| Time |
Nick |
Message |
| 01:15 |
|
leat joined #minetest-dev |
| 01:27 |
|
johnnyjoy joined #minetest-dev |
| 01:38 |
johnnyjoy |
Are there any plans to make the screen size of the client visible in lua on that server? |
| 01:39 |
est31 |
johnnyjoy, what would be the usecase? |
| 01:40 |
est31 |
Tesseract, its ok for me if hmmmm agrees. |
| 01:40 |
johnnyjoy |
Same user case as a browser. The UI would be able to adjust to the screen better. |
| 01:40 |
est31 |
pixel number doesnt help |
| 01:40 |
est31 |
pixels can be different size |
| 01:40 |
est31 |
look at retina screens |
| 01:41 |
johnnyjoy |
That makes sense. |
| 01:41 |
johnnyjoy |
Is there any short cut for full screening something like inventory? |
| 01:41 |
est31 |
no, formspecs aren't fullscreen |
| 01:41 |
johnnyjoy |
I did not think so. |
| 01:41 |
johnnyjoy |
Thanks. |
| 01:42 |
est31 |
they take as much as they need, if screen isn't enough, there is a problem |
| 01:43 |
est31 |
johnnyjoy, the pr is still not ok with git |
| 01:44 |
johnnyjoy |
Wouldn't it be easier for me to delete my repo, and start with a new fork? |
| 01:44 |
est31 |
johnnyjoy, it seems you have two "same" commits differing in a detail, and then you made a merge commit |
| 01:44 |
johnnyjoy |
I really am not comfortable with using up your time. |
| 01:44 |
est31 |
no problem, now i have time |
| 01:45 |
johnnyjoy |
What can I do about the 2 commits? |
| 01:45 |
est31 |
so first lets find out what the commits differ |
| 01:45 |
est31 |
there is the `git diff` tool |
| 01:46 |
est31 |
run git diff 05e51c5 f8def41 |
| 01:46 |
johnnyjoy |
It returned nothing. |
| 01:46 |
est31 |
thats good |
| 01:46 |
johnnyjoy |
However with that argument, it did. |
| 01:47 |
johnnyjoy |
Small change to 2 files. |
| 01:48 |
johnnyjoy |
Changes to builtin/game/chatcommands.lua & CMakeLists.txt |
| 01:57 |
|
est31 joined #minetest-dev |
| 01:57 |
est31 |
johnnyjoy, back |
| 01:58 |
johnnyjoy |
Glad to see you that you are. |
| 01:58 |
est31 |
seems everything in your rebase worked fine except the final step |
| 01:59 |
est31 |
putting the "branch pointer" onto it |
| 01:59 |
est31 |
one of the two commits is "good" the other one "bad" |
| 01:59 |
est31 |
the "bad" one has changes we dont want to include |
| 02:00 |
johnnyjoy |
Okay. |
| 02:00 |
est31 |
seems the good one is 05e51c5507 |
| 02:00 |
est31 |
just checkout the branch of your PR, then git reset --hard the commit name |
| 02:01 |
est31 |
then push -f |
| 02:01 |
johnnyjoy |
Just to be straight, git reset --hard 05e51c5507 |
| 02:01 |
est31 |
yes |
| 02:03 |
johnnyjoy |
It's done. |
| 02:04 |
est31 |
nice |
| 02:04 |
est31 |
now one thing, the commit's author is "root" |
| 02:04 |
johnnyjoy |
That is a positive word. I like where this is going. |
| 02:05 |
est31 |
you should perhaps change the author to you again. |
| 02:05 |
johnnyjoy |
Will do. |
| 02:06 |
|
leat joined #minetest-dev |
| 02:06 |
johnnyjoy |
git commit --amend --author="Author Name <email address.com>" Would something like that work? |
| 02:07 |
est31 |
yes :) |
| 02:07 |
johnnyjoy |
It's done. |
| 02:41 |
hmmmm |
[05:05 PM] <paramat> hi hmmmm another one for your consideration sometime #2935 |
| 02:41 |
ShadowBot |
https://github.com/minetest/minetest/issues/2935 -- Mgv7: Use density noise + density gradient for mountain terrain by paramat |
| 02:41 |
hmmmm |
isn't that sort of like making an entirely new mapgen at this point? |
| 02:41 |
hmmmm |
i mean i realize it's as simple as adding a line and changing a multiplication to division, but it completely changes the shape of everything |
| 02:42 |
hmmmm |
there are a surprising number of users who like mapgen v7 as it exists right now... i wouldn't do anything too radical with it if i were you |
| 02:42 |
hmmmm |
[09:40 PM] <est31> Tesseract, its ok for me if hmmmm agrees. <--- the threading cleanup? looking now |
| 02:42 |
hmmmm |
[05:34 PM] <Sokomine> i get crashes where gdb reports in Mapgen::propagateSunlight(irr::core::vector3d<short>, irr::core::vector3d<short>) () when calling vm:calc_lighting(tmin, tmax) sometimes. am i using it wrong? |
