| Time |
Nick |
Message |
| 00:00 |
RealBadAngel |
Cybersoft, look in lua_api.txt for examples |
| 00:00 |
RealBadAngel |
you need to change visual_scale for node |
| 00:01 |
RealBadAngel |
or define nodebox to have different sized node |
| 00:02 |
RealBadAngel |
better to define smaller node and apply modified textures to it |
| 00:03 |
RealBadAngel |
or if you want just to apply the texture from mc, cut borders in gimp and resize rest to 16x16 again |
| 00:03 |
SiegeWizard |
not a good idea to increase images sizes |
| 00:05 |
Cybersoft |
I'm using a 128x texture |
| 00:05 |
Cybersoft |
the problem is that in mc the cactus is about 80% of a block, and judging from the transparent gaps on all size it seem to be 100% in minetest |
| 00:11 |
RealBadAngel |
it is 100% |
| 00:13 |
RealBadAngel |
if you want some better textures use those for example: https://github.com/RealBadAngel/HavenTP/archive/master.zip |
| 00:13 |
RealBadAngel |
or https://github.com/VanessaE/128px-realistic-textures/archive/master.zip |
| 00:14 |
RealBadAngel |
both are 128x, VanessaE has also different sized version of it |
| 00:17 |
Cybersoft |
thanks! |
| 01:14 |
Cybersoft |
is there a todo list somewhere? |
| 01:14 |
Cybersoft |
or a list of bugs? requests? |
| 01:15 |
Exio |
in github :P |
| 01:16 |
Exio |
a big list of issues |
| 01:22 |
Cybersoft |
I'll see if there is something relatively easy I could do before I start throwing ideas around |
| 01:22 |
Cybersoft |
and I need to read more on voxels... |
| 01:34 |
Cybersoft |
anyway the whole thing about spheres in #minetest was to code dynamic blocks in the engine |
| 01:34 |
Cybersoft |
ie: blocks that would be generated on the fly instead of pre-defined |
| 01:35 |
Cybersoft |
and which couldn't be destroyed/built without destroying/bulding the whole structure |
| 03:52 |
Cybersoft |
would it be useful if I propose a code to change the part where the size of heart.png is hardcoded? so using a 32x heart texture will not work? |
| 05:01 |
|
jyfl987 joined #minetest-dev |
| 08:10 |
|
sfan5 joined #minetest-dev |
| 08:32 |
|
sfan5 joined #minetest-dev |
| 08:42 |
|
SiegeWizard joined #minetest-dev |
| 09:59 |
celeron55 |
so this is today's thing http://forum.minetest.net/viewtopic.php?id=4243 |
| 10:00 |
celeron55 |
http://dev.minetest.net/MinetestWiki:TODO |
| 10:05 |
celeron55 |
#minetest-doc is currently relatively empty and almost everybody in there is also in here, so i'll talk here until a bit more people come on #-doc, as this channel has logs and so (while the other doesn't) |
| 10:06 |
celeron55 |
i'm currently considering whether i'd be able to write a more design-oriented "Engine structure" page (compared to the simple "list of things" style of "Core Architecture") |
| 10:07 |
celeron55 |
i wonder how much of history it should contain |
| 10:17 |
SiegeWizard |
celeron |
| 10:17 |
SiegeWizard |
as i said yesterday |
| 10:17 |
SiegeWizard |
i think that |
| 10:17 |
SiegeWizard |
more important that documenting it on the wiki |
| 10:17 |
celeron55 |
i say you should use less lines |
| 10:18 |
SiegeWizard |
is documenting it with coments |
| 10:18 |
SiegeWizard |
ok |
| 10:18 |
SiegeWizard |
someone proposed doxygen-style comenting |
| 10:19 |
SiegeWizard |
that would transform the coments inside the code into a wiki too, killing two birds with a single shoot |
| 10:20 |
celeron55 |
comments for the sake of comments, only duplicating what the function definitions already say, are useless - and also, i have seen more people ask for overall documentation rather than that of small details |
