Test of continuing KER developments [0.24.2]

[Darth Lazarus]

i was curious about RAPIER's, cause my recent Soyuz replica use them .... works very fine ! thanks for teh code RAPIER's are fixed

[dlrk]

All looks good.

[Mike the Mechanic]

Great job! This tool saves time and lives Just one small thing, the ion engine does not work properly yet.

Edited April 3, 2014 by Mike the Mechanic
wrong info ....

[Padishar]

Great job! This tool saves time and lives Just one small thing, the ion engine does not work properly yet.

In what way does the ion not work? Are you using the latest version downloaded from the first post? Can you provide a craft and a description of what you think is wrong?

[Mike the Mechanic]

Ignore my post, it works perfectly - I messed up some staging ... Sorry

[Mike the Mechanic]

However, there is a problem I encountered:

Javascript is disabled. View full album

This happened when I was trying to remove the leg. The error occurs every time I load the ship. Hope this helps to make the plugin even better

http://imgur.com/Yvh9kEu

[Mike the Mechanic]

The error is reproducible, the craft file is here:

https://www.dropbox.com/s/h6t7dsiu8uvd7r8/Xenon%20Impulse%20II.craft

Edited April 3, 2014 by Mike the Mechanic
I suck at forums ...

[Padishar]

Excellent, I've only ever seen my in-game exception reporting working when deliberately provoked in the code... As for the actual bug, are all the parts still attached or has the editor done something weird? This looks like it might be a core game bug but I'll check it out if you can upload the craft file...

Edit: cheers... I'll take a look now...

Edited April 3, 2014 by Padishar

[Padishar]

There is something very wrong with your craft. You will notice that the staging is showing 4 engines but only 3 are "visible". One is clipped inside another. Also there is a complete copy of the leg with it (so 4 legs too) and if I try to launch the craft only has 2 legs and falls over.

Did you have clipping enabled when you built it? This looks like a case of the VAB getting badly confused and breaking your craft. There is no way I know of that KER could have caused this.

I'll try and work out why KER is giving an exception and make the message more friendly/helpful but that won't help with your craft. You will probably need to rebuild it from scratch...

[Mike the Mechanic]

Clipping was not enabled, but I used the symmetry function excessively - I may have broken the VAB. Yes, I realized that I have to rebuild it, but I'm glad it's not the engineer in this case. I could not determine whether anything is clipped or not, because the game breaks down as soon as I try to remove anything ...

[AlexisBV]

I didn't see an official release yet on SpacePort... if not already done, can you please please fix the typo on the RDV tab, where it says "Orbital Preiod"? Every time I look at it, it feels like it's laughing at me...

[Mischief]

Removed MechJeb due to sudden bursts of lag in the VAB after 23.5, added engineer redux with the new patch represented here. No lag in VAB, no issues with any of the ships I've produced, launched two missions since its installation using a mix of both original stock and ARM engines and boosters and had no issues, deltaV estimates were completely accurate.

[xcq]

I've tried the fix with the DLLs dated to Apr 03, 14:09. Mostly no issues, but I have noticed two things so far:

1. In comparison with the original Engineer, it seems the Thrust displays have gotten a little bit jerky. I guess it's more performant that way, it just doesn't look as... fluent, I guess.

2. Turbojets seem to produce wrong Thrust information which results in wrong TWRs which results in critical mission failure. At first this happend to me with a B9 part, but I just tested the stock Turbojet, see here, the discrepancy is almost factor 2. Oddly enough the Turbojets say max thrust 225kN in the tooltip and basic jet engines do not seem to be affected.

[Padishar]

I didn't see an official release yet on SpacePort... if not already done, can you please please fix the typo on the RDV tab, where it says "Orbital Preiod"? Every time I look at it, it feels like it's laughing at me...

This has been fixed in this test version. I'm not sure when (or even if) Cybutek will be releasing an updated version based on 0.6.2.3 for 0.23.5 as he has been working on a major new version that will become v1.0. He is currently waiting on the changes I have been doing to the deltaV calculation code so it can be integrated into the new version. In the meantime I would recommend that everyone running 0.6.2.3 switch to using this test version as a large number of improvements have been made.

I've tried the fix with the DLLs dated to Apr 03, 14:09. Mostly no issues, but I have noticed two things so far:

1. In comparison with the original Engineer, it seems the Thrust displays have gotten a little bit jerky. I guess it's more performant that way, it just doesn't look as... fluent, I guess.

2. Turbojets seem to produce wrong Thrust information which results in wrong TWRs which results in critical mission failure. At first this happend to me with a B9 part, but I just tested the stock Turbojet, see here, the discrepancy is almost factor 2. Oddly enough the Turbojets say max thrust 225kN in the tooltip and basic jet engines do not seem to be affected.

