Collapses Happening Even When Structural Integrity Rules Appear to be Met

[Stranded_Napkin]

So, I've put in over 3k hours into this game. I get how things work. Each block has its own set weight and horizontal support limit. For some reason, the game has been ignoring this at times. I've now had two collapses when upgrading from wood frames to wood blocks and from cobble to concrete. Both of those situations are from one block to another that have the exact same specs for weight and horizontal support.

 

Does the game add extra weight during the process of upgrading the blocks or something? I never had this issue before so I'm at a loss. I mean, I'm not saying it's a bug. I do not know if there were additional changes that I'm just not aware of. Just frustrating to see everything collapse when you have everything already in place.

[meganoth]

I don't understand. Wood frames and wood don't have the same specs (AFAIK, didn't check), and neither have cobble and concrete the same specs (very very sure of that). If you upgrade from cobble to concrete the weight of that block definitely changes

 

If you meant that the collapse happened between two blocks of the same weight, that isn't a surprise either, one was probably slightly nearer to his nearest support and therefore had just enough support and the other didn't. Also collapses often don't happen at the block which is out of support but on blocks that are part of the "support chain".

 

 

 

Edited August 20, 2023 by meganoth (see edit history)

[pApA^LeGBa]

They have the same stats now. That´s interesting.

 

I know for a fact that there was a diference before, you couldn´t do the same amount of blocks without vertical support for cobble and concrete. 100%.

 

And i can do 12 blocks of concrete or cobblestone with no vertical support. For wood and building blocks it´s 10 for both.

 

I can also upgrade the 10 and 12 blocks without support from either building blocks to wood or from cobblestone to concrete with no problems.

 

Edited August 20, 2023 by pApA^LeGBa (see edit history)

[theFlu]

1 hour ago, meganoth said:

Wood frames and wood don't have the same specs (AFAIK, didn't check), and neither have cobble and concrete the same specs (very very sure of that). If you upgrade from cobble to concrete the weight of that block definitely changes

They've all been uniform for a couple alphas now, basically three "weight" tiers: woods, stones and metals. Would've taken screenies, but I haven't updated to A21.1 yet.

[Stranded_Napkin]

Yeah, I don't know how long it's been but they have the same specs now. Wood Frames and wood blocks both have a mass of 5 and a horizontal support of 40. That gives you a horizontal block support of 8 blocks. The cobble and the concrete both state a mass of 10 and a horizontal support of 120. This gives you 12 blocks of horizontal support.

 

These numbers are as stated in the game. I've never found it to be different in practice either. That being said, I get that my build is likely pushing those limits. Did they just change it and forget to change the localization file or something?

 

Don't get me wrong, this issue only appears to happen when you upgrade blocks away from the supports. If you upgrade at the supports and then progressively move away from them, it doesn't seem to cause any collapse. It just doesn't make sense to me if the mass and horizontal support numbers aren't changed. It really shouldn't matter what blocks you upgrade as long as those numbers are the same. That is unless there is an unknown mechanic where it momentarily adds an extra bit of mass during the process of upgrading. I can't think of another reason why it would act like this.

 

Edit: spelling correction

Edited August 20, 2023 by Stranded_Napkin (see edit history)

[theFlu]

2 minutes ago, Stranded_Napkin said:

That is unless there is an unknown mechanic where it momentarily adds a extra bit of mass during the process of upgrading. I can't think of another reason why it would act like this.

I wouldn't think there's an intentional addition of weight; but I wouldn't be too surprised if there was some sort of bug where a block can "vanish" momentarily from SI calc during upgrade. That would affect any and all at-limits upgrades though, so at least it would need to be rare. Can you reliably repro the original collapse in creative, or is the build just too complex to try?

[meganoth]

Well, that's a surprise. If it was already in A20 I am even more surprised. I remember checking on these values some time ago and it to be more than 2 years ago is the embodiment of "time flies away"

 

Another bug possibility: Maybe the algorithm doesn't always find the best support for a specific block and then drops the ball and with it the blocks.

 

[Riamus]

27 minutes ago, Stranded_Napkin said:

If you upgrade at the supports and then progressively move away from them, it doesn't seem to cause any collapse.

That is why I always upgrade from support outwards.  I don't want to risk any collapse by not doing it that way and since it doesn't take any more time to do it that way, it just makes sense.  😀

[Stranded_Napkin]

I did some in game testing. I can only guess at what's causing it. The concrete and the cobble blocks hold the same. Hell, I even tried to look at the code (turns out it's not in the xml, figures though). I figured it might be that extra mass thing. I built multiple supports with horizontal blocks built out to max.

 

First one, was control so I upgraded normally.

Second one, I upgraded from the end back to the support. Nothing.

Third one, I upgraded randomly. Nothing.

Fourth one, I built a support, 12 blocks off of it, a second support offset from it, connected the two paths and then built up to max. Upgraded end block first, then randomly. Nothing.

 

I am now thinking it's just the phantom block issue. Sometimes random blocks just have issues. My luck, it's the ghost of a tree haunting me.

14 minutes ago, Riamus said:

That is why I always upgrade from support outwards.  I don't want to risk any collapse by not doing it that way and since it doesn't take any more time to do it that way, it just makes sense.  😀

Yeah, I just got lazy twice and it cause me frustration in the end. One would think that I'd have learned my lesson by now. The base is still in mostly working order. Small fix to my pathway and then the side hallway that allowed for the electric fences to be repaired. Could always add some extra supports as well. 

[meganoth]

It wouldn't be easy to reproduce with a few test blocks since SI is tested daily by hundreds of players and the code for it seems not to have been changed for years.

 

Your only hope is to use the building where you seem to have had the problem and try to recreate by reenacting what you did last time.

 

Edited August 20, 2023 by meganoth (see edit history)

[Riamus]

6 minutes ago, meganoth said:

It wouldn't be easy to reproduce with a few test blocks since SI is tested daily by hundreds of players and the code for it seems not to have been changed for years.

 

Your only hope is to use the building where you seem to have had the problem and try to recreate by reenacting what you did last time.

 

A good way to test would be to save the base as a POI with that problematic area left not built.  Then place the base somewhere else and see if you have the same problem.  If not, it could be some support issue being caused by the terrain under it.

[pApA^LeGBa]

I am quite surprised that it is like that for a few alphas now. I could swear that at least in A19 they weren´t the same and also was quite sure they were still different in A20.

 

Or do we have a mandela effect here, due to the long wait for A21?

Edited August 20, 2023 by pApA^LeGBa (see edit history)

[Rotor]

Although the last reference here:

https://7daystodie.fandom.com/wiki/Structural_Integrity

Says A15, i thought there was thread where one of the moderators said they would update it.  But that was like last year and I could be confusing things.

[meganoth]

The block system was updated, sure, for example brick and cobblestone merged and their weights changed at another time. The latter is a very simple xml change that can't really introduce any bugs, at max it could unearth already existing ones in the code.

But I don't remember anyone saying in recorded history 😉 that the SI algorithm itself would be changed or looked at.

 

[Toops144]

I'm

3 hours ago, Riamus said:

That is why I always upgrade from support outwards.  I don't want to risk any collapse by not doing it that way and since it doesn't take any more time to do it that way, it just makes sense.  😀

Yeah this is what I do too.

 

I also test things out in the debug mode to see the various stress levels to highlight problematic blocks, as well as to experiment with structures before implementing them in the 'real world'.