| 10:21 |
celeron55 |
but of course if somebody adds some actually useful doxugen comments, they are then actually useful 8) |
| 10:22 |
SiegeWizard |
thats the point i tried to make yesterday, you don't need to be yourself comenting it, at least not what is already coded, but comenting the code you do from now on would be a way of getting things done |
| 10:22 |
SiegeWizard |
the rest of us could help comenting the code that is already implemented |
| 10:22 |
celeron55 |
i haven't been doing much of minetest in a longish time |
| 10:23 |
celeron55 |
you're talking to the wrong guy 8) talk to new contributors |
| 10:23 |
SiegeWizard |
PS: i don't know how to add another channel, could you explain so i join the new one? |
| 10:24 |
celeron55 |
/join #minetest-doc will get you on that one; it's very very quiet at the moment though |
| 10:25 |
celeron55 |
#minetest is the general-purpose channel that is usually very active |
| 10:25 |
SiegeWizard |
should we start talking or is better to wait for more of us? |
| 10:26 |
celeron55 |
marktraceur is the one organizing most of this, but he's still sleeping; i don't know exactly what he has in mind |
| 10:27 |
SiegeWizard |
quoted from http://dev.minetest.net/MinetestWiki:TODO: Engine documentation List here things we need to document, how we will document those things, and more. Maybe use Doxygen-style comments. Yes, do that. |
| 10:47 |
|
rsiska joined #minetest-dev |
| 11:01 |
|
Calinou joined #minetest-dev |
| 11:02 |
celeron55 |
SiegeWizard: you should probably get it started then, so that people can do it |
| 11:02 |
SiegeWizard |
and what should i start |
| 11:02 |
SiegeWizard |
i was wondering if someone could explain git/github to me as I haven't use them ever |
| 11:03 |
Calinou |
SiegeWizard: there is a guide on github ;-) |
| 11:03 |
Calinou |
http://github.com/ |
| 11:03 |
celeron55 |
i assume doxygen requires some kind of configuration or something, and people require a file that is doxygenized as an example |
| 11:04 |
SiegeWizard |
but everyone doesn't need to create the doxygen files |
| 11:05 |
SiegeWizard |
they only owuld be created every "semi-important" version |
| 11:05 |
SiegeWizard |
people needs to know how to comment it so that doxygen udnerstands the comments |
| 11:05 |
celeron55 |
but there needs to be a consistent way of generating it; it's not likely that somebody will personally do it for years to go |
| 11:06 |
celeron55 |
people come and go - that is the first rule of libre open source stuff |
| 11:06 |
celeron55 |
code stays |
| 11:07 |
SiegeWizard |
a guide on how to generate the doxygen documentation could be fine |
| 11:07 |
Calinou |
http://paste.ubuntu.com/1548711/ |
| 11:07 |
celeron55 |
SiegeWizard: do it! |
| 11:07 |
Calinou |
(not perfect at all, feel free to comment :p) |
| 11:09 |
SiegeWizard |
ok, but i think it's more important to tell people how to coment first |
| 11:09 |
SiegeWizard |
if there are no coments, there's no sense in getting a tutorial on how to make the documentation of those coments |
| 11:09 |
celeron55 |
people won't comment before they can see what comes out of the comments |
| 11:09 |
|
kaeza joined #minetest-dev |
| 11:10 |
SiegeWizard |
that's a good point |
| 11:18 |
|
Infinity_69 joined #minetest-dev |
| 11:19 |
Infinity_69 |
hi guys :D |
| 11:19 |
Infinity_69 |
VanessaE |
| 11:19 |
Infinity_69 |
Celeron55 & 56 |
| 11:19 |
Infinity_69 |
kaeza |
| 11:19 |
Infinity_69 |
loggingbot |
| 11:19 |
Infinity_69 |
sfan5 |
| 11:19 |
Infinity_69 |
sfan5|OFF |
