What is keeping you from starting your KSP2 Mod?

[Lisias]

On 8/19/2023 at 4:52 AM, Falki said:

there are no competing standards for third-party injectors

That you are aware!

 

On 8/19/2023 at 4:52 AM, Falki said:

As for breaking changes, so far (6 months) there was only one big change that affected most mods and it was quickly addressed in a single afternoon with a patch on the common modding API SpaceWarp.

You are new around here  . There's no way to calm down these fears with **one** example - mainly because the game are still pretty bleak in content, with a very small surface of exposition for regression bugs.

These guys have years of experience on handling insidious bugs created by new releases of the previous game - some of them only really diagnosed years after the initial affected release.

And since we know that a good part of the labourforce that worked on that troubling times is now working on the sequel, it's not only comprehensible  but even logical to think that things will be like that again.

 

On 8/19/2023 at 4:52 AM, Falki said:

Honestly, existing modders are having a blast modding the game. So many ideas, so little time 

Being condescending and dismissive with the old guys (even when they deserve it sometimes, not the case here) definitively is not going to help on luring them in the Scene.

KSP¹ is a different game: you will find it running on kids' notebooks, but also on NASA engineers' computers!

You will not lure engineers to the new game by using tactics that work for kids (except candies, candies will work) - you will get exactly the opposite result by doing that.

Edited August 27, 2023 by Lisias
Some entertaining grammars made less entertaining.

[Royalswissarmyknife]

Game just needs like maybe around 10 years to mature.

And a Fire Spitter port to Ksp 2.

[Falki]

17 hours ago, Lisias said:

That you are aware!

I'm following the scene from day 1 and have been actively contributing to it (my forum post count is not indicative as I don't use it much).

So I'll repeat: there are no competing standards for third party injectors.

BepInEx is the mod loader with an optional SpaceWarp API to make things easier for modders. Sure, nobody can stop people from creating their own mod loaders, but same can be said for KSP1 . Point is that all mods now use BepInEx, and that's been the case for months now.

  

17 hours ago, Lisias said:

Being condescending and dismissive with the old guys (even then they deserve it sometimes, not the case here) definitively is not going to help on luring them in the Scene.

You will not lure engineers to the new game by using tactics that work for kids (except candies, candies will work) - you will get exactly the opposite result by doing that.

I'm sorry you interpreted it that way. My point is that we (the modders) are really enjoying writing mods for the game, modeling new engines, new tanks, writing UI with UITK, so if there are people who are thinking about jumping in, go right ahead, you won't be sorry 

[Lisias]

3 hours ago, Falki said:

So I'll repeat: there are no competing standards for third party injectors.

If you say so.

 

3 hours ago, Falki said:

I'm sorry you interpreted it that way. 

Once bitten, twice shy. This is going to happen lot - in fact, I'm absolutely sure that at least 50% of the bad blood I'm seeing "out there" (as in Reddit, but no only) is due unfortunate choice of words while communicating.

Ok, it's true that "I'm only responsible for what I say, not for what you understand", but since you (AFAIK) have a stake on what we understand, it looks more productive to me to modulate the message to be better understood by your intended audience.

 

3 hours ago, Falki said:

My point is that we (the modders) are really enjoying writing mods for the game, modeling new engines, new tanks, writing UI with UITK, so if there are people who are thinking about jumping in, go right ahead, you won't be sorry 

Now I understand the message.

As a side note, we old guys also had a hell of a good time doing the same in the past - but something changed, and part of the fun had spoiled into misery (sometimes, almost all of it). I'm reasonably convinced that a good part of the reluctance on adopting KSP2 is due most people still grieving about.

It's easy to forget that there's nothing new under the Sun!

[munix]

3 hours ago, Falki said:

BepInEx is the mod loader with an optional SpaceWarp API to make things easier for modders.

Just to add on to this: since version 1.4, SpaceWarp supports using either BepInEx or the official mod loader as the backend for mod loading, and we're actively working on making it so that when the official mod loader gets finished, modders can very easily switch from the BepInEx backend to the official one with very little changes.

