openmw issueshttps://gitlab.com/OpenMW/openmw/-/issues2023-11-29T15:02:18Zhttps://gitlab.com/OpenMW/openmw/-/issues/6371Crash To Desktop entering Redoran Council in shadows ConvexHull findInternalE...2023-11-29T15:02:18ZEvgeny FursovCrash To Desktop entering Redoran Council in shadows ConvexHull findInternalEdges with NaNsOpenMW-0.47.0-win64
Running on win10
Heavily modded
CTD attempting to enter Redoran Council under the Scar. Tried again couple dozen times from the last save, reproduced only once. Attaching crash dump.OpenMW-0.47.0-win64
Running on win10
Heavily modded
CTD attempting to enter Redoran Council under the Scar. Tried again couple dozen times from the last save, reproduced only once. Attaching crash dump.https://gitlab.com/OpenMW/openmw/-/issues/6362Terrain artifacts along tile borders2023-02-26T20:29:39ZCody GlassmanTerrain artifacts along tile bordersThere are noticeable seams around terrain tiles, confirmed by at least @cmaugust. This is not to do with textures at all, disabling texturing andshading the terrain a solid color gives the same artifacts. Presumably it's more to do with ...There are noticeable seams around terrain tiles, confirmed by at least @cmaugust. This is not to do with textures at all, disabling texturing andshading the terrain a solid color gives the same artifacts. Presumably it's more to do with having some fight between vertices with exact position, or it could be some precision issue, I am unsure of cause.
MSAA helps with these artifacts, but it is still present. All photos/videos are showing terrain rendered pure black to better highlight the seams. If this is a regression it is not a recent one, as these artifacts appear in versions dating back to at least 0.39 release.
To debug with black terrain simply add this to main function in `terrain_fragment.glsl`:
```
void main()
{
gl_FragData[0] = vec4(0,0,0,1);
return;
```
This is quite hard to see on a video capture, but I've done my best to capture a small segment where you see the artifacts, which again happen happen along the tile borders. This can be confirmed by enabling `debug chunks` and observing that the artifacts disappear, this is because the debug geometry is hiding the seams.
https://streamable.com/fxdrgi
![screenshot039](/uploads/22cc6ffdf12f1476a5f0356096640045/screenshot039.png)https://gitlab.com/OpenMW/openmw/-/issues/6359Refraction and non-refraction shader water look extremely different2021-11-03T14:22:49ZwareyaRefraction and non-refraction shader water look extremely differentThe refraction toggle in the water shader options is not just a refraction toggle. It changes a lot of different techniques that the water shader uses around, and also their mixing/blending values, for reasons that have nothing to do wit...The refraction toggle in the water shader options is not just a refraction toggle. It changes a lot of different techniques that the water shader uses around, and also their mixing/blending values, for reasons that have nothing to do with refraction itself. For example, refraction water is significantly less reflective than non-refractive water, even in areas of the visible water surface where the refracted underwater areas are invisible and have no influence on what the player can see.
If both versions are desirable, each difference should be individually configurable. Otherwise, the two versions of shaded water should be brought in line and unified. The only differences should be 1) whether the underwater surface has actual refraction effects applied 2) applying fog below the water surface in refraction mode (since doing so for non-refraction water would be kind of difficult).
This issue is distinct from the various general refraction and non-refraction water bugs: it only concerns the fact that they are *in general* so different, not the individual differences.https://gitlab.com/OpenMW/openmw/-/issues/6352FPS drop when standing on unoptimized modded large outdoor urns2023-07-13T06:09:10ZDave CousineauFPS drop when standing on unoptimized modded large outdoor urnsBasically just standing on any urn (as found in cities; not sure about the urns that are in ancestral tombs) makes the FPS drop from ~75 all the way down to 15-20. It does not happen when standing on crates or baskets.
On the F3 graph t...Basically just standing on any urn (as found in cities; not sure about the urns that are in ancestral tombs) makes the FPS drop from ~75 all the way down to 15-20. It does not happen when standing on crates or baskets.
