Sex Mod Framework - Development Discussion

[Martin56]

Definitely a Sexlab approach with new features like more interactions, branching and control during the scenes is the way to go.

 

Having tools like integrated voice emulators, standardised sub, dom, victim parametrrs etc, standardised system of tags etc from the beginning will render the system next gen and help content creators make quality content from the right up!

 

It should be based on FNIS/sexlab and definitely not on AAF

[TheStranger]

Nice to know that a sex framework will be made for Starfield. I really feel the game has a lot of potential when it comes to mods.

 

I really hope that high performance will be a goal while developing this system. The flow of the game really breaks when a scene takes more than a few seconds to start. The difference between regular SexLab and SexLab p+ is quite amazing and I hope this framework will be quick to start scenes as well. Perhaps you and Scrab can work together (if they are also interested in Starfield).

 

As far as more specific features go, I have a few things I'd like to see:

Support for using animations on already existing furniture in the world. But limit it to furniture that can be interacted with which is also within the players line of sight (to avoid being teleported into places you shouldn't be and animations playing on obstructed furniture).

 

A built in system for filtering out animations that doesn't fit if your character is restrained in some way (handcuffed, gagged, etc.).

 

Also, making the start and end of scenes more immersive would be great. Perhaps some built in standard undressing/dressing animations would work (one set for headgear and one set for the body would be enough). Also animations for a character stripping a different character (with both friendly and aggressive variants). It would also be neat if the characters would walk to the scene location instead of just teleporting. Also an optional quick fade to black to mask the final positioning of the characters (perhaps also between animation sets) would be great.

 

Another feature I'd like to see (which I understand is probably not going to happen) is support for transition animations that go from/to the default standing and crouched idle animations in the game. That way, if an animator wants, they could have a potentially smooth transition from the regular game idle animations into their sex animations. It would allow a scene to use the default idle animations as a "link" between different animation sets. The animation sets could have their own internal variations and branches as long as they end up in the idle position. But it would probably also require the involved characters to potentially move around while in the idle animation stage if the relative position between the characters aren't standardized, which complicates things, especially when the environment around the characters could contain obstructions. So a scene composed of two animation sets would function like this: [Scene start!] -> [Idle + positioning] -> [Animation set A] -> [Idle + positioning] -> [Animation set B] -> [Idle] -> [Scene completed!]. I see now that I've typed it all out how incredibly difficult it would be to build that kind of system. Oh well, I'll just post it anyway. Perhaps it can be of use to someone even if it's just a pipe dream.

 

Back to a more realistic idea, how about support for masturbation animations while a character is walking. Not very important but could be potentially funny.

 

Anyways, thanks for the work you've done on SexLab, and I wish you luck in creating the Starfield framework!

[TheCaptn]

56 minutes ago, TheStranger said:

Support for using animations on already existing furniture in the world. But limit it to furniture that can be interacted with which is also within the players line of sight (to avoid being teleported into places you shouldn't be and animations playing on obstructed furniture).

 

Based.

I hope furniture gets a lot more consideration as part of the framework and isn't left as an afterthought... I honestly don't remember how SL even handled it now, but AAF was super clunky, with the scanning, and then a menu full of identical items with different distances... I really hope it'll be possible to make all of that much more seamless.

 

Like if you initiate a Rough Standing scene, it should be able to transition to a Rough Blowjob scene on the chair that's right next to you without having to scroll through a series of menus to get there... Likewise if you initiate in intimate kissing scene, there'd be a good reason for it to skip over nearer furniture and move the scene to a Bed Makeout if there's a player owned bed nearby... Stuff like that.

[AJsAlterEgo]

https://www.nexusmods.com/starfield/mods/106 script extender...

[Leito86]

Exciting times.  Here's to hoping creating animations won't be so convoluted.  Doubtful, but one can hope.

Edited September 5, 2023 by Leito86

[jg8ts]

On 9/2/2023 at 4:34 AM, kaxat said:

You are on to something with wanting to thoroughly plan things out. But if you are a software PM you realize how fluid software dev is. Waterfall design frequently backfires. The more agile strategy would be to build modules alongside the core framework. From here you can dogfood and iterate the core framework. Adding hooks, events, and additional API functions, as the modules require. If the modules are open source they can be forked once they reach maturity. If the API calls and events are public those forks will have no problems growing into their own. If the module is truly optional some users may choose to disable it. To me this is the best strategy for developing a versatile mod ecosystem with as many features as SexLab. However I would have a different opinion if I had not watched how AAF unfolded. AAF would have benefited from better planing and discussion like you suggest. But that is the issue with waterfall design. You can always benefit from having more foresight than you actually had. What you have is never enough.

 

@kaxat I agree with this. I haven't done waterfall for close to ten years - just scrum, kanban or some variation. When I said planning it was more of a set of conceptual capabilities that we know we might want.