[shelshok]

On 6/13/2023 at 11:05 AM, Lisias said:

Users!

Makes no sense on investing time (and modding is a huge time investment!!!) on a thingy with less than 100 users while I get way more results on supporting my add'ons on a game with 1260 users at the same time.

I would imagine there are many users like myself who moved KSP 1 out of the steamapps directory a long time ago, and aren't even counted. Not sure if that's the case for KSP 2.

[Lisias]

15 hours ago, shelshok said:

I would imagine there are many users like myself who moved KSP 1 out of the steamapps directory a long time ago, and aren't even counted. Not sure if that's the case for KSP 2.

Indeed, and the PD Launcher also is a hell of an incentive to do such. And there's also Epic, that did a huge giveaway some months ago, and we don't have the slightest idea of how many of that new users are still around.

But since Steam have about 70% of the market (or at least it was so last time I checked), and since probably more or less the same proportion of KSP2 would have the game moved from the steamapps as KSP¹, the Steam numbers besides not being a precise information, it's still a good way to infer (taken or given) how the games perform compared with each other.

So, besides the absolute number being highly imprecise, the proportions hold - and so it's reasonable to affirm that KSP¹ has 10 to 12 times more players than KSP2, what means that you will have 10 to 12 times better chances on getting engagement on KSP¹.

[Alpha_star]

I personally would say that it's mainly because of two things:

1.KSP2 gets updated a LOT, and as a student I do not have time after EVERY update to re-compile and adapt my mods in order for them to work.

2.Learning how to do KSP1 mods is pretty hard for me, and I don't want to start this all over again for twice, one for the current BelpinEx+ Spacewarp and one for the official KSP2 mod loader.

Still, I believe that the KSP2 development team's main job as of this post is to fix the bugs and improve the optimization, rather than to make a mod loader for their game. Doing so would mean that all mods would have to be re-written and having less or slower content updates/ bug fixes.

Maybe I'll join KSP2's modding community after the official mod loader, but probably not NOW.

Edited August 28, 2023 by Alpha_star
Added the last two paragraphs.

[munix]

53 minutes ago, Alpha_star said:

I don't want to start this all over again for twice, one for the current BelpinEx+ Spacewarp and one for the official KSP2 mod loader.

I would just like to reiterate what I said above:

54 minutes ago, Alpha_star said:

1.KSP2 gets updated a LOT, and as a student I do not have time after EVERY update to re-compile and adapt my mods in order for them to work.

As for this point, none of my mods needed any (patch-related) updates since the release of 0.1.0.0, and that's the case for most mods that have come out so far, I believe.

[Alpha_star]

OK, thanks for mentioning! Still, school starts in several days,  and I won't have too much time for learning.

Edit: I'm a huge fan of your work! Please keep on going!

Edited August 28, 2023 by Alpha_star

[gfdgfherytrey]

Dont know how to code

[munix]

22 hours ago, gfdgfherytrey said:

Dont know how to code

Part mods can be made completely without any coding 

[gfdgfherytrey]

What do you mean?

 

[Superfluous J]

On 8/21/2023 at 4:49 PM, shelshok said:

I would imagine there are many users like myself who moved KSP 1 out of the steamapps directory a long time ago, and aren't even counted. Not sure if that's the case for KSP 2.

Not that an anecdote is the same as data, but as the one user I know of the most I launched KSP2 from Steam *once*.

By accident.

And then I closed the loader immediately to run the game from the command line. Because the loader is stupid and running it makes me feel dirty.

Now whenever I load the game it's from CKAN, and I will never show up on the Steam counts.

...unless they swap the loader for CKAN. Then maybe I'll run it from Steam.

1 hour ago, gfdgfherytrey said:

What do you mean?

 

All you need to make parts is knowledge of how to use whatever 3d software people use to make parts. I don't remember the name because I tried it once and I don't usually hate things but I actually hated it.

