[KSP 1.9.0] [x] Science! Continued - KSP Science report and checklist (V5.23)

[Armor9]

No there is nothing with that name.  I have event search the files and nothing comes up.  Under X or even using the word Science. 

It is very odd.

[linuxgurugamer]

2 hours ago, Armor9 said:

When I am trying to install this via CKAN I am getting the error listed below.  I currently do not have [x]science installed.  Any ideas how to fix?  Thanks.

 

About to install...

 * [x] Science! Continued 5.22 (cached)

The following inconsistencies were found:
* xScienceContinued wishes to install GameData/[x] Science!/[x] Science!.dll, but this file is registered to xScience
* xScienceContinued wishes to install GameData/[x] Science!/[x] Science!.version, but this file is registered to xScience
* xScienceContinued wishes to install GameData/[x] Science!/bubbles.wav, but this file is registered to xScience
* xScienceContinued wishes to install GameData/[x] Science!/CHANGES.md, but this file is registered to xScience
* xScienceContinued wishes to install GameData/[x] Science!/LICENSE.txt, but this file is registered to xScience
* xScienceContinued wishes to install GameData/[x] Science!/MiniAVC.dll, but this file is registered to xScience
* xScienceContinued wishes to install GameData/[x] Science!/PluginData/[x] Science!/settings.cfg, but this file is registered to xScience
* xScienceContinued wishes to install GameData/[x] Science!/README.md, but this file is registered to xScience
* xScienceContinued wishes to install GameData/[x] Science!/science.cfg, but this file is registered to xScience
Error during installation!
An unknown error occurred, please try again!

 

What version of KSP are you running?

Also, could you start KSP, get to the main menu and exit, and then provide a link to the log file?

[Armor9]

I am running the latest version of KSP with both DLC. 

Where do I get the link to the log file? 

[Jognt]

Outputlog.txt is in your user folder. Go to %appdata%, one folder up, then LocalLow, Squad, Kerbalspaceprogram. KSP.log should be in a folder in your game directory.

Edited June 5, 2019 by Jognt

[linuxgurugamer]

1 hour ago, Armor9 said:

I am running the latest version of KSP with both DLC. 

Where do I get the link to the log file? 

My signature has a link to complete instructions 

[Armor9]

Thanks.  The log is uploaded.

https://www.dropbox.com/s/22xp9dksf1l2r0k/output_log.txt?dl=0

[Tekaoh]

@Flupster I submitted a PR to remove the filter for Kerbin Splashed Down science in Highlands and Mountains. These were apparently filtered out because they were (presumed to be) impossible, but as @casper88 and @BlackHat mentioned on page 1, these are indeed possible and therefore probably don't need to be filtered out.

As a side note, I wonder if we can look into compatibility with new DLC science like Mun stones and stuff. (I don't think I've even discovered all the new DLC science yet.)

Edited June 6, 2019 by Tekaoh

[Armor9]

Anyone have any luck seeing the issue after reviewing the log?  I'm baffled. 

[Jognt]

9 hours ago, Armor9 said:

Anyone have any luck seeing the issue after reviewing the log?  I'm baffled. 

No idea sorry
All I can say is the usual base advice: Clear caches, reinstall CKAN, reboot PC, sacrifice a turnip at sunset, that sort of thing. (actually, clearing the CKAN cache might work? Maybe?)

[drake127]

Hi,

during my gameplay I noticed that I was unable to complete an atmosphere analysis experiment without many many tries even when [x] Science! reports green full bar. After long investigation I found out that science gain in [x] Science! is calculated incorrectly and in reality it's nearly impossible to reach full bar. E.g. after six collection of atmosphere analysis on KSC runway you are still missing 0.68 science points (from 7.2 total) and gaining just 0.05 per collection and declining.

I propose two changes:

1. Fix the formula for calculating collected science onboard. It prevents filling the bar completely when there's plenty of science left.
2. The [x] Science! considers an experiment to be completed (collected) when there is less than 0.1 science to obtain. That's almost impossible to get for most recurring experiments, so I would also consider an experiment completed when you would recover at least 0.1 science points during single run.

What do you think? I can create PR if needed.

Edited June 9, 2019 by drake127

[Jognt]

On 6/9/2019 at 2:26 PM, drake127 said:

Hi,

during my gameplay I noticed that I was unable to complete an atmosphere analysis experiment without many many tries even when [x] Science! reports green full bar. After long investigation I found out that science gain in [x] Science! is calculated incorrectly and in reality it's nearly impossible to reach full bar. E.g. after six collection of atmosphere analysis on KSC runway you are still missing 0.68 science points (from 7.2 total) and gaining just 0.05 per collection and declining.

I propose two changes:

1. Fix the formula for calculating collected science onboard. It prevents filling the bar completely when there's plenty of science left.
2. The [x] Science! considers an experiment to be completed (collected) when there is less than 0.1 science to obtain. That's almost impossible to get for most recurring experiments, so I would also consider an experiment completed when you would recover at least 0.1 science points during single run.

What do you think? I can create PR if needed.