| 11:19 |
Infinity_69 |
thexyz |
| 11:19 |
Infinity_69 |
marktraceur |
| 11:19 |
celeron55 |
are you in need of a ban? |
| 11:21 |
|
john_minetest joined #minetest-dev |
| 11:24 |
|
john_minetest left #minetest-dev |
| 11:29 |
|
Oblivion1500 joined #minetest-dev |
| 11:29 |
Infinity_69 |
no |
| 11:30 |
Infinity_69 |
Oblivion1500 |
| 11:30 |
Infinity_69 |
can you do me a favour? |
| 11:30 |
Oblivion1500 |
whats that? |
| 11:30 |
Infinity_69 |
what os do you run atm? |
| 11:30 |
Oblivion1500 |
im running the linux mint 14 distro |
| 11:30 |
Infinity_69 |
fuck... |
| 11:31 |
Oblivion1500 |
why? |
| 11:31 |
Infinity_69 |
Celeron55, do you run Windows? |
| 11:31 |
Infinity_69 |
a program doesnt start :/ |
| 11:31 |
Oblivion1500 |
what program? |
| 11:32 |
Infinity_69 |
a cs 1.6 decal converter |
| 11:32 |
Infinity_69 |
wait here |
| 11:32 |
SiegeWizard |
he really wants a ban |
| 11:32 |
Infinity_69 |
http://forum.minetest.net/viewtopic.php?id=4496 |
| 11:32 |
Infinity_69 |
no i dont |
| 11:33 |
SiegeWizard |
u think the dev IRC is for that kind of support? |
| 11:33 |
|
Infinity_69 was kicked by celeron55: wrong channel |
| 11:46 |
|
PilzAdam joined #minetest-dev |
| 12:04 |
|
darkrose joined #minetest-dev |
| 12:04 |
|
darkrose joined #minetest-dev |
| 12:06 |
celeron55 |
i wonder if this is going to the right direction http://dev.minetest.net/Engine_structure |
| 12:17 |
celeron55 |
...at least the wrong channel |
| 12:37 |
|
john_minetest joined #minetest-dev |
| 12:58 |
|
Gambit joined #minetest-dev |
| 13:20 |
|
SpeedProg joined #minetest-dev |
| 13:37 |
|
sfan5 joined #minetest-dev |
| 13:40 |
|
rubenwardy joined #minetest-dev |
| 14:09 |
|
klunk joined #minetest-dev |
| 15:10 |
|
iqualfragile joined #minetest-dev |
| 15:18 |
|
Cybersoft joined #minetest-dev |
| 15:28 |
|
Gambit joined #minetest-dev |
| 15:29 |
|
klunk left #minetest-dev |
| 15:42 |
|
Gambit joined #minetest-dev |
| 16:19 |
|
sfan5 joined #minetest-dev |
| 16:19 |
|
Progers joined #minetest-dev |
| 16:20 |
Progers |
Hi |
| 16:20 |
sfan5 |
hi |
| 16:21 |
Progers |
I want to ask,MineTest was coded on QT or Eclipse? |
| 16:21 |
Cybersoft |
Everyone probably use their own prefered IDE |
| 16:21 |
Progers |
I know |
| 16:22 |
Progers |
But wanted to know on what IDE minetest was coded |
| 16:23 |
sfan5 |
IIRC visual studio |
| 16:23 |
Progers |
Ok |
| 16:23 |
sfan5 |
but celeron decided to switch to cmake |
| 16:24 |
Exio |
I think c55 used notepad at the first! |
| 16:24 |
Progers |
:D |
| 16:24 |
Progers |
Nano :D |
| 16:24 |
PilzAdam |
vim ftw |
| 16:24 |
sfan5 |
what about vim? |
| 16:25 |
celeron55 |
what |
| 16:26 |
celeron55 |
i code in vim |
| 16:26 |
celeron55 |
since long before i started minetest |
| 16:26 |
Progers |
good |
| 16:27 |
Progers |
I started learn C++ 4 month ago =) |
| 16:28 |
Progers |
The best language |
| 16:28 |
Cybersoft |
that's could be discussed at lenght |
| 16:28 |
sfan5 |
^ i agree |
| 16:28 |
Cybersoft |
*that |
| 16:28 |
sfan5 |
but c has the fastest compiler (tcc) |
| 16:29 |
Cybersoft |
If I want to do something easily I use C#. If I want it to be robust I use C++, if I want to do insanely fast math libs I might use ocaml |
| 16:29 |
Progers |
How about java? java is better than C# i think? No? |
| 16:31 |
Progers |
sorry for bad english,i am from Russia :D |
| 16:31 |
darkrose |
java is better than c# in the same way that horse sh** is better than dog sh** |
| 16:32 |
Cybersoft |