On the F3 graph thing, it seems to be related to "phys" which goes from about 2 to about 258.
On the F10 profiler it says, I think, that the convexSweepTest takes about 240-260ms. (When not standing on the urn it's 3-5ms.)
This does not happen in my non-OpenMW, though it's pretty far from true vanilla (MW FPS Optimizer + Code Patch).
This is OpenMW 0.46.
An example urn in Ald'ruhn:
```
Report time: 2021.10.20 05:51:44 UTC
Content file: Morrowind.esm
RefNum: 49298
RefID: urn_02
Memory address: 000001A1480EBD20
Cell: Ald-ruhn
Grid: -3 6
Coordinates: -16711.4 54538.4 2046.72
Model: meshes\o\Contain_urn_02.NIF
```
I notice that there are three different urn models here and all three cause the issue.https://gitlab.com/OpenMW/openmw/-/issues/6318Storage::getHeightAt doesn't take in account which way the triangles are facing2022-01-25T14:39:36ZNelsson HuotariStorage::getHeightAt doesn't take in account which way the triangles are facingThis may result in errors when getting the terrain heights.
https://gitlab.com/OpenMW/openmw/-/blob/master/components/esmterrain/storage.cpp#L497This may result in errors when getting the terrain heights.
https://gitlab.com/OpenMW/openmw/-/blob/master/components/esmterrain/storage.cpp#L497https://gitlab.com/OpenMW/openmw/-/issues/6317Navigator and collision system treat slopes in a different way2022-02-15T07:48:55ZAndrei KortunovNavigator and collision system treat slopes in a different wayAn example is Infidelities quest from Tribunal. Marked area is not walkable, despite navigator assumes that it is, so Velyna Seran stuck:
![Screenshot_20210930_121622](/uploads/b49b0796e3f4cf8b1cf678761dc18a6b/Screenshot_20210930_121622...An example is Infidelities quest from Tribunal. Marked area is not walkable, despite navigator assumes that it is, so Velyna Seran stuck:
![Screenshot_20210930_121622](/uploads/b49b0796e3f4cf8b1cf678761dc18a6b/Screenshot_20210930_121622.png)
When actors stuck, it can break quests, so we need somehow either to make sure that navigator matches collisions system or add some kind behaviour to AI to work around this bug.
Note that in case of Infidelities quest this bug is masked by bug #6307 in 0.47.
The bug can probably be a some kind of precision mismatch - a normal z part in the `isWalkableSlope` is 0.655944 in marked area, while it needs to be more than 0.656059 to be walkable.https://gitlab.com/OpenMW/openmw/-/issues/6293Crashing on latest nightly when looking straight ahead/downwards at buildings...2022-11-02T20:26:06ZCaleb HixonCrashing on latest nightly when looking straight ahead/downwards at buildings in VivecThis has been pretty strange, I've been using my laptop more recently and just noticed that whenever looking at different buildings in Vivec, especially straight on or downwards, the game freezes, stutters side-to-side, then after clicki...This has been pretty strange, I've been using my laptop more recently and just noticed that whenever looking at different buildings in Vivec, especially straight on or downwards, the game freezes, stutters side-to-side, then after clicking a bunch of different keys, I start getting these graphical errors/artifacts, I then have to hard reboot the laptop. To go in the buildings, I need continuously look upwards and strafe my character towards the door I need to go into. I have modded it using the I Heart Vanilla mod-list with not many other mods. I don't get these crashes anywhere else in the game as far as I can tell. This is on pop_os 21.04. No errors in openmw.log as far as I can tell.
