Kethane Pack 0.9.2 - New cinematic trailer! - 1.0 compatibility update

[NeoAcario]

If I leave a scanner in orbit when it is turned on will it continue to scan even if I go fly other crafts? and Does the scan that is generated visible every immediately or do i need to transmit it back to base somehow?

No, you have to baby the scanner, it's a KSP limitation (though with a lot of effort (and maybe FPS loss) it might be doable, but...).

It's actually not a KSP limitation. It works perfectly fine in KSP Interstellar:

http://forum.kerbalspaceprogram.com/threads/43839-0-21-KSP-Interstellar-%28Now-with-Heat-Radiators%21%29-Version-0-6-1-%28Beta%29

Well... it mostly does. He's just did / is still ironing out a few bugs.

~Steve

[forsaken1111]

KSP interstellar just has to remember the last time you accessed the craft and then retroactively generate the resources. This would need to track every cell it crossed over during that time which is much harder to do.

You're comparing apples to rabbits.

[Fractal_UK]

It's theoretically possible, you could read the state of the kethane scanner from the save file. Then, if you had some code that you could execute every frame or every few frames, you could check the position of the vessel with the Kethane scanner (the position of the vessel itself is one of the few things that is actually loaded when a vessel is a outside of physics range) then update which cells have been scanned accordingly.

Not really sure what the calculation overhead would be on this approach, I can't imagine it would be too bad but it might be a lot of effort to go to for a relatively minor change.

[NeoAcario]

It would be a tremendous change. I know many people whom choose to let scanners just run all night since they take so long. This way you could just set it and then when your ship arrives you'd have most of the planet mapped.

This would further be awesome because you wouldn't need to sit with your miner. You can simply set it and forget it until you need it.

I'm the type that likes to multitask as many things at a time as possible. I tend to play KSP more like an RTS than a simulator. That's a big part of what pushed me to trying your mod in the first place, Fractal_UK - it runs in the backround.

~Steve

[Galane]

I scanned Minmus at 50x speed while keeping an eye out across space at the progress of the Kleveland kethane miner/rover towards Mun. I got to learn how to use the fine tune closest approach function of MechJeb because it lost the Hohmann transfer plot on the Munbound ship.

Once Kleveland was in a tight orbit, I put it on 2x warp and went to bed. Next morning it had enough mapped to pick out a spot to land.

Edited October 11, 2013 by Galane

[Gaius]

KSP interstellar just has to remember the last time you accessed the craft and then retroactively generate the resources. This would need to track every cell it crossed over during that time which is much harder to do.

You're comparing apples to rabbits.

It's not that much harder. Unfocused satellites don't maneuver. Given a satellite's orbital parameters and current position, you can tell exactly where it was two seconds ago, five minutes ago, three days ago, two years ago, etc. Exactly what terrain it flew over between any two points in time is not a mystery, regardless of whether you were "looking" at the time or not. Cells could easily be marked retroactively.

Edited October 11, 2013 by Gaius

[Galane]

Do cells have to be rescanned to show the reduced amount as kethane is extracted? So far I've only pulled 4K each from two cells and the amount on the moseover popup on the grid didn't drop.

[taniwha]

It's not that much harder. Unfocused satellites don't maneuver. Given a satellite's orbital parameters and current position, you can tell exactly where it was two seconds ago, five minutes ago, three days ago, two years ago, etc. Exactly what terrain it flew over between any two points in time is not a mystery, regardless of whether you were "looking" at the time or not. Cells could easily be marked retroactively.

Right, start coding. Be sure to take into account the scanner(s) going in and out of range, whether or not there is sufficient power for the scanner(s), the variable^* rotation of the planet under the scanner(s), and probably many other things. Also, don't forget that a ship skimming along at 70km above Kerbin can be running at 100000x timewarp if a different ship is in focus: that's somewhere between 55 and 83 orbits in one second.

^*From the scanner's point of view.

