Time |
Nick |
Message |
01:31 |
|
proller__ joined #minetest-dev |
01:31 |
|
proller joined #minetest-dev |
02:10 |
|
vampirefrog joined #minetest-dev |
04:00 |
|
MTDiscord joined #minetest-dev |
04:27 |
|
cranez joined #minetest-dev |
05:10 |
|
YuGiOhJCJ joined #minetest-dev |
05:25 |
celeron55_ |
I'm having a ridiculously busy month now so if it's up to me I'd rather extend this for another month as I won't have time to review properly https://github.com/orgs/minetest/discussions/93 |
08:22 |
|
Warr1024 joined #minetest-dev |
09:27 |
|
YuGiOhJCJ joined #minetest-dev |
11:10 |
|
grorp joined #minetest-dev |
11:13 |
grorp |
merging #14606, #14608 in 10 min |
11:13 |
ShadowBot |
https://github.com/minetest/minetest/issues/14606 -- TouchScreenGUI: Fix shootline not being updated if press and release happen in the same step by grorp |
11:13 |
ShadowBot |
https://github.com/minetest/minetest/issues/14608 -- Add object's own position for each collision to moveresult by grorp |
11:41 |
|
Lupercus joined #minetest-dev |
11:54 |
|
MTDiscord joined #minetest-dev |
12:14 |
sfan5 |
merging #14550, #14539 in 10m |
12:14 |
ShadowBot |
https://github.com/minetest/minetest/issues/14550 -- Allow quoting hypertext attribute values by rubenwardy |
12:14 |
ShadowBot |
https://github.com/minetest/minetest/issues/14539 -- Gate disabling of fog or camera update behind 'debug' priv by sfan5 |
12:28 |
|
Noisytoot joined #minetest-dev |
13:10 |
|
behalebabo joined #minetest-dev |
13:43 |
|
stormchaser3000 joined #minetest-dev |
14:01 |
sfan5 |
writing this down took longer than expected |
14:01 |
sfan5 |
#14613 |
14:01 |
ShadowBot |
https://github.com/minetest/minetest/issues/14613 -- The data structure problem with active objects |
14:05 |
rubenwardy |
yeah that's been a problem for ages. numberZero tried to fix it but ran into issues during feature freeze, was reverted |
14:05 |
rubenwardy |
#6453 |
14:05 |
ShadowBot |
https://github.com/minetest/minetest/issues/6453 -- Improve object-object collision performance |
14:06 |
rubenwardy |
josiah and luatic were also trying to fix it with a spatial parititon library |
14:09 |
Krock |
rubenwardy: your answer is too late |
14:09 |
Krock |
8) |
14:10 |
Krock |
adding an entity LUT to mapblocks would make sense in this case |
14:10 |
Krock |
allows quicker area-based lookups |
14:12 |
|
Lupercus joined #minetest-dev |
14:13 |
|
stormchaser3000 joined #minetest-dev |
14:17 |
rubenwardy |
yeah, I did the same in my non-MT games - I just stored entities in mapblocks, then had both a double linked list and an entity id lookup table. Mapblocks are an existing fixed size partition, and in my games I didn't expect huge numbers of entities in a single block and search radiuses rarely exceeded a blocks or so. So wouldn't have benefited from octtrees |
14:32 |
|
Warr1024 joined #minetest-dev |
15:15 |
|
Lupercus joined #minetest-dev |
15:30 |
|
turtleman joined #minetest-dev |
15:40 |
|
Sharpman joined #minetest-dev |
15:53 |
MTDiscord |
<luatic> note that the approach i am proposing is not fixed-size. it would adapt well to various distributions of entities. |
16:07 |
sfan5 |
I just posted the idea I had today which I think is the most complete one so far |
16:07 |
sfan5 |
but there might be a hole in my design that means it can't work |
16:11 |
MTDiscord |
<luatic> I think my idea is relatively "complete". I'm already 300 lines of code in. Take it with a grain of salt because I've only worked on getting the data structures down and not on how to integrate it. I also haven't begun testing yet. |
16:21 |
sfan5 |
well I'm worrying more about the bookkeeping/invalidation than structure design itself |
16:44 |
MTDiscord |
<bastrabun> sfan5: I put up a readout of position and entityname of what's currently on YL under the URL known to you. Around 4500 entities, most of which are static (signs, signs_lib, itemshelf, itemframes and more). That ratio changes by a lot when there's a large scale battle for a city |
16:47 |
sfan5 |
great |
16:48 |
sfan5 |
looks like you have 70/30 split of static and dynamic entities |
16:58 |
sfan5 |
if my calculations are right this split basically doesn't matter though. O(n²) is so terrible that any minimally decent solution will perform better. |
16:59 |
sfan5 |
(the tipping point is something like 97% of objects move every step) |
17:23 |
sfan5 |
what's the right label for something that has been confirmed but it's not clear if it's a bug |
17:23 |
sfan5 |
? |
17:24 |
Krock |
Bug + Feature labels :3 |
17:25 |
MTDiscord |
<luatic> Bug + Documentation labels. Either it's an implementation or a documentation bug. |
17:26 |
MTDiscord |
<luatic> (or both) |
17:26 |
Krock |
of it's for #14614 I'd consider it a bug |
17:26 |
ShadowBot |
https://github.com/minetest/minetest/issues/14614 -- Entity spawned in `on_step` not visible to `get_objects_inside_radius` |
17:26 |
Krock |
a low-priority one, but still something that should be solved |
17:27 |
sfan5 |
feel free to tag it bug then |
17:27 |
MTDiscord |
<luatic> yeah that's a bug |
18:55 |
|
stormchaser3000 joined #minetest-dev |
20:38 |
sfan5 |
planning to merge #14560 in 15m |
20:38 |
ShadowBot |
https://github.com/minetest/minetest/issues/14560 -- Add unit tests for MapBlock (de)serialization by sfan5 |
20:55 |
sfan5 |
@appguru would you mind if I merged the two MTG PRs without a second review? |
20:56 |
sfan5 |
or maybe quickly look at the API of the second one |
21:48 |
MTDiscord |
<luatic> sfan5: game#3112 is definitely fine, let me take a quick look at the other one |
21:48 |
ShadowBot |
https://github.com/minetest/minetest_game/issues/3112 -- Add API to weather mod by sfan5 |
22:21 |
|
loggingbot_ joined #minetest-dev |
22:21 |
|
Topic for #minetest-dev is now Minetest core development and maintenance. Chit-chat goes to #minetest. https://dev.minetest.net/ https://irc.minetest.net/ https://github.com/minetest |
22:33 |
|
panwolfram joined #minetest-dev |
22:51 |
|
Lupercus joined #minetest-dev |
23:24 |
|
Noisytoot joined #minetest-dev |