Skip to main content

On Sale: GamesAssetsToolsTabletopComics
Indie game storeFree gamesFun gamesHorror games
Game developmentAssetsComics
SalesBundles
Jobs
TagsGame Engines

I couldn't replicate it yet .. at first I thought it's probably all the crates generated
but the number of objects don't seem to be increasing after a while

hmm.. can you send over the save file somehow?

(+1)

Here it is.

https://bytes.one/upload/slot0.zip

I'm not sure how it happened but somehow a 158mb large decal mesh was generated 🥴
(how it not just crashed is anyones guess)
it's rather strange - there is a hard limit how many triangles a decal can make
and all mem seems to be initialised properly..  so I have no idea of the cause
anyway I better put in some safeguards/sanity checks just in case

for now here is a fixed version of the save:
https://drive.google.com/file/d/1dJHbPpHYxNT81XvVTynfIDeqdV_GoVJV/view

(+1)

Now I opened the Windows Explorer to see the filesize and hammered F6 (sometimes F9)  in the gameplay and it is inconspicuous until I come to the dump in E2M3, here for example after some time the save file seems to double its size every time or at least grow faster every time.

this bug is so weird - it worries me a lot - but I just cannot seem to reproduce it 🤔
anyway I added a new hotfix to Patches.zip (Build 1890)
please check if it still happens with this version
(it double checks if the decal mesh seems correct)

(+1)

It's getting worse but maybe the way it does helps you figure it out. I played the same map E2M3 at the same spot. This time I just saved 2 or 3 times until the game crashes. Now this is interesting, the quicksave from 1890 crashes 1890 but works in 1878. So it is somehow corrupted but you only stopped the uncontrolled growth. https://bytes.one/upload/e2m3_quicksave.zip

It's no big deal for me but this could be one of the more serious bugs which affects at least some people.

Can you confirm the crash with my savegame in 1890 on your computer?

yes, save files help me a lot - I can actually debug and see where the crash is 👍
of course in this case an opengl call seem to not like something about the mash data
so all I know is that  a decal mesh that got malformed somewhere
I added more sanity checks and it seem to have worked
( it now attempts to fix the mesh on load more aggressively - or throws away the decal altogether if it's beyond saving)

anyway I added the fix to Patches.zip (Build 1892)
also thank you for finding this and of course helping fixing it!
this is one of those - how nobody have noticed this before bugs 😔

(+1)

It looks good, couldn't break it anymore!