| Time |
Nick |
Message |
| 01:13 |
|
Ganome317 joined #minetest-hub |
| 01:39 |
ANAND |
rubenwardy: Me too :D |
| 01:40 |
ANAND |
I've gotten as far as transformation matrices (gonna begin playing around with cameras anytime soon) |
| 01:41 |
ANAND |
Abstraction will help clean up the clutter that is OGL |
| 01:42 |
ANAND |
i.e. Making everything classes would make the code so much cleaner than having all OGL calls in main() |
| 02:00 |
|
ssieb joined #minetest-hub |
| 02:52 |
|
Ganome317 joined #minetest-hub |
| 04:11 |
|
jas_ joined #minetest-hub |
| 05:28 |
|
calcul0n joined #minetest-hub |
| 06:29 |
|
CWz_ joined #minetest-hub |
| 10:26 |
|
Fixer joined #minetest-hub |
| 12:26 |
|
aerozoic joined #minetest-hub |
| 14:15 |
ircSparky |
how does one use the profiler to see the time taken for ABMs? I have some that take ~3000ms every 20-30 seconds |
| 14:16 |
ircSparky |
I set profiler.load = true but it dosnt seem to save any abm stuff |
| 14:20 |
ircSparky |
ah, nvm, there is a command to save moreinfo that what is saved automatically |
| 15:05 |
|
kilbith joined #minetest-hub |
| 15:31 |
|
lisac joined #minetest-hub |
| 16:56 |
|
Krock joined #minetest-hub |
| 17:04 |
|
DS-minetest joined #minetest-hub |
| 17:21 |
ircSparky |
I recently slowed down farming by a lot on my server, I use farming_plus, which uses node timers to handle growth. now, I was hoping this would mean that plants would grow even if there are no players around, but plants that are near populated areas still grow while ones farther away do not. is this something to do with node timers, or do I just need to emerge the blocks when growing the plant and starting new timers? |
| 17:24 |
Krock |
node timers don't re-trigger while being inactive |
| 17:25 |
Krock |
you'd need to get the last timestamp and compare it with the current time to call the growth function |
| 17:26 |
ircSparky |
ok |
| 17:27 |
ircSparky |
what do you mean by "re-trigger"? |
| 17:28 |
shivajiva |
I've seen cascades where a crash at node timer stops them all |
| 17:29 |
ircSparky |
ok, so they arent very reliable |
| 17:30 |
shivajiva |
I just make sure I check the are running once all mods are loaded, in my case |
| 17:30 |
shivajiva |
+y |
| 17:31 |
ircSparky |
ok, thanks for the help |
| 17:31 |
ircSparky |
I think my best option would be to give each plant a "grow at this time" meta and an abm to handle the rest |
| 17:35 |
shivajiva |
hmm I think node timers are a better option once you understand where they fall over, timestamp is good because you can trigger that amount of growth, then it's transparent to the player |
| 17:36 |
Krock |
timestamp + node timer. check growth on_timer, repeat growing steps until the current time is reached |
| 17:36 |
shivajiva |
that's what Krock was pointing out |
| 17:38 |
ircSparky |
I would think a node timer for every crop would be heavier than a single abm |
| 17:38 |
ircSparky |
whick in farming plus is used anyways to make sure timers are going amont other things |
| 17:39 |
ircSparky |
which*amount* |
| 17:39 |
ircSparky |
ugh, among not amount |
| 17:41 |
ircSparky |
though i could use an lbm too |
| 17:41 |
shivajiva |
yea an intelligent deployment with a radius of scope would be reqd shifting the load around and not solving much with node timers :) |
| 17:42 |
Krock |
ircSparky: question is the complexity of finding the right nodes to trigger |
| 17:42 |
Krock |
ABMs are run for all active mapblocks, checking each node |
| 17:43 |
Krock |
node timers are run for the specific mapblock, checking only the triggered node |
| 17:44 |
Krock |
hence there's no node lookup for node timers, giving it a speed advantage regardless on how many of them you use |
| 17:48 |
ircSparky |
ok sweet. and having hundreds of node timers wont eat a ton of RAM, right? |
| 17:50 |
Krock |
all it needs is a content id (2 bytes), a unix timestamp (8 bytes) and a position (6 bytes) plus some management overhead |
| 17:51 |
Krock |
didn't check the code, though. |
| 17:54 |
ircSparky |
awesome, thanks for the help Krock and shivajiva |
| 17:55 |
ircSparky |
oh |
| 17:56 |
ircSparky |
what do you mean by "timestamp" I had in mind a metadata entry with the os.time() |
| 17:59 |
Krock |
that |
| 17:59 |
Krock |
note that it's saved as text |
| 18:45 |
Krock |
Are StackOverflow developers using their own platform to program it? |
| 18:45 |
Krock |
Sorry, that was a strange thing to ask. |
| 19:21 |
|
ssieb joined #minetest-hub |
| 19:32 |
|
garywhite joined #minetest-hub |
| 19:34 |
garywhite |
hello all |
| 19:34 |
|
garywhite1 joined #minetest-hub |
| 19:36 |
garywhite |
hello folks |
| 19:39 |
Krock |
hi garywhite |
| 20:08 |
|
scr267 joined #minetest-hub |
| 20:24 |
|
garywhite joined #minetest-hub |
| 20:41 |
ircSparky |
node meta is detroyed when the node is right? |
| 20:43 |
shivajiva |
yes |
| 22:47 |
|
benrob0329 joined #minetest-hub |
| 23:30 |
|
lisac joined #minetest-hub |
