| Time |
Nick |
Message |
| 00:41 |
VanessaE |
before 0.4.6 goes out, that acceleration/speed bug (huge memory usage -> crash on realyl big values) needs fixed. |
| 00:42 |
|
dexter0 joined #minetest-dev |
| 00:43 |
Exio |
how can i "cause" that bug? |
| 00:43 |
Exio |
i want to reproduce it locally |
| 00:43 |
|
dexter0 joined #minetest-dev |
| 00:43 |
VanessaE |
try using jeija's jumping mod (trampolines) |
| 00:44 |
VanessaE |
the right placement of trampolines will cause you go off into orbit (Y+200K or so) |
| 00:44 |
VanessaE |
when you jump on them or run into them hard |
| 00:44 |
VanessaE |
particularly on the highest setting |
| 00:45 |
Exio |
mis pressed multiplayer tab |
| 00:46 |
Exio |
with this lag => mt stuck |
| 00:47 |
VanessaE |
it's a bug in that mod that it should cause those high accel values, but it's a bug in the engine that it should use 4-6GB of memory and crash because of it :) |
| 00:52 |
Exio |
are all trampolines the same in terms of acceleration? |
| 00:53 |
VanessaE |
dunno. |
| 00:53 |
VanessaE |
a tramp has like 6 settings, from little more than a pillow to like 50% faster going up than you hit it going down |
| 00:54 |
VanessaE |
with one tramp at the highest setting, you'll begin taking damage after three or four bounces |
| 00:54 |
VanessaE |
here's a trick guaranteed to crash your client: |
| 00:54 |
Exio |
what color is the "best"? |
| 00:55 |
Exio |
the red? |
| 00:55 |
VanessaE |
place a tramp on the ground. set it to max. Place four dirt blocks in a column next to it. Place another tramp against the column so that it's 3 blocks above the first. set them both to max (yes, red). Now jump onto the lower one. |
| 00:58 |
Exio |
hm... |
| 00:58 |
Exio |
should i kill it now or it will do a bad::alloc before it eats my 4gb ram + 16swap? |
| 00:59 |
VanessaE |
the client? |
| 00:59 |
VanessaE |
which bug did you trigger? the client crash or the server hogging RAM+CPU? |
| 00:59 |
Exio |
the client got stucked when doing that |
| 00:59 |
VanessaE |
oh that, just kill it. |
| 01:00 |
VanessaE |
I've never bothered to wait that one out. |
| 01:00 |
Exio |
but but, wait |
| 01:00 |
VanessaE |
but you can see that there's an issue. :) |
| 01:00 |
Exio |
3 bugs for one thing? |
| 01:00 |
Exio |
"client crash, client stuck, server eating your ram/cpu" |
| 01:01 |
VanessaE |
client crash = client stuck |
| 01:01 |
Exio |
ah |
| 01:01 |
VanessaE |
the server eating RAM/CPU is a different but related issue |
| 01:01 |
|
Taoki joined #minetest-dev |
| 01:01 |
VanessaE |
now, lemme see if I remember this right |
| 01:01 |
Exio |
how was for enabling the "debug" release? |
| 01:01 |
VanessaE |
hey Taoki. |
| 01:01 |
VanessaE |
next test "should" cause the server to go insane: |
| 01:02 |
VanessaE |
make a flat region at least 6x6 nodes. |
| 01:03 |
VanessaE |
Place 12 trampolines on the ground, side-by-side in a square. That is, four to a side, thus with a 2x2 hole in the middle. |
| 01:03 |
VanessaE |
set them all to max. |
| 01:03 |
VanessaE |
fly up and over and drop into the middle. |
| 01:03 |
VanessaE |
turn off flying. |
| 01:03 |
VanessaE |
turn fast on. |
| 01:03 |
VanessaE |
move to one corner. |
| 01:04 |
VanessaE |
hold your 'fast' key ("E" for me) and run around in the 2x2 open space |
| 01:04 |
VanessaE |
you should shortly get catapulted out into nowhere, perhaps into the -200k/+200k range in one of your coordinates. |
| 01:05 |
Exio |
how was for enabling the debug stuff? |
| 01:05 |
Exio |
i'll try to see what the hell happens |
| 01:05 |
VanessaE |
oh, for debug? cmake . -DCMAKE_BUILD_TYPE=Debug |
| 01:06 |
VanessaE |
(plus the RUN_IN_PLACE and whatever else you usually use) |
| 01:06 |
Exio |
grazie :P |
| 01:07 |
VanessaE |
np |
| 01:11 |
Exio |
i see |
| 01:11 |
Exio |
Program received signal SIGTSTP, Stopped (user). |
| 01:11 |
Exio |
0x00007ffff4807c50 in *__GI___libc_malloc (bytes=24) at malloc.c:3658 |
| 01:11 |
Exio |
ctrl-z and stuck at malloc, perfect :P |
| 01:12 |
VanessaE |
ah, so it happens to the client too? |
| 01:12 |
Exio |
i'm searching for the client stuck |
| 01:12 |
VanessaE |
oh ok |
| 01:24 |
Exio |
haha |
| 01:24 |
Exio |
when looking at the debug.txt |
| 01:24 |
Exio |
22:11:04: INFO[main]: collisionMoveSimple: WARNING: Loop count exceeded, aborting to avoid infiniite loop |
| 01:24 |
Exio |
22:11:04: INFO[main]: Pointing at [nothing] |
| 01:33 |
hmmmm |
http://ompldr.org/vaHluNQ |
| 01:35 |
kaeza |
wat |
| 01:35 |
VanessaE |
hmmmm: looks like the stuff we used to see in 0.3.x |
| 01:35 |
kaeza |
old 0.3.x mountains? |
| 01:35 |
VanessaE |
albeit a little "sharper" |
| 01:36 |
Exio |
am i the only who still is loading the image? |
| 01:37 |
kaeza |
also... is that thing gravel? |
| 01:39 |
Exio |
$1 = <value optimized out> |
| 01:39 |
Exio |
fuck |
| 01:40 |
VanessaE |
Exio: make clean ; CXXFLAGS=-O0 make -j4 |
| 01:41 |
VanessaE |
(that's "dash oh zero" of course) |
| 01:41 |
Exio |
i know, but i want a proper fix |
| 01:41 |
VanessaE |
I mean that'll disable the optimization so you can see what was optimized out :) |
| 01:41 |
Exio |
("proper" == edit the CMakelists |
| 01:41 |
Exio |
) |
| 01:41 |
VanessaE |
oh. |
| 01:41 |
VanessaE |
heh |
| 01:42 |
Exio |
ok, compiled |
| 01:46 |
hmmmm |
this is no 0.3.x, boy! |
| 01:47 |
Exio |
this is v7! </hmmmm> |
| 01:47 |
hmmmm |
too bad nobody'll get the reference :/ |
| 01:47 |
Exio |
well |
| 01:47 |
Exio |
what reference? ^^ |
| 01:47 |
hmmmm |
nvm |
| 01:49 |
hmmmm |
anyway, the dirt is from 3d noise, the stone is from 2d |
| 01:49 |
hmmmm |
i tried the idea i had about how to make ridges from 2d noise but it didn't work out as planned at all |
| 01:50 |
hmmmm |
3d noise is basically the only way to make ridges |
| 01:51 |
hmmmm |
what i really want the 3d noise to do is to form mostly on higher parts, like the mountains and what not |
| 01:51 |
hmmmm |
a problem with mapgen v5 was that you almost never had flat terrain to work with... it looked nice, but you couldn't build crap |
| 01:51 |
hmmmm |
my usage of 2d noise as a base eliminates that |
| 01:53 |
Exio |
haha |
| 01:53 |
Exio |
this is funny |
| 01:53 |
Exio |
now i get the "stuck" ... later! with the same error |
| 01:54 |
VanessaE |
great, now exio has a new toy to play with :) |
| 01:59 |
hmmmm |
actually i think the negative of this terrain is more interesting |
| 01:59 |
hmmmm |
i made the 3d noise air |
| 01:59 |
hmmmm |
to carve out chunks of the 2d generated terrain |
| 02:00 |
VanessaE |
screenshot? |
| 02:00 |
hmmmm |