![viveccrash_smallererest](/uploads/852a2ec56cecb23c42b283c04d792c70/viveccrash_smallererest.mp4)https://gitlab.com/OpenMW/openmw/-/issues/6267Frost salt mesh from "Fire and Frost Salts" doesn't render correctly2022-08-01T06:10:48ZHristos N. TriantafillouFrost salt mesh from "Fire and Frost Salts" doesn't render correctlyThe frost salt mesh included in the [Fire and Frost Salts](https://www.nexusmods.com/morrowind/mods/47784) mod renders as a purple box:
![Screenshot_2021-09-07_17-10-56](/uploads/9f8020cbdf706a48ecf6599669c3964d/Screenshot_2021-09-07_17...The frost salt mesh included in the [Fire and Frost Salts](https://www.nexusmods.com/morrowind/mods/47784) mod renders as a purple box:
![Screenshot_2021-09-07_17-10-56](/uploads/9f8020cbdf706a48ecf6599669c3964d/Screenshot_2021-09-07_17-10-56.png)
It renders fine in Morrowind.exe and also looks OK to me in Nifskope (though, I only opened it and visually inspected it). There's nothing special in the openmw log relating to this.
OpenMW sha: aef6ec464https://gitlab.com/OpenMW/openmw/-/issues/6260Cell loading causes visibly longer frame duration2023-10-29T12:54:29ZelsidCell loading causes visibly longer frame durationWhen player crosess a cell border engine unloads and deactivates some cells, loads and activates other. Ideally all content should be preloaded but in practice it doesn't work good enough. With enabled `OPENMW_OSG_STATS_FILE` it looks li...When player crosess a cell border engine unloads and deactivates some cells, loads and activates other. Ideally all content should be preloaded but in practice it doesn't work good enough. With enabled `OPENMW_OSG_STATS_FILE` it looks like this:
![cell_loading](/uploads/1800c770abf476b33d8dbf4e3f05fa61/cell_loading.png)
Single spike:
![cell_loading_zoom](/uploads/36ef5db5c2a9adb0db426189b618b06d/cell_loading_zoom.png)
There are multiple reasons for it and they are not related to the physics system directly. It just happened to be that `PhysicsSystem::moveActors` indirectly calls `Scene::changeCellGrid`:
1. Terrain preloading may not finish when cell is started to load. Here is a sample of a frame when this happens:
| function | duration | calls number | duration, % (frame) |
| ------------------------------------------------------------------------------------- | ----------------------- | ----------------------- | ----------------------- |
| frame@/home/elsid/dev/openmw/apps/openmw/engine.cpp:257 | 143041 | 1 | 100.0 |
| moveActors@/home/elsid/dev/openmw/apps/openmw/mwphysics/physicssystem.cpp:812 | 140914 | 1 | 98.51301375130207 |
| changeCellGrid@/home/elsid/dev/openmw/apps/openmw/mwworld/scene.cpp:598 | 140807 | 1 | 98.4382100236995 |
| preloadTerrain@/home/elsid/dev/openmw/apps/openmw/mwworld/scene.cpp:1283 | 93437 | 1 | 65.32183080375556 |
| insertCell@/home/elsid/dev/openmw/apps/openmw/mwworld/scene.cpp:1025 | 35777 | 8 | 25.011709929320965 |
| activateCell@/home/elsid/dev/openmw/apps/openmw/mwworld/scene.cpp:400 | 19451 | 3 | 13.598199117735474 |
| loadInactiveCell@/home/elsid/dev/openmw/apps/openmw/mwworld/scene.cpp:510 | 17426 | 5 | 12.182521095350284 |
| addObject@/home/elsid/dev/openmw/apps/openmw/mwphysics/physicssystem.cpp:493 | 15131 | 824 | 10.578086003313736 |
| getInstance@/home/elsid/dev/openmw/components/resource/bulletshapemanager.cpp:192 | 7925 | 824 | 5.540369544396362 |
| placeObject@/home/elsid/dev/openmw/apps/openmw/mwworld/worldimp.cpp:1436 | 5074 | 4 | 3.547234708929608 |
| createInstance@/home/elsid/dev/openmw/components/resource/bulletshapemanager.cpp:204 | 3909 | 413 | 2.732782908396893 |
