The v1.2 Hype Train Thread - Prerelease is Out

[taniwha]

5 hours ago, GregroxMun said:

On the orbital mechanics side of things, have you ever noticed how close approaches sometimes jitter back and forth? That's because there's more than just two potential close points. Now all of the orbit crossings (if they exist) will be plotted.

 

I just want to point out that half of those point sets are utterly useless to the targeter as they cross the orbit and you won't possibly get an encounter from them. Of the remaining half, half are furthest points and not closest points (furthest points may be useful for refining SoI encounters): For orbits around a common body, there really are only zero, one, or two intercept points, or more usefully, one or two closest points (there's always at least one, except for the case of two perfectly coplanar perfectly circular orbits). My work made it possible to be more aggressive about searching for those closest points. Also, the found points will be more stable as an algorithm more suited to the problem was used for finding them.

I actually "cheated" with those shots: 1200 samples to find the regions that hold the points, and then a pile more samples to actually find those points (up to 54 each point pair (so up to 232 sampes)). However, 1200 samples is quite expensive: around 50ms. That might not seem like much, but when you consider everything else KSP does, that gets quite expensive. (note: times are for blender+python, and were fairly jittery, so take with a pinch of salt).

Still, things are muchbetter than they were in 1.1.x, and even 1.0.5 .

[Sharpy]

I'm really happy about the encounters. Just two days ago I had KSP mark a point with 40km separation, while utterly failing to mark one with 100 meters separation. It's especially poor at marking really good encounters with not only minimal separation but also minimal speed difference - doing a Hohmann transfer from 230 to 250km coplanar orbit, getting it to show the closest approach was really hard. But if I raised the transfer orbit apoapsis to 300km, making it clearly overshoot past the target altitude, the nodes were firm and clear.

[Brainlord Mesomorph]

"Ping Part Prefab"??

"Autostrut"??

What be they?

 

[taniwha]

Sadly, the major part of the problems in 1.1.x was caused by a very embarrassing mistake on my part when cleaning up some swap code (ie, when swapping the two encounters so the nearest one would always be first). That said, I've seen various targeting problems back in 0.23.5 so...

[Sharpy]

28 minutes ago, Majorjim! said:

I don't think fuel flow through parts has changed at all. Just the indication of how they flow is new. Amiright?

Not entirely. There's also setting priority of which tanks are emptied first. The routing of fuel is pretty much unchanged though, but you can e.g. balance fuel for SSTO. Or - say, you have have two droptanks on the upper stage with a good vacuum engine, and a lower stage with strong engine and a large tank that is not quite sufficient to reach the orbit. You can now set it up so the lower stage will first empty its own tank, then reach into the droptanks, without touching the upper stage's supply. Then after separation you have the upper engine, its tank full and drop-tanks partially depleted.

Normally, either there would be no crossfeed from the upper stage at all, or the droptanks would be depleted first, then the upper stage tank, and by the time you should perform separation you'd be busy pumping all the fuel from the bottom stage back to upper tanks.

[Arsonide]

1 minute ago, Brainlord Mesomorph said:

"Ping Part Prefab"??

That is a developer tool that we have available in the Unity editor, it is not available in game when we build. It's not terribly interesting, don't worry.

[Brainlord Mesomorph]

8 minutes ago, Arsonide said:

That is a developer tool that we have available in the Unity editor, it is not available in game when we build. It's not terribly interesting, don't worry.

and "Autostrut"?

[Vlyan]

yay getting my ticket too !

it seem RemoteTech and ScanSat might have a lot of work to do this time

Edited September 6, 2016 by Vlyan

[Arsonide]

2 minutes ago, Brainlord Mesomorph said:

and "Autostrut"?

Autostruts were a fix for wheels that automatically created a joint to the heaviest part on the vessel when the part loads. This joint is not a physical thing, so it's a bit different than a normal strut. We also use something similar to secure payloads in fairings. In 1.2 the wheel fix is being expanded to an advanced tweakable for every part, if someone wants to enable autostruts everywhere.

[GoSlash27]

You'll find me traveling in style and luxury aboard the Hyper Chief!