| 02:43 |
hmmmm |
sounds like paramat should take a look at this one since he's the last one to modify LuaVoxelManip::l_calc_light ;-) |
| 02:44 |
hmmmm |
it's likely that sokomine is providing invalid input, but no lua api should ever crash with invalid input |
| 02:50 |
hmmmm |
hmmm, https://github.com/minetest/minetest/pull/2613/files#diff-8f01816ac62a60ae302d07174b96e5e8R28 why move the standard lib headers to the bottom? is there a reason in particular? we never really decided on a formal header include ordering, but I put global includes at the top and then local includes |
| 02:50 |
est31 |
hmmmm, all 3 of the cleanups |
| 02:50 |
est31 |
network, threads, logging |
| 02:50 |
hmmmm |
oh wow I forgot about the network cleanup |
| 02:51 |
est31 |
the mod profiler cleanup is very small, but depends on logging |
| 02:53 |
hmmmm |
what's up with these |
| 02:53 |
hmmmm |
#define ATOMIC_ADD_EQ(T, v, x) return __sync_add_and_fetch(&(v), x); |
| 02:53 |
hmmmm |
why the return statement?? |
| 02:53 |
hmmmm |
in fact, all of those macros look like they need work |
| 02:54 |
est31 |
btw, is style class ClassName { |
| 02:54 |
hmmmm |
yea |
| 02:54 |
est31 |
or is style class ClassName |
| 02:54 |
est31 |
{ |
| 02:54 |
hmmmm |
the first |
| 02:54 |
est31 |
ok |
| 02:55 |
hmmmm |
the bracket going on the next line is only for functions |
| 02:56 |
|
johnnyjoy joined #minetest-dev |
| 03:04 |
johnnyjoy |
est31, please let me know if there is anything else I can do. |
| 03:10 |
|
est31 joined #minetest-dev |
| 03:10 |
|
Hunterz joined #minetest-dev |
| 03:32 |
hmmmm |
alright I added some comments on the threading cleanup |
| 03:33 |
hmmmm |
est31: merge https://github.com/minetest/minetest/pull/2828 ? |
| 03:34 |
hmmmm |
jenkins-ci/android-nrz is kinda flakey lately |
| 03:34 |
hmmmm |
we can ignore that failure |
| 03:36 |
est31 |
hmmmm, its because sf.net is down |
| 03:36 |
hmmmm |
still!? |
| 03:37 |
est31 |
http://sourceforge.net/blog/sourceforge-infrastructure-and-service-restoration/ |
| 03:37 |
est31 |
"We’re holding SCM service restoration for last, and will be prioritizing Git service to be first within that process based on its fast verification path. Holding SCM restoration for last allows us to take a cautious approach and to free our staff to interact with developers if any concerns exist when the service is re-enabled." |
| 03:37 |
est31 |
still. |
| 03:37 |
|
Hunterz joined #minetest-dev |
| 03:39 |
hmmmm |
oh man there's so much stuff to review |
| 03:39 |
hmmmm |
it could be a full time job |
| 03:58 |
est31 |
why do we have InvalidPositionException? |
| 03:58 |
hmmmm |
in reference to what? |
| 03:58 |
est31 |
most of the times we do a catch(InvalidPositionException &e){} anyways |
| 03:58 |
est31 |
its just stupid to bother the exception handler |
| 03:59 |
est31 |
its not zero cost when it has to handle exceptions |
| 04:00 |
est31 |
but shrug not that important |
| 04:11 |
Sokomine |
hmmmm: i provided the edges of the area for which light was to be calculated - and those included the shell. without parameters (and without the shell), it doesn't crash, but lighting is not updated correctly |
| 04:16 |
hmmmm |
celeron did go a bit exception-crazy at first |
| 04:16 |
hmmmm |
it wasn't until much later it was determined that exceptions are super freaking slow and also hard to deal with effectively |
| 04:17 |
hmmmm |
minetest started out as this coding exercise where C++ flavor-of-the-month "best practices" were used everywhere |
| 04:18 |
est31 |
well its still around, thanks to the fact that not the "language-of-the-month" was used |
| 04:43 |
hmmmm |
honestly, i've found that the best C++ code eschews all features but the STL and classes |
| 04:43 |
hmmmm |
there is a lot of crap that's more troublesome than it's worth |
| 04:44 |
hmmmm |
in general the rule is to don't listen to what ##C++ tells you to do |
| 04:52 |
|
jin_xi joined #minetest-dev |
| 05:06 |
|
cheapie joined #minetest-dev |
| 05:17 |
|
leat joined #minetest-dev |
| 05:26 |
|
