EVA bug in 1.1.3

[bewing]

I just saw a pretty serious bug in 1.1.3 and I wanted to get ahead of things by making a post.

I was sending a rocket to the Mun, and did a few science experiments. I have all my equipment right below a MK1 command pod on the rocket stack. Sent a scientist out on EVA onto the little ladder and grabbed the data -- all normal so far.

Then I hit B to board. The scientist remained outside the rocket in space, and control returned to the rocket. I was completely unable to return focus to the scientist to try to move or board again. When I hit [], the game said "Focus: Sally Kerman", but I was still controlling the rocket with the keys.

Edited June 22, 2016 by bewing

[Padishar]

11 minutes ago, bewing said:

I just saw a pretty serious bug in 1.1.3 and I wanted to get ahead of things by making a post.

You should always include a log file (output_log.txt or player.log) when reporting any issue and your report could also do with more detail about the situation.  Was your 1.1.3 install updated from 1.1.2 or was it a clean install?  Is it definitely unmodded?  Have you been able to repeat it?  Where was the vessel (i.e. which SoI)?  Did you happen to change SoI while on EVA?  Had you quicksaved and reloaded during the mission or gone back to KSC and then returned to the vessel?  Could you return to KSC after the bug occurred and did that change anything?

[Vaporized Steel]

Strange, I cannot reproduce this bug. I tested out the same conditions. A kerbal (scientist) on the Mk1 ladder. With Mystery Goo, gravmax and thermometer on the fuel tank beneath. I took all the science without leaving the ladder and then pressed B to board. The science and the Kerbal entered the Mk 1 command pod low orbit around the Mun.

Can you post a picture of your craft and the time it is happening. Maybe it's a bug only with Sally Kerman or other scientist other then Bob who doesn't seem to suffer this issue.

Edited June 22, 2016 by Vaporized Steel

[bewing]

I just replicated it, and I will post all the log files etc. in the bugtracker, of course. Duh. The install is completely unmodded, of course, and was updated from 1.1.2. Verifies OK from Steam.

The ship was well within Kerbin's SOI at the time (1.5Mm), and I had watched it all the way from launch -- no quicksaves or reloads or focus switching. I did not want to try switching back to KSC, because then I would lose my ability to revert the flight -- which I seemingly needed to do to avoid losing the scientist.

This is the first space EVA I've done in 1.1.3 -- so I hope it doesn't happen all the time! But I did a bunch of EVAs on the ground first, and those worked OK.

Turns out that you can't go back to the Space Center during the bug -- because it says you are on a ladder, so the game's state can't be saved.

Edit: Odd ... for some reason, I can always duplicate the bug if I launch from the ground, but I can't duplicate it from a savefile taken just before the EVA.

And it appears to only happen for ships that have been staged .... and it's not every rocket that does it, it's only some of them.

filed: http://bugs.kerbalspaceprogram.com/issues/10007

Edited June 22, 2016 by bewing

[Vaporized Steel]

^Under those conditions, same result!
I hope we can nail it down to what action specifically triggers it as long as there is no fix for this.

[bewing]

Well, the workaround is just to make a savegame before your EVAs, if you've watched the rocket go all the way up from the ground. If you can't board, then just reload the savegame and it'll work.

[Padishar]

I've had a bit of a look through your log and there are a number of exceptions (all very similar) at different points.  The first one appears to happen when you stage off the boosters and it appears to be something to do with trying to destroy temperature gauges.  Were the boosters (or any other parts) showing the gauges when you staged?  Does this affect whether the bug occurs?

A similar exception (in the same bit of KSP code but called from a different place) happens when switching to Jeb when you EVAd.  This one doesn't appear to cause any immediate problem.

Then another happens when you board again and the active vessel is switched.  This is happening during the processing of the "board" and is apparently preventing some important part of the code from running (e.g. the bit that actually switches the vessel over) causing the failure to correctly board and, presumably, the other accompanying issues too.

[bewing]

1 hour ago, Padishar said:

Were the boosters (or any other parts) showing the gauges when you staged?

It's always a little hard for me to tell which part is showing a heat bar/temp gauge (especially in the middle of a giant ball of plasma) -- but I'm pretty confident that there was only one heat bar showing at the moment of staging (during my example testflight), and it was on the SciJr that is just below the command pod.

1 hour ago, Padishar said:

 Does this affect whether the bug occurs?

I waited a few seconds after that one last heat bar vanished before I staged, this time. I think that is the test that you mean. No, the bug still happened.

