[WIP] Random Failures (for your masochism needs XD)

[HoneyFox]

Well the thread title tells all.

Current progress:

Basic framework for random failures: can support multiple types of failures now, still lacking "Repair/Replenish" & "Mission Controller additional quality-assurance expense affecting probability of failures" designs though.

Some failures implemented:

Decoupler random activation due to overheat,

Engine random explosion/shutdown due to resonate or exceeded vibration,

Engine random explosion due to overheating,

Resource leakage due to Gee load and/or exceeded amount of fuel in a tank with a very low dry mass (balloon tank)

Source-codes are on GitHub: github.com/HoneyFox/RandomFailures/

Currently the development pace is slow so don't expect it to have many new failures coming out very soon.

Also I'm expecting other coders to join and help.

[TheCrimsonCorndog]

Don't engines already explode due to overheating?

[Guest]

Engines do. Decouplers, however, are harder to overheat.

I have another proposition. Accident investigation simulation. Track whether a single part had failed on any particular design. After that (assuming a flight is not reverted) disable failures on that particular part on this craft, and every later iteration (made by loading the .craft file and re-saving it). After 3 failures of the same type of part, disable failures for all identical parts on the entire ship. Additionally, there could be an option to also make the entire design unavailable for a while for this all to happen (you usually ground the entire vehicle family after a major failure).

[HoneyFox]

Don't engines already explode due to overheating?

Stock game logic will and only will explode when it just exceeds the maxTemp, this will however have probability when temperature is getting close to maxTemp.

[skykooler]

I have been waiting for a mod like this to come along! Two things I would want to be able to fail though:

- Decoupler fails to activate - this can happen in RL space programs as they depend on all of the explosive bolts going simultaneously

- Engine fails to start. This would give us an actual reason to put launch clamps on a separate stage from the engines!

[Scripto23]

Yes! This will actually mean there will be uses for launch escape systems, engine out capability, redundant parts, etc. I was hoping someone would make a mod like this. And of course it is HoneyFox who makes some of the most underrated mods in this game, they might not be flashy part mods with ramjets and such, but they really make for a greatly improved gaming experience.

To add to what skykooler already mentioned, some features I would like to see:

-Parachute failures (or maybe partial failures, so they don't completely open meaning decreased drag)

-Life support failure. Maybe hard to do, will depend on which LS mod being used

-Micrometeor impacts. Could be integrated into leaking of resources.

-Solar panels getting "stuck" while opening

I'll probably add others as I think of them.

[Helix935]

you may want to look into this although it is abandoned

micrometeorites

Edit: you can also try doing things like solar flares with radiation shelters lined with lead or ice for this

Edited March 8, 2014 by Helix935

[Camacha]

I like the idea of failures, as long as the player has a way of either dealing with them or influencing them. Of course there will be unexpected failures in real life, but there are more options for dealing with them as well. I can not force my Kerbals to do additional safety checks or redesign my engines.

The idea that things are more likely to fail when hot sounds good - that is something you can influence and seems reasonable. The other stuff sound like things that can be dealt with too

I have always felt that KSP was not really appropiately fully portraying the uncertainties and hazards of space travel - when you venture into a for humans highly dangerous environment like space and you fully depend on technology to survive, any defect could be a major problem

Edited March 8, 2014 by Camacha

[Space_Kraken]

Ksp has it's own set of bugs that cause this already, except it is more frustrating fixing the buggy parts.

[Scripto23]

I think it would be cool if there was a function that could be called so you could easily add failures to specific parts that aren't already covered in the mod via modulemanager. And even the type of failure and rate of failure and other parameters, for example

@Part[KethaneDrill]
{
explosionFailure = false
functionFailure = true
failureRate = .01
evaFixable = true
randomActivation = false
randomResourceLeak = false
}

[CaptainKipard]

I love this idea, I really do, but could you please limit the failures to the kinds of things that have already happened to spaceships and aeroplanes? I know Kerbals are crappy engineers but don't go overboard. I'd like this to be reasonably realistic.

[Scripto23]

I love this idea, I really do, but could you please limit the failures to the kinds of things that have already happened to spaceships and aeroplanes? I know Kerbals are crappy engineers but don't go overboard. I'd like this to be reasonably realistic.

I don't understand, isn't it already like that? And that's why I was suggesting the failures be tweakable via modulemanager so you could set the severity and type of failure etc.

[CaptainKipard]

I don't understand, isn't it already like that? And that's why I was suggesting the failures be tweakable via modulemanager so you could set the severity and type of failure etc.

Yeah, maybe. I admit my post was largely preventative, and not in response to anything really. I just wouldn't want something stupid like wings breaking. I seriously doubt something like that would happen to NASA.

[biohazard15]

I seriously doubt something like that would happen to NASA.

Actually it did, and in a quite Kerbal way.

[curiousepic]

I think it would be cool if there was a function that could be called so you could easily add failures to specific parts that aren't already covered in the mod via modulemanager.

Oh I hope this is possible! Just disallow a part module for a failing part until it's fixed.

Some suggestions:

- Engines failing to cut off or change throttle (due to overheating or otherwise)

- Individual RCS block failures similar to engine ones

- Any deployable part failing to deply/retract

[skykooler]

Actually it did, and in a quite Kerbal way.

Well, NASA's problem is obvious; they need more struts.

Does Helios count as an infiniglider?

[frizzank]

I like the idea of failures for poorly or over designed rockets

[Scripto23]

I like the idea of failures for poorly or over designed rockets

I think that's a good idea, but how could that be implemented? In other words, what will constitute a poor/over designed rocket?

[Amaroq]

I think it would be cool if there was a function that could be called so you could easily add failures to specific parts that aren't already covered in the mod via modulemanager. And even the type of failure and rate of failure and other parameters, for example

@Part[KethaneDrill]
{
explosionFailure = false
functionFailure = true
failureRate = .01
evaFixable = true
randomActivation = false
randomResourceLeak = false
}

Very nice idea, Scripto23. Opening it up to the community in that way sounds pretty much perfect.

[Meatsauce]

I like the idea of forcing direct or random failures on my designs to see how they hold up. Such failure initiation could be as simple as adding a 'destroy' context menu to each part.

Not sure if there is anything out there that does this.

And of course, for large vessels, causing a critical failure would result in rapid and catastrophic disassembly; I would expect nothing less than a glorious display of fragments. To this end, it would be fun to render fire and smoke from fuel-bearing and other volatile componants.

[Scripto23]

That would be cool. You could have an included cfg with a line like "debug = true" which will then display the "destroy" context menu to force failures on certain parts for testing

[8749236]

Perhaps relates the chance of failure to the total part counts..

And the chance of failure increases exponentially for every parts you add =D