 

Given the historic issues you raised I'd agree that in this environment it might be better to have a larger core framework that then contains more of the scope, just so long as that scope boundary and framework interface is clear to allow mod developers to start their work. They can manage prioritisation of that backlog based on some mix of demand for their own internal needs and what mod developers would need to get their mods into general availability.

 

Oh, and I'd pay for Patreon for the core framework if their T&Cs allow it.

[SecretFreya]

Once the kinks of a SL framework and nude models are fleshed out (yes that's a double-cringe pun), I look forward to seeing a pregnancy mod along the lines of Soul Gem pregnancy or Hentai pregnancy.

[leonardo7th]

On 9/4/2023 at 4:30 AM, robert_d_negro said:

- the boostpack is the new archery... followers equip it at random moments even in safe zones

I have a hunch about why that is. Currently, you can go into your inventory, and toggle an option that will "hide" your spacesuits in settlements. This will also hide the helmet and boostpack. But you're still wearing them. So when you double jump in a settlement, you can still activate your boostpack, and it appears briefly before it disappears again.

 

My theory is that your followers have the same toggle permanently on, so that they're always wearing their spacesuits/boostpacks. When they try to follow you somewhere they think they need to boost to, their boostpack will appear. At least, that's been my experience.

[Good Provider]

On 9/4/2023 at 5:16 AM, Martin56 said:

Definitely a Sexlab approach with new features like more interactions, branching and control during the scenes is the way to go.

 

Having tools like integrated voice emulators, standardised sub, dom, victim parametrrs etc, standardised system of tags etc from the beginning will render the system next gen and help content creators make quality content from the right up!

 

It should be based on FNIS/sexlab and definitely not on AAF

 

While interactions/branching would be a nice feature, the most important feature is support, or ability to port, existing animations. 

[Akira1364]

On 9/3/2023 at 11:45 AM, asdt123123 said:

Yup decompiled it, it's using Havok.

It's not (for animations), like the annotations are in totally separate files and the behavior data structure is utterly different... They do seem to be using Havok for collison data still, however.

 

TLDR none of FNIS or Nemesis will work with it, a totally new behavior tool will be needed.

Edited September 6, 2023 by Akira1364

[Lordchaosc]

Question though: how does one have sex in space

[epitaph78]

1 hour ago, Lordchaosc said:

Question though: how does one have sex in space

with his/her genitals

[Halstrom]

I hope it doesn't need a FNIS and can just work from the Sexmod, which I'm hoping takes the best features it can from all the different Sex animation mods done in the Beth engine.

[Fl0g]

12 hours ago, Lordchaosc said:

Question though: how does one have sex in space

Just another docking maneuver.

[LunaPapilio]

7 hours ago, Halstrom said:

I hope it doesn't need a FNIS and can just work from the Sexmod, which I'm hoping takes the best features it can from all the different Sex animation mods done in the Beth engine.

Not likely.

 

Unsure about Starfield, but for Fallout 4 for example, the behaviour files are exposed so you can add custom animations directly unlike skyrim which is why it has to be "hacked" to allow it, hence why FNIS is required.

 

Starfield does have it's own skeleton added by bethesda as well, so that will be a significant factor in what is possible to do. Skyrim is the only bethesda game to require a tool like FNIS to inject animations. The animation situation is strange in Starfield.

 

We have .rig files for animation rigs, .as files for animations, and .asx for animation annotations, for both fallout and skyrim, annotations were baked in .hkx files. Starfield is set up like Havok animations were, but it uses a different format and the animation engine to play them is not havok, it's an in-house system built by bethesda. We'll only know more once the official modding tools are out and we can see what we have to work with.

 

Also, if we are lucky, we already know that Creation engine 2 uses The Forge, cross-platform rendering framework, it might include features like the UI or the animation engine, we don't really know currently.

Edited September 6, 2023 by LunaPapilio

[ebbluminous]

Whatever animation solution comes out... Just hoping that the modder(s) will allow it to be fairly open with it so other modders can do a good job of adding to it rather than it get throttled by someone who wants supreme control of it...

[HentaiGnome]

On 8/26/2023 at 11:28 AM, xXDeliteXx said:

What's wrong with going for a more of an Ostim kind of UI and functionality?

 

IMO, OStim is unintuitive as all hell. I must've spent almost an hour setting up OStim only to remove it after 15 minutes of actually trying to get used to it lol. The interface is aesthetically nice to look at it, but functionally it's clunky. It requires a bunch of button presses to get anything done, while with Sexlab it's just straight to the point. To quote a certain famous man, "it just works" with Sexlab. So my desired options are: a) Go the Sims 4 and Wicked Whims route where you can just use the mouse to interact with the UI and get things done, AND/OR b) implement a full native gamepad support (since I'm a gamepad user anyway), AND/OR c) just go with what already is a tried-and-true method, the classic Sexlab way.

