Page 6 of 9
Re: [Modpack] Plantlife [git rolling release][plantlife_modp
Posted:
Sun Sep 07, 2014 14:43
by Inocudom
I really don't think the vines mod will be worked on ever again. Maybe it could be added to this modpack?
Re: [Modpack] Plantlife [git rolling release][plantlife_modp
Posted:
Sun Sep 07, 2014 16:52
by VanessaE
I could possibly do that.
Re: [Modpack] Plantlife [git rolling release][plantlife_modp
Posted:
Mon Sep 08, 2014 11:00
by balthazariv
VanessaE wrote:Ok, remove your change and try updating plantlife. See if the change I just pushed fixes it.
It works again.
Thanks
Re: [Modpack] Plantlife [git rolling release][plantlife_modp
Posted:
Wed Sep 10, 2014 21:27
by VanessaE
Update: I have imported the vines-and-rope mod.
Re: [Modpack] Plantlife [git rolling release][plantlife_modp
Posted:
Thu Oct 02, 2014 08:36
by VanessaE
Update: For those who have not been watching the commit log, I have pushed a performance update in the past couple of days that massively smooths out the latency of this modpack. Unlike previous attempts, this one actually has numbers to back it up. Just figured I would hang back a bit and let some choice folks test it more before announcing. :-D
What the code does now is, rather than adding stuff in one big gulp when the engine's mapgen runs, the code just logs the mapgen's activities and exits without changing anything. At the same time (more or less), the code "plays back" the current contents of that log via a globalstep, and it's at this time that the code adds all of the plants, saplings, trees, molehills, and everything else, one "registered action" per globalstep. Many thanks to sapier for this idea.
The total throughput is about the same as before, so it'll still take just as much real time for the terrain to be fully populated, but now the maximum lag is now only a tiny fraction of what it was: the numbers I've seen via Minetest's builtin profiler now hover at around 200 or so milliseconds in this new code, versus upwards of 15 seconds in the old code.
This means that you can actually DO stuff while the terrain is being created (e.g. inventory operations, chat commands, build/dig actions, etc), and it keeps the code from stalling out a multiplayer server nearly as easily when players are exploring new terrain.
This comes with a small cost: instead of everything just appearing in its fully-completed form after some time, the foliage is now laid down in layers, and you can actually see this happening. Unless you're using a 486, this probably won't bother you.
There are still more performance tweaks to be made, as well as perhaps some further anti-lag tuning if it turns out this is needed. Watch this space for updates.
Re: [Modpack] Plantlife [git rolling release][plantlife_modp
Posted:
Fri Oct 03, 2014 02:39
by ExeterDad
Oh! Time for a update. I can always tell when my kids are off exploring. My pathetic server starts tasting it's spleen. Anxious to give you a followup. Ummmm... Will this be safe to update on a existing world?
Re: [Modpack] Plantlife [git rolling release][plantlife_modp
Posted:
Fri Oct 03, 2014 02:49
by VanessaE
Yep, it should be fine on old worlds too - it has been tested in this way also.
Re: [Modpack] Plantlife [git rolling release][plantlife_modp
Posted:
Fri Oct 03, 2014 03:12
by Zeno
ExeterDad wrote:Ummmm... Will this be safe to update on a existing world?
I've been testing it for about 6 or 7 days on Onez' Server and not noticed anything unexpected.
Re: [Modpack] Plantlife [git rolling release][plantlife_modp
Posted:
Sat Oct 04, 2014 20:22
by Michael Eh?
New bush please... coffee!
return of grapes?
craft coffee into pot and cup of coffee. Paper cups instead of clay unless cups are reusable.
How about just 'IVY' and not just poison ivy for decorative on buildings?
How about Apple pies, banana bread, chocolate cake and hot chocolate to make use of cocoa, banana and apples already out there?
Re: [Modpack] Plantlife [git rolling release][plantlife_modp
Posted:
Sun Oct 05, 2014 20:16
by VanessaE
Ivy (like Virginia creeper or kudzu?) might be a good idea. Grapevines too, perhaps. Coffee is doable also. Those various foods, though, seem like they belong (or are already) in rubenwardy's food mod though.