| getPagedRefnums@/home/elsid/dev/openmw/apps/openmw/mwrender/renderingmanager.cpp:1450 | 1942 | 1 | 1.3576527009738466 |
| getPagedRefnums@/home/elsid/dev/openmw/apps/openmw/mwrender/objectpaging.cpp:825 | 1934 | 1 | 1.3520598989101027 |
| deactivateCell@/home/elsid/dev/openmw/apps/openmw/mwworld/scene.cpp:343 | 1863 | 16 | 1.302423780594375 |
| objectAddedToScene@/home/elsid/dev/openmw/apps/openmw/mwlua/luamanagerimp.cpp:219 | 1681 | 450 | 1.1751875336442 |
2. Not all meshes are cached. `createInstance` is slow when there is cache miss:
| function | duration | calls number | duration, % (frame) |
| ------------------------------------------------------------------------------------- | ----------------------- | ----------------------- | ----------------------- |
| frame@/home/elsid/dev/openmw/apps/openmw/engine.cpp:257 | 66548 | 1 | 100.0 |
| moveActors@/home/elsid/dev/openmw/apps/openmw/mwphysics/physicssystem.cpp:812 | 65223 | 1 | 98.008955941576 |
| changeCellGrid@/home/elsid/dev/openmw/apps/openmw/mwworld/scene.cpp:598 | 65156 | 1 | 97.908276732584 |
| insertCell@/home/elsid/dev/openmw/apps/openmw/mwworld/scene.cpp:1025 | 49924 | 8 | 75.01953477189397 |
| loadInactiveCell@/home/elsid/dev/openmw/apps/openmw/mwworld/scene.cpp:510 | 36194 | 5 | 54.38781030233816 |
| addObject@/home/elsid/dev/openmw/apps/openmw/mwphysics/physicssystem.cpp:493 | 32667 | 776 | 49.08787642002765 |
| getInstance@/home/elsid/dev/openmw/components/resource/bulletshapemanager.cpp:192 | 25500 | 776 | 38.31820640740518 |
| createInstance@/home/elsid/dev/openmw/components/resource/bulletshapemanager.cpp:204 | 21262 | 557 | 31.949870769970545 |
| activateCell@/home/elsid/dev/openmw/apps/openmw/mwworld/scene.cpp:400 | 14600 | 3 | 21.93905151169081 |
| placeObject@/home/elsid/dev/openmw/apps/openmw/mwworld/worldimp.cpp:1436 | 7274 | 4 | 10.930456212057463 |
| preloadTerrain@/home/elsid/dev/openmw/apps/openmw/mwworld/scene.cpp:1283 | 5765 | 1 | 8.66291999759572 |
| getPagedRefnums@/home/elsid/dev/openmw/apps/openmw/mwrender/renderingmanager.cpp:1450 | 1761 | 1 | 2.6462102542525696 |
| getPagedRefnums@/home/elsid/dev/openmw/apps/openmw/mwrender/objectpaging.cpp:825 | 1753 | 1 | 2.6341888561639717 |
| objectAddedToScene@/home/elsid/dev/openmw/apps/openmw/mwlua/luamanagerimp.cpp:219 | 1276 | 379 | 1.917412995131334 |
| unloadInactiveCell@/home/elsid/dev/openmw/apps/openmw/mwworld/scene.cpp:317 | 1211 | 5 | 1.8197391356614774 |
There are might be other problems but so far mentioned two are hiding them.
See https://gitlab.com/OpenMW/openmw/-/issues/6260#note_898206200 for the most recent state.https://gitlab.com/OpenMW/openmw/-/issues/6257Enchanted darts cause momentary game freeze2023-05-13T00:24:54ZAbdu SharifEnchanted darts cause momentary game freezeThis doesn't seem to happen with 0.46, but it affects both 0.48 and the 0.47 RC build I have:
![projectile_freeze](/uploads/9da93e335da82cdb8f25c7e0102cd308/projectile_freeze.mp4)
A save to reproduce the issue with:
[Ranged_Problem.omws...This doesn't seem to happen with 0.46, but it affects both 0.48 and the 0.47 RC build I have:
![projectile_freeze](/uploads/9da93e335da82cdb8f25c7e0102cd308/projectile_freeze.mp4)
A save to reproduce the issue with:
[Ranged_Problem.omwsave](/uploads/300e994e7515eb42b25c931e6e5c2f2c/Ranged_Problem.omwsave)https://gitlab.com/OpenMW/openmw/-/issues/6245OpenCS - incorrect implementation for moved references2022-01-25T15:04:14Zcc9ciiOpenCS - incorrect implementation for moved referencesMR !1051 attempted to fix #3514 and #4752. However, the implementation may have wrongly generated a deleted reference.
