puetzk

I'm seeing this as well, but there does seem to be something relevant in the log: "It is illegal for a join to connect a rigid body to itself". Seems to happen if I have a physics-less part attached to the fairing base (in my case, it was the RTShortAntenna1 from RemoteTech), or attached to something radial-attached. Turning off auto-stuts (or ejecting the fairing) releases the rocket. So my guess is something in the auto-strut logic is following the part tree back through physics-less or radial attachments to find pick a rigid body, and manages to strut the fairing base to itself. The fairing base then becomes an immovable object, at least until the fairing is jettisoned and the auto-struts disconnect.

I'm really loving the new 1.x GUI, great work! One cosmetic bug report (in 1.0.10): mean Anomaly is being reported in radians, but labeled as degrees. I'm not sure if this was different in 0.6 or not, this is the first time I've gotten fancy enough to use it (trying to launch a tetrahedral commsat constellation). No strong preference whether this gets fixed by converting to degrees or by labeling as radians; it's not a real angle anyway. Secondly, may I request Right Ascention/Celestial Longitude (relative to the body center being orbited) as an addition to the Orbital category? Off by default, since it's pretty specialized, but it would be useful for tasks like scheduling a plane change such that it results in a particular Longitude of AN). It would also be really nice to have a Epoch field/mode (maybe with a "Mark" button that sets it to current UT) that freezes the time-varying numbers (e.g. the 3 anomaly values, the "Angle to X" values, maybe also their "Time To X" counterparts). This would make it a lot easier to compute the necessary adjustments to phase satellites that are in different planes (within a plane one can just use the rendezvous phase angle, of course). To avoid confusion, maybe make the "Epoch" field and the values its affecting turn a different color than the usual green when it's active, to show they are linked. UI-wise, this is a bit like [ktf].VtD's request for info after a maneuver node (that would also be great!), but it should be a lot simpler to implement since it's the same conic; just evaluate the current elements at the epoch instead of the current time.

This works in deep-space (beyond Kerbin SOI), and in low orbit, but It doesn't work well at all in between. If you're doing any missions in cis-Munar space (beyond omnidirectional range, but not actually in Mün/Minimum SOI - e.g. something orbiting Kerbin at the same altitude as Minimus, but 60° ahead), it's incredibly frustrating. You basically can't use time warp because you'd need to figure out which relay will be up at the desired time before beginning the warp, and you won't get a chance to fix it if you overshoot. Right now about all you can do is put multiple dishes on the craft, so you can always have one pointed at each relay. This feels like a silly way to design the communication when the dishes are all steerable :-) Some thoughts that don't feel too cheat-y that would fix this: 1) make the "pointed at X=pointed at all Y" mechanism work based on the angle between the relay and the selected target being less than some threshold, rather than SOIs. This might probably prohibit some things that work today though (the common setup with 3 evenly-spaced Keostationary relays has them are far enough out, and far enough apart, that at opposition they'd be 32° apart - a receiver in Mün orbit probably shouldn't be able to just target them all generically). Even from Minmus it would be hard (they'd be 8° apart). Being generous, a 2.5m Ku-band dish would have a -10dB beam width of about 1°, a 1m dish about 2.5°. A view 30° wide is basically an omni, but the Minmus angles are in plausible range of gain/beamwidth for a Yagi antenna. So hey, more parts :-) 2) be able to select multiple targets when aiming a dish, as long as they are in the same SOI, and have the dish take automatic action on LOS to iterate through each selected target (at some interval) until contact is regained? This could be kind of fun, especially with the new tracking dishes; you'd get reminded of the comm needs each time you got LOS because your satellite set, then you'd watch and wait while the dish hunts, without it ruining the mission. Either of those feels realistic to me, and some brief gaps for the hand-off would be OK (ISS broadcasts often have dropouts when transitioning to a different TDRS).