Change Log for v7.2 Discussion

[ardem]

Hopefully never. 

[Gyrfalcon]

If the maintenance clock continues to tick, can commercial hangers be used in conjunction with fighters? Like many others, my fighters aren't designed to let their maintenance clock tick for more then a week at most. Hauling them in a commercial hanger where the clock is still ticking will result in delivering a squadron of scrap metal if they need a month to reach their destination.

[Steve Walmsley]

If the maintenance clock continues to tick, can commercial hangers be used in conjunction with fighters? Like many others, my fighters aren't designed to let their maintenance clock tick for more then a week at most. Hauling them in a commercial hanger where the clock is still ticking will result in delivering a squadron of scrap metal if they need a month to reach their destination.

The commercial hangar bay is mainly for repairs. However, I have had fighters sat in orbit for months waiting for a carrier with no maintenance and not had any failures. Although they have no engineering spaces, their chance of failure is very low due to their small size. For example, here is the relevant portion of the US fighter from my current campaign:

F-40 Starfury class Fighter    300 tons     2 Crew     53.4 BP      TCS 5.99  TH 42  EM 0
7011 km/s     Armour 1-3     Shields 0-0     Sensors 1/1/0/0     Damage Control Rating 0     PPV 2.25
Maint Life 0 Years     MSP 0    AFR 59%    IFR 0.8%    1YR 3    5YR 43    Max Repair 13 MSP
Intended Deployment Time: 0.5 months    Spare Berths 0   

The class AFR is 59% per year and that is the rate when the fighter has already been in space for a full year (it starts at 0% and will get to 59% after one year of deployment and continue getting higher after that). A fighter with a low deployment clock will have very little chance of failure. A 3 month trip aboard an auxiliary carrier would be very low risk.

[Felixg]

7. 2 looks awesome! I can't wait to play it!  Going to hold off on starting my next game until its out!

[Steve Walmsley]

7. 2 looks awesome! I can't wait to play it!  Going to hold off on starting my next game until its out!

It may be a while. I have a few other changes in mind and with several fundamental changes I need to test it for a while.

[alex_brunius]

Although they have no engineering spaces, their chance of failure is very low due to their small size. For example, here is the relevant portion of the US fighter from my current campaign:

F-40 Starfury class Fighter    300 tons     2 Crew     53.4 BP      TCS 5.99  TH 42  EM 0
7011 km/s     Armour 1-3     Shields 0-0     Sensors 1/1/0/0     Damage Control Rating 0     PPV 2.25
Maint Life 0 Years     MSP 0    AFR 59%    IFR 0.8%    1YR 3    5YR 43    Max Repair 13 MSP
Intended Deployment Time: 0.5 months    Spare Berths 0   

The class AFR is 59% per year and that is the rate when the fighter has already been in space for a full year (it starts at 0% and will get to 59% after one year of deployment and continue getting higher after that). A fighter with a low deployment clock will have very little chance of failure. A 3 month trip aboard an auxiliary carrier would be very low risk.

Does this mean the wiki is wrong? It claims the following:
(AFR, the chance of a spacecraft component failing based on a one year duration)

You seem to suggest here that AFR is the chance each increment that something will fail here after it's gone a year deployment?


Even with 59% per year it should mean 15% per 3 months, which means 15% of your fighters should have damaged components after that low risk trip. If the failing component happens to be engines on any of them they are at big risk for secondary explosions due to normal high power of the engines.

[Steve Walmsley]

Does this mean the wiki is wrong? It claims the following:
(AFR, the chance of a spacecraft component failing based on a one year duration)

You seem to suggest here that AFR is the chance each increment that something will fail here after it's gone a year deployment?

Even with 59% per year it should mean 15% per 3 months, which means 15% of your fighters should have damaged components after that low risk trip. If the failing component happens to be engines on any of them they are at big risk for secondary explosions due to normal high power of the engines.

I haven't read the wiki

The base failure rate increases as a ship spends more time away from port. The chance of a failure is equal to Base Class Failure Chance * Years Since Overhaul * Portion of Year in 5-day Increment.

For example, if a class has an annual failure rate of 50% and has been away from port for exactly three months when a production phase takes place. The failure rate at that point is 50% * 0.25 = 12.5%. If the 5-day increment at that point is exactly five days, the chance of failure in that single increment is 5/360 * 12.5%.

