Realism in KSP

[regex]

I will reiterate on this topic of "gatekeepers" that they are in no way dictating the direction of the game. They are however providing a volunteer service of performing triage on the many, many issue reports that we receive on the tracker. Their personal opinion on the game does not in any way come into their duty, they provide an objective service and adhere to guidelines and documentation provided by myself. Every one of the Managers on the public tracker project is either an Experimental Team or a QA Team member and I know each and every one of them well.

If you honestly feel that there is a bias being enacted on the tracker triage, that you are being unfairly targeted due to your personal opinions on the development of KSP and that your issue reports are being supressed to ensure they don't get read, please don't hesitate to contact me and I will deal with the matter with complete and utter sincerity.

If you're directing this at me, there is already a (locked) thread where we've gone over this; I am satisfied that I am not being targeted (per se, rather, I now think that was a knee-jerk reaction from someone who vehemently opposes any sort of "realism" fixes in this game, as demonstrated many times on this forum).

My issue now is with the classification of what is an obvious inaccuracy in the math KSP uses as a "feature request". However, as noted, it may not actually be considered a bug and there are reasons for it to not be considered a "bug" (even if I think they're pretty weak), but at least the ticket is still open instead of being dismissively closed. I mean, really, if you guys are going to call it "isp", use the math correctly so people can rely on it. If not, call it "fnorp" or something because it's not "isp".

[Ted]

If you're directing this at me, there is already a (locked) thread where we've gone over this; I am satisfied that I am not being targeted (per se, rather, I now think that was a knee-jerk reaction from someone who vehemently opposes any sort of "realism" fixes in this game, as demonstrated many times on this forum).

My issue now is with the classification of what is an obvious inaccuracy in the math KSP uses as a "feature request". However, as noted, it may not actually be considered a bug and there are reasons for it to not be considered a "bug" (even if I think they're pretty weak), but at least the ticket is still open instead of being dismissively closed. I mean, really, if you guys are going to call it "isp", use the math correctly so people can rely on it. If not, call it "fnorp" or something because it's not "isp".

Fantastic. I was just not entirely sure if you felt that component of the issue was resolved and wanted to ensure that we were all on the same page with it.

You feel that the classification of issue types is "pretty weak"? In your opinion, what defines as bug versus a feature request or perhaps a feedback report?

On the topic of the nomenclature of in-game mechanics, I can appreciate how one would find it confusing when they delve deeper into the internal workings of the mechanic, but for the vast majority of players I would wager they prefer a mechanic to be named according to how it appears through gameplay. I'm of the school of thought that if you named mechanics for how they work and not how they appear to the player, many, many things in a lot of games would have to be named entirely differently.

[Franklin]

I'm of the school of thought that if you named mechanics for how they work and not how they appear to the player, many, many things in a lot of games would have to be named entirely differently.

Not at all trying to quibble, I know you're just here to keep the peace, but if this were a fantasy MMO sure, but not in something toted as a teaching tool. :/

edit: I mean maybe the average, casual player is a larger market, and so will be of more value to a game producer that's fair, but please remember who the game was originally targeted to: space, math and physics dorks. I mean early concepts of KSP were floated around the Orbiter forum, correct?

Clearly they love the game, enough to make it semi-iconic, so why dismiss inaccuracies as something the broad-market wouldn't care about, so as to not be worth addressing. I think that's where the frustration is coming from.

Edited September 8, 2014 by Franklin

[regex]

You feel that the classification of issue types is "pretty weak"? In your opinion, what defines as bug versus a feature request or perhaps a feedback report?

KSP calculates (rocket engine) thrust with a total disregard to the isp of the engine, per the formula found here. This is proven math and KSP leaves out an important part of it. The fact that g0 being incorrect within KSP's ModuleEngines and ModuleEnginesFX is accepted as a bug while the missing part of the equation is considered a "feature request" is quite puzzling to me, which is why I question it.

On the topic of the nomenclature of in-game mechanics, I can appreciate how one would find it confusing when they delve deeper into the internal workings of the mechanic, but for the vast majority of players I would wager they prefer a mechanic to be named according to how it appears through gameplay. I'm of the school of thought that if you named mechanics for how they work and not how they appear to the player, many, many things in a lot of games would have to be named entirely differently.

Hey, fair enough, but the point really is that "isp" within KSP really isn't "isp", it's something else.

[Sky_walker]

On the topic of the nomenclature of in-game mechanics, I can appreciate how one would find it confusing when they delve deeper into the internal workings of the mechanic, but for the vast majority of players I would wager they prefer a mechanic to be named according to how it appears through gameplay. I'm of the school of thought that if you named mechanics for how they work and not how they appear to the player, many, many things in a lot of games would have to be named entirely differently.

