Most temples (at least of the types "witch hut", "desert temple", "jungle temple", and "igloo") are not being generated when the chunks containing them are generated. Experiments using 1.18.1 show they were being generated at that time, but they no longer are.
Steps to Reproduce:
Note: After navigating to a location, you can use /locate temple to verify that a temple is supposed to generate there.
In 1.18.10, create a new world with seed 5522698.
Navigate to -280, 312. Verify that there is no desert temple there,
Navigate to 312, 72. Verify that there is no desert temple there,
Navigate to -264, -392. Verify that there is no witch hut there.
Navigate to 216, -280. Verify that there is no witch hut there.
Expected Results:
Most temples generate where expected, though a small percentage may be canceled because of terrain incompatibility, conflict with other structures, etc.
Observed Results:
The number of missing temples within 1000-2000 blocks of spawn is larger than expected.
Additional Information:
A test using a different seed with a more extensive list of missing temples can be found below in a comment. In this test, almost all of the temples are missing.
Original description:
When using the /locate temple there is frequently no structure at the location given. In particular the desert temple, jungle temple and witch hut structures often don't generate.
Steps to Reproduce:
Load a new world with the seed 5522698
/tp -280 100 312
/locate temple
Note that the coordinates are the same and no structure exists
/tp 312 100 72
/locate temple
Note that the coordinates are the same and no structure exists
/tp -264 100 -392
/locate temple
Note that the coordinates are the same and no structure exists
/tp 216 100 -280
/locate temple
Note that the coordinates are the same and no structure exists
Observed Results:
At all 4 locations the structure didn't generate where the locate command says it should.
Expected Results:
Structures should exist at the locations the /locate command outputs
Notes:
This issue seems to have been introduced at some point between 1.18.0.24 and 1.18.10.20. I loaded up a seed which I had previously noted had a witch hut at a specific location and its not there in 1.18.10.20.
Linked issues
is duplicated by 11
Attachments
Comments 21
Something people don't always understand is that the /locate command doesn't tell you where structures are generated in the world. Instead, it tells you where an attempt will be made to generate them. These locations are calculated from the world seed, but they can't be 100% accurate because sometimes the calculated location can be unsuitable for that type of structure, for some reason. This doesn't happen very often, but it's not rare, either.
So when a single /locate command points to a location but there's no structure there, that isn't necessarily a bug. To decide whether it is, you have to check a bunch of locations. If many of the structures aren't generated, that's a good sign there's a bug involved.
The original description on this report gave 4 examples, which would have been pretty convincing on its own but we already had the similar issue in MCPE-141249 and it misdirected us. We now think the issue in 141249 was entirely different and was related to the fact that world generation was in the process of changing when that report was written. That change was completed in 1.18, and this problem (which we've tested and confirmed with a total of around 20 missing temples now) began after 1.18 was released, so we're ready now to confirm this and submit it to Mojang. I will be modifying the summary and description because we understand now that this is a problem with structure generation, not with the /locate command.
More examples of missing temples
Note that this is not a curated list, it's just locations near spawn picked at random. For each location, I teleported there, then issued a /locate temple command to verify that I was at the coordinates where a temple should be generated. Of these 16 locations, the only one that actually generated a temple was the first jungle temple location.
Seed: 1889557986
Jungle temple locations:
-2472, 72
2648, -5928
Witch hut locations:
-1000, 312
-4328, 1176
2792, -3016
-792, -2360
3144, -2536
2936, -2440
-664, -4872
-968, -5944
-680, -5272
-280, -5592
24, -5464
Desert temple locations:
-1528, 1192
-1720, 1272
Igloo location:
872, 8
I also tested these locations using the 1.18.1 release. All of these temples generated except for the igloo, which was canceled because the terrain was hilly, grassy, and tree-covered (possibly as a result of biome blending).
Actually I've noticed while trying to push the limits of 64 bit seeds that something seems not quite right. Specifically I've been using (cough abusing) the fact dripstone and lush caves are allowed biomes for woodland mansions to try to create a 6 way structure collision.
Anyway what I've noticed is the high occurrence of temples not being there where the /locate command says they should. While I do agree with Auldrick somewhat regarding the locate command not being perfect, I do suspect there are things that can be done to help align these things. In general, structures appearing in /locate but not generating would lead me to think the structure is passing the "isFeatureChunk" check but failing in "postProcess". I'm certainly happy to help if more digging is required.
Major, is that on 1.18.12? If so it probably won't work until 1.18.30 releases, which given previous release patterns could be this Tuesday.
Note that temples are generated only during chunk generation. This means that if this bug prevented a temple from generating in 1.18.12 or earlier, it's not going to suddenly be there after you upgrade the world to 1.18.30. Also, if you did a lot of checking for temples in your survival world, you caused the chunks they should have been in to be generated without them, so those temples will never exist in that world, which could mean after upgrading to 1.18.30 you'll have to travel even farther to find new temples. You might want to consider restoring the world from a backup taken before you did all that exploring, even if it means some work is lost.
Update
In Bedrock Edition Release 1.18.30, 3 of the 4 listed temples that were not generated are generating now in the latest version, so it means this issue seems to be partially fixed.
–
Reply from @unknown:
We let the developers decide when an issue is fixed. This is mainly because an issue may appear fixed to you but still be affecting somebody with a different world. So while we appreciate that you want to help, comments like this aren't really helpful.
Also, as far as this particular issue goes, the fact that one temple didn't generate for you doesn't by itself prove that the issue isn't fixed. The world seed is used to generate a bunch of coordinate sets where the game will try to generate temples, but sometimes it's not possible to generate a particular temple near those coordinates. In that case, world generation just skips it. This isn't a bug, it's a consequence of two things: the extraordinarily vast variability of how nearby biomes can affect one another's generations, and the fact that /locate typically calculates the coordinates before the chunks in the area are even generated (since you haven't visited the temple yet), so /locate doesn't actually know whether the game will be able to generate the temple there.
Got emailed comments on this issue that don't appear to be here! anyway, it looks to me like this issue is fixed also, but will wait for official confirmation, It definitely has changed in 1.18.30 and for the better.
Reply from @unknown:
We removed that and similar contents because they're not helpful and might be misleading. We know people are just trying to be helpful, but we can't assume that if it's fixed on your platform and device then it must be fixed for everybody. We have to rely on Mojang to tell us when it's fixed, which they do by listing the report on the changelog. They didn't list this ticket yet, so we have to assume they still have more work to do.
So thank you for being involved, but please don't try and tell us when to close our reports.
This report was initially resolved as a duplicate of MCPE-141249. After further study and additional reports of the issue, we have realized that there is a separate issue here that began in 1.18.10 (or possibly 1.18.2) but was definitely not occurring in the 1.17 releases affected by the other issue. We are therefore restoring this report and will be using it to control the newly recognized issue.