Preferred CommNet Difficulty

[Blaarkies]

2 minutes ago, klesh said:

For clarity, I meant no additional dishes beyond the ones at KSC.

Ah i see. I do that too. Also, reduce the DSN range modifier to like 0.40 so that it is only as strong as the best relay dish...this way, the super long range DSN wont let your satellite relay network feel un-used 50% of the time

[Jestersage]

20 hours ago, Chaia said:

The occlusion factor means basicly how far the network connection between two vessels can slip into a planet for simulation over-the-horizon sending methods etc.

Setting it to 1 means, that you need line of sight between two vessels for connection. As example for kerbin: Set it to 1 and the connection line has to be outside of the atmosphere. Any lower and the connection may slip into the atmosphere. Same for air-less bodies, except the line may slip into hard rock.

Why did they set those not to 1, especially for objects with no atmosphere? I can understand about the atmosphere to simulate the bouncing of signal by atmosphere...

[KerikBalm]

For lower difficulty... also due to diffraction, you can get signal without line of sight, but close to it.

Just think of the pattern of light that you get when it passes through a narrow slit. There will be places that are illuminated which have no "line of sight" to the light source

[klesh]

9 hours ago, Blaarkies said:

Ah i see. I do that too. Also, reduce the DSN range modifier to like 0.40 so that it is only as strong as the best relay dish...this way, the super long range DSN wont let your satellite relay network feel un-used 50% of the time

 

Good idea, I will try that.  Indeed, all the additional stations (and I hadn't thought of range yet) seemed to make it not much of a challenge.

[monstah]

On 19/10/2016 at 10:34 PM, FullMetalMachinist said:

I'm using it with 100% occlusion, and the DSN modifier bumped up to 20, and vessel antenna modifier taken down to .1

IMO it's silly, bordering on ridiculous, to have to use relays to boost the range because of distance from Kerbin, especially considering KSP's tiny solar system. Relays should only be needed for communication with things that are occluded by a planet/moon. 

Never thought of it that way, and I like this! Will probably try using this myself.

[Tidus Klein]

95% occlusion for atmo bodys an 105% for vac bodys. ill let a little atmo bouncing threw and i just hate when you can clearly see a green line go right threw a mountain, plasma black out on, total control lose on.

 

might just get RT again an run with no delay

Edited October 21, 2016 by Tidus Klein

[waterlubber]

On 10/20/2016 at 3:33 AM, Blaarkies said:

I am really intrigued about it now. I set my Atmospheric occlusion to 99%, to kind of simulate the effect that microwave transmissions can reflect against the atmosphere...and mostly because i am building a ground network to the north pole, this would require 10+ relay probes placed on mountains like the beacons of Gondor. 99% should make it so that i need about 5 probes

Microwave transmissions don't really reflect too well, but shortwave ones can go all the way around easily from sea level

[Jestersage]

11 hours ago, KerikBalm said:

For lower difficulty... also due to diffraction, you can get signal without line of sight, but close to it.

Just think of the pattern of light that you get when it passes through a narrow slit. There will be places that are illuminated which have no "line of sight" to the light source

So what's the more realistic occulsion, 0.90/0.75, 1/1, or some other number?

[moogoob]

On, multiple ground stations, default occlusion and other settings.

Oh, and no plasma black out or full control loss, cuz I'm not that concerned about "realism" and those settings didn't sound fun.

Didja know that with default occlusion there's no dead spot during launch to LKO? Like it was tuned for that or somethin!

Edited October 21, 2016 by moogoob

[swjr-swis]

I keep going back and forth on this.

I chose to start my career with CommNet on for a bit more challenge, plasma blackout on, and occlusion a bit higher because the visible lines clipping into the surfaces bother me more than the slightly more realistic allowance of refraction/reflection. I do want line of sight to matter for transmissions, but not needing signal for control, because I feel probes should be either fully capable of performing pre-programmed maneuvers or fully independent AI anyway, neither of which are proper options in this implementation. The limited control feels too much like a forced game punishment mechanic to me, I'll never play with that on.

On my sandbox game, I have it all off because the fun to bother ratio is not really favourable, and I find that it doesn't really feel like I'm missing anything. But going back to career, I am happily planning and setting up relays.

I'm so glad it's optional.

[KerikBalm]

13 hours ago, Jestersage said:

So what's the more realistic occulsion, 0.90/0.75, 1/1, or some other number?