I don't know why you seem to have a problem with that Ted. It's not like regex is asking for anything drastically increasing learning curve, or anything like that - simply: instead of teaching a wrong thing you'll start teaching a right thing while investing relatively little effort into it. It's basically a win-win situation as far as I see.

To be honest I just thought it's an oversight or a temporary placeholder, as plenty are made in the games with early access, but seeing that you actually try to defend it as something that should stay in KSP is.... well... disturbing.

Edited September 8, 2014 by Sky_walker

[ferram4]

Even more so, it'll make the engine module less computationally intensive.

Currently, it solves T / (g0 * Isp) = fuel_mass_flow. With this change, it'll solve g0 * Isp * fuel_mass_flow = T. Since multiplications are faster than divisions, this will be an overall performance boost.

[sal_vager]

Franklin, the game was originally 2D, and apart from patched conics it had very little anyone could call "accurate", it was shown on the Orbiter forums because Felipe played Orbiter and was active there but KSP was designed to be fun, not Orbiter Lite.

Showing KSP on the Orbiter forums does not make KSP Orbiter.

Regex, the way KSP handles Isp is intentional behaviour and the developers have chosen to use that term, KSP gives people a taste of what orbital mechanics are like to spark imagination and curiosity, not to be a dry and 100% accurate simulation.

Strict accuracy will send KSP down a path the developers, specifically Felipe, do not intend to take with the game, there are other games that offer accuracy and I know you are tired of that argument, others also get tired of the repeated calls for KSP to be made into something it's not, it's hard to appeal to everyone but Squad are trying, so that means having to take liberties with realism for gameplay.

Another response I'm sure you don't want to hear is that KSP is moddable, and players are free to write plugins to make engines work according to their specifications.

Not everything can or will be realistic, nor will it be stock, but anyone can add to the base that Squad has made and change it if they put some effort in to learn how.

As for a teaching tool, there is KSPedu, and KSP already does help people learn, if not more than the basics of orbital flight then it can inspire people to learn to code, to model and to texture.

[Ippo]

Wait a second, fixing a glaring mistake in an equation is now sending "KSP down a path the developers, specifically Felipe, do not intend to take with the game"!? Guys this is not a request for simulator-like features. It's not about Nbody physics or radiation belts or stuff like that.

You call it Isp, but it is not. That's a fact. You don't want to fix it? Not cool, but at least don't try to defend this choice.

You can't get an equation wrong on purpose and then tell US we are wrong for wanting it fixed.

[ferram4]

Strict accuracy will send KSP down a path the developers, specifically Felipe, do not intend to take with the game, there are other games that offer accuracy and I know you are tired of that argument, others also get tired of the repeated calls for KSP to be made into something it's not, it's hard to appeal to everyone but Squad are trying, so that means having to take liberties with realism for gameplay.

What about when taking such liberties not only result in something that is inaccurate, but is also more computationally expensive than doing things correctly? You and I both know the game suffers from crippling performance issues, why not make a change that will lessen the computational load and provide a slight change in reality as well?

Honest question Sal, what is the benefit to leaving things as they are? It's less accurate and it's more expensive to compute. The only change is that suddenly, we need to be given SL thrust as well as Vac thrust, but considering that the stock game doesn't provide TWR or dV readouts anyway, that lack of information shouldn't be anything new to players, and providing it would, (if I remember the quote correctly) "remove the mystery and fun" of trial and error gameplay. So why not make the change?

[regex]

Another response I'm sure you don't want to hear is that KSP is moddable, and players are free to write plugins to make engines work according to their specifications.

The fact that people have to mod the game to fix basic math is just ... bad.

[Ted]

Not at all trying to quibble, I know you're just here to keep the peace, but if this were a fantasy MMO sure, but not in something toted as a teaching tool. :/

edit: I mean maybe the average, casual player is a larger market, and so will be of more value to a game producer that's fair, but please remember who the game was originally targeted to: space, math and physics dorks. I mean early concepts of KSP were floated around the Orbiter forum, correct?

Clearly they love the game, enough to make it semi-iconic, so why dismiss inaccuracies as something the broad-market wouldn't care about, so as to not be worth addressing. I think that's where the frustration is coming from.

Very good points. I should emphasise that I'm not dismissing anything here, and my apologies if it came off that way.

KSP calculates (rocket engine) thrust with a total disregard to the isp of the engine, per the formula found here. This is proven math and KSP leaves out an important part of it. The fact that g0 being incorrect within KSP's ModuleEngines and ModuleEnginesFX is accepted as a bug while the missing part of the equation is considered a "feature request" is quite puzzling to me, which is why I question it.

