[1.4] StageRecovery - Recover Funds+ from Dropped Stages - v1.8.0 (March 11, 2018)

[magico13]

10 hours ago, Loren Pechtel said:

I can see the point about letting you know if it was recoverable but the way the message is worded it looks like it tried it.  As I said, it's only a message bug, the code did what it should have.

Question:  If I included some fuel on board could SR use it for a retroburn to avoid burning up?

Not yet. At some point that will be an option. Maybe for 1.2, but it might wait for when I rewrite SR.

[Alexology]

I installed the 1.2 version (https://github.com/magico13/StageRecovery/releases) but it does not seem to be working. The button is appearing in the SPH and it does all of the calculations there but I am not getting a message in flight when the debris leaves render range. Can I send you anything to help? I understand of course we are pre-1.2 here and it was unlikely to work fully but if my save file is useful I am more than happy to send it to you. 

[linuxgurugamer]

38 minutes ago, Alexology said:

I installed the 1.2 version (https://github.com/magico13/StageRecovery/releases) but it does not seem to be working. The button is appearing in the SPH and it does all of the calculations there but I am not getting a message in flight when the debris leaves render range. Can I send you anything to help? I understand of course we are pre-1.2 here and it was unlikely to work fully but if my save file is useful I am more than happy to send it to you. 

That was compiled 3 weeks ago.  there have been a number of changes, it's unlikely to work without at least a recompile

[Alexology]

22 minutes ago, linuxgurugamer said:

That was compiled 3 weeks ago.  there have been a number of changes, it's unlikely to work without at least a recompile

I expected as much, however I can send any data that would be helpful. Just say the word. 

[linuxgurugamer]

After looking at the code and the release notes, I can state with certainty that the version on Github won't work.

However, I did some playing, and have a compiled DLL here if you would like to try it:

https://www.dropbox.com/s/bzf7k9zkclp7piu/stageRecoveryDll.zip?dl=0

@magico13 All the changes were in RecoveryItem.cs, I found the following note in the Build 1548 notes and used it to make the fixes:

Quote

• Fix ProtoPartResourceSnapshot to not rely on strings. This means modders: you should use the double amount/maxAmount in it, not the resourceValues node (which is now inaccessible).

I had a problem getting a clone of the repo, so I don't think I can (yet) create a PR for you.  I'll keep working on it, but I can always send you the file.

Let me know

Going to do a PR for you

Edited October 9, 2016 by linuxgurugamer

[magico13]

9 hours ago, linuxgurugamer said:

After looking at the code and the release notes, I can state with certainty that the version on Github won't work.

However, I did some playing, and have a compiled DLL here if you would like to try it:

https://www.dropbox.com/s/bzf7k9zkclp7piu/stageRecoveryDll.zip?dl=0

@magico13 All the changes were in RecoveryItem.cs, I found the following note in the Build 1548 notes and used it to make the fixes:

I had a problem getting a clone of the repo, so I don't think I can (yet) create a PR for you.  I'll keep working on it, but I can always send you the file.

Let me know

Going to do a PR for you

Thanks for that. PR pulled in and I've got a few other minor changes I want to get in and then I'll post a new build. Most likely later tonight.

 

And done (for tonight at least). A few minor changes, and pulled in @linuxgurugamer's PR. I've got a few other things I might try to get in before 1.2 drops officially, but I also have to switch over to KCT at some point.

Edited October 10, 2016 by magico13

[magico13]

I haven't tested the 1.2 pre-release build with the 1.2 release, but it might work. I'm planning on getting a full release out tonight after I test RealChute and FAR (assuming they're up to date). I've replaced the code I used for Stock velocity estimates, but it's not matching up as nicely as I had hoped with actual experiments. Stock may still be an approximation in this release, but it'll hopefully be better than what I was using before.

[Scuwr]

6 hours ago, magico13 said:

I haven't tested the 1.2 pre-release build with the 1.2 release, but it might work. I'm planning on getting a full release out tonight after I test RealChute and FAR (assuming they're up to date). I've replaced the code I used for Stock velocity estimates, but it's not matching up as nicely as I had hoped with actual experiments. Stock may still be an approximation in this release, but it'll hopefully be better than what I was using before.

