Jump to content

[PART, 1.0.2] Anatid Robotics / MuMech - MechJeb - Autopilot - Historical thread


r4m0n

Recommended Posts

Kudo's on the Hohmann transfers, they've tightened up tremendously. I use them for vessel rendezvous and I have to be careful about when I burn for velocity match or I'll collide. I've already taken out solar panels.

Link to comment
Share on other sites

Are there any issues with the auto landing since the last 4-5 patches? I hate to be "that guy", but its getting worse and worse for me. MJ ist constantly misscalculating his burns (crashing into the ground), waits to the very last second for burns (crashing into the ground), tries to to burn corrections and keeps hanging at the last 0-1% (doing nothing noticeble, and consequently crashing into the ground). The good thing is, that i`ve gotten used to it (at least with non-stock engines, ksp interstellar engines for instance is something that MJ completely doesnt understand most of the time), and therefore rather good for my needs to land myself in most situations. But i always liked the perfectionism of MJ autolanding and the offer to handle lots of standard situations, and most of all "precision landings" near bases. What happened? In what cases can i trust MJ, and in which i can not?

Another thing... picked a landing spot on Moho, MJ makes his burns, Auto-Warp is on, and now MJ hangs with the Statusmessage "Deorbit burn complete: waiting for the right moment to start braking". For 20 minutes?

Again ... i love MJ, and everything its thaught me. Thanx for that. But landing with anything more complex than the standard stocklander from career start is for me right now some kind of russian roulette. Is there any development going on?

Link to comment
Share on other sites

@smart013 -- I'm sure the LGAP will get some attention on these issues. Heaven knows I've mentioned some problems with it a time or two...

Sarbian, one consistent thing I've noticed (at least with Kerbin landings) is the map view "projected target" doesn't get updated unless I've aborted MJ to do the final burn manually. If MJ is doing it, the target indicator (the blue one) doesn't move once it shows up in map view. It also pops itself onto the map far to the west of the target, instead of the intersect point of the orbit with the surface. MJ seems to take the data from the indicator when plotting out the last part of the burn, which is why it wants to burn pro-grade. Maybe this is at the root of the problem with all of these landing modes, whether at Kerbin or Moho, or some other body.

Side note - Feature request: For better rendezvous, could we get an option about when MJ starts the velocity match burn? Perhaps something that allows MJ to begin the node burn earlier than the projected start time (typically half the burn duration ahead of the node). I've had occasions where I've screamed past asteroids at multi-hundreds of m/s at closures of less than 300 m, simply because the burn started later than it should (happens a lot) or because my ship's engine is a low-TWR type that takes minutes to bring the velocity down (instead of seconds). A checkbox to begin the burn at the full-burn-duration ahead of the node would be helpful; a slider allowing us to vary the time between half-duration and full-duration ahead of the node would be great.

Link to comment
Share on other sites

One little detail, not really a bug, but maybe something to investigate. If you activate "Smooth throttle" in the Utilities a lot of burns for rendezvous and such, and of course landings go horrible wrong ... ok, thats to expect. When using MJ and do such things i regularly uncheck this otherwise great feature. There could of course be some kind of security check before starting precision maneuvers, or do i get this wrong?

(no, this was not the cause of my landing rant above, i learned these error of my own a while ago :)

Link to comment
Share on other sites

Is there any way to get MJ to reliably produce the same [censored] node time for interplanetary transfers? It's almost impossible to plan ahead when it randomly chooses departures anywhere from 200 days away to 1 year 180 days and almost every figure in between.

Link to comment
Share on other sites

I think Sarbian may be taking a break from Mechjeb. Considering how much work he's put in and how long he's been at it, I don't blame him.

Aye ! I'm enjoying others games ATM :) But I'm still around and I'll resume work when .24 is here (64bit!)

Is there any way to get MJ to reliably produce the same [censored] node time for interplanetary transfers? It's almost impossible to plan ahead when it randomly chooses departures anywhere from 200 days away to 1 year 180 days and almost every figure in between.

You [censored] need to [censored] the [censored] config file. Savy ?

Link to comment
Share on other sites

Aye ! I'm enjoying others games ATM :) But I'm still around and I'll resume work when .24 is here (64bit!)

I guess I need to catch up on my Kerbal News...is this true? .24 will include x64 support for Windows?

Link to comment
Share on other sites

Aye ! I'm enjoying others games ATM :) But I'm still around and I'll resume work when .24 is here (64bit!)

You mean [gasp!] there are other games besides KSP? :confused::D

Actually, I understand fully. My other activity when I'm not screaming toward asteroids trying to slow down, or desperately trying to reach one with my remaining dV, is building RC airplanes from scratch. Quite therapeutic...

