Technical questions re military hardware.

I've read things too quickly quite a few times, too. Had to quickly delete posts or add Oopses far too often.

SWM wrote:Actually, the astronomical term is that Sphinx is in conjunction with Manticore. 'Opposition' is when they are on the same side of the star. But I wasn't going to bother puting on my astronomer hat just to bring up the terms.

Doesn't it depend on the point of reference? I was thinking of a (purely hypothetical, of course) viewpoint on Manticore-A, would Manticore and Sphinx be in opposition to that viewer, with Manticore-A and one of the planets in a superior(?) conjunction when viewed from the other one?

Dafmeister wrote:

SWM wrote:Actually, the astronomical term is that Sphinx is in conjunction with Manticore. 'Opposition' is when they are on the same side of the star. But I wasn't going to bother puting on my astronomer hat just to bring up the terms.

Doesn't it depend on the point of reference? I was thinking of a (purely hypothetical, of course) viewpoint on Manticore-A, would Manticore and Sphinx be in opposition to that viewer, with Manticore-A and one of the planets in a superior(?) conjunction when viewed from the other one?

Okay, time to get out the astronomer hat.

All of these terms are taken from the point of view of a planet or other body orbiting the star; it does not make sense to talk about a viewpoint from Manticore-A.

From that viewpoint planet, any other planet which is in the direction of the star is In Conjunction, and any planet which is in the opposite direction of the star is In Opposition; in other words, it means "in opposition with the star" or "in conjunction with the star" from that viewpoint. That's why it does not make sense to talk about conjunctions and oppositions from the viewpoint of the star.

Obviously, a planet in opposition must have an orbit outside the viewpoint.

In the case of a conjunction, if the planet has an orbit outside the viewpoint then it is simply called a conjunction. If the planet has an orbit inside the viewpoint, then it is called an inferior conjunction or superior conjunction depending on whether the planet is between the viewpoint and the star or on the other side of the star.

So let's look at the example of the Manticore system. If Manticore and Sphinx are on the same side of Manticore-A:
* from the view of Manticore, Sphinx is in opposition
* from the view of Sphinx, Manticore is in inferior conjunction

If Manticore and Sphinx are on opposite sides of Manticore-A:
* from the view of Manticore, Sphinx is in conjunction
* from the view of Sphinx, Manticore is in superior conjunction

From Manticore, Sphinx can never be in inferior or superior conjunction--it can be in conjunction or opposition. From Sphinx, Manticore can never be in opposition, it can be in superior conjunction or inferior conjunction.

Ah, I see. Thanks for the explanation .

Yeah, thank you. That's very clear and interesting.

Btw, just before that specific passage it states that by this time Sphinx has moved out of the RZ.

Jonathan_S wrote:I remember it, though I believe the passage you're thinking of was actually in Shadows of Saganami (when Helen was asked the plot a least time course from the Lynx Terminus to the Spindle System). I had it in mind when I alluded to being forced to drop into a lower band (because of something like a rouge wave).

But the transit times compared to the velocity multipliers RFC gave us for each hyper band show that in this specific instance there isn't time for the ships to have had to go into a lower hyper band. They could barely make the stated transit times if they bee-lined straight there in the Theta bands at full speed. There's barely time for them to accelerate; much less to drop a band or two, come back up, and reaccelerate.


Though I'm still unclear why a warship in a hurry would ever want to follow a wave into a lower hyper band. Yes, avoiding a rouge wave might force you to. But Helen's example made no sense to me; least time courses pretty much mandate that you keep to the highest hyper band you can (to keep your velocity multiplier up) and that you deviate from the straight line path as little as practical. (Though at the beginning it can be worth following a grav wave a little out of the way where the 10x accel bonus outweighs the extra distance. But once up to cruising speed there's no speed advantage to being in a wave)

If there are areas to avoid then the routing gets more interesting. How big an area, does it extend to all bands, is it quicker the pay the energy losses and reduction in velocity multiplier to dive "under" in in lower bands and have to climb back - or to add lots of distance to go around in in your current band. But that didn't seem to be what Helen was calculating <shrug>

(And More than Honor had a background essay that included some historical limitations of sails; like having to tack when going "up wind"; but gave the impression those basically didn't apply anymore)

Now a warship on a long patrol might prioritize fuel over speed; so you'd make tradeoffs for following grav waves whenever that didn't overly delay you (where that's a bit of a subjective determination). But that wasn't the request in SoS - that was to plot a least time course.

