I wonder why Squad doesn't want to give us a dV and TWR readout

[severedsolo]

On 31/03/2016 at 5:30 PM, regex said:

Is there a KSPedia entry about delta-V in 1.1?  I haven't checked into that, tried some interplanetary transfers on night one and updated some code last night...

[regex]

Needs moar Tsiolkovsky...

[John JACK]

2 hours ago, Nich said:

Ok if calculating DV is so easy give me the pseudo code that can calculate the DV of a craft with x stages y engines and z tanks.   Don't forget to account for thurst limiter and any engine can be used for any stage or stages

Thrust does not matter for dV. Only ISP and fuel mass.

Start from the bottom of staging. Look what engines are started and what tanks they are under. ISP - check, mass - check, math - done. Decouple now-empty tanks with engines, fire next stage, repeat.
Right, in R-7 or Atlas or Shuttle style rockets there are different engines working simultaneously for different time with different ISP and mass flow, and that greatly complicates things. But default KSP staging is serial, not parallel. First fire solid boosters, decouple them, then fire central liquid stage, decouple, fire upper stage. Simplest dV calculation will work for simplest rockets and greatly help new players. Next steps may vary, from showing dV of only lower stage, approximating by adding up dV of parallel blocks, to counting burn times of every engine or actually accepting KER code or something. But just basic dV readout is that 20% of work that gives 80% of results to 95% of players who actually need it.

[sarbian]

2 hours ago, Foxster said:

As I said before, this has already been done successfully at least 3 times in different open-source mods, so no need to start this from scratch again. At the very least, these could be reverse engineered for the calculations, maybe even lifted in their entirety if the authors can be contacted and would agree. 

How many page of moders telling you it is a lot more complex that you think do you need ? We have 2 (totally different) implementation, one of with 2 modders working on it and we still find plenty of bugs and edge cases. Squad would not have an easier time implementing it and they CAN NOT use our code because the GPL would force them to publish their game code. 

And the stock implementation would not work with mods, as ours do, so we would still have to maintains ours mods. I prefer Squad to work on actually new features, or integrating mods that actually bring something new (like they did with Arsonide's and RoverDude's) than doing what 2 mods already do. 

3 minutes ago, John JACK said:

Thrust does not matter for dV. Only ISP and fuel mass.

Start from the bottom of staging. Look what engines are started and what tanks they are under. ISP - check, mass - check, math - done. Decouple now-empty tanks with engines, fire next stage, repeat.
Right, in R-7 or Atlas or Shuttle style rockets there are different engines working simultaneously for different time with different ISP and mass flow, and that greatly complicates things. But default KSP staging is serial, not parallel. First fire solid boosters, decouple them, then fire central liquid stage, decouple, fire upper stage. Simplest dV calculation will work for simplest rockets and greatly help new players. Next steps may vary, from showing dV of only lower stage, approximating by adding up dV of parallel blocks, to counting burn times of every engine or actually accepting KER code or something. But just basic dV readout is that 20% of work that gives 80% of results to 95% of players who actually need it.

You forgot the really simple step of finding out the ressource available for each engine. I am waiting for your peudo code for that.

[Rocket In My Pocket]

A space simulation game with no Delta V/TWR readout makes about as much sense as:

• A racing game with no speedometer/rpm
• A football game with no score display
• An RPG where you don't know what level you are
• An FPS with no ammo counter
• An RTS with no display for collected resources or population cap
• A sims game with no "needs and wants" meters
• A fighting game with no health bars

I could go on...and on...and on...

[Padishar]

3 minutes ago, Rocket In My Pocket said:

I could go on...and on...and on...

Yes, you could, and it would still be just as pointless.  This thread isn't about whether or not it is needed, that has been discussed time and again in this forum.  Squad have said that they plan to add some kind of deltaV readout, we just have to wait...

[John JACK]

2 minutes ago, sarbian said:

You forgot the really simple step of finding out the ressource available for each engine.

I remembered to ignore fuel pipes. Plumbing is researched much longer into a tech tree, when player already should understand more than basics (i.e. can find a mod).

Simple step of finding out resource available for each engine is simple. Each engine can use fuel only from tanks it's mounted on. Just like that is found in-flight. That does not account for radial-attached engines, but they again appear later. Shortly, finding resources is not a problem — code for handling fuel flow is already there.

For next steps dV calculator may actually simulate burning and staging. I suppose that's what mods do.
- check takeoff mass, take weighted average ISP of all lower stage engines;
- remove fuel according to tweaked consumption of engines until some tanks are empty, check mass, count dV;
- stage decouplers if any, check mass, retake average ISP of remaining or freshly started engines;
- continue removing fuel again.

Yes, that must be repeated every time when player change some parts in VAB. But that's just math and modern computers do math like it in no time.
Also realism. A game must be not harder than reality. In reality every Head Designer does not grab a slide rule for every pencil stroke on a draft, they have an army of interns with slide rules to do math for them. KSP is already harder than real life for you can't calculate orbits and maneuvers in advance with any kind of precision. Eyeballing that is part of the fun, but running out of dV is frustrating. And dV readout at least in VAB will help to not to run out of dV most of the way out there without grossly overengineering rockets.