Edited September 6, 2023 by HentaiGnome

[SeranaUser]

50 minutes ago, HentaiGnome said:

 

IMO, OStim is unintuitive as all hell. I must've spent almost an hour setting up OStim only to remove it after 15 minutes of actually trying to get used to it lol. The interface is aesthetically nice to look at it, but functionally it's clunky. It requires a bunch of button presses to get anything done, while with Sexlab it's just straight to the point. To quote a certain famous man, "it just works" with Sexlab. So my desired options are: a) Go the Sims 4 and Wicked Whims route where you can just use the mouse to interact with the UI and get things done, AND/OR b) implement a full native gamepad support (since I'm a gamepad user anyway), AND/OR c) just go with what already is a tried-and-true method, the classic Sexlab way.

 

I'd also be a fan of doing as much as practical through menus and dialogue rather than the keyboard. In time there will be enough other mods with the same default keys complicating things.

[SomeUserNameIGuess]

2 hours ago, HentaiGnome said:

 

IMO, OStim is unintuitive as all hell. I must've spent almost an hour setting up OStim only to remove it after 15 minutes of actually trying to get used to it lol. The interface is aesthetically nice to look at it, but functionally it's clunky. It requires a bunch of button presses to get anything done, while with Sexlab it's just straight to the point. To quote a certain famous man, "it just works" with Sexlab. So my desired options are: a) Go the Sims 4 and Wicked Whims route where you can just use the mouse to interact with the UI and get things done, AND/OR b) implement a full native gamepad support (since I'm a gamepad user anyway), AND/OR c) just go with what already is a tried-and-true method, the classic Sexlab way.

 

The new OStim Standalone (no more OSA dependency, implements a significant amount of stuff in C++) is pretty nice IMO and does have a properly gamepad-compatible UI.

[ShadeAnimator]

Hey guys. 

 

On animation front - let me know if there's anything I can help with. I do have very limited time now (work, life, family), so it's unlikely that I will be able to perform any deep hands-on digging, not for a while anyway. But I'm willing to help with anything I can spare.

 

I don't frequent llab, but feel free to find me on Discord, same name on LLab's server. Replies from me will be more frequent there.

[VersuchDrei]

10 hours ago, HentaiGnome said:

a) Go the Sims 4 and Wicked Whims route where you can just use the mouse to interact with the UI and get things done, AND/OR

We had considered this when creating the UI for OStim Standalone, but encountered the problem that in Skyrim the mouse is already used to move the camera.

 

10 hours ago, HentaiGnome said:

c) just go with what already is a tried-and-true method, the classic Sexlab way.

OStim has always had an auto mode. OStim Standalone greatly improved its quality. So you were never forced to use the UI if you didn't want to.

 

 

We have had plans to port OStim to Starfield ever since we first heard about the game. And we will start on that as soon as the CK and Address Library for Starfield release. We are open for cooperation with the SL team, regardless of whether it'll just be a common API or an entirely merged framework. At the end our top priority is the user experience and the users are probably going to appreciate not having to choose between two frameworks because some mods only exist for the one and some only for the other.

So @Ashal if you're up for it we can sit down together and figure something out.

[zaira]

4 hours ago, VersuchDrei said:

At the end our top priority is the user experience

That is wrong - IMHO. Top or at least even priority should be modder support. What good is a nice looking framework without modders using it and delivering (stable) content?

There are enough bad designed, over-engineered and over-complex frameworks in the world that demonstrate this.

Edited September 7, 2023 by zaira

[cheese76]

I am looking forward to and willing to donate

- sex with the corpse

[VersuchDrei]

2 hours ago, zaira said:

Top or at least even priority should be modder support.

I see modability as part of the user experience, the more and better mods there are the better the experience of the end user.

 

2 hours ago, zaira said:

over-complex

With precision always comes complexity. OStims API has some complex parts, especially when it comes to metadata. But those are very powerful tools once you've figured out how to use them. The trick is to find the balance between making a tool powerful and making it easy to use.

[Varithina]

3 hours ago, zaira said:

That is wrong - IMHO. Top or at least even priority should be modder support. What good is a nice looking framework without modders using it and delivering (stable) content?

There are enough bad designed, over-engineered and over-complex frameworks in the world that demonstrate this.

 

Nope no1, is a solid and reliable, as well as stable framework, that does not kill itself when somebody updates their animation packs, but instead just goes, nope animation pack xx is out, write out a error log stating this animation pack is not being used as it is not compatible, with perhaps codes saying why it is not, now let us get on with actually getting the game running, rather than just sitting their going nope not doing anything, I will just sit here and wait and wait and wait.