[0.7.6.6] Silo errant behavior. {& [0.7.6.7] Base drops resources when away from it}

Adriano

PAYLOAD STUDIOS
May 16, 2016
2,594
1,352
570
38
@ZeroGravitas Haha, indeed I was informed there was some mathematical reason for that figure. And (0,0) is at the centre of the starting tile, so (192,0) would be the edge of the starting tile.

I've just been informed that resources coming off of silos should be fixed in the Unstable. About to test it myself. I think most reports of it still happening have been in 0.7.6, so it could well be fixed.
 
  • Like
Reactions: ZeroGravitas

ZeroGravitas

Breaker of Games
Jun 29, 2017
3,938
5,759
675
37
UK
www.youtube.com
I've just been informed that resources coming off of silos should be fixed in the Unstable.
The current unstable (0.7.6.8)? If so - lol!:

In creative mode, I planted this bad boy (I borrowed from the laggy crafting bases thread) and planted it on {192, 192}, just to be a git.:eek::D It crashed the game almost immediately, after switching to mobile tech and starting to move (log attached, no dump made).

Reloading game (from save just before the crash), time has been fast forwarded from night to day but all the lights are on on the techs still.

It also detaches all the additional base anchors upon reloading, or returning to the area.

Resources all over the floor upon returning from 1km away (via relay towers).

Silo shuffling and uncollectable resource chunks.

Bonus floating rock debris sprite atop the base, that just happened to be exactly on one of the tile seams (and I'd had to break to place the base).


(Full res.)

Will try a few more things...

[Edit: oh, and going back to that first save, the GSO scrapper is animating with debris and smoke, despite never having been used...]
 

Attachments

Last edited:

ZeroGravitas

Breaker of Games
Jun 29, 2017
3,938
5,759
675
37
UK
www.youtube.com
So, second time around I did a control experiment: an identical base placed mid-tile, near the origin, was (of course) fine. While the other, now stating it's co-ordinates as precisely {192, 192} threw it's chunks and glitched the rock seam, again.

I actually caught the base loading back in on video, while flying back, and flinging resources in the air that were already in a pool on the ground, by the look of it. (So, yeah, the main issue appears to occur on un-loading the bases.)


Before and after video saves attached in Zip with log file.

Bonus log file of crash a while later, while tabbed out - I've had this happen quite a few times in recent unstables.
 

Attachments

  • Like
Reactions: Adriano

Adriano

PAYLOAD STUDIOS
May 16, 2016
2,594
1,352
570
38
Thanks @ZeroGravitas , I had had similar results in my own testing. Not all but a lot of resources off the silos when returning to my base. Will add you save to the ticket as well. With regards to the other issues you mention...
Reloading game (from save just before the crash), time has been fast forwarded from night to day but all the lights are on on the techs still.
This one is a known issue, but doesn't require an urgent fix I don't think.
It also detaches all the additional base anchors upon reloading, or returning to the area.
Not sure about this one, might be a limitation, or intended. The terrain you spawn it on will dictate which anchors are anchored though, i.e. if it's on an incline some anchors might not reach the ground where others might. Are you sure you saw a change with the anchors after a reload?
It crashed the game almost immediately, after switching to mobile tech and starting to move (log attached, no dump made).
Couldn't repeat this, but will log it with your output log.
Silo shuffling and uncollectable resource chunks.
I think this is as a consequence of resources coming off silos. I had it every time the resources fell off. Must be an issue with the tracking.
Bonus floating rock debris sprite atop the base, that just happened to be exactly on one of the tile seams (and I'd had to break to place the base).
This is the same issue as with Erudite reported elsewhere. I also had it when testing this issue.
 

ZeroGravitas

Breaker of Games
Jun 29, 2017
3,938
5,759
675
37
UK
www.youtube.com
Couldn't repeat this, but will log it with your output log.
Yeah, I think it's probably not related. The game's just being a little extra crashy for me, this unstable. Have made a separate report, here, for a my recent cashes.

Are you sure you saw a change with the anchors after a reload?
Yes. I'd added additional anchors manually, before wandering off and back. None of them were attached on return, only the original main, central one. As per these before and after screenshots:


Tile re-load and world re-load do this to the 2 new test bases I have on seams, but not the test base I have set up on the origin point (different terrain, but set up the same). Seems like only the anchor closest to the ground re-attaches, when this fails.

