Does the "optimize world" function with clear cache turned on have a risk of removing entities? I'm asking because we've just had the second case of all entities disappearing from certain, seemingly random chunks. Horses, armour stands, maps on item frames, paintings. In both cases, the removal was permanent and was discovered shortly after the "optimization" was completed.
Yesterday we discovered that multiple chunks in the Nether - nearly half of the chunks near 0,0 - have gone dark with this bug. Again, ours is a world that was updated straight from 1.13.2 to 1.14 - and did not experience any of the snapshots. This bug is now reoccurring at too fast a pace for me to just keep fixing it every few days with the "Optimize World" function. Strangely, only the same region of the world, my city project to be precise, seems to be getting this bug with any regularity. The bug that occurred on the Nether is the only one that hasn't occurred in the same area. I can also confirm that the lighting bug is not connected to slime chunks in our case.
I too am very disappointed that the 1.14.1 pre-release has no mention of the lighting bug. Even though I can sympathize with the issue that the bug is difficult to reproduce on demand - my own attempts at reproducing it on new worlds has not met with success - the bug itself is happening rather too often in my main world for me to be appreciative of efforts that are yet fruitless.
I am also seeing this in 1.14; we updated from 1.13.2 to 1.14 full release.
EDIT: I applied the fix outlined in the bug header and it fixed the lighting bug for me.
RE-EDIT: The bug returned after a while, and I suspect this will keep reoccurring.
I have tried, up to this point, restarting the game, restarting both the game and the launcher, restarting my internet router, restarting my computer, and also re-installing the game.
The game console output looks like this:
Completely ignored arguments: [--nativeLauncherVersion, 286]
[19:12:23] [Client thread/INFO]: Setting user: meto30
[19:12:23] [Client thread/INFO]: (Session ID is token:566d687b522d4a44af6eb81f941d665e:79fbdb6a007447c6b6555870da44e7aa)
[19:12:25] [Client thread/INFO]: LWJGL Version: 2.9.4
[19:12:26] [Client thread/INFO]: Reloading ResourceManager: Default
[19:12:27] [Sound Library Loader/INFO]: Starting up SoundSystem...
[19:12:27] [Thread-6/INFO]: Initializing LWJGL OpenAL
[19:12:27] [Thread-6/INFO]: (The LWJGL binding of OpenAL. For more information, see http://www.lwjgl.org)
[19:12:27] [Thread-6/INFO]: OpenAL initialized.
[19:12:27] [Sound Library Loader/INFO]: Sound engine started
[19:12:30] [Client thread/INFO]: Created: 512x512 textures-atlas
[19:12:32] [Realms Notification Availability checker #1/INFO]: Could not authorize you against Realms server: Invalid session id
[19:12:38] [Client thread/INFO]: Realms library version == 1.7.48
[19:12:38] [MCO Availability Checker #1/INFO]: Could not authorize you against Realms server: Invalid session id
[19:12:38] [MCO Availability Checker #1/ERROR]: Couldn't connect to realms:
I have this problem as well. Could not log in for the last few hours.
Welp, ocean monument is there on those exact coordinates on a regenerated 1.8 world, so... yeah, looks like I just made a dupe report. Sorry about the mess.
Hm, I went to another Stronghold, also under the sea, to check if this happened there too, but no guardians spawning there. Therefore, it does seem like a dupe of MC-65214 as Sonic has pointed out... I'll try generating the world with the same seed to check if a temple is genned on the coordinates.
After prohibiting all use of Nether Portals as a temporary measure, I'm pleased to note that there have been no new occurrences of the block light bug in my server. We're still running 1.14 full. As such, I think I can also testify, albeit indirectly, that using the portal is linked to the bug. Perhaps related to the massive lag spike? None of the other methods of travel - including long-distance teleport commands - have yet to cause chunks to go dark. It's been a week since we prohibited the use of portals.