well, i mean, i need to make the 3d noise more sparse now |
| 02:00 |
hmmmm |
but sure, i'll take a screenshot as-is |
| 02:01 |
hmmmm |
with this you get the benefits of both 3d and 2d noise and none of the disadvantages |
| 02:02 |
hmmmm |
minetest-0.4.5 is the 5th most popular file on omploader :) hah |
| 02:02 |
VanessaE |
too bad 0.4.6 is out soon :D |
| 02:03 |
hmmmm |
what on earth |
| 02:03 |
hmmmm |
both zimg and omploader are having problems |
| 02:03 |
Exio |
how i thinked |
| 02:04 |
hmmmm |
well, whatever. |
| 02:04 |
VanessaE |
the whole net's fucked up today |
| 02:04 |
Exio |
added a printf("--stuff--\n", dtime); before the "collision detection" |
| 02:04 |
Exio |
get mt gets stuck |
| 02:05 |
Exio |
=> my term full of --stuff-- and more --stuff-- :D |
| 02:12 |
Exio |
wtf, printf("..\n", dtime); |
| 02:12 |
Exio |
i need to read what i write |
| 02:12 |
VanessaE |
oops? :) |
| 02:16 |
hmmmm |
http://ompldr.org/vaHluZQ |
| 02:22 |
kaeza |
https://github.com/minetest/minetest/pull/589 |
| 02:23 |
RealBadAngel |
kaeza, im pretty sure this will fuck up stargates ;) |
| 02:23 |
kaeza |
how so? |
| 02:23 |
RealBadAngel |
8/9 segments of it doesnt have selection box |
| 02:24 |
RealBadAngel |
9th covers them all |
| 02:24 |
kaeza |
hmm... |
| 02:24 |
RealBadAngel |
can you test it? |
| 02:25 |
kaeza |
sure |
| 02:25 |
kaeza |
lemme fetch technic |
| 02:25 |
RealBadAngel |
i may be wrong, maybe the bigger one will override it |
| 02:26 |
RealBadAngel |
hmmmm, is this supposed to be cavegen? |
| 02:26 |
hmmmm |
no |
| 02:26 |
hmmmm |
that's just terrain |
| 02:26 |
VanessaE |
hmmmm: I *like* that last screenshot |
| 02:26 |
VanessaE |
THAT is the sort of terrain most have been wanting, I think |
| 02:26 |
VanessaE |
except those distant mountains are a little too simple, maybe |
| 02:27 |
hmmmm |
no shit |
| 02:27 |
VanessaE |
(but you said this was indeed simple for the moment) |
| 02:27 |
hmmmm |
like i'm not trying to fix that |
| 02:28 |
kaeza |
wait... RealBadAngel you said 8/9 of your stargates do not use a selection box. you mean a nil selection box or a 0x0x0 selection box? |
| 02:29 |
RealBadAngel |
not defined at all |
| 02:29 |
RealBadAngel |
only bottom middle node has the selection box to cover 3x3 area |
| 02:30 |
kaeza |
well... let's see if this breaks 'em |
| 02:32 |
Exio |
i think i got *where* is the problem |
| 02:33 |
RealBadAngel |
is my system broken then? |
| 02:34 |
Exio |
haha |
| 02:35 |
Exio |
enviroment.cpp:2025-2103 (at least here, but if you see |
| 02:35 |
Exio |
gets a infinite loop |
| 02:35 |
Exio |
if you see the code you will get where it is) * |
| 02:36 |
Exio |
now to check the code and see what really is wrong |
| 02:37 |
kaeza |
RealBadAngel, how do I get/craft a stargate? |
| 02:37 |
RealBadAngel |
no craft recipe yet |
| 02:37 |
kaeza |
itemstring? |
| 02:37 |
RealBadAngel |
use givme stargate:stargate |
| 02:37 |
kaeza |
ok |
| 02:38 |
kaeza |
cannot give an unknown item |
| 02:38 |
RealBadAngel |
i havent made one yet cause im not sure stargates should be standalone or tied to technic |
| 02:38 |
RealBadAngel |
ahh, wait |
| 02:39 |
kaeza |
well... I'd prefer stargates to be standalone.. but it's your mod anyway :P |
| 02:39 |
RealBadAngel |
stargate:gatenode8_off |
| 02:40 |
kaeza |
that worked |
| 02:40 |
kaeza |
there isn't any conflict with my change and the stargates it seems |
| 02:41 |
RealBadAngel |
good :) |
| 02:46 |
kaeza |
probably unrelated: 23:43:53: ERROR[main]: ServerError: LuaError: error: ...sting/mods/calinou_mods/moreblocks/redefinitions.lua:191: attempt to index local 'ret' (a nil value) |
| 02:46 |
kaeza |
that happened when right-clicked the stargate after activation |
| 02:48 |
Exio |
hahahaha |
| 02:48 |
Exio |
got teleported to -90 with the trampolines |
| 02:48 |
Exio |
VanessaE: i got that stuff |
| 02:48 |
Exio |
-nan! |
| 02:49 |
VanessaE |
hah! |
| 02:49 |
VanessaE |
but did it send the server to never-never land? |
| 02:49 |
Exio |
never-never land? no idea, my coords are, actually: 832, -nan, 327 |
| 02:49 |
Exio |
and everything is grey :) |
| 02:50 |
VanessaE |
as in the RAM+CPU insanity :-) |
| 02:50 |
Exio |
one cpu at 100% the others normally, between 10 to 30% |
| 02:50 |
Exio |
% |
| 02:51 |
Exio |
and nope, not eating my ram |
| 02:52 |
ShadowNinja |
hmmmm: https://github.com/minetest/minetest/pull/590 |
| 02:57 |
Exio |
-215000 |
| 02:57 |
Exio |
my y cord VanessaE |
| 02:57 |
VanessaE |
heh |
| 02:57 |
VanessaE |
well at least you can see one of the bugs :) |
| 02:57 |
Exio |
and 4 FPS |
| 02:57 |
VanessaE |
I'm kinda surpri....... |
| 02:57 |
VanessaE |
uh oh. |
| 02:57 |
VanessaE |
better check your RAM/CPU usage :) |
| 02:58 |
Exio |
it wasn't high |
| 02:58 |
Exio |
i added a "limit" to the infinite loop causing that |
| 02:58 |
VanessaE |
oh? |
| 02:58 |
Exio |
i'm debugging it :P |
| 02:59 |
VanessaE |
have fun :-) |
| 03:00 |
Exio |
10 FPS! when limiting to 250 instead of inf |
| 03:01 |
VanessaE |
heh |
| 03:01 |
Exio |
it is funny because my player is now in, .. nonwhere? |
| 03:01 |
VanessaE |
yup, you're in The Voice |
| 03:01 |
VanessaE |
fuck! |
| 03:01 |
VanessaE |
yup, you're in The VOID |
| 03:01 |
Exio |
i leaved the matrix |
| 03:01 |
Exio |
>:D |
| 03:01 |
* VanessaE |
takes some of exio's red pills |
| 03:02 |
Exio |
haha |
| 03:02 |
Exio |
wooo |
| 03:03 |
Exio |
8309249873236578271260856811520 |
| 03:03 |
Exio |
my y coord |
| 03:03 |
VanessaE |
WTF? |
| 03:03 |
VanessaE |
O_O |
| 03:03 |
Exio |
yep |
| 03:04 |
Exio |
i see why mt gets stuck xD |
| 03:04 |
Exio |
so, the problem is not that while |
| 03:04 |
ShadowNinja |
oh, make sure you enable text-wrap for the coords :-) |
| 03:05 |
Exio |
hahaha |
| 03:16 |
Exio |
if(bouncy_jump && control.jump){ |
| 03:16 |
Exio |
m_speed.Y += movement_speed_jump*BS; |
| 03:16 |
Exio |
i think i've found what is wrong |
| 03:18 |
VanessaE |
which is? |
| 03:18 |
VanessaE |
(I don't know how that code works) |
| 03:19 |
Exio |
it just adds the "speed jump" to the actual speed |
| 03:19 |
Exio |
i added a "printf" after that for the value of m_speed.Y |
| 03:19 |
Exio |
http://dpaste.com/1043754/ |
| 03:20 |
VanessaE |
lol |
| 03:20 |
VanessaE |
um |
| 03:20 |
VanessaE |
yeah that would be the problem |
| 03:20 |
Exio |
well, that makes me feel a bit less useless ^^ |
| 03:21 |
Exio |
now to get a proper, or at least decent, fix |
| 03:21 |
VanessaE |
m_speed += movement_speed_jump*BS; if ( m_speed > 5000 ); then m_speed = 500; fi |
| 03:21 |