It´s specifically stated that if she just let the computer do a "straight line" course, it could cause a much slower trip in total.

(currently rereading exactly that book)

The hyperspace version of weather patterns seems to be how it works.

Tenshinai wrote:It´s specifically stated that if she just let the computer do a "straight line" course, it could cause a much slower trip in total.

Computers are actually very good at solving those kind of routing problems and are about the only way to come up with the optimal solution. Though if I was a military I probably wouldn't want to always use the optimal solution, a near optimal solution that was less predictable would seem like a good option to have.

kzt wrote:Computers are actually very good at solving those kind of routing problems and are about the only way to come up with the optimal solution. Though if I was a military I probably wouldn't want to always use the optimal solution, a near optimal solution that was less predictable would seem like a good option to have.

Seriously? Why do you think the "travelling salesman" problem is still a classic within computer programming?

Because NOONE has still managed to do software than can reliably solve it(and compared to a human going over it, computers are bloody useless).

Just look at common strategy games, even good ones move AI sometimes end up moving units around like a complete idiot.

Or the follies of following GPS directions without checking to make sure they are sane, the only reason they are no longer (USUALLY!!!) sending people off of cliffedges is that library preset routes and node detail have improved immensly.

And it´s still pathetically common that GPS equipment directions will have you go a longer route than it takes a human about 10 seconds to figure out.

And that´s in 2 dimensions, hypernavigation in honorverse is at least a 4 dimensional problem depending on how you count.


No, computers utterly suck at solving routing problems.
The only way they can do it well is by testing potential, more or less randomised routes a few million(or trillion) times and then pick the best result.

A computer only sees the numbers, not what they represent, so unlike humans it is completely unable to make connections based on the "complete/big picture" information.
It´s why computers are so poor at playing go, good pattern recognition is extremely difficult to program.

What i found when i was working with a random/genetic pathfinder program was that the amount of code increased exponentially for every tiny little bit of improved "skill" of the software in getting anywhere ASAP.

And again, that was in TWO dimensional ubersimplistic mapping, with only simple "yes/no" terrain obstacles, even just adding the option of "yes/no/slow" to the terrain made the software become drastically less efficient, more like stupid beyond belief.

And fixing it became essentially futile, the amount of code required just shot up to insane levels.


Modern gaming move AIs can look really neat and all with pathing, up until you take a moment to realise that they´re throwing around thousands and thousands of pathing nodes together with plenty of cheating, despite being nowhere near reality in scale.

It´s as if a human had to guess a few hundred times every time you lifted your foot, where you would put it down next to take a single step.

Routing and work scheduling are among the things computers are the most terrible at.
They have no common sense to tell them the obvious.

Yeah, and you'll never get a computer program that can defeat a chess grandmaster. Solutions for the traveling salesman problems have been made for problem sets of > 80,000 cities, and the scope of the problem posed is a lot simpler and people have had centuries to write algorithms to solve it better. Computers are very, very good at looking at a lot of alternatives very, very fast. This is particularly true given that in the Honorverse you have a totally absurd amount of computing capability on a ship.

Yeah, and you'll never get a computer program that can defeat a chess grandmaster.

Noone has managed it yet without cheating to the point where playing is utterly pointless.

When the computer relies on a library with hundreds of TB worth of data about previously played games and variations on them rather than processing power, what´s the point of having the computer there for anything beyond searching the database anyway?

You´re not playing against a computer any longer by that point.

Computers are very, very good at looking at a lot of alternatives very, very fast.

Oh yes, never said anything else. Problem is they suck at coming up with good alternatives to look at in the first place.

Garbage in, garbage out.


And textev already makes it blatantly clear that processing power alone isn´t good enough anyway.



Why don´t you try a "little" experiment?
Get 2 identical sailboats, have an expert crew on the one, put a computer in charge of the other, then see which one can get through a course faster.

Experiments almost like that has been done. Rarely does the computers fare well.

Solutions for the traveling salesman problems have been made for problem sets of > 80,000 cities, and the scope of the problem posed is a lot simpler and people have had centuries to write algorithms to solve it better.

And still, a 10 year old with a map, pen and ruler can usually solve it better in a tiny fraction of the time.

Someone highly spatially skilled with better tools, well computers simply doesn´t beat them except with absurd once in a million luck.

And once more, 2 dimensional problem vs many dimensional with a lot of extra complexity added, which means computer power required to even reach the current level of "incompetence" goes up literally astronomically higher.