leat joined #minetest-dev |
| 05:37 |
|
leat joined #minetest-dev |
| 05:47 |
|
Hunterz joined #minetest-dev |
| 05:57 |
|
VanessaE joined #minetest-dev |
| 06:04 |
|
nore joined #minetest-dev |
| 06:05 |
Hunterz |
wow too many commits and PRs last weekend... nice |
| 06:07 |
|
Krock joined #minetest-dev |
| 06:15 |
|
est31 joined #minetest-dev |
| 06:25 |
est31 |
Tesseract, time to review my areastore pr? |
| 06:30 |
|
freemint_ joined #minetest-dev |
| 07:00 |
VanessaE |
paramat: are you aware that default acacia wood isn't being registered with the stairs mod? |
| 07:08 |
freemint_ |
In lua_api.txt i read that minetest can nnot play steroe sounds positionally is this limitation still in place? |
| 07:10 |
VanessaE |
it is still the case. |
| 07:10 |
VanessaE |
limitation of OpenAL I think. |
| 07:11 |
freemint_ |
ok, |
| 07:11 |
freemint_ |
thx |
| 07:20 |
|
nrzkt joined #minetest-dev |
| 07:40 |
est31 |
hmmmm, can you remember discussions about interleaving coordinates for mapblocks? |
| 07:40 |
est31 |
aka using z order curves |
| 07:41 |
hmmmm |
not really |
| 07:42 |
hmmmm |
is this for building your areastore more efficiently? |
| 07:42 |
est31 |
no, for loading mapblocks |
| 07:42 |
est31 |
if we store them in interleaved manner, the database perhaps aligns them on the disk that way too |
| 07:43 |
est31 |
I dont know whether we have that guarantee |
| 07:43 |
est31 |
if we have however, z order curves are a fine thing |
| 07:43 |
est31 |
esp. on hard disks which have seeking time |
| 07:43 |
hmmmm |
no we definitely do not have that guarantee |
| 07:44 |
hmmmm |
and Z order curves aren't being used right now |
| 07:44 |
est31 |
yes |
| 07:44 |
hmmmm |
some insane algorithm that doesn't quite make much sense but somehow guarantees uniqueness is used instead |
| 07:44 |
hmmmm |
if it were up to me i'd keep things extremely simple and it would just be the concatenation of each 10 bit value |
| 07:44 |
hmmmm |
rather, 12 bit |
| 07:45 |
est31 |
z order curves are concatenation too but per bit |
| 07:45 |
hmmmm |
right, so X Y and Z are interleaved |
| 07:45 |
est31 |
and closeness in the map means also closeness in ids |
| 07:46 |
hmmmm |
I see how it works yeah |
| 07:46 |
hmmmm |
I'm just wondering whether there would be any tangible benefit from that |
| 07:47 |
est31 |
what do databases align the blocks after |
| 07:47 |
hmmmm |
whatever they feel like doing.. |
| 07:49 |
|
OldCoder joined #minetest-dev |
| 07:53 |
est31 |
<hmmmm> if it were up to me i'd keep things extremely simple and it would just be the concatenation of each 10 bit value |
| 07:53 |
est31 |
isnt it that right now? |
| 07:53 |
est31 |
https://github.com/minetest/minetest/blob/990a96578f20244626b6b9f67f8e79a7e2e614ea/src/database.cpp#L51 |
| 07:53 |
hmmmm |
it would seem like that, wouldn't it |
| 07:54 |
hmmmm |
consider negative components |
| 07:55 |
est31 |
oh |
| 07:56 |
est31 |
err |
| 07:56 |
est31 |
no |
| 07:56 |
est31 |
its casted |
| 07:56 |
hmmmm |
exactly |
| 07:56 |
est31 |
doesnt that mean "take raw memory value" |
| 07:56 |
hmmmm |
well not quite |
| 07:56 |
hmmmm |
for x86/64, yes |
| 07:57 |
hmmmm |
so consider what (u64)-1 looks like in memory |
| 08:01 |
|
Yepoleb_ joined #minetest-dev |
| 08:10 |
|
julienrat joined #minetest-dev |
| 08:17 |
|
Calinou joined #minetest-dev |
| 08:28 |
|
leat joined #minetest-dev |
| 08:28 |
Calinou |
est31, can I make the website competition topic a sticky? |
| 08:28 |
est31 |
yes please |
| 08:28 |
est31 |
thanks |
| 08:29 |
Calinou |
done |
| 08:31 |
|
julienrat joined #minetest-dev |
| 08:35 |
|
julienrat left #minetest-dev |
| 08:51 |
|
Darcidride joined #minetest-dev |
| 08:58 |
|
leat joined #minetest-dev |
| 09:46 |
|
celeron55 joined #minetest-dev |
| 09:56 |
|
julienrat joined #minetest-dev |
| 09:57 |
|
julienrat left #minetest-dev |
| 10:11 |
|
celeron55 joined #minetest-dev |
| 10:23 |
|
celeron55 joined #minetest-dev |
| 10:38 |
|
rubenwardy joined #minetest-dev |
| 10:45 |
rubenwardy |
http://irc.minetest.ru/minetest-dev/2015-07-20#i_4331860 |
| 10:45 |
