Time Nick Message 01:21 octacian #1592 01:21 ShadowBot https://github.com/minetest/minetest/issues/1592 -- ConnectionReceiveThread::Thread(): Assertion '0' failed. 01:22 octacian sofar, nore, paramat, anyone else xD ^ 01:22 octacian * game#1592 01:22 ShadowBot https://github.com/minetest/minetest_game/issues/1592 -- Add workbench and nametag to rename items by octacian 01:22 octacian I keep forgetting to put "game" first lol 01:30 paramat ok 01:34 octacian Also, game#1589 and game#1590 are ready to be merged from my testing. 01:34 ShadowBot https://github.com/minetest/minetest_game/issues/1589 -- Keys: Show owner in description by octacian 01:34 ShadowBot https://github.com/minetest/minetest_game/issues/1590 -- Books: Show title in description by octacian 03:03 sofar octacian: merged one of them 03:03 octacian sofar: ok. Adding screenshot to workbench PR 03:04 sofar I'm gonna be honest with you - I really don't like the workbench as is 03:06 octacian sofar: Added formspec screenshot. 03:06 octacian And what do you not like about it? 03:07 sofar in due time 03:07 sofar but, almost all of it, lol 03:07 octacian Specifics? 03:07 sofar the need for a workbench entirely 03:07 sofar the fact that it's a nodebox 03:07 octacian Ah, I see. 03:07 octacian You'd prefer a model? 03:07 sofar the fact that one of the slots just always contains a nametag 03:08 sofar for mtg, I'd prefer a normal node, just like a chest or a furnace 03:08 octacian IDK, it'd be much harder to create a nice texture for a normal node. 03:09 sofar plenty of workbench textures around that are usable 03:09 octacian Also, the basic reason is that doing it with just a nametag using the method that books use for copying books would require that I register recipes with nametags for every single item in the game 03:09 octacian True. 03:09 octacian Could you give me a link. 03:10 sofar xdecor has one from originally pixelbox I think 03:10 sofar plus there's a few cc-by-sa MC texture packs around that are very usable as well 03:10 octacian OK, I'll take a look. I'd be willing to change it. 03:10 sofar e.g. pixelperfection, isabellaII 03:10 sofar etc. 03:11 octacian Now, IMO having a workbench is the best way to do it. 03:12 octacian As you mentioned, something should be consumed in the renaming process as it should not be "free" to rename an item 03:12 kaeza just a minor thing: should it be named "workbench"? 03:12 kaeza may cause confusion for people who come from MC 03:12 kaeza I'm fine with it in any case 03:13 octacian So, I thought that keeping the nametag and requiring that it be in the workbench almost as "fuel" would be a good way to add cost 03:13 sofar I think you can do it without all the recipes, and do it on the crafting grid 03:13 octacian kaeza: I thought about that. I didn't have any better ideas though, aside from directly using anvil which doesn't even make sense anyways 03:13 octacian sofar: it's true that you can, however, I personally think that the workbench is more, IDK, "elegant" maybe 03:14 sofar MC solves the issue with the anvil, but the anvil has many purposes 03:14 sofar the problem is that we've implemented tool repair without an anvil 03:14 sofar otherwise I'd say, make an anvil instead 03:15 octacian Yes, that's why I introduced the workbench. I was going to open a PR after this allowing you to repair tools only inside the workbench 03:16 kaeza meh. I say leave it as-is 03:16 octacian Though I guess an anvil is a "block with a hard surface on which another object is stuck" so it would kinda make sense to use an anvil 03:20 octacian sofar: anyways, about the slot for nametags, I didn't really have a better idea as to how to require fuel 03:26 sofar seriously though, I think we need to try a little harder to do this on the craftgrid 03:26 octacian We could. 03:27 octacian I'm open to giving it another try. 03:27 octacian I still have the old code for the nametag on_use formspec. 03:27 octacian Though I still think that overall a workbench would be better that would be used both for this and for repairing tools 03:28 octacian Thought it would probably be actually easier to do it in the craftgrid. 