Distant Lands and OpenMW

General discussion regarding the OpenMW project.
For technical support, please use the Support subforum.
davidcernat
Posts: 232
Joined: 19 Jul 2016, 01:02

Re: Distant Lands and OpenMW

Post by davidcernat » 15 Mar 2020, 22:45

Xenuria wrote:
15 Mar 2020, 21:10
If you have at least 32GB of Ram and a SSD you can do it but it runs at 3-8 fps and takes longer to load a save than GTA Online.
So it's the wrong approach and you shouldn't even present it as any kind of viable option.

User avatar
akortunov
Posts: 758
Joined: 13 Mar 2017, 13:49
Location: Samara, Russian Federation

Re: Distant Lands and OpenMW

Post by akortunov » 16 Mar 2020, 08:43

AnyOldName3 wrote:
10 Mar 2020, 21:34
adding a NiLODNode to the Nif where the near version has the mesh and the far version doesn't will be enough to exclude it from distant versions of any combined meshes it would be part of.
I told several times that wide using of NiLODNode is a bad idea - it adds an update callback to determine if there is a need to switch a LOD level.
A lot of callbacks may lead to a quite large performance drop, when you limited by the main thread.
We will need to use a different approach to store simplified meshes, while NiLODNode should be used to simplify complex objects from active cells (e.g. to replace particle systems by billboards or animated meshes by non-animated).
AnyOldName3 wrote:
10 Mar 2020, 21:34
The polygon count of a whole Morrowind cell isn't wildly different from the polygon count of an overly detailed armour mod for Skyrim or Fallout 4, so we shouldn't need to be very aggressive with culling small objects from distant versions of combined meshes
Well, when I tested bzzt's batching, I got about 10 millions of triangles in scene with about 5 cells viewing distance (while with MGE many players use higher viewing distances). I used smoothed trees and rocks meshes. Interesting, but with such setup OpenMW worked for me much faster without batching at all.

User avatar
AnyOldName3
Posts: 1912
Joined: 26 Nov 2015, 03:25

Re: Distant Lands and OpenMW

Post by AnyOldName3 » 16 Mar 2020, 22:08

I'm not saying we'd use NiLODNodes as-is, or encourage them for lowering the polygon count of things that are already low-poly enough that taking away triangles would give no performance benefit. It would be completely reasonable to make a near version and a far version of a batched cell, though. Your examples of swapping out particle systems and animated meshes for simpler representations based on LOD level are pretty much analogous to my example of swapping an already simple mesh for no mesh at all.

If my rough sketch is accurate, a 90° viewing frustum with five cells of viewing distance would mean about 35 cells might be visible at once, so that's about 300,000 triangles per cell. CBBE for Fallout 4 is around 20,000 triangles and isn't considered to have too many triangles (except for the bit of the engine that handles dismemberment), so I think it's feasible that a badly optimised armour mod would have 15 times that, which means my guesstimate wasn't that far out, which is nice. Anyhow, five million triangles is a perfectly reasonable number even on engines with much more going on than we have, so it's pretty clearly a case of not drawing what we're drawing as efficiently as we could be that's holding us back.
AnyOldName3, Master of Shadows

Xenuria
Posts: 45
Joined: 26 Feb 2017, 22:35

Re: Distant Lands and OpenMW

Post by Xenuria » 19 Mar 2020, 19:06

davidcernat wrote:
15 Mar 2020, 22:45
Xenuria wrote:
15 Mar 2020, 21:10
If you have at least 32GB of Ram and a SSD you can do it but it runs at 3-8 fps and takes longer to load a save than GTA Online.
So it's the wrong approach and you shouldn't even present it as any kind of viable option.
It's not the "wrong" approach it's a optimistic demonstration of how much hard work and effort has gone into making OpenMW. The simple fact that we can at least in theory have a fully rendered world in all it's rich detail without memory leaks or crashes is nothing short of a miracle. In a perfect world, yes we would just render the entire game and build work arounds for AI like the AI distance setting. This isn't a perfect world so we use LOD and distant land. Running with 5 grids loaded is a viable option if you have the Actor distance down to about a 1.5-2 grid distance. You can get 25-30 fps this way provided you have optimized meshes and good hardware. It is not only a "viable option" it's something we get closer and closer too every release.

BloopsCrusts
Posts: 4
Joined: 27 Mar 2020, 18:37

Re: Distant Lands and OpenMW

Post by BloopsCrusts » 27 Mar 2020, 18:55

Hi. I just wanted to reg an account real quick to say I've checked in on this project dozens of times over the last ten years. I think you guys are awesome and I really appreciate everything you're doing. However, I've only installed OpenMW once, back in the mid 2010s I think. The biggest reasons I didn't stay with it are one, the project isn't 1.0 yet and that bothers me a lot, and two, lack of support for distant land. As long as I can get a better graphics experience using MGE with distant statics, that's probably what I'm going to stick with if I ever want to play Morrowind.