rubenwardy |
Isn't it the second? The logic for braces on new line for functions is that you can't next functions |
| 10:45 |
rubenwardy |
oh wait, you can nest classes |
| 10:45 |
rubenwardy |
nevermind |
| 10:59 |
|
leat joined #minetest-dev |
| 11:11 |
|
rubenwardy joined #minetest-dev |
| 11:19 |
|
blaise joined #minetest-dev |
| 11:22 |
|
Darcidride joined #minetest-dev |
| 11:23 |
|
celeron55 joined #minetest-dev |
| 11:28 |
|
Player_2 joined #minetest-dev |
| 11:53 |
|
celeron55 joined #minetest-dev |
| 11:54 |
|
kilbith joined #minetest-dev |
| 11:59 |
|
H-H-H joined #minetest-dev |
| 12:30 |
|
leat joined #minetest-dev |
| 12:42 |
|
johnnyjoy joined #minetest-dev |
| 12:44 |
|
kilbith joined #minetest-dev |
| 13:03 |
|
Routhinator joined #minetest-dev |
| 13:03 |
|
leat joined #minetest-dev |
| 13:04 |
|
Zeno` joined #minetest-dev |
| 13:05 |
Routhinator |
An odd question as I cannot find anything on this. Is there a maximum number of Biomes allowed? Ive been trying to figure out why a few are not generating and trying to rule out "too many biomes" as a possibility. |
| 13:06 |
|
kilbith left #minetest-dev |
| 13:11 |
Zeno` |
I don't think there should be |
| 13:12 |
Zeno` |
(well, there would be a limit -- everything has a limit -- but if you're encountering it something would be wrong) |
| 13:13 |
|
Routhinator joined #minetest-dev |
| 13:13 |
freemint_ |
The Biomes are stored inside a LUA table as long you do not have more biomes then a lua numer (something like 10^42) you must only fear that your machine has not enough ram |
| 13:14 |
Routhinator |
OK so maybe it's my humidity and temp settings. Will have to review them tonight. |
| 13:15 |
freemint_ |
maximum number on a 64 bit 5.6* 10^300 |
| 13:16 |
freemint_ |
humidity and temp are depircated why do you use them? |
| 13:18 |
Routhinator |
They are? I was basing my biomes off paramats biomesdev. It's still used in his. If this settings are deprecating then how do you define the humidity and temp of a biome? |
| 13:19 |
freemint_ |
this feature was dropped |
| 13:20 |
freemint_ |
The weather functions were dropped at Apr 18, 2014. -wiki |
| 13:20 |
freemint_ |
http://dev.minetest.net/minetest.get_humidity http://dev.minetest.net/minetest.get_heat |
| 13:21 |
freemint_ |
:-) sorry |
| 13:21 |
|
leat joined #minetest-dev |
| 13:22 |
kahrl |
IANAMD but I'm pretty sure biome generation is still based on two noise maps called "heat" and "humidity" |
| 13:23 |
kahrl |
they are/were unrelated to the ex-API functions get_heat and get_humidity |
| 13:23 |
freemint_ |
Maybe it is, yes, but these are not exposed to the modding api |
| 13:25 |
kahrl |
not directly, but heat_point and humidity_point in a biome def affect it |
| 13:27 |
freemint_ |
ok you are rigth, look in the latest chagelog : |
| 13:27 |
freemint_ |
Made biome heat and humidity noise parameters user-configurable (kwolekr) |
| 13:27 |
freemint_ |
Maybe this helps. |
| 13:34 |
|
H-H-H joined #minetest-dev |
| 13:45 |
|
est31 joined #minetest-dev |
| 13:45 |
est31 |
Zeno`, you really think #2913 is caused by the pop_front function? |
| 13:45 |
ShadowBot |
https://github.com/minetest/minetest/issues/2913 -- Unexplained, random crashes (segfaults, aborts, OOM) |
| 13:46 |
est31 |
Yes it appears in all gdb output, but not inside the crashing thread's stack |
| 13:46 |
est31 |
hrmm perhaps its zeno humor |
| 13:47 |
Zeno` |
est31, I haven't traced it myself |
| 13:47 |
Zeno` |
it seems a good place to start though |
| 13:48 |
Zeno` |
for several reasons... |
| 13:48 |
Zeno` |
that Mutexed thing was reimplemented not all *that* long ago |
| 13:48 |
nrzkt |
hi Zeno` ! |
| 13:48 |
Zeno` |
howdy :) |
| 13:49 |
Zeno` |
nrzkt, you reimplemented it :P |
| 13:49 |
est31 |
from how I see it, the function is just "lets wait for something to happen on the producer queue" |
| 13:49 |
Zeno` |
is pop_front properly protected and checks are in place for empty queue? |
| 13:50 |
est31 |
my theory its blocking |
| 13:50 |
Zeno` |
est31, possibly |
| 13:50 |
est31 |
at least there is sem_timedwait() call from libpthread.so higher in the stack |