If the same ship has been away from port for three years, the chance would 50% * 3 = 150% * 5/360. If you check the ship display rather than the class, the exact current failure rate is displayed

So a fighter with an annual failure rate of 59% would only have a failure rate of 4.9% after a month in space and 14.75% after three months.

So for a single fighter with an annual failure rate of 59%, the chance of at least one failure in a 60-day journey (starting with zero deployment and assuming exact 5-day increments) would be about 0.88%

[alex_brunius]

Extremely interesting information! Thanks for taking the time to write it. I think the wiki needs some updating

[TMaekler]

This is very complex because the program would have to calculate if the last order of the conditional command made sense if followed by your existing first order. Because of the huge variety of potential orders you could have set, this is likely to result in logic-breaking situations. Therefore the orders are removed so the human player can intervene.

Hmm, that's true. Maybe inserting a buffer for the old orders (instead of deleting them buffering them) from which you then can manually choose to reenter if you see that they fit?!? Although might not be less complex for the program to check if those orders are still valid, I guess... .

[bean]

It may be a while. I have a few other changes in mind and with several fundamental changes I need to test it for a while.

Noooooo!!!!!
Why must you do this to me, Steve?  Why?

Or would it make the testing go faster if you had help? 
Also, to a first approximation, failure chance seems to go up with the square of time spent in space, assuming you start at 0.  That seems a very useful thing to keep in mind, although not terribly realistic.

[TheDeadlyShoe]

Also, to a first approximation, failure chance seems to go up with the square of time spent in space, assuming you start at 0.  That seems a very useful thing to keep in mind, although not terribly realistic.

IMO, it's within the bounds of reason.  older components fail more and take more and more work to fix.  the flip side though is that inevitably some new components are lemons, which isn't modeled in Aurora, not that it needs to be.

this is why it can be important to have good engineering lifetimes on your ships;  all else being equal, it makes them cheaper to deploy for the same length of time because you will face less failures.
/unless im bad at math

[bean]

IMO, it's within the bounds of reason.  older components fail more and take more and more work to fix.  the flip side though is that inevitably some new components are lemons, which isn't modeled in Aurora, not that it needs to be.

this is why it can be important to have good engineering lifetimes on your ships;  all else being equal, it makes them cheaper to deploy for the same length of time because you will face less failures.
/unless im bad at math

Well, the usual curve of failures is U-shaped.  There's a period of high failures during burn-in, then a while when the failure rate is pretty much constant and you're getting essentially random failures, and then the failure rate goes up as stuff wears out.  This model seems to dump us straight into the last phase.  There are mathematical methods to look at when you should do preventative replacements to reduce overall cost.  Some components work differently.  For instance, preventative replacement of vacuum tubes is pointless.  They fail randomly before they wear out, so there's no spike and no point in replacing them before they burn out.  (Which tells you just how old the book I got this from is.)
I'll have to look through that book and see if there might be a good equation to use here if Steve's interested.  The first thought would be to skip burn-in (which may or may not be significant for a given system) and then have a period of pretty constant failures (which would depend on the ship), and then have the failure rate start up.

[Mor]

Does this mean the wiki is wrong? It claims the following:
(AFR, the chance of a spacecraft component failing based on a one year duration)

You seem to suggest here that AFR is the chance each increment that something will fail here after it's gone a year deployment?

Unless I am reading it wrong, both are "true" according to Steve tutorial. AFR on the Class design summary tab assume a ship has one year on its maintenance clock. While the individual ship pages show actual maintenance clock data.

So you'll have to be more specific as to the info on the wiki you refer to, and if you did spot a mistake, please comment here to make sure they see it.

[iceball3]

I wonder if, with a low enough failure rate, if it would be more economic to have a ship deployed for the lower part of it's failure portion, then come back in for overhauls periodically. Or does the overhaul make up for lost time by consuming extra minerals/maintenance supplies equal to the rewound time of normal maintenance?

[linkxsc]

Couple thoughts.
With the new governing rules for maintenance, perhaps the maintenance module could add a couple thousands worth of MSP storage to a maintenance ship (partially to help newbies who might forget to put maintenance storage, and partially so that you can just outright store a bit more)


Meanwhile I wonder. If you had a single station or ship with maintenance modules and civilian hangars. Would those maintenance modules maintain the ships in the hangars? Also would they maintain fighters (which often get very grumpy if not maintained for more than a month or so since I rarely build maint into them)