[Rocket In My Pocket]

1 minute ago, Padishar said:

Yes, you could, and it would still be just as pointless.  This thread isn't about whether or not it is needed, that has been discussed time and again in this forum.  Squad have said that they plan to add some kind of deltaV readout, we just have to wait...

Doesn't that make your response pointless?

Do you have a source for this supposed claim on Squads part?

I was under the impression they were on the fence about it, something about scaring new players off with lots of numbers and taking away the fun of experimenting on your own. (Which I don't really agree with...but it's their game.)

[sarbian]

11 minutes ago, John JACK said:

Simple step of finding out resource available for each engine is simple. Each engine can use fuel only from tanks it's mounted on.

Each ressource can use one of the 7 flow mode. Each engine can override that flow mode. If you ignore surface attachement you get player who complain that your display numbers are wrong.

But keep saying it is simple. Saying it is the only simple thing here.

[John JACK]

2 minutes ago, sarbian said:

If you ignore surface attachement you get player who complain that your display numbers are wrong.

Sure, player complaining that display numbers are wrong in some not-so-common situation is much worse that player refunding a game because there is no display numbers at all to make any sense for not rocket-geek. Actually to complain about wrong numbers one must know right numbers, and the point of stock dV readout is to help players who don't know any numbers. Just a "deltavee" word from KSPedia. And to complain one must go to forum, right by Add-on Section.

Isn't overriding of flow modes handled already by game code? The same code that can be used for simulation? There is no need to handle ALL situations, like fiddling with thrust limiting in flight, manual transfer or early staging. dV readout should display just maximum design performance or something close. Jet and monoprop engines can be just ignored, for there is only one Puff, and dV is useless for aircraft.

I'm not saying that calculating dV for any random craft is simple. Thought it's certainly possible. But I insist that in stock there must be any dV readout. Even simplest one (that may be not correct with more complex design) will introduce player to crucial foundation of rocket science and will give a base to trial-and-error from.

[sarbian]

40 minutes ago, John JACK said:

The same code that can be used for simulation?

Do you really think we would write 10k lines piece of code if all there was to do was calling the stock methods ? The stock call are made to handle simulation per physic frame (0.02s). To get any kind of simulation that don't chew your frame rate you need to have yours simulation do each stage in a couple of calls. That means finding out how fast your current ressources are being used by each of the active engines at their respective rates. No stock call does it and writing one that does is a large part of the hard work. 

Padishar, others and I keep explaining it ain't that easy. We wrote it so we do know how complex it is. Now maybe there is some magically simpler way to do it than the way KER and MJ does (and as they said earlier they don't do it the same way). After all we are mere humans, but I sincerely doubt it.

[regex]

2 hours ago, John JACK said:

But default KSP staging is serial, not parallel.

There is no "default" staging in KSP.

[John JACK]

Okay, writing perfect simulation to compute dV up to 0.01 m/s is not simple. Thought it may be possible to "simulate launch" on demand, second or two delay after pressing a button is not a problem. Framerate isn't important in VAB at all, and KER somehow does calculate dV without affecting it slightest.

But taking default staging with similar engines and calculating dV from mass with and without fuel IS. And it gives good enough guess whether going to space today or not. There is a reason not to give us perfect dV readout, but there is a reason to include simplest dV readout in stock as soon as possible.

2 minutes ago, regex said:

There is no "default" staging in KSP.

And what exactly staging sequence is if I didn't adjust it after adding all parts top-to-down or if I clicked a "reset" button?

[regex]

9 minutes ago, John JACK said:

But taking default staging with similar engines and calculating dV from mass with and without fuel IS. And it gives good enough guess whether going to space today or not. There is a reason not to give us perfect dV readout, but there is a reason to include simplest dV readout in stock as soon as possible.

That's unfortunately not good enough.  The beauty of, and the problem with, KSP is that you can build pretty much anything, which means that the delta-V calculator has to handle tons of edge cases.  You may think the most simple solution is totally acceptable and that edge cases should not be accounted for but the rest of the world will point at "Delta-V: N/A" as a software flaw.

Quote

And what exactly staging sequence is if I didn't adjust it after adding all parts top-to-down or if I clicked a "reset" button?

Ah, so you're talking about how the game automatically stacks engines?  Is that how everyone builds rockets, including the newbies?  That's the only solution we have to calculate for?  The first craft I build in a career mode is a pod with a parachute and a Flea, which means I have to adjust staging if I don't want to watch the parachute drag the craft back to the ground.  You learn to check and adjust staging pretty quickly, and when you go to build your "replica" of a favorite iconic rocket you're going to wonder pretty quickly why your simple edge case, moving an engine around, wasn't calculated for.