| 13:51 |
Zeno` |
Why are these hard ones always so hard to reproduce? heh |
| 13:51 |
Zeno` |
higher in the stack doesn't mean much |
| 13:51 |
est31 |
higher, lower, depends on how you look at it |
| 13:51 |
Zeno` |
the function higher in the stack may have been passed an invalid pointer |
| 13:52 |
est31 |
you are in australia, everything thats up for me is down for you :) |
| 13:52 |
Zeno` |
lmao :) nice :P |
| 13:52 |
Zeno` |
est31, have you reproduced that bug? |
| 13:52 |
est31 |
nope |
| 13:53 |
Zeno` |
damn :( |
| 13:54 |
est31 |
most of the crashes are e.g. inside constructors |
| 13:55 |
est31 |
sign for that something terrible has happened to the heap |
| 13:55 |
est31 |
perhaps the line before, perhaps far away |
| 13:55 |
Zeno` |
well my main reasons for suspecting that are: a) it was changed since the last release; b) it seems like a race condition (which point to some kind of mutex fault) |
| 13:56 |
Zeno` |
It's hard to say though because I haven't been able to reproduce. It's just a guess |
| 13:56 |
nrzkt |
Zeno`, in reimplement what ? |
| 13:57 |
Zeno` |
the mutexed queue |
| 13:57 |
nrzkt |
oO |
| 13:57 |
nrzkt |
where ? |
| 13:57 |
Zeno` |
#2913 |
| 13:57 |
est31 |
I think container.g |
| 13:57 |
ShadowBot |
https://github.com/minetest/minetest/issues/2913 -- Unexplained, random crashes (segfaults, aborts, OOM) |
| 13:57 |
est31 |
.h* |
| 13:58 |
Zeno` |
hmm |
| 13:58 |
* Zeno` |
looks at the dumps again |
| 13:58 |
nrzkt |
i only switch from std::list to std::queue into MutexedQueue to have a faster container |
| 13:59 |
nrzkt |
you said taht std::queue is thread_safe ? |
| 13:59 |
est31 |
no, I dont think we have guarantee for that |
| 13:59 |
est31 |
it has to be mutexed I think |
| 13:59 |
Zeno` |
I said it was thread safe? |
| 14:00 |
nrzkt |
i may have misunderstood what you said |
| 14:00 |
Zeno` |
possibly, because I don't believe it would be at all |
| 14:00 |
nrzkt |
std::queue is not threadsafe |
| 14:00 |
Zeno` |
(language barrier possibly) |
| 14:01 |
Zeno` |
nrzkt, I am not pointing the finger at you btw. I don't *know* if that's even the cause. It's just a guess |
| 14:02 |
Zeno` |
and even if it was part of that change then I don't care as long as it's fixed :) |
| 14:02 |
|
kilbith joined #minetest-dev |
| 14:02 |
Zeno` |
It could be anything. Those traces are not 100% useful because they are from release builds :( |
| 14:02 |
kilbith |
est31: you're aware of that simpler implementation of PostGreSQL by SN ? https://forum.minetest.net/viewtopic.php?p=184869#p184869 |
| 14:03 |
Zeno` |
and I'd rather catch the fault and backtrace manually, but I can't reproduce |
| 14:03 |
est31 |
Zeno`, VE does have a debug build |
| 14:03 |
est31 |
kilbith, now i am, thanks |
| 14:04 |
Zeno` |
hmm |
| 14:04 |
nrzkt |
Zeno` no problem :p |
| 14:05 |
Zeno` |
it's a tricky one :( |
| 14:05 |
Zeno` |
you might be right though, est31 |
| 14:06 |
Zeno` |
wait time seems a bit high |
| 14:06 |
Zeno` |
but... nah |
| 14:06 |
Zeno` |
dammit! I will have to pay for a VPS again |
| 14:07 |
Zeno` |
seems like the only way to reproduce heh |
| 14:07 |
Zeno` |
hold on |
| 14:07 |
Zeno` |
data = {vmanip = 0x0, seed = 0 |
| 14:07 |
Zeno` |
is vmanip a pointer? |
| 14:08 |
est31 |
lotsa 0s in there |
| 14:08 |
est31 |
vmanip = 0x0, seed = 0, blockpos_min = {X = 0, Y = 0, Z = 0}, blockpos_max = {X = 0, Y = 0, Z = 0}, blockpos_requested = {X = 0, Y = 0, Z = 0} |
| 14:08 |
Zeno` |
what's vmanip though? |
| 14:08 |
Zeno` |
I don't have the src handy |
| 14:09 |
est31 |
I guess its short for "voxel manipulator" |
| 14:09 |
est31 |
so most likely a pointer |
| 14:09 |
Zeno` |
yeah I know that much, but does that function check for NULL ptr? |
| 14:10 |
Zeno` |
or maybe it gets the real manipulator from somewhere else, but then why is it passed as a param? |
| 14:11 |
est31 |
hrmm |
| 14:11 |
est31 |
data is only declared at line 457 |
| 14:11 |
est31 |
in theory its out of scope |
| 14:11 |
est31 |
(as stack variable) |
| 14:11 |
est31 |
"BlockMakeData data;" |
| 14:11 |