03:28 octacian If you've got any ideas as to how to do it, I'll try and implement it. Prob in a new PR. 03:28 octacian That way both options are available for review. 03:33 sofar my idea would be to make a craft recipe for the nametag and then use a group for renameable items 03:33 sofar groups = { renamable = 1 }, 03:33 sofar then the recipe for naming can refer to "group:renamable" 03:34 octacian Yes, you could. However, chances are we'd want to make all items renamable. It'd be better to automatically assume renamable unless group = {renamable = 0} 03:34 sofar and then we can selectively using groups mark items that are safe to rename 03:34 octacian Yes. I think that that should be implemented either way. 03:34 sofar nobody is going to make stone renamable 03:34 sofar keys are the main thing imo 03:35 sofar "front door key" 03:35 octacian Yes, somebody will in a custom map. 03:35 sofar well sure, but in a custom map they can toss a worldmod that marks special items renamable and actually names them too 03:35 octacian With the ability to rename anything, you could do some pretty interesting stuff 03:35 sofar so we don't need to worry about that much 03:35 octacian It's preferable if map creators don't have to use worldmods though 03:36 octacian However, I agree at this point with renaming in the craftgrid. 03:36 sofar why? they're highly convenient and can ship with the map in one location 03:36 octacian It wouldn't make sense unless repair was done in the workbench as well, which should definitely be left for later. 03:37 octacian Yes, but IMO, if I was making a map and specifically wanted to prevent renaming items, there would only be a small amount of items that I'd want to prevent 03:37 sofar well, it's moot anyway since you can just make named items irregardless 03:38 octacian Why, well because when making maps I'd personally prefer not having to restart the world every time I wanted to change something 03:38 octacian Which is why I think we should also have a utilities mod with stuff like a command block 03:38 sofar you're argueing the wrong way 03:38 octacian That's probably true :rotfl: 03:38 sofar map creators want to use special tools to make special items 03:38 sofar and then remove those so the map is "adventure mode" 03:39 octacian True. Still though, I think stuff should be automatically renamable. 03:39 sofar can't see why we couldn't have an admin-only command 03:40 octacian What do you mean? to do what? rename items? 03:40 octacian Oh. I understand. 03:40 octacian I don't see an issue with that either, but it just doesn't make since to not be able to rename items. 03:41 sofar well, here's what I think 03:41 sofar I think we should have a relatively small basic set of items that can always be renamed 03:41 sofar things like craftitems and some tools 03:41 sofar but probably not nodes 03:41 sofar leave those not nameable for now 03:42 sofar it's easy enough to generate named nodes irregardless 03:42 sofar any 2 lines of lua can make them 03:42 sofar books are already solved 03:42 sofar keys are badly needed 03:43 sofar but I could care less if people named a mushroom or a loaf of breakd 03:43 sofar bread* 03:43 sofar but weapons, no 03:43 sofar I don't see why we would allow so by default 03:43 octacian Yes, I agree. However I still see no reason why we should prevent people from naming nodes or weapons. 03:43 sofar it's only going to conflict with mods 03:43 * sofar ponders 03:43 sofar well, maybe not 03:44 sofar do we have a group:tool ? 03:44 octacian I'm not sure. 03:44 octacian No. 03:44 octacian we do have minetest.registered_tools don't we though? 03:44 octacian Yes, we do. 03:45 sofar the craft stuff works for us if we use groups 03:45 octacian Could use a simple loop to iterate through and add renamable though I still don't like preventing renaming some stuff. 03:45 octacian But again, that's true, it only works if we use groups. 