Build beaming the bases up and back down fixes the extra anchors. As does de-anchoring and re-anchoring - which also throws some resources off the bases, due to the more violent jerks (but not as many as reloading on a seam).

I've been re-testing this with 3 copies of a custom made resource base. I found that, with the bases on the north edge (and NE corner) of the origin tile, flying to the west and back did not cause any resource drops. But travelling to the north and back triggered both bases on the seam. (Video to follow.)

Not sure if a directional thing (being OK when both tiles that the bases straddle get reloaded at the same time), or a rate of travel thing (with there being a pretty big hang when rapidly returning via transit tower at 200m/s, or so)...
 

ZeroGravitas

Breaker of Games
Jun 29, 2017
3,938
5,759
675
37
UK
www.youtube.com
As promised:


Part 1 (0m0s) - Flying West and back: Note the deployment of the additional anchors on all the bases. With the bases on the tile seams, only one of the anchors is re-attached (upon tile reload), while the base mid-tile is back to an identical state. No chunks dropped from any base though, which is interesting... (May warrant me retesting, if you are interested @Adriano?)

Part 2 (1m47s) - Returning from trip to north via transfer towers:
Many chunks dropped/flung from both bases on map tile seams. Anchors had been reset and again only one reattached on the 2 affected bases.

Part 3 (2m50s) - Showing re-anchoring: raising the partially de-anchored bases up and down on build beam reattached all anchors. As does just clicking the anchor button twice. This also shakes resources free if the base falls a fair way, to uneven ground. Not as many as lost on a reload, though.


Scintillating stuff, eh... o_O:rolleyes:

Edit:
oh, before and after (Northern trip) save game files now attached.
 

Attachments

Last edited:

ZeroGravitas

Breaker of Games
Jun 29, 2017
3,938
5,759
675
37
UK
www.youtube.com
I hypothesised that tiles might be unloading before the techs. And my predictions from this seemed confirmed:

When flying East, only the base tech on the Eastern edge of the home tile got upset. Flying South (or West) had no affect on either of the 2 bases on the Northern edge, whereas a Northern trip threw off both.

The extra anchors were always failing to re-attach though on the edge bases, though.

But then I also had a couple instances where bases on an edge running parallel to my travel direction also messed up a little. So maybe it's more to do with the time between tile load for bases straddling those 2 tiles...


Inconsistent game load: I've found that the same save game can't be loaded to the same state reliably - it tends to drop a few chunks on one base. One time, loading directly from a game that was lagging due to lots of resources on the ground, 3 of the test bases dropped a bunch of resources.

So I wondered if PC performance is be a factor here? With weaker specs, or busy CPU meaning that there's a bigger gap in time between a tech loading and being properly anchored, falling further during that time...? But I couldn't provoke a problem by loading my CPU with folding@home or Cinebench, so probably not.


I should stop hypothesising and rambling now. I realised that maybe the dev tools allow turning off the distance fog (and weather) so that you can see exactly what's happening, as it happens...?

Edit: test base tech attached, in case it is helpful, + later 5 base test map ready to test.
 

Attachments

Last edited:

Mr-Vagabond

Well-Known Member
Dec 21, 2015
188
117
445
23
=o But this is my crafting base ! The old version of it though, but my base nontheless ! aha hope you have fun testing bug with it *_*
 

Deepincision

Well-Known Member
Sep 18, 2017
17
26
115
34
Australia
Howdy folks, new to the forums and I've had a blast with this game and have thus far had a bug-free experience until today where I encountered glitches as per this thread:

I decided to relocate my base by converting resources to blocks, saving it, sending all blocks to inventory and placing a copy of my base in a pillars biome. All my components were transported to the new site via tech. When I setup at the new location, the base's collectors gathered up my components as expected as well as a few raw resources I picked up along the way and sent them all to storage. I left my base, and upon return I find all resources (newly gathered/refined resources and older transported components) have been duplicated on the ground while silos were still full. All resources on the ground would not respond to collectors and required me to either manually place them on collectors or actually drive over them, only then would they respond. Up till now, I had not seen this behaviour but I have never attempted to save and copy a tech as large as my base before, or one with storage/refining/manufacturing capacity.

