How Many People Code/Develop on the KSP Team?

[Javster]

Just a quick question. How many people are there at Squad who actually work on the game, and not the other things related to it?

I don't mean server guys or graphic designers or sound persons or community managers, I mean people who code directly for the game. Programmers, basically.

Edited July 2, 2014 by Rowsdower
Edited title to make it more topical

[Linear]

2/3. Programming is a small chunk of the game, there really isn't much that needs to be done. It's about everything coming together.

If you add more programmers, like adding boosters, the result isn't of a massive gain and you're going to blow something up.

[WeirdCulture]

If you add more programmers, like adding boosters, the result isn't of a massive gain and you're going to blow something up.

hihi, thats right

[xcorps]

2/3. Programming is a small chunk of the game, there really isn't much that needs to be done. It's about everything coming together.

If you add more programmers, like adding boosters, the result isn't of a massive gain and you're going to blow something up.

If that is in fact the case, then that is exactly what they should be doing.

[Caelib]

Adding more developers actually has the opposite effect - it slows things down quite a bit in the short term.

[Linear]

If that is in fact the case, then that is exactly what they should be doing.

...Strange, KSP has this effect.

You are correct.

Hire ALL the developers.

[Robotengineer]

Moar devs?

[WinkAllKerb'']

welcome to the mess yiiiiiiiiii ahhhhhhh

[TNM]

I have to disagree strongly here. Yes, if you have 200 on a project and add another 100 you have diminishing returns asls you can't indefinitely split up a project. For KSP we are talking 2-3 coders. If you add another 2 you could do double. No, not double fast for the same code, but simply "more" code, meaning you can prototype ALOT more stuff to get an idea what you then will code for the next update. Think of it like the guy prototyping multiplayer right now. He is not "slowing down KSP developement"...he is testing out what/how multiplayer can work before the rest of the team commits to it.

What i agree on is that a small team like KSP has to grow organically, not getting 10 new programmers at a time but adding them 1 by 1 to fit it. Remember, KSP just lost a few guys, i hope they get replaced fast.

[KerbMav]

Somewhere we could read that adding new coders would be to difficult this late in the development process, as it would take to much time to fill them in on everything, so there would be no additions to the team in this area.

[Guest]

At this point, Squad doesn't really needs coders, perhaps except for an aerodynamics expert hired specifically to rewrite the atmosphere model. The current one sucks and from what I've seen, Squad devs are no physicists, and aerodynamics is a complicated field. An addition to "program" dev team I'd like to see would be a dedicated physics/mathematics expert able to overhaul certain key aspects of KSP physics. For example, N-body simulation, always decried by Squad as "impossible", not only could be implemented in a playable way, but such a system is being developed as a plugin (and already works for most part, to boot, though it has trouble interacting with the rest of the game). Drag, buoyancy, heat... all the stuff that currently acts ridiculous should be fixed before 1.0. It'd be great if they had an actual physicist to sort out those problems. If KSP is meant to be an educational game, it should not teach things that are blatantly wrong (if I wasn't already a student of physics, I'd have some pretty bad misconceptions about what flying a rocket actually entails...).

[xcorps]

if I wasn't already a student of physics, I'd have some pretty bad misconceptions about what flying a rocket actually entails.

I know what you mean. I was scared and confused the first time I got elected for mayor. It was NOTHING like SimCity. I mean nothing bro. And I must have played that damn game for like 4 years.

[Red Iron Crown]

I know what you mean. I was scared and confused the first time I got elected for mayor. It was NOTHING like SimCity. I mean nothing bro. And I must have played that damn game for like 4 years.

"You must spread some Reputation around before giving it to xcorps again." XD

The concept some have been nosing around in this thread is Brooks' Law. The idea being that adding additional developers can actually slow development, as the original developers spend more time bringing the new ones up to speed and more time managing them. Not sure if it applies to Squad and KSP, I don't know enough about the internal workings there to really guess.

As for n-body physics, I say no, thank you. I like that orbits are easily predictable and stable and I don't really care about LaGrange points (the only really useful thing n-body would add). Patched conics are a realistic enough approximation for a game and permit physicsless high time compression.

[JimmyAgent007]

"You must spread some Reputation around before giving it to xcorps again." XD

I covered you.

And I agree with everything else you said. Im a programmer and explaining your code to even other programmers can be a pain in the ass. Also N-Body would limit what computers could run the game smoothly, for very little payoff. Would be easier to fake it with SOI just planted in place, but the devs stated they wont do that.

[Lilleman]

@Dragon01: KSP is a videogame, not a simulator. It must be fun to play. I'm sure a lot of people will be happy if a mod including real orbital mechanics is released, but that's not what I want from KSP, I like the spirit of those physics, and I'm pretty sure I'm not the only one:

It's simple, but not too much. A nice balance for challenge. I don't want to go back to school and get another diploma before I can play a game...

Reentries? OK, that should be in stock: everyone's been surprised seeing a ship doing a reentry at 3000m/s without problem. But for the rest of you propositions, I have to disagree, sorry.