Re: [Modpack] Plantlife [git rolling release][plantlife_modp
Posted:
Mon Oct 06, 2014 00:02
by Sokomine
Michael Eh? wrote:New bush please... coffee!
return of grapes?
Mauvebics hydro mod has coffee, grapes, habaneros, peas, tomatoes and rose bushes. Hydro was aimed mostly at greenhouses or underground plantations. Two or three node high grapes - which regrow their fruits after some time - would be very decorative in the landscape. That is, as manually planted plants.
Re: [Modpack] Plantlife [git rolling release][plantlife_modp
Posted:
Mon Oct 06, 2014 00:09
by ExeterDad
VanessaE wrote:Ivy (like Virginia creeper or kudzu?) might be a good idea. Grapevines too, perhaps. Coffee is doable also. Those various foods, though, seem like they belong (or are already) in rubenwardy's food mod though.
I'm a fan of Virginia Creeper. Especially this time of the year when it goes scarlet.
Re: [Modpack] Plantlife [git rolling release][plantlife_modp
Posted:
Mon Oct 06, 2014 01:05
by ExeterDad
BTW I don't have numbers... but your latest performance update has DRASTICALLY improved my server performance when new map is generated. How much I can't be sure since when updating, I also installed the snow mod which surely would slow things down.
Before (old plantlife. and no snow) when flying in new world... I could go for about 15 seconds til I hit the edge of the world. I would have to wait for 20 seconds or so before more map would come in. Now I can fly and new world shows before I cross the line. And this is with the new snow mod running as well.
To be fair, my box is way under spec until I hit the lottery. But we've dealt with it. But now, it's getting downright pleasant to move around :P
You're a Goddess. Thanks :D
Re: [Modpack] Plantlife [git rolling release][plantlife_modp
Posted:
Mon Oct 06, 2014 01:18
by VanessaE
Sounds like I did something right ;-)
Re: [Modpack] Plantlife [git rolling release][plantlife_modp
Posted:
Mon Oct 06, 2014 13:20
by DarkgarX
VanessaE wrote:Sounds like I did something right ;-)
This mod is beautiful! Great work, I will definitely be using this in our second season streaming setup in Nov-Dec-ish. So lush and fabulous :)
Re: [Modpack] Plantlife [git rolling release][plantlife_modp
Posted:
Tue Oct 14, 2014 08:09
by Fritigern
I know that flowers are part of this modpack, but i can't seem to find any in newly generated worlds. I wonder why.
Re: [Modpack] Plantlife [git rolling release][plantlife_modp
Posted:
Tue Oct 14, 2014 08:14
by Topywo
Fritigern wrote:I know that flowers are part of this modpack, but i can't seem to find any in newly generated worlds. I wonder why.
Flowers is a mod of minetest_game.
Re: [Modpack] Plantlife [git rolling release][plantlife_modp
Posted:
Tue Oct 14, 2014 08:24
by Fritigern
Topywo wrote:Fritigern wrote:I know that flowers are part of this modpack, but i can't seem to find any in newly generated worlds. I wonder why.
Flowers is a mod of minetest_game.
Are you hovering around the forum waiting to pounce people that appear to have made a mistake? Way to get your number of posts up.
If you read closely, you will see that i said "flowers
are ..." not "flowers
is ...". So I was obviously
NOT talking about the mod called flowers.
Now please, take your snark elsewhere, or actually TRY to be helpful.
Re: [Modpack] Plantlife [git rolling release][plantlife_modp
Posted:
Tue Oct 14, 2014 14:29
by VanessaE
Indeed, flowers are supplied by minetest_game, not by Plantlife. What this modpack includes with it is waterlilies, seaweed, and one or two other similar things - only the stuff that used to be part of the standard flowers mod before those basic flowers were imported into minetest_game, and whatever I've added since.
Since waterlilies and so forth got left out of the game's version of the flowers mod, I kept them in Plantlife and renamed my version of that mod to flowers_plus.
I don't know for sure why you're not seeing any regular flowers, but I know you use the mg mod on some of your worlds. Since that mod supplies a custom mapgen, could it be the reason? I don't know what its settings are regarding flowers.
Re: [Modpack] Plantlife [git rolling release][plantlife_modp
Posted:
Wed Oct 15, 2014 04:18
by Fritigern
VanessaE wrote:I don't know for sure why you're not seeing any regular flowers, but I know you use the mg mod on some of your worlds. Since that mod supplies a custom mapgen, could it be the reason? I don't know what its settings are regarding flowers.
mg definitely added flowers to the (green) biomes, but with plantlife added, the only flowers are the ones found in village gardens.
Re: [Modpack] Plantlife [git rolling release][plantlife_modp
Posted:
Thu Oct 16, 2014 18:00
by Minetestforfun
Hi VanessaE,
Many players (with me), think the pebble have too many nodes, it's not very "minetest like"...
For the moment the pebble have around 10/15 nodes, but it's very ugly...
Please, can you change the model of the pebble or, give us a possibility to have a 1/2/3 node pebble model ?
Re: [Modpack] Plantlife [git rolling release][plantlife_modp
Posted:
Thu Oct 16, 2014 18:03
by jp
Minetestforfun wrote:give us a possibility to have a 1/2/3 node pebble model ?
Here is your easy solution :
viewtopic.php?f=14&t=2840
Re: [Modpack] Plantlife [git rolling release][plantlife_modp
Posted:
Thu Oct 16, 2014 18:29
by VanessaE
I'm pretty happy with the shape and size of the pebbles actually. I see no reason to change them right now, especially with mesh nodes coming VERY soon (as in, I've played with the code myself just in the past 24 hours). Once those go into mainline, we will no longer be restricted to boxy, pixelated-looking stuff; I may redesign the pebble then. We'll see.
Re: [Modpack] Plantlife [git rolling release][plantlife_modp
Posted:
Sun Oct 19, 2014 21:06
by Carnildo
While working on reconciling plants_lib's biomes with those of the Watershed map generator, I noticed that plants_lib uses PerlinNoise() rather than minetest.get_perlin(). As I understand it, PerlinNoise() ignores the world seed, which would mean that plants_lib's biomes don't change from world to world. Is there a reason plant_lib does things this way?
Re: [Modpack] Plantlife [git rolling release][plantlife_modp
Posted:
Sun Oct 19, 2014 22:46
by VanessaE
Actually there's no particular reason at all - in fact you can see that I use both methods within the code.
The temperature and humidity noises written that way was probably just an artifact of learning how to use perlin noise. I just never rewrote it.
Re: [Modpack] Plantlife [git rolling release][plantlife_modp
Posted:
Fri Nov 14, 2014 15:56
by afflatus
I love this mod, thank you so much for all the work you've put into it!
Due to the order things get spawned in different mapgens, I often find things growing where there isn't enough light. So I added this hack to woodsoils/nodes.lua - it is far from perfect, it messes up the subtleties around ferns and results in little growth under oak trees. I may modify it to differentiate between conifers and deciduous trees, but for now it has the right sort of effect. Currently it kills off anything in group:flora, which may leave wild veg from farming* untouched.
Your phone or window isn't wide enough to display the code box. If it's a phone, try rotating it to landscape mode.
- Code: Select all
-- Die-off ABM
minetest.register_abm({
nodenames = {"default:dirt",
"woodsoils:dirt_with_leaves_1",
"woodsoils:dirt_with_leaves_2",
"woodsoils:grass_with_leaves_1",
"woodsoils:grass_with_leaves_2",
"default:dirt_with_grass"
},
interval = 2,
chance = 200,
action = function(pos, node)
local above = vector.add(pos, {x=0, y=1, z=0})
local name = minetest.get_node(above).name
local nodedef = minetest.registered_nodes[name]
if nodedef and (nodedef.sunlight_propagates or nodedef.paramtype == "light")
and nodedef.liquidtype == "none" then
if (minetest.get_node_light(above, 0.5) or 0) < 13 then
if minetest.find_node_near(above, 9, {"group:tree","group:sapling"})
or minetest.find_node_near(above, 3, {"group:leaves"}) then
if (minetest.get_node_light(above, 0.5) or 0) >= 11 then
minetest.set_node(pos, {name = "woodsoils:grass_with_leaves_1"})
elseif (minetest.get_node_light(above, 0.5) or 0) >= 9 then
minetest.set_node(pos, {name = "woodsoils:grass_with_leaves_2"})
elseif (minetest.get_node_light(above, 0.5) or 0) >= 7 then
minetest.set_node(pos, {name = "woodsoils:dirt_with_leaves_1"})
if minetest.get_node_group(name, "flora") > 0 then
minetest.remove_node(above)
end
else
minetest.set_node(pos, {name = "woodsoils:dirt_with_leaves_2"})
if minetest.get_node_group(name, "flora") > 0 then
minetest.remove_node(above)
end
end
end
elseif minetest.find_node_near(above, 3,
{"ferns:fern_04","ferns:fern_03","ferns:fern_02","ferns:fern_01","group:sapling"}
) == nil then
if(minetest.get_node_light(above, 0.5) or 0) < 7 then
minetest.set_node(pos, {name = "default:dirt"})
elseif (minetest.get_node_light(above, 0.5) or 0) >= 13 then
minetest.set_node(pos, {name = "default:dirt_with_grass"})
end
end
end
end
})
Re: [Modpack] Plantlife [git rolling release][plantlife_modp
Posted:
Fri Nov 14, 2014 16:05
by afflatus
Of course if we were going to do this job properly we'd have to map soil acidity as well as temperature and humidity ... :D
Re: [Modpack] Plantlife [git rolling release][plantlife_modp
Posted:
Fri Nov 14, 2014 16:10
by VanessaE
That's kinda what the generic "can grow here" perlin layer is for. It's sort of a soil "fertility factor". I'm not sure I like the ABM-method of reworking the plants under the trees, but it's a good start though.
Re: [Modpack] Plantlife [git rolling release][plantlife_modp
Posted:
Fri Nov 14, 2014 16:50
by afflatus
Sure additional ABMs are not ideal. It's a hack. I just upgraded mods and notice that paragenv7 now plays much better with plantlife and moretrees.
Re: [Modpack] Plantlife [git rolling release][plantlife_modp
Posted:
Thu Nov 20, 2014 13:31
by Fritigern
VanessaE, the latest Git version of minetest checks for uninitialized globals. Plantlife has a few of these instances, the worst one being the one in plants_lib. see my fixes below. they stopped the warnings.
I hope this helps.
Message: "
03:32:58: WARNING: Undeclared global variable "spawn_plants" accessed at ...orlds/Pacata/worldmods/plantlife/plants_lib/init.lua:502"
When: spammed several times per second after logging into my local server.
OLD
Your phone or window isn't wide enough to display the code box. If it's a phone, try rotating it to landscape mode.
- Code: Select all
if type(spawn_plants) == "string" then
assert(loadstring(spawn_plants.."(...)"))(pos)
NEW
Your phone or window isn't wide enough to display the code box. If it's a phone, try rotating it to landscape mode.
- Code: Select all
if type(biome.spawn_plants) == "string" then
assert(loadstring(biome.spawn_plants.."(...)"))(pos)
Notes: This was probably a remnant of an earlier version where the biome namespace did not yet exist. Overlooked whilst introducing the biome namespace (it's called a namespace, right? If not, then i am sure what i meant to say)
Message: "
04:04:24: WARNING: Assignment to undeclared global "walldir" inside a function at .../../worlds/Pacata/worldmods/plantlife/vines/init.lua:260."
When: Occurs once during startup.
OLD
Your phone or window isn't wide enough to display the code box. If it's a phone, try rotating it to landscape mode.
NEW
Your phone or window isn't wide enough to display the code box. If it's a phone, try rotating it to landscape mode.
Notes: Walldir is a helper variable, so it only exists so that the next line can do some calculations with it without having to change the value of node.param2. Therefore should not be a global but a local. I think that this was a simple case of "I meant to change that but I forgot"
Message: "
04:08:20: WARNING: Assignment to undeclared global "bush_branch_type" inside a function at ...../worlds/Pacata/worldmods/plantlife/bushes/init.lua:169."
When: Occurs once after logging into my local server.
OLD
Your phone or window isn't wide enough to display the code box. If it's a phone, try rotating it to landscape mode.
NEW
Your phone or window isn't wide enough to display the code box. If it's a phone, try rotating it to landscape mode.
- Code: Select all
local bush_branch_type = 0 -- This will not remain 0, so should be safe use 0 to initialize the var.
Notes:This was intended to be a local variable, but two lines that were both supposed to initialize the local var we commented out. Instead of uncommenting one of the two, i decided to add my own :-)