1. When you say it is jerky, do you mean it updates too fast? The performance of the new simulation code should be better though it does now split the work between the main thread and a background thread and the code in the main thread needs to create more complex data structures for use by the background thread to avoid various nasty race conditions with the core game code (e.g. the vessel changing while the background thread is running the simulation). I have also changed the way it decides how long to wait between runs of the simulation. The old code runs the simulation and then waits for at least 10 times as long as the simulation took to run before running it again. The new code currently tries to maintain approx. 10 updates per second (which is a bit high really) and usually seems to manage this. I want to slow it down to between 2 and 4 updates per second which should be fine if that is what you mean.

2. I haven't explicitly done anything to the handling of jets though I am aware that they may not be giving correct values. The probable cause is that the thrust generated by jets depends on the speed the plane is flying and this can't be easily simulated in the VAB/SPH. Perhaps there should be a way of adjusting the speed used similar to how you can switch between vacuum and atmosphere. This could either have a settings value to set the speed or maybe even have a slider in the build UI. I don't really want to show another field for high-speed thrust/TWR as I want to add a "max TWR" field anyway (to let people see the TWR at stage burnout) and the window is quite crowded already.

[xcq]

1. When you say it is jerky, do you mean it updates too fast? The performance of the new simulation code should be better though it does now split the work between the main thread and a background thread and the code in the main thread needs to create more complex data structures for use by the background thread to avoid various nasty race conditions with the core game code (e.g. the vessel changing while the background thread is running the simulation). I have also changed the way it decides how long to wait between runs of the simulation. The old code runs the simulation and then waits for at least 10 times as long as the simulation took to run before running it again. The new code currently tries to maintain approx. 10 updates per second (which is a bit high really) and usually seems to manage this. I want to slow it down to between 2 and 4 updates per second which should be fine if that is what you mean.

On the contrary, I find it updating too slow. :-) Mainly because it is inconsistent: The orbital part seems to go fluent and change each value with almost each frame (I guess something between 30 and 60 ups?), while I can see the 10 updates / second in the vessel part and it seems like it is stuttering to me. Might just be personal preference.

2. I haven't explicitly done anything to the handling of jets though I am aware that they may not be giving correct values. The probable cause is that the thrust generated by jets depends on the speed the plane is flying and this can't be easily simulated in the VAB/SPH. Perhaps there should be a way of adjusting the speed used similar to how you can switch between vacuum and atmosphere. This could either have a settings value to set the speed or maybe even have a slider in the build UI. I don't really want to show another field for high-speed thrust/TWR as I want to add a "max TWR" field anyway (to let people see the TWR at stage burnout) and the window is quite crowded already.

I can live with the fact that VAB/SPH values are not overly specific, because I think pre-flight Engineer is just a tool to estimate your performance. Your ideas sound good. However, just to be clear: What I'm pointing out is that there's a difference between right-clicking the turbojet engine mid-flight and the thrust displayed in the Flight Engineer, see the screenshot from my previous post. The Flight Engineer claims it delivers e.g. 220kN when in fact the engine tooltip says there is only half of that amount, e.g. 110kN.

[Padishar]

On the contrary, I find it updating too slow. :-) Mainly because it is inconsistent: The orbital part seems to go fluent and change each value with almost each frame (I guess something between 30 and 60 ups?), while I can see the 10 updates / second in the vessel part and it seems like it is stuttering to me. Might just be personal preference.

The orbital pane doesn't need to do much in the way of calculations so it can easily update every time the plugin is asked to redraw the window. The vessel tab has to run complex code that simulates the complete burning of all the stages of the vessel so it is deliberately throttled. I will have another play with the timings and see if basically removing the delay has any bad effect on the game. We certainly don't want to go back to the days when the vessel pane had a noticeable impact on fps...

I can live with the fact that VAB/SPH values are not overly specific, because I think pre-flight Engineer is just a tool to estimate your performance. Your ideas sound good. However, just to be clear: What I'm pointing out is that there's a difference between right-clicking the turbojet engine mid-flight and the thrust displayed in the Flight Engineer, see the screenshot from my previous post. The Flight Engineer claims it delivers e.g. 220kN when in fact the engine tooltip says there is only half of that amount, e.g. 110kN.

Yes, I understand. The values shown in KER are the max the engine is capable of. The value in the tweakable window takes into account how fast the vessel is moving. A stationary turbojet produces much less power than a fast moving one. All that should be required is that the values in KER are adjusted according to the velocityCurve of the engine. The vessel simulation code currently has inputs for gravity and atmospheric pressure so it shouldn't be difficult to add a velocity input too.

