sfence

This is still blocking #14129. Would be nice if core developers found a conclusion about safe/unsafe to store passwords in pure text and merge or close this PR. It makes...

@HybridDog Thanks for checking. Should be fixed now.

Windows build failed due to: ``` CMake Error at irr/src/CMakeLists.txt:267 (message): SDL2 is too old, required is at least 2.0.10! ```

If I get it well, APi of Guids should be updated, related to https://github.com/minetest/minetest/pull/11050#issuecomment-824519570 (@sorcerykid or anybody else)?

> (Unofficial conceptual review, I'm not a core dev) > > Concept looks good to me. However the implementation of the GUID generator feels overengineered. Why not use random UIDs,...

So I have made some changes, based on the discussion here and under the previous PR. `get_guid` returns the player name for the player object. GUId for luaentities looks like...

@fan5 Use of UUIDv4 will end with code similar to: https://github.com/crashoz/uuid_v4/blob/master/uuid_v4.h So, it will be probably again over-engineered as @appgurueu mentioned before. And much worse performance can be expected from...

So, last commit... GUId generator state is no longer stored in `env_meta.txt` file. The return value from `porting::getTimeS()` function is used as the initial counter value after the server starts...

So, I have done some research about this. It looks like it should be ok to use `std::random_device` when `std::random_device::entropy` returns a non-zero value. But some go around should be...

So a new variant is available. It uses std::random_device. Usage depends on the entropy of the random source to eliminate the risk of duplicates on systems with poor random_device.