OpenMW 0.45.0

Anything related to PR, release planning and any other non-technical idea how to move the project forward should be discussed here.
User avatar
AnyOldName3
Posts: 1086
Joined: 26 Nov 2015, 03:25

Re: OpenMW 0.45.0

Post by AnyOldName3 » 25 Sep 2018, 23:04

It's one of those maybe-requires-a-major-overhaul-of-a-whole-system type issues which is easier to deal with before the system is merged, so I'll only take that approach if my hand is forced.
AnyOldName3, Master of Shadows

User avatar
wareya
Posts: 218
Joined: 09 May 2015, 13:07

Re: OpenMW 0.45.0

Post by wareya » 26 Sep 2018, 05:16

I checked and vanilla does show shadows for invisible NPCs: https://imgur.com/a/3Ik4WTc
paying attention to #1751/#1816 #2473 #3609 #3862/#3929 #3807 #4297

User avatar
akortunov
Posts: 390
Joined: 13 Mar 2017, 13:49
Location: Samara, Russian Federation
Github profile: https://github.com/akortunov

Re: OpenMW 0.45.0

Post by akortunov » 26 Sep 2018, 08:01

If shadows from invisible actors are blocker, can we just ignore such actors in shadows calculations (by using masks or something like that)?

User avatar
AnyOldName3
Posts: 1086
Joined: 26 Nov 2015, 03:25

Re: OpenMW 0.45.0

Post by AnyOldName3 » 26 Sep 2018, 12:02

The shadows for invisible actors issue in and of itself is minor. The real issue is that my efforts to solve it have been thwarted by other bugs that aren't a problem otherwise, but seem pretty major.

Using nodemasks would be pretty invasive as we'd have to give every actor an 'actor that isn't invisible' mask and take it away when an invisibility effect is applied and bring it back when it goes. Some invisibility effects leave actors translucent (it might just be the player, though, so you can see where you are), so non-shadow-map rendering would need to use the existing mask so that's visible. I think in cases where an invisibility effect makes something actually invisible, the current implementation does actually ignore it in the shadow map.
AnyOldName3, Master of Shadows

User avatar
wareya
Posts: 218
Joined: 09 May 2015, 13:07

Re: OpenMW 0.45.0

Post by wareya » 26 Sep 2018, 12:42

Using the nodemask system is definitely the right way to go for something like this, even if it seems gross. "Doesn't cast shadows" is a pretty general thing. For example, future full conversions would probably love to do something like apply it to animated foliage and have AO simulate their internal shadows instead.
paying attention to #1751/#1816 #2473 #3609 #3862/#3929 #3807 #4297

User avatar
AnyOldName3
Posts: 1086
Joined: 26 Nov 2015, 03:25

Re: OpenMW 0.45.0

Post by AnyOldName3 » 26 Sep 2018, 14:29

We can't have a "doesn't cast shadows" nodemask as they're ORed with the traversal mask. We'd have to have a duplicate of every mask such that one was "<thing> which doesn't cast shadows" and the other was "<thing> that does cast shadows", with <thing> being actor, static, effect etc.
AnyOldName3, Master of Shadows

User avatar
wareya
Posts: 218
Joined: 09 May 2015, 13:07

Re: OpenMW 0.45.0

Post by wareya » 27 Sep 2018, 00:08

That sounds like a problem.
paying attention to #1751/#1816 #2473 #3609 #3862/#3929 #3807 #4297

WhiteGuar
Posts: 6
Joined: 10 May 2017, 14:20

Re: OpenMW 0.45.0

Post by WhiteGuar » 14 Oct 2018, 10:45

Will ranged criticals be enabled in 0.45? Sorry I didn't know where to look

User avatar
Capostrophic
Posts: 428
Joined: 22 Feb 2016, 20:32

Re: OpenMW 0.45.0

Post by Capostrophic » 14 Oct 2018, 11:06

They're already "enabled" in the sense that they work in the current code - and nightly builds - and they work the same as melee critical hits. There are three conditions:
1) The enemy is not in combat
2) The enemy doesn't see you (line of sight check fails for whatever reason, e.g. you're invisible)
3) The hit is successful

Then the hit is considered critical and has a damage multiplier, a special sound and a message.
shitty lingua anglica grammar ftw

WhiteGuar
Posts: 6
Joined: 10 May 2017, 14:20

Re: OpenMW 0.45.0

Post by WhiteGuar » 14 Oct 2018, 11:49

Thank you this is awesome. I read somewhere this would have been implemented after v 1.00. Now that I think about it I'm still using 0.43 so maybe that's why it doesn't work (chameleon 80% + sneak 100 should be more than enough)

Post Reply

Who is online

Users browsing this forum: No registered users and 5 guests