| Time |
Nick |
Message |
| 00:17 |
kilbith |
someone explains me this bug in the water? https://youtu.be/ZjdCKAfamBk |
| 00:17 |
kilbith |
the player cannot go forward if I look down in the water |
| 00:24 |
MTDiscord |
<GreenXenith> kilbith: If you have pitch fly enabled, Minetest thinks you are swimming downwards rather than walking |
| 00:24 |
MTDiscord |
<GreenXenith> pitch fly/movement/whatever |
| 00:26 |
kilbith |
it's enabled yes but I should walk normally when I'm not full drown |
| 00:27 |
MTDiscord |
<GreenXenith> Its probably polling the player position which is at the feet for liquid |
| 01:11 |
|
behalebabo joined #minetest-dev |
| 01:22 |
|
T^4im joined #minetest-dev |
| 02:22 |
MTDiscord |
<exe_virus> Can't we turn off shadows on entities specifically from other entities? That would fix the problem, no entity to entity shadows |
| 02:29 |
|
queria joined #minetest-dev |
| 02:30 |
|
queria^clone joined #minetest-dev |
| 03:22 |
VanessaE |
I dunno about who casts shadows on what, but if I could turn them off per-entity I'd like that -- for signs_lib |
| 03:23 |
VanessaE |
probably that one item drawers mod needs it also, I forget who's mod it is or even the proper name |
| 03:23 |
VanessaE |
item frame in home decor also, perhaps. |
| 03:23 |
VanessaE |
whose* |
| 03:24 |
|
MTDiscord joined #minetest-dev |
| 05:12 |
|
Sokomine_ joined #minetest-dev |
| 08:01 |
|
specing_ joined #minetest-dev |
| 08:16 |
|
YuGiOhJCJ joined #minetest-dev |
| 09:52 |
|
YuGiOhJCJ joined #minetest-dev |
| 10:13 |
|
calcul0n__ joined #minetest-dev |
| 10:23 |
|
hecks joined #minetest-dev |
| 10:35 |
|
YuGiOhJCJ joined #minetest-dev |
| 10:53 |
|
Fixer joined #minetest-dev |
| 11:35 |
hecks |
so since we do not support toggleable shader variants at all, functionality like shadows is hard-baked into the shaders when the client connects with no way to switch it off later... |
| 11:43 |
sfan5 |
shouldn't be so hard to support, should it? |
| 11:49 |
hecks |
shader variants actually are a mildly difficult task if you want to do them right |
| 11:50 |
hecks |
suddenly it's not one E_MATERIAL_TYPE you have to handle, but a whole matrix of them |
| 11:50 |
hecks |
I'm making a crude fix just for shadows but the next time this comes up, it must be redone |
| 11:52 |
hecks |
also all the code happily assumes that all it needs to know about a shader is its s32 handle |
| 11:53 |
hecks |
anyway the first thing I'll do is make the shadows go away if the sun is off |
| 11:53 |
hecks |
both sun and moon |
| 11:53 |
hecks |
this is still less than ideal because a skybox might have a baked sun but it works for me |
| 11:54 |
MTDiscord |
<Warr1024> I have a skybox with no sun or moon at all, so I have no idea how silly the game is going to look if players can't turn off shadows... |
| 11:57 |
|
entuland joined #minetest-dev |
| 12:06 |
sfan5 |
https://0x0.st/-99Q.png about like this, no sun but shadows mysteriously appear |
| 12:16 |
hecks |
yeah, that |
| 12:16 |
hecks |
actually it seems like the server generates sun/moon packets even if those apis aren't touched directly |
| 12:16 |
hecks |
so all it takes is handling those |
| 12:17 |
hecks |
i already did that and it works, the problem is turning off the usage of nonexistent shadows on nodes and objects |
| 12:17 |
hecks |
this requires variants |
| 12:17 |
hecks |
simply setting the shadow intensity to 0 is not acceptable, shadow sampling is ridiculously expensive |
| 12:18 |
hecks |
so expensive in fact that I would recommend making unfiltered shadows the default mode, and requiring a separate setting for PCF |
| 12:35 |
MTDiscord |
<Warr1024> Being able to blacklist shaders, similar to how games can blacklist mapgens, might be more ideal, but I'd at least like a workaround, like if I can set time_speed to 0 and just set a particular time-of-day at least... |
| 12:37 |
MTDiscord |
<Warr1024> I guess I'm going to have to figure out how to build this thing myself again; haven't had it working since the irrlicht fork. Are there instructions out there somewhere how to build this thing now? |
| 13:04 |
MTDiscord |
<josiah_wi> No good ones The README has some footnote in the build section. |
| 13:08 |
|
hecks joined #minetest-dev |
| 13:09 |
hecks |
ShaderInfo.base_material - this is a fallback when shaders are disabled? |
| 13:11 |
MTDiscord |
<josiah_wi> My PR is broken on CI, but it works on a normal build environment and makes it pretty easy. It also contains build instructions in the description. |
| 13:51 |
|
Lone_Wolf joined #minetest-dev |
| 14:34 |
Krock |
will merge #11092, #11333 and #11368 in 10 minutes |
| 14:34 |
ShadowBot |
https://github.com/minetest/minetest/issues/11092 -- Make chat web links clickable by pecksin |
| 14:34 |
ShadowBot |
https://github.com/minetest/minetest/issues/11333 -- Add min_y and max_y checks for Active Block Modifiers (ABM) by sfence |
| 14:34 |
ShadowBot |
https://github.com/minetest/minetest/issues/11368 -- Inventory: show error on invalid list names by SmallJoker |
| 14:50 |
nrz |
instead of blacklisting, permit to control the rendering to enable/disable for servers with some meteo for example, it's better |
| 14:57 |
MTDiscord |
<Warr1024> I was able to work around them by just forcing timeofday to 5:00AM always. Players with the shader enabled may still be paying the performance cost, but at least the game looks right again. |
| 14:58 |
MTDiscord |
<Warr1024> Having servers or games able to override selected player options that don't make sense seems like it's part of a larger issue. |
| 15:19 |
Krock |
... reminder didn't work.. merging. |
| 15:22 |
Krock |
done |
| 15:50 |
|
Extex joined #minetest-dev |
| 16:14 |
|
Lone_Wolf joined #minetest-dev |
| 17:59 |
|
hecks joined #minetest-dev |
| 18:51 |
|
v-rob joined #minetest-dev |
| 18:59 |
|
hecks joined #minetest-dev |
| 18:59 |
|
kilbith joined #minetest-dev |
| 19:02 |
kilbith |
chat web links are clickable but formspec's not |
| 19:02 |
kilbith |
how consistent |
| 19:07 |
|
kilbith_ joined #minetest-dev |
| 19:12 |
v-rob |
What? Formspecs are supposed to be consistent? I never would have guessed :P |
| 19:13 |
kilbith_ |
I mean, there would only need core.open_url accessible from mods and weblinks would be clickable from hypertext |
| 19:14 |
kilbith_ |
idk why it isn't in the modding API |
| 19:14 |
rubenwardy |
would require a confirmation dialog |
| 19:14 |
specing |
yeah and we need javascript in formspecs |
| 19:14 |
specing |
/s |
| 19:14 |
v-rob |
Calling it in a loop would be bad |
| 19:15 |
rubenwardy |
might instead be worth having button_web to ensure that UI interactions are needed to open the url |
| 19:15 |
v-rob |
Better yet than just JavaScript, make formspecs a W3C standard so they can get even more confusing and worse |
| 19:16 |
rubenwardy |
My original PR to add core.open_url to the mainmenu had a button_web element available to server-side mods |
| 19:16 |
rubenwardy |
was removed due to concerns of security, ie: the need to show a confirmation dialog |
| 19:17 |
|
YuGiOhJCJ joined #minetest-dev |
| 19:18 |
kilbith_ |
pretty sure professional-level engines have a open-url function in their API and they are not concerned about security |
| 19:18 |
kilbith_ |
`OpenURL` method in Unity for example |
| 19:18 |
rubenwardy |
they also ship separate games with client-side code |
| 19:18 |
kilbith_ |
https://docs.unity3d.com/ScriptReference/Application.OpenURL.html |
| 19:19 |
rubenwardy |
Minetest is different in that "games" can run on servers, and you don't have a separate client for each game |
| 19:19 |
kilbith_ |
true |
| 19:21 |
MTDiscord |
<appguru> a confirmation dialog would be a clean solution |
| 19:27 |
kilbith |
and ideally we should be able to open .onion links and make Tor a dependency |
| 19:42 |
|
kilbith joined #minetest-dev |
| 19:45 |
MTDiscord |
<Warr1024> I absolutely hate when shit asks me for confirmation to open a link without any way to turn off confirmation (like Discord). If I change my mind and don't want to open the link after I see the URL then I just won't pick a browser to open it with. |
| 19:46 |
|
kilbith joined #minetest-dev |
| 19:56 |
hecks |
onion link support when we don't even encrypt game traffic :DDDD |
| 20:01 |
|
specing_ joined #minetest-dev |
| 20:04 |
hecks |
URLs in forms are a little more tricky because of clickjacking potential |
| 20:04 |
hecks |
you cannot trick someone into clicking a link they don't want from the chat, but you can do this with a form |
| 20:05 |
hecks |
best case the destination is just some pay per click thing, worst case it's a zero day in a common browser |
| 20:06 |
v-rob |
That's why a confirmation dialog would be important, yeah |
| 20:06 |
v-rob |
Unfortunate, but necessary |
| 20:06 |
hecks |
furthermore MT traffic, as simple as it is, is quite vulnerable to MITM |
| 20:06 |
hecks |
and while nobody does that *now*, clickable links in forms would actually provide an incentive for it |
| 20:33 |
MTDiscord |
<Jordach> why don't we just do things like steam |
| 20:33 |
MTDiscord |
<Jordach> this is not an official minetest website |
| 20:33 |
MTDiscord |
<Jordach> it means the engine itself wraps the link with a rerouting link |
| 20:43 |
|
jonadab joined #minetest-dev |
| 21:53 |
|
v-rob joined #minetest-dev |
| 21:59 |
|
kilbith joined #minetest-dev |
| 22:00 |
|
T4im joined #minetest-dev |
| 22:25 |
|
wsor4035 joined #minetest-dev |
| 22:34 |
|
MTDiscord joined #minetest-dev |
| 23:06 |
|
Alias joined #minetest-dev |
| 23:58 |
|
v-rob joined #minetest-dev |
