| Time |
Nick |
Message |
| 00:21 |
rubenwardy |
It's limited to 8 on Android |
| 00:31 |
VanessaE |
sure, on android it makes sense, but I have for example room for 27 slots on my screen (if Areas mod's HUD doesn't overlap). |
| 01:05 |
|
ANAND joined #minetest-dev |
| 03:01 |
|
Ruslan1 joined #minetest-dev |
| 04:00 |
|
reductum joined #minetest-dev |
| 06:27 |
|
calcul0n joined #minetest-dev |
| 09:15 |
|
calcul0n joined #minetest-dev |
| 09:54 |
|
YuGiOhJCJ joined #minetest-dev |
| 10:06 |
|
calcul0n joined #minetest-dev |
| 10:16 |
|
Beton joined #minetest-dev |
| 11:23 |
|
p_gimeno joined #minetest-dev |
| 11:38 |
|
calcul0n joined #minetest-dev |
| 11:54 |
|
Fixer joined #minetest-dev |
| 12:36 |
|
Fixer joined #minetest-dev |
| 13:58 |
|
hecks joined #minetest-dev |
| 14:07 |
|
Beton_ joined #minetest-dev |
| 14:50 |
|
pgimeno_ joined #minetest-dev |
| 14:58 |
|
p_gimeno joined #minetest-dev |
| 15:48 |
|
Fixer joined #minetest-dev |
| 16:47 |
|
Krock joined #minetest-dev |
| 16:56 |
|
Gael-de-Sailly joined #minetest-dev |
| 17:02 |
|
paramat joined #minetest-dev |
| 17:19 |
Krock |
Hijiri: they'll have to ship compatibility code (if possible) on their own or wait for the next release |
| 17:20 |
Krock |
in the past, when mods depended on dev features, then the mod topic stated "needs a recent -dev build of 0.4.x or later" |
| 17:20 |
Krock |
although, it gets more complicated when old code stops working |
| 18:02 |
|
hecks joined #minetest-dev |
| 18:22 |
Hijiri |
I was thinking of something like dimensions, where basically anywhere a mod was using positions becomes suspect |
| 18:23 |
Hijiri |
though conceivably you could first implement everything except the actual modding interface for creating multiple dimensions (just only have one actual dimension around) |
| 18:24 |
Hijiri |
another issue with dimensions specifically I guess is that there is some disagreement on how it should be implemented |
| 18:25 |
Hijiri |
though maybe that is true about any major feature |
| 18:26 |
paramat |
dimensions don't need to break mods. and changing co-ords to 4D is silly |
| 18:27 |
paramat |
most just want different realms, not true 4D space |
| 18:30 |
Hijiri |
I don't think a non-breaking API can be made |
| 18:30 |
Hijiri |
for example, set node on a position won't go to the right dimension if it's not specified |
| 18:30 |
Hijiri |
like if a mod wants to place dirt next to a player or something, if that player's not in some default dimension then the dirt will be placed in the wrong place |
| 18:31 |
Hijiri |
forgot to mention, the mod used get_position on the player in my mind |
| 18:32 |
sfan5 |
there's a simple but stupid way to have non-breaking api, a separate lua state for each realm |
| 18:32 |
sfan5 |
that brings some problems with it however |
| 18:32 |
Hijiri |
that is true |
| 18:32 |
Hijiri |
but I want a shared lua state |
| 18:32 |
Hijiri |
keeping shared state consistent could be a lot of trouble for modders |
| 18:32 |
Hijiri |
consistent across multiple lua states I mean |
| 18:38 |
hecks |
are we discussing parallel universes? |
| 18:38 |
hecks |
because a 4D player and node coord *is* in fact the easiest way to implement it |
| 18:39 |
hecks |
it's totally backwards compatible as long as you don't try to do arithmetic on "w" |
| 18:39 |
