| Time |
Nick |
Message |
| 00:00 |
sofar |
well that took a strange turn |
| 00:20 |
|
nolsen joined #minetest-dev |
| 00:30 |
RealBadAngel |
sorry for a break |
| 00:32 |
RealBadAngel |
but displayin meshes instead of hacky and dirty tricks (see android case) is getting things done correct way |
| 00:34 |
Fixer |
texture tear problem |
| 00:34 |
* Fixer |
runs |
| 00:36 |
kahrl |
render to texture is a hacky and dirty trick?! |
| 00:46 |
|
sdzen joined #minetest-dev |
| 00:47 |
rubenwardy |
I guess rendering a mesh to a texture, the rendering a mesh with a texture over it as part of drawing is a little ugly |
| 01:45 |
|
eman_ joined #minetest-dev |
| 03:11 |
|
sdzen joined #minetest-dev |
| 03:59 |
|
paramat joined #minetest-dev |
| 04:01 |
paramat |
now merging game 780 |
| 04:06 |
paramat |
merged |
| 04:07 |
|
paramat left #minetest-dev |
| 04:14 |
RealBadAngel |
kahrl, when you can just set viewport on the current render, set draw material and order to draw meshbuffer? yes |
| 04:16 |
RealBadAngel |
its just overengineering |
| 04:18 |
RealBadAngel |
http://pastie.org/10684770 |
| 04:18 |
RealBadAngel |
that code just draws the mesh |
| 04:21 |
RealBadAngel |
as you can see, theres no need to set new scenemanager, add scene nodes, set cameras, lighting, draw to the texture, cut texture, store and render it again later on |
| 06:29 |
sofar |
anyone around that can help me figure out how to add something specific to script/common/c_content.cpp? |
| 06:29 |
sofar |
specifically, I need to add an array of strings |
| 06:29 |
sofar |
connects_to = { "string", "string" } |
| 06:30 |
sofar |
https://gist.github.com/sofar/b429738f068b06be7ffd came up with that, but I'm not entirely sure |
| 06:31 |
sofar |
(actually fairly sure it's not correct) |
| 06:35 |
|
Hunterz joined #minetest-dev |
| 06:36 |
RealBadAngel |
textures are using that? |
| 06:36 |
RealBadAngel |
line #150 |
| 06:36 |
sofar |
I've nosed around, but it's using strange offsets I don't understand |
| 06:37 |
RealBadAngel |
special tiles are also accepting strings inside of tables |
| 06:37 |
RealBadAngel |
line #306 |
| 06:39 |
RealBadAngel |
but #422 looks like is doing what you want exactly |
| 06:39 |
sofar |
assuming I can ditch the CF_SPECIAL_COUNT section |
| 06:40 |
sofar |
so -1 index then |
| 06:40 |
RealBadAngel |
if(i==CF_SPECIAL_COUNT) that makes sure you dont have more than 6 tiles here |
| 06:41 |
sofar |
yeah, I'm just adding all values |
| 06:41 |
sofar |
irregardless of size |
| 06:41 |
sofar |
of the array |
| 06:41 |
RealBadAngel |
so drop that line |
| 06:42 |
RealBadAngel |
or make there some limit like 256 or something |
| 06:42 |
RealBadAngel |
just to make sure somebody havent pasted here a book ;) |
| 06:43 |
sofar |
only mods can provide the info, so, screw them :) |
| 06:43 |
RealBadAngel |
hehe |
| 06:43 |
sofar |
ok, that compiles already |
| 06:43 |
sofar |
now serialize and deserialize this |
| 06:45 |
RealBadAngel |
btw, i saw on the screenie some wires |
| 06:46 |
RealBadAngel |
you made them connected to some central spot, right? |
| 06:46 |
sofar |
there is a small nodebox in the center for an unconnected node |
| 06:46 |
sofar |
just like the pole for a fence |
| 06:46 |
sofar |
just, very small |
| 06:46 |
RealBadAngel |
i see |
| 06:46 |
sofar |
2x2x1 pixel, in the wire case |
| 06:46 |
sofar |
but, it's optional |
| 06:47 |
RealBadAngel |
so this kind of connection will be useable for tubes and technic wireing |
| 06:47 |
sofar |
you can have any number of fixed nodeboxes that are optional |
| 06:47 |
RealBadAngel |
not mesecons |
| 06:47 |
sofar |
yes, it's useful for tubes |
| 06:47 |
RealBadAngel |
mesecons are kind different |
| 06:47 |
sofar |
right, the 10-fold connecting thing |
| 06:47 |