[pxi]

Agree that in principle, increasing the size of the dev team may be a good idea. However there are many subtle issues with this seemingly simple solution, as already highlighted in other posts above.

Furthermore, considering the small size of the dev team, I would predict that we would need to be prepared for a slowdown in the pace of development while the new guys are brought up to speed. It would however be an extremely opportune time for code-review / massive bugfixing.

[Alshain]

It's not just getting "up to speed" on the code either. 2 to 3 programmers can discuss things among themselves. When you start getting teams of programmers you have to start hiring managers to divide task load and handle the differences of opinion, so now you've added more overhead cost. Adding more developers could be good if the project is large enough but can actually add cost and delay releases on small projects.

Plus, even if you got more coders and were churning out code faster, well now you need more testers. It's a balancing game of bottlenecks. Depending on the size of the project there is a point of diminishing returns and on one like KSP, they could probably add 1 maybe 2 more programmers if they needed to, but too many and it's more of a hindrance than a help.

(I am a developer by trade, not for video games, but yeah I know what I'm talking about)

Edited July 2, 2014 by Alshain

[Linear]

I have to disagree strongly here. Yes, if you have 200 on a project and add another 100 you have diminishing returns asls you can't indefinitely split up a project. For KSP we are talking 2-3 coders. If you add another 2 you could do double. No, not double fast for the same code, but simply "more" code, meaning you can prototype ALOT more stuff to get an idea what you then will code for the next update. Think of it like the guy prototyping multiplayer right now. He is not "slowing down KSP developement"...he is testing out what/how multiplayer can work before the rest of the team commits to it.

What i agree on is that a small team like KSP has to grow organically, not getting 10 new programmers at a time but adding them 1 by 1 to fit it. Remember, KSP just lost a few guys, i hope they get replaced fast.

Have you ever been in a development firm, or co-developed, because you'd realise you were wrong.

KSP does not need more developers. Most of the work isn't in the programming.

[TNM]

Oh the internet...so easy to just say "you are wrong". You didn't adress 1 point i made, not 1.

Just think about this to get your braincells going: on this planets larger companies then SQUAD exist that do more with more devs. Now if we could just figure out how that magic works...

[Red Iron Crown]

Just think about this to get your braincells going: on this planets larger companies then SQUAD exist that do more with more devs. Now if we could just figure out how that magic works...

I think the critical part of Brooks' Law is not the size of the coding team but instead how far along the project is when more coders are added. I.e. it's about how much time must be spent getting up to speed. As we don't really know how close we are to being feature complete, it's hard to say whether it would be worthwhile to bring more coders in or not.

[Creature]

We COULD clone the entire dev team. They could work in 8 hour shifts 24/7. One clone works and the others play around in ball pit until it's their shift again (this is what I imagine devs do in their free time).

[Karriz]

@Dragon01: KSP is a videogame, not a simulator. It must be fun to play. I'm sure a lot of people will be happy if a mod including real orbital mechanics is released, but that's not what I want from KSP, I like the spirit of those physics, and I'm pretty sure I'm not the only one:

It's simple, but not too much. A nice balance for challenge. I don't want to go back to school and get another diploma before I can play a game...

Reentries? OK, that should be in stock: everyone's been surprised seeing a ship doing a reentry at 3000m/s without problem. But for the rest of you propositions, I have to disagree, sorry.

Well, having better physics wouldn't really make the game any less fun.

At the moment, player doesn't have to worry about any calculations, everything is presented visually in the map.

It's true that mods like FAR show lots of numbers which I understand nothing about. In the stock game, we'd just need the aerodynamics, not the info displays.

[Pecan]

This thread is lots of fun for those of us in software development ;-0

When developing new software, rather than making another clone of someone else's idea, you are spending 60%+ of your time working out what you really want to do, ~20% how you can do it and very little actually doing it - a bit like *** (relations with the opposite gender).

The important thing about this is that early design decisions can turn out to be mistakes much later on and you only really know what you should have done once version 1 is finished. If ever.

[Linear]

Oh the internet...so easy to just say "you are wrong". You didn't adress 1 point i made, not 1.

Just think about this to get your braincells going: on this planets larger companies then SQUAD exist that do more with more devs. Now if we could just figure out how that magic works...

You'd be correct if these guys were making call of duty, sitting on their ass all day. The entirity of KSP is new ground, everything is new. They're spending such a small amount of time programming it's unreal - you must realise that.

You answered your own question, SQUAD isn't a large company compared to others, and it is NOT a games studio that requires many programmers.

At this stage in the SDLC there is no point adding more developers, you want actual designers and people who structure the game.

[Curveball Anders]

Just a quick question. How many people are there at Squad who actually work on the game, and not the other things related to it?

I don't mean server guys or graphic designers or sound persons or community managers, I mean people who code directly for the game. Programmers, basically.

Graphic designers, 3D modellers, audio designers, content/gameplay designers make up the majority of devs in any studio or development office.

The heroic coding cowboys are almost always a minority (but most studios are yelling for more, and better educated, programmers).

This is of course even more a fact if you base your game on an existing engine or framework.