Confusing error: ServerEnvironment::deactivateFarObjects()

[Wuzzy]

In MineClone 2 (and maybe other subgames, not sure), I sometimes get these error messages in chat:

2017-05-27 12:24:29: ERROR[Server]: ServerEnvironment::deactivateFarObjects(): id=24 m_static_exists=true but static data doesn't actually exist in (-9,-3,-41)
2017-05-27 12:24:29: WARNING[Server]: StaticObjectList::remove(): id=24 not found
2017-05-27 12:24:29: ERROR[Server]: ServerEnvironment::deactivateFarObjects(): id=23 m_static_exists=true but static data doesn't actually exist in (-9,-3,-41)
2017-05-27 12:24:29: WARNING[Server]: StaticObjectList::remove(): id=23 not found

Can somebody please explain me what is going on here? When does this occour, what does it mean and how can I fix the problem?

[theFox]

I'm also having this quite often.
Looks like the Static data of some objects have been deleted or never existed, but the data is listed in m_static_exists does.

[azekill_DIABLO]

it does when i unload chuncks and then deconnect ... the data of some objects still exists but not them.

[theFox]

In MineClone 2 (and maybe other subgames, not sure)

I had a world with only my factory mod enabled and got these messages.
I guess (dropped) items cause this problem.

[azekill_DIABLO]

all entities does it.

[Wuzzy]

There are GitHub issues for this weird error message:

https://github.com/minetest/minetest/issues/6249
https://github.com/minetest/minetest/issues/194

I still have no real idea what is going on. I talked to core developers, they told me that this is probably a bug in the engine, but this is not 100% sure. They don't want to exclude the possibility of a mod bug either.
I am also pretty sure that this error message is related to entities.
And this is all I know.

If you want to help research this bug, you could maybe tell us if you know of any mod in which the error occours (and under which circumstances).
Maybe this helps us in isolating this bug.

[Lejo]

I also get this on my own subgame.
So I think it's bug in the engine.
On my server there is often auto done /clearobjects quick
may this doesn't clears the m_static_exists

[joe7575]

I have the same issues. I identified a luacontroller with connected digiline display as source.
The simple program:

Code: Select all

interrupt(1)
digiline_send("lcd", "Test")

in an unloaded area leads to this errors.

When you leave the area, the error messages are starting like this:

Code: Select all

2018-04-23 19:34:14: WARNING[Server]: StaticObjectList::remove(): id=277 not found
2018-04-23 19:34:14: ERROR[Server]: ServerEnvironment::deactivateFarObjects(): id=278 m_static_exists=true but static data doesn't actually exist in (14,0,76)
2018-04-23 19:34:14: WARNING[Server]: StaticObjectList::remove(): id=278 not found
2018-04-23 19:34:16: WARNING[Server]: ServerEnv: Trying to store id = 280 statically but block (14,0,76) already contains 64 objects. Forcing delete.
2018-04-23 19:34:18: WARNING[Server]: ServerEnv: Trying to store id = 281 statically but block (14,0,76) already contains 64 objects. Forcing delete.
2018-04-23 19:34:18: WARNING[Server]: ServerEnv: Trying to store id = 282 statically but block (14,0,76) already contains 64 objects. Forcing delete.

And when you come back, it looks like:

Code: Select all

2018-04-23 20:09:39: WARNING[Server]: ServerEnv: Trying to store id = 1053 statically but block (14,0,76) already contains 65 objects. Forcing delete.
2018-04-23 20:09:41: WARNING[Server]: ServerEnv: Trying to store id = 1055 statically but block (14,0,76) already contains 65 objects. Forcing delete.
2018-04-23 20:09:43: ERROR[Server]: suspiciously large amount of objects detected: 65 in (14,0,76); removing all of them.

The luacontroller keeps running even if the display is not available.

But I was able to reproduce the same errors with my SmartLine controller and display.
When the display is far away in an unloaded area and the controller writes cyclically to that display, I will get the same warnings.

It seams to be a minetest core bug

[gpcf]

I also can reproduce this bug. a quick fix would be stopping the lua controllers if they get unloaded, but this is tricky since the lua controllers themselves don't seem to get unloaded.

[joe7575]

I also can reproduce this bug. a quick fix would be stopping the lua controllers if they get unloaded, but this is tricky since the lua controllers themselves don't seem to get unloaded.

I revised the Lua controller, so that it gets unloaded, but this is hacky somehow...

[gpcf]

can you send me the mod? I failed while trying to get the luacontroller unloaded.

[joe7575]

can you send me the mod? I failed while trying to get the luacontroller unloaded.

You got a PM