Time |
Nick |
Message |
00:13 |
benrob0329 |
(A texture pack I'm working on) https://gitlab.com/benrob0329/gameboy |
01:04 |
IcyDiamond |
benrob0329: nice |
01:06 |
_Xenon |
Indeed |
01:07 |
benrob0329 |
Thanks |
01:56 |
GreenDimond |
!tell |
01:56 |
GreenDimond |
gee at least tell me if thats the right cmd |
01:57 |
GreenDimond |
!tell Krock Base64 encode worked, thanks! |
01:57 |
GreenDimond |
:v |
01:57 |
GreenDimond |
sheesh |
01:57 |
GreenDimond |
Fine, be like that |
02:55 |
|
FrostRanger joined #minetest-hub |
06:09 |
|
bwarden joined #minetest-hub |
06:30 |
|
ANAND joined #minetest-hub |
06:54 |
|
lumberJ joined #minetest-hub |
07:31 |
|
CWz joined #minetest-hub |
09:21 |
ANAND |
CDB seems to be down? |
09:21 |
ANAND |
"content.minetest.net’s server IP address could not be found" |
09:22 |
ANAND |
Nvm, works now |
10:49 |
|
ANAND joined #minetest-hub |
11:37 |
|
Fixer joined #minetest-hub |
11:47 |
|
jluc joined #minetest-hub |
12:39 |
|
_Xenon joined #minetest-hub |
12:50 |
rubenwardy |
No, I didn't touch mod storage at all |
12:50 |
rubenwardy |
It was in JSON files from the start |
12:51 |
rubenwardy |
All I did is tell nerzhul to use the abstract API MetadataRef |
12:53 |
rubenwardy |
The API I refactored was player attributes into player meta, but the backend wasn't touched |
13:28 |
|
jas_ joined #minetest-hub |
13:29 |
|
lumidify joined #minetest-hub |
13:37 |
|
Krock joined #minetest-hub |
13:37 |
Krock |
hi imaginary tenplus1 |
13:37 |
jas_ |
hi krock |
13:37 |
jas_ |
and tenplus1 who's not here, as usual |
13:40 |
Krock |
hi jas_ |
13:40 |
jas_ |
the magic of the internet: https://imgur.com/a/1kIO2di |
13:40 |
jas_ |
lol that's the actual title. anyway, hi |
13:45 |
Krock |
that's treason |
13:45 |
Krock |
running bash on Windows :3 |
13:47 |
jas_ |
i don't know powershell |
13:48 |
jas_ |
how's minetest? |
13:48 |
jas_ |
you did a lot of work |
13:48 |
jas_ |
all enjoyable i hope |
13:49 |
jas_ |
paramat too, i hope he's ok |
13:50 |
jas_ |
y'all are just swell, and i wanted to express my sincere gratitude. thanks for minetest! |
13:53 |
ANAND |
:_) |
13:53 |
ANAND |
How touching :') |
13:55 |
Krock |
ok |
15:25 |
nerzhul |
rubenwardy, thanks for explanation |
15:31 |
|
Gael-de-Sailly joined #minetest-hub |
15:35 |
|
Ruslan1 joined #minetest-hub |
18:26 |
rud0lf |
aint nobody f** with rust :) |
18:30 |
NathanS21 |
Even fewer with LUA |
18:32 |
|
behalebabo joined #minetest-hub |
18:51 |
Krock |
> LUA |
18:51 |
Krock |
IT'S NOT LUA PLS |
19:01 |
|
aerozoic joined #minetest-hub |
20:02 |
|
IhrFussel joined #minetest-hub |
20:05 |
|
Fixer joined #minetest-hub |
20:05 |
IhrFussel |
I'm testing num_emerge_threads with only 1 now (was 3 before) ... not sure if the performance is worse |
20:13 |
IhrFussel |
Why is the default 1 though? minetest/src/defaultsettings.cpp:settings->setDefault("num_emerge_threads", "1"); |
20:13 |
IhrFussel |
And somewhere else it tries to choose an appropriate value based on CPU cores |
20:14 |
IhrFussel |
But that never happens, because this default setting exists |
20:16 |
IhrFussel |
No it only checks "numberofprocessors" but I try to to find out why it shouldn't work with a multicore machine too |
20:17 |
IhrFussel |
For example if I set 4 mapgen threads and all of them use some CPU doesn't it mean they work independently even with just 1 CPU? |
20:18 |
IhrFussel |
And the OS chooses which cores to use for which thread of a process...or am I wrong? |
20:34 |
|
Gael-de-Sailly joined #minetest-hub |
20:57 |
IhrFussel |
I asked in #ubuntu and they pretty much say "The OS (kernel) decides which core to use for which thread and the software itself has no say over that" ... so it *should* help to use multiple emerge threads even with just 1 CPU that has multiple cores |
20:58 |
Calinou |
you mean 1 physical core with HyperThreading? |
20:58 |
Calinou |
that's… very uncommon today |
20:58 |
Calinou |
old Intel Atoms were like that |
20:58 |
Calinou |
IhrFussel: also, setting more than 1 emerge thread can result in map generation not being deterministic |
20:58 |
Calinou |
(for instance, caves may be cut early according to minetest.conf.example) |