The attached file was generated with vanilla CS. Two objects were moved by editing their positions. In neither cas...MR !1051 attempted to fix #3514 and #4752. However, the implementation may have wrongly generated a deleted reference.
The attached file was generated with vanilla CS. Two objects were moved by editing their positions. In neither case `DELE` records were generated.
[move_test.ESP](/uploads/ed54fa8a09016303a525e225384633e5/move_test.ESP)
The test file loads correctly in OpenMW but not in OpenCS (the `crate_01_food_misc01` is still shown in cell -3,-2).https://gitlab.com/OpenMW/openmw/-/issues/6240State sharing prevents the use of the same texture file for different purpose...2024-02-21T05:42:43ZDaniil BaturinState sharing prevents the use of the same texture file for different purposes in shadersOpenMW built from commit 1a2efd441cee2f255b0b5586a72e785538b6dd19
The [Mace of Aevar the Stone-Singer](https://en.uesp.net/wiki/Bloodmoon:Mace_of_Aevar_Stone-Singer) from Bloodmoon is supposed to have a "vapor" animation, but it's not d...OpenMW built from commit 1a2efd441cee2f255b0b5586a72e785538b6dd19
The [Mace of Aevar the Stone-Singer](https://en.uesp.net/wiki/Bloodmoon:Mace_of_Aevar_Stone-Singer) from Bloodmoon is supposed to have a "vapor" animation, but it's not displayed in recent builds.
![Screenshot_at_2021-08-24_00-41-19](/uploads/19c904f8c5ef0dff3e0235a19727268b/Screenshot_at_2021-08-24_00-41-19.png)https://gitlab.com/OpenMW/openmw/-/issues/6239Launcher is unaware of --replace2023-11-22T11:28:10ZAnyOldName3Launcher is unaware of --replaceFor list-like settings (e.g. the `data` and `content` lists) you can add them to the `replace` list and then instead of composing them from lower-priority config sources, the list will be replaced. For example, if you specify the followi...For list-like settings (e.g. the `data` and `content` lists) you can add them to the `replace` list and then instead of composing them from lower-priority config sources, the list will be replaced. For example, if you specify the following command line, you'll just end up with `Morrowind.esm` loaded no matter what's in the user, global and local `openmw.cfg`:
```
openmw --replace content --content Morrowind.esm
```
This can also be done via the config files, so if you specify `replace=content` in the user `openmw.cfg`, it'll knock out any content files specified in the global and local `openmw.cfg`s.
If you do that, while the engine will ignore the relevant entries, the launcher will still treat them as any others, and display the contents of any data directories and tick any content files from the low-priority source.
I'd intended to implement this as part of !1076, but it got merged before that happened, and then I've not had time to do it since then. I'm putting it on the tracker so it doesn't get forgotten forever.https://gitlab.com/OpenMW/openmw/-/issues/6226Curing disease removes it from the spell list instead of just removing its ef...2023-07-13T06:07:48Zcc9ciiCuring disease removes it from the spell list instead of just removing its effects~~The importer simply copies `NPCS` sub-record into `ESM::NPC::mSpells`. But vanilla `.ess` savefile includes *all* spells the player had ever known, including diseases.~~
~~Vanilla must be re-evaluating the current disease status else...~~The importer simply copies `NPCS` sub-record into `ESM::NPC::mSpells`. But vanilla `.ess` savefile includes *all* spells the player had ever known, including diseases.~~
~~Vanilla must be re-evaluating the current disease status elsewhere, whereas OpenMW simply uses whatever is in `ESM::NPC::mSpells`.~~
~~A quick fix might be to remove all entries that are not found under `SPLM` record but then we may need to store the lost info somewhere else.~~
As per the new title.cc9ciicc9ciihttps://gitlab.com/OpenMW/openmw/-/issues/6223Some Constant Effect Bound Items inconsistencies2022-08-04T09:56:24ZAbdu SharifSome Constant Effect Bound Items inconsistenciesHere are some of them:
- If you equip a boot enchanted with a CE Bound Boots enchantment, the boot gets unequipped but no Bound Boots are equipped nor will you find any in your character's inventory, but in Morrowind once the boot is eq...Here are some of them:
- If you equip a boot enchanted with a CE Bound Boots enchantment, the boot gets unequipped but no Bound Boots are equipped nor will you find any in your character's inventory, but in Morrowind once the boot is equipped Bound Boots will take it's place and the enchanted boot will stay in the inventory (as normal). By the way, dropping the enchanted boot will not remove the Bound Boots.