I suspect the effect is rather small, and a planetary body isn't the same as a sharp edge. Also, the signal when theres not direct line is much weaker, and KSP seems to model all or nothing(I could be wrong). Ionospheric propagation isn't super reliable, and I think KSP already doesn't model hills and mountains as blocking LOS, just a sphere of the planets defined radius.

I can't say what is realistic of not, just that perfect LOS is not always needed with radio waves:

https://en.wikipedia.org/wiki/Radio_propagation#Diffraction

Myself... I play with 0.95 for atmospheric bodies, and 1.0 for vacuum.

From the "Knife edge diffraction" linked to in the article I linked too... this is assuming a barrier blockingthe right half of the signal... signal can stil larrive at places without LOS:

 

Edited October 22, 2016 by KerikBalm

[Renegrade]

17 hours ago, moogoob said:

Oh, and no plasma black out or full control loss, cuz I'm not that concerned about "realism" and those settings didn't sound fun.

Full control loss isn't realism at all anyhow.  Signal delay makes direct control meaningless anywhere in the system aside from LEO (or for certain remote lunar rovers during the Space Race).   Probes are pre-programmed with maneuvers LONG before they reach the target.

The round-time trip to Pluto at c is about nine hours (give or take), for example.   Even to Mars it's 4 minutes to 24 minutes (depending on the positions of Earth and Mars). Real-time control is pretty much out of the question. 

That's some EPIC level lag there.  Three million ping to Mars (worst case), thirty two million ping to Pluto.  Two thousand to the Moon...

[Gaarst]

4 minutes ago, Renegrade said:

Probes are pre-programmed with maneuvers LONG before they reach the target.

This. While I like the idea of a comm network, the fact that we still can't plan manoeuvres doesn't make any sense. There should be a way to plan nodes when you have comms available so that you can still execute them (automatically maybe, with no direct control if they go wrong) even if your craft is on the dark side of the Mun, Duna, Jool, Sun, whatever.

[moogoob]

1 hour ago, Gaarst said:

This. While I like the idea of a comm network, the fact that we still can't plan manoeuvres doesn't make any sense. There should be a way to plan nodes when you have comms available so that you can still execute them (automatically maybe, with no direct control if they go wrong) even if your craft is on the dark side of the Mun, Duna, Jool, Sun, whatever.

IIRC Mechjeb handles like this if you use the maneuver planner - or at least it does with the settings I use. It's performed maneuvers on the far side of the mun for me before I was able to get up relays.

[pandaman]

I have just turned the occlusion up to 1 as thought it looked silly having the lines passing through quite a lot of the planet, I have plasma blackout turned on too

[Blaarkies]

On 10/22/2016 at 3:57 PM, Renegrade said:

Two thousand to the Moon...

Why 2000? Is it not 1300ms one-way?

[IronCretin]

With the higher level probes, or even the lower ones if the maneuver is simple enough, you can execute preprogrammed maneuvers. Plot your node, tell the probe to hold to node, if you can, or if not, manually point it at the node, and then hammer on the throttle when the time comes.

[Renegrade]

8 hours ago, Blaarkies said:

Why 2000? Is it not 1300ms one-way?

Control requires feedback of varying kinds (which could actually more-than-double it, depending on how bad the protocol in question is).

Ex. if you were driving that Soviet lunar rover, you'd make a motion with the controls, which would be transmitted, and the result of that movement would be seen AFTER a video frame returned to Earth (rover->base).   You could blindly send commands with only the one-way latency, but there'd be no correction then..

(and yeah, it's about 1255ms one-way surface-to-surface in ideal locations on both bodies; I tend to lazily round that down to a second)

[Archgeek]

On 10/22/2016 at 8:57 AM, Renegrade said:

Full control loss isn't realism at all anyhow.  Signal delay makes direct control meaningless anywhere in the system aside from LEO (or for certain remote lunar rovers during the Space Race).   Probes are pre-programmed with maneuvers LONG before they reach the target.

The round-time trip to Pluto at c is about nine hours (give or take), for example.   Even to Mars it's 4 minutes to 24 minutes (depending on the positions of Earth and Mars). Real-time control is pretty much out of the question. 

That's some EPIC level lag there.  Three million ping to Mars (worst case), thirty two million ping to Pluto.  Two thousand to the Moon...