Hijiri |
Why not just have both a spatial coordinate and a dimension inside some non-position object |
| 18:40 |
Hijiri |
then you can handle the coordinate part as just a 3D coordinate |
| 18:41 |
hecks |
my point is, everything would work as it did before |
| 18:41 |
Hijiri |
define "everything" and "work" |
| 18:41 |
hecks |
while having a separate "realm" id attached to everything could cause issues |
| 18:41 |
Hijiri |
because we would still have the dirt problem I mentioned |
| 18:42 |
hecks |
then consider this |
| 18:42 |
hecks |
mod wants to place dirt next to player |
| 18:42 |
hecks |
it gets player position, adds { x = 1, y = 0, z = 0 } to the position |
| 18:42 |
paramat |
the demand is mostly only because of minecraft, and note that the only reason minecraft has dimensions is because it doesn't have the vertical height to put the nether deep down and the aether or 'end' high up |
| 18:42 |
hecks |
unknown to the mod, the player's position has { w = 1 } as well |
| 18:42 |
hecks |
the dirt's new position is, in fact { foo, bar, baz, 1 } |
| 18:43 |
hecks |
and thus it *is* placed in the right dimension |
| 18:43 |
Hijiri |
suppose that the mod didn't use the addition function, and instead they did { x = v.x, y = v.y + 1, z = v.z } |
| 18:43 |
paramat |
so ironically, dimensions are a result of minecraft lacking the feature MT has |
| 18:43 |
Hijiri |
Then it's still broken |
| 18:43 |
Hijiri |
paramat: That's true, for the original use case |
| 18:43 |
Hijiri |
however minecraft mods have expanded its use so that they can't be implemented in minetest |
| 18:43 |
Hijiri |
If we're comparing minetest to minecraft we need to compare it to minecraft forge as a modding platform, since minetest is intending to be modded |
| 18:44 |
Hijiri |
obvious example is mystcraft, where you can travel to other worlds using books (like in myst) |
| 18:44 |
Hijiri |
I think dimensional doors mod also allows players to create dimensions? |
| 18:45 |
Hijiri |
forgot an important point, in mystcraft each new book goes to a new procedurally-generated dimension |
| 18:45 |
hecks |
if float precision wasn't an issue, you could just encode dimensions as any position outside of the mapgen boundary |
| 18:46 |
Hijiri |
yes |
| 18:46 |
hecks |
and sure enough, you can use horizontal slices of minetest's obscene amount of unused space if you're stubborn |
| 18:46 |
hecks |
I might just do that, myself |
| 18:46 |
Hijiri |
actually if in lua you could represent a larger integer then I would support just making minetest have 32 bit y coordinate |
| 18:46 |
Hijiri |
and then use stacked dimensions |
| 18:47 |
hecks |
lua can represent some pretty huge numbers, lua's type is double unless it's compiled to use something else |
| 18:47 |
Hijiri |
if the y coordinate was 32 bits then the hashed position would be 64 bits, you can't fit a 64 bit integer in a double |
| 18:47 |
paramat |
dimensions is low priority, the work needed and disruption makes it not worth doing in the short/medium term. we can barely cope with the essentials currently, in fact, we can't even do that. however it's cool to discuss it :) |
| 18:47 |
Hijiri |
what essentials are we struggling with |
| 18:48 |
Hijiri |
CSM comes to mind, but not sure if it's essential |
| 18:48 |
hecks |
I think it's a non-issue given how much space in +Y you have to fit your dimensions in |
| 18:48 |
hecks |
{ foo, 10000+, bar } will work just fine |
| 18:48 |
Hijiri |
currently there's only space for 64 dimensions of height 1024 |
| 18:48 |
hecks |
floatlands are at 1200ish and that's really, really far |
| 18:49 |
Hijiri |
what if 64 players make a myst book |
| 18:49 |
Shara |
Has there been any practical case of someone needing more than 64? |
| 18:49 |
Hijiri |
then you'll need 65 dimensions (1 for the main world, assuming there is one) |
| 18:49 |
Hijiri |
Shara: In Minetest? |
| 18:49 |
hecks |
if you want an infinite number of full, persistent dimensions each having a whole map, then oh well, check out and start coding |
| 18:50 |
Hijiri |
I gave some examples of mods not from Minetest, that would probably still be fun and usable if they were done in Minetest |
| 18:50 |
Shara |
I would assume that MT is what we're talking about :P |
| 18:50 |
hecks |
but for something like temporary instances for players to grind in, we have enough space |
| 18:50 |
hecks |
I think you're even allowed to create mapblocks outside of the 32k radius, is that correct? |
| 18:50 |
Shara |
And I can't imagine a player needing the full x/z distance for themselves? |
| 18:51 |
Hijiri |
even if you don't need the full x/z distance, it's better if you're not stopped after walking a certain distance in one direction |
| 18:51 |
Hijiri |
in mystcraft the additional worlds are conceptually the same size as the main world |
| 18:52 |
hecks |
minetest will eventually stop you, even on the main map... but the boundary takes 25 minutes to reach at flying speed |
| 18:52 |
Hijiri |
you're not going to generate most of the world but you are able to walk as far as you would be in the normal world |
| 18:52 |
paramat |
we might eventually move to s32 co-ords, which would be good, then you have more vertical space for stacked dimensions than you could ever need |
| 18:52 |
Hijiri |
I agree with that |
| 18:52 |
Hijiri |
that would be great |
| 18:53 |
Hijiri |
Only thing really blocking that is the numeric position hash exposed to lua, but maybe removing that would be less disruptive than changing everything to do with positions |
| 18:53 |
hecks |
I just wonder what do you need so many dimensions for, other than "meinkraft has it" |
| 18:53 |
Shara |
How maps are saved and the space needed feels like an issue that would need to be sorted out first to make it practical either way |
| 18:53 |
Hijiri |
hecks: if someone wanted to implement mystcraft in minetest |
| 18:54 |
paramat |
anyway, the actual concept and desire here is new realms, not 4D space. new realms can be done without breaking API |
| 18:54 |
Hijiri |
Shara: A block in another dimension doesn't have to take any more space that a block in the default dimension |
| 18:54 |
paramat |
also VAEs are in the TODO and could possibly be used for smaller dimensions |
| 18:54 |
hecks |
I hardly know what mystcraft is, I just wonder, what for... |
| 18:55 |
hecks |
MT worlds are big enough that you can never meet another player if you want to |
| 18:55 |
Hijiri |
it's a mod where you craft books, each of which takes you two a unique dimension |
| 18:55 |
hecks |
yeah, and... what for? |
| 18:55 |
Hijiri |
and you can make custom books to find a world with specific properties if you want |
| 18:55 |
Shara |
Hijiri: I run a server where I limit that map at 7k from 0 in all directions.. and it easily reached 20GB in size. If you feel the current space is not enough, how large will maps get regardless of number of dimensions? |
| 18:55 |
hecks |
if it's for infinite random grind, there are ways to do that with more modest requirements |
| 18:55 |
Hijiri |
hecks: I don't know, why play minetest? |
| 18:55 |
hecks |
to play with blocks, then get bored, and spend the rest of your week programming in lua |
| 18:55 |
hecks |
why else |
| 18:56 |
Hijiri |
Shara: infinite dimensions doesn't mean players will explore infinitely many more blocks |
| 18:56 |
paramat |
i think a reachable edge is more interesting than no practically reachable edge. the edge is interesting to visit |
| 18:56 |
Shara |
Personally I don't feel like managing 100GB+ maps just so players can each get their own personal dimension to walk along until they hit the map edge |
| 18:57 |
hecks |
I play minetest for the moe anime girls https://a.uguu.se/GTDUpp3TM0jJ_ffgsfds.png |
| 18:57 |
Hijiri |
I guess don't install mysttest on a large server then |
| 18:58 |
Shara |
The number of players who go wandering is pretty high. When they are all in one dimension, you hit a point where the map was already generated, so the size increase of the map over time starts to drop. If those wanderers all had their own dimension, you'd not get that drop... mapsize growth could easily be crazy |
| 18:58 |
Hijiri |
that's also assuming that every player is going to use the mod |
| 18:58 |
hecks |
if you want your own dimension... play sp |
| 18:58 |
Shara |
Well, not using it on a busy server invalidates your point about needing 64+ dimensions then |
| 18:58 |
Hijiri |
One player could create more than one dimension |
| 18:59 |
Hijiri |
In that case you don't have the same problem as a bunch of players creating their own dimension, because that player is probably not going to walk to the edge of all their created dimensions |
| 18:59 |
Shara |
One player could also make multiple worlds. Sorry - it's just my opinion, but it seems really low priority compared to other things |
| 19:00 |
Hijiri |
I just said that a player could make multiple worlds, are you responding to that? |
| 19:00 |
Hijiri |
I don't understand your message (the first part) |
| 19:00 |
Shara |
Well, the end part is what matters |
| 19:01 |
Hijiri |
Ok, I can understand that |
| 19:01 |
hecks |
by the way, this is a topic more for #minetest |
| 19:02 |
hecks |
dev is where you'd discuss things you're actually doing |
| 19:02 |
Hijiri |
probably it would fit there better |
| 19:02 |
Hijiri |
this began as a question about development logistics |
| 19:03 |
paramat |
anyway, see the issue, it's all very non-trivial. for 'dimensions' i would only support dividing up our vertical space, which can be done now with a new core mapgen and no complex changes |
| 19:04 |
paramat |
yeah another channel is better :) |
| 19:04 |
Hijiri |
probably should have each mod declare a max number of dimensions it will create too, to avoid overbooking the dimensions |
| 19:12 |
Krock |
merging #2221 |
| 19:12 |
ShadowBot |
https://github.com/minetest/minetest/issues/2221 -- minetest.get_craft_recipe returns different table than lua_api claims |
| 19:12 |
Krock |
game#2221 |
| 19:12 |
ShadowBot |
https://github.com/minetest/minetest_game/issues/2221 -- map: Use wood group crafting recipe by SmallJoker |
| 19:13 |
Krock |
done |
| 19:14 |
|
behalebabo joined #minetest-dev |
| 19:20 |
|
Lymkwi joined #minetest-dev |
| 19:37 |
|
Taoki joined #minetest-dev |
| 20:02 |
paramat |
#7651 test fine in inventory and using unified inventory, no fuzziness that krock gets |
| 20:02 |
ShadowBot |
https://github.com/minetest/minetest/issues/7651 -- Simpler software inventorycube by numberZero |
| 20:15 |
paramat |
reviews needed |
| 20:26 |
|
mrchiantos joined #minetest-dev |
| 20:30 |
|
Taoki joined #minetest-dev |
| 21:22 |
|
paramat joined #minetest-dev |
| 21:35 |
|
bigfoot547 joined #minetest-dev |
| 21:35 |
|
bigfoot547 joined #minetest-dev |
| 21:45 |
sfan5 |
paramat: was the distortion with that fixed yet? |
| 22:02 |
paramat |
oh .. |
| 22:02 |
paramat |
well |
| 22:02 |
paramat |
krock sees distortion when using unified inventory, but i don't |
| 22:02 |
paramat |
sfan5 ^ |
| 22:03 |
paramat |
that is, the PR as it is, not the hgher res alternative |
| 22:03 |
sfan5 |
hm |
| 22:03 |
sfan5 |
will have to take another look |
| 22:04 |
paramat |
so i don't think the high res version is needed |
| 22:26 |
|
ewong4 joined #minetest-dev |
| 22:26 |
ewong4 |
Αfter thе aϲq∪іsitiοn by Priᴠɑte Internet Acceѕs, Frᥱеnode іs ᥒow being usеⅾ tο p∪ѕһ IСO sсaⅿs httⲣs:⧸/ᴡww．coindesk.cοⅿ/hanⅾshɑkе－revᥱalеd╴ⅴcs-bɑck-plɑᥒ-to-gⅰve-аᴡaỿ−100˗mⅰlⅼⅰഠᥒ－ⅰᥒ-crуptο/ |
| 22:27 |
ewong4 |
"Αⅼl tolԁ, Hаndѕһаkе ɑіmѕ tο gⅰve $ᒿ50 worth of іts tokenѕ to ﹡еaϲh＊ ᥙser of tһе websiteѕ tһe comрany haѕ рɑrtᥒᥱrѕhⅰрs wⅰtһ – GitHubᛧ thе Ⲣ2P ᖴoundɑtіon and *ᖴREENⲞDᎬ*, a ϲhɑt cһanᥒеⅼ fഠr pᥱer˗to-рeer рroјeсts. Ꭺs such, ... |
| 22:27 |
ewong4 |
ԁе⋁eⅼoрᥱrs ᴡho һɑvе existinɡ ɑϲcounts οᥒ eɑϲh coulԁ receive up tο $750 wortһ of Handshake tokᥱᥒѕ." |
| 22:27 |
ewong4 |
Hɑᥒԁshake crỿptⲟⅽurrency sϲam is oреrɑteⅾ bу Αᥒdrew Ⅼее (276⎼88-05ℨᏮ), tһe fraᥙdster ⅰᥒ chiᥱf ɑt Рrіvatе Iᥒternеt Aⅽcesѕ wһicһ ᥒow owᥒѕ Frеenoԁе |
| 22:27 |
ewong4 |
ᖴreeᥒοdе ⅰs rᥱgiѕtered аs a ＂privɑte comⲣany limiteⅾ by ɡᥙarantᥱᥱ ᴡitһο∪t sharе ⅽapital" performing "ɑϲtіvitіeѕ of ഠtһer ⅿеmbеrshiр оrɡɑᥒiѕɑtiⲟns not ᥱlsewһere clаѕsified"ˏ wіth Ꮯhrіѕtel and Аᥒⅾrеᴡ ᒪeе ﹙PIA's fo∪ᥒdеr) ɑѕ ⲟffісerѕ, aᥒԁ Aᥒⅾrew Ꮮеe hаving thе ⅿаjοritỿ of votinɡ riɡhts |
| 22:27 |
sfan5 |
lmao |
| 22:37 |
|
YuGiOhJCJ joined #minetest-dev |
| 22:48 |
paramat |
oh man |
| 22:53 |
|
longerstaff13_ joined #minetest-dev |
| 22:54 |
hecks |
is that thing using emoji to scramble letters? |
| 22:54 |
hecks |
cause I see a lot of dead chars |
| 22:54 |
sfan5 |
you mean s/emoji/unicode/, yes |
| 23:04 |
sfan5 |
paramat: the pr is certainly okay in the normal inventory, haven't tested unified inv yet |
| 23:05 |
sfan5 |
yep looks "fuzzy" in unified inv too for me |
| 23:08 |
|
ssieb joined #minetest-dev |
| 23:09 |
|
Ruslan1 joined #minetest-dev |
| 23:10 |
paramat |
odd |
| 23:10 |
paramat |
https://github.com/minetest/minetest.github.io/pull/142 updated |
| 23:16 |
|
ircSparky joined #minetest-dev |
| 23:24 |
paramat |
however i feel we should investigate why image buttons are fuzzy |
| 23:25 |
hecks |
continuing the huge model problem; compression would help a little... https://a.uguu.se/FiisAMKx0crU_filesize.png |