VanessaE |
or something similar |
| 03:21 |
VanessaE |
I don't know enough C++ to get that right the first time :-) |
| 03:21 |
VanessaE |
er, that was supposed to be a 500, not 5000 |
| 03:21 |
Exio |
a speed_cap? |
| 03:21 |
Exio |
i'll do that and send a pull request |
| 03:21 |
VanessaE |
yeah. assuming that's nodes/sec, 500 is already quite insane |
| 03:22 |
VanessaE |
might want to define some constant at 500 or 1000 or something and use that, so it can be easily referenced later. e.g. your speed_cap idea |
| 03:23 |
VanessaE |
but first see if that works ;) |
| 03:23 |
Exio |
i'll check what it does when i limit the m_speed.Y=5000 and then talk |
| 03:23 |
Exio |
hehe, yeah |
| 03:24 |
VanessaE |
yeah, make both 5000 |
| 03:25 |
Exio |
it seems a "perfect" fix |
| 03:25 |
VanessaE |
I have to assume that X and Z will need similar limits eventually, though for now it's okay |
| 03:25 |
Exio |
the only change what is "done" in the "bouncy stuff" is for the Y coord, not the X or Z |
| 03:25 |
VanessaE |
oh ok |
| 03:25 |
VanessaE |
right, that makes sense. duh. :D |
| 03:26 |
Exio |
ehm... i wonder what happened when i started writing here.. |
| 03:26 |
Exio |
now it is in ... -nan other time |
| 03:26 |
VanessaE |
eh? |
| 03:27 |
Exio |
exactly, no idea what happened :P |
| 03:28 |
VanessaE |
but it's okay now? |
| 03:28 |
Exio |
nah, now it crashes, but after more time of jumping |
| 03:28 |
VanessaE |
(I assume the NaN is from a divide-by-zero error or so) |
| 03:28 |
Exio |
s/crashes/gets stuck/ |
| 03:29 |
Exio |
new value of m_speed!!!!!!!!11111!!one: 5000.000000 |
| 03:29 |
Exio |
new value of m_speed!!!!!!!!11111!!one: -1464338519404052480.000000 |
| 03:29 |
Exio |
VanessaE: +5000 and -5000 |
| 03:30 |
VanessaE |
oh my |
| 03:30 |
VanessaE |
oh yeah, I guess you need a negative cap of 0 |
| 03:30 |
VanessaE |
or is m_speed more of a delta thing than a raw speed? |
| 03:31 |
Exio |
i'm not fully sure |
| 03:31 |
Exio |
but i'd say it is more like a raw speed, but i repeat |
| 03:31 |
VanessaE |
what's the variable type being used here? |
| 03:31 |
Exio |
NOTSURE :P |
| 03:31 |
hmmmm |
ENOTSURE |
| 03:31 |
VanessaE |
trying to figure out how it's wrapping around to -14 bazillion |
| 03:31 |
Exio |
double, i think, if not, float :P |
| 03:31 |
hmmmm |
what are you doing |
| 03:32 |
VanessaE |
hmmmm: he's fixing the client hang/crash that happens with certain usages of the jumping code |
| 03:32 |
hmmmm |
the correct way to fix the collision thing is to add bounds to how much is allocated |
| 03:32 |
VanessaE |
place a jeija "jumping" trampoline, place another 3 nodes above it. set both to max. jump on the bottom one. You die + client crash. |
| 03:32 |
VanessaE |
this is a different bug |
| 03:33 |
VanessaE |
he found it before the collision bug :D |
| 03:33 |
hmmmm |
ah |
| 03:33 |
|
jojoa1997|Tablet joined #minetest-dev |
| 03:33 |
Exio |
is there other bug about this? |
| 03:33 |
VanessaE |
Exio: there's a bug in the collision code too, related but different than this |
| 03:34 |
VanessaE |
hmmmm: did you see his debug output? |
| 03:34 |
VanessaE |
new value of m_speed!!!!!!!!11111!!one: 26659528970266221230541500217032704.000000 |
| 03:34 |
VanessaE |
Exio: I think it's an absolute speed, so a lower-end cap of 0 should be fine. |
| 03:34 |
hmmmm |
yes, i see that in the channel, but i don't know what code is doing that so i can't comment on it |
| 03:35 |
Exio |
hmmmm: localplayer.cpp:325~ (as i modified some lines so not sure :P) |
| 03:36 |
Exio |
VanessaE: i didn't see any <0 value in legits jumps, i think the same |
| 03:36 |
hmmmm |
anyway, i hope you fixed the actual cause of the crash |
| 03:36 |
hmmmm |
because if not, somebody could make an exploit for it, if you're just fixing what causes the crash |
| 03:36 |
hmmmm |
fixing a situation in which the crash happens* |
| 03:37 |
Exio |
yep, the actual cause is just a "raw" + |
| 03:37 |
Exio |
what adds until infinite |
| 03:39 |
VanessaE |
hmmmm: the reason this came up, btw, was my server had another accel+collision crash |
| 03:41 |
Exio |
what should be the commit message? |
| 03:43 |
VanessaE |
"Add cap to jump speed" maybe? |
| 03:43 |
VanessaE |
that's what I'd say anyway |
| 03:43 |
VanessaE |
"Add jump speed cap, stops client crash with certain uses of jumping code." |
| 03:43 |
hmmmm |
hmm |
| 03:44 |
hmmmm |
for the biome calculation i'm going to do something pretty different from what i was originally going to do, by celeron's suggestion |
| 03:45 |
hmmmm |
biome boundaries are calculated as with a voroni diagram from points such as heat/humidity/etc. |
| 03:46 |
Exio |
VanessaE: the only "code" what can cause the crash i fixed is with the bouncy stuff, btw https://github.com/EXio4/minetest/commit/231ec89c1d9a2e095bad7f2f4db1477a5ecdd060 |
| 03:46 |
Exio |
anything to change or i just throw the pull request? |
| 03:46 |
hmmmm |
i didn't really like this because adding one biome messes up all the others, and you might want to control biome frequency |
| 03:46 |
hmmmm |
exio, question for you, what was the actual crash message? |
| 03:47 |
VanessaE |
Exio: looks good to me, short and simple. |
| 03:47 |
VanessaE |
hmmmm: it hangs/locks up the client |
| 03:47 |
Exio |
that |
| 03:47 |
hmmmm |
thiught you said it crashes |
| 03:47 |
hmmmm |
thought |
| 03:47 |
VanessaE |
yeah, sorry. not the right word |
| 03:47 |
VanessaE |
the client locks up solid and has to be killed. |
| 03:47 |
VanessaE |
(to me, that's a form of a crash) |
| 03:47 |
Exio |
and when you reconnect you are at 1334534783243463745734085564580345 |
| 03:48 |
Exio |
or -238324349856085354093845-453465 |
| 03:48 |
VanessaE |
lol yep |
| 03:48 |
Exio |
real numbers |
| 03:49 |
Exio |
hmmmm: it gets "hanged" in a infinite (while) loop at enviroment.cpp:2025~ |
| 03:51 |
Exio |
(when you reconnect you get "hanged" in that infinite while loop, too) |
| 03:52 |
Exio |
a way to see that is adding a && loopcount < 100 or 75 or so for keeping the client with more than 5~10 FPS and a printf or whatever with the "loopcount" variable, values between 1 to 10 'max', suddenly, the 100 |
| 03:53 |
hmmmm |
ah |
| 03:53 |
hmmmm |
with a huge player speed, dtime_max_increment would become very small but not 0 |
| 03:53 |
Exio |
Setting this to 0 (no -=dtime_part) disables an infinite loop |
| 03:53 |
Exio |
when dtime_part is so small that dtime_downcount -= dtime_part |
| 03:53 |
Exio |
does nothing |
| 03:53 |
hmmmm |
yeah i can see that |
| 03:54 |
hmmmm |