03:45 sofar I'm sure we can compromise on that 03:46 octacian Meaning that the only way to allow renaming everything would be to iterate through registered_items adding renamable = 1 to the groups if it wasn't already equal to 0 03:46 sofar we'd just add it to every node reg 03:47 sofar override_item makes no sense inside default 03:47 octacian True. 03:47 sofar sadly, some parts do :/ 03:47 octacian But what about other mods outside MTG. 03:47 octacian For example, the books in homedecor. 03:48 octacian Defaulting to renamable would allow players using homedecor to take advantage of the new features and rename books even before homedecor is updated 03:48 octacian (the same goes for other mods that introduce like items) 03:48 sofar ah see 03:48 sofar I don't think we should allow books to be renamed 03:49 sofar that'd be a copyright violation 03:49 octacian rly.. :rotfl: 03:49 sofar besuides, the persson writing can already give a title 03:49 sofar and even change it 03:51 octacian True. But they can't change the description tooltip 03:51 octacian It would be appreciated for organization, just as the new functionality in default will be. 03:52 octacian This, IMO, leaves two options: implement using craftgrid and groups and iterate through registered_items setting groups.renamable = 1 03:52 octacian or use workbench approach 03:52 octacian ^^ only if groups.renamable ~= 0 ofc 03:52 octacian That's what I see anyway. 03:53 sofar separate mechanism from nodes/items 03:53 octacian What do you mean? 03:54 octacian registered_items contains a table of all nodes, craftitems, and tools 03:54 octacian * list, not table 03:54 sofar iterating and modifying every registration == bad 03:54 octacian Yes, I agree. 03:54 sofar so, dont 03:54 octacian I just don't see a better way to do it aside from limiting the items which I personally do not like either. 03:54 sofar just implement naming based on groups 03:55 sofar then we can argue the second part 03:55 octacian Will do. I'll open a second PR tomorrow. 03:55 sofar ppl will disagree with me as well, lol 03:55 octacian Leave them both open ofc though. 03:55 octacian somebody will ALWAYS disagree with you xD 03:56 paramat sofar are you ok with game#1587 ? 03:56 ShadowBot https://github.com/minetest/minetest_game/issues/1587 -- Biomes: New surface node for rainforest by paramat 03:57 sofar paramat: do it, :+1: 03:59 octacian I've gtg though. https://github.com/minetest/minetest_game/pull/1592#issuecomment-282935503 04:11 paramat ok 04:26 paramat hm 'litter' seems to be a more suitable term than 'debris' 04:28 paramat there's 'detritus' and 'duff' but those are less clear language 04:28 octacian paramat: that font is just installed locally 04:28 octacian I didn't hardcode it, in fact, is that even possible? 04:29 octacian https://github.com/minetest/minetest_game/pull/1592#issuecomment-282939018 04:30 paramat ok 04:30 paramat no problem 04:31 paramat i'm going to use 'rainforest litter'. i get anxious when naming nodes, it has to be right otherwise we have to use dreaded aliases to support old worlds 04:32 octacian paramat: BTW, do you prefer the crafting grid approach to renaming items, or the workbench (which easily allows renaming anything but nametags)? 04:45 paramat i'll think on that 04:46 octacian OK 05:30 paramat merging game#1587 in a moment 05:30 ShadowBot https://github.com/minetest/minetest_game/issues/1587 -- Biomes: New surface node for rainforest by paramat 05:32 paramat merging .. 05:35 paramat .. complete 05:56 lhofhansl How can I determine whether a piece of code (say in Map or MapBlock) is run on the server or the client? #ifdef SERVER does not work when the client is starting the server. 