I don't think any language is better or worse than any other language. I'm a professionnal C# developper, so for me java would be a worse language because I would have to learn it before I could code with it. C++ is harder because it has pointers and memory management |
| 16:32 |
Cybersoft |
and the C++ standard library is currently evolving, so what you learned in school or from book will need updating |
| 16:35 |
|
Calinou joined #minetest-dev |
| 16:36 |
Progers |
i am now learning (Object Oriented Programming - C++ Lafore) |
| 16:38 |
Cybersoft |
and at some point you have to decide what you will learn and what you will not... I'm an ex-linux user so I decided to focus on c++ on .net and microsoft technologies (because I have limited time) |
| 16:39 |
Cybersoft |
so if I worked for Oracle it wouldn't be an ideal fit, but working for a microsoft dev shop is a better choice |
| 16:39 |
Progers |
You are Expert Linux User? |
| 16:41 |
Cybersoft |
No., "ex" means I used it previously, but I don't it anymore |
| 16:41 |
Cybersoft |
*I don't use it anymore |
| 16:42 |
Progers |
Oh |
| 16:43 |
Progers |
What did you use? Kubuntu? or Ubuntu? |
| 16:43 |
Progers |
:D |
| 16:43 |
Cybersoft |
Slackware, Redhat, SuSE ... mostly suse when it was an independant company |
| 16:43 |
darkrose |
guys, take the chatter elsewhere |
| 16:43 |
Cybersoft |
I sometimes use ubuntu in virtual machines |
| 16:43 |
Progers |
I liked Kubuntu =) |
| 16:44 |
Progers |
But i had some errors with QT |
| 16:44 |
Progers |
Who use Eclipse? |
| 16:47 |
|
hmmmm joined #minetest-dev |
| 16:47 |
Calinou |
eclipse for C++? |
| 16:47 |
Calinou |
erm, what did I say |
| 16:47 |
Calinou |
eclipse for any programming language? |
| 16:47 |
Progers |
Yes! |
| 16:48 |
Calinou |
Cybersoft: .net is da hell |
| 16:48 |
Calinou |
or da shit/the suck/anything |
| 16:50 |
Cybersoft |
maybe, maybe not, but it is what I'm paid to do |
| 16:51 |
Cybersoft |
so I gotta drink the kool-aid |
| 16:51 |
Cybersoft |
when you are paid to code your perspective is different... |
| 16:52 |
Progers |
=) |
| 17:08 |
Calinou |
i'd write c# only if i was paid for doing that :P |
| 17:13 |
Progers |
hehe |
| 17:38 |
|
berome joined #minetest-dev |
| 17:38 |
Cybersoft |
I'd use linux only if I was paid for doing that ;) |
| 17:38 |
sfan5 |
:( |
| 17:39 |
thexyz |
are you all sure this is appropriate channel to discuss all this minetest-unrelated stuff? |
| 17:43 |
VanessaE |
celeron55: is there any chance of getting you to tweak the register_on_generated() call to return 1 mapblock at a time instead of 5x5x5 ? |
| 17:44 |
VanessaE |
this adds incredibly huge latency when checking through the returned nodes for something. |
| 17:44 |
thexyz |
>Documentation today in #minetest-doc and http://dev.minetest.net |
| 17:44 |
thexyz |
you a word |
| 17:49 |
|
klunk left #minetest-dev |
| 17:57 |
|
doserj joined #minetest-dev |
| 17:57 |
|
john_minetest left #minetest-dev |
| 18:00 |
celeron55 |
VanessaE: it will cause too much latency anyway; the right thing to do currently is to figure out what kind of functions should be implemented in C++ to facilitate the things that are slow to do in lua |
| 18:02 |
VanessaE |
celeron55: mind if I make a suggestion to that end: find_open_spaces(minp, maxp, radius, {list of nodes to avoid}) |
| 18:03 |
VanessaE |
should return a list of positions that are more than <radius> from all items in {list} within the minp/maxp area |
| 18:04 |
VanessaE |
effectively, a density function. |
| 18:04 |
VanessaE |
I'd point you to the best lua implementation I've come up with for that, but I assume you have no time now. |
| 18:05 |
Cybersoft |
