Stock fairings: Procedural or not?

[regex]

I'm sure that procedural tanks would cause a big number of people to be upset (sigh )

Which is really sad to me. Procedural tanks are literally the single most parts-count-reducing improvement mod in the game right now. If I want a 5x35m tank for an orbiter, BAM, I have it. If I need a 0.75m radius circular tank for a lander, BAM, I have it. If I want a nosecone-shaped tank for the front of my spaceplane, BAM, I have it. Because of this, most of my launchers for RSS come in under 25 parts. I built a launcher that could take nearly 600 tons to LEO and the entire thing was (IIRC) 52 parts of which 23 were engines, and including fairings. If I had tried to build the 30m wide lower stage with stock tanks I would surely have had a parts count upwards of 300, which is patently ridiculous.

[DBowman]

One trouble I've had with procedural fairings is sometimes it is too 'greedy' and although I use the tweaking menu to set a specific size (a radius) in order to make some stuff 'stick out' of the fairing it ignores my instructions and encloses the whole thing. So I think it's important that the procedural stuff is still reliably controllable or you can end up fighting the editor - which is not the fun part...

Also I don't think you can attach things to the current procedural fairings - which makes sense I guess if you think of them as 'pure fairings' - but if you think of them as a 'fancy building block' then it would be nice to be able to attach stuff. e.g. Von Braun's lunar plans involved using the the sides of the cargo containers as dwellings (think half cylinder fairings as huts) - which would be kind of cool to be able to do.

[Nitrous Oxide]

What is this terrible argument of "IRL, customers have to fit their payload inside certain fairing diameters"? Hello! We're a space program! We build & designs our own rockets, obviously we need the ability to design and build our own fairings to fit our payloads.

I can't even believe some people would rather have 18 parts that confine them to a fixed-style of play vs 6 parts that allow you to still build stuff in traditional KSP-fashion...

People who don't want procedural parts just have a phobia of having to make their own choices it seems. Always easier to work inside the box than to think outside of it, metaphorically.

Edited February 16, 2015 by Nitrous Oxide

[9911MU51C]

I'm for procedural but then again I'm a casual player

[RobotsAndSpaceships]

I'm for procedural but then again I'm a casual player

How does that make a difference?

[cpast]

What is this terrible argument of "IRL, customers have to fit their payload inside certain fairing diameters"? Hello! We're a space program! We build & designs our own rockets, obviously we need the ability to design and build our own fairings to fit our payloads.

An argument that has been advanced, many times, for procedural fairings above procedural other parts is "in real life, it's not that hard to just build a bigger fairing, so there's no reason it should be that hard in KSP." Basically, people conceded that procedural engines don't really make tons of sense, and that in real life tanks are built to fixed sizes and can't be tweaked that easily, but maintained that fixed-size fairings present a constraint that isn't there in real life. You see it when you see people talking about why procedural fairings makes sense but aren't willing to say other procedural parts make sense. Earlier in this thread, I've been pointing out that that's blatantly untrue, and procedural fairings is a major departure from real-world rocket constraints. That's not to say there aren't arguments that it's more fun to have procedural fairings, but...well, it's been a discussion topic lately where the line should be drawn between fun and realism, no?

Edited February 16, 2015 by cpast

[Superfluous J]

I gotta say having used the SETI Balance mod for just a little while, I am already having a hard time just LOOKING at the vast array of parts in the VAB in my non-procedural install.

I think that once I finish my current YouTube series, I'll never use stock fuel tanks (or SRBs, or nosecones, or...) again.

[RobotsAndSpaceships]

I hate to be blunt, but you could just install the procedural part mods and let people who prefer the stock game do their thing. Procedural parts cut down on the creativity IMO.

IMHO, it actually increases the challange. With stock lego tanks, you either have a few hundred m/s of delta v over what you need, or a few under, rarely the exact amount number. Procedural parts let you get the exact amount of delta v you need in order to complete your mission. This means you have to carefully plan your launch windows, burn at the exact heading and time, and be sure to not screw anything up. Once I overshot a burn to the Moon in RSS with Procedural tanks meaning that I had just enough delta v to get there and circularize, and I had to start dumping hydrazine that I wouldn't need in order to increase my delta v, meaning that I had even more limited maneuvering capability. Procedural parts don't limit creativity, they increase it. Now if only they could get some decent textures.

Procedural wings are a whole other topic. If I want to try to rebuild the SR-17 in KSP, a plane which has a very unique shape and is difficult to make, I can try to make one with twenty stock wing pieces, that only looks OK, or one with Bac9s pwings, meaning I can change the color, the size the root, whatever, to be anything I like, along with using less parts. If I had a choice, which one do you think I'd pick? The clipped, lego'd one, or the sleek, customizable one?