Additional observations:
  • I noticed that on spawning the base, it placed all blocks slightly clipping into the terrain, then raised it higher than the terrain. My original base had conveyors, refineries, fabricators etc on the ground, this time they were elevated in the air.
  • Upon return, only 3 out of 7 anchor blocks were actually in the ground.
  • I've since moved my base a 2nd time to a salt flats biome with reproducible results.
  • Once the duplicated resources were sent back into storage, they'd keep "shifting" or cycling around within the silo even though no new resources were coming in.
I'll try to build a new base from scratch and reintroduce the duplicated resources and see if it's still glitching out.
 
  • Like
Reactions: ZeroGravitas

Deepincision

Well-Known Member
Sep 18, 2017
17
26
115
34
Australia
[Update]
After reading through the forums a little more, I've realised I'm running stable 0.7.7 so this isn't in the correct board... anyway I've gathered some further information that'll hopefully be helpful:
  • Building a new base from scratch did not solve the problem, as once I left draw distance I'd return to a messy pile of resources.
  • Tried selling all my glitched resources+components through Matryoshkas, and replacing them with newly collected raw resources. Left draw distance and returned to non-glitched resources.
  • Tried refining about half the resources so it was split between raw and refined, since the original glitch for me occurred with entirely refined resources or components. Left draw distance and returned, same non-glitched result.
  • Tried saving the base, despawning and copying it a short distance away. Transferred all refined+raw resources to silos. Left draw distance, and again non-glitched...
TLDR, seems selling the offending resources is the way to solve it for me. Not ideal but at least I've got a workaround.
1709181727.jpg
 

ZeroGravitas

Breaker of Games
Jun 29, 2017
3,938
5,759
675
37
UK
www.youtube.com
Hi, @Deepincision. Thanks for taking the time to report on your game's buggy behaviour.:)
I'm running stable 0.7.7 so this isn't in the correct board
I'd say this thread is still the perfect place for describing your problems, as these exact issues are still occurring in 0.7.7.0.
on spawning the base, it placed all blocks slightly clipping into the terrain, then raised it higher than the terrain
Yes, I saw this as standard (in my creative mode tests above). The taller the tech, the further through the floor it spawns in.
Left draw distance and returned to non-glitched resources.
The tech will (usually) de-spawn just beyond maximum draw distance (~600 meters), but the world tile it's on may still be loaded in, if you have happened to travel across it's width (384 meters). So, as @Adriano has said, above, best to travel at least 1Km (1000 meters), to be sure everything there unloads and then reloads. (Although, with my tests above, I could tell for sure when the full unload occurred due to a 1-2 second game hang - but extreme circumstances.)

So, while selling the glitched resources stops the silo shuffling, as you observed, it shouldn't stop the base re-dropping resources again and creating newly glitched chunks.

Direction of travel away from (and back to...? Hmmm.) the base also seems to have an effect. For re-testing the above base, I would ask that you travel south and/or north (1Km, to be sure, but I think 600m will do it).

From your screen shot (thanks for including it - great stuff!:) And nice twin prop plane.;)): I can calculate that your base is sat on a tile seam located at 9024 North (a fault line that runs East-West under the left hand side of your base). Hence flying north/south will cause the 2 tiles your base is straddling to unload/reload at different times (for sure). Afterwards, moving the base an entire base width north or south should stop any repeat of this problem.

You can calculate where all the seams are for yourself: the first ones are at 192m from origin, and then every 384m thereafter (in each compass direction). So input 192 to your calculator and keep adding 384 until the answer you get is close to one of the co-ordinates shown under the mini-map (when controlling the base tech!). If a seam is within half your base width of base's location, then it's straddling a seam. (1 tech block is 1 meter wide.)
 

ZeroGravitas

Breaker of Games
Jun 29, 2017
3,938
5,759
675
37
UK
www.youtube.com
@Django - hey.:) You've previously commented on treadmill world glitches, so might you work on these issues too (and have insight regardless)? Have any of my/our investigations, above, been potentially useful for you coders? Is there anything you'd like us to re-test/report in particular? Or is this on the back burner for now, and you'll get back to us, perhaps?
 

Django

Well-Known Member
Apr 21, 2016
278
572
505
London, UK
Hey @ZeroGravitas I have actually been keeping an eye on this thread after Ade brought it to my attention. I've not been able to read through all of the replies, although it seems like there's a lot of useful information there.