hmm how to fix that isn't immediately obvious to me |
| 03:55 |
hmmmm |
i guess a lower bound can be set for dtime increments |
| 03:56 |
Exio |
fix what of that? the 'infinite loo' at that thing? |
| 03:56 |
hmmmm |
yes |
| 03:57 |
Exio |
starting for players a 20000 or more blocks of the world limit... |
| 03:57 |
Exio |
i really think there should be a way to enforce the "be" inside the world_limit (as, ANYTHING will NEVER be outside the map, if yes, it gets back |
| 03:57 |
hmmmm |
you know, i would put a clamp on player_speed too if i were you |
| 03:58 |
Exio |
"clamp"? |
| 03:58 |
hmmmm |
yeah |
| 03:58 |
hmmmm |
like on line 2007 |
| 03:58 |
ShadowNinja |
hmmmm: did you see my pull? |
| 03:58 |
hmmmm |
player_speed = MYMIN(player_speed, 5000) |
| 03:58 |
hmmmm |
shadowninja, i did |
| 03:59 |
Exio |
ah, got it |
| 03:59 |
Exio |
should i add it to my commit too? |
| 04:07 |
hmmmm |
erm |
| 04:07 |
hmmmm |
nah |
| 04:08 |
hmmmm |
prestidigitator seems to be watching us from the channel logs |
| 04:08 |
Exio |
why you say that when i already push -f |
| 04:08 |
hmmmm |
hi, prestidigitator, great job at being creepy :) |
| 04:08 |
Exio |
hahahaha |
| 04:09 |
hmmmm |
i thought he was "done with this political BS" |
| 04:09 |
Exio |
in any case, does this seems ok for a pull request? https://github.com/EXio4/minetest/commit/48cbec2bf30f042e38442d78a4ca31b0bec55682 |
| 04:09 |
hmmmm |
yeah, that's fine |
| 04:09 |
hmmmm |
actually |
| 04:10 |
Exio |
yep? |
| 04:10 |
hmmmm |
you know, it's probably better if the server had hard limits on things like speed too |
| 04:10 |
hmmmm |
but i don't know, can someone really screw up the server if they go too fast? |
| 04:10 |
hmmmm |
from what you said before, it seemed to be only the client having the problems |
| 04:10 |
Exio |
ask VanessaE |
| 04:10 |
Exio |
i only had problems with the server |
| 04:10 |
Exio |
err, client |
| 04:11 |
Exio |
but no idea really |
| 04:14 |
Exio |
can anyone access to coursera.org? |
| 04:14 |
Exio |
or a ping, at least |
| 04:15 |
Exio |
25 packets transmitted, 0 received, 100% packet loss, time 24193ms |
| 04:15 |
VanessaE |
hmmmm: correct - the bug he was addressing is, afaik, strictly client-side |
| 04:15 |
VanessaE |
I've never crashed my server from it anyway |
| 04:17 |
Exio |
now i need help from VanessaE for saying in a decent english what was the bug and the fix |
| 04:17 |
Exio |
because i'm going to write nonsense :P |
| 04:18 |
kaeza |
+continue |
| 04:18 |
kaeza |
:P |
| 04:18 |
VanessaE |
Exio: "Sometimes jumping speed can exceed the limits of XXXXX - cap speed to 0 <= N <= 5000. Fixes client lockup with certain uses of jumping devices such as trampolines." |
| 04:18 |
VanessaE |
Change XXXX to the variable type in question. |
| 04:18 |
VanessaE |
(int or whatever) |
| 04:19 |
VanessaE |
too verbose? :) |
| 04:19 |
Exio |
nah, perfect |
| 04:20 |
Exio |
but it is not only or when jumping, it is when "bouncy jumping" if we can say |
| 04:20 |
VanessaE |
right |
| 04:20 |
VanessaE |
add in "bouncy" then |
| 04:20 |
Exio |
(check the code of the trampolines, bouncy = stuff) |
| 04:23 |
Exio |
well, something |
| 04:24 |
Exio |
here we go, a very cool and nice pull request https://github.com/minetest/minetest/pull/591 |
| 04:27 |
ShadowNinja |
hmm, 52 pull requests, maybe a few more active pushers would be a good idea, kwolekr seems to be doing it all |
| 04:32 |
Exio |
there is some stuff what is delayed because the feature freeze and similar |
| 04:35 |
ShadowNinja |
well yes, but I havn't seen it below 40, he is just barely keeping up |
| 04:38 |
Exio |
https://github.com/minetest/minetest/pull/326 https://github.com/minetest/minetest/pull/575 + luajit should be for the 0.4.7 "rc" |
| 04:40 |
Exio |
and https://github.com/minetest/minetest/pull/585 too |
| 04:40 |
Exio |
about the last, just as "fixing mem leaks" |
| 04:43 |
hmmmm |
i got it to 39 once |
| 04:45 |
ShadowNinja |
well that is still just once, it hasn't been steadily decreasing |
| 04:47 |
Exio |
hmmmm: suddenly, more features :P |
| 04:48 |
ShadowNinja |
heh, everyone aims to get them in before feature-freeze and just misses :-) |
| 04:53 |
Exio |
well, i'm going to sleep now |
| 04:53 |
Exio |
good night |
| 06:16 |
celeron55 |
hmmmm: you should both add and subtract with 3D noise - both are good |
| 06:21 |
celeron55 |
--- |
| 06:21 |
celeron55 |
i think all Exio's fixes are wrong |
| 06:22 |
celeron55 |
the speed should be limited in collisionMoveSimple, not elsewhere |
| 06:22 |
celeron55 |
this way you're doomed to get the bug somewhere else and cluttering other code with copypasta checks that don't belong there |
| 06:23 |
celeron55 |
the problem is the collision and movement code, not the speed that some entity happens to have; the collision and movement code should 100% handle any values given |
| 06:24 |
celeron55 |
and then there will be no more problems, ever |
| 06:30 |
VanessaE |
I don't even see where it's calling that function at all here.. |
| 06:33 |
|
Jordach joined #minetest-dev |
| 06:33 |
|
Jordach joined #minetest-dev |
| 07:14 |
kaeza |
you know what? screw that selection box fix. It would be better to use the node box directly as selection box.otherwise things like technic or mesecons wires willseeno benefit at all |
| 07:14 |
kaeza |
will see no* |
| 07:18 |
VanessaE |
kaeza: what's needed is a way to set the selection box based on the nodeboxes, such that the selection box exactly follows the nodebox shape -- but... |
| 07:18 |
VanessaE |
... with the selection box blown out a bit. |
| 07:18 |
kaeza |
hmm... |
| 07:18 |
VanessaE |
maybe by like 25% or some configurable number. like the "grow selection" function in an image editor |
| 07:19 |
VanessaE |
selection_box = { type = "auto", auto = "25%" } |
| 07:19 |
VanessaE |
or something like that |
| 07:20 |
kaeza |
that wouldrequire C++ changes, no? |
| 07:20 |
kaeza |
I ain't touching that |
| 07:20 |
VanessaE |
it could theoretically be done entirely in Lua |
| 07:20 |
VanessaE |
but it would be a bitch. |
| 07:20 |
kaeza |
not hard to do in Lua, but somebody would complain it's "too hacky" |
| 07:22 |
kaeza |
doing that would be as follows: recognize "auto" in the sel box on register_item(), calc the boxes (based on the "auto" param), and replace the sel box type to "fixed" and the list of boxes |
| 07:23 |
kaeza |
too hacky? |
| 07:23 |
VanessaE |
actually that sounds good. the issue is calculating each of the boxes' coordinates. |
| 07:24 |
VanessaE |