Link to comment
Share on other sites

You [censored] need to [censored] the [censored] config file. Savy ?

OK, I deserved that. :) Seriously, if there's a config switch to force it to select the same transfer node time, I'd appreciate knowing what it is.

Link to comment
Share on other sites

Having played with the DAP for a while, I've noticed that if I try to engage too closely, DAP goes a bit nuts. But if I start, say, 45m directly in front of or at an angle to the target, DAP is all sweet and friendly (most of the time, not all of the time...).

Is there an "optimum minimal distance" and an "optimal maximum approach angle" for setting up the DAP? Is this calculated for each set of docking vessels, or is there a "hard-coded" set of values in DAP?

Link to comment
Share on other sites

Having played with the DAP for a while, I've noticed that if I try to engage too closely, DAP goes a bit nuts. But if I start, say, 45m directly in front of or at an angle to the target, DAP is all sweet and friendly (most of the time, not all of the time...).

Is there an "optimum minimal distance" and an "optimal maximum approach angle" for setting up the DAP? Is this calculated for each set of docking vessels, or is there a "hard-coded" set of values in DAP?

I dont think its range related. I think it just randomly flakes out. cancel and reengage if it doesnt immediately make sensible moves

Link to comment
Share on other sites

Can someone tell me why my craft just keeps going around and around the other craft in docking autopilot?
I dont think its range related. I think it just randomly flakes out. cancel and reengage if it doesnt immediately make sensible moves

Just so....

Link to comment
Share on other sites

Could it be possible to make/use fixed number of digits for orbital data please (AP/PE time) ?

When timewarp (even low), usage of 1 then 2 then 1 digits for time values make the full text "moving" in a very => :confused: way

EDIT: version 2.2.1.238

Edited by Justin Kerbice
added version
Link to comment
Share on other sites

Since mechjeb cases attached to a vessel count as controlpods... And with the perspective that recovering parts in .24 is a good thing... What kind of problems would i get if for example i put mechjebcases at boosters, so that they can be controlled and brought down on parachutes after separation? Part count? Memory? Problems with more that one instance of MJ? What would happen with a mj equipped booster after reaching the 2.3km distance while i keep controlling the mainvessel during launch?

Ok, the main question is about performance, maybe i should test the distance problem myself. Or has someone already the answer?

Link to comment
Share on other sites

Since mechjeb cases attached to a vessel count as controlpods... And with the perspective that recovering parts in .24 is a good thing... What kind of problems would i get if for example i put mechjebcases at boosters, so that they can be controlled and brought down on parachutes after separation? Part count? Memory? Problems with more that one instance of MJ? What would happen with a mj equipped booster after reaching the 2.3km distance while i keep controlling the mainvessel during launch?

Ok, the main question is about performance, maybe i should test the distance problem myself. Or has someone already the answer?

You would get no problems - Mechjeb is fine with having more than one instance present, only one will run. However, once anything (your boosters or your main craft) is over 2.5km away and at more than 0.01 atm of pressure, it'll be deleted, just like it is now.

Link to comment
Share on other sites

You would get no problems - Mechjeb is fine with having more than one instance present, only one will run. However, once anything (your boosters or your main craft) is over 2.5km away and at more than 0.01 atm of pressure, it'll be deleted, just like it is now.

Ah, the pressure criteria is something that i dont knew about till now. Thanks!

Link to comment
Share on other sites

Did you select control from here on the docking port you want to use

Yes I don't think it can be engaged any other way

Just so....

I tried that, numerous times actually. I spent roughly an hour last night trying to dock a large fuel tank to my waiting space craft. Turning autopilot on and off repeatedly, eventually it go so close and then started going silly so I quickly turned it off and nudged in myself randomly.

But to get to that point it would literally circle around the target craft randomly. Should I take a video next time and post?

Is the other craft stable? (I.E. has some form of attitude control engaged?)

Um, the target craft I had on a 90,90,0 surface smart sas alignment if that's what you mean? Is that what I am suppose to do? There is no tutorial on docking autopilot in the manual so I might just be doing it wrong. For example I might be engaging too far out, I usually engage autopilot as soon as I can target the target docking port (so can be like 150m etc).

What I think might be happening just from looking at it, is that docking autopilot is using too much RCS thrust, like it hauls ass haha and then when it decides it needs to stop its going too fast in one direction and the RCS cant stop it in time so its flawed and starts again or something. Like it look visually like its maneuvering at full throttle and cant correct fast enough.

I have tried the speed limiter but doesn't seem to make much difference.

Don't get me wrong, great mod lol and I'm totally thankful, this part just frustrates me a little lol. Sometimes it works flawlessly.

Link to comment
Share on other sites

Guest
This topic is now closed to further replies.
×
×
  • Create New...