- - - Updated - - -

-snip

Funny, the same people who have been crusading against realism are now all for it in the form of fairings.

[cpast]

Incidentally, my view on this has evolved a bit reading responses. I now agree that some sort of procedural thing is ultimately the only workable option, especially as compacting payload is a lot harder in KSP when you don't have the proper tools. However, I still think it makes sense to have a bit of "trying to make it fit," which one good way to do that is to have procedural fairings at fixed size increments - say, 0.625 m increments of diameter with increasing cost as diameter increases, so you get an outsized advantage making your payload fit in the next-smaller size without actually needing multiple parts, but you don't have to worry about that unless you feel like it, and you never pay too big a penalty for having a slightly-too-large ship.

- - - Updated - - -

Funny, the same people who have been crusading against realism are now all for it in the form of fairings.

Sort of the opposite, actually - it seems like the more vocal people are on the side of proc fairings. However, I really don't think "crusading" is a useful word in describing any argument, really ever. It's fairly pejorative, and doesn't contribute to constructive debate.

[cantab]

With 20:20 hindsight, a quality "procedural" system might have been better for KSP than the lego block style system it's used. But the lego blocks are what it's gone with, and I think a major change in how you build rockets is too much for 1.0. Even perhaps something better for a KSP 2, that might then include more sophisticated considerations than having parts be perfect rigid bodies. The issue then is whether we want a mishmash with some things "procedural" and others pure lego blocks or whether it's better for a unified approach.

As far as gameplay goes, personally I find it fun designing my payload to fit inside the fairing I have. That's why I ditched Procedural Fairings - they're very powerful and can do things I've not found possible with other fairing mods, but they were just too automagic for me. In career it restricts the base size, but last I checked still lets the fairing mushroom out as wide as you like. By the same token, I find it fun designing a vehicle that not only does its mission but does it within the mass, size, and part count limits in career. Because of those limits I read up on the real aerodynamics of flying wings then built one in KSP.

But really, my only strong view is that plenty of approaches to fairings work well. Contrary to what Maxmaps has said, fairings don't have to be "procedural" to be useful.

[RobotsAndSpaceships]

You need an algorithm to morph the fairings to the payload... Could be pretty complicated.

Its already been done.

[Brotoro]

Incidentally, my view on this has evolved a bit reading responses. I now agree that some sort of procedural thing is ultimately the only workable option, especially as compacting payload is a lot harder in KSP when you don't have the proper tools. However, I still think it makes sense to have a bit of "trying to make it fit," which one good way to do that is to have procedural fairings at fixed size increments - say, 0.625 m increments of diameter...

Like other things in KSP, it would be nice if this choice was determined by the setting of the little icon you toggle with the C key (icon showing polygon...click-stop to standard diameter fairings; icon showing circle...drag to whatever diameter you want).

- - - Updated - - -

Also -- For these threads, it would be nice if, once the Devs decide on something, they could simply mention it in the thread so people could quit arguing back and forth about something that may have already been decided and locked down.

[Bill Phil]

Its already been done.

yeah but we're talking about Squad here...

I kid I kid.

- - - Updated - - -

Which is really sad to me. Procedural tanks are literally the single most parts-count-reducing improvement mod in the game right now. If I want a 5x35m tank for an orbiter, BAM, I have it. If I need a 0.75m radius circular tank for a lander, BAM, I have it. If I want a nosecone-shaped tank for the front of my spaceplane, BAM, I have it. Because of this, most of my launchers for RSS come in under 25 parts. I built a launcher that could take nearly 600 tons to LEO and the entire thing was (IIRC) 52 parts of which 23 were engines, and including fairings. If I had tried to build the 30m wide lower stage with stock tanks I would surely have had a parts count upwards of 300, which is patently ridiculous.

Perhaps a compromise can be made.

Have procedural tanks be rather expensive (compared to the pre-set ones) and make them have a lower size limit, as well as moving them further in the tech tree and perhaps they could have a mass fraction less than the LEGO-tanks.

I would like huge launchers, but I don't 5m parts...

I see why it's a good idea, but it doesn't fit that well with with KSP's LEGO theme. Unless you made it so the payload had to not use the proc-tanks. My launchers take forever to perfect, and build, and are my largest part count builds, so it would make it easier to build launchers while keeping the LEGO style for the payloads ( not ginormous payloads...)

I just think it needs limits. Same with proc-fairings.

