Simplest way to reduce parts count?

[Arugela]

Is there a simple way they could reduce parts count in stock?

Since it will take along time more than likely to get physics to be unlimited or even increased at this point. What simple solution could be implemented to allow a temp/band-aid solution in stock?

 

My first thought is to have it where you can combine parts and make them one part and simply put them into the VAB/SPH after via existing tools/importing? Could this be easily done to allow multiple stock parts to be stuck together with the same basic stats or effects it would otherwise. Just in this case the connecting physics removed and counted as one part. Would it be hard to make a simple tool(3D) or simple edit tool(2D/text) to edit the stats for this if needed? Could it be done in a way that would be easily kept in game from version to version?

I would think the most complex tool needed would be an updater that could try to change values per part then resave the part for each version manually. Basically something the player has to do but with a slight aide to get them started.

This could be insanely useful for making giant wings and make them act like large single objects. Same with big player designed fuselages and engine pods or cargo containers or anything else you would want a single object. Slap together/save to remove part count then stick on the plane like normal via nodes or radial attachments. A base part could be saved in the part with root before saving as a single object or something and nodes turned off/on if possible to say which ones are active. This could define some parts parameters to make it act as desired. Lots of more creative things could be done with this also.

I imagine the physics externally and the lift/drag and stuff could be left how it would be if they were separate parts. Unless it's easy to modify this also.

Edited December 9, 2016 by Arugela

[paul23]

Parts don't bend in KSP - joints do.

 

So combining many parts together will give a very stiff structure - while the separate parts would actually be weak. This "suggestion" hence would make KSP way easier.

[Arugela]

What if they introduced stats for bend for single parts eventually. It could be modded in the process if desired. If not the point is an optional/unnecessary way to reduce parts count if desired until other options are developed.

It would mostly be for the sake of large parts counts. Bend and joints would be done for realism. You could always make a wing out of multiple parts if you knew the physics to make it bend exactly how much you wanted to add to the realism. It could enhance it for people who know what they are doing potentially. So, maybe it could be used to aid in design for those working on more complex objects. Or at the least ones that don't have immediate or feasible solutions currently in stock. Even modded parts could benefit from this as it is a general tool.

Edited December 9, 2016 by Arugela

[tater]

19 minutes ago, paul23 said:

Parts don't bend in KSP - joints do.

So combining many parts together will give a very stiff structure - while the separate parts would actually be weak. This "suggestion" hence would make KSP way easier.

Because real rockets are incredibly floppy, it's a wonder they get to space, right?

Or because the problem is part count, and the way to "fix" wobbly rockets is to spam struts everywhere... adding more parts.

Procedural parts would be the way. They can be unlocked in the same sizes at the same times as stock. This would not change the number of parts profoundly, but later nodes could unlock longer tanks than currently exist. They could also have different diameters, which would allow more precise options without adding little tanks.

SSTU does this very well. From an aesthetic POV, it also adds nose and mount options, so a tank otherwise very much like a stock one can have variant looks.

[steuben]

There's a mod for that...

 

 

[Raptor9]

42 minutes ago, Arugela said:

Is there a simple way they could reduce parts count?
[snip]

One of the reasons the physics track each part is so if a craft has a collision with another object or the ground, individual parts can be separated as debris.  Or an entire Rapid Unplanned Disassembly event can occur.  Or, if you break off a landing leg due to an improper touchdown, and your lander is now on it's side, you may be stuck on the surface until a rescue mission can arrive.

If you're a stock only kind of guy like me, the offset tool will probably be your best friend.  The offset tool, especially when used in conjunction with the shift key, allows you to place parts in rather far off places from their parent part, but still give the appearance of them being attached through other parts in between.  This saves you from making an internal structure with parts to mount other objects to, which increases your part count.  In the first two images below, the part count was reduced from 34 to 9 using the offset tool instead of direct attachment.