[NeoAcario]

Right, start coding. Be sure to take into account the scanner(s) going in and out of range, whether or not there is sufficient power for the scanner(s), the variable^* rotation of the planet under the scanner(s), and probably many other things. Also, don't forget that a ship skimming along at 70km above Kerbin can be running at 100000x timewarp if a different ship is in focus: that's somewhere between 55 and 83 orbits in one second.

^*From the scanner's point of view.

The code is already written... for the game. You just have to code to note the last time you focused the probe... and then plot it's course since then. This would take all of <1 second during load/ship change.

~Steve

[forsaken1111]

The code is already written... for the game. You just have to code to note the last time you focused the probe... and then plot it's course since then. This would take all of <1 second during load/ship change.

~Steve

His point was that if its so easy, go ahead and do it. You guys keep saying its possible and its really easy when you really have no idea what it would require. So go ahead and make this change!

[Majiir]

There are two separate issues here.

Tracking satellites that aren't currently controlled is easy. However, since vessels aren't simulated unless they're in physics range, flow of resources (including electric charge) isn't simulated. The scanning system would have to simulate electrical generation, capacity and consumption, and there are a number of ways this could break.

Allowing high-resolution scanning at high timewarp levels is mathematically and computationally difficult. Leaving a ship in orbit and then back-calculating what cells it's passed over is not a trivial task. We had a long discussion about this on IRC, and the phrase "harder than n-body" was uttered near the end. Leaving logarithmic timewarp attenuation in place makes things significantly easier, but it leaves us with that "gaps" problem.

I'll be tweaking a few scan numbers for 0.8, but improving the underlying mechanics isn't as easy as everyone seems to think. (Not without making some serious compromises in other areas, anyway.)

[forsaken1111]

There are two separate issues here.

Tracking satellites that aren't currently controlled is easy. However, since vessels aren't simulated unless they're in physics range, flow of resources (including electric charge) isn't simulated. The scanning system would have to simulate electrical generation, capacity and consumption, and there are a number of ways this could break.

Allowing high-resolution scanning at high timewarp levels is mathematically and computationally difficult. Leaving a ship in orbit and then back-calculating what cells it's passed over is not a trivial task. We had a long discussion about this on IRC, and the phrase "harder than n-body" was uttered near the end. Leaving logarithmic timewarp attenuation in place makes things significantly easier, but it leaves us with that "gaps" problem.

I'll be tweaking a few scan numbers for 0.8, but improving the underlying mechanics isn't as easy as everyone seems to think. (Not without making some serious compromises in other areas, anyway.)

I'm glad you chimed in.

As a stopgap, would you consider a more power-intensive scanning part that scans all adjacent cells as well? So you get a wider area scanned per pass. Perhaps with a minimum altitude for scanning, like a minimum of 1 planetary radius out from the surface.

[Commander Gizmo]

I'm glad you chimed in.

As a stopgap, would you consider a more power-intensive scanning part that scans all adjacent cells as well? So you get a wider area scanned per pass. Perhaps with a minimum altitude for scanning, like a minimum of 1 planetary radius out from the surface.

I would like to see something like this. The EL ore scanner is interesting because it is huge (over 2.5m), forces the end of stack (if placed there), and heavy (4 tons). I had quite a time building something that could scan at 1000x and not miss a hex. But it was fun trying. I'd love to see something that big for Kethane that could scan maybe four times better (and maybe even wider) than the current large scanner, as long as it was a real challenge to deploy any number of them. Make us work for the privilege.

[stregone]

There are two separate issues here.

Tracking satellites that aren't currently controlled is easy. However, since vessels aren't simulated unless they're in physics range, flow of resources (including electric charge) isn't simulated. The scanning system would have to simulate electrical generation, capacity and consumption, and there are a number of ways this could break.

Allowing high-resolution scanning at high timewarp levels is mathematically and computationally difficult. Leaving a ship in orbit and then back-calculating what cells it's passed over is not a trivial task. We had a long discussion about this on IRC, and the phrase "harder than n-body" was uttered near the end. Leaving logarithmic timewarp attenuation in place makes things significantly easier, but it leaves us with that "gaps" problem.