[Superfluous J]

Have procedural tanks be rather expensive (compared to the pre-set ones) and make them have a lower size limit, as well as moving them further in the tech tree and perhaps they could have a mass fraction less than the LEGO-tanks.

I hate ideas like this. Why should I pay more to work with a more logical setup? I'd be like putting PreciseNode in the game but charging $10 Kerbucks every time you use it simply to encourage us to use the worse way to do it.

And in case you don't know, procedural tanks are already limited by tech. You can only make tanks that are about as big as those you'd unlock with each tech tree node.

[Bill Phil]

Why penalize you? It's not the core of the game. It's not the SPIRIT behind the game. It's like cheating.

EDIT:

Anything specially built is generally more expensive than the mass produced ones.

Edited February 16, 2015 by Bill Phil

[regex]

Anything specially built is generally more expensive than the mass produced ones.

Specially built? Are you aware that every carrier rocket on the planet has been specially built for a specific purpose? Engineers at NASA, ESA, or POCKOCMOC have never pulled pre-built tanks off the shelves, stacked them on top of one-another, slapped an engine on the bottom, and launched to space. Each rocket had a payload in mind (or a mass, width, and length), was built to specific specifications, a fairing (or three) was designed that fit the upper stage, and an engine was probably designed for that rocket unless one already fit the bill (actually a fairly rare occurrence IRL, judging from the sheer number of rocket engines that exist). Each stage would have a single large tank for each propellant needed for the stage engine encased in a structural shell.

KSP employs a useful abstraction in stacking parts but there is nothing wrong with using a single tank part to create your rocket stage. In fact, it's more realistic and much, much better for performance. Certainly in career there should be certain limits to width and length of tanks in order to provide a progression, and Procedural Tanks does this, but why constrain the sandbox players? Not to mention that part counts in a game with memory leaks and bad GUI design on an overloaded parts selection windows is actually a really big deal.

Seriously, if you're still not convinced, check out ATK's SRM Catalog. There is a bewildering array of solid rocket motors available for all kinds of applications from a single company. Procedural parts make game sense and they make IRL sense.

[Superfluous J]

Why penalize you? It's not the core of the game. It's not the SPIRIT behind the game. It's like cheating.

EDIT:

Anything specially built is generally more expensive than the mass produced ones.

Ah. I see. You think I want Squad to put this in the game in spite of it being against the core spirit of the game.

This is not what I want. I want procedural-everything-that-can-be-procedural to BE the core spirit of the game.

Regarding specially produced vs mass produced, the MILLISECOND that a second launch of the same rocket is cheaper than the first, I'll rethink my stance that every single rocket is specially built.

[Daze]

Which is really sad to me. Procedural tanks are literally the single most parts-count-reducing improvement mod in the game right now. If I want a 5x35m tank for an orbiter, BAM, I have it. If I need a 0.75m radius circular tank for a lander, BAM, I have it. If I want a nosecone-shaped tank for the front of my spaceplane, BAM, I have it. Because of this, most of my launchers for RSS come in under 25 parts. I built a launcher that could take nearly 600 tons to LEO and the entire thing was (IIRC) 52 parts of which 23 were engines, and including fairings. If I had tried to build the 30m wide lower stage with stock tanks I would surely have had a parts count upwards of 300, which is patently ridiculous.

Yeah i know it's pretty sad, i play too in RSS + RO, but as you can see people start complaining even if it's simple an idea...

I think that (like for people who "hate" FAR) is dictated by the fact that they never tried the mod that we are talking about.

[Jouni]

There's two kinds of creativity going on, I think. There's the type of creativity that involves adapting a finite series of parts to meet some goal, the type used so far in stock KSP. Then there's the type that ignores those sort of limitations for greater freedom of expression, the type to which procedural parts cater.

I've been advocating a third view. Using tools for their intended purposes isn't creative, no matter how flexible or limited the tools are. To be creative, you have to surprise yourself.

Which is really sad to me. Procedural tanks are literally the single most parts-count-reducing improvement mod in the game right now. If I want a 5x35m tank for an orbiter, BAM, I have it. If I need a 0.75m radius circular tank for a lander, BAM, I have it. If I want a nosecone-shaped tank for the front of my spaceplane, BAM, I have it. Because of this, most of my launchers for RSS come in under 25 parts.

That's why I don't want procedural tanks, at least if you can make them arbitrarily big. If a rocket has less than 50 parts, it feels like a small rocket, no matter whether it lifts 1 or 1000 arbitrary units of mass to an orbit that requires 3500 or 9500 arbitrary units of delta-v. To feel large, a rocket has to approach the limits of what's possible within the current design constraints. Because part count is the most significant design constraint in the game, and because I enjoy pushing the limits, launching large payloads should require rockets with high part counts.

