minimum jump distance from main station

[Commander McLane]

Through some tests I just found out that I can jump out of the system without being mass locked by the main (Coriolis) station from just 4450 meters in front of the station entrance. This feels wrong to me. I was under the impression that the minimum jump distance is about 10000 meters, roughly where the beacon is located, and I seem to remember that in previous versions of Oolite I got mass locked by the station if I was just short of the beacon.

I just fired up Oolite 1.65. No textures are displayed at all, only flashers and exhausts, which gives a strange visual impression. (MacBook Pro, 2.4 GHZ Core i5, NVIDIA GeForce GT 330M, OS X 10.6.6)

Anyway, I launched and tried to jump. At 4450 meters I wasn't able. I could jump from about 6775 meters, which is still short of 10000 (so maybe I misremembered), but considerably more than 4450.

Which leads me to the question: has anything changed with regard to calculating when a jump is cancelled due to mass lock? 4450 meters from the station entrance, not even halfway to the beacon, doesn't seem like a good place to create a wormhole. It just doesn't feel right. The system-leaving AIs usually let the NPC get some more distance between himself and the station, so I don't think the risk for the player of launching from the station and immediately getting sucked into a wormhole is very real. Still, I feel it shouldn't even be possible.

[Zireael]

I was under the impression that the minimum jump distance is about 10000 meters, roughly where the beacon is located, and I seem to remember that in previous versions of Oolite I got mass locked by the station if I was just short of the beacon.

I had the same impression. I don't know when it changed, I always hit H after arriving close to the beacon...

[Switeck]

With both my Cobra 3 and Boa 2, so long as I move away from the station at max throttle I can start the hyperspace countdown the moment I first gain control after leaving the station...and not get mass-locked.

[JensAyton]

I just fired up Oolite 1.65. No textures are displayed at all, only flashers and exhausts, which gives a strange visual impression. :shock: (MacBook Pro, 2.4 GHZ Core i5, NVIDIA GeForce GT 330M, OS X 10.6.6)

Yeah, I noticed this the other day. I don’t intend to do anything about it, but after MNSR I do plan to keep the stable branch maintained while working on the Next Big Thing.

Which leads me to the question: has anything changed with regard to calculating when a jump is cancelled due to mass lock?

I can’t remember a deliberate change. Possibly it was modified to be mass-dependent at some point? (Bug #17940)

[Cody]

Probably not relevant, but you have to get well clear of the Torus station in order to jump.
Usually it has to be near the edge of the scanner.

[aegidian]

Might be worth someone recalculating the value of 'k' in ShipEntity.m

Code: Select all

- (int) checkShipsInVicinityForWitchJumpExit
{
        // checks if there are any large masses close by
        // since we want to place the space station at least 10km away
        // the formula we'll use is K x m / d2 < 1.0
        // (m = mass, d2 = distance squared)
        // coriolis station is mass 455,223,200
        // 10km is 10,000m,
        // 10km squared is 100,000,000
        // therefore K is 0.22 (approx)

        int result = NO_TARGET;

        GLfloat k = 0.1;

        int                     ent_count =             UNIVERSE->n_entities;
        Entity**        uni_entities =  UNIVERSE->sortedEntities;       // grab the public sorted list
        ShipEntity*     my_entities[ent_count];
        int i;

        int ship_count = 0;
        for (i = 0; i < ent_count; i++)
                if ((uni_entities[i]->isShip)&&(uni_entities[i] != self))
                        my_entities[ship_count++] = (ShipEntity*)[uni_entities[i] retain];              //      retained
        //
        for (i = 0; (i < ship_count)&&(result == NO_TARGET) ; i++)
        {
                ShipEntity* ship = my_entities[i];
                Vector delta = vector_between(position, ship->position);
                GLfloat d2 = magnitude2(delta);
                if ((k * [ship mass] > d2)&&(d2 < SCANNER_MAX_RANGE2))  // if you go off scanner from a blocker - it ceases to block
                        result = [ship universalID];
        }
        for (i = 0; i < ship_count; i++)
                [my_entities[i] release];       //              released

        return result;
}

0.1 seems a little arbitrary and off (assuming this code is still in use!)

[JensAyton]

Interesting. According to svn blame, though, that line was last touched in r170, shortly after 1.55 was tagged, so it can’t be the proximate cause. (The code is in use, and hasn’t been changed recently. I suspect ship mass calculations may have been changed at some point. The mass of a Coriolis station in trunk is 242,106,560.)

[Commander McLane]

I guess having the mass of a Coriolis halved would indeed explain its smaller mass lock ability.

[Eric Walch]

The mass of a Coriolis station in trunk is 242,106,560.)

That weight decrease is called progress. We get stronger and stronger materials over time, allowing us to build stations with thinner walls. That happened already with 1.74. In 1.73 the stations were twice as heavy.

I also measured the weight of a Behemoth type-1. 6.70624e+07 in 1.73 and 6.3323e+07 in 1.74. So, here the difference is not so drastic. Probably because it is a fighting machine and they decided to not build a lighter ship but use the stronger material to make a stronger hull.

[JensAyton]

Thinking about it, this is probably because of the higher resolution collision octree generation. mass is calculated from “volume”, which is actually the number of octree cells that intersect the surface of a ship. We could probably get the old values by limiting recursion depth in “volume” calculations; I’ll try it for 1.75.1.