1 hour ago, Padishar said:

A similar exception (in the same bit of KSP code but called from a different place) happens when switching to Jeb when you EVAd.  This one doesn't appear to cause any immediate problem.

I suppose the command pod is still warm after the liftoff, and there is that whole deal with the EVA kerbal being set to the same temp as the internal temp of the pod. But the bug can happen after 10 minutes in space, so the command pod is not particularly warm.

1 hour ago, Padishar said:

Then another happens when you board again and the active vessel is switched.  This is happening during the processing of the "board" and is apparently preventing some important part of the code from running (e.g. the bit that actually switches the vessel over) causing the failure to correctly board and, presumably, the other accompanying issues too.

Hmmmm. Interesting. 

[AzraelDR]

Interesting to know that this happens in stock--I had assumed that one of the mods had caused this problem as well. A minor workaround besides reloading that I found is to go to the space center BEFORE going on EVA. I am trying to play hardcore career mode and recording my progress, so I disabled quicksaving; this is how I found the workaround. 

[Gojira1000]

Confirmed - managed to have the same bug arise last night. Had quicksaved and a load did fix the issue. Ship had performed a rendezvous so was likely quite cool. Switched to the derelict, EVAed the pilot over, got stuck on the ladder.

[Slaghead]

I had this happen this morning as well.  Also near the Mun.  I had Jeb EVA to collect some science before loading him into the lander can.  I had him board the command pod and I could control the ship but couldn't get into map view and his body was floating next to the ship.  I had to hit escape and go into the space center and reload the ship and he was where he was supposed to be.  Then, I moved him over to the lander can and detached and tried to use RCS to inch away from the main ship and even though Jeb was in the can, the main ship was firing its engine instead.  It was very confusing.  I had to go back to the space center and reload the lander and was able to control the lander from there.

My install is not unmodded though.  I just saw this thread and was all 'hey, that's that annoying thing that happened to me!

[DMac5000]

I've been having this same bug every single time I EVA since I updated to 1.1.3. I've got the Kerbal Engineer, PlanetShine, and RCS Build Aid mods. I just found a workaround that kind of helped me out (this time anyway). When the bug occurred, I Alt+tabbed out of the window then back into the game again. This time Jeb still appeared to be outside my craft, but I could return to the tracking station without having to revert to an autosave.

[fommil]

are there any plans to release a patch to the 1.1 series with a fix for this? It makes the game incredibly frustrating to play. FYI I created a duplicate before spotting this 

 

[swjr-swis]

2 hours ago, fommil said:

are there any plans to release a patch to the 1.1 series with a fix for this?

Yes, it's called 1.2, and it's scheduled to drop tomorrow.

(in other words: no, they don't do retroactive patching of older branches).

[Kerbal101]

IMHO this is caused by incorrectly attached fairing just below the pod. Its upper border clips into the door and prevents kerbals to enter back.

Basically this.

[fommil]

6 hours ago, Kerbal101 said:

IMHO this is caused by incorrectly attached fairing just below the pod. Its upper border clips into the door and prevents kerbals to enter back.

Basically this.

No, is not this. Although possibly related. No point in me sharing my craft if it isn't going to get fixed in 1.1 line. I'll just have to live with it. I have a long running career mode running and I went be updated to 1.2 for a few patches because I'll have to start again.

[Kerbal101]

7 hours ago, fommil said:

No, is not this. Although possibly related. No point in me sharing my craft if it isn't going to get fixed in 1.1 line. I'll just have to live with it. I have a long running career mode running and I went be updated to 1.2 for a few patches because I'll have to start again.

Sure you can share the craft.

[Vaporized Steel]

Attention people: This thread was posted on June 22nd and refers to a problem that will be fixed in v1.2. Seems to me that opening a new thread would be a better idea. Since hijacking a thread is like posting your own thread but now we have to read something from months ago, which doesn't even apply anymore by tonight. Making a new thread and notifying that the issue still applies gets you better attention, and sadly no help because 1.1.3 will not be patched so your at a dead end unfortunately.

@Kerbal101

As for starting a new game. If you want to have your old savegame in place. You can see if you can copy/past the craft files from 1.1.3 to 1.2 (if that is compatible)
If not rebuild them and atleast hyperedit your more valuable and necessary equipment to the places you'd send them. My condolences if you have hundreds of vessels in the solar system, good luck!

It really doesn't seem cheating to me because you just recreate the universe you already created. And it's not like your put on a public cheat list so no worries there.
This can be quite time consuming but alteast it's a option. Although I doub't you'd do it just for that 1 single bug.
It won't work if you have mods.
But then you really cannot complain, it's just the consequence of having newer versions every 3 to 6 months.