It amuses me immensely that you've put that in terms of ping numbers there.  'Seems future online games will need seperate Earth, Luna, and Mars servers.  You'd think they'd be able to abuse entanglement states to get around that, but recent research implies that it doesn't actually work that way -- I forget if it causes the particles to disentangle, or if the "spooky action at a distance" has a delay commensurate with light speed, but either way, get set for some serious lag if you decide to connect to a lunar CoD 12: This Time It's Actually Fish server.

[suicidejunkie]

1 hour ago, Archgeek said:

It amuses me immensely that you've put that in terms of ping numbers there.  'Seems future online games will need seperate Earth, Luna, and Mars servers.  You'd think they'd be able to abuse entanglement states to get around that, but recent research implies that it doesn't actually work that way -- I forget if it causes the particles to disentangle, or if the "spooky action at a distance" has a delay commensurate with light speed, but either way, get set for some serious lag if you decide to connect to a lunar CoD 12: This Time It's Actually Fish server.

"recent research"?

I'm under the impression that it has been known for quite a long time and expected to be so even longer, but sci-fi authors have always been quite willing to ignore such inconvenient things for the sake of the story.

 

A simple way to put it is that the two ends of the connection can get a series of related but random bits from measuring their entangled particles.

You still have to send a conventional message from point A to point B, but you can rig the results of the entangled particles into a high tech one-time-pad that can't be intercepted.

[Archgeek]

1 hour ago, suicidejunkie said:

"recent research"?

I'm under the impression that it has been known for quite a long time and expected to be so even longer, but sci-fi authors have always been quite willing to ignore such inconvenient things for the sake of the story.

 

A simple way to put it is that the two ends of the connection can get a series of related but random bits from measuring their entangled particles.

You still have to send a conventional message from point A to point B, but you can rig the results of the entangled particles into a high tech one-time-pad that can't be intercepted.

Last I checked a few years back there was hope for more serious shenanigans.  If there were any control over the states at one point, though, you'd think there'd be a way to say, set a message packet of a few hundred entangled particles and T/R bit at point a, prior to an agreed-upon time t, and then check the T/R bit at point b come time t, grabbing the message if its been set and setting it back to R, with notable gains so long as the t interval is less than the time it'd take light to make the distance between a and b.  However, inconveniently, it seems that there is no control of the state at point a, only a guarantee that the a and b will have predictably related states.  So yeah, no information teleportation, but sweet for encryption.

[septemberWaves]

I'll require signal for any control, I'll have plasma blackout turned on (any probes dropping into atmospheres will use one of the surface-mounted antennas) and I'll have the extra ground stations turned on. I'll keep the range and occlusion modifiers that the Normal difficulty sets.

[Blaarkies]

14 hours ago, Renegrade said:

Control requires feedback of varying kinds (which could actually more-than-double it, depending on how bad the protocol in question is).

Ex. if you were driving that Soviet lunar rover, you'd make a motion with the controls, which would be transmitted, and the result of that movement would be seen AFTER a video frame returned to Earth (rover->base).   You could blindly send commands with only the one-way latency, but there'd be no correction then..

(and yeah, it's about 1255ms one-way surface-to-surface in ideal locations on both bodies; I tend to lazily round that down to a second)

 

On 10/22/2016 at 3:57 PM, Renegrade said:

Full control loss isn't realism at all anyhow.  Signal delay makes direct control meaningless anywhere in the system aside from LEO (or for certain remote lunar rovers during the Space Race).   Probes are pre-programmed with maneuvers LONG before they reach the target.

The round-time trip to Pluto at c is about nine hours (give or take), for example.   Even to Mars it's 4 minutes to 24 minutes (depending on the positions of Earth and Mars). Real-time control is pretty much out of the question. 

That's some EPIC level lag there.  Three million ping to Mars (worst case), thirty two million ping to Pluto.  Two thousand to the Moon...

Earth to mars: 3minutes ping at minimum distance. So double that gives us 4:24 minutes?
You didn't explicitly state that "Two thousand to the Moon" is measured as ping, but I assumed it was because of "ping to Mars", "ping to Pluto" previously in the sentence.

"Three million ping to Mars": That is 50minutes worth of ping, or a distance at c of 900 000 000km...while the maximum distance between earth and Mars is 401 000 000km. How and where did you pull these numbers out?

I am not even going to double check the Pluto calculation at this point. Just please stop spreading the misinformation, please? Your comments look scientifically correct, people believe that and then argue against actual facts because they are pretty sure they read somewhere about the communication latency between planets.