- In Morrowind, if you equip for example a left gauntlet enchanted with a CE Bound Gloves enchantment, it equips two Bound Gauntlets (left and right). I'll not mention the OpenMW behavior because it's unfathomable and inconsistent, but the summary is only one Bound Gauntlet gets equipped for one left or right enchanted gauntlet you equip, please test it OpenMW to know the full behavior otherwise.
Note that in Morrowind, If you equip two gauntlets (left and right) at the same time then close the inventory (remember Morrowind only casts magic effects after closing the inventory unlike OpenMW); the game will crash :)
- Everytime you equip for example a boot with CE Bound Boots enchantment it should add/equip a new Bound Boot everytime, like what's mentioned in #6087
Some test saves:
[Daedric_Enchants.omwsave](/uploads/ff4aaf792df8924c11d5383a4c0f52cc/Daedric_Enchants.omwsave)
[Daedric_Enchants___Boots.omwsave](/uploads/e184cd56d21698f4a8896776b87d64f1/Daedric_Enchants___Boots.omwsave)
[Daedric_Enchants___Gloves.omwsave](/uploads/5e6cf4dae529b239294a4f5c1c5a9609/Daedric_Enchants___Gloves.omwsave)https://gitlab.com/OpenMW/openmw/-/issues/6210Object paging hides objects in non-active cells in gridlike fashion2022-01-30T21:35:26ZNelsson HuotariObject paging hides objects in non-active cells in gridlike fashionIt seems like some far cells are skipped in gridlike fashion. The test case is far from the origin (coordinates 65,65), using simple Collada test models, and using extreme view distance (200k). In order to get as many far objects to rend...It seems like some far cells are skipped in gridlike fashion. The test case is far from the origin (coordinates 65,65), using simple Collada test models, and using extreme view distance (200k). In order to get as many far objects to render, I'm using setting `object paging min size = 0.00000001`. With these settings some objects are rendered in extremely far cells, even if cells at moderate distance are skipped. In some cells, object rendering is cut at some randomly seeming x or y land vertex (also visible in the screenshot).
Using Linux Manjaro, git after commit b8878cb. I've tested also with changes brought by !1118 and !1117 , and the bug is always there.
I'm using Example Suite/Template -based game. I haven't found a Morrowind case where this would happen.
![Screenshot_2021-08-10_00-26-02](/uploads/8472043f9890337b1e801c153985c797/Screenshot_2021-08-10_00-26-02.png)https://gitlab.com/OpenMW/openmw/-/issues/6202MR !1085 seems to have caused some physics problems2023-07-13T06:08:50ZHristos N. TriantafillouMR !1085 seems to have caused some physics problemsWhile playing I am noticing that my player character (and all other actors) stop moving, but the scene keeps rendering (my frametime graph doesn't stop). After some time I'll begin to move again (usually), and if I pressed any movement k...While playing I am noticing that my player character (and all other actors) stop moving, but the scene keeps rendering (my frametime graph doesn't stop). After some time I'll begin to move again (usually), and if I pressed any movement keys while I was stuck, I'd suddenly "teleport" to where I would have been if moving normally.
~~I did a git bisect and found that `bd1bf4ec731dfc0e8cb50c3476361707e6b2aa23` seemed to be the bad commit.~~
cc @fr3dz10 since we talked about this on Discord.