05:57 lhofhansl (and I do not want to use virtual methods between ClientMap and ServerMap because this is hot path and the dispatch might slow things) 06:47 lhofhansl NM. Found another way. 07:14 lhofhansl #5320 07:14 ShadowBot https://github.com/minetest/minetest/issues/5320 -- [For Testing] Allow server side occlusion culling. by lhofhansl 15:05 octacian sofar: I believe the books PR is ready to merge (rubenwardy gave it a thumbs up) 17:59 Thomas-S Can someone please review #5302? This is very much needed, I think this would solve the mining drill problem with the technic mod, too. 17:59 ShadowBot https://github.com/minetest/minetest/issues/5302 -- Store legacy metadata separate from new item meta data by rubenwardy 18:20 Krock Thomas-S, would be more helpful to know why numberZero downvoted on it 18:21 Krock sofar, FYI. I'm keen to fix the issues in #4304 and open a new pull for it - but I guess this won't happen before this weekend 18:21 ShadowBot https://github.com/minetest/minetest/issues/4304 -- Add fading sounds by Bremaweb 18:22 sofar Krock: neat, I can wait 18:23 Krock ^^ 18:54 VanessaE paramat, sofar: will param2 or hardware coloring ever be applied to default wool? (need to know for a future plan) 18:54 VanessaE s/ or /\// 18:54 sofar my position is 18:54 sofar it's contingent on getting coloring of itemstacks 18:55 sofar if we can get itemstack meta coloring 18:55 sofar then we can have colored wool 18:55 paramat indeed 18:55 sofar and even handle digging and placing with a simple API 18:55 VanessaE I ask because I want to pipe wool through Unified Dyes, given the extensive color range it has. 18:55 VanessaE UD has those, btw. 18:56 VanessaE I tried to cover everything that could be wanted with this sorta thing (colored itemstacks aside) 18:56 sofar without colored itemstacks, we can't have red wool in the inventory next to blue wool 18:57 paramat but also some are concerned about how this limits texture packs ability to choose their own exact hues and use individual textures instead of colourising a single texture 18:57 sofar and I absolutely dislike the "drop item+dye" mantra 18:57 sofar paramat: they can replace the colormap too 18:57 VanessaE sofar: only sane way to do it without colored itemstacks 18:57 sofar paramat: so that solves that problem entirely 18:57 paramat oh yeah oops 18:58 sofar VanessaE: I can't politely say how bad I find that method 18:58 sofar so I would have never done it 18:58 sofar colored itemstacks, problem solved 19:00 VanessaE sofar: ok so how would you have done it? 19:00 VanessaE assume no colored itemstacks 19:00 sofar not 19:00 sofar that's what I said 19:00 sofar so I would have never done it 19:00 VanessaE nono I mean, 19:01 VanessaE how would you have done colorized ... anything ... without managing the dyes etc? 19:01 sofar I would have *not* done it, and instead solve the colored itemstack problem first 19:01 VanessaE I see. 19:01 paramat ^ indeed 19:01 sofar now you have everyone learning a weird new way 19:02 sofar I just want a red wool block in my inventory when i dig a red wool block 19:02 sofar I mean, I didn't see the problem beforehand either :) 19:02 VanessaE and then I presume colored itemstacks, you would import the param2 value to the itemstack, expl... yes, that. 19:02 sofar the only exception is landscape coloring 19:02 sofar that doesn't need colored itemstacks imho 19:02 VanessaE I'm fine with that idea, and as soon as it becomes possible, I'll implement it. 19:03 VanessaE or rather, use it 19:03 sofar afk a bit 19:03 paramat a minor issue is that colourising a single base texture is less flexible in appearence than tuning individual textures in Gimp 19:03 VanessaE paramat: true enough, and default wool would have some difficulty there because yellow wool is more of a yellow-with-gold-highlights 19:04 VanessaE the others, I was able to get reasonably close (look at castles++ mod's tapestries for an example) 19:04 paramat and there are only 14 wools, so i don't see much need to use hardware colouring 19:04 VanessaE 14 wools, hardware coloring, suddenly you have the equivalent of 256. 19:04 VanessaE since they'd all be reduced to a single node. 19:04 VanessaE (node def that is) 19:05 paramat you mean add new colours? 19:06 VanessaE well you could, but "new colors" comes with the territory 19:06 VanessaE no one says the extra colors have to be in the inventory 19:06 VanessaE I do that with UD for example. the default 15 dyes are there, but the 240 others added by UD are not. 19:06 paramat ah the palette .. 19:06 VanessaE yes. 19:06 VanessaE it's just a side effect, one you'd be advised to take advantage of ;) 19:10 cheapie OK, so param2-ize the wool, then make it work with colored itemstacks as well once that's possible. I don't see the issue here. 19:11 VanessaE point is, I'm going to write a mod, or maybe cheapie will, that's gonna param2-ize the wool, one way or another. 19:11 VanessaE and it'll run through UD 19:11 cheapie VanessaE: I already failed miserably at that. You do it :P 19:11 VanessaE the whole point of that mod is a central API and way of doing things. one change there and everything that uses it follows. 19:11 VanessaE cheapie: no worries, I'll figure it out. 19:12 juhdanad Hi all! 19:13 VanessaE hey 19:13 juhdanad Finally I could connect... 19:15 juhdanad So, I don't know how itemstack metadata works. Is it possible to return an itemstack with a field 'param2' set when a node is dug? 19:15 VanessaE no idea, though if there's meta then I don't see why not 19:16 paramat not sure 19:16 juhdanad And if you have two items with the same metadata values, can you merge them? 19:19 Krock I'm not aware of such a check 19:19 paramat Krock if core devs agree to this, would you be interested in being a mtgame dev? 19:20 Krock paramat, well, I'm not sure. Either I'd want to join both or none 19:20 paramat ah it could be both 19:20 paramat normally it would be actually 19:20 Krock just because Zeno also already talked to me a while back 19:21 Krock I'm not sure what's to expect from a dev and the instructions on the wiki are quite vague 19:22 paramat your blood 19:22 VanessaE heh 19:22 Krock doing a blood brother ritual xD 19:23 paramat there's no obligation to work when you don't feel like it, many devs disappear often. it's mostly code review and approvals 19:25 Krock ah, good to know that. Well, for me this doesn't have any priority at all but if you think there should be a voting then I wouldn't say no :) 19:26 paramat we'll see what devs think 19:26 Krock juhdanad, AFAIK the metadata is cleared when two items are merged, thus all of these specific items have a stack max of 1 or use wear 19:29 juhdanad Then metadata is not a good solution for storing colors... 19:29 VanessaE well my vote means exactly bupkis but I'm okay with Krock being added as a dev. 19:29 Krock VanessaE, but you're also quite long into this game and surely have got plenty of knowledge ;) 19:30 paramat i would be +1 for engine and game dev 19:31 sofar juhdanad: itemstacks with different meta don't merge 19:31 sofar juhdanad: yes, you can (in lua already) make an itemstack contain a param2 when digging that reflects the node param2 19:31 sofar as a matter of fact, we could retain the digging param2 for certain nodes 19:31 sofar but, easy enough to do in lua 19:32 juhdanad Great! Then it is easy to render the item according to its metadata (I think so...) 19:32 VanessaE fwiw, UD stores the color in the node's metadata as well as its param2 19:32 sofar but remember, items ~= nodes 19:33 sofar craftitem properties do not have tiles 19:33 sofar instead, inventory_image 19:33 VanessaE (easier to reference the stored dye name then to look up the color from the param2 value) 19:33 sofar you may have to add a color = property 19:33 sofar as well, to make it symmetric with nodes 19:34 sofar VanessaE: that's a mod choice, though 19:34 VanessaE yes. 19:34 VanessaE just pointing it out for reference. 19:34 sofar if we're reserving itemstack meta names we should be consistent 19:34 sofar description is now already used officially 19:34 juhdanad No, param2 is needed. This ensures that mods can replace palettes. 19:34 sofar we can do param2, it seems the best 19:34 sofar but are we going to do paramtype2 for craftitems? 19:34 sofar it would be most logical 19:36 sofar also, you need to provide a palette 19:36 juhdanad There would be two fields, I suppose: 'param2', which is effective for nodes, and 'color', which overrides everything. 19:37 sofar so a palette = param is needed 19:37 juhdanad That way you can also rotate facedir nodes in the inventory. 