 

It looks like Squad changed Science with the most recent update. It's no longer possible to cap science the way we used to so if you're running an up-to-date client, could be that that's what you're experiencing.

[drake127]

That's right! I was wondering why I have some recurring experiments fully completed in my save file.

Now if we just fix the equation, we are not able to complete any (recurring) experiment. That makes UI quite unfriendly when filtering completed experiments. So where should we draw a line?

I was thinking about 90 % of science, 0.1 science earned per measurement, 1 % science earned per measurement or something similar? There's no point to clutter list when full science is unreachable...

[Jognt]

Honestly? I found the mod Science - Full Value years back, and I've not played without it since. (it makes it so you get the full value the first time, no need to repeat) So I really can't say anything about balancing.

I did find that Squad changed Science a bit with 1.7.1 in that you can't gather the full science worth of an experiment anymore, which is what you're experiencing. [x] Science would need to be updated for this.
Since this change defaults to being applied, it also affects every experiment we're used to. The cynic in me assumes this is one of those "Get our users to never be done" carrot-on-a-stick thing.

===============================

If you want to go back to the old way of gathering science where 3 to 4 samples got you the max science for an experiment you'll have to edit the EXPERIMENT_DEFINITIONs for those experiments. The value you need to add is applyScienceScale = false.
This will return the old functionality to experiments that haven't been run yet. Experiments that you have run are saved to your persistent.sfs with their DIMINSHING_RETURNS (not my typo) value. You can try and edit these values (asc = false & scv = 1), but that's at your own risk.

Quick test results:
I used 8x Mystery Goo and used them all in the same spot to see what the max science was that I could get with the new system (no applyScienceScale, it defaults to true):
10 data value 0.30 - 3.0 science
10 data value 0.02 - 0.2 science
10 data value 0.01 - 0.1 science

The remaining experiments (5x) did not give any science at all, despite there still being science left to gather.

I then added applyScienceScale = false to the Mystery Goo EXPERIMENT_DEFINITION and used the same craft to get Goo values from another place at the KSC:
10 data value 0.30 - 3.0 science
10 data value 0.07 - 0.7 science
10 data value 0.02 - 0.2 science

The remaining 5x experiments were again not listed anymore, but this time we maxed out the Goo experiment with the first three. Like always.

Conclusion: Without mods, you will always only be able to collect a part of the collectible science, never the whole chunk, no matter how small.

=====================

With this new system Squad added a variable to the persistent.sfs savefile called "asc" which remembers the applyScienceScale value. Here are all the values the game saves to persistent.sfs with their meaning attached:

Science
{
	id = mysteryGoo@KerbinSrfLandedKSC		// Identifier
	title = Mystery Goo™ Observation from KSC 	// speaks for itself..
	dsc = 1 		// dataScale.
	scv = 0.162092224	// Subject Value. Multiplier on the collected science AFTER the old multipliers are done. Is set to 1 for applyScienceScale = false

	sbv = 0.300000012	// Celestial Body multiplier
	sci = 3.26784039	// Amount of science already gathered
	asc = True		// applyScienceScale
	cap = 3.9000001		// scienceCap for this experiment
}

I'm guessing that applyScienceScale utilizes the new DIMINSHINGRETURNS entries in Serenity/resources/deployedscience.cfg. Time will tell whether Squad meant to only use this new feature for the new experiments and forgot to add applyScienceScale = false for the old experiments, or if this is an intentional change across the board with only a few exceptions like the new ROCScience_ experiments.

Pinging @Sir Mortimer, @SilverState, and @Tekaoh since this covers our recent chat in the Science - Full Transmit thread.

Edit: Rewrote/cleaned up this post. It was a total trainwreck.

Edited June 11, 2019 by Jognt

[drake127]

I created a PR that calculates science for onboard experiments correctly. I have also prepared fixes for GUI but I am not sure if we can agree on some specific threshold that marks experiment completion (PR).

Edited June 11, 2019 by drake127

[Tonka Crash]

5 hours ago, drake127 said:

I created a PR that calculates science for onboard experiments correctly. I have also prepared fixes for GUI but I am not sure if we can agree on some specific threshold that marks experiment completion (PR).

Can you put a threshold in the science.cfg file for this mod? Set it at a recommended value and let the users change it if they want something different.

[drake127]

3 minutes ago, Tonka Crash said:

Can you put a threshold in the science.cfg file for this mod? Set it at a recommended value and let the users change it if they want something different.

Sure, I didn't think of that. However, should it be percentage of science cap (~90 %), experiment gain percentage (~1-5%), experiment gain (~0.1)? I didn't test all experiments nor Breaking Ground expansion so I am not sure what scale is the most reasonable.

For standard science on Kerbin, each of those figures lead to three to five measurements. I am not sure how it scales with high-yield science though. I may look into it later.

[Tonka Crash]

@drake127 I don't think it matters how it's done as long as it's clear what the value represents. To me the science cap is probably simplest to get across to users, i.e.:

ScienceCap = .90 // [x] science will see an experiment with at least 90% science collected as complete

[Jognt]

18 minutes ago, drake127 said:

Sure, I didn't think of that. However, should it be percentage of science cap (~90 %), experiment gain percentage (~1-5%), experiment gain (~0.1)? I didn't test all experiments nor Breaking Ground expansion so I am not sure what scale is the most reasonable.