Details on my system and how I produce the issue:
* This happens while walking around, with mods or totally vanilla Morrowind.
* I spawn in Seyda Neen and walk up the western coast to Gnaar Mok, then I walk back south. It usually happens after some time (though it has happened in the first few minutes at least once)
* The message `Error in frame: moveTo: object is not in this cell` is repeated spammed in the logs while playing.
* I'm playing on Linux and using LTO (gcc 10.2.1)
* This is my settings.cfg:
[settings.cfg](/uploads/5f04c014a75f44b0bdcba955877c06e9/settings.cfg)https://gitlab.com/OpenMW/openmw/-/issues/6182Implement proper x and y angle modification behaviour2021-07-27T18:17:38ZKuyondoImplement proper x and y angle modification behaviourCurrently _setangle x_ or _setangle y_ doesn't work as in vanilla.
In OpenMW, these commands did nothing useful apart from strange 1st person camera swaying.
In Vanilla, both the body and 1st person camera move accordingly.
This is mo...Currently _setangle x_ or _setangle y_ doesn't work as in vanilla.
In OpenMW, these commands did nothing useful apart from strange 1st person camera swaying.
In Vanilla, both the body and 1st person camera move accordingly.
This is more obvious if it is applied to the player. It would be neat also if NPCS and Objects can behave similarly as an extra feature.
![aa](/uploads/d40ed5637b5ca22e6b4a06d503be92e9/aa.mp4)
Plugin to test, type startscript aaaa on the player. [a.ESP](/uploads/e1f4e0db6784d9ba171f136bc7b66d6e/a.ESP)https://gitlab.com/OpenMW/openmw/-/issues/6175Some modded meshes are invisible2023-03-08T10:44:54ZRadish StrangerSome modded meshes are invisibleI am running OpenMW 0.46 on a macOS 10.15.6 (Catalina) with a CD version of Morrowind, Tribunal & Bloodmoon (English).
The problem occurs with some mods where a mesh is replaced by a new .nif file ([New Boots of Blinding Speed](https://...I am running OpenMW 0.46 on a macOS 10.15.6 (Catalina) with a CD version of Morrowind, Tribunal & Bloodmoon (English).
The problem occurs with some mods where a mesh is replaced by a new .nif file ([New Boots of Blinding Speed](https://www.nexusmods.com/morrowind/mods/42042) and [SM_Mace of Aevar Stone-Singer](https://www.nexusmods.com/morrowind/mods/50054) are good mod examples). The modded element simply doesn't show and becomes invisible.
The problem occurs even when removing all other mods and esps (both from mod folder and openmw.cfg) and starting up from a content list containing only Morrowind, Tribunal & Bloodmoon.
The problem doesn't occur on OpenMW 0.45, but it does on 0.46 up to the latest unreleased build marked 0.48.
The problem doesn't occur in vanilla Morrowind or the Windows versions of OpenMW I was able to test through CrossOver.
My openmw.log file ([issue.zip](/uploads/9d4eb3908bf312111712c77f1aa9b36b/issue.zip)) doesn't seem to show any related errors.https://gitlab.com/OpenMW/openmw/-/issues/6153Underwater blended objects incorrectly rendered with refraction2021-07-24T18:00:54ZCody GlassmanUnderwater blended objects incorrectly rendered with refractionUnderwater blended meshes do not render correctly with water refraction. I have not yet confirmed if this is a regression.
![screenshot453](/uploads/a15c9c4000295c01d1371668a1743957/screenshot453.png)
**Reproducing**
Tested with 5357c...Underwater blended meshes do not render correctly with water refraction. I have not yet confirmed if this is a regression.
![screenshot453](/uploads/a15c9c4000295c01d1371668a1743957/screenshot453.png)
**Reproducing**
Tested with 5357c67c
1. Ensure you have no graphical mods enabled (some popular ones change blending to testing)
2. Have water shaders and refraction enabled
3. Load [this](/uploads/3e85fa863ca1f3db40ae79efdc97e269/blended.omwsave) save and observe the underwater tree