A user from the Great White North, recently returned to the game is experiencing memory corruption in the original USNI BattleSet, Scenario 5.0.  He has experienced this both in 2020.009 and 2020.017.  

Some saved games are attached. I have tested extensively using AutoTest and have not replicated either issue (game exiting with heap corruption error [USNI5.zip] nor game exiting without any warning [NoErrorCrash.zip]).  Right after the USNI5.zip saves a Slava is sunk but does not trigger a crash for me.  

The user has mentioned doing a lot of air group splitting and joining while playing.  I view that as a valuable hint as very few players exercise that functionality and it is an area of the game prone to bugs.  The latest hint is that the crash may occur when an air group is landing, perhaps an hour of play after the USNI5.zip saves when loading from the saves.  

Quote

Hi Tony,

I did reload the savefile autosave.0040 yesterday and I did crash again about after a other hour of play.  Same error message same line of code.

I am thinking a something.

I just look back at all the screen capture that I did already send to you. It seems that the crash always happen when a group of plane just finish there landing.

Any help narrowing down these crashes is greatly appreciated! 

 

USNI5.zip NoErrorCrash.zip

Played 66 minutes of autosave.0036 with out a crash.

Sank Red TG and shot down AC and Helos

Played saved game at 5 second time.

In addition to multiple runs using autoplay in both 2020.009 and 2020,017, I was not able to crash the GE during a manual play of the game from the USNI 5 0040 saved game.  I'm attaching my saved game from when all of the Red bases were destroyed for the heck of it.  

I made sure to do some air group splitting during that manual play, no issues.

USNI5_RedBasesDestroyed.zip

I forgot to add in my post last night that I also split air groups and did not have any issues.

I did more tests. You will find more information in the attach files. The detail information of my third try is within the document.

 

HUCE crash with USNI 5.0 build 2020.017.docx

I'm still very rusty, maybe that helps, I did manage to get a crash with error message after about an hour of game time from supplied auutosave.006. I wasn't doing autosaves or any logging and my eyelids were dropping (I forgot a screen shot even) so I'll see what I can do to repeat with more info but perhaps we have some common elemnt in our computers, so I'll grab that info too and post again.

 

Don

11 hours ago, donaldseadog said:

I'm still very rusty, maybe that helps, I did manage to get a crash with error message after about an hour of game time from supplied auutosave.006. I wasn't doing autosaves or any logging and my eyelids were dropping (I forgot a screen shot even) so I'll see what I can do to repeat with more info but perhaps we have some common elemnt in our computers, so I'll grab that info too and post again.

 

Don

Thank you Don.  What security software do you use?

Hi Tony,

the Friday's Zoom session was interesting. Thanks.

As you already know I had a similar crash playing the Red side on different Battelset.

I did the test again on a different computer with a different security software. This time the security software was ESET Antivirus.

same result.

 

JEAN.0059.hp1 JEAN.0060.hp1 JEAN.0057.hp1 JEAN.0058.hp1

On 1/21/2021 at 5:18 AM, TonyE said:

Thank you Don.  What security software do you use?

AVG free.

I'm on windows 7 home, an old (now ) pentium laptop and 2020.017 game build.

None of my game saves have created reproducible crashes. They are a bit random in nature, on one occasion when I restarted the old game seemed to have been saved but a whole surface group of mine had disappeared. I've now started playing other battlesets to see if they are stable.

I've also noticed I have work to do on my DLL s

Don

AVG free.

I have ran a new game and from many of the saved games and had no crashed for hours of play. I am using McAfee and was requested to post for reference. I am also downloaded from HUCE starting with the 2009.50 version to the latest.

TonyD

Thanks all, keep it up.  I won't be able to test again for 3-7 days so anything you can continue to refine is appreciated!

On 1/25/2021 at 9:52 PM, donaldseadog said:

AVG free.

I'm on windows 7 home, an old (now ) pentium laptop and 2020.017 game build.

None of my game saves have created reproducible crashes. They are a bit random in nature, on one occasion when I restarted the old game seemed to have been saved but a whole surface group of mine had disappeared. I've now started playing other battlesets to see if they are stable.

I've also noticed I have work to do on my DLL s

Don

I don't seem to be getting crashes from any other scenarios/battlesets.

As this is a legacy DB could it be corrupted on his machines? All his machines are from the download from the same source. Just throwing out ideas. Can he be sent someone's legacy DB and replaced in one of his machines to test for testing?

20 minutes ago, Tony said:

As this is a legacy DB could it be corrupted on his machines? All his machines are from the download from the same source. Just throwing out ideas. Can he be sent someone's legacy DB and replaced in one of his machines to test for testing?

My knee-jerk reaction is "that is absurd, the file hasn't changed for 20 years or more."  On the other hand, it is an easy enough thing to try.  Tony, can you zip and send your USNI.RES and USNI.RSR files to Jean?  Let me know if you need any help.  

Your thought does reinforce a thought that has been bouncing around the back of my head in that there is a reasonable probability that the memory corruption originates from data conversions applied to that ancient BattleSet, not just legacy but ancient <g>.  In theory I could compare many of the structures in that BattleSet to try to spot data conversion errors while loading the USNI.RES file.

The problem is not limited to USNI battleset. I had the same crash twice on the "bee hive" scenario of the battleset HDSA.

Reloaded using 

autosave.0050.hp2

then

 

Crash landing a plane on a carrier both time.

I will happy to test.