Ah, I can see the root of your confusion and that is a very fair point. I didn't realise you had created two reports, similar in nature and they had been treated differently. I've applied the same action to your g0 issue report as I have confirmed that the value is the intended value.

[Technical Ben]

I don't know why you seem to have a problem with that Ted. It's not like regex is asking for anything drastically increasing learning curve, or anything like that - simply: instead of teaching a wrong thing you'll start teaching a right thing while investing relatively little effort into it. It's basically a win-win situation as far as I see.

To be honest I just thought it's an oversight or a temporary placeholder, as plenty are made in the games with early access, but seeing that you actually try to defend it as something that should stay in KSP is.... well... disturbing.

It is how it always goes. Leave something for long enough, and it turns from a choice/incidental quick fix, to a requirement. With that all the social and economic pressures to not change it. The best option is to start on the correct footing, the harder option is to change something later.

Take atomic electron orbitals described to be the same as planetary orbitals, as an example. The teachings examples are effectively wrong to begin with, but as there are lots of people invested in the model and illustration, it keeps traction and stays around.

The Wrong: http://www.chemguide.co.uk/atoms/properties/orbits.gif

The Right: http://en.wikipedia.org/wiki/Atomic_orbital

Edited September 8, 2014 by Technical Ben

[regex]

Ah, I can see the root of your confusion and that is a very fair point. I didn't realise you had created two reports, similar in nature and they had been treated differently. I've applied the same action to your g0 issue report as I have confirmed that the value is the intended value.

Well played sir.

[Ted]

I don't know why you seem to have a problem with that Ted. It's not like regex is asking for anything drastically increasing learning curve, or anything like that - simply: instead of teaching a wrong thing you'll start teaching a right thing while investing relatively little effort into it. It's basically a win-win situation as far as I see.

To be honest I just thought it's an oversight or a temporary placeholder, as plenty are made in the games with early access, but seeing that you actually try to defend it as something that should stay in KSP is.... well... disturbing.

It's not the case that I have an issue with, and in all honesty I wouldn't say I have an issue with anything in this thread other than posters being concerned with management of the bug tracker.

I'm just letting my general opinion on matters such as this known, and it should be noted that this is my personal opinion, not one that bleeds over to KSP's development. I didn't intend for it to come across as my professional opinion about the development of KSP or how certain features should be implemented, apologies if it did.

Edited September 8, 2014 by Ted

[Red Iron Crown]

Isp in KSP is most definitely specific impulse as defined in real rocket science, all equations that use it save one work correctly. It's just the handling of its variation that is a bit unrealistic. Currently, the fuel pumps work at maximum flow when at 1 atm of pressure and reduce their flow as pressure decreases. The more realistic representation would be for the pumps to always give full flow and have thrust vary.

This would be such a minor change in game play terms that I'm surprised it's meeting with such resistance. I guess I don't see the gameplay benefits of doing Isp variation the way it's done now.

[sal_vager]

Ferram, If it was my choice I'd change it, I already have ideas of what engines should be good for sea level, high altitudes and vacuum, but that's not the way KSP is.

The benefit, as far as I can work out, is simplicity, you know a Mainsail will give you 1500 thrust everywhere, it's not going to let you down, it'll just drink more at sea level compared to space.

You guys are seasoned KSP vets, you know how to calculate ejection angles and Isp, new players don't and most don't want to have to, just having it in-game can confuse people and they'll think "this is too hard for me, I'll pass"

The math isn't wrong, KSP just isn't using the math you expect, suggest it be changed by all means, make your case to the developers, but remember that politeness moves mountains, demands rarely get you what you want.

[Ippo]

"The math isn't wrong, KSP just isn't using the math you expect"

I'll try this justification during my next physics exam and let you know how it works out for me.

[Red Iron Crown]

You guys are seasoned KSP vets, you know how to calculate ejection angles and Isp, new players don't and most don't want to have to, just having it in-game can confuse people and they'll think "this is too hard for me, I'll pass"

That's a part of the problem, IMO. All the rocket science calculation I know how to do I learned from or because of KSP. I learned how rocket engines behave in atmospheres wrongly because of how KSP calculates it.

Given that the game doesn't calculate TWR for you and seems unlikely to do so, variable thrust isn't a problem for new players. Does it lift off? Good, you have enough thrust.

[Rowsdower]

"The math isn't wrong, KSP just isn't using the math you expect"

I'll try this justification during my next physics exam and let you know how it works out for me.

And with that, the rails are off. Thread closed.