I have however been able to track down the issue - All resources currently held get 'duplicated' in a bugged state (they share the same ID and the game doesn't quite know what to do with them) when the tech is unloaded.
There was a new edge case introduced where the tech would get unloaded but not the resources it was holding along with it. Nothing to do with treadmill this time ;)

I've got a fix ready on my machine but wanted to do a little more testing before I push it to main branch :)

Again super thanks for helping narrow this down. A fix should be in by the next unstable - if there's any remaining issues with it please let us know :)
 

ZeroGravitas

Breaker of Games
Jun 29, 2017
3,938
5,759
675
37
UK
www.youtube.com
A fix should be in by the next unstable
Cool. I'll give my Creative mode test site (save attached above) another poke once that's out.
All resources currently held get 'duplicated' in a bugged state
Note, I don't think that it's ever appeared that *all* base resources get dropped/duplicated. Just enough to make a big mess, but usually not enough to be certain there's additional (unless all silos were already brim filled).

There's also the 'edge case' bases mostly de-anchoring. But anchoring has other problems too, perhaps, and we can circle back to that anyway, if it remains; I'll let you get on with your fixes (good work).:)
 

Deepincision

Well-Known Member
Sep 18, 2017
17
26
115
34
Australia
Hi, @Deepincision. Direction of travel away from (and back to...? Hmmm.) the base also seems to have an effect. For re-testing the above base, I would ask that you travel south and/or north (1Km, to be sure, but I think 600m will do it).
Thanks for the feedback. I've since run a few more experiments and have had some interesting results:
  • With my new scratch built and stable base containing newly harvested or recovered resources from bricks, I've tried travelling ~1500 units north, south and random short distances with no replication of the glitch.
  • I've also tried to replicate the glitch by copying my test base with a mix of resources and components carried on my tech. Same travelling conditions, still no glitch.
Here's where it gets interesting:
  • I found a stable location in the desert biome nearby but away from any tile seams, copied the test base, same travelling conditions and no glitch.
  • I returned to the original site of the glitch in the pillars biome on the seam, copied the base, using the same resources and reproduced the glitch under the same travelling conditions.
  • I then went to the stable location in the desert, copied the base and used the glitched resources, same travelling conditions and reproduced the glitch again. The glitch seems to be triggered by the seam location, but once "tagged" the resources retain their behaviour when moved to a non-seam location.
  • The resources were constantly shifting around on my tech as I was travelling. If a 3x3 silo contains less than 9 resources, it won't shift. Shifting only seems to occur when 1 or more "layers" of resources are present.
  • Same results were obtained using 3x3 and 2x2 silos.
  • Effects are the same for raw, processed and components.
 
  • Like
Reactions: ZeroGravitas

streak1

Well-Known Member
Aug 30, 2015
2,291
2,139
570
23
steamcommunity.com
Title is self explanatory. Seems to happen every time I load the save. The tech is just a simple thing to hold resources and components before I make a base. Mostly GSO 3x3 silos, all of them full.

Save and log attached
 

Attachments

ZeroGravitas

Breaker of Games
Jun 29, 2017
3,938
5,759
675
37
UK
www.youtube.com
I've just merged @streak1's new thread titled "[0.7.7.2] Resources fly out of silos upon save load" with this existing one. Please read below and see if you agree this makes sense.:)

I opened up your save, and yup, resources and components all over. Once re-collected, with a little difficulty, they were shuffling on the silos and fell again upon reload. Typical of the silo resource dropping bug reported above.


Selling all the resources that dropped stopped this behaviour. Reload were then fine and unloading the world tile by leaving 1km and returning was fine too.

However, the base, in the save file, was not on a world tile seam. There are some within the base area, south and east, bisecting the wall techs.

My hypothesis would be that the bugged (duplicate) resources were created when a base holding them was on a world seam, in 0.7.7.1 (or before), then collected to this base... (ye, no?). And that @Django's coding fix seems ok, still.


However, that save file is majorly screwed up, anyway (and deserving of a thread to itself). Flying out to the east, I found a pile of large mission enemies, duplicated and intersecting. This caused my game to hang for 10 seconds or so when they loaded in, then totally lock up when entering their agro range (output log attached, save with plane attached).
 

Attachments

Last edited: