| Time |
Nick |
Message |
| 00:44 |
galaxie |
How do I increase the pointing distance? As in, make it so I can dig/place nodes from a further distance. |
| 00:46 |
rubenwardy |
range in the item de |
| 00:46 |
rubenwardy |
+f |
| 00:46 |
galaxie |
Oh, so it can be changed depending on what you are pointing at? |
| 00:46 |
rubenwardy |
on what you're pointing with |
| 00:47 |
paramat |
in tool definition |
| 00:47 |
rubenwardy |
the hand is "" |
| 00:47 |
rubenwardy |
so, to change it for all players you can do: minetest.override_item("", { range = 300 }) |
| 00:47 |
rubenwardy |
actually |
| 00:47 |
rubenwardy |
hmmm |
| 00:47 |
paramat |
https://github.com/minetest/minetest_game/blob/dd3f3b2032242e57d23fc183430cebcbf6424b00/mods/creative/init.lua#L47 |
| 00:47 |
rubenwardy |
yeah |
| 00:48 |
paramat |
beware larger ranges affect performance (i think) |
| 00:48 |
rubenwardy |
note that minetest.register_item(":", { }) is exactly the same as minetest.override_item("", { range = 300 }) |
| 00:48 |
rubenwardy |
paramat: not that much |
| 00:48 |
rubenwardy |
it was heavily optimised |
| 00:48 |
paramat |
ok |
| 00:48 |
rubenwardy |
you can have 100s now |
| 00:48 |
rubenwardy |
and it's client side, so it only selected loaded blocks |
| 00:49 |
rubenwardy |
so loading blocks is more of a bottleneck than the pointing |
| 00:50 |
galaxie |
rubenwardy: I suppose using override_item would make more sense, still. |
| 00:51 |
rubenwardy |
oops |
| 00:51 |
rubenwardy |
note that minetest.register_item(":", { range = 300 }) is exactly the same as minetest.override_item("", { range = 300 }) |
| 00:51 |
rubenwardy |
but the latter should be used |
| 00:57 |
galaxie |
Is there a nice mod that allows you to, say, print out the metadata of a clicked node to aid in debugging? Or any other mods that a mod dev might want? |
| 00:58 |
kaeza |
I think inspector is what you want. |
| 00:58 |
kaeza |
!mod inspector |
| 00:58 |
MinetestBot |
kaeza: Chest Inspector [chest_inspector] by tjnenrtn - https://forum.minetest.net/viewtopic.php?t=17692 |
| 00:58 |
kaeza |
Uh, not that one. |
| 01:00 |
kaeza |
!mod inspect |
| 01:00 |
MinetestBot |
kaeza: Inspect nodes! [inspect] by bigfoot547 - https://forum.minetest.net/viewtopic.php?t=17092 |
| 01:01 |
galaxie |
What about https://github.com/minetest-mods/inspector ? |
| 01:01 |
kaeza |
That one. |
| 01:02 |
kaeza |
>An in-game tool to inspect node parameters and metadata. |
| 01:04 |
galaxie |
When you use set_pos and move_to, is there a way to make it move a set of blocks, so as to make it less clunky? |
| 01:07 |
ANAND |
galaxie: set_pos and move_to only work on ObjectRefs (players and Lua entities), and won't work on nodes. |
| 01:07 |
ANAND |
As of now, it's impossible to "move" nodes from one place to another |
| 01:07 |
rubenwardy |
because that wouldn't make sense |
| 01:08 |
rubenwardy |
nodes are directly tied to a position |
| 01:08 |
p_gimeno |
except with pistons (in full node units) |
| 01:08 |
galaxie |
So just remove and place again? |
| 01:09 |
p_gimeno |
that's what pistons do |
| 01:10 |
galaxie |
Hmmm, OK. Is it possible to make a node/entity light on and off? I saw light_source but it appears to be static and for all nodes. |
| 01:11 |
ANAND |
Can be done to entities, but not possible to do to nodes as you observed. |
| 01:11 |
ANAND |
For entities, it's simply a matter of setting `glow` in object properties |
| 01:12 |
galaxie |
Alright. I assume you can make an entity look exactly like a node? |
| 01:12 |
ANAND |
e.g. object:set_properties({glow = 10}) |
| 01:12 |
rubenwardy |
no |
| 01:12 |
rubenwardy |
use two different node definitions |
| 01:12 |
ANAND |
Not possible |
| 01:12 |
ANAND |
Entities could resemble a node, but won't behave like one |
| 01:13 |
paramat |
yeah use 2 nodes, one lit, and replace one with the other |
| 01:13 |
ANAND |
^ |
| 01:13 |
kaeza |
See e.g. furnaces. |
| 01:14 |
galaxie |
And perhaps save the definition to a local table then change the light_source field for the second node? |
| 01:14 |
ANAND |
No, both nodes have to be registered at startup |
| 01:15 |
ANAND |
But yes, you can save the definition to a local table, and change `light_source` before registering the second node |
| 01:15 |
rubenwardy |
could be useful to have minetest.extend_item(oldname, newname, overrides) |
| 01:15 |
galaxie |
That exists? Or were you suggesting it? |
| 01:15 |
ANAND |
It was a suggestion |
| 01:15 |
rubenwardy |
that function doesn't exist |
| 01:15 |
kaeza |
Don't do that. Make a copy of the table. |
| 01:16 |
ANAND |
But that's actually a good idea. |
| 01:16 |
rubenwardy |
you'll need to table.copy |
| 01:17 |
kaeza |
I think register_* stores a reference to the telephonic you pass, so you need two independent copies. |
| 01:17 |
rubenwardy |
telephonic? |
| 01:17 |
kaeza |
... How did that telephonic get there |
| 01:17 |
rubenwardy |
on a phone? |
| 01:17 |
kaeza |
Table* |
| 01:17 |
kaeza |
Yes :p |
| 01:17 |
rubenwardy |
as in, are you using a device with autocorrect? |
| 01:17 |
rubenwardy |
not making a pun about telephonic |
| 01:18 |
rubenwardy |
because that was a telephonic mistake |
| 01:18 |
kaeza |
Hah |
| 01:18 |
ANAND |
lol |
| 01:19 |
MinetestBot |
[git] rubenwardy -> minetest/minetest: Prevent multi-line chat messages server-side (#8420) d0a1a29 https://git.io/fjU37 (2019-03-26T01:18:52Z) |
| 01:23 |
galaxie |
For this sample code, is the minetest.after strictly necessary/recommended? https://pastebin.com/XgmJBQke |
| 01:24 |
rubenwardy |
it shouldn't be needed |
| 01:24 |
rubenwardy |
by on joinplayer, they should be in the game |
| 01:24 |
rubenwardy |
prejoinplayer is before they join the game |
| 01:26 |
|
testman joined #minetest |
| 01:27 |
|
walle303 joined #minetest |
| 01:32 |
galaxie |
So just replace the inside of the function with the insides of func but replace name with player:get_player_name() ? |
| 01:36 |
galaxie |
Is there some way to make a node that can only be dug by an admin? I know there is area protection but I was hoping for something simpler, like a group or something. |
| 01:50 |
Emerald2 |
Like in maptools mod? |
| 01:50 |
ANAND |
One way to do this is to not set any group, except an arbitrary group (e.g. groups = {immortal = 1}), and register a custom pickaxe which can dig nodes with immortal group |
| 01:51 |
ANAND |
with the arbitrary group* |
| 01:51 |
Emerald2 |
https://github.com/minetest-mods/maptools |
| 01:51 |
Emerald2 |
It has an admin pickaxe that can dig anything. |
| 01:51 |
ANAND |
galaxie: See ^ |
| 01:51 |
ANAND |
That mod has a bunch of indestructible nodes. |
| 01:52 |
ANAND |
and an admin pick, ofc |
| 01:52 |
ANAND |
Btw, hey Emerald2 :) |
| 01:54 |
|
Miner_48er joined #minetest |
| 01:54 |
Emerald2 |
Hey ANAND. |
| 01:56 |
|
DI3HARD139 joined #minetest |
| 01:58 |
Emerald2 |
I remember having maptools stuff back in my MM-Survival days. :) |
| 01:58 |
Emerald2 |
We had a bunch of fullclip nodes to keep the players out of the staff area. |
| 02:20 |
|
Cornelia joined #minetest |
| 02:25 |
|
Lia joined #minetest |
| 02:30 |
Emerald2 |
I guess "vel" and "acc" got deprecated in 5.0? |
| 02:40 |
|
milkt joined #minetest |
| 02:41 |
Emerald2 |
Poor admin. People sprinting spams the server console badly. XD |
| 02:46 |
|
cddepppp256 joined #minetest |
| 03:04 |
|
XanT joined #minetest |
| 03:05 |
XanT |
Hi there, is there a team member floating around I can talk to regarding the mod API support within Minetest? |
| 03:05 |
rubenwardy |
regrettably, I'm off to sleep now |
| 03:05 |
rubenwardy |
(most of the devs are european) |
| 03:06 |
XanT |
it can wait till tomorrow then if that's better :) |
| 03:07 |
rubenwardy |
it may be worth just asking |
| 03:07 |
|
botch joined #minetest |
| 03:17 |
|
Nezrok_ joined #minetest |
| 03:23 |
|
ndrsn joined #minetest |
| 03:24 |
|
paramat joined #minetest |
| 03:57 |
MinetestBot |
[git] paramat -> minetest/minetest: Dungeons: Do not remove nodes that have 'is_ground_content = false' (… 5e7662c https://git.io/fjUGc (2019-03-26T03:56:57Z) |
| 03:59 |
|
fling joined #minetest |
| 04:09 |
paramat |
XanT go ahead and ask your question |
| 04:38 |
|
amenji joined #minetest |
| 04:46 |
|
kaeza joined #minetest |
| 05:16 |
|
Ruslan1 joined #minetest |
| 05:16 |
|
lumberJ joined #minetest |
| 05:24 |
|
paramat joined #minetest |
| 05:52 |
|
est31 joined #minetest |
| 06:26 |
|
nowhere_man joined #minetest |
| 06:33 |
|
SanskritFritz joined #minetest |
| 06:47 |
|
Destructo joined #minetest |
| 06:51 |
Destructo |
anyone online? |
| 06:52 |
Destructo |
I have a problem |
| 06:52 |
Emerald2 |
What's the problem? |
| 06:53 |
Destructo |
the one I was facing yesterday..... I am not getting any solution. |
| 06:54 |
Emerald2 |
Oh yes, you're having trouble signing up to the forum. |
| 06:54 |
Destructo |
yes |
| 06:54 |
|
milkt joined #minetest |
| 06:55 |
Destructo |
I am not using any vpn or proxy |
| 06:56 |
Destructo |
still it kicks me out. : / |
| 06:56 |
Emerald2 |
This is probably not the best time of day to find somebody to help. |
| 06:56 |
lumberJ |
is your ip banned for some other reason? :) |
| 06:56 |
Destructo |
no |
| 06:56 |
ANAND |
Not sure about this, but forum moderators might be able to whitelist your IP/account |
| 06:57 |
Emerald2 |
US area are sleeping and UK/Europe probably not here yet. |
| 06:57 |
ANAND |
Europeans are probably still brushing their teeth :) |
| 06:57 |
Destructo |
it just says that a lot of spams have got access through my ip though i have not done anything |
| 06:57 |
Destructo |
i am indian |
| 06:57 |
Emerald2 |
Probably share an IP range with forum spammers then. |
| 06:58 |
Destructo |
i do not understand such technical things |
| 06:58 |
ANAND |
I faced a similar issue with Freenode recently - got banned in the middle of an IRC session, for connecting from a blacklisted IP. |
| 06:58 |
Emerald2 |
Same internet service provider maybe. |
| 06:58 |
* Emerald2 |
shrugs. |
| 06:59 |
Destructo |
hey anand how did you got access to forums |
| 06:59 |
Destructo |
i think you are from india too? |
| 06:59 |
ANAND |
Uh... I signed up? :) |
| 06:59 |
Destructo |
then why cant I ? |
| 06:59 |
Emerald2 |
Vanessa pinged Calinou but he appears to be afk. |
| 06:59 |
|
VanessaE left #minetest |
| 07:00 |
Destructo |
i am not any spammer |
| 07:00 |
Emerald2 |
Oops I think I scared her. :D |
| 07:00 |
|
VanessaE joined #minetest |
| 07:00 |
ANAND |
Yes, I am from India. But fortunately, I (probably) don't share my IP with spammers |
| 07:00 |
Destructo |
someone help me sign in..... |
| 07:00 |
Emerald2 |
Only certain forum moderators can help you, and I don't think they are available right now. |
| 07:01 |
Destructo |
i dont do anything like spamming |
| 07:01 |
ANAND |
sfan5 and rubenwardy are forum moderators too. I hope they might be able to help you. |
| 07:01 |
Destructo |
I prefer peace.... :) |
| 07:01 |
ANAND |
(And I hope that pings them :P) |
| 07:01 |
ANAND |
I prefer peas.... |
| 07:01 |
Destructo |
are they online.. |
| 07:01 |
Emerald2 |
So is Shara. |
| 07:02 |
Destructo |
any of them? |
| 07:02 |
ANAND |
Yes, Shara too. |
| 07:02 |
ANAND |
(uh-oh, double-pinged) |
| 07:02 |
Emerald2 |
Well, if they answer they're here. Otherwise, nope. |
| 07:02 |
Emerald2 |
I think it is still a bit early in their morning. |
| 07:03 |
Destructo |
ahhhhhh |
| 07:04 |
Destructo |
its already midday here :/ |
| 07:04 |
ANAND |
You can come here again a little after 9PM IST to catch them when they're awake. |
| 07:04 |
Destructo |
9PM ? |
| 07:04 |
Emerald2 |
Some of them might be here in 2 hours? |
| 07:04 |
Destructo |
<i have dinner then> |
| 07:05 |
ANAND |
They'll surely be around in their evenings |
| 07:05 |
ANAND |
Not many hang out on IRC in the mornings, as far as I can tell |
| 07:06 |
Destructo |
oh! |
| 07:06 |
ANAND |
Bah, time-zones... |
| 07:06 |
ANAND |
So confusing :) |
| 07:06 |
Destructo |
yeah :D |
| 07:07 |
Destructo |
ummmmmm then can someone create an account for me? |
| 07:07 |
lumberJ |
if your ip is banned it won't matter |
| 07:07 |
Destructo |
or tell me how to change IP address |
| 07:09 |
|
Destructo left #minetest |
| 07:09 |
lumberJ |
lol, no time for answers |
| 07:10 |
lumberJ |
was going to suggest a vpn :P |
| 07:12 |
VanessaE |
don't hide behind proxies or VPNs to get on the forums. |
| 07:12 |
lumberJ |
was a joke :) |
| 07:17 |
Emerald2 |
Speaking of forum spammers. We sure had some word soup ones this past 2 days. |
| 07:18 |
Emerald2 |
Or should I say 'word salad'. |
| 07:18 |
Emerald2 |
Word soup makes as much sense as the spammers. :P |
| 07:36 |
|
amenji joined #minetest |
| 07:48 |
|
CWz joined #minetest |
| 08:03 |
|
lumberJ left #minetest |
| 08:13 |
|
Beton joined #minetest |
| 08:15 |
|
SanskritFritz left #minetest |
| 08:38 |
|
karamel joined #minetest |
| 08:53 |
|
ensonic joined #minetest |
| 09:20 |
|
proller joined #minetest |
| 10:02 |
|
nowhere_man joined #minetest |
| 10:03 |
|
Hirato_ joined #minetest |
| 10:04 |
|
fling joined #minetest |
| 10:05 |
|
twoelk joined #minetest |
| 10:18 |
|
Ruslan1 joined #minetest |
| 10:43 |
|
YuGiOhJCJ joined #minetest |
| 10:52 |
|
proller joined #minetest |
| 10:54 |
Calinou |
Emerald2: I had a look but I don't think I can create forum accounts (unless I do it manually for you, but I think your IP may still prevent you from logging in, not sure) |
| 10:59 |
|
twoelk|2 joined #minetest |
| 11:02 |
|
SmugLeaf joined #minetest |
| 11:03 |
Emerald2 |
Calinou it's not me having the issue. |
| 11:04 |
Emerald2 |
It's Destructo having the issue. |
| 11:04 |
Calinou |
I know |
| 11:11 |
Emerald2 |
So he just can't get a forum account? Or somebody could whitelist him? |
| 11:16 |
|
riff-IRC joined #minetest |
| 11:16 |
Emerald2 |
Ugh more forum spam. |
| 11:16 |
Calinou |
I don't know if we have any control on the whitelist |
| 11:16 |
Calinou |
(or the blacklist, actually) |
| 11:19 |
|
twoelk|2 left #minetest |
| 11:19 |
|
twoelk joined #minetest |
| 11:19 |
|
Fixer joined #minetest |
| 11:20 |
|
proller joined #minetest |
| 11:21 |
|
ndrsn joined #minetest |
| 11:21 |
p_gimeno |
I have the same problem but I do connect with Tor |
| 11:22 |
ndrsn |
Hello folks! |
| 11:27 |
|
ndrsn left #minetest |
| 11:35 |
|
Fulgen joined #minetest |
| 11:55 |
|
calcul0n joined #minetest |
| 12:07 |
|
sonicpp joined #minetest |
| 12:22 |
Emerald2 |
Calinou this spammer keeps posting new crap. |
| 12:26 |
|
kurtzmusch_ joined #minetest |
| 12:37 |
|
sonicpp joined #minetest |
| 12:45 |
|
Emerald2 joined #minetest |
| 12:45 |
|
Emerald2 joined #minetest |
| 12:49 |
|
JDCodeIt joined #minetest |
| 12:49 |
JDCodeIt |
Hi All, does minetest-mods have its own channel or do we discuss here? |
| 13:04 |
|
aheinecke joined #minetest |
| 13:07 |
calcul0n |
you can discuss here, afaik there's no specific chan for mods |
| 13:28 |
|
Wuzzy joined #minetest |
| 13:44 |
JDCodeIt |
VanessaE noted a problem with moreblocks, which I also hit recently. See https://github.com/minetest-mods/moreblocks/issues/130 |
| 13:45 |
JDCodeIt |
The issue might actually be in minetest.clear_craft - somehow corrupting the recipes. |
| 13:45 |
|
milkt joined #minetest |
| 13:49 |
|
ndrsn joined #minetest |
| 14:16 |
|
kaeza joined #minetest |
| 14:24 |
|
nowhere_man joined #minetest |
| 14:56 |
|
nowhere_man joined #minetest |
| 15:02 |
|
Megaf_ joined #minetest |
| 15:24 |
galaxie |
What's the state of integrating LuaJIT in MT? |
| 15:24 |
|
proller joined #minetest |
| 15:28 |
sfan5 |
"integrating"? |
| 15:31 |
galaxie |
Well, doesn't MT come bundled with Lua 5.1 by default? You have to build it yourself with LuaJIT and some of the forum posts seem to suggest not all the Lua code works with LuaJIT. |
| 15:33 |
sfan5 |
building it with luajit is really easy on both linux and mac, and for windows the official downloads come with luajit |
| 15:33 |
sfan5 |
never heard of the issue with "code not working" |
| 15:35 |
JDCodeIt |
Anyone aware of a bug in clearCraftRecipesByInput() when it is deleting two or more crafts that match? |
| 15:36 |
galaxie |
sfan5: Next time I should double-check dates. That was back in 2013, lol. But then, how much more performant is LuaJIT now that we have 5.0? |
| 15:38 |
|
benrob0329 joined #minetest |
| 15:38 |
sfan5 |
what do you mean? LuaJIT has always been far more performant than "normal" Lua and is recommend for use whenever available |
| 15:38 |
galaxie |
sfan5: For instance, in https://dev.minetest.net/Lua_Optimization_Tips, for one of the tips, "Some benchmarks have shown the opposite to be true when using LuaJIT, this must be verified." |
| 15:40 |
rubenwardy |
LuaJIT is pretty amazing |
| 15:40 |
rubenwardy |
but rotting :( |
| 15:40 |
sfan5 |
galaxie: i'm not sure how this edge case would say anything about the performance of LuaJIT in general |
| 15:43 |
|
zarcade_droid joined #minetest |
| 15:49 |
|
milkt joined #minetest |
| 15:51 |
JDCodeIt |
..because my question goes back to the core code, should I take the discussion in the minetest-dev channel? |
| 15:52 |
p_gimeno |
JDCodeIt: if you've found a bug, the best place is the issue tracker |
| 15:52 |
|
calcul0n joined #minetest |
| 15:52 |
p_gimeno |
I'm interested, btw - have you found more details? |
| 15:53 |
JDCodeIt |
I didn't locate it specifically, but have a question about a certain function where there may be a bug. |
| 15:53 |
benrob0329 |
Quality of code is often more important for performance than raw execution speed. |
| 15:54 |
JDCodeIt |
p_gimeno: did you have a look at the function that I mentioned above? |
| 15:56 |
p_gimeno |
JDCodeIt: I'm starting to look into the code, trying to understand the chain of events, I'm pretty sure I'll end up in that function but I wanted some context first |
| 15:56 |
p_gimeno |
just opened l_craft.cpp |
| 15:56 |
JDCodeIt |
moreblocks call clear_craft with input {"default:papyrus", "default:papyrus", "default:papyrus"} |
| 15:57 |
JDCodeIt |
If I comment that out, the dark_green wool recipe is not corrupted (produces dark_green wool instead of green wool) |
| 15:59 |
p_gimeno |
yes, I got that from moreblocks#130 |
| 16:01 |
JDCodeIt |
At that point in the code there are two recipes defined with input of three papyrus - one produces 1 paper, the other produces 4 paper |
| 16:03 |
|
nri joined #minetest |
| 16:03 |
JDCodeIt |
...misspoke - just the 1 paper craft recipe is there. So the code is clearing a craft and stating a new one with 4 paper output |
| 16:04 |
nri |
Hello |
| 16:12 |
AceNovo |
Heyas |
| 16:16 |
p_gimeno |
JDCodeIt: I can't reproduce the bug in the issue in the first place. I get dark green wool. |
| 16:17 |
p_gimeno |
ah wait, stable-0.4 |
| 16:18 |
JDCodeIt |
another clue - 5.0.0 doesn't happen? |
| 16:19 |
p_gimeno |
I couldn't reproduce it there, is all I can say. I'm preparing to compile stable-0.4 now. |
| 16:26 |
p_gimeno |
reproduced in stable-0.4 |
| 16:27 |
p_gimeno |
is stable-0.4 going to get any more commits? |
| 16:28 |
rubenwardy |
unlikely |
| 16:28 |
JDCodeIt |
So even if we find the bug, we wouldn't be able to backport? |
| 16:29 |
p_gimeno |
in a private branch, sure |
| 16:29 |
rubenwardy |
it can be backported to 5.0.1 |
| 16:30 |
p_gimeno |
ummmm, wait |
| 16:30 |
p_gimeno |
I can reproduce with 5.0, I wasn't using a clean enough environment |
| 16:30 |
|
proller joined #minetest |
| 16:31 |
p_gimeno |
this is head-scratching |
| 16:31 |
JDCodeIt |
we got a live one, then! |
| 16:31 |
p_gimeno |
let me try to identify the definitive cause |
| 16:32 |
p_gimeno |
I still have MTG rolled back to the commit specified in that issue, so there are still some factors to consider |
| 16:34 |
|
YuGiOhJCJ joined #minetest |
| 16:41 |
p_gimeno |
these two MTG commits fixed the issue: https://github.com/minetest/minetest_game/commits/6f80fd3737cd4de30e548dd48bc0ab8c288d9930 |
| 16:43 |
|
puzzlecube joined #minetest |
| 16:44 |
JDCodeIt |
To get coral to generate in mgv7 are there any specific flags needed? |
| 16:48 |
JDCodeIt |
p_gimeno: Color grouping...so upgrade MTG to get these improvements? |
| 16:50 |
JDCodeIt |
p_gimeno: But still not quite clear why clear_craft causes the side-effect. |
| 16:59 |
p_gimeno |
JDCodeIt: I can try to get to the bottom of it, but I wonder if it's worth the effort |
| 17:00 |
JDCodeIt |
Seems rare, and if you say upgrading MTG fixes it, we can wait for another example, but like the stray pointer bugs, it could be hard to find |
| 17:03 |
|
FrostRanger joined #minetest |
| 17:16 |
|
Gael-de-Sailly joined #minetest |
| 17:29 |
|
calcul0n joined #minetest |
| 17:31 |
JDCodeIt |
Anyone have problems to generate coral reef in mgv7? |
| 17:35 |
kurtzmusch_ |
can i find an explanation of each mapgen somewhere? like what features it tryes to generate, aswell as what types of noise it uses... |
| 17:43 |
scr267 |
Anyone has any experience using Shara's other_worlds? |
| 17:44 |
p_gimeno |
JDCodeIt: I believe I've found one cause |
| 17:44 |
JDCodeIt |
Yes, how so... |
| 17:46 |
JDCodeIt |
scr267: there is some info on the dev.minetest.net wiki about mapgen. I've tried to update it based on what I found in lua_api.txt |
| 17:47 |
scr267 |
JDCodeIt, Thanks - I'm just asking because I'm having difficulties getting asteroids to generate. Thanks for the link! |
| 17:48 |
p_gimeno |
JDCodeIt: https://github.com/minetest/minetest/blob/5e7662ca168b47ed3e81901d53bff2eb712f571c/src/craftdef.cpp#L1005-L1013 |
| 17:49 |
JDCodeIt |
That's where I was looking - off by one bug? |
| 17:49 |
p_gimeno |
JDCodeIt: that loop works by adding the items that are not cleared to new_vec_by_input, then swapping it (don't ask me why) |
| 17:49 |
|
jluc joined #minetest |
| 17:49 |
p_gimeno |
(instead of just removing the unwanted items) |
| 17:49 |
|
twoelk left #minetest |
| 17:49 |
p_gimeno |
BUT it iterates in reverse, while the items are added forwards |
| 17:50 |
p_gimeno |
therefore, the new items are in reverse order, and that does make a difference in the case of dark green dye (not sure why) |
| 17:51 |
nri |
scr267: did you ever get my link from VE about the day/night setting? |
| 17:51 |
JDCodeIt |
what is that .swap that is inside the for loop? |
| 17:51 |
p_gimeno |
JDCodeIt: I bet that if a new clear_recipe() is called, the issue disappears |
| 17:51 |
scr267 |
nri: yes I sure did. and I tried to respond to you to thank you. |
| 17:51 |
scr267 |
Thanks very much! |
| 17:51 |
nri |
NP |
| 17:52 |
nri |
scr267: Did you figure out how to use it? |
| 17:52 |
JDCodeIt |
ahhh - call it twice? I'll try it. |
| 17:52 |
scr267 |
nri: Not yet, I did put it on my local dev but I'll get around to it :) |
| 17:52 |
nri |
ok |
| 17:53 |
p_gimeno |
JDCodeIt: because that will reverse it twice (if my hypothesis of what's going on is correct) |
| 17:53 |
p_gimeno |
JDCodeIt: https://en.cppreference.com/w/cpp/container/vector/swap |
| 17:53 |
|
nerzhul joined #minetest |
| 17:53 |
p_gimeno |
I have to study the code better, but I think there's a bug right there |
| 17:54 |
p_gimeno |
(in the swap) |
| 17:54 |
JDCodeIt |
why will got_hit be true on the second call? |
| 17:55 |
p_gimeno |
JDCodeIt: I mean to add another recipe just to be deleted |
| 17:55 |
JDCodeIt |
Yes, could do... |
| 17:56 |
p_gimeno |
JDCodeIt: that would also mean that there are TWO definitions of the recipe, one producing dark green wool and the other producing bright green wool |
| 17:56 |
|
proller joined #minetest |
| 17:56 |
p_gimeno |
changing the order implies changing which one is picked |
| 17:57 |
p_gimeno |
I think that duplication is probably what the MTG commits I mentioned earlier fixed |
| 17:58 |
|
ssieb joined #minetest |
| 18:01 |
JDCodeIt |
Well at least for that recipe - it is confirmed - cleared the paper recipe, added the paper recipe, then cleared it a second time - dark green wool |
| 18:03 |
JDCodeIt |
Dark green wool results also if you add back the recipe - so the twice cleared system works? But did it simply destroy a recipe I'm not looking at? |
| 18:05 |
JDCodeIt |
I guess the funny(odd) thing is that I'm deleting the paper recipe, and it affects the dark green wool recipe. |
| 18:06 |
JDCodeIt |
You're saying I think that the dark green wool has a wrong recipe that simply changes order by that routine. |
| 18:06 |
p_gimeno |
the issue is that every time clear_recipe() clears something, the list order is inverted |
| 18:07 |
JDCodeIt |
I always wondered about that - if a craft recipe is registered, I guess there is nothing that looks to see if there is a conflict - and issues a warning. |
| 18:07 |
p_gimeno |
that makes the "bad" wool recipe to be found before the "good" one |
| 18:08 |
p_gimeno |
yeah, duplicate recipes are possible |
| 18:08 |
p_gimeno |
I'm creating a mod to dump the craft definitions of dark green wool and light green wool |
| 18:08 |
JDCodeIt |
...and does the engine answer back with two predicted outputs or just one? |
| 18:10 |
|
sonicpp joined #minetest |
| 18:12 |
p_gimeno |
the first match it finds, I guess |
| 18:14 |
p_gimeno |
ok, found the recipes |
| 18:14 |
p_gimeno |
Recipes for wool:green: |
| 18:14 |
p_gimeno |
group:dye,basecolor_greengroup:wool |
| 18:14 |
p_gimeno |
Recipes for wool:dark_green: |
| 18:14 |
p_gimeno |
group:dye,unicolor_dark_greengroup:wool |
| 18:14 |
JDCodeIt |
The second is of course "wrong" |
| 18:15 |
p_gimeno |
the second? |
| 18:15 |
p_gimeno |
it looks to me that the problem is that dark green dye matches basecolor_green |
| 18:16 |
JDCodeIt |
sorry. my misread there - it is that the basecolor_green is both green and dark_green? |
| 18:16 |
p_gimeno |
I would say so |
| 18:16 |
JDCodeIt |
So inverting the list of recipes is not a bug. |
| 18:17 |
p_gimeno |
well, it causes side effects, as can be seen above |
| 18:17 |
p_gimeno |
so it's kind of a bug |
| 18:17 |
JDCodeIt |
Any particular reason you can think of that the list is processed in reverse? |
| 18:18 |
JDCodeIt |
I guess there could be some expectation that the first registered recipe takes precedence |
| 18:18 |
p_gimeno |
as I said, the loop at lines 1005-1006 loops in reverse, but elements are added to the new list in the forward direction (line 1011) |
| 18:19 |
p_gimeno |
as a quick test, I replaced push_back(def) with insert(new_vec_by_input.begin(), def), and I got dark green wool as expected |
| 18:20 |
p_gimeno |
registering a recipe twice might be already caught somewhere - haven't checked. But in this case, the recipes were *different*. |
| 18:20 |
JDCodeIt |
Just wondering if the processing order of reverse was a legacy thing - the author was removing nodes from the middle? |
| 18:21 |
JDCodeIt |
but now a new list is built and swapped |
| 18:21 |
p_gimeno |
to clarify, the processing order has not changed, the list is reversed in both stable-0.4 and master |
| 18:21 |
JDCodeIt |
Yes, understood your solution. |
| 18:22 |
p_gimeno |
what has changed, is how these recipes are registered, let me check the new definitions after these commits |
| 18:23 |
JDCodeIt |
I was just wondering - why even process in reverse if you are not touching the original list. |
| 18:24 |
p_gimeno |
Recipes for wool:green: |
| 18:24 |
p_gimeno |
group:dye,color_greengroup:wool |
| 18:24 |
p_gimeno |
Recipes for wool:dark_green: |
| 18:24 |
p_gimeno |
group:dye,color_dark_greengroup:wool |
| 18:24 |
JDCodeIt |
Seems very assembly - DEC, then JPZ |
| 18:24 |
p_gimeno |
JDCodeIt: no idea... probably patch after patch after patch :) |
| 18:25 |
p_gimeno |
what I wonder is why to make a new list in the first place |
| 18:25 |
JDCodeIt |
Good point - what type of list is this? linked list? |
| 18:26 |
p_gimeno |
std::vector, it's up to the implementation |
| 18:26 |
p_gimeno |
that method may work better if massive deletions are expected, but in most cases, I doubt it |
| 18:27 |
JDCodeIt |
by override or the base? |
| 18:28 |
p_gimeno |
I don't understand the question |
| 18:28 |
JDCodeIt |
Is it a template class or it has some base implementation? |
| 18:30 |
p_gimeno |
it's part of the STL, it does have a base implementation that depends on the C++ library, but I think there are guidelines about the expected runtime of each operation (like e.g. std::vector::swap() does not depend on the number of elements etc) |
| 18:30 |
JDCodeIt |
I read that std:vector<> is guaranteed to occupy contiguous memory - sounds like not linked list - but I would think you could override that |
| 18:32 |
JDCodeIt |
Maybe was speed - but really, clear_craft should be called at startup, not while running, eh? |
| 18:34 |
p_gimeno |
yeah |
| 18:40 |
p_gimeno |
well, I sorta get it. Doing it this way is O(n) while doing it with vector::erase() it's O(n²). |
| 18:49 |
p_gimeno |
remove_if <https://en.cppreference.com/w/cpp/algorithm/remove> seems like what should have been used here |
| 18:51 |
JDCodeIt |
Where are we at C++11 - this is supported? |
| 18:52 |
p_gimeno |
remove_if is previous to C++11, anyway Minetest has already switched to C++11 |
| 18:53 |
JDCodeIt |
it even supports parallelism - nice. |
| 18:55 |
p_gimeno |
seems that the reverse loop is there since the beginning: https://github.com/minetest/minetest/commit/281e9f39fdffd106e79122ae23e2b5443a5daea3 |
| 18:55 |
JDCodeIt |
I'll add our findings to the original issue report - main thing is that there are essentially duplicate recipes in play. |
| 18:58 |
p_gimeno |
yes, apparently the problem is that dark green dye matches basecolor_green |
| 19:00 |
p_gimeno |
and that on top of that, clear_craft is reversing the internal recipes list |
| 19:01 |
p_gimeno |