RealBadAngel |
they shall "climb" the walls of adjacent nodes |
| 06:48 |
sofar |
you can probably still use this nodetype and massively reduce the amount of nodes |
| 06:48 |
sofar |
and like I said, maybe later we can add the 4 extra connections as well |
| 06:48 |
RealBadAngel |
propably, but it will be still more than just one |
| 06:49 |
sofar |
but I want it working with 6x connections first :) |
| 06:49 |
RealBadAngel |
but imho you shall stop there |
| 06:49 |
RealBadAngel |
wires are way more complicated |
| 06:49 |
sofar |
crossing wires with shielding |
| 06:50 |
sofar |
I don't think this mod can do that |
| 06:50 |
sofar |
this patch* |
| 06:50 |
RealBadAngel |
have you actually saw RedPower2's wireing? |
| 06:50 |
sofar |
long time ago |
| 06:51 |
RealBadAngel |
once is enough ;) |
| 06:51 |
RealBadAngel |
the one youre developing is good for in the centre connections |
| 06:51 |
RealBadAngel |
walls have way more fucked up logic |
| 06:53 |
RealBadAngel |
but if you wanna look at it, the mod is recreated open source |
| 06:54 |
RealBadAngel |
http://projectredwiki.com/wiki/Main_Page |
| 07:00 |
sofar |
cool I think I got the lua part reading working. Got the correct amount of entries in my array |
| 07:03 |
RealBadAngel |
hah and i found "nice" bug in my code |
| 07:04 |
RealBadAngel |
the dragged item under the cursor got streched when moved into the corner of the screen |
| 07:04 |
RealBadAngel |
it looks funny :) |
| 07:04 |
Skyduskguy |
I just want conveyer belts and auto crafting :( |
| 07:04 |
RealBadAngel |
Skyduskguy, install pipeworks and technic just? |
| 07:05 |
sofar |
an open conveyor belt would be neat |
| 07:05 |
sofar |
one that moves entities |
| 07:05 |
Skyduskguy |
there was one >.> but its ded |
| 07:05 |
Skyduskguy |
actually it was the first thing i searched for |
| 07:05 |
Skyduskguy |
https://forum.minetest.net/viewtopic.php?f=11&t=9277&hilit=conveyor&start=50 |
| 07:06 |
Skyduskguy |
yeah pipes are not belts sadly |
| 07:06 |
RealBadAngel |
it shouldnt be that hard to replace tubes with other GFx |
| 07:06 |
RealBadAngel |
and change offset of the items travelling slightly |
| 07:06 |
sofar |
yup |
| 07:07 |
RealBadAngel |
that could be done propably in less than 10 minutes |
| 07:07 |
sofar |
the belt can even use their own entities, but it would be nice if they could be dropped and picked up like normal itemstack entities |
| 07:07 |
Skyduskguy |
well yeah |
| 07:08 |
RealBadAngel |
ok, 15 minutes ;) |
| 07:08 |
Skyduskguy |
if it didnt do that it wouldnt be belts |
| 07:08 |
Skyduskguy |
dont tease :o |
| 07:08 |
RealBadAngel |
im on my own code, dont try to point at me ;P |
| 07:09 |
Skyduskguy |
well yeah im sure ppl who know howto this stuff could |
| 07:09 |
Skyduskguy |
but im faaar from that |
| 07:09 |
Skyduskguy |
im still trying to learn how not to have my mod fry the poor console |
| 07:10 |
RealBadAngel |
learnig lua is quite fun |
| 07:11 |
exio4 |
j |
| 07:11 |
exio4 |
oops <.< |
| 07:11 |
RealBadAngel |
now you do have lotsa examples out here |
| 07:11 |
RealBadAngel |
a few yrs ago we were crying out loud to accomplish simple tasks ;) |
| 07:12 |
Skyduskguy |
no way id attempt such a task, im certine id get frustraited and quit all together |
| 07:13 |
Skyduskguy |
actually a small community im part of is waiting on a threshold for minetest to drop mc alltogether |
| 07:13 |
RealBadAngel |
oh cmon, reward (the feeling) when you have it working is better when it wasnt that easy |
| 07:13 |
Skyduskguy |
im currently the only one working tword learning modding to make it happen |
| 07:14 |
Skyduskguy |
im not good at commiting to long things |
| 07:14 |
Skyduskguy |
I drop projects all the time, im not proud of it |
| 07:15 |
Skyduskguy |
just personal stuff |
| 07:15 |
RealBadAngel |
try to keep focus |
| 07:17 |
Skyduskguy |
its a motivation thing |
| 07:18 |
|
celeron55 joined #minetest-dev |
| 07:18 |
Skyduskguy |
I just try to keep my stuff small, its the only way i can see it through |
| 07:20 |
RealBadAngel |
good for starters |
| 07:20 |
RealBadAngel |
with time each project grows |
| 07:21 |
RealBadAngel |
ive started Technic with simple chest modification, now its the most complex mod for mt ever |
| 07:22 |
RealBadAngel |
propably minetest_game is a simpler mod than technic alone ;) |
| 07:22 |
Skyduskguy |
forum link |
| 07:23 |
RealBadAngel |
https://forum.minetest.net/viewtopic.php?t=2538 |
| 07:23 |
sofar |
ok, deserialization and serialization done too |
| 07:23 |
sofar |
now for node matching |
| 07:24 |
sofar |
CNodeDefManager::getIds excellent |
| 07:28 |
|
celeron55 joined #minetest-dev |
| 07:29 |
RealBadAngel |
hmm, i think that one callback for collision boxes would be handy, the one triggered on actual collision |
| 07:30 |
sofar |
for your meshes? |
| 07:30 |
RealBadAngel |
in general i think |
| 07:30 |
RealBadAngel |
if player walks into something |
| 07:30 |
sofar |
sure |
| 07:30 |
sofar |
return what it has walked into |
| 07:30 |
RealBadAngel |
atm we are using abms for that |
| 07:30 |
sofar |
walk into water => play water splash |
| 07:30 |
RealBadAngel |
thats costly |
| 07:31 |
sofar |
walk into cactus node => instant death |
| 07:31 |
sofar |
walk into lava => melt face |
| 07:31 |
RealBadAngel |
it shall be something like on_collide |
| 07:31 |
RealBadAngel |
so you can redefine it |
| 07:31 |
sofar |
why not on_collision? |
| 07:32 |
RealBadAngel |
whatever, just naming :) |
| 07:32 |
Skyduskguy |
can i move th eplayer on collision with a particle? |
| 07:32 |
sofar |
no, particles can only collide with the environment |
| 07:32 |
sofar |
there's no callbacks either |
| 07:32 |
sofar |
particles are not entities |
| 07:33 |
Skyduskguy |
darn |
| 07:33 |
RealBadAngel |
sofar, it shouldnt be hard, theyre already checked (player cant walk into some nodes) |
| 07:33 |
sofar |
if you ask me, animated particle textures would be a fantastic addition |
| 07:33 |
RealBadAngel |
so just add there callback |
| 07:33 |
sofar |
right, should not be insane to code up |
| 07:33 |
sofar |
could be costly though |
| 07:34 |
sofar |
you'd want that handled client side |
| 07:34 |
RealBadAngel |
propably it should be somehow limited |
| 07:34 |
sofar |
question is, what are you going to use it for? |
| 07:34 |
RealBadAngel |
like first hit or something |
| 07:34 |
RealBadAngel |
mainly? |
| 07:34 |
Skyduskguy |
fan nodes |
| 07:34 |
RealBadAngel |
step in teleporters |
| 07:34 |
sofar |
right, but I hope not particle collisions :) |
| 07:35 |
Skyduskguy |
well that cant exist |
| 07:35 |
Skyduskguy |
you just said right? |
| 07:35 |
RealBadAngel |
use cases can be different of course, thats just one i have in mind |
| 07:35 |
Skyduskguy |
id imagine drawing line of sight to player along a ray? |
| 07:35 |
|
turtleman joined #minetest-dev |
| 07:35 |
Skyduskguy |
vector? |
| 07:35 |
RealBadAngel |
i was thinkin about stargates being sometimes responding too slow (thx to abms usage) |
| 07:36 |
Skyduskguy |
idk what im talkin here |
| 07:36 |
RealBadAngel |
Skyduskguy, theres some function bout it in builtin i think |
| 07:37 |
Skyduskguy |
yeah there was a fan block mod, moved items players mobs |
| 07:37 |
Skyduskguy |
loved a combination with store on hit chests |
| 07:38 |
|
celeron55 joined #minetest-dev |
| 07:38 |
Skyduskguy |
ohhh does anything like hopper exist? |
| 07:38 |
Skyduskguy |
I think it does |
| 07:38 |
sofar |
pipeworks "filter" |
| 07:38 |
RealBadAngel |
Skyduskguy, lua_api.txt line #2063 |
| 07:39 |
Skyduskguy |
spawn tree? |
| 07:39 |
RealBadAngel |
awww |
| 07:39 |
|
nrzkt joined #minetest-dev |
| 07:40 |
RealBadAngel |
https://github.com/minetest/minetest/blob/master/doc/lua_api.txt#L2065 |
| 07:40 |
Skyduskguy |
i was on https://github.com/celeron55/minetest/blob/master/doc/lua_api.txt |
| 07:41 |
Skyduskguy |
nice anyway |
| 07:44 |
sofar |
alright... I think I have a great way to do connecting nodes where you have fine-grained control |
| 07:44 |
sofar |
and it won't be slow to match either |
| 07:45 |
sofar |
I'll use a std::vector<content_t> |
| 07:45 |
sofar |
actually |
| 07:45 |
Skyduskguy |
howdo you change an entity's velocity? |
| 07:45 |
|
Obani joined #minetest-dev |
| 07:46 |
Skyduskguy |
ok its just setvelocity |
| 07:46 |
RealBadAngel |
http://i.imgur.com/0sIS7vA.png |
| 07:47 |
sofar |
gifv that shit |
| 07:47 |
RealBadAngel |
items on the left are shown with meshes |
| 07:47 |
RealBadAngel |
on the right with usual textures |
| 07:47 |
RealBadAngel |
ive added a slight rotations to them |
| 07:47 |
sofar |
ahh they're not rotating? |
| 07:48 |
RealBadAngel |
they do if selected |
| 07:48 |
|
celeron55 joined #minetest-dev |
| 07:48 |
RealBadAngel |
but im talkin now about still representation |
| 07:48 |
RealBadAngel |
theyre rotated and shaded the same way as nodes |
| 07:49 |
RealBadAngel |
imho they DO look more 3d now |
| 07:49 |
RealBadAngel |
compare mese pickaxe for example |
| 07:52 |
|
eman__ joined #minetest-dev |
| 07:53 |
RealBadAngel |
uploading now video |
| 07:56 |
|
Lunatrius joined #minetest-dev |
| 07:56 |
RealBadAngel |
sofar, https://youtu.be/lC3SAyiD2UI |
| 07:58 |
|
celeron55 joined #minetest-dev |
| 07:59 |
sofar |
cure |
| 07:59 |
sofar |
cute* |
| 08:01 |
Skyduskguy |
thats very nice |
| 08:02 |
RealBadAngel |
imho that brings consistence, items looks the very same way in inventories, on the ground and when wielded |
| 08:03 |
RealBadAngel |
if they spin when selected is just a side effect ;) |
| 08:04 |
Skyduskguy |
well sofar imade a few pushmits |
| 08:04 |
Skyduskguy |
but im about drained and tierd 2am here |
| 08:04 |
RealBadAngel |
pushmits? |
| 08:04 |
Skyduskguy |
push commits |
| 08:05 |
Skyduskguy |
idk im stillg etting the words mixed up |
| 08:05 |
sofar |
I'm out. midnight. *poof I'm a pumpkin* |
| 08:05 |
Fritigern |
It's been an eventvul day, Skyduskguy |
| 08:06 |
RealBadAngel |
sofar, where do you live? (its 9am here lol) |
| 08:06 |
Skyduskguy |
I leave you with besige gif |
| 08:06 |
Skyduskguy |
http://i.imgur.com/0dbQAd8.webm |
| 08:08 |
|
celeron55 joined #minetest-dev |
| 08:24 |
|
red-001 joined #minetest-dev |
| 08:29 |
|
red-001 joined #minetest-dev |
| 08:52 |
|
hmmmm joined #minetest-dev |
| 09:07 |
|
jin_xi joined #minetest-dev |
| 09:27 |
|
blaze joined #minetest-dev |
| 09:55 |
|
thatgraemeguy joined #minetest-dev |
| 09:55 |
|
thatgraemeguy joined #minetest-dev |
| 09:55 |
|
eman__ joined #minetest-dev |
| 10:26 |
|
Megaf joined #minetest-dev |
| 10:27 |
Megaf |
Hi |
| 10:29 |
|
guest365 joined #minetest-dev |
| 10:31 |
guest365 |
when will the next major version be released |
| 10:32 |
|
est31 joined #minetest-dev |
| 11:28 |
Taoki |
RealBadAngel: Lovely work again! It does work with all drawtypes though, does it? |
| 11:28 |
|
kaeza joined #minetest-dev |
| 11:28 |
Taoki |
Also, does that mean 3D plants are at last going to be in themselves soon? :) |
| 11:49 |
|
Calinou joined #minetest-dev |
| 12:08 |
|
Fixer joined #minetest-dev |
| 12:26 |
|
sdzen joined #minetest-dev |
| 13:00 |
|
Megaf_ joined #minetest-dev |
| 13:15 |
|
sofar_ joined #minetest-dev |
| 13:15 |
|
H-H-H joined #minetest-dev |
| 13:16 |
|
est31 joined #minetest-dev |
| 13:39 |
|
sdzen joined #minetest-dev |
| 13:40 |
|
turtleman joined #minetest-dev |
| 14:25 |
|
celeron55 joined #minetest-dev |
| 14:39 |
|
Lunatrius joined #minetest-dev |
| 14:42 |
|
Megaf_ joined #minetest-dev |
| 14:51 |
|
rubenwardy joined #minetest-dev |
| 15:00 |
|
jin_xi joined #minetest-dev |
| 15:27 |
|
hmmmm joined #minetest-dev |
| 15:43 |
|
red-001 joined #minetest-dev |
| 15:49 |
|
behalebabo joined #minetest-dev |
| 16:11 |
RealBadAngel |
Taoki, yes it will work with all drawtypes. but thats not related to plantlike, it is using provided, regular or extruded one meshes, just like wielded |
| 16:11 |
RealBadAngel |
so it is consistent with everythin in game, item looks exactly the same in hand, in inventory or on the ground |
| 16:12 |
RealBadAngel |
same meshes, same textures and the shaders |
| 16:17 |
|
kaadmy joined #minetest-dev |
| 16:26 |
|
CraigyDavi joined #minetest-dev |
| 16:26 |
|
Calinou joined #minetest-dev |
| 16:44 |
|
Krock joined #minetest-dev |
| 16:54 |
|
Hunterz joined #minetest-dev |
| 16:55 |
|
void7 joined #minetest-dev |
| 17:03 |
|
Amaz joined #minetest-dev |
| 17:21 |
|
Obani joined #minetest-dev |
| 17:28 |
|
rubenwardy joined #minetest-dev |
| 17:35 |
|
DFeniks joined #minetest-dev |
| 17:51 |
|
nrzkt joined #minetest-dev |
| 17:55 |
|
Robert_Zenz joined #minetest-dev |
| 18:03 |
|
eman__ joined #minetest-dev |
| 18:03 |
|
sofar joined #minetest-dev |
| 18:06 |
sfan5 |
x.x.x.x - - [12/Jan/2016:18:05:56 +0000] "GET /list?proto_version_min=13&proto_version_max=25 HTTP/1.1" 200 185755 "-" "cgcraftingguide/1.0 (Linux/3.4.0+ armv7l)" "-" |
| 18:06 |
sfan5 |
so apparently someone named an android ripoff "cgcraftingguide" |
| 18:06 |
sfan5 |
these names are only getting bettert |
| 18:07 |
sfan5 |
also: "cgbuildahouse" |
| 18:07 |
sfan5 |
and "buildavillage" |
| 18:08 |
sfan5 |
and someone is trying to announce a server from "modernhousebuild" |
| 18:11 |
Krock |
xD |
| 18:14 |
sfan5 |
so i just checked 4 random games that are on google play when searching for "build craft" |
| 18:15 |
sfan5 |
none of them mention any licensing stuff or provide their modified source |
| 18:16 |
sofar |
do you have any significant contributions to minetest gpl parts? |
| 18:17 |
red-001 |
I found about 5 minetest builds on google play once and only two even said anything about minetest |
| 18:17 |
sofar |
okay hardcode minetest-core C++ devs, I have some questions I could use some help with |
| 18:18 |
sfan5 |
sofar: yes but i dont feel like sending a dmca |
| 18:18 |
sofar |
who'd be the person to ask? hmmmm? |
| 18:18 |
sfan5 |
probably |
| 18:18 |
sofar |
I'm attempting to pack a std::set<content_t> as part of the nodedef |
| 18:18 |
sofar |
but I'm having |
| 18:18 |
sofar |
some issues regarding when to create the set |
| 18:19 |
sofar |
and how to serialize it. apparently writeU16() doesn't work |
| 18:19 |
sofar |
plus my iterator hates me |
| 18:19 |
sfan5 |
if writeU16() doesn't work then you are using it wrong |
| 18:20 |
sofar |
that's what I think to |
| 18:20 |
sofar |
mapnode.h:typedef u16 content_t; |
| 18:20 |
sofar |
it's... really a u16 though |
| 18:20 |
sofar |
content_mapnode.h:extern content_t trans_table_19[21][2]; |
| 18:20 |
sofar |
but what is that? |
| 18:21 |
sofar |
nvm |
| 18:22 |
sofar |
sfan5: no idea... It's a U16 but it hates... |
| 18:23 |
sofar |
actually, huh, my bad I fixed that part |
| 18:23 |
sofar |
I'm stuck at my iterator hate |
| 18:23 |
sofar |
+ for (std::set<content_t>::const_iterator it = ids.begin(); |
| 18:23 |
sofar |
+ it != ids.end(); it++) |
| 18:23 |
sofar |
+ connects_to_ids.insert(*it); |
| 18:24 |
sofar |
last line throws an error |
| 18:24 |
sofar |
error: no matching function for call to 'std::set<short unsigned int>::insert(const short unsigned int&) const' |
| 18:24 |
sfan5 |
try (u16) *it |
| 18:25 |
sofar |
already did before |
| 18:26 |
sofar |
error: no matching function for call to 'std::set<short unsigned int>::insert(short unsigned int) const' |
| 18:28 |
sofar |
plus, and this is more important than anything |
| 18:28 |
sfan5 |
idk |
| 18:28 |
sofar |
I need to know that I can serialize and deserialize std::set<content_t> values as they remain the same on the client and the server |
| 18:30 |
sofar |
otherwise I need each client to convert nodenames into content_t themselves |
| 18:30 |
sofar |
which means sending node names over the wire, and not ids |
| 18:30 |
sofar |
yes, these are pretty core questions :) |
| 18:30 |
sfan5 |
sofar: have you tried using clang |
| 18:30 |
sfan5 |
maybe it spits out some more useful stuff than gcc |
| 18:31 |
sofar |
gheh, ew |
| 18:31 |
sofar |
I despise other compilers |
| 18:31 |
sfan5 |
clang > gcc |
| 18:32 |
sofar |
the messages are very useful in this case, I just follow examples of others that I've found... and they don't work |
| 18:32 |
sofar |
besides, I'd have to install and figure out how clang works |
| 18:33 |
sfan5 |
apt-get install -y clang |
| 18:33 |
kahrl |
sofar: see the const at the end of the error message. the set is const, so you can't modify it |
| 18:33 |
sfan5 |
and then re-run cmake with CC=clang CXX=clang++ |
| 18:33 |
sfan5 |
not that hard |
| 18:34 |
kahrl |
I've been wondering about std::set<content_t> vs. std::vector<content_t> performance lately |
| 18:34 |
kahrl |
(in the context of ABMs, INodeDefManager::getIds, etc.) |
| 18:34 |
sofar |
kahrl: yeah, that's it. do I change the iterator to a non-const iterator then? |
| 18:34 |
kahrl |
I expect that std::vector<content_t> is much faster at least for small-ish sets |
| 18:34 |
kahrl |
sofar: no, the set itself |
| 18:35 |
sofar |
uhm, it's not const |
| 18:35 |
kahrl |
but does anybody know of benchmarks? |
| 18:35 |
sofar |
nodedef.h:std::set<content_t> connects_to_ids; |
| 18:35 |
kahrl |
sofar: it's const when the object that contains it is const |
| 18:35 |
sofar |
kahrl: yeah, I thought of that too. Sets remove duplicates, which is why I picked it |
| 18:36 |
sfan5 |
kahrl: isn't std::vector likely faster because it doesn't need to order the items? |
| 18:36 |
sofar |
kahrl: hmmmm |
| 18:36 |
kahrl |
sfan5: that and fewer allocations needed |
| 18:36 |
kahrl |
and better cache locality |
| 18:36 |
sfan5 |
no idea about the allocations |
| 18:37 |
sfan5 |
it depends on the c++ stdlib |
| 18:37 |
sfan5 |
os x and iOS use a different one e.g. |
| 18:37 |
kahrl |
doesn't std::set in most implementations allocate each node of the red black tree individually? |
| 18:37 |
kahrl |
but yeah idk, that's why I asked for benchmarks |
| 18:37 |
sfan5 |
i think freebsd also uses libc++ (from the llvm project) |
| 18:38 |
kahrl |
sofar: the method that the erroneous line is in - is it marked as const? |
| 18:38 |
kahrl |
sofar: that means it is operating on a const object |
| 18:38 |
sofar |
insert? |
| 18:39 |
kahrl |
sofar: could you paste the code? it would be easier to talk about then |
| 18:40 |
sofar |
sec |
| 18:42 |
sofar |
https://gist.github.com/sofar/50c040e78a8f2b691fe1 |
| 18:43 |
sofar |
the concept is simple |
| 18:43 |
sfan5 |
i see nothing wrong with that code |
| 18:43 |
sofar |
this is for my connected nodeboxes patch |
| 18:43 |
sfan5 |
but then again i don't use C++ that often |
| 18:43 |
sofar |
each nodedef has a list of node name matches |
| 18:43 |
sofar |
strings |
| 18:43 |
Taoki |
RealBadAngel: Nice. Hope they can reach master soon... that spinning effect in inventory is pretty neat :) |
| 18:44 |
sofar |
the server converts them with getIds() to content_t numbers |
| 18:44 |
sofar |
and serializes them for the client |
| 18:44 |
kahrl |
sofar: ContentFeatures::serialize is a const method |
| 18:44 |
sofar |
client gets them and knows how to connect each connected node with that set |
| 18:44 |
kahrl |
so it operates on a const ContentFeatures |
| 18:44 |
kahrl |
which you can't modify the connects_to_ids field of |
| 18:44 |
rubenwardy |
(sofar: if you call the file something.patch it adds highlighting) |
| 18:44 |
sofar |
right, so my first FIXME is key |
| 18:45 |
sofar |
I need to move the segment that converts to ids to another place in the code |
| 18:45 |
sofar |
it needs to be *after* lua is done reading nodedefs |
| 18:45 |
sofar |
but I haven't found where that would be a great place to put yet |
| 18:45 |
sofar |
any ideas? |
| 18:46 |
sofar |
I'm assuming at that point the const problem disappears |
| 18:46 |
|
Megaf joined #minetest-dev |
| 18:47 |
|
DFeniks joined #minetest-dev |
| 18:48 |
* sofar |
eyeballs game::startup |
| 18:49 |
sofar |
um, that's the client |
| 18:49 |
sofar |
I need the server startup |
| 18:49 |
kahrl |
do it when INodeDefManager::runNodeResolveCallbacks is run? |
| 18:49 |
RealBadAngel |
ok, http://pastie.org/10686100 |
| 18:49 |
kahrl |
either by adding a new NodeResolver or by doing it directly in there |
| 18:49 |
kahrl |
the former is clearner |
| 18:49 |
kahrl |
cleaner* |
| 18:49 |
RealBadAngel |
those are coords (x, y, z) of the simple cube |
| 18:50 |
RealBadAngel |
dumped right from the engine |
| 18:50 |
rubenwardy |
sofar, search for the texture overrides thing |
| 18:50 |
rubenwardy |
1 sec |
| 18:50 |
sofar |
server.cpp has m_nodedef->runNodeResolveCallbacks(); |
| 18:50 |
rubenwardy |
https://github.com/minetest/minetest/blob/master/src/server.cpp#L313 |
| 18:50 |
rubenwardy |
around here |
| 18:50 |
RealBadAngel |
kahrl, can you see whats wrong here? |
| 18:50 |
rubenwardy |
by that point all nodes have been defined |
| 18:51 |
sofar |
ahh yes, this looks like the right context |
| 18:51 |
sofar |
-1.5e+37 lol |
| 18:51 |
RealBadAngel |
we do have some rounding/conversion problem in the engine |
| 18:51 |
sofar |
overflow, possibly? |
| 18:51 |
sofar |
or bad typing |
| 18:51 |
RealBadAngel |
that may be the reason for rendering artifacts |
| 18:52 |
kahrl |
RealBadAngel: can't say without knowing the code that prints that |
| 18:52 |
RealBadAngel |
ive put dump in scale mesh code |
| 18:52 |
RealBadAngel |
so instead of scaling it dumps actual vertices |
| 18:53 |
RealBadAngel |
doing nothing to them |
| 18:54 |
RealBadAngel |
http://pastie.org/10686111 |
| 18:54 |
RealBadAngel |
the code that dumps it |
| 18:55 |
RealBadAngel |
if we do have nans, some weird vaules here, that explains those white dots on dark areas |
| 18:55 |
RealBadAngel |
or lines between nodes |
| 19:01 |
kahrl |
RealBadAngel: can't reproduce. I added the same dstream lines to scaleMesh and it prints reasonable values |
| 19:02 |
RealBadAngel |
kahrl, and thats weird |
| 19:02 |
RealBadAngel |
irrlicht version related? |
| 19:03 |
kahrl |
sounds unlikely |
| 19:03 |
kahrl |
though possible of course |
| 19:04 |
RealBadAngel |
btw, i was able to reproduce wielded glitch |
| 19:04 |
RealBadAngel |
when rendering the whole scene to texture it is always present |
| 19:04 |
kahrl |
are you on master or a different branch? because I get no output from scaleMesh after digging default:sand |
| 19:05 |
RealBadAngel |
oops, ive put that code elsewhere, on my modified one |
| 19:05 |
RealBadAngel |
gimme a few minutes, i will upload the branch with the changes |
| 19:15 |
|
void7 joined #minetest-dev |
| 19:24 |
|
DFeniks joined #minetest-dev |
| 19:30 |
|
ElectronLibre joined #minetest-dev |
| 19:39 |
* sofar |
struggles with virtual |
| 19:40 |
RealBadAngel |
kahrl, i will test that on embeeded 1.9 irrlicht |
| 19:43 |
|
eman__ joined #minetest-dev |
| 20:09 |
sofar |
kahrl: rubenwardy: came up with this solution |
| 20:09 |
sofar |
https://gist.github.com/sofar/2d12d56ef35a1f6c78d0 |
| 20:09 |
sofar |
still fighting iterators here, doesn't compile |
| 20:25 |
sofar |
oh, type mismatch |
| 20:25 |
sofar |
[ 4%] Linking CXX executable ../../bin/minetest |
| 20:25 |
sofar |
[100%] Built target minetest |
| 20:25 |
sofar |
heh, neat |
| 20:27 |
Megaf |
:) |
| 20:27 |
sofar |
fracking awesome, that's a major step done |
| 20:27 |
Megaf |
It's such a good feeling when you change the code and it compiles |
| 20:27 |
sofar |
permitted the order is not changing :) |
| 20:28 |
Megaf |
I'd like to understand how the windows zip is generated when compiling using mingw |
| 20:28 |
Megaf |
Run CPack packaging tool... |
| 20:28 |
Megaf |
CPack: Create package using ZIP |
| 20:28 |
Megaf |
this step, I'd like to customize it |
| 20:55 |
|
pozzoni joined #minetest-dev |
| 20:57 |
|
Amaz joined #minetest-dev |
| 21:01 |
|
EvergreenTree joined #minetest-dev |
| 21:12 |
|
PenguinDad joined #minetest-dev |
| 21:12 |
|
sdzen joined #minetest-dev |
| 21:19 |
|
Robert_Zenz joined #minetest-dev |
| 21:24 |
Skyduskguy |
is a mesh just a b3d? |
| 21:24 |
Skyduskguy |
or obj |
| 21:25 |
sofar |
can be both |
| 21:25 |
sofar |
I think b3d are animated meshes |
| 21:26 |
sofar |
obj are just static |
| 21:26 |
kaadmy |
.x handles animations but animations are broken |
| 21:26 |
sofar |
I usually make blender files (.blend) and then export to obj |
| 21:26 |
Skyduskguy |
ok i jusr reg a node with drawtype of mesh n use an obj? |
| 21:26 |
kaadmy |
they're warped up in-game |
| 21:26 |
kaadmy |
Skyduskguy: that should work afaik |
| 21:26 |
Skyduskguy |
kewl ty |
| 21:27 |
Skyduskguy |
im about to go play some cards but i got an idea cooking |
| 21:27 |
sofar |
Skyduskguy: example mod: github.com/minetest-mods/warps |
| 21:27 |
Skyduskguy |
i was lookin at simply slopes for the ref |
| 21:27 |
Skyduskguy |
seems legit |
| 21:27 |
sofar |
there are many slope mods around, you may not need to make a new model |
| 21:28 |
Skyduskguy |
oh not for the model |
| 21:28 |
Skyduskguy |
just how the node is regged |
| 21:32 |
|
eman_ joined #minetest-dev |
| 22:07 |
|
pozzoni joined #minetest-dev |
| 22:25 |
|
pozzoni joined #minetest-dev |
| 22:39 |
|
Robert_Zenz joined #minetest-dev |
| 22:44 |
|
Robert_Zenz joined #minetest-dev |
| 22:46 |
|
Player2 joined #minetest-dev |
| 23:01 |
|
sdzen joined #minetest-dev |
| 23:07 |
|
pozzoni joined #minetest-dev |
| 23:09 |
|
rubenwardy joined #minetest-dev |
| 23:23 |
|
luizrpgluiz joined #minetest-dev |
| 23:26 |
luizrpgluiz |
hello, put support for colorful chat in minetest? I looked and saw that the github closed the item was opened in github minetest |
| 23:26 |
luizrpgluiz |
sorry for my english |
| 23:26 |
luizrpgluiz |
:/ |
| 23:27 |
rubenwardy |
Colored chat is long overdue |
| 23:28 |
rubenwardy |
It's a small feature, but makes a world of difference for server environmets |
| 23:37 |
|
pozzoni joined #minetest-dev |
| 23:38 |
luizrpgluiz |
but it would be possible to put the next version of minetest? |
| 23:48 |
exio4 |
hey luizrpgluiz |
| 23:48 |
exio4 |
aunts and friends are going to brazil tonight |
| 23:48 |
luizrpgluiz |
:) |
| 23:48 |
exio4 |
do you live close to the big cities? |
| 23:48 |
luizrpgluiz |
no |
| 23:48 |
exio4 |
or in the other 90% of brazil? |
| 23:48 |
exio4 |
:P |
| 23:54 |
luizrpgluiz |
exio4:well, I live in a small town in the state of sao paulo |
| 23:54 |
exio4 |
nice :) |
| 23:55 |
exio4 |
oh, you aren't in the main channel :( |
| 23:56 |
luizrpgluiz |
? |
| 23:56 |
luizrpgluiz |
oh,yes |
| 23:57 |
luizrpgluiz |
I came here to take some questions I've seen on github that are still open there |
