[QUESTION][AND PARTIAL FIX] Which Generic Missions Still Not Closing Out MD Cues?

RoverTX · Post by **RoverTX** » Sun, 19. Jun 16, 23:25

I have some results for L2M119, Escort Convo.

The issue is under the Cleanup queue. The remove Escort Ships fires and compeletes, BUT, the Wait For Ship Clean up does not! This of course is the que which actually cancels everything.

It will only fire when the CleanupShips, Escort, and COM01Ships groups have all been destroyed. One or more of these are not being destroyed full, I am guessing COM01Ships. Digging in further now!

Edit 1: I turned on SETAX10 walked away for 15 or so minutes and came back and all the escort ship missions had closed out. I think this means that the timeout que is firing and thus causing the cleanup que to fire, BUT, the cleanup que doesn't actually cause a full clean up, so it waits until the ships reach their end point before closing out the que.

This results in dog pile of ques being left open over time. The safer bet might be to just add removal ques for the two above groups, and the parent group as it where L2M119.Ships. That way after cleanup has started and the player is out of sector the ships will be removed no matter what and the que chain will end on time.

RoverTX · Post by **RoverTX** » Tue, 21. Jun 16, 02:10

Here are the changes I have made so far, haven't been able to test them out yet, but this should cause a quicker cleanup of the escort missions.

https://www.diffchecker.com/nlwhzbot

You can download this "fixed" director here, warning have not properly tested this myself yet,

https://s3.amazonaws.com/forums-rover/2 ... Convoy.xml

I am assuming at this point that the really long lived ques are because the cargo ships never reaching their end destination for a number of reasons. When this happens the only way for them to get cleaned up via the "go to director" will never happen and the ques never close.

With the above edit not only are the escorts removed after timeout, but so are the cargo ships. This way the dog pile and the possibility of becoming stranded are not possible.

MrFiction · Post by **MrFiction** » Wed, 22. Jun 16, 14:20

@RoverTX: does never accepting these missions help?

And if I remember correctly you could also fix this issue by going to an unknown sector, one without stations and waiting there for 2 hours. Can't remember if it's 2 hours with or without seta but I've read that clears out the missions as well.

Elucca · Post by **Elucca** » Wed, 22. Jun 16, 17:30

MrFiction wrote:@RoverTX: does never accepting these missions help?

From my experience, I don't believe so. I almost never actually play the missions and my game slows down all the same.

RoverTX · Post by **RoverTX** » Wed, 22. Jun 16, 17:32

Does not. Actually doing the missions might help just a tiny bit.

MrFiction · Post by **MrFiction** » Thu, 23. Jun 16, 10:03

Yes, but waiting in an empty sector for a few hours should let the garbage collector kick in, that's how I understand it.

RoverTX · Post by **RoverTX** » Sat, 25. Jun 16, 22:33

I have been testing the fix for the Escort missions I posted above and it seems to be working. To use MrFiction's vocabulary on the subject, it seems to be ensuring and kicking in sooner the "garbage collection". Which really is a good way to think about this. The "unused/unreferenced objects/cues/ships" are being culled effectively.

Some third party confirmation of this would be much appreciated.

I have already moved on to the scan asteroid missions. Which cause seems far less obvious.

RoverTX · Post by **RoverTX** » Sat, 25. Jun 16, 23:36

Ok I have a theory about the Scan Asteroid Missions.

There are two groups created "this.Unfiltered Roids" and "this.Roids" Under 3/4 cancel ques that happen both of these groups are un-referenced via remove_group before the que is canceled. But when there are less than 5 roids in the sector resulting in the mission offer being aborted the Unfiltered groups is not un-referenced.

See the below diff for the changes I have made so far, and a clearer view of what I am talking about.

https://www.diffchecker.com/igjaa13h

Not sure if this could be a problem, but could cause a bunch of orphaned groups, unless canceling the que already causes them to be un-referenced. Not sure how to test this, because the mission director won't be telling me any info about ques that have already been canceled...

Edit: Actually was reading that wrong, there is one instance where Riods group is not un-referenced but the Unfiltered is, and one where the Unfiltered is not. Please check the updated diff checker below.

https://www.diffchecker.com/omurhrhb

Bandaro249 · Post by **Bandaro249** » Thu, 30. Mar 17, 16:01

@RoverTX

Will your file work for tc plots aswell? If not, is there any existing file that can help improve my perfomance in tc plots without removing my plot progress?

Edit: I hope i didn't resurrect too old thread.

RoverTX · Post by **RoverTX** » Thu, 30. Mar 17, 19:43

The upcoming patch from egosoft address all the issues this patch attempted to fix and more. Look to that instead of this.

Bandaro · Post by **Bandaro** » Thu, 30. Mar 17, 23:14

Do you mean the 3.2 beta patch? Currently on steam i got patch 3.1 and it didn't seem like it fixed anything cause when i reset mission director i got steady 60 fps in comparison to 44 on seta1 and 27 on seta 10.

RoverTX · Post by **RoverTX** » Thu, 30. Mar 17, 23:14

Yes the beta

egosoft.com

egosoft.com

[QUESTION][AND PARTIAL FIX] Which Generic Missions Still Not Closing Out MD Cues?