I'll be tweaking a few scan numbers for 0.8, but improving the underlying mechanics isn't as easy as everyone seems to think. (Not without making some serious compromises in other areas, anyway.)

How about doing something simple like scanning random cells while out of focus. I mean already the optimal scanning strategy isn't based on actual in-game physics or rules. It is almost completely governed by meta-game rules (The idiosyncrasies of the way time warp works). So making it work 'less realistically' isn't really. The way it works now is like designing a fps mechanic around the way the pause menu works.

I think making it less tedious, even if it isn't 'realistic' would be a net gain. You still need to do the same stuff (get a scanner into orbit of the body), minus leaving the game time warping while you sleep. And the end result is the same (a scanned planet).

[NeoAcario]

We had a long discussion about this on IRC, and the phrase "harder than n-body" was uttered near the end.

Thanks for the clarifications! I was simply curious why I never saw talk of it on these forums before... that I can recall.

~Steve

[Galane]

What about having scanning work only at 1x when the focus isn't on the scanner? Then at warp it quits. If you want faster scanning you have to sit there and watch it scan.

If that's a possible compromise it would be a good balance and make it more worthwhile to deploy scanners across the system so that at least some of the time they'd be working when you're not watching paint dry staring at it scanning.

[DeathSabre]

Download link appears to be down. Unless its just me?

[forsaken1111]

Download link appears to be down. Unless its just me?

Disable your antivirus briefly. Its been flagged because the ISP/hosting company was used by someone else to host malware, but its safe.

[Psycix]

There are two separate issues here.

Tracking satellites that aren't currently controlled is easy. However, since vessels aren't simulated unless they're in physics range, flow of resources (including electric charge) isn't simulated. The scanning system would have to simulate electrical generation, capacity and consumption, and there are a number of ways this could break.

Allowing high-resolution scanning at high timewarp levels is mathematically and computationally difficult. Leaving a ship in orbit and then back-calculating what cells it's passed over is not a trivial task. We had a long discussion about this on IRC, and the phrase "harder than n-body" was uttered near the end. Leaving logarithmic timewarp attenuation in place makes things significantly easier, but it leaves us with that "gaps" problem.

I'll be tweaking a few scan numbers for 0.8, but improving the underlying mechanics isn't as easy as everyone seems to think. (Not without making some serious compromises in other areas, anyway.)

So retroactive scanning is a no-go.

What about server-client, so that one server can simulate multiple scanning probes?

+1 to the wider scan path using a heavy, advanced scanner (or, a multitude in combination!)

Edited October 12, 2013 by Psycix

[Majiir]

What about server-client, so that one server can simulate multiple scanning probes?

That wouldn't do anything from a technical standpoint. For eliminating timewarp scaling, the simulation is difficult no matter how you expose it in gameplay. For electrical simulation, you could pretend the "server" part is somehow taking over all the electrical responsibilities of the individual satellites, but that would feel really weird.

[OopsThatNotWork]

Love the mod. Just checking, 0.8 coming with 0.22 right?

[Majiir]

Just checking, 0.8 coming with 0.22 right?

Yep. Maybe earlier. It won't have everything I originally planned for 0.8, but the improved grid is really nice and it might have other goodies you'll like.

[togfox]

Hey majir, if a new release is coming out, have you considered restricting your stock scanners to look for kethane only? Your scanners scan for all resources by default which discentifies resource makers from including their own models/scanners.

Of course, the flip side is that people will want to make a new resource without the hassle of making new models.

Ive been thinking about this for a while with my bacteria resource and im undecided which approach benefits the community the most. Im personally thawrted by the fact your scanner makes my bacteria soil probe redundant, while recognising few ppl make resources and even less can model. Tough call really.

[Majiir]

Hey majir, if a new release is coming out, have you considered restricting your stock scanners to look for kethane only? Your scanners scan for all resources by default which discentifies resource makers from including their own models/scanners.

They only detect Kethane as of 0.7.7.

[LOLCAT254]

Avast is saying that the download page has a risk of giving me a virus.