[1.5.0 <-> 1.8.1] Kerbalism v3.2

[JebIsDeadBaby]

12 minutes ago, Sir Mortimer said:

I very much like to hear the story behind this one

The story is called Oh Scrap! Meets Kerbal Construction Time, starring a leaking nitrogen tank that was supposed to fuel RCS. Before the rescue vessel got to the orbit, the tank was empty, so no RCS control. This in itself wouldn't be a catastrophy, but the launch was so well timed and the Ap so well guessed that the stranded pod passed my ship by less than 2,3 km just before circularization burn (ending about 20 km away after the burn), which enabled resources consumption. In a last ditch effort I tried to use the main engine to fine tune the orbit but then I lost the signal, and since CommNet does not throttle down to 0 in such a case, my ship fell down back to Kerbin. I needed about 4 days to build another rescue vessel, by that time the batteries on the stranded pod went dead and Kerbal soon died due to CO2 poisoning. I then terminated this ship but the mission remained active. I had to terminate the mission manually. 

IIRC if a rescued Kerbal dies during reentry, the mission fails automatically, so I was surprised it's not the case when it's Kerbalism death (not to mention ship termination that followed). 

[Retro Starship]

1 hour ago, JebIsDeadBaby said:

Dunno if it should be considered a bug and it's definitely a minor problem but kerbals dying does not auto-terminate rescue contracts. 

This happened to me but I do not think its a bug. I think your just supposed to rebuild the ship and try again with a new Kerbal.

Oh you said (if a rescued Kerbal dies during reentry) sorry, I got that wrong.

Edited November 14, 2019 by Saturn5tony

[Beetlecat]

5 hours ago, JebIsDeadBaby said:

The story is called Oh Scrap! Meets Kerbal Construction Time, starring a leaking nitrogen tank that was supposed to fuel RCS. Before the rescue vessel got to the orbit, the tank was empty, so no RCS control. This in itself wouldn't be a catastrophy, but the launch was so well timed and the Ap so well guessed that the stranded pod passed my ship by less than 2,3 km just before circularization burn (ending about 20 km away after the burn), which enabled resources consumption. In a last ditch effort I tried to use the main engine to fine tune the orbit but then I lost the signal, and since CommNet does not throttle down to 0 in such a case, my ship fell down back to Kerbin. I needed about 4 days to build another rescue vessel, by that time the batteries on the stranded pod went dead and Kerbal soon died due to CO2 poisoning. I then terminated this ship but the mission remained active. I had to terminate the mission manually. 

IIRC if a rescued Kerbal dies during reentry, the mission fails automatically, so I was surprised it's not the case when it's Kerbalism death (not to mention ship termination that followed). 

oh! interesting that a terminated ship doesn't "register" for the contract, but maybe that's an edge case, since terminating like that is sort of a "god-mode" action. Does Kerbalism not *poof* kerbals when their LS fails / or any other death-causing issue?

[lordcirth]

21 minutes ago, Beetlecat said:

oh! interesting that a terminated ship doesn't "register" for the contract, but maybe that's an edge case, since terminating like that is sort of a "god-mode" action. Does Kerbalism not *poof* kerbals when their LS fails / or any other death-causing issue?

It does kill the Kerbal, the problem is that the kill doesn't seem to fail the mission.

[Sir Mortimer]

5 hours ago, Beetlecat said:

Does Kerbalism not *poof* kerbals when their LS fails / or any other death-causing issue?

It does...

5 hours ago, lordcirth said:

It does kill the Kerbal, the problem is that the kill doesn't seem to fail the mission.

well...

11 hours ago, JebIsDeadBaby said:

IIRC if a rescued Kerbal dies during reentry, the mission fails automatically

The core problem at the base of this is that stock KSP doesn't expect Kerbals to just die for no good reason, and the only good reason KSP knows about is the destruction of the part a Kerbal happens to be in. In the stock game, they cannot die from anything other than a RUD event. All other cases are not expected and thus not handled by stock KSP, which is why those contracts don't react to the death of a Kerbal. I don't think there's much we can really do about this.

[Autolyzed Yeast Extract]

The new engine launch failures system makes recovering 2.5m or larger lifter rockets statistically impossible. Even if you make your engines high quality, you will probably have to reignite them around 5 times when trying to land the booster because of how powerful the lifters are even at minimum thrust (not necessarily simulated in this mod, but my point stands because that's realistic behavior). I am using Realism Lite so high quality is unfeasible even if it did make a difference (this is why the engines have minimum thrust values). 20 minutes of burn time means little if it can only ignite twice before failing.