Edited April 3, 2016 by regex

[Foxster]

2 hours ago, sarbian said:

How many page of moders telling you it is a lot more complex that you think do you need ? We have 2 (totally different) implementation, one of with 2 modders working on it and we still find plenty of bugs and edge cases. Squad would not have an easier time implementing it and they CAN NOT use our code because the GPL would force them to publish their game code. 

And the stock implementation would not work with mods, as ours do, so we would still have to maintains ours mods. I prefer Squad to work on actually new features, or integrating mods that actually bring something new (like they did with Arsonide's and RoverDude's) than doing what 2 mods already do. 

You forgot the really simple step of finding out the ressource available for each engine. I am waiting for your peudo code for that.

I do bow to your far superior wisdom in these matters. 

However, I really don't think that working out a dV and TWR display is beyond Squad, especially with the maths already available from the open-source solutions available now.

[regex]

Just now, Foxster said:

However, I really don't think that working out a dV and TWR display is beyond Squad

No one has ever claimed it's beyond Squad, what we're claiming is that it's lots of hard work to implement for a "production" game.  It has to be right the first time.

[Foxster]

1 minute ago, regex said:

No one has ever claimed it's beyond Squad, what we're claiming is that it's lots of hard work to implement for a "production" game.  It has to be right the first time.

I think we can now leave that to Squad to decide as we are now guessing as to their abilities and priorities. 

[regex]

1 minute ago, Foxster said:

I think we can now leave that to Squad to decide as we are now guessing as to their abilities and priorities. 

  No, you're totally right, two people who have experience with writing this sort of routine and a Squad dev saying "Because there's definitely thousands of lines of C# involved, and many tears." equates to guessing at how much work is involved for Squad.

[Foxster]

1 minute ago, regex said:

No, you're totally right

I suspected that might be the case. 

[sarbian]

1 minute ago, Foxster said:

I suspected that might be the case. 

Indeed. I am blinded by your righteousness 

[Foxster]

10 minutes ago, sarbian said:

Indeed. I am blinded by your righteousness 

Please be kind, it is actually very challenging to be perfect (especially after a bottle of Wolf Blass Yellow Label).

Edited April 3, 2016 by Foxster

[John JACK]

27 minutes ago, regex said:

You may think the most simple solution is totally acceptable and that edge cases should not be accounted for but the rest of the world will point at "Delta-V: N/A" as a software flaw.

I think that KSP must have at least tutorial-level dV calculator. Maybe just more detailed page in KSPedia; with interactive Equation if possible. To tell new player how to guess dV manually without lurking forums or studying books.

And I understand reasoning for making it right at first try. Thought I do not agree, there was and still are things not perfect even after 1.0 release. Right now porting to Unity 5 is absolute priority, but the point of this thread is that stock dV readout was important for a long time. No dV at all is bad and not noob-friendly at all. Before making replicas new player (not rocket-geek with a slide rule) should at least be able to make some rocket and fly it somewhere and beck with only ingame knowledge. And even for experienced one addon readouts may be too overloaded.

Perfect dV readout is not simple. But dV readout is necessary and please make it a priority.

[regex]

6 minutes ago, John JACK said:

I think that KSP must have at least tutorial-level dV calculator. Maybe just more detailed page in KSPedia; with interactive Equation if possible. To tell new player how to guess dV manually without lurking forums or studying books.

Now this, I am all for.  As for the perfect delta-V calculator, I disagree.  Players will point to ENG and MJ and go "Why isn't Squad doing it right?"  It needs to at least be on the same level as those two.

[boborene]

On 4/1/2016 at 4:03 AM, StrandedonEarth said:

I do know that a while back Harvester himself said he didn't want dV readouts, because he liked the trial-and-error way of the kerbals. He appears to have since come around, but I can see his point.

That was valid in 0.9 with a kerbin so easy to fly and correct. try that today with almost all physics implemented is pure stupidity, not trial and error. We got error marge so narrow, no need to make game frustrating.

[EladDv]

I am sure there's a very small minority of people here that have ever written 1500 lines of code for a platform so interconnected as a game, not to talk about the endless possibilities of ships,staging events,fuels,symmetry parts and engine thrust limits. Those guys over here care about the game and it's state,they care about the community and they come here and listen to you and give thoughtful responses and try to explain to the masses the difficulty of programming such an endeavour. NathanKell and sarbian are here not because they are obligated to but because they care about that subject- as they said- they too play with dV readings on their personal games and would love to see it go stock but it's just not the right time to pour all the resources needed for it to work completely on the first time, And QA? oh don't get me started on that, MAYBE if they wanted to test something like that they would release a closed mod to get this thing tested but as a full release? i mean are you serious? just thinking about the amount of edge cases makes my head spin.

In conclusion, while it'd be a nice feature to have around and a massive addition to the toolset we have to work with i think we should stop demanding it to be done now and while not forgetting this- move on to new features,requests and ideas.

Thank you for your time, EladDv

Edited April 3, 2016 by EladDv