[TK421d]

not sure if its just me, also maybe it could be from the new toolbar plugin from a few days ago, i dont know.

anyway: the data window is fine, but the icon on the toolbar rapidly flashes on/off and is (nearly) impossible to toggle. both in assembly and flight

[hawk_za]

works awesomely for me thank you for your continued support to this mod without it i would still be only getting into orbit and not much further it has added allot to the game.

[Padishar]

not sure if its just me, also maybe it could be from the new toolbar plugin from a few days ago, i dont know.

anyway: the data window is fine, but the icon on the toolbar rapidly flashes on/off and is (nearly) impossible to toggle. both in assembly and flight

I'll check this out. I did use the toolbar button in the VAB to hide the window yesterday and it seemed to work fine but I don't think I've actually updated the toolbar plugin since 0.23.5 so perhaps there is some weird conflict with the latest version.

[TK421d]

I'll check this out. I did use the toolbar button in the VAB to hide the window yesterday and it seemed to work fine but I don't think I've actually updated the toolbar plugin since 0.23.5 so perhaps there is some weird conflict with the latest version.

oddly enough, today its slightly more stable, but its still flashing. no changes on my end so i'm mystified about this.

also i only just noticed since i usually display it that way anyway, while all the delta-v etc are working fine, it is stuck in "show all stages mode" whether its pressed or not.

[Arrowmaster]

Using MFT together with AIES and KW. Both of those part mods have nosecones and MFT lets the nosecones store fuel. Starting with empty nosecones and adding fuel to them, KER telling me I have lost 999m/s of dV while MJ is saying I've gained 46m/s of dV.

Could this be because the nosecones are "module = Strut" while normal tanks are "module = Part".

[AndersW]

I found something that causes Kerbal Engineer to flip out. I was using the new 2.5 meter engine that comes with its own fuel as a booster. But I found that after they were dropped I didn't have enough thrust and started losing speed. I wanted a way to get the boosters and center stage to share fuel and act like a single stage. To do this I ran fuel lines from the boosters to the center, and from the center to the boosters. It worked, but I noticed that my total delta-v was Increasing as I flew. I went back and checked, and found that between the VAB and the launchpad Kerbal Engineer lost almost 2000m/s of delta-v.

[Padishar]

Using MFT together with AIES and KW. Both of those part mods have nosecones and MFT lets the nosecones store fuel. Starting with empty nosecones and adding fuel to them, KER telling me I have lost 999m/s of dV while MJ is saying I've gained 46m/s of dV.

Could this be because the nosecones are "module = Strut" while normal tanks are "module = Part".

Whatever the actual reason it certainly does sound like the parts aren't providing fuel for the simulation. I'll need to take a look with the logging build...

I found something that causes Kerbal Engineer to flip out. I was using the new 2.5 meter engine that comes with its own fuel as a booster. But I found that after they were dropped I didn't have enough thrust and started losing speed. I wanted a way to get the boosters and center stage to share fuel and act like a single stage. To do this I ran fuel lines from the boosters to the center, and from the center to the boosters. It worked, but I noticed that my total delta-v was Increasing as I flew. I went back and checked, and found that between the VAB and the launchpad Kerbal Engineer lost almost 2000m/s of delta-v.

From your description you may just be seeing the difference between atmospheric and vacuum deltaV. Can you post your craft file?

[Godot]

Works fine for me, maybe with one exception ...

I had an Apollo style Capsule / Lander setup for a Mun landing,

both (i.e. capsule and lander) with their own KER Flight engineer.

When I separated the lander from the main spaceship, it didn´t show any dV for the landers engines/fuel tanks.

(I might add that the configuration was a little bit faulty: Planned was, that a central tank would supply 4 outlying engines (who also had heir own fuel tanks) ... but I foirgot the fuel lines leading to the outlying engine/fuel tank combos ... nevertheless, KER should have shown me the dV for the fuel in the outlying tanks, as they, correctly, supplied the engines)

Edited April 5, 2014 by Godot

[xcorps]

Just reinstalled after the KSP update caused many failures with KER.

Everything looks to be working to me, I don't do maths so I can't tell you if the calcs are correct, but at least it's making calcs. It's getting the staging correctly for the most part, but it still combines two separate stage engines if there is no separator. You can see the TWR and Time, just not the Dv. The Dv gets combined into the next stage up.

IE, if you place a stage 6 SRB on a stage 5 SRB then tweak the thrust limiter on stage 6, you see the TWR and time on stage 6 but no Dv. The Dv gets added to stage 5. I don't remember if it was doing that on the last version.

Thanks very much for keeping this great utility running.

Edited April 5, 2014 by xcorps