution.
Hasn't a lot of research gone into this for VR stuff? I recall reading some articles about tricking people into thinking they're turning and moving more/less than they actually are. I'm sure there are some differences since that has to do with tricking your body to feel something based on actual movement vs purely visual feedback like we traditionally have.

ezze
Posts: 428
Joined: 21 Nov 2013, 13:20

Re: Portals to interiors (post 2.0 feature)

Post by ezze » 08 Aug 2019, 08:29

Here are some of the thoughts about this topic from almost a decade ago.

It was a good idea then, it is still a good idea. Teleporters need to be replaced with prey-like portals; secondly where now is a teleporter will need to put a functional door that open and closes.

There are quite a few of challenges, at very least:
0. to change all the teleporters with a functional door and a portal is long and tedious
1. the AI should learn how to see through the portals
2. Morrowind teleporters are one-way
3. Internal cells and externals cells behave differently

Now, why those are problems:
0. There are many doors, obviously. However, considering the length of this project and perseverance of the developers this may be a non-problem
1. There are cases where the AIs pass through the teleporters. Those cases at very least must be adapted for correctness. However, it is also a point of gameplay, it would be quite immersion-breaking if moving one step ahead (inside the interior) you get an enemy come yelling at you "NOW YOU DIE!" and just doing one step behind he stops (because you are now outside)
2. In most cases the teleporters come in pairs so it should be possible to replace them with a portal, but there is at least a stronghold where it does not happen. For Indoranyon we would need some kind of one-way portal: from ouside you see the inside but once stepped in there is nothing behind you.
3. Switching behaviors without loading may bring unwanted effects?


Oldie, but still good example

mjmax
Posts: 8
Joined: 06 May 2012, 22:31

Re: Portals to interiors (post 2.0 feature)

Post by mjmax » 08 Aug 2019, 13:03

Ravenwing wrote:
07 Aug 2019, 21:02
Hasn't a lot of research gone into this for VR stuff? I recall reading some articles about tricking people into thinking they're turning and moving more/less than they actually are. I'm sure there are some differences since that has to do with tricking your body to feel something based on actual movement vs purely visual feedback like we traditionally have.
I think you're thinking of redirected walking. Here's a research paper on it. It's about tricking a user into staying within a finite play space by subtly changing how their real-life motion corresponds to the in-game motion. The consensus seems to be that it requires a really large play space to be convincing to the player and minimize contact with the play space boundaries.

Maybe you could use this research to get a sense of whether players would find the portals jarring, but the interior/exterior misalignment represents a bit of a different sort of mismatch. I can't think of a way you could use any of these sort of techniques to make the portals less jarring, though.

Post Reply