On occasion, I've searched the forums to see if there are any discussions about these issues, especially the poor graphics capabilities as compared with MGE. I've noticed anytime a member of these forums has brought it up in the past, they've usually been shouted down and the thread's been locked. I just want to say up front, I'm not trying to ruffle any feathers so please don't flame me. I know it's a volunteer project and I'm really happy you guys are so devoted to keeping MW alive. But the reality is, I hope you can understand, graphics are really important to me. I totally get there might not be a lot of talent for coding distant land and statics on the team. But there's probably a lot of other people like me out there who really value the graphics and aren't willing to trade them off, so I hope that makes sense and this can maybe eventually be given a high priority.

User avatar
AnyOldName3
Posts: 1912
Joined: 26 Nov 2015, 03:25

Re: Distant Lands and OpenMW

Post by AnyOldName3 » 27 Mar 2020, 19:45

It's one of the things with the highest priority, but it's a big job (especially as the point of OpenMW is to be objectively better than things like MGE XE) and has a lower priority than acquiring money to buy food and stuff, eating and sleeping. We have a manpower shortage, so it's really a case of being the change you want to see in the world - I realised that shadows weren't going to get done in a hurry, so I learned the necessary skills and gained the necessary knowledge to implement them myself, then did it during a part of my life when I had enough time to do that. That's how a lot of things get done with open-source stuff - someone gets bored of waiting and does it themselves.
AnyOldName3, Master of Shadows

mikeprichard
Posts: 52
Joined: 16 Dec 2018, 19:42

Re: Distant Lands and OpenMW

Post by mikeprichard » 27 Mar 2020, 20:21

As another person who would love to see this feature, it's exactly as AnyOldName said, and always will be, especially with massive complex projects like this, but equally applicable to every single bugfix and new feature added: until someone steps up with the time and skills (or willingness to learn) and actually does the work, it won't get done. I'm an absolute beginner when it comes to all things technical, so I'm not part of the solution either, and I did recently ask for my own curiosity whether there is anyone actively working on this already (there isn't, but if there had been, I'd have known there was something specific I could follow for updates on git). However, I'd like to reinforce the fact that with this or any other feature (regardless of how many users may or may not be interested in seeing the feature implemented), the only thing any of us on the sidelines can do of any relevance or value is to continue to advertise OpenMW as widely as possible and try to get skilled people involved. Thanks again to all who have contributed to OpenMW for your hard work!

BloopsCrusts
Posts: 4
Joined: 27 Mar 2020, 18:37

Re: Distant Lands and OpenMW

Post by BloopsCrusts » 28 Mar 2020, 15:25

AnyOldName3 wrote:
27 Mar 2020, 19:45
has a lower priority than acquiring money to buy food and stuff, eating and sleeping.
I have looked for where to donate to this project in the past to help grease the wheels of development, but for some reason I've never seen a link to a Patreon, and the OpenMW FAQ says the project refuses to accept any kind of donations. I don't know the first thing about programming. But I will happily put up a $500 bounty right now for anyone who can program in distant land and statics on par with what MGE offers, that is to say, able to see all the way from Vivec to Red Mountain at 60+ FPS. You say "be the change you want to see," offering money is about the only way I know how to do that.

User avatar
cc9cii
Posts: 511
Joined: 28 Mar 2013, 04:01

Re: Distant Lands and OpenMW

Post by cc9cii » 28 Mar 2020, 22:56

Sorry to hijack the thread but does anyone know anything about the distantlod\*.lod files and meshes\landscape\lod\*.nif and how they are used in Oblivion?

User avatar
AnyOldName3
Posts: 1912
Joined: 26 Nov 2015, 03:25

Re: Distant Lands and OpenMW

Post by AnyOldName3 » 29 Mar 2020, 00:14

I have a Patreon, and so do a couple of other developers. Even when I was averaging several hours a day on OpenMW, it wasn't bringing in anywhere close to enough to live on. It's certainly nice to have money that otherwise wouldn't be coming in, but very few people make anything more than a bit of extra beer money from Patreon without producing pornographic content of some kind.

I'm sure whoever ends up implementing distant statics will be very grateful for a $500 donation, but it's very unlikely to be the one thing that makes someone do it who wouldn't have done it otherwise. If we were to commission a contractor or consultant to do it, I'd be surprised if anyone competent enough to actually do it would quote under $25,000, and unsurprised if they quoted over $50,000.

Software development takes a lot of time and some of it can only be done by very clever people.
AnyOldName3, Master of Shadows

Post Reply