Another way this is useful is so you can make craft components modular and easy to swap out.  Let's say you're building an SR-71 looking plane like in the above two images.  The conventional way to build such a craft would be to attach the engine nacelles to the inner wing segment, and then attach the outer wing segment to the engine nacelles.  If you want to completely change the engine nacelle out with different fuel tank and/or engine combinations, you would need to first remove the outer wing segment.  By attaching the outer wing segment directly to the inner wing segment and then offsetting it so there is a gap between the two, you can swap out engine nacelles freely without reattaching any other parts.

This may not be the solution you were looking for in your post, but these building techniques are useful for keeping your part count low.

[paul23]

5 minutes ago, tater said:

Because real rockets are incredibly floppy, it's a wonder they get to space, right?

Or because the problem is part count, and the way to "fix" wobbly rockets is to spam struts everywhere... adding more parts.

Procedural parts would be the way. They can be unlocked in the same sizes at the same times as stock. This would not change the number of parts profoundly, but later nodes could unlock longer tanks than currently exist. They could also have different diameters, which would allow more precise options without adding little tanks.

SSTU does this very well. From an aesthetic POV, it also adds nose and mount options, so a tank otherwise very much like a stock one can have variant looks.

Real rockets are indeed quite easy to bend - with a lot of development going into the structural strength. (Though given, it's more a problem of buckling than bending). What's more: the amount of mass spent to "stiffening" the rocket (buckling prevents) grows with the length of the rocket squared: 

 

 

The acute numbers don't really matter (it's a game-universe anyways). However the idea should still be there: the longer the rocket the more stiffening you need.

[Racescort666]

2 hours ago, paul23 said:

Real rockets are indeed quite easy to bend - with a lot of development going into the structural strength. (Though given, it's more a problem of buckling than bending). What's more: the amount of mass spent to "stiffening" the rocket (buckling prevents) grows with the length of the rocket squared: 

 

 

The acute numbers don't really matter (it's a game-universe anyways). However the idea should still be there: the longer the rocket the more stiffening you need.

I literally just finished writing a report with this equation in it. 

It's also worth noting that many rockets use pressure stabilized tanks and I'm not sure how that effects the buckling equation.

[tater]

2 hours ago, paul23 said:

Real rockets are indeed quite easy to bend - with a lot of development going into the structural strength. (Though given, it's more a problem of buckling than bending). What's more: the amount of mass spent to "stiffening" the rocket (buckling prevents) grows with the length of the rocket squared: 

 

 

The acute numbers don't really matter (it's a game-universe anyways). However the idea should still be there: the longer the rocket the more stiffening you need.

Yeah, and the dry mass of stock tanks is grossly higher than the real world. Assume the structure is built in. Or have procedural parts that scale dry mass properly to account for the required stiffening.

[FL-T400]

another fix idea: Once a craft is in vaccum, combine like parts(interconnected fuel tanks,cargo holds, even crew modules.) This would make it signifigantly fewer parts, less complicated to do fuel balncing, and even simpler to preform actions. This would also be well balnced, because if you can get it to space it must be pretty solid

[Sharpy]

On 9.12.2016 at 8:08 PM, Raptor9 said:

If you're a stock only kind of guy like me, the offset tool will probably be your best friend.  The offset tool, especially when used in conjunction with the shift key, allows you to place parts in rather far off places from their parent part, but still give the appearance of them being attached through other parts in between. 

This guy has it right. The offset tool is your friend that helps you avoid tons of structural components. Another such friend is the Autostrut.

Also, learn to abuse parts "not for their original purpose". Wing segments, radial intakes, static radiators as structural. Docking ports as decouplers. Stack separators instead of multiple radial decouplers. Many parts have multiple secondary functions and small hidden features that allow you to save on part count massively. Heatshield with its three nodes allows for some extreme clipping while maintaining node-attachment. Fuel pipes are almost entirely obsolete except rare obscure cases; so are standard struts. Docking ports are often superior to decouplers of their "native" size. A probe core like HECS2 will remove the need to carry batteries. Elevons don't need to be attached to a wing to work. A cargo bay makes an excellent airbrake and a decent set of landing legs. Multi-couplers not only can introduce wild vibration problems, they bring very little to the table that can't be achieved through surface attachment.