On 11/15/2019 at 2:06 AM, Sir Mortimer said:

The core problem at the base of this is that stock KSP doesn't expect Kerbals to just die for no good reason, and the only good reason KSP knows about is the destruction of the part a Kerbal happens to be in. In the stock game, they cannot die from anything other than a RUD event.

this is false
g forces can kill them if set in difficulty settings

Edited November 16, 2019 by Autolyzed Yeast Extract

[Bersagliere81]

Is there a way to make TAC-LS food/water/oxy storage compatible?

[Gordon Dry]

@Bersagliere81the resources (Food, Waste, Water, WasteWater, Oxygen, CarbonDioxide) have exact the same names - so they're compatible from scratch.

Kerbalism just adds some resources like WasteAtmosphere.

 

You only have to make sure that patches are not strictly looking for TAC-LS.

If it's "only" a patch that adds above mentioned resources without some module-fiddling, you can just check for

NEEDS[Kerbalism|TacLifeSupport]

and that's it.

[The-Doctor]

The engine failure rate isn't that realistic, for example the RS-25 can ignite 25 times with little to no need for refurbishment, different engines in real life have different failure rates and different numbers of ignitions, meanwhile the md makes their be 2 ignitions for practically all the engines. I have made reusable boosters but using tweakscale and cluster engines

[Sir Mortimer]

6 hours ago, The-Doctor said:

The engine failure rate isn't that realistic, for example the RS-25 can ignite 25 times with little to no need for refurbishment, different engines in real life have different failure rates and different numbers of ignitions, meanwhile the md makes their be 2 ignitions for practically all the engines. I have made reusable boosters but using tweakscale and cluster engines

The engine reliability settings are done with just a few MM patches that try to determine the relight and burn duration numbers based on  some rough guesses. It's a very broad brush we use there, and it isn't perfect - but it works. If you (or anyone, really) is willing to provide engine-specific numbers that are somewhat realistic and work for the game, we'll happily use that

[mindstalker]

Has anyone had any luck making a probe that sinks in water in the current release. Until I unlock Ore tanks it seems everything floats even with shielding turned up to max.

[Xt007]

4 hours ago, mindstalker said:

Has anyone had any luck making a probe that sinks in water in the current release. Until I unlock Ore tanks it seems everything floats even with shielding turned up to max.

Would you happen to have FAR installed?

[HansAcker]

I can't get Discord signed in on this computer, so I'll ask here. It's just eye-candy, nothing serious. 3.1 is a great release and praise all round to SirMortimer and gotmachine for their work

In flight, the file manager display is tidy and shortened:

Spoiler

while in the Space Center scene, it is not:

Spoiler

I looked at the code and found that this seems to be the intended behaviour (IsFlight check here and there). Is the wrap-around a quirk of my particular screen resolution or could this be left over from a wider UI? I didn't really play with Kerbalism before 3.1, so I don't remember anything from before the rework.

There's another in-flight limit for overly long vessel names, too, but only for a select few panels (and even the in-flight limit is too wide on my screen, a horizontal scroll bar appears).

 

[plausse]

On 11/19/2019 at 9:14 AM, Sir Mortimer said:

The engine reliability settings are done with just a few MM patches that try to determine the relight and burn duration numbers based on  some rough guesses. It's a very broad brush we use there, and it isn't perfect - but it works. If you (or anyone, really) is willing to provide engine-specific numbers that are somewhat realistic and work for the game, we'll happily use that

And here I was really impressed by how good the guesswork was. That said.. Is there anything that influences chances of failure? Because I think on average a good majority  of engines breaks down for me well before the end of its rated service life (either burn time or ignitions). This would mean a lander doing a normal mun or minmus landing would either go through a few redundant engines or need several inflight repairs. I love the feature, it just seems a tad aggressively tuned. Can this be modified in config?

*Edit - engines that will need restarting would probably be flagged as high quality anyway, so changed the general "QualityScale = 4.0"                modifier to 8 on high quality parts. Also changed "the rated_operation_duration = 600" tp 1000. I will test if that gives a bit more juice in those high impact cases where one would happily pay more for a bit extra.

Edited November 20, 2019 by plausse
Looked at the config files

[mindstalker]

18 hours ago, Xt007 said:

Would you happen to have FAR installed?

Nope, in the 3.1 release they drastically reduced the weight that shielding causes. Everything floats now, or I might just be missing something.

 

[Sir Mortimer]

19 minutes ago, plausse said:

I love the feature, it just seems a tad aggressively tuned. Can this be modified in config?

There is a setting in preferences that tweaks the probability of an ignition failure; caption is "Engine Ignition Failure Chance" and it should be a slider. But that doesn't affect burn durations.

The rest can be tweaked in the configuration. Magic happens in KerbalismConfig/System/Reliability.cfg.

If you don't want to change that file, you could also drop this somewhere in your GameData. This doubles ignitions and the burn duration for all engines.