you have to blow each coordinate pair out by twice the given value, then re-center that pair against the original nodebox, then repeat for every coordinate pair that's specified |
| 07:24 |
kaeza |
that shouldn't be hard. it's just calc'ing the center, the width (and height and length) and adding as much as needed to every side |
| 07:24 |
VanessaE |
It could be done in Lua, and it wouldn't be slow at all (not that it matters much at startup). |
| 07:25 |
VanessaE |
*shrug* give it a shot :_0 |
| 07:25 |
VanessaE |
:-) |
| 07:31 |
kaeza |
actually, I don't need the centers :) |
| 07:44 |
VanessaE |
bbl |
| 07:48 |
kaeza |
forget about that. It looks ugly >.< |
| 07:54 |
kaeza |
https://github.com/minetest/minetest/pull/592 |
| 09:26 |
|
proller joined #minetest-dev |
| 09:29 |
|
Topic for #minetest-dev is now **FEATURE-FREEZE FOR 0.4.6 IN EFFECT** | Minetest core development and maintenance. Chit-chat goes to #minetest. Consider this instead of /msg celeron55. http://irc.minetest.ru/minetest-dev/ http://dev.minetest.net/ |
| 09:32 |
|
Calinou joined #minetest-dev |
| 10:36 |
|
PilzAdam joined #minetest-dev |
| 11:05 |
|
proller joined #minetest-dev |
| 11:06 |
|
jojoa1997|Tablet joined #minetest-dev |
| 11:08 |
|
proller joined #minetest-dev |
| 11:17 |
|
jojoa1997_Tablet joined #minetest-dev |
| 11:23 |
|
jojoa1997|Tablet joined #minetest-dev |
| 11:26 |
|
jojoa1997_Tablet joined #minetest-dev |
| 12:04 |
|
darkrose joined #minetest-dev |
| 12:04 |
|
darkrose joined #minetest-dev |
| 12:09 |
|
jojoa1997|Tablet joined #minetest-dev |
| 12:21 |
|
jojoa1997|Tablet joined #minetest-dev |
| 12:22 |
|
jojoa1997|Tablet joined #minetest-dev |
| 12:35 |
|
rubenwardy joined #minetest-dev |
| 12:55 |
|
RealBadAngel joined #minetest-dev |
| 13:16 |
|
iqualfragile joined #minetest-dev |
| 13:36 |
|
IceCraft joined #minetest-dev |
| 13:59 |
Exio |
celeron55: for me, having an entity or anything at THAT speed is not something what should be "ok" |
| 14:01 |
Exio |
there should be a limit, even if it is high, (even if it should be in other place, i don't know "the whole" code of MT about that stuff, i just tried to look around) |
| 14:38 |
|
hmmmm joined #minetest-dev |
| 14:46 |
hmmmm |
well |
| 14:46 |
hmmmm |
i'm not going to reply to the last prestidgititoatr message on the forums because i think he made my point well enough, and i decided that i won't respond more than twice anyway |
| 14:50 |
hmmmm |
i was supposed to go to school today but my alarm clock was set for the wrong time :( meh, it's just one class today |
| 14:51 |
Exio |
/ctcp hmmmm time |
| 14:54 |
hmmmm |
i am going to work on the mapgen until 1pm, then i'm going to cut it short and get school-related things done |
| 14:55 |
Exio |
what time where you live? |
| 14:56 |
hmmmm |
EST zone |
| 14:58 |
Exio |
ah, a bit more of time than here (nearly 12AM) |
| 15:04 |
|
Calinou joined #minetest-dev |
| 15:08 |
|
jojoa1997|Tablet joined #minetest-dev |
| 15:15 |
|
jojoa1997_Tablet joined #minetest-dev |
| 15:17 |
|
jojoa1997|Tablet joined #minetest-dev |
| 15:21 |
|
jojoa1997_Tablet joined #minetest-dev |
| 15:23 |
|
jojoa1997|Tablet joined #minetest-dev |
| 15:28 |
|
jojoa1997_Tablet joined #minetest-dev |
| 15:30 |
|
jojoa1997|Tablet joined #minetest-dev |
| 15:30 |
|
jojoa1997_Tablet joined #minetest-dev |
| 15:31 |
|
jojoa1997|Tablet joined #minetest-dev |
| 15:32 |
|
jojoa1997|Tablet left #minetest-dev |
| 15:42 |
|
jojoa1997|Tablet joined #minetest-dev |
| 15:42 |
celeron55 |
PilzAdam: |
| 15:42 |
celeron55 |
Hello Celeron, |
| 15:42 |
celeron55 |
My name is neron on the minetest forums, and I have been banned by |
| 15:42 |
celeron55 |
user account and IP with the message "getminecraftforfree spam" I |
| 15:42 |
celeron55 |
believe it was due to the post I made to PilzAdam about MiniTest. I |
| 15:42 |
celeron55 |
think it was misunderstood that I was spamming. I was actually hoping |
| 15:42 |
celeron55 |
that PilzAdam might be able to more accurately redesign minecraft by |
| 15:42 |
celeron55 |
using a free copy. I am sorry that it looked liked I was spamming. |
| 15:42 |
celeron55 |
Would you please consider lifting the ban? Thanks, |
| 15:43 |
celeron55 |
neron |
| 15:43 |
celeron55 |
oh god, not this newline crap again |
| 15:43 |
celeron55 |
well, i guess 'lll learn to copy-paste some day |
| 15:44 |
PilzAdam |
oh, I remember his post |
| 15:44 |
PilzAdam |
dunno who banned him |
| 15:45 |
celeron55 |
well, sounds like sfan5 or Calinou; they have a track record of doing silly things |
| 15:49 |
Calinou |
sfan5 banned him, not me |
| 15:49 |
Calinou |
xyz* |
| 15:50 |
|
dexter0 joined #minetest-dev |
| 15:50 |
Calinou |
PilzAdam: should his ban be removed? |
| 15:51 |
PilzAdam |
dunno |
| 15:51 |
PilzAdam |
I guess yes |
| 15:54 |
celeron55 |
throw him a PM then so he knows (i think the forum sends those by e-mail by default) |
| 15:56 |
hmmmm |
hrmm |
| 15:56 |
hmmmm |
http://www.minecraftwiki.net/wiki/Data_values#Biome_IDs |
| 15:56 |
hmmmm |
notice how there's no pattern to the biome ids |
| 15:57 |
hmmmm |
pretty dumb if you ask me... i'm making biomes a bitfield with a value for each of the three attributes |
| 15:58 |
celeron55 |
well, it's just made by something like this http://www.minecraftwiki.net/wiki/File:BiomesGraph.png |
| 15:59 |
hmmmm |
if you do add terrain variance to that, it doesn't correlate at all |
| 16:00 |
celeron55 |
i don't know how they do that |
| 16:00 |
celeron55 |
i guess it's been reworked a lot since the early biomes |
| 16:01 |
hmmmm |
anyway i need to go with your voroni diagram idea.. i realized that I can get around the problem of a biome occuring more often than you'd like by being able to place points that are compared against outside of the range of acceptable values |
| 16:03 |
hmmmm |
i feel like i should also include some kind of utility for people to see the voroni diagram with their biome setup, so they can visualize how common a biome really is |
| 16:04 |
hmmmm |
if we're going to include minetestmapper in the core, it'd make sense to do this as well |
| 16:07 |
celeron55 |
points could have weights to them though too |
| 16:07 |
celeron55 |
it doesn't complicate it much i think |
| 16:08 |
hmmmm |
it does make it less predictable though, which shouldn't be a problem if people can actually see the distribution of their biomes |
| 16:08 |
celeron55 |
how are you going to visualize it, as the map is going to be 3d? |
| 16:09 |
celeron55 |
(temprature, humidity, and height or something related to it) |
| 16:09 |
hmmmm |