Edited October 26, 2016 by Blaarkies

[Renegrade]

11 minutes ago, Blaarkies said:

 

Earth to mars: 3minutes at minimum distance. So double that gives us 4:24 minutes?
You didn't explicitly state that "Two thousand to the Moon" is measured as ping, but I assumed it was because of "ping to Mars", "ping to Pluto".

"Three million ping to Mars": Thats 50minutes worth of ping, or a distance at c of 900 000 000km...while the maximum distance between earth and Mars is 401 000 000km. How and where did you pull these numbers out?

I am not even going to double check the Pluto calculation at this point. Just please stop spreading the misinformation, please? Your comments look scientifically correct, people believe that and then argue against actual facts because they are pretty sure they read somewhere about the communication latency between planets.

1. The 4 to 24 minutes is actually one-way-trip time.  I should have been clearer about that (actually I should have doubled them beforehand).  It's 8 to 48 minutes round-trip-time.  The "pings" are specified in RTT exclusively.
2. Yes, 'two thousand to the moon' is in milliseconds (known simply as 'ping' to the internet at large.  I can't say I've always supported this nomenclature but, well, it happens.  No point fighting it anymore).  Very approximate milliseconds, mind you, although it really doesn't matter if it's 2000 or 2400 or 2510 etc.  It's still massive latency.
3. The numbers mostly came from Google's infoboxes or (in the case of the Moon) my memory.  They're a bit loosey-goosey as I doubt anybody needs to know exact times, given that there's practically nothing to talk to at any of the destinations mentioned (only a handful of probes on/around the Moon or Mars, and currently ZERO at Pluto - New Horizons is already almost four AU away from it [SOURCE: The Johns Hopkins University Applied Physics Laboratory]).  Plus I'd assume the people who do need EXACT times already know 'em.

Here's some more accurate figures, stated more clearly, if you wish:

Earth SMA is 149.6GM [Source: NASA], Mars SMA is 227.9GM [Source: NASA].  Mars has a fair degree of eccentricity in it's orbit (almost 0.1), which I'm ignoring as I have better things to do than perform calculations all night.    That means that (again, ignoring eccentricity), the distance to Mars would be 78.3GM at nearest approach, and 377.5GM when it's on the opposite side of the Sun.   That's 522.4s and 2518.4s round-trip travel time at c, respectively, or 0.5 million to 2.5 million "ping" (ie, milliseconds) round-trip-time at c. (2.675 million using your 401GM figure, which may be taking eccentricity into account).  Sure, 3 million 'ping' is outside that range, but only by a bit.

Pluto is currently about 33.26 AU away from the Sun [Source: Celestia], which is about 16,600 light-seconds away (4.6 hours).  Thus, round trip time (at c) is about 33,200s (~9.2h).   The original source I used said '9 hours round trip time', which is obviously a bit less (32,400s or 32,400,000ms or "ping").  A direct line to Earth would be 33.564au [Source: Celestia], resulting in very much the same figures (Pluto and Earth are almost at right angles right now).

The figure I'm using for lightspeed is 299792458 m/s [Source: Wikipedia.  My memory recalls this as 2.9979e+8 m/s, so Wikipedia can't be too wrong here].  For AU, I'm using 1.4960e+11 (149.6GM) [Source: Wikipedia. Also seems to match what I recall (~150GM) ].

The distance to the Moon is > 300MM [Source: my memory, which loves inequalities, and hates exact figures], so it's at least 2s (> 2000 ms or "ping") round-trip time.  I've already stated a more accurate computation in a previous post.

TL;DR: That's some EPIC level lag there.  Two point six seven five million ping to Mars (worst case), thirty two million ping to Pluto.  Twenty five hundred the Moon...

Sources:

NASA: http://mars.nasa.gov/allaboutmars/facts/#?c=inspace&s=distance (matches Wikipedia's data, more or less)

The Johns Hopkins University Applied Physics Laboratory: http://pluto.jhuapl.edu/Mission/Where-is-New-Horizons/index.php

Celestia: http://celestiaproject.net/

[Blaarkies]

29 minutes ago, Renegrade said:

Mars SMA is 227.9GM

Mars SMA is 227.9Gm (meter notation should always be lowercase)

 

29 minutes ago, Renegrade said:

Here's some more accurate figures, stated more clearly, if you wish:

Sorry, it just seemed wildly inconsistent, to the point where I was thought you were doing it almost on purpose. That was mostly due to latency ping vs roundtrip-time. Thanks for the in-depth info here

I apologize if the "misinformation" comment was offensive