@PART[*]:HAS[@MODULE[Reliability]]:NEEDS[FeatureReliability]:AFTER[KerbalismDefault]
{
	@MODULE[Reliability]
	{
		%rated_operation_duration *= 2
		%rated_ignitions *= 2
	}
}

(Note this is utterly untested, but I think it should work)

3 minutes ago, mindstalker said:

Nope, in the 3.1 release they drastically reduced the weight that shielding causes. Everything floats now, or I might just be missing something.

Yup, that's probably it. I don't know how to get anything to submerge - never did any submarining in KSP.

Edited November 20, 2019 by Sir Mortimer

[panarchist]

I don't seem to be able to use the Magnetometer Boom in orbit - I also have SCANsat installed, and the boom appears to only want to find ore. I haven't used it since prior to the new science system in Kerbalism, but my understanding was that it should work for measuring magnetic fields in space. Am I doing something wrong? What should I be checking?

[HansAcker]

50 minutes ago, panarchist said:

I don't seem to be able to use the Magnetometer Boom in orbit - I also have SCANsat installed, and the boom appears to only want to find ore

Did you check the "Auto" tab in the Kerbalism window or the "Science" tab in the Magnetometer PAW? I don't have SCANsat installed and the Magnetometer Boom works for me:

 

[Gordon Dry]

That could possibly be related to this:
https://github.com/Kerbalism/Kerbalism/pull/540

Before that fix I had weird duplicates and/or stock experiments in Bluedog_DB science parts.

[panarchist]

2 hours ago, HansAcker said:

Did you check the "Auto" tab in the Kerbalism window or the "Science" tab in the Magnetometer PAW? I don't have SCANsat installed and the Magnetometer Boom works for me:

 

The Auto Tab, under "MODULE DEVICES", does not list the Magnetometer Boom per se, and there is no Science tab in the Magnetometer PAW. In the Auto Tab, I see "SCANresource Scanner on Magnetometer Boom", and it says "disabled". It's got to be a mod conflict, I'll take a better look.

[Gordon Dry]

I also recognized this, the Magnetometer boom got no science tab.

[panarchist]

5 hours ago, panarchist said:

The Auto Tab, under "MODULE DEVICES", does not list the Magnetometer Boom per se, and there is no Science tab in the Magnetometer PAW. In the Auto Tab, I see "SCANresource Scanner on Magnetometer Boom", and it says "disabled". It's got to be a mod conflict, I'll take a better look.

Verified this in an install with only Kerbalism, Kerbalism Config, and DMagic Science - all latest 1.8.1 versions. The MagBoom.cfg hasn't changed between versions, but GameData\KerbalismConfig\System\ScienceRework\ModSupport\DMagicOrbitalScience.cfg has changed quite a bit since Kerbalism 3.0.2. I'd do some more troubleshooting, but that's a bit more involved a MM patch than I typically work with. 

[KallangoVerde]

Is it possible to install the mod keeping the default settings of the experiments?

[Sir Mortimer]

2 hours ago, KallangoVerde said:

Is it possible to install the mod keeping the default settings of the experiments?

Yes, see https://github.com/Kerbalism/Kerbalism/wiki/FAQ#q-how-can-i-turn-off-kerbalisms-science-system-or-failures-or-radiation

6 hours ago, panarchist said:

Verified this in an install with only Kerbalism, Kerbalism Config, and DMagic Science - all latest 1.8.1 versions. The MagBoom.cfg hasn't changed between versions, but GameData\KerbalismConfig\System\ScienceRework\ModSupport\DMagicOrbitalScience.cfg has changed quite a bit since Kerbalism 3.0.2. I'd do some more troubleshooting, but that's a bit more involved a MM patch than I typically work with. 

Just to make sure, did you perhaps copy the contents of the updated KerbalismConfig zip file into an existing, older KerbalismConfig folder in your GameData? If so, you might have duplicate configurations because a few files have been renamed / moved. If that is the case, I suggest delting your Kerbalism and KerbalismConfig folders from your game data and then reinstall the latest release.

[AVaughan]

18 hours ago, plausse said:

Is there anything that influences chances of failure? Because I think on average a good majority  of engines breaks down for me well before the end of its rated service life (either burn time or ignitions).

Somewhere on the wiki they mentioned that Radiation can also damage an engine, and increase the chances of it failing before the end of it's rated life.  After a couple of Munar missions that crashed because the lander engine failed during the landing burn, I've come to the conclusion that engines that are mission critical and will be used beyond low kerbin orbit should be high quality.  (Somewhere on the wiki I think they also said that apart from ignition failures, engines are unlikely to fail before about 35(?)% of their rated burn time.  So for engines that are mission critical and beyond LKO, I'm using high quality engines and planning to only use them for about 30% of their rated burn time.