Zeno` |
wait... |
| 14:12 |
est31 |
I guess the compiler did some reorderings |
| 14:12 |
Zeno` |
where is BlockMakeData declared? |
| 14:13 |
Zeno` |
depending on the size of an object of that type, whacking it on the stack could certainly cause a segfauly |
| 14:13 |
est31 |
emerge.h |
| 14:14 |
Zeno` |
well it doesn't look all that big :D |
| 14:14 |
Zeno` |
we have another queue though... |
| 14:15 |
est31 |
also VE's server has tons of RAM |
| 14:15 |
Zeno` |
RAM would not matter. Do char array[29402942094]; even if you have the RAM it will segfault |
| 14:15 |
est31 |
why |
| 14:16 |
Zeno` |
because the stack size is set by the OS |
| 14:16 |
est31 |
ok |
| 14:16 |
Zeno` |
it's allocated on the stack and not the heap |
| 14:16 |
Zeno` |
yeah |
| 14:16 |
est31 |
I've thought stack size is unlimited, until stack and heap have a meetup. |
| 14:19 |
Zeno` |
pretty sure it's not... write a quick test case ;) |
| 14:23 |
Zeno` |
although with -O2 is might mitigate it |
| 14:25 |
Zeno` |
same as it mitigates stack overflow for tail end recursion |
| 14:26 |
est31 |
kilbith, for the record I think the older implementation is by johnnyojoy, because his copyright headers date to 2014, and his pr wasnt rebased at the start, but he didnt release it until now. |
| 14:26 |
est31 |
fwiw |
| 14:30 |
|
FR^2 joined #minetest-dev |
| 14:30 |
Zeno` |
http://codepad.org/sggxINAq |
| 14:32 |
Zeno` |
I doubt the array needs to be *that* large |
| 14:37 |
|
luizrpgluiz joined #minetest-dev |
| 14:38 |
|
Hunterz joined #minetest-dev |
| 14:38 |
|
Hunterz left #minetest-dev |
| 14:39 |
|
rom1504_ joined #minetest-dev |
| 14:43 |
|
Sketch2 joined #minetest-dev |
| 14:43 |
|
Icedream joined #minetest-dev |
| 14:43 |
|
Icedream joined #minetest-dev |
| 14:43 |
|
OldCoder joined #minetest-dev |
| 14:43 |
Zeno` |
maybe the queue pre-allocates some memory (but it'd probably do that on the stack so that's a red herring) |
| 14:43 |
Zeno` |
*sigh* |
| 14:44 |
Zeno` |
s/on the stack/on the heap |
| 14:44 |
* Zeno` |
assigns est31 this bug |
| 14:46 |
kilbith |
est31: ok, but SN's one looks cleaner from my standpoint |
| 14:47 |
est31 |
what are the points you see SN did better? |
| 14:50 |
kilbith |
"It automatically sets up the tables and only requires one extra setting (postgresql_connection_info)." |
| 14:51 |
est31 |
yea, thats more appealing |
| 14:51 |
kilbith |
also SN is master in polishment :p |
| 14:51 |
est31 |
yea |
| 14:51 |
|
leat joined #minetest-dev |
| 14:52 |
nrzkt |
if prefer SN implementation too |
| 14:52 |
nrzkt |
it's very clear |
| 14:52 |
nrzkt |
but i hope this use authentication |
| 14:52 |
|
leat joined #minetest-dev |
| 14:53 |
Zeno` |
night all |
| 14:53 |
Zeno` |
I'll be back tomorrow |
| 14:54 |
kilbith |
nite |
| 14:54 |
kilbith |
i guess SN intends to put it in PR (rebased recently) |
| 14:55 |
est31 |
we should make benchmarks, and compare between interleaved, split up like in his pr, or the old "just throw it together" approaches |
| 14:56 |
|
jin_xi joined #minetest-dev |
| 14:56 |
est31 |
I wonder what SN suggests about doing with storing mod data in the database |
| 14:57 |
est31 |
whether to make extra table |
| 14:58 |
|
luizrpgluiz left #minetest-dev |
| 14:59 |
|
Hunterz joined #minetest-dev |
| 15:00 |
TBC_x |
have you tried to compile it with sanitizers? |
| 15:00 |
est31 |
nrzkt, #2931 needs second +1 |
| 15:00 |
ShadowBot |
https://github.com/minetest/minetest/issues/2931 -- Allow random menu images for subgames by sfan5 |
| 15:01 |
est31 |
gtg |
| 15:04 |
nrzkt |
est31: i agree but this is annoying for redis |
| 15:06 |
nrzkt |
est31: validated |
| 15:06 |
nrzkt |
please note i shutdown the jenkins project for building PR for android because irrlicht source download has a problem |
| 15:12 |
|
rubenwardy joined #minetest-dev |
| 15:15 |
H-H-H |
nrzkt according to the tweets it shouldnt be to long before its back up |
| 15:15 |
nrzkt |
H-H-H thanks for the review |
| 15:15 |
nrzkt |
does they add spyware into source with the next release ? :p |
| 15:16 |
H-H-H |