should I put a bug on github before I attempt to fix it? |
| 18:07 |
thexyz |
it depends, if that's some urgent security-related issue, better talk to c55 privately |
| 18:09 |
Cybersoft |
no, it's just being unable to use a texture with any size for hearts |
| 18:09 |
Cybersoft |
because it's hardcoded to be 16x, so if it's bigger it won't resize it and there will be display issues |
| 18:10 |
|
kaeza_ joined #minetest-dev |
| 18:10 |
thexyz |
well, if you can (and will) fix it [in reasonable amount of time], then just do that and submit a pull request |
| 18:14 |
VanessaE |
actually my suggestion might be useful to some, but it won't help in my particular case :-/ |
| 18:16 |
Cybersoft |
VanessaE- that function sounds AIish |
| 18:16 |
VanessaE |
Cybersoft: not really. in Lua it's trivial to do, but slow |
| 18:16 |
Cybersoft |
it's exactly the kind of function you would use if you wanted an NPC to avoid hazards like holes or lava |
| 18:17 |
VanessaE |
hm... I suppose it would be useful for that. |
| 18:17 |
Cybersoft |
what is a node in minetest? a block? part of a block? |
| 18:17 |
VanessaE |
a 1m cube e.g. a piece of dirt or cobble |
| 18:17 |
PilzAdam |
http://dev.minetest.net/terminology |
| 18:18 |
thexyz |
how about renaming it to "Terminology"? |
| 18:18 |
PilzAdam |
+1 |
| 18:18 |
thexyz |
done |
| 18:18 |
Cybersoft |
how many voxels are there in a block? |
| 18:19 |
VanessaE |
depends on how you want to define voxel. |
| 18:19 |
PilzAdam |
Cybersoft, http://dev.minetest.net/terminology |
| 18:19 |
VanessaE |
commonly it seems to equate to a single node |
| 18:20 |
VanessaE |
but if you wanna be picky, you could say it's 4096 per node (since they use 16px textures by default) |
| 18:20 |
Cybersoft |
so if I wanted to deform a block in an experiment, I would have control of 4096 voxels for a single block? |
| 18:21 |
VanessaE |
nope |
| 18:21 |
VanessaE |
you'd control the node as a whole. |
| 18:21 |
VanessaE |
it's still one voxel the way it's drawn, it just *looks* like 4096 because of the texture resolution. |
| 18:22 |
Cybersoft |
I think I'm asking the wrong questions |
| 18:23 |
VanessaE |
celeron55: if you wanna take a look at how I implemented it in Lua, here: https://github.com/VanessaE/plantlife/blob/mapgenbased/plants_lib/init.lua#L81 |
| 18:23 |
VanessaE |
the code above this point is as fast as I could have wanted, but from this point to line 88, it's slow for reasons I am sure are obvious. |
| 18:23 |
VanessaE |
so what's needed is a way to do that ^^^^ in a single C++ function call. |
| 18:27 |
Cybersoft |
I think a better question would be, how much control do I have on the shape of a block from within the c++ code without using lua nodebox with coordinates |
| 18:27 |
Cybersoft |
could I do csg operation to remove a sphere from a box for example from within the code if your basic building block is a voxel? |
| 18:51 |
doserj |
Cybersoft: in c++ you can do pretty much anything you want, you just have to build the 3d-mesh for it in the client code. |
| 18:52 |
Cybersoft |
yeah but more or less work is involved depending on how minetest does thing |
| 18:52 |
Cybersoft |
from mesh.cpp I can see that the only kind of basic mesh you use are blocks |
| 18:52 |
Cybersoft |
*cubes |
| 18:55 |
doserj |
enum NodeDrawType in nodedef.h shows you what kind of things the engine currently knows about |
| 18:56 |
doserj |
plantlike, for example, is not a block ;) |
| 18:56 |
Cybersoft |
it's a billboard / sprite no? |
| 18:57 |
Cybersoft |
I mean the way they look ingame looks like sprites or billboards |