For standard science on Kerbin, each of those figures lead to three to five measurements. I am not sure how it scales with high-yield science though. I may look into it later.

Note that experiment gain percentage is no longer like it was. It grows smaller with every measurement of the experiment. Due to the way the game calculates it now I'd honestly go for baseValue. As long as an experiment has collected its baseValue (so it's been run once) it's completed.

Any other value is going to be arbitrary in the sense that it's "Am I done after 4 samples? 10? 3?" or "How many runs does it take to get to 90%?" whereas the user is either going to consider an experiment done when they've done it once (maybe twice, if it has a nice payoff, which it no longer has) or when they've done it as much as they need to to get the full scienceCap. Getting the scienceCap is not going to happen anymore, so you may as well go for the "have run once" value and allow the user to show experiments that have any remaining value.

Ps. I've rewritten my trainwreck of a post 3 posts up. I hope it's more readable now.

Edited June 11, 2019 by Jognt

[Tonka Crash]

@Flupster KSP 1.7.1 with BG. I was looking for a different error and noticed a warning I don't recall seeing before and I'm pretty familiar with what warnings/errors/exceptions get thrown on load for my save. All mods current against CKAN.

• DMagic Orbtal Science 1.4.2
• ScanSat v18.10
• [x] Science Continued 5.22

[LOG 09:47:19.128] [6/12/2019 9:47:19 AM [x] Science!]: <Trace> (ScienceChecklistAddon) - Start
[WRN 09:47:19.148] [6/12/2019 9:47:19 AM [x] Science!]: <Info> (DMAPI) - DMAPI.deployDMExperiment method signature has changed. [x] Science may not work for DMagic experiments
[LOG 09:47:19.159] [6/12/2019 9:47:19 AM [x] Science!]: <Trace> (ScienceChecklistAddon) - Done Start

KSP.Log

[Jognt]

3 hours ago, Tonka Crash said:

@Flupster KSP 1.7.1 with BG. I was looking for a different error and noticed a warning I don't recall seeing before and I'm pretty familiar with what warnings/errors/exceptions get thrown on load for my save. All mods current against CKAN.

• DMagic Orbtal Science 1.4.2
• ScanSat v18.10
• [x] Science Continued 5.22

[LOG 09:47:19.128] [6/12/2019 9:47:19 AM [x] Science!]: <Trace> (ScienceChecklistAddon) - Start
[WRN 09:47:19.148] [6/12/2019 9:47:19 AM [x] Science!]: <Info> (DMAPI) - DMAPI.deployDMExperiment method signature has changed. [x] Science may not work for DMagic experiments
[LOG 09:47:19.159] [6/12/2019 9:47:19 AM [x] Science!]: <Trace> (ScienceChecklistAddon) - Done Start

KSP.Log

I can confirm [x] Science still not working with DMagic stuff despite the recent update. Oh well. Atleast manual clicks still work

Edit: I'm still bummed out about the science change.. Would it be an idea to add an optional file/setting to revert back to the 'old ways'? I'm guessing that'd be out of scope for [x] science, but thought I'd mention the idea.

Edited June 12, 2019 by Jognt

[Tekaoh]

5 hours ago, Jognt said:

Edit: I'm still bummed out about the science change.. Would it be an idea to add an optional file/setting to revert back to the 'old ways'? I'm guessing that'd be out of scope for [x] science, but thought I'd mention the idea.

Give it a try in 1.7.2. The impression I got from the changelog was that the change was unintentional and has now been fixed to work as it always has.

[Jognt]

16 minutes ago, Tekaoh said:

Give it a try in 1.7.2. The impression I got from the changelog was that the change was unintentional and has now been fixed to work as it always has.

That’s the plan! 

Well, as soon as I start playing again rather than digging through code and configs. 

[Tekaoh]

1 hour ago, Jognt said:

That’s the plan! 

Well, as soon as I start playing again rather than digging through code and configs. 

Haha, isn't that half the fun?

[drake127]

It looks like 1.7.2 fixes science gain issue. :-) Unfortunately, asc (applyScienceScale) is still True in save file.

@Jognt  Can you check BG experiments how they behaves? I was hoping to calculate science based on asc, but it seems that it's ignored now?

[Jognt]

5 hours ago, drake127 said:

It looks like 1.7.2 fixes science gain issue. :-) Unfortunately, asc (applyScienceScale) is still True in save file.

@Jognt  Can you check BG experiments how they behaves? I was hoping to calculate science based on asc, but it seems that it's ignored now?

ASC state was/is set at the time of experiment deployment for that cb/biome/state/experimentid. Was that ASC setting in your save for a previously-deployed or a new experiment?

I’m still on 1.7.1 at the moment, but I’m guessing it’s a matter of checking sciencedefs.cfg/rocscience.cfg for 1.7.2.