[munix]

11 hours ago, gfdgfherytrey said:

What do you mean?

 

All you need is a 3D model, textures, and to configure the part modules in Unity (stuff like setting the mass of the part, what category of part is it, etc.) and then you just click a button to export the mod. No need to code anything.

[HebaruSan]

15 hours ago, Superfluous J said:

...unless they swap the loader for CKAN. Then maybe I'll run it from Steam.

You can configure CKAN to launch KSP2 via Steam:

1. Settings → Game command line
2. steam://run/954850   (← this also works from the OS's run dialog)

... by default that will run the launcher, but any of the available methods for bypassing that should still work:

Edited November 29, 2023 by HebaruSan

[R-T-B]

If you honestly want to know what's keeping me, it's some of the game breaking bugs that I encountered on day 1.  I'm sure they're better now, of course, but really, I am honestly waiting for science modes launch to take a second look at the status of the game.  Hopefully the improvement will be large.

Edited November 29, 2023 by R-T-B

[HomingPigeon]

Long Long time lurker, 4 time poster.  Heh.

I came to this thread seeking clear documentation to augment a language model aimed at facilitating KSP2 code generation for modding (side note: the quickest way to code a mod is to have an LLM help you do it). Instead, what I found is a climate that dampens the spirit of collaboration and innovation. The overt cynicism from a vocal minority of users has really derailed the overall intent of this thread.  Why invest time and energy to bring others innovation, problem solving, and exploration down?  Especially when producing as many potentially viable software candidates as possible is what is inarguably best for the game.

As a seasoned KSP1 enthusiast with over 1000 hours and a senior engineer in machine learning, I urge us to foster not just intra-game collaboration but also cross-game interactions between KSP1 and 2. This approach is crucial for the growth and health of our beloved game.

I challenge everyone here to encourage all avenues of growth. Dismissing ideas and projects prematurely stifles the very creativity we rely on to build something remarkable. Let's iterate, explore, and support each other in this journey.

[Lisias]

1 hour ago, HomingPigeon said:

I came to this thread seeking clear documentation to augment a language model aimed at facilitating KSP2 code generation for modding (side note: the quickest way to code a mod is to have an LLM help you do it). 

Please mind the license. GPL derivative code should be licensed under GPL, don't matter if it was coded using a human brain or a LLM.

[HomingPigeon]

5 hours ago, Lisias said:

Please mind the license. GPL derivative code should be licensed under GPL, don't matter if it was coded using a human brain or a LLM.

False

I am feeding the unofficial docs into an open source and free to use LLM that uses the contents of the documents as a form of RAG.   The code generated from this model is 100% mine and I can license it how I see fit.  

Regardless, my intentions are to make this free as a modding tools for the community.   

Edited December 1, 2023 by HomingPigeon
Clarified and removed statement

[Lisias]

On 11/30/2023 at 9:05 PM, HomingPigeon said:

False

Don't be so "optimistic".

Quote

Apart from the a possibility that saving an LLM output may cause verbatim non-free content to be carried over to the article, these models can produce derivative works. For example, an LLM can rephrase a copyrighted text using fewer, the same, or more words than the original – editors should mind the distinction between a summary and an abridgement.

https://en.m.wikipedia.org/wiki/Wikipedia:Large_language_models_and_copyright

Look carefully what Disney is going to do about AI generated art derivated from Disney I.P.

Disney had paced the Copyright practices in the past (a lot of weird Copyright rules on Software are consequence from  Disney stablishing precedents), and there's a good chance they are going to do it again.

Software is still under the same Copyright Rules as books and art.

— — POST EDIT — — 

That said, I'm not telling you that it's impossible to build a LLM for copyright free code - as long you can prove that you are using only copyright free data to train it.

I suggest you to, also, read about:

Quote

But generative AI has a copyright problem.

The technology relies heavily on large language models (LLMs), which hoover up large swaths of data from every corner of the internet. It uses this data to generate responses to user prompts. 