that's the thing, i don't know if it's going to be 3d or not |
| 16:09 |
celeron55 |
i think it kind of has to be |
| 16:09 |
hmmmm |
height itself is a more unrelated factor |
| 16:09 |
celeron55 |
or, well... |
| 16:09 |
hmmmm |
it just won't work at all for hell/sky biomes |
| 16:10 |
hmmmm |
height variance on the other hand, i see that as something that could be an attribute on top of temperature/humidity |
| 16:10 |
celeron55 |
if the height is simply taken into account for temperature and humidity, i guess it will work |
| 16:10 |
celeron55 |
like, higher is colder and dryer |
| 16:10 |
hmmmm |
that would make sense, but i'd rather not do that |
| 16:10 |
celeron55 |
what will you do then? |
| 16:11 |
hmmmm |
i don't have everything figured out, i've only been thinking about this for a couple hours |
| 16:11 |
hmmmm |
i've tossed some ideas around passively, but i just added them to the list of ideas and never went to fully consider them until now when i'm actually working on it |
| 16:11 |
celeron55 |
i still think this latest suggestion of mine would work well |
| 16:12 |
hmmmm |
i can try it, but i have a feeling it might make things too boring |
| 16:13 |
celeron55 |
well; maybe - it's the issue of real life being boring |
| 16:13 |
celeron55 |
but it's a trivial solution which has a good chance of working 8) |
| 16:13 |
hmmmm |
what i have now for terrain generation is really great but i have a feeling people will bitch about there being a couple floating blocks |
| 16:14 |
hmmmm |
wonder if i should do a post-generation gravity check |
| 16:14 |
celeron55 |
i suggest not caring about that bitching; there are way more people who love flying blocks than those who hate them |
| 16:14 |
celeron55 |
but only those who are unhappy will say anything |
| 16:14 |
hmmmm |
also, it would be interesting if you were in a desert, and then above you is a single floating blob with snow and ice |
| 16:15 |
celeron55 |
dunno |
| 16:16 |
celeron55 |
if the desert is very dry and hot, not really; of course if it's a chilly-ish desert, then it would make some sense (and would actually work via my suggestion) |
| 16:16 |
hmmmm |
in real life, all deserts are cold at nighttime |
| 16:18 |
|
BlockMen joined #minetest-dev |
| 16:18 |
BlockMen |
hello |
| 16:18 |
BlockMen |
celeron55, have you seen the pull request for homepage? |
| 16:19 |
celeron55 |
yes; i'm not really in the mood of going through it now |
| 16:19 |
hmmmm |
i saw that pull request, i don't really want to be "glorified" like that |
| 16:19 |
celeron55 |
it's a lot of code |
| 16:20 |
|
Zeg9 joined #minetest-dev |
| 16:20 |
BlockMen |
ok, np |
| 16:20 |
BlockMen |
hmmm, what do you mean? |
| 16:20 |
BlockMen |
*hmmm |
| 16:20 |
thexyz |
wut |
| 16:20 |
hmmmm |
all the developers on the front page? :/ |
| 16:20 |
thexyz |
so, now spamming about "free" minecraft is allowed, right? |
| 16:21 |
thexyz |
sorry, didn't know that |
| 16:21 |
BlockMen |
on the community page |
| 16:21 |
hmmmm |
please tell me that i misread it and that was the about page, or better, a contributor's page |
| 16:21 |
BlockMen |
so a sub-page |
| 16:21 |
celeron55 |
thexyz: was it his first post? |
| 16:21 |
hmmmm |
alright, better i guess |
| 16:22 |
thexyz |
celeron55: no, i guess |
| 16:22 |
celeron55 |
if not, then banning is kind of too much; if yes, then whatever; we have too many spammers to be any more thorough in decisions |
| 16:23 |
celeron55 |
if somebody new but one who has already posted legit stuff is misbehaving on the forum, such should be warned about first; ban comes second |
| 16:24 |
thexyz |
ok |
| 16:24 |
thexyz |
i just got really tired by those free minecraft ads at our forum |
| 16:26 |
|
BlockMen_ joined #minetest-dev |
| 16:26 |
BlockMen_ |
thexyz, can you add the views to extern.php->rss (at the forum)? |
| 16:28 |
thexyz |
so, what's the point of asking a question and leaving right after? |
| 16:28 |
thexyz |
oh |
| 16:28 |
thexyz |
i'm blind |
| 16:28 |
thexyz |
BlockMen_: well, send me a patch |
| 16:29 |
thexyz |
BlockMen_: i'm not going to mess with it |
| 16:29 |
BlockMen_ |
thexyz, sry, client crashed |
| 16:29 |
BlockMen_ |
so if i code that you would implement? |
| 16:30 |
thexyz |
sure, just tell me what did you mean by saying "views" |
| 16:33 |
BlockMen_ |
thexyz, i mean the views of a topic (like on viewforum.php) |
| 16:33 |
BlockMen_ |
and great, i will let you know when i coded |
| 16:33 |
thexyz |
oh, so, count of views |
| 16:34 |
BlockMen_ |
yes |
| 17:03 |
hmmmm |
hmm? pilzadam, you seem to be generating a lot of buzz with the whole being-able-to-see-yourself modification |
| 17:03 |
hmmmm |
how did you do it that's so hacky, exactly? |
| 17:04 |
PilzAdam |
I just made the CAO of the local player visible and set the animation and position in its on_step |
| 17:04 |
PilzAdam |
but we cant really know the animation it |
| 17:04 |
PilzAdam |
we even dont know if it uses a mode at all |
| 17:05 |
hmmmm |
do you think you'll polish that up? |
| 17:05 |
hmmmm |
i think it'd be a great addition, and it has much positive feedback |
| 17:05 |
celeron55 |
have you made some kind of a shortcut or is it lagged by the network? |
| 17:06 |
PilzAdam |
celeron55, Its client side |
| 17:06 |
PilzAdam |
I hard coded the animation frames |
| 17:06 |
PilzAdam |
hmmmm, I dont think I can polish it |
| 17:06 |
hmmmm |
i think it's better without any server interaction |
| 17:06 |
PilzAdam |
it would require some restructure of the whole player model thing |
| 17:07 |
hmmmm |
hmm |
| 17:07 |
celeron55 |
hmmmm: the key here is that to make the player animate and move immediately, you have to do it locally - but the player model and it's animations can really be *anything*, as said by the server |
| 17:07 |
PilzAdam |
we should add a minetest.set_player_animation({dig_start=167, dig_end=178;…}) |
| 17:07 |
celeron55 |
hmmmm: so you can just guess |
| 17:07 |
hmmmm |
guessing isn't really that bad |
| 17:07 |
sfan5 |
celeron55: :/ |
| 17:08 |
celeron55 |
i guess there should be a static specification of animations (like what frames or some kind of an animation name (whatever a format supports) is walking, digging and whatever) |
| 17:08 |
hmmmm |
pilzadam, ? digging animations setable from within lua? |
| 17:08 |
* sfan5 |
does not like the "they have a track record of doing silly things" part |
| 17:08 |
|
serengeor joined #minetest-dev |
| 17:08 |
celeron55 |
and that specification should be sent from the server to the client |
| 17:08 |
PilzAdam |