Also, watch your parts' impact speed rating and adapt your landing systems appropriately. MK2 parts very rarely need more than one parachute to land safely.

[Magzimum]

Just keep things simple.

My most complex ships got so complex because a design did not satisfy the requirements. So, (for example) I added a little staged droptank onto the payload here and there. And I split some stages into 2 smaller asparagus-stages. Then of course the launcher did not satisfy anymore, so that needed a couple of extra SRBs, but because I had to put those on awkward places, they needed some sepatrons to separate safely. And then I had to strut all that wobbliness together. Bottom line is that before you know it, you've added 250 parts.

The actual solution is of course to go back to the drawing board, and keep things simple.

 

[Arugela]

What about when you enjoy making large planes for fun? Or when it is needed for scale on something you are replicating and not just designing.

I personally make large planes because I usually make set of engine and see how far they can go.

Edited December 14, 2016 by Arugela

[Sharpy]

On 14.12.2016 at 6:46 PM, Arugela said:

What about when you enjoy making large planes for fun? Or when it is needed for scale on something you are replicating and not just designing.

I personally make large planes because I usually make set of engine and see how far they can go.

Then you should definitely abandon "pure stock" and use some awesome large airplane part packs. There are things stock is very good at, and things it's not going to be good at, ever, because one of the absolute requisites of stock is "not to confuse newbies".

"There's a mod for that" is a perfectly good answer to most of problems of that origin. Requiring stock to adapt to your playstyle, because your playstyle has deviated too much from what stock caters to, is really not the way. There are way too many playstyles and to try to make the game support all of them would create an unsustainable mess. And there's absolutely no reason why your playstyle should get stock support while other "niche" ones won't. Mods are exactly what is there to cater to them. And if you insist not to use mods - it's kinda like "I like to build airplanes, but they all fly such a short time because rocket engines run out of fuel too fast - and I refuse to use jet engines instead, because I love airplanes that are propelled by rocket engines. Please add rocket engines that are a good propulsion for airplanes." No. The functionality is there already. Use it.

Edited December 16, 2016 by Sharpy

[NSEP]

Use autostrut instead of normal struts. That can help a bit.

[EpicSpaceTroll139]

On 12/14/2016 at 0:46 PM, Arugela said:

What about when you enjoy making large planes for fun? Or when it is needed for scale on something you are replicating and not just designing.

I personally make large planes because I usually make set of engine and see how far they can go.

I'm not opposed to the idea of part welding entirely, (though at a certain point, it becomes kinda cheaty due to infinite stiffness). Not having it didn't stop me from making this absurdity.

Now part count reduction would be awesome, but this is just to point out with sufficient autostruts, regular struts, and rigid attachment, you can make some pretty big structures.

[Leafbaron]

I think a simple work around would be the option to merge all parts to one when saving sub assemblies. When saving you could be prompted to save as one part or save as indivdual parts. This way you could make wings and save them as one part or other various structures and save them as one part. Then when using them from the sub assembly it would only add one part. This would also limit the cheatyness of infinite stiffness to a point. 

[mystik]

On 28/12/2016 at 11:02 PM, EpicSpaceTroll139 said:

I'm not opposed to the idea of part welding entirely, (though at a certain point, it becomes kinda cheaty due to infinite stiffness). Not having it didn't stop me from making this absurdity.

Now part count reduction would be awesome, but this is just to point out with sufficient autostruts, regular struts, and rigid attachment, you can make some pretty big structures.

LOL, DAT NAME, you win the Top Internets Award!

Edited December 30, 2016 by mystik