Limited procedural parts could be ok, as long as they feel like Lego blocks, where large parts are just integer multiples of small parts.

[BudgetHedgehog]

If a rocket has less than 50 parts, it feels like a small rocket, no matter whether it lifts 1 or 1000 arbitrary units of mass to an orbit that requires 3500 or 9500 arbitrary units of delta-v. To feel large, a rocket has to approach the limits of what's possible within the current design constraints. Because part count is the most significant design constraint in the game, and because I enjoy pushing the limits, launching large payloads should require rockets with high part counts.

Not everyone has that good of a computer. I simply don't see the point in suffering through low FPS just it should feel big. Look through the suggestions forum and see how many posts suggest or advocate welding/joining parts or otherwise reduce part count. To not only willingly subject yourself to that, but actually prefer it, that's just so weird to me. I honestly don't see why that's preferable. Yeah increase immersion or whatever, but actually enjoying playing the game? To force yourself to sit through launches that take 5 times as long as they would normally take? I think you have a weird sense of enjoyment.

[Daze]

Not everyone has that good of a computer. I simply don't see the point in suffering through low FPS just it should feel big. Look through the suggestions forum and see how many posts suggest or advocate welding/joining parts or otherwise reduce part count. To not only willingly subject yourself to that, but actually prefer it, that's just so weird to me. I honestly don't see why that's preferable. Yeah increase immersion or whatever, but actually enjoying playing the game? To force yourself to sit through launches that take 5 times as long as they would normally take? I think you have a weird sense of enjoyment.

A very weird i add. You are basically trying to justify: "because i've a powerful PC i can launch a bigger rocket, if you have a low specs PC you can't" Wow.

[Scoundrel]

Specially built? Are you aware that every carrier rocket on the planet has been specially built for a specific purpose?

Regarding specially produced vs mass produced, the MILLISECOND that a second launch of the same rocket is cheaper than the first, I'll rethink my stance that every single rocket is specially built.

These statements are certainly true for launch vehicles pre-SpaceX, but the era of the bespoke rocket as the norm is slowly coming to a close.

From http://www.spacex.com/news/2013/09/24/production-spacex

"...production of the Dragon spacecraft has increased significantly... No other American company is mass producing spacecraft at the same rate... Each Falcon 9 is identical, regardless of the type of mission it will fly..."

The trend is that the cost per ship (dry) is going down thanks to a combination of economy of scale and improved component recovery methods, which in turn translates into reduced launch costs.

That said, I think the Skylon koolaid is tastier, but that's just me.

[Jouni]

Not everyone has that good of a computer. I simply don't see the point in suffering through low FPS just it should feel big. Look through the suggestions forum and see how many posts suggest or advocate welding/joining parts or otherwise reduce part count. To not only willingly subject yourself to that, but actually prefer it, that's just so weird to me. I honestly don't see why that's preferable. Yeah increase immersion or whatever, but actually enjoying playing the game? To force yourself to sit through launches that take 5 times as long as they would normally take? I think you have a weird sense of enjoyment.

I don't have a fast computer either. There's usually noticeable lag with ~150 parts at launch, and the game becomes annoyingly slow at ~250 parts. In orbit, I usually get the same performance with ~50 more parts.

I guess I've been playing KSP long enough that I usually can't find anything interesting to do beyond LKO. Designing, building, and launching ships has become more fun than flying the missions. While the solar system itself feels boring and repetitive, there's still much left to explore in what can be achieved with the 150-200 part ships that remain enjoyable to fly. Reducing part counts with bigger parts seems rather pointless, as I would just be building bigger ships with the same part counts with them.

[ratchet freak]

Why not have pre-set stackable fairing "cans" like one use cargo bays that you have to build everything in, then using the new cargo shielding setup decide the aerodynamics. (then stage/eject outside the atmo)

That way fairings aren't procedural but still you can still lengthen them, You are just limited to the diameter of what you put in it.

[cantab]

Specially built? Are you aware that every carrier rocket on the planet has been specially built for a specific purpose?

Not really. Sure, they're hardly "mass produced", but my understanding is most real rockets are built for a general "x kg to LEO, y kg to GTO" kind of thing. Payloads have to be built to fit the mass limits of the rocket, and yes to fit inside the fairing. Right now if I design a satellite that's 8 metres wide, no launcher can put that in orbit.