hmmmm, just the frame ranges to set the animations for the default actions |
| 17:08 |
hmmmm |
oh |
| 17:09 |
hmmmm |
yeah i get it now |
| 17:09 |
hmmmm |
now i think that's great, and we send it to the client all in a single packet |
| 17:09 |
celeron55 |
sfan5: it's my experience of things - of course it's not so bad i'd remove the moderator status though |
| 17:09 |
hmmmm |
as long as it can be displayed instantly |
| 17:11 |
hmmmm |
ah |
| 17:11 |
hmmmm |
can't you get the walking and what not from ClientEnvironment? |
| 17:11 |
|
rubenwardy1 joined #minetest-dev |
| 17:12 |
hmmmm |
i don't think what you did is really that hacky, but to do it right, it'd require another packet (or tack the fields onto an existing packet, for that matter...) |
| 17:12 |
PilzAdam |
in the current state it is hacky |
| 17:13 |
PilzAdam |
but it can be made better |
| 17:13 |
celeron55 |
is there something else that is hacky than the hard-coded animations? |
| 17:13 |
hmmmm |
i just looked at the code and honestly it doesn't seem so |
| 17:14 |
celeron55 |
just make them somehow mod-defined (just like the model) then and it's good to go |
| 17:14 |
PilzAdam |
Ill see if I can improve it |
| 17:16 |
|
rubenwardy1 left #minetest-dev |
| 17:17 |
celeron55 |
you may want to consider one thing: if someobdy wants to make a game where the player is a... ehm, bird or something, he should be able to make the client not do any animations on it's own and just wait for the server to tell them |
| 17:17 |
celeron55 |
but feel free not to; it's not really in the scope of today |
| 17:18 |
BlockMen_ |
thexyz, http://pastie.org/private/mlpqayxmfhfep6djzx0ua |
| 17:23 |
hmmmm |
ahhhhhhh, i am regretting two of my design choices right now |
| 17:24 |
hmmmm |
one, not making flat its own mapgen, because now BiomeSuperflat won't work the way I originally intended it to |
| 17:24 |
hmmmm |
two, actually beliving that i'd be able to generate all the terrain I needed at once in a single pass |
| 17:25 |
hmmmm |
(well, i *could*, but you're not able to do many interesting things with this setup) |
| 17:31 |
BlockMen_ |
thexyz, same but highlighted changes now http://pastie.org/private/mlpqayxmfhfep6djzx0ua#137,448,493 |
| 17:35 |
thexyz |
BlockMen_: done |
| 17:35 |
thexyz |
i'd prefer .patch though |
| 17:40 |
BlockMen_ |
thexyz, that was fast |
| 17:40 |
BlockMen_ |
ty :) |
| 18:32 |
PilzAdam |
in wich file should I add minetest.set_player_animation()? |
| 18:38 |
hmmmm |
scriptapi.cpp |
| 18:39 |
hmmmm |
it doesn't really matter, sapier would take care of where it's placed when he updates the scriptapi cleanup to 0.4.6 for merging |
| 18:42 |
hmmmm |
adam, i think we might want to add the trampoline bugfix before 0.4.6, since it's rather critical, doesn't change much, and most people running clients only (windows users) would be using 0.4.6, not the latest git build |
| 18:42 |
PilzAdam |
hm |
| 18:43 |
PilzAdam |
yea, throw it in :-) |
| 18:43 |
VanessaE |
hmmmm: celeron55 said he doesn't like it |
| 18:43 |
hmmmm |
what's wrong... |
| 18:44 |
VanessaE |
[04-02 02:22] <celeron55> i think all Exio's fixes are wrong |
| 18:44 |
VanessaE |
[04-02 02:22] <celeron55> the speed should be limited in collisionMoveSimple, not elsewhere |
| 18:44 |
VanessaE |
[04-02 02:23] <celeron55> this way you're doomed to get the bug somewhere else and cluttering other code with copypasta checks that don't belong there |
| 18:44 |
VanessaE |
[04-02 02:24] <celeron55> the problem is the collision and movement code, not the speed that some entity happens to have; the collision and movement code should 100% handle any values given |
| 18:44 |
VanessaE |
[04-02 02:24] <celeron55> and then there will be no more problems, ever |
| 18:44 |
VanessaE |
(however, I'd have done it the same way exio did) |
| 18:44 |
hmmmm |
i did say something similar, but this is still acceptable |
| 18:45 |
hmmmm |
(then again, i didn't know that collisionMoveSimple had anything to do with the player speed, common sense tells you that it would only deal with position) |
| 18:45 |
VanessaE |
that's what I thought. |
| 18:46 |
hmmmm |
perhaps it checks the collision with the current position + next step in speed |
| 18:46 |
hmmmm |
cba to check |
| 18:47 |
PilzAdam |
particles also use collision_move_simple() |
| 18:48 |
VanessaE |
yeah but this issue is in the bouncy code adding huge acceleration values that blow out even over 64 bit int without capping it. |
| 18:48 |
VanessaE |
not directly related to the server huge RAM+CPU+crashing issue |
| 18:50 |
hmmmm |
hmm, "better to both add and subtract 3d noise"? i'm very satisfied with the terrain currently being generated and i'd rather not screw it up |
| 18:50 |
hmmmm |
i'll experiment later |
| 18:53 |
Exio |
it is an "non controlled" speed problem, there should be a limit somewhere, maybe i just did it in the wrong place |
| 18:53 |
hmmmm |
yes exio, he clearly said that the right place to do it in is collisionMoveSimple |
| 18:54 |
VanessaE |
this fix breaks my suicide-client-crasher booth though :D |
| 18:54 |
Exio |
ah |
| 18:54 |
Exio |
yes, read |
| 18:55 |
|
Leoneof joined #minetest-dev |
| 18:55 |
Leoneof |
hello devs! |
| 18:56 |
Leoneof |
i've downloaded minetest from git, and compiled, but i can't run it, i get these errors http://pastebin.com/CtVqFFbG |
| 18:56 |
Jordach |
Leoneof, is nouveau installed? |
| 18:56 |
hmmmm |
sounds more like a driver problem |
| 18:56 |
Leoneof |
Jordach: no, it's nvidia driver from offical website |
| 18:57 |
Jordach |
Leoneof, i would have probably installed the driver through your normal software channels |
| 18:57 |
VanessaE |
PilzAdam: regarding your partially-visible player patch, I like that idea. |
| 18:59 |
VanessaE |
there's been a pretty big call for making a player's wielded item visible to others. Any chance we can get that feature into the 0.4.7 cycle? |
| 19:02 |
Jordach |
mmmm, yeah |
| 19:02 |
Jordach |
it does make it easier to detect items like lava buckets, potentially finding griefers easier |
| 19:03 |
PilzAdam |
its not easy since the "default" model isnt part of the engine |
| 19:04 |
Jordach |
PilzAdam, minetest.conf as default for builds |
| 19:04 |
PilzAdam |
we could just move the current model into the engine |
| 19:04 |
Jordach |
which makes this less painful |
| 19:04 |
Jordach |
and more of a polished game |
| 19:17 |
RealBadAngel |
Leoneof, googled for your error and found this: https://forum.kag2d.com/threads/crashing-before-game-even-runs-d.1275/ |