I just checked on the forum for FAR and it doesn't seem to have been updated yet..

[JeffreyCor]

RealChute is also in progress for a compatibility update. Traditionally it is supposed to be updated after SR, so that it can break SR and make a mess of things

[magico13]

2 hours ago, Scuwr said:

I just checked on the forum for FAR and it doesn't seem to have been updated yet..

 

1 hour ago, JeffreyCor said:

RealChute is also in progress for a compatibility update. Traditionally it is supposed to be updated after SR, so that it can break SR and make a mess of things

I may wait until they're released to officially update then. The timing is a bit bad because I might not get a chance to test for a few days.

[JeffreyCor]

I will keep an eye on the status. Starwaster is working on the compatibility update and as soon as he makes something available I will grab it. As always I'll be ready to have you break a save / test the status of the update

[Tybot]

18 hours ago, magico13 said:

I may wait until they're released to officially update then. The timing is a bit bad because I might not get a chance to test for a few days.

Is FAR/RS updating a hurdle for releasing or a prerequisite to recompiling?

[magico13]

36 minutes ago, Tybot said:

Is FAR/RS updating a hurdle for releasing or a prerequisite to recompiling?

Not necessary for recompiling since I use reflection to access the things I need, but they have a frequent tendency to change everything I access so that StageRecovery no longer works with it, meaning I have to redesign and rebuild. I'd rather put the build out after I get to test them, since adding StageRecovery and then adding RealChute/FAR later could actually break things.

The latest pre-release probably works with the 1.2 release. It doesn't have the latest version of the Stock parachute handler though.

One day I'll break all of this out into separate .dlls so that I only have to update one thing without breaking things.

Edited October 13, 2016 by magico13

[JeffreyCor]

Testing of the current dev version of SR seems to be working very well with the stock parachute in the 1.2 release.

 

[Trevisan]

11 hours ago, JeffreyCor said:

Testing of the current dev version of SR seems to be working very well with the stock parachute in the 1.2 release.

 

How we can use this without the toolbar mod? Tried to use it but nothing shows up in game...

[JeffreyCor]

9 hours ago, Trevisan said:

How we can use this without the toolbar mod? Tried to use it but nothing shows up in game...

It has built in stock toolbar support. Toolbar mod hasn't been required for a very long time.

 

RealChute was just updated so I know what I'll be testing. Will let you all know how this goes

EDIT: To my pleasant surprise we have joy joy feelings! SR seems to be playing nicely with RealChute

Edited October 14, 2016 by JeffreyCor

[MaxwellsDemon]

Hope it all goes smoothly.   Campaign play just isn't the same without KCT and SR!  

Edited October 19, 2016 by MaxwellsDemon

[magico13]

On 10/14/2016 at 11:33 AM, JeffreyCor said:

RealChute was just updated so I know what I'll be testing. Will let you all know how this goes

EDIT: To my pleasant surprise we have joy joy feelings! SR seems to be playing nicely with RealChute

Darn, I missed this update somehow. Alright, I'll see about a StageRecovery release tonight and then try to do something with KCT this weekend. At some point I have to look at the modlets as well.

[magico13]

Updated! Hopefully nothing is totally broken!

[Apollo13]

Did not work for me in 1.2.

I just tested the 1.2 version with a simple rocket.  I attached a parachute to the probe and to its booster(Rt-5 Flea).  Launched.  Separated the booster stage after burnout, and watched probe and spent booster float back to the ground under parachutes.  SR did not report that the booster was recovered.

The booster speed under the parachute was 4.8 m/s, so it survived. I actually launched it again and switched to it to verify it was not destroyed upon landing.  SR did not report the probe landed safely

EDIT:  on a later launch with bigger, more complex vessel, the spent boosters did not have parachutes.  SR reported their destruction.

Edited October 21, 2016 by Apollo13

[Loren Pechtel]

3 hours ago, Apollo13 said:

Did not work for me in 1.2.