20:59 |
Calinou |
if you don't care about it, go ahead and set it to an higher value :) |
21:00 |
IhrFussel |
Calinou, conf.example mentions that it only helps when you have *multiuple CPUs* but in theory that should also apply with multiple cores |
21:00 |
IhrFussel |
multiple* |
21:01 |
Calinou |
it's a number of threads, so setting it to the number of logical cores (instead of physical cores) can help a bit |
21:01 |
Calinou |
(hyperthreading typically has a 25% best-case speedup, sometimes it can actually worsen performance) |
21:06 |
IhrFussel |
I don't mean hyperthreading... I mean if my CPU has 8 physical cores (real cores) and I set the mapgen threads to 4 then in theory those 4 threads should be able to share the mapgen workload among multiple cores |
21:07 |
IhrFussel |
Which cores are used is decided by the kernel of course |
21:09 |
Calinou |
yes, affinity is decided by the OS |
21:10 |
IhrFussel |
But why do the docs only mention that this works with "multiprocessor systems"? |
21:10 |
IhrFussel |
Multiprocessor systems are rather uncommon AFAIK |
21:18 |
Calinou |
"multiprocessor" can interchangeably mean "multi-core" or "multi-CPU" :) |
21:18 |
Calinou |
people are often not very accurate with terminology, or its meaning changes over time |
21:19 |
Calinou |
and yes, desktop PCs with multiple physical CPUs are very uncommon (but they do exist) |
21:20 |
IhrFussel |
Calinou, is 1 CPU with 2 cores the same as separate CPUs connected regarding processing of threads? |
21:21 |
IhrFussel |
The same as 2 singlecore CPUs* |
21:21 |
Calinou |
usually, yes. The OS is usually aware that they're just logical CPUs though |
21:21 |
Calinou |
(if you check the task manager in Windows 10, you'll be able to see the difference) |
21:22 |
IhrFussel |
So... as long as a certain process has lots of threads the OS will always be able to use those for different cores...correct? |
21:22 |
Calinou |
I'm not 100% sure but that would make sense |
21:30 |
|
benrob0329 joined #minetest-hub |
21:43 |
|
Calinou joined #minetest-hub |
21:56 |
benrob0329 |
are there any good mods for renaming items? |
21:59 |
Krock |
there was a workbench PR from octacian for MTG |
22:06 |
benrob0329 |
was abandoned unfortunetly, and is more ocmplex than what I'm looking for |
22:06 |
benrob0329 |
*complex |
22:08 |
benrob0329 |
I may just end up making a mod myself for this |
22:08 |
benrob0329 |
call it, "Branding" |
22:10 |
Fixer |
still remember those TYAN motherboard from 2001 or so with two sockets |
22:10 |
Fixer |
motherboards* |
22:10 |
Fixer |
dual pentium 3 |
22:14 |
Krock |
solid 200W heat |
22:28 |
Fixer |
state of the art japanese style heating (2001) |
22:48 |
luk3yx |
Can I check if a mod is installed on a server from a CSM without updating the server mod? |
22:54 |
rubenwardy |
no |
22:55 |
rubenwardy |
btw, CPUs == cores, IhrFussel |
22:56 |
rubenwardy |
!mod [rename |
22:56 |
MinetestBot |
rubenwardy: Could not find anything. |
22:56 |
rubenwardy |
!mod [renam |
22:56 |
MinetestBot |
rubenwardy: Could not find anything. |
22:56 |
rubenwardy |
https://forum.minetest.net/viewtopic.php?f=9&t=21658 |
22:57 |
rubenwardy |
benrob0329^ |
22:57 |
IhrFussel |
rubenwardy, do you have definite knowledge about whether or not num_emerge_threads will help with mapgen on multicore machines? |
22:57 |
rubenwardy |
not sure if good |
23:07 |
* luk3yx |
could use some hacky chat message handler and /help, but if a CSM that overrides chat is installed the handler won't be called |
23:09 |
rubenwardy |
/mods |
23:09 |
rubenwardy |
exists |
23:10 |
luk3yx |
If something catches the /mods chat message it will break. |
23:10 |
rubenwardy |
table.insert(callbacks_table, 1, func) |
23:10 |
rubenwardy |
you can make yourself go first like that |
23:10 |
rubenwardy |
also, why do you need this?? |
23:20 |
benrob0329 |
rubenwardy: thats not a very "good" mod tbh |
23:31 |
rdococ |
Thought: Semi-permadeath with items that allow you to respawn (or a "cloning machine" you own, that if it isn't inactivated, you respawn at) |
23:38 |
|
Gael-de-Sailly joined #minetest-hub |