| 18:57 |
doserj |
not a billboard |
| 18:58 |
doserj |
(billboard always face the camera) |
| 18:59 |
Cybersoft |
Well I was under the impression the plants were always facing the camera |
| 19:00 |
doserj |
then look again |
| 19:04 |
Cybersoft |
hum, but plants are just a texture with alpha on a square mesh no? |
| 19:05 |
RealBadAngel |
drawtype plantlike are two textures 90 deg to each other |
| 19:05 |
RealBadAngel |
btw, http://ompldr.org/vaDJiMQ/screenshot_907329753.png |
| 19:06 |
RealBadAngel |
4 textures plantlike with 45 deg |
| 19:06 |
Cybersoft |
I see |
| 19:29 |
VanessaE |
celeron55: I figured out the problem - my mapgen code is now quite fast but I need that stupid above-ground-caves-in-air fix put into place please |
| 19:29 |
VanessaE |
it is destroying trees here and there after they're generated. |
| 20:19 |
|
BackupCoder joined #minetest-dev |
| 20:52 |
iqualfragile |
"Examples include mobs, dropped items, falling sand or gravel, primed TNT." found on http://dev.minetest.net/terminology. It looks like that was taken from a text describing minecraft, but i might be wrong |
| 21:01 |
thexyz |
iqualfragile: → #minetest-doc |
| 22:35 |
Progers |
Hm 2:35 AM |
| 22:35 |
Progers |
=) |
| 23:03 |
Taoki |
RealBadAngel: I like how those leaves look like! |
| 23:05 |
RealBadAngel |
Taoki, please wait a few hours |
| 23:05 |
RealBadAngel |
VanessaE is finishin the code to place all trees in the world |
| 23:05 |
Taoki |
ok. Anyway I'd like to see that change in master |
| 23:05 |
Taoki |
awesome |
| 23:06 |
RealBadAngel |
before mapgen v7 they will propably stay as a mod |
| 23:06 |
Taoki |
ok. Waiting for the LUA-hooked mapgen too :) |
| 23:07 |
RealBadAngel |
but i can say i tested out version with 2 kinds of trees |
| 23:07 |
RealBadAngel |
it worked just fine |
| 23:07 |
RealBadAngel |
now shes having total of 11 trees |
| 23:08 |
RealBadAngel |
and tuning map generation for them |
| 23:08 |
Taoki |
The change in that screenshot doesn't look related to any mapgen BTW. More like it changes leaves from blocks to X shapes, which looks better |
| 23:08 |
RealBadAngel |
its a complete overhaul |
| 23:09 |
RealBadAngel |
new trees, trunks, leaves, saplings, planks... |
| 23:09 |
RealBadAngel |
and fruits |
| 23:09 |
RealBadAngel |
i finished my part today, now VanessaE is makin use of it |
| 23:17 |
VanessaE |
I'm slowly getting there... |
| 23:18 |
VanessaE |
busy plugging numbers into the biome defs. |
| 23:23 |
Taoki |
Still, is there something keeping the current tree leaves from working when set to an X shape instead of a block drawing type (just thatn part)? |
| 23:23 |
Taoki |
**that |
| 23:24 |
VanessaE |
probably noy |
| 23:24 |
VanessaE |
not* |
| 23:24 |
VanessaE |
they're just another node |
| 23:24 |
VanessaE |
how the fancy/new style leaves would play into that, dunno |
| 23:25 |
Taoki |
Maybe they can be set that way for the existing mapgen trees. That would be pretty :) |
| 23:25 |
VanessaE |
try it :) |
| 23:44 |
doserj |
I would guess plantlike leaves are somewhere between fast and fancy leaves in mesh complexity. fancy leaves-> 6 faces. plantlike -> 2 faces. fast leaves -> anything between 1 and 6 faces on the surface, but 0 faces in the interior. |
| 23:47 |
doserj |
oh, the 6 faces for fancy leaves is actually only 3 for interior ones |
| 23:48 |
|
iqualfragile1 joined #minetest-dev |
| 23:50 |
doserj |
Anyways, I am not sure I really like plantlike leaves. |
| 23:52 |
kaeza |
night all |
| 23:53 |
|
rsiska joined #minetest-dev |