I'll add a report for the latter |
| 19:01 |
JDCodeIt |
Issue https://github.com/minetest-mods/moreblocks/issues/130 comments added. |
| 19:07 |
|
sonicpp joined #minetest |
| 19:08 |
|
karamel joined #minetest |
| 19:15 |
scr267 |
Anyone have issues with connected glass showing strange slab-shaped artifacts? |
| 19:22 |
p_gimeno |
#8427 |
| 19:22 |
ShadowBot |
https://github.com/minetest/minetest/issues/8427 -- clear_craft() is reversing the recipes list |
| 19:24 |
p_gimeno |
galaxie: whoever made that claim had no idea of how LuaJIT operates |
| 19:24 |
|
troller joined #minetest |
| 19:25 |
p_gimeno |
I wish there was a MediaWiki blame tool... |
| 19:27 |
|
LMD joined #minetest |
| 19:29 |
|
LMD joined #minetest |
| 19:30 |
galaxie |
I saw the page on MVC programming for MT mods and all, but I'd still like to know a way to mock minetest. Just a simple table that does nothing for functions that get called, etc. |
| 19:31 |
p_gimeno |
anyway, I've removed that sentence |
| 19:32 |
rubenwardy |
galaxie: see the person that made the profiler |
| 19:32 |
rubenwardy |
hmm |
| 19:32 |
rubenwardy |
T4im |
| 19:32 |
rubenwardy |
https://github.com/t4im?tab=repositories&type=source |
| 19:32 |
rubenwardy |
iirc there's some mocks in one of these repos |
| 19:33 |
rubenwardy |
would be nice to have a standard mocking library for Minetest |
| 19:34 |
JDCodeIt |
Can't seem to get the coral reef decoration to appear in mgv7 - I even commented out the biomes requirement - not looking hard enough or is there a trick? |
| 19:35 |
galaxie |
rubenwardy: Which one would be the profiler? shakedown? coretest? Or is it integrated with MT? |
| 19:35 |
|
Taoki joined #minetest |
| 19:37 |
rubenwardy |
galaxie: the profiler is built in to minetest |
| 19:38 |
|
AntumDeluge joined #minetest |
| 19:41 |
galaxie |
ruberwardy: Apparently there's a minitest/mock |
| 19:49 |
kurtzmusch_ |
guise, where can i configure the parameters for mapgen? |
| 19:55 |
JDCodeIt |
mapgen params : https://dev.minetest.net/Mapgen_Parameters |
| 20:00 |
kurtzmusch_ |
tx |
| 20:00 |
|
nowhere_man joined #minetest |
| 20:08 |
|
nri joined #minetest |
| 20:15 |
|
Darcidride joined #minetest |
| 20:31 |
|
FreeFull joined #minetest |
| 20:39 |
|
nri joined #minetest |
| 20:49 |
nri^AFK |
Nri went afk |
| 20:52 |
kaeza |
AFK nicks/notices are cancer. |
| 20:53 |
kaeza |
Also greetings. |
| 20:53 |
nri^AFK |
lol |
| 20:54 |
nri^AFK |
test |
| 20:54 |
nri^AFK |
afk?! |
| 20:55 |
|
nri^AFK left #minetest |
| 20:55 |
|
nri joined #minetest |
| 20:59 |
|
cddepppp256 joined #minetest |
| 21:08 |
JDCodeIt |
Can anything be controlled on the priority of the emerge threads? Or anything to make less impact on the lag it can cause? |
| 21:14 |
p_gimeno |
not really |
| 21:14 |
p_gimeno |
the lag is due to locking |
| 21:15 |
p_gimeno |
or that's what I presume |
| 21:19 |
p_gimeno |
any idea what is/was worldedit_infinity / worldedit_limited for? |
| 21:40 |
JDCodeIt |
size of the area to edit I think |
| 21:45 |
|
YuGiOhJCJ joined #minetest |
| 21:46 |
p_gimeno |
I mean, they are mods in the WorldEdit modpack |
| 21:46 |
p_gimeno |
er, were |
| 21:47 |
p_gimeno |
it keeps working when omitting both |
| 21:48 |
JDCodeIt |
limited was "limited" to 30x30x30 volume and made a safe wrapper for the functions |
| 21:48 |
JDCodeIt |
infinity has some kind of ray calculator I never used |
| 21:48 |
p_gimeno |
ahh thanks |
| 21:54 |
|
diemartin joined #minetest |
| 22:01 |
|
riff-IRC joined #minetest |
| 22:19 |
|
nowhere_man joined #minetest |
| 22:39 |
|
YuGiOhJCJ joined #minetest |
| 22:49 |
|
kaeza joined #minetest |
| 22:56 |
nerzhul |
no lag is not due to locking |
| 22:56 |
nerzhul |
we loose many time on lua callbacks generally |
| 23:08 |
|
argyle77 joined #minetest |
| 23:12 |
|
walle303 joined #minetest |
| 23:42 |
Emerald2 |
Is there a way to stop the technic chainsaw massacring so much of the forest? |
| 23:42 |
|
paramat joined #minetest |
| 23:48 |
paramat |
!tell JDCodeIt coral reefs are intentionally slightly rare, see https://github.com/minetest/minetest_game/pull/2254#issuecomment-438866092 |
| 23:48 |
MinetestBot |
paramat: I'll pass that on when JDCodeIt is around |
| 23:49 |
|
cddepppp256 joined #minetest |