lol nah just some bloatware instead usually :P |
| 15:16 |
* H-H-H |
ducks |
| 15:23 |
TBC_x |
what version of irrlicht is the android built against? |
| 15:25 |
TBC_x |
I got the zip here http://en.osdn.jp/projects/sfnet_irrlicht/downloads/Irrlicht%20SDK/1.8/1.8.1/irrlicht-1.8.1.zip/ |
| 15:25 |
TBC_x |
checksum matches with sf zip |
| 15:29 |
freemint_ |
minetest.get_voxel_manip([p1, p2]) -- "wiki" what does the [ ] mean in this context? Or is this bullshit in the wiki? |
| 15:36 |
nrzkt |
it's the latest if i remember |
| 15:36 |
nrzkt |
look at jenkins logs |
| 15:48 |
jin_xi |
freemint_: i think its interval notation, meaning its inclusive p1 and p2 |
| 15:49 |
freemint_ |
How is the usage? |
| 15:51 |
freemint_ |
jin_xi {p1, ..... , p2} ? |
| 15:59 |
|
leat joined #minetest-dev |
| 16:02 |
|
leat joined #minetest-dev |
| 16:10 |
|
Hunterz joined #minetest-dev |
| 16:13 |
|
leat joined #minetest-dev |
| 16:21 |
|
MinetestForFun joined #minetest-dev |
| 16:25 |
rubenwardy |
updated #2908 |
| 16:25 |
ShadowBot |
https://github.com/minetest/minetest/issues/2908 -- Make map limit configurable by rubenwardy |
| 16:46 |
|
rubenwardy_ joined #minetest-dev |
| 16:46 |
TBC_x |
whoops |
| 16:47 |
TBC_x |
double free corruption in network code |
| 16:47 |
TBC_x |
coredump lost for some reason |
| 16:47 |
TBC_x |
but got libc stack trace |
| 16:48 |
TBC_x |
client |
| 16:55 |
|
eeew joined #minetest-dev |
| 17:00 |
|
CraigyDavi joined #minetest-dev |
| 17:03 |
|
Robert_Zenz joined #minetest-dev |
| 17:30 |
TBC_x |
oh boy |
| 17:35 |
|
hmmmm joined #minetest-dev |
| 18:15 |
|
blaise joined #minetest-dev |
| 18:26 |
|
RealBadAngel joined #minetest-dev |
| 18:31 |
TBC_x |
damn: df -h /home |
| 18:31 |
TBC_x |
Filesystem Size Used Avail Use% Mounted on |
| 18:31 |
TBC_x |
/dev/sda3 274G 260G 385M 100% /home |
| 18:38 |
|
nore joined #minetest-dev |
| 18:51 |
|
freemint_ joined #minetest-dev |
| 18:52 |
|
Jasper_ joined #minetest-dev |
| 18:55 |
|
MinetestForFun_ joined #minetest-dev |
| 19:01 |
|
paramat joined #minetest-dev |
| 19:02 |
paramat |
hi sfan5, please can i merge game#582 ? |
| 19:02 |
ShadowBot |
https://github.com/minetest/minetest_game/issues/582 -- Default: Add biome API tree schematics and enable growing by sapling ABMs by paramat |
| 19:02 |
sfan5 |
looks good |
| 19:10 |
paramat |
thanks |
| 19:11 |
paramat |
also if you have time, comments on game#581 ? |
| 19:11 |
ShadowBot |
https://github.com/minetest/minetest_game/issues/581 -- New stairs and slabs discussion |
| 19:22 |
|
MinetestForFun joined #minetest-dev |
| 19:30 |
paramat |
VanessaE yes i will add acacia to stairs soon, ^ that's the discussion issue |
| 19:33 |
paramat |
Routhinator, maximum is 256 biomes in the biome API |
| 19:34 |
paramat |
heat and humidity have not been dropped |
| 19:34 |
hmmmm |
no, don't tell me somebody wants more than 256 biomes |
| 19:35 |
paramat |
what was removed was dynamic weather, which was a dynamic variation of heat and humidity |
| 19:35 |
hmmmm |
if so that's just dumb |
| 19:35 |
hmmmm |
i will not modify any biome code to accomodate an absurd number of biomes |
| 19:35 |
paramat |
no they don't, they were just checking if they had exceeded a low limit |
| 19:35 |
paramat |
http://irc.minetest.ru/minetest-dev/2015-07-20#i_4332375 |
| 19:35 |
hmmmm |
oh ok |
| 19:35 |
paramat |
heh |
| 19:36 |
paramat |
256 is more than enough |
| 19:38 |
|
leat joined #minetest-dev |
| 19:43 |
|
leat joined #minetest-dev |
| 19:47 |
|
paramat left #minetest-dev |
| 20:06 |
RealBadAngel |
so, any more comments on #2934? |
| 20:06 |
ShadowBot |
https://github.com/minetest/minetest/issues/2934 -- Add wielded (and CAOs) shader by RealBadAngel |
| 20:11 |
hmmmm |
i approve |
| 20:33 |
|
nrzkt joined #minetest-dev |
| 20:39 |
|
stormchaser3000 joined #minetest-dev |
| 20:39 |
nrzkt |
i looked at environment code and i found a very very strange loop in the code, i don't know exactly why this is like this but i think this loop should be removed: https://github.com/minetest/minetest/blob/master/src/environment.cpp#L1174 |
| 20:39 |