19:37 sofar yup 19:37 sofar well nodes should be simpler 19:37 sofar since they already have palette, paramtype2 properties 19:38 juhdanad The problem is that palettes are only loaded if a node requests them. 19:42 sofar so that would have to be added for craftitems then 19:42 VanessaE all I care about is that whatever is in the inventory, it looks exactly the same as it does in the world (drops/forced inventory colors aside). :) 19:44 juhdanad Well, I think I can do that, but then item definitions will also have palettes. 19:46 juhdanad BTW when do you plan to release the next version of Minetest? 19:46 paramat perhaps may-june 19:47 juhdanad Then I'll have a plenty of time. 19:51 juhdanad paramat: did you notice that your 'chambers' mod does not always manage to generate the full chamber if you fly around fast? 19:52 sofar juhdanad: that's acceptable 19:53 sofar juhdanad: wanted, even, I think 19:54 juhdanad It is because the map generator copies some blocks, then the mod generates the chamber, finally the mapgen blits back the unchanged data. 19:54 juhdanad This also causes the 'moretrees' mod's partially generated trees. 20:03 diemartin #1596 20:03 ShadowBot https://github.com/minetest/minetest/issues/1596 -- Wrong letters shown in Czech translation (V0.4.8)- Mageia, GNU/Linux 20:03 diemartin game#1596 20:03 ShadowBot https://github.com/minetest/minetest_game/issues/1596 -- Add desert/silver sandstone-related blocks. by kaeza 20:06 Fixer i really waiting for coloured baked clay in default, i miss it so much (very popular building block in minecraft btw, wool texture is not good for buildings and nature materials are limiting) 20:06 VanessaE unifiedbricks has that. 20:08 VanessaE (it isn't baked, but it is clay blocks and looks good) 20:09 Fixer hardened/backed/whatever - texture is important 20:10 VanessaE there's a mod for that, too 20:10 VanessaE (I should fork it and convert it to use UD) 20:11 VanessaE https://forum.minetest.net/viewtopic.php?id=8232 or https://forum.minetest.net/viewtopic.php?id=8890 20:18 paramat my 'catacomb' mod? 20:18 paramat that mod doesn't work in 'on generated though' 20:20 paramat oh you mean when a nearby mapchunk is generated 20:21 paramat hm i think i noticed a few chambers missing a wall 20:37 paramat well nevermind seems too difficult and complex to fix 22:05 octacian sofar: FYI, I think the books PR is ready to merge, rubenwardy approved. 23:39 diemartin paramat, should I also change the "normal" sandstonebrick? 23:40 paramat no, that would need an alias and we want to avoid those 23:41 paramat trees have inconsistent naming too for old and new tree types 23:44 diemartin I can use a simple alias, duplicated code, or an unreadable mess 23:45 paramat hm yes looks like old sandstone would have to be separate from your register functions 23:45 paramat well, for crafting at least 23:46 diemartin no, because I also register the nodes, and I must make a special exception for the normal one 23:46 paramat for nodedefs you can expand the prefix to become the whole name? 23:47 diemartin I can do that. don't have to like it though :( 23:48 diemartin uh, actually, you'd need to pass three names or at least just the "stone" name which will quickly get a mess 23:48 diemartin what's bad about aliases? 23:50 paramat yes i can see this would affect your clean code, however i think some code duplication is fine, i would suggest splitting off the old sandstone and using your code for the 2 new ones 23:51 paramat well, i personally don't like them unless essential, and here we are doing it just for name consistency (non essential) and your registering method (nice but non essential) 23:53 diemartin so, just opinion, no technical reasons? 23:53 paramat for just a few new nodes i would have been fine with duplicated code 23:53 paramat well technical too yes 23:55 paramat duplicated code for nodes is actually a little clearer to read and understand 23:55 paramat sorry btw 23:56 diemartin I'm just trying to get some comments before I change it and somebody complains "why not create a function for this" or something :) 23:58 diemartin I disagree. I find it perfectly readable. also, duplicating code makes it harder to change things (like groups or whatever) if needed in the future as you need to change it N times instead of just once