 

 

[Sharpy]

9 minutes ago, Arsonide said:

Autostruts were a fix for wheels that automatically created a joint to the heaviest part on the vessel when the part loads. This joint is not a physical thing, so it's a bit different than a normal strut. We also use something similar to secure payloads in fairings. In 1.2 the wheel fix is being expanded to an advanced tweakable for every part, if someone wants to enable autostruts everywhere.

Can you elaborate how they work?

I was getting various rumors, but no solid data. "Drawn to nearest, heaviest part" etc - so nearest, or heaviest? How does fuel contribute - will autostrut change once a tank ceases to be heaviest? What if the heaviest part is staged? Say, I'm launching a pretty heavyweight rover, or a horizontal landing space shuttle that could definitely benefit from autostrut, but all autostruts will go to the launch stage tank. Generally, a lot of partial, incomplete or outright incorrect information floating around...

[ThatPilot]

1 hour ago, cfds said:

 

I guess I will grab a ticket for the Patience Ferry and wait for 1.2.4 and my favourite mods.

 

What an adult and responsible thing to say 

[Superfluous J]

13 minutes ago, Sharpy said:

Can you elaborate how they work?

I was getting various rumors, but no solid data. "Drawn to nearest, heaviest part" etc - so nearest, or heaviest? How does fuel contribute - will autostrut change once a tank ceases to be heaviest? What if the heaviest part is staged? Say, I'm launching a pretty heavyweight rover, or a horizontal landing space shuttle that could definitely benefit from autostrut, but all autostruts will go to the launch stage tank. Generally, a lot of partial, incomplete or outright incorrect information floating around...

Most massive, as chosen the last time the ship part tree changed. So on first load, docking/undocking, staging away something, or any form of unplanned rapid disassembly caused by aero- or lithobraking. Draining your most massive tank will NOT affect strutting until you change the part tree in some way.

I believe you can also set to autostrut to the COM but I see no indication in that screenshot, and therefore assume it's not a thing.

Edited September 6, 2016 by 5thHorseman
added clarity. I hope.

[DunaRocketeer]

23 minutes ago, Arsonide said:

Autostruts were a fix for wheels that automatically created a joint to the heaviest part on the vessel when the part loads. This joint is not a physical thing, so it's a bit different than a normal strut. We also use something similar to secure payloads in fairings. In 1.2 the wheel fix is being expanded to an advanced tweakable for every part, if someone wants to enable autostruts everywhere.

What state are landing gear, landing legs and wheels in for 1.2? I haven't been keeping up with the latest changes, but I'm given to understand that the gear skitters around on runways and the legs slide on terrain. I've heard of odd behaviour with rover wheels as well. I'm thinking of finally upgrading from 1.0 to 1.2, but would like to know if these issues are resolved first.

Thanks.

[Red Iron Crown]

One ticket please.

Also: Stickied. The reaction image rule is suspended in this thread, I've added that to the OP (with apologies to @GregroxMun).

[The_Rocketeer]

Happy 3rd Forum Anniversary to meeeee!

Thanks Squad for this awesome present!

[steve_v]

I will be observing this train from the safety of the station until I am quite certain it won't crash with "Error in `./KSP.x86_64': double free or corruption (out)".
 

[Dafni]

[Sharpy]

11 minutes ago, DunaRocketeer said:

What state are landing gear, landing legs and wheels in for 1.2?

FIXED!!

[Jimmidii]

 

One ticket please!

Edited September 6, 2016 by Jimmidii

[Azimech]

The new flow controls are exactly what I need!

_{*Insert mad laughter here, scientist is optional.}

[Starhawk]

I'm definitely on board for this!

I'm very much looking forward to finally having a use for antennae.  And the fuel flow priority tweakable.

Most especially I eagerly await the changes and optimizations under the hood.  I have more than Reasonable Expectations of a significantly improved gameplay experience.

Happy landings!

[Kerbart]

One ticket, please!

[Crocket]

Tickets? Pfft, please. I'll be ridin' this one King-of-the-North-Pole style. I'll be making sock coffee if you need me.

[steuben]

I'm not climbing on the hype train. Sure the club car is nice and serves nice drinks of my choice of ethane hydroxide, methylated methanol, or basic grain alcohol with various adulterants; and the dining car serves nice snacks.

I'm going to be riding along side on my kickback powered rocket sled.