nrzkt |
i don't see why we need to loop over active blocks multiples times to apply abm |
| 20:41 |
nrzkt |
somebody can enlighten me about this loop ? |
| 20:42 |
nrzkt |
i removed it from my code into AppleTree server and i didn't notice any problem with many loaded mapblocks with abm (i have many mods, like mobs from tenplus1 |
| 20:44 |
nrzkt |
hmmmm: have you got an idea ? |
| 21:07 |
|
Routh_ joined #minetest-dev |
| 21:07 |
|
Jasper joined #minetest-dev |
| 21:07 |
|
Jasper left #minetest-dev |
| 21:32 |
nrzkt |
if someone can review #2940 before merge it could be good. It's a minor cleanup in the code |
| 21:32 |
ShadowBot |
https://github.com/minetest/minetest/issues/2940 -- Remove profiler.h include where it's not needed and some unreachable code. by nerzhul |
| 21:33 |
nrzkt |
if a dev agree, merge it |
| 21:41 |
|
Megaf_ joined #minetest-dev |
| 21:51 |
|
paramat joined #minetest-dev |
| 21:54 |
paramat |
now pushing game#582 |
| 21:54 |
ShadowBot |
https://github.com/minetest/minetest_game/issues/582 -- Default: Add biome API tree schematics and enable growing by sapling ABMs by paramat |
| 21:57 |
|
stormchaser3000 joined #minetest-dev |
| 22:01 |
Tesseract |
hmmmm: I'm not sure how to fix one of your comments. I have to implement something like this: """(mutex.lock(); T _val = (op); mutex.unlock();, _val)""" (in a macro). It has to be an expression and has to lock and unlock the mutex before returning the value of "op". |
| 22:02 |
Tesseract |
(The return-in-macro comment) |
| 22:03 |
paramat |
push complete |
| 22:03 |
Tesseract |
I can do something with an inline function call, but I'd like something more concise. |
| 22:04 |
|
blaise joined #minetest-dev |
| 22:06 |
nrzkt |
Tesseract can you review #2940 please ? |
| 22:06 |
ShadowBot |
https://github.com/minetest/minetest/issues/2940 -- Remove profiler.h include where it's not needed and some unreachable code. by nerzhul |
| 22:27 |
Routh |
hmmmm: No no ridiculous 256 biomes haha, just wanted to see if 18 was too much. |
| 22:27 |
Routh |
paramat has me sorted now though, I had too many biomes with similar heat/humidity |
| 22:28 |
paramat |
and you can adjust the average size of biomes by adjusting the noise 'spreads' of heat and humidity noises |
| 22:28 |
paramat |
by adding lines into .conf |
| 22:28 |
Routh |
Gotcha, I'll look into that once I get them all working properly. Lots of tweaking left to do. |
| 22:30 |
paramat |
if you want vertically separated biomes to appear vertically aligned they need identical heat/humidity values |
| 22:30 |
Routh |
identical heat/humidity but distinct y_min and y_max? |
| 22:31 |
Routh |
I was wondering, so that means I can have a 'mountain top' biome? |
| 22:32 |
paramat |
yes |
| 22:32 |
paramat |
if separated vertically you can use identical biome points |
| 22:32 |
paramat |
my systems are based on https://upload.wikimedia.org/wikipedia/en/2/29/PrecipitationTempBiomes.jpg |
| 22:33 |
paramat |
plus a glacier biome in extreme cold |
| 22:35 |
Routh |
Yes I reused some of yours directly, hope you don't mind. |
| 22:36 |
paramat |
of course not =) |
| 22:37 |
Routh |
The funny thing is this is the easy part of my plan. The City Ruins biome will be the hardest for my subgame, and thus I'm saving it for last. |
| 22:38 |
paramat |
nice idea, i like generative architecture |
| 22:45 |
|
paramat left #minetest-dev |
| 22:52 |
hmmmm |
Tesseract: Hmm, I see your plight |
| 22:52 |
hmmmm |
this might be a good situation for inline functions |
| 22:53 |
hmmmm |
but then again you need actual text substitution here |
| 22:57 |
hmmmm |
Tesseract: Oh well, it can't be helped |
| 22:59 |
hmmmm |
nrzkt, that is not a loop, it's a silly way to avoid goto |
| 23:00 |
hmmmm |
usually whenever you see a do { } while (0) in the actual code (not a macro definition) this is a good candidate to be made into a function |
| 23:44 |
TBC_x |
wtf? I just found a 3.2 GB plaintext file on my disk |
| 23:44 |
TBC_x |
and it is not a log |
| 23:48 |
TBC_x |
it is random sequence of bytes in ascii... looks like xxd output |
