stock procedural parts and welding

[jab136]

I would think that making the tanks and engines procedural and only putting one model in would help performance, it would also give us better control by putting different buttons on the right click menu for length and diameter of the tanks, then we wouldn't need multiple tanks to make the rocket longer, this would also avoid the issues of weak joints in the stock game. also if we could weld parts it would help with performance on computers with lower performance. Also, as a side note, would it be possible to make "grains" select able on SRB's? so that we could select if we want constant burn rate, or a higher burn rate at ignition that tapers off as it burns.

Edit: forget the procedural engines, except maybe for tweakable grains and procedural SRB's, also, procedural wings would be really helpful

Edited September 10, 2015 by jab136

[panzer1b]

Id love procedural wings and fuel tanks, as for welding, if we get proc parts it wouldnt be needed. That said, squad hates procedural parts and thinks its against KSP's style so i HIGHLY doubt anything will ever be implemented (after the fairings that were added in because well they were kinda necessary with the aero). Ohh, and the fairings are horribly unoptimized too, so much lag and they wobble like nothing else.

[aidyn89]

As far as I know, this wouldn't be possible without major modifications to the game, as it is not using a procedural system in any fashion. Also, even if it was, it wouldn't change performance, only the size of the game. Reducing variety doesn't change individual complexity. This would also make the game (in my opinion) much less interesting, as everything would always be the same.

Also, about your point on welding parts together, can you elaborate, because I don't see how that could/would help performance. I can only see this reducing them. If what you mean to say is to weld, let's say, fuel tanks together to increase structural integrity, that is already being done by the game, but as with any welding done, there is a weight limit that can be applied: more than that and the welding breaks. It is the case in this game too, which is why you can place Strut connectors: they greatly enhance stability and can eliminate any chance of the vessel collapsing on itself under its own weight.

[Red Iron Crown]

As far as I know, this wouldn't be possible without major modifications to the game, as it is not using a procedural system in any fashion.

Struts, fuel lines, and fairings are already procedural. Mods already do procedural tanks and wings, among many other things.

Also, even if it was, it wouldn't change performance, only the size of the game.

Procedurals reduce part count, which is the largest single factor in KSP performance.

Reducing variety doesn't change individual complexity. This would also make the game (in my opinion) much less interesting, as everything would always be the same.

I honestly think there would be more variety, we would not be limited to the relatively few parts we have now but a near infinite variety.

Also, about your point on welding parts together, can you elaborate, because I don't see how that could/would help performance. I can only see this reducing them. If what you mean to say is to weld, let's say, fuel tanks together to increase structural integrity, that is already being done by the game, but as with any welding done, there is a weight limit that can be applied: more than that and the welding breaks. It is the case in this game too, which is why you can place Strut connectors: they greatly enhance stability and can eliminate any chance of the vessel collapsing on itself under its own weight.

Part count reduction, same as with procedurals.

For a long time I've been against procedurals in the stock game, as I think Lego-like construction is more approachable for players. For some things, they do make a lot of sense, though, like wings and the existing stock fairings. I'm starting to come around to procedural tanks, with the very least being content selection (this solves a bunch of problems and opens new possibilities). Procedural engines I'm less keen on, to a large degree selection of engine is the biggest design factor in a craft.

[jab136]

if they did the procedural system correctly it would actually increase variety, they could just put a few textures in that can be swapped. as for welding, it would basically entail taking a fuselage and a wing and turning them into one part instead of two, which would allow for fewer physics calculations (you could still put some sort of failure criteria on the weld, but it would still probably help) also, the other benefit of procedural parts is that surface attach would work better for stacks of parts or wings, instead of just having one fuselage part attached via surface and the rest hanging off of it, they could surface attach along the entire length, thus increasing strength with less struts necessary, and thus less parts to model and apply physics to.

Edit: yah, come to think of it, engines probably wouldn't work for the most part, but tanks and wings would benefit greatly

Edited September 10, 2015 by jab136

[Kegereneku]

Just to counter-arguments,

Procedural content, especially if they are active game-mechanic (engine & tank) rather than passive one (fairing), are so much of a pain to balance right that you recognize a game built around them by their simplicity and monotonous gameplay (MMO like Eve Online or Elite Dangerous).

This is not to say that procedural generation do not have its use, quite the contrary. But it is hard to keep a game balanced if several features can now fail in an infinite number of way.

- The fairing being procedural is... mostly complicated because of aerodynamic. Just to make sure everybody know, Realism don't necessarily equate to "more fun".

- Tank being procedural cause problems on the little balance there is, you would have to balance the system for Technological-progression anyway, and prevent the game from becoming too easy, or some Exploit (at least not make clipping easier)

- Engine being procedural, as much as I see an interest, would be the trickiest as any change to them or the ability to massively parallel them is enough to break the game.

Another unforseen consequences is that you need point/form of reference to estimate the size and "range" (dV bugdet) of a spaceship, yours or seen in a screenshot. Procedural many-thing, even with standardized capsule/hatch can make this hard. Equally there is an innate interest in SOLVING THE PUZZLE that are Standardized Part Size & Spec into whatever you want which isn't available the same with purely procedural.

So I mostly wanted to remind people that LEGO-style construction is a perfectly legitimate and voluntary game-design choice and not any sign of weakness or bad/incorrect design, not everybody have the time (or knowledge) to finely shape the fuel & aerodynamic until it work.

In the end : I would rather improve the TWEKABLE than go further on the procedural route.

(and this post got longer than wanted, that's all from me)

[jab136]

counter-counter argument

it would be easy to implement limits on both tank radius and length, wing size wouldn't be a balance issue as all those do is add mass and drag, with the tradeoff being that you can now get lift, so you could easily have no limits on those after they are unlocked and not mess with the balance. I agree that the "lego" style is a design choice, however it comes at a significant performance cost, and also makes the vehicles themselves structurally weaker. the lego style would remain, just with a few more procedural parts and therefore significantly more design options, which would only serve to make an already good game that much better. I also edited the OP, as the procedural engines is probably not a great idea.

[cfds]

"Automatic welding" could even provide a compromise between LEGO and procedural:

- Build a tank from (for example) 0.5m segments, the whole tank becomes a single part as far the physics are concerned.

- Build an engine from a basic core (fuel pumps and burn chamber), a nozzle (light and low expansion for lower stage engine, heavy and high expansion for upper stage engine) and an optional gimbal module. Kind of "clicking the modules together".