This process poses an intellectual property risk because much of this data may be copyrighted or otherwise protected by law. Worse, the creators of the original content are rarely paid when LLMs scrape their work. 

<…>

For generative AI to be used responsibly and ethically, a system must be used to identify the provenance of copyrighted data collected by LLMs so that appropriate steps can be taken — either for proper compensation or to avoid lawsuits.

https://blockworks.co/news/ai-copyright-blockchain

Of course I'm not advocating the use of blockchain, other methodologies can accomplish the same goal - take the link above as an evidence that LLM and Copyrights are a problem that you will need to cope with.

— — POST POST EDIT — — 

MOAR LINKS:

https://venturebeat.com/ai/the-copyright-case-against-ai-art-generators-just-got-stronger-with-more-artists-and-evidence/

Edited December 3, 2023 by Lisias
POST POST EDIT

[gfdgfherytrey]

Ok, starting to attempt to make a mod that adds a crude model fo the orignal ksp1 tv-30 engine

[gfdgfherytrey]

22 hours ago, gfdgfherytrey said:

Ok, starting to attempt to make a mod that adds a crude model fo the orignal ksp1 tv-30 engine

Ok, what am I supposed to do, IDK how to use unity lol

Edited December 3, 2023 by gfdgfherytrey
added context

[etmoonshade]

Just some thoughts from someone who's been playing KSP1 since the 0.7.3 days.

I'm not really a modder, in the sense of coding. My first mod was a set of landing skids for a plane, because we didn't have wheels yet.   My second one was a terribly modeled ion engine, because we didn't have those at the time either. They were both, quite frankly, crap (although I could land on the skids... and bounce from 3000m since I had to bump up the crash velocity... )  Neither one was released.

I've got one "released" mod for KSP1, and it's a set of - you guessed it - MM patches. I also have about 20 MM patches I run on my personal game, and those patches cover a lot more than 20 parts. I run about 200 mods, give or take, and I do a lot of changes to get the type of game I want, using the work of modders as a base.

@Lisias hit the nail on the head for me with these parts of their post, regarding my initial experience with KSP modding though:

On 7/8/2023 at 10:54 PM, Lisias said:

And not only that, they made the thing extremely friendly to newbies. Notepad and PaintBrush, is all what you really needed to toy around in your first custom part set (reusing current meshes, of course).

<snip>

Addionaly, for each new modder that published something, you have about 20 (rhetorical number, I pulled it from my… hat..) that did the same for themselves at home but didn't bored to publish their changes.

<snip>

We didn't did what we did because of the MM's many flaws, we did besides them. We did that because MM made it very easy to anyone to do that.

And while I don't have enough context about KSP2 modding yet to strictly agree with this statement by @Kavaeric, it definitely brings up at least a related point for me:

On 7/8/2023 at 11:54 AM, Kavaeric said:

I feel as though much of the effort and design of modding tools for KSP2 have been engineered by and for programmers and coders, with little input or consideration for those who do not come from that background—and there are a lot of us of that background.

I'm one of the folks who saw KSP2 and said "lol no i'll wait." A discussion of my reasoning is beyond the scope of this thread, but I will say that without some equivalent of MM, I won't get KSP2 at this point - I like making way too many tweaks now, and don't want to have to do them every time I update a mod*.

* My personal pain with KSPI-E/WarpPlugin's MM patch setup and weird changes it makes is why we have installation filters in CKAN, in fact - I whined to the CKAN devs 'til they implemented SOMETHING I could use to avoid having to go through and delete those specific patches every time I updated - thanks for that by the way, @HebaruSan - I never said that directly.

[munix]

On 12/9/2023 at 9:47 AM, etmoonshade said:

I'm one of the folks who saw KSP2 and said "lol no i'll wait." A discussion of my reasoning is beyond the scope of this thread, but I will say that without some equivalent of MM, I won't get KSP2 at this point - I like making way too many tweaks now, and don't want to have to do them every time I update a mod*.

We do have a mod for that: Patch Manager. You can see some mods that make use of it here, here and here.