I just tested the 1.2 version with a simple rocket.  I attached a parachute to the probe and to its booster(Rt-5 Flea).  Launched.  Separated the booster stage after burnout, and watched probe and spent booster float back to the ground under parachutes.  SR did not report that the booster was recovered.

The booster speed under the parachute was 4.8 m/s, so it survived. I actually launched it again and switched to it to verify it was not destroyed upon landing.  SR did not report the probe landed safely

EDIT:  on a later launch with bigger, more complex vessel, the spent boosters did not have parachutes.  SR reported their destruction.

With something that small you might not get outside physics range in order to trigger the recovery. 

[Apollo13]

@Loren, you might be onto something.  My probe and the spent booster landed less than 200 meters from each other.

This is one of my favorite mods from previous KSP versions; really been missing it in 1.2 pre-release and 1.2

Edited October 21, 2016 by Apollo13

[Loren Pechtel]

6 minutes ago, Apollo13 said:

@Loren, you might be onto something.  My probe and the spent booster landed less than 200 meters from each other.

This is one of my favorite mods from previous KSP versions; really been missing it in 1.2 pre-release and 1.2

That certainly would cause your failure.  I've given up on recovering the booster used in space and jettisoned during re-entry because of this--it usually stays close enough that it goes splat rather than recover.  It's not worth the chutes for the low chance they actually do something.

Note, also, that while boosters discarded high enough up merely need the necessary equipment on board a booster that falls off early actually needs to be able to properly land.  Stage Recovery assumes the chutes work as desired but if it's in physics range you actually have to make sure they work.  You need proper activation criteria and you must stage them properly.  Note that it doesn't matter the separation at staging, but rather that your rocket gets far enough away before they go splat.

[magico13]

4 minutes ago, Loren Pechtel said:

Note that it doesn't matter the separation at staging, but rather that your rocket gets far enough away before they go splat.

Yeah, it's an issue I've considered fixing in the past by monitoring all loaded vessels that aren't the currently active one and triggering their chutes when they get low. The reason I haven't done it is because you can set restrictions on the pressure needed for pre-deployment, so you can stage and then not worry about the chutes. I thought 1.2 added some additional logic to prevent them from deploying when going too fast, but I'm not positive about that. With those stock features all that's required is a bit of planning, so I haven't felt a pressing need to implement an alternative. RealChute of course has for a long time had excellent features, such as only deploying when the vertical velocity is negative, that make this pretty trivial to handle as a user.

[Loren Pechtel]

32 minutes ago, magico13 said:

Yeah, it's an issue I've considered fixing in the past by monitoring all loaded vessels that aren't the currently active one and triggering their chutes when they get low. The reason I haven't done it is because you can set restrictions on the pressure needed for pre-deployment, so you can stage and then not worry about the chutes. I thought 1.2 added some additional logic to prevent them from deploying when going too fast, but I'm not positive about that. With those stock features all that's required is a bit of planning, so I haven't felt a pressing need to implement an alternative. RealChute of course has for a long time had excellent features, such as only deploying when the vertical velocity is negative, that make this pretty trivial to handle as a user.

Yeah, with RealChute it's not a problem at all once you have upgraded the VAB.  It's just something you have to be aware of.  (With an unupgraded VAB it can be a pain to set the chutes properly.)

Before tackling something like this how about my dream list for SR:

Make a part equivalent to a probe core, any rocket bit for which the only guidance is such a part attempts recovery after 30 seconds regardless of location.  (The delay to let it fall away from your rocket, it would look wrong if the boosters vanished immediately.)

Calculate the remaining delta-v.  This is allocated:

1)  Whatever is needed to slow it from it's falling speed to the fully safe touchdown speed.

2)  If within Kerbin's SOI and heading out:  To lower apoapsis to the SOI.

3)  If within Kerbin's SOI:  To lower periapsis to 40km.  Figure the burn occurs at apoapsis.

4)  If not within Kerbin's SOI, to pass 40km from the surface of Kerbin.

5)  The remainder is used to shed velocity down to 2000 m/s when it's 70km up.

And for the icing on the cake, if anything's left after that, try to guide it as close to KSC as possible.  That's probably harder than the rest of it combined, though.