| 19:21 |
|
jojoa1997|Tablet joined #minetest-dev |
| 19:34 |
|
proller joined #minetest-dev |
| 19:52 |
|
BlockMen joined #minetest-dev |
| 19:53 |
BlockMen |
thexyz, could you replace 50 with 300 at this position (http://pastie.org/private/mlpqayxmfhfep6djzx0ua#287)? |
| 20:04 |
hmmmm |
hmmm... I am changing the BiomeDefManager to be entirely contained within the EmergeManager |
| 20:05 |
hmmmm |
even though it does have some non-mapgen-related things to (in the future) it's just better not directly in Server |
| 20:05 |
jojoa1997|Tablet |
hmmmm would it be possible to make the biome definer use lua to define the params but C++ to do the actual work |
| 20:05 |
Exio |
he is working in that stuff |
| 20:05 |
hmmmm |
jojoa, that's the entire point of the new mapgen... |
| 20:06 |
Exio |
i'll look at the speed stuff and limit the speed in the collisionMoveSimple |
| 20:06 |
hmmmm |
ok |
| 20:06 |
jojoa1997|Tablet |
oh ok |
| 20:06 |
jojoa1997|Tablet |
i cant wait to make a snow desert |
| 20:07 |
jojoa1997|Tablet |
snow sould be added to minetest for biomes |
| 20:07 |
hmmmm |
a lot more textures than that will have to be added |
| 20:07 |
jojoa1997|Tablet |
also is there any chance that new trees will be added |
| 20:08 |
hmmmm |
there is a chance |
| 20:08 |
hmmmm |
it'll be what you want, just trust me |
| 20:08 |
jojoa1997|Tablet |
i dont care as long as they are not as big as moretrees |
| 20:08 |
jojoa1997|Tablet |
and as numerous |
| 20:09 |
hmmmm |
they're going to be moretrees |
| 20:09 |
hmmmm |
but if you don't like that, you can make your own trees and use those instead |
| 20:09 |
jojoa1997|Tablet |
that is a heavy mod why do you want that |
| 20:10 |
hmmmm |
because that was part of the plan |
| 20:10 |
jojoa1997|Tablet |
it creates tons of lag why not make regular tree sized ones |
| 20:10 |
hmmmm |
the size of the tree isn't what causes the lag, it's the manner in which it's done |
| 20:11 |
jojoa1997|Tablet |
ok but one thing is maybe make them a lot rarer cause on my world there is no room to build |
| 20:11 |
hmmmm |
lol |
| 20:11 |
hmmmm |
i don't have anything do even do with that part, honestly |
| 20:11 |
jojoa1997|Tablet |
i know but edit the version of more trees to |
| 20:12 |
jojoa1997|Tablet |
have trees per biome and not be as clumped unless in jungle? |
| 20:12 |
hmmmm |
moretrees isn't being added as a mod in the default game |
| 20:12 |
hmmmm |
you misunderstand |
| 20:12 |
hmmmm |
i'd rather not explain it |
| 20:12 |
hmmmm |
you'll see |
| 20:12 |
jojoa1997|Tablet |
oh |
| 20:12 |
jojoa1997|Tablet |
so you are making a mod or not? |
| 20:13 |
PilzAdam |
lol |
| 20:13 |
PilzAdam |
mapgen v7 as a mod :-) |
| 20:13 |
hmmmm |
hahahah |
| 20:13 |
PilzAdam |
can you get it to run on 0.3.1? |
| 20:14 |
jojoa1997|Tablet |
oh it is the v7 mapgen |
| 20:15 |
hmmmm |
"mapgen v7" is really just an umbrella term for a completely new terrain generator that makes much more interesting landforms, biomes, a way to efficiently place arbitrary figures onto the map, and all cool game content that you can make happen with these abilities |
| 20:35 |
RealBadAngel |
hmmmm, for those who says trees are to big recommend those: http://forum.minetest.net/viewtopic.php?id=5387 |
| 20:35 |
hmmmm |
bonzai trees 8) |
| 20:36 |
RealBadAngel |
good for dwarves biome ;) |
| 20:38 |
Jordach |
or, we could use that for a sapline |
| 20:38 |
Jordach |
sapling* |
| 20:39 |
RealBadAngel |
thats actually nice idea with multinodeboxes together |
| 20:40 |
RealBadAngel |
divide one node into 16x16x16 little cubes (maybe more) and make mini lsystem trees for saplings |
| 20:41 |
RealBadAngel |
hell its genial for little plants! |
| 20:42 |
RealBadAngel |
all flowers, grass, saplings whatever |
| 20:44 |
RealBadAngel |
makin smaller cubes one could generate ultra realistic plants |
| 20:44 |
RealBadAngel |
in mods, not the core |
| 20:45 |
* ShadowNinja |
would love that as the sapling in default |
| 20:46 |
hmmmm |
except saplings don't look like that at all.. |
| 20:46 |
hmmmm |
bonzai trees, sure.. but.... |
| 20:46 |
RealBadAngel |
ofc, they would need another axiomes |
| 20:47 |
RealBadAngel |
i really love the idea |
| 20:47 |
ShadowNinja |
well better than the "plantlike" drawtype |
| 20:48 |
RealBadAngel |
hell better |
| 20:48 |
RealBadAngel |
most funny, to get HD texture pack modder will have just to redefine the cuboid |
| 20:49 |
RealBadAngel |
and the code will generate plants in HD instead of default 16x16x16 |
| 20:50 |
ShadowNinja |
that would be realy cool |
| 20:51 |
thexyz |
BlockMen: done |
| 20:52 |
BlockMen |
thexyz, thanks! |
| 20:53 |
RealBadAngel |
so for example to get grass or flower in 128x you redefine default:grass cuboid_resolution=128 and done :) |
| 20:54 |
RealBadAngel |
in fact in this textures will be just colors of the pixels in 3d model |
| 20:55 |
* ShadowNinja |
wonders how you will apply textures properly to a random model |
| 20:56 |
RealBadAngel |
same as for trees |
| 20:56 |
RealBadAngel |
but at this level texture will be color of the 3d pixel |
| 20:58 |
Jordach |
does anyone remember this type of mapgen: http://www.zimagez.com/zimage/screenshot-020413-215803.php |
| 21:01 |
ShadowNinja |
Jordach: no, it usualy doesn't generate cloud... |
| 21:02 |
Jordach |
ShadowNinja, that's taken from when minetest's mapper was a little more crazy |
| 21:02 |
Jordach |
0.3.x days -> early 0.4 days |
| 21:03 |
ShadowNinja |
yea, I have seen flying chunks of rock and dirt :-) |
| 21:04 |
ShadowNinja |
but I wish we had stuff like that, everything is a plain with a clif here and there, we should have mountains |
| 21:04 |
Jordach |
ShadowNinja, very early versions of 0.4 generate those types of map |
| 21:04 |
Jordach |
and everything below 0.3.x generates it too |
| 21:06 |
hmmmm |
:( |
| 21:06 |
hmmmm |
the order in which things happen is such a pain in the butt |
| 21:06 |
hmmmm |
i wish that the environment was created before the scripts run |
| 21:34 |
BlockMen |
gd n8t everyone |
| 21:35 |
|
BlockMen left #minetest-dev |
| 23:12 |
|
jojoa1997|Tablet joined #minetest-dev |
| 23:18 |
|
rarkenin joined #minetest-dev |
| 23:20 |
Exio |
well |
| 23:20 |
Exio |
finally, stopped playing redeclipse and stuff, i'll check about the stuff |
| 23:27 |
Exio |
well |
| 23:29 |
Exio |
is this https://github.com/minetest/minetest/pull/591 right, then? |
| 23:30 |
Exio |
just tried and didn't get any hang |
| 23:30 |
Exio |
(aka, "works") |
| 23:42 |
|
jojoa1997|Tablet joined #minetest-dev |