[Kerbal101]

@Vaporized Steel I don't understand a single word you wrote. The problem in this thread is current. 1.2 stable has not been released yet. Squad does not use hot-patching, all bugfixes (and new bugs (tm)) are released as next version, for example 1.1.2 is bugfix of 1.1.1, 1.1.3 is bugfix of 1.1.2, 1.2pre is bugfix of 1.1.3, 1.2 stable will be bugfix of 1.2pre. The craft from older versions is compatible with newer KSP version, but craft created in newer version will not open in older KSP.

Still waiting for the problematic craft to be shared.

Edited October 11, 2016 by Kerbal101

[Vaporized Steel]

@Kerbal101

Can you elaborate on what there is not to understand about? Or not want to understand because you seem Finicky about your problem.
Technically a new version is added content with bugfixes like you say. That is how I understood it and it has always been this way, you have thereby answered why bug fixing is applied this way. So you already know there won't be any fixes.

So is this a suggestion to do bugfixes prior to a new release?
If so then why not open a new thread to suggest it then use a new thread that reports the problem, knowing that there is no setup in place yet to do post release bug fixing of any kind.
And even if frequent bug fixing in between main releases does get implemented then it will start today and the implementation of these bug fixes as soon as V1.2

Also I was not referring to the age of the problem but the age of the thread.
Also to call the problem "current" 1 day before a new release is a big stretch although accurate for a few more hours. Your current probem will be solved in a few hours, and I advice a alternative solution for your current savegame. A solution you probably don't appreciate since you didn't mention it. Unless that is the part where you thought you were reading Chinese. IF it does, I'm not a native English speaker.

While I hope for everyone including myself that you get your way, because bugfixing is important it is not going to happen in 1.1.3.

 

Edited October 11, 2016 by Vaporized Steel

[Kerbal101]

@Vaporized Steel I don't have this problem, I am trying to diagnose it.

[fommil]

Kerbal101 I've posted my craft to the bug tracker, thanks for linking to it.

[Kerbal101]

2 hours ago, fommil said:

Kerbal101 I've posted my craft to the bug tracker, thanks for linking to it.

It would be cool, if you'd remove all the mod parts from the craft.

Nevertheless, I sanitized craft file using text editor - and can confirm the issue on my machine.

tip: angling the craft on launchpad at 85 degrees saves you some fuel.

Edited October 11, 2016 by Kerbal101

[Kerbal101]

@fommil and all: this issue is caused by fairing alone. Heatshield plays no role. Solar panel plays no role. Decoupler plays no role.

It seems that the game does not account for fairing state - wither it was deployed or not, so long its attached to the vehicle and whilst it is/was - it has covered the pod even once - it applies the limits around the area and these limits persist until its separated from the vehicle.

Means, if one decouples the second stage (fairing is also part of that) from that vehicle, the bug never appears.

A hackish way to get back, is to EVA the kerbal and push against the rocket in retrograde direction using EVA propellant (like stopping car with bare hands) until 40k periapsis mark.

 

I have not yet tested the case where fairing would make sufficient room for kerbal in EVA (ie. "FAT" fairing).

Yes, it looks like fairing bug.

[fommil]

14 hours ago, Kerbal101 said:

@fommil and all: this issue is caused by fairing alone. Heatshield plays no role. Solar panel plays no role. Decoupler plays no role.

It seems that the game does not account for fairing state - wither it was deployed or not, so long its attached to the vehicle and whilst it is/was - it has covered the pod even once - it applies the limits around the area and these limits persist until its separated from the vehicle.

Means, if one decouples the second stage (fairing is also part of that) from that vehicle, the bug never appears.

A hackish way to get back, is to EVA the kerbal and push against the rocket in retrograde direction using EVA propellant (like stopping car with bare hands) until 40k periapsis mark.

 

I have not yet tested the case where fairing would make sufficient room for kerbal in EVA (ie. "FAT" fairing).

Yes, it looks like fairing bug.

wow, I would never have guessed! Thanks for looking into this. I'll just design my rockets differently and discard fairings earlier... which might be a bit tricky for this particular rocket design :-/

It would be really amazing if a 1.1.4 patch came out with critical bugfixes like this. I'm in the middle of a career campaign on Difficult mode (although I've loaded previous savegames when this has stung me) and have no plans to upgrade until I've completed the tech tree.

Another workaround might be to use the Procedural Fairings plugin instead of stock.

@Kerbal101 can you please confirm if this bug is still present in Kerbal 1.2?

Edited October 12, 2016 by fommil