Time Nick Message 01:37 SitDog I want to setup minetest development on my mac. Anything I should know before I start? 01:41 hmmmm n0p3 02:30 SitDog i have been in dll hell on my PC getting setup. i think the mac is probably a better choice for me and I don't have to worry about juggling DLLs for my other projects. 10:38 Polaque Hello 10:38 jin_xi hi 10:38 Polaque Bug with disappear door if current select terrain is not empty is detect ? 17:14 hmmmm hrmm 17:15 hmmmm other minecraft clones have some pretty advanced server-side features that would be really nice 17:15 hmmmm lattice network 17:15 Calinou [citation needed] 17:15 hmmmm for loadbalancing 17:16 hmmmm eventually, maybe in the far future, i'd like to see something like that so minetest servers can really scale 17:17 hmmmm i mean let's talk about what actually needs to be done 17:17 hmmmm in the server-to-server protocol, you'd have broadcast not only to clients, but to other servers as well, block set, player position change, some other things 17:17 hmmmm actually that's not too bad 17:18 hmmmm they'd have to transfer newly generated blocks as well 17:19 hmmmm but it might be faster to simply not, and in the event that a block modification is broadcast to a server where a nonexisting block is modified, it'd lazily generate it and then apply the change 17:25 celeron55 i'm wondering what we should take as the base for minetest 0.5 17:25 celeron55 what can we afford to rewrite and what is worthy of keeping? 17:26 celeron55 some interested and capable individual should start trying to figure out 17:27 hmmmm the more i think about it, the more i want to dump the database entirely and just have a huge memory mapped file we do an indirect indexing from 17:27 hmmmm and store block metadata separately 17:28 hmmmm i'm going to be able to do so much more now that school's almost over 17:28 hmmmm farmesh i think is not worthy of keeping at all 17:28 hmmmm sorry 17:28 celeron55 i will probably have time to do some more massive things after next summer 17:29 kotolegokot https://github.com/celeron55/minetest/issues/323 17:29 celeron55 depending on random things 17:30 hmmmm hey 17:30 celeron55 actually, not probably, byt maybe 17:30 hmmmm did you ever consider only storing the difference between blocks 17:31 hmmmm and having the client have a mapgen too so it generates the map and then applies the changes sent to it by the server 17:31 hmmmm imagine how much less you'd have to store 17:31 celeron55 i have considered not saving map generation results unless they have been modified, but it hasn't been practical with the generation speeds 17:31 hmmmm right, just a little bit sooner and it might be practical 17:32 hmmmm really, the new perlin noise stuff is a game changer 17:33 hmmmm i decided against the map preview feature since my reference perlin noise implementation can already do the same 17:33 hmmmm in order to test different parameters, what i did was make a small GTK application with the perlin noise stuff to test/debug and stuff, and also the gist of what a map would look like 17:33 celeron55 for 0.5, one thing we'll have to have is flexible creation of new "voxel spaces" for the purposes of... many things 17:33 hmmmm huh? like what 17:34 celeron55 i would like to experiement with eg. allowing building of ships from nodes 17:34 celeron55 and other moving things 17:34 hmmmm ah 17:34 hmmmm so you mean a huge ship made out of nodes that can move just like that 17:34 hmmmm that's awesome 17:34 celeron55 and it would also allow some miniaturized mesecon stuff or something 17:35 hmmmm yes, that can obviously simplify pistons 17:35 celeron55 that's what people would like, but the current codebase is hard to make do those 17:36 celeron55 also one thing to consider is whether it is practical to retain the 0.4 scripting api or are there reasons to break compatibility 17:36 celeron55 there'd be work for many full time employees 8) 17:36 hmmmm what changes are you thinking of in particular? 17:38 celeron55 dunno, really 17:38 celeron55 it's just a question to keep in mind when figuring out new things 17:40 celeron55 i'm a bit unsure about if we have enough experience for designing the basis for a next-gen minetest yet 17:40 hmmmm i can't really see a situation where it'd be more beneficial to break compatibility than it is to just add a new version of the function and slap an Ex at the end 17:41 celeron55 that is probable 17:42 celeron55 ...mostly thanks to the design principle of making the api good no matter what the insides look like 8) 17:43 celeron55 then there is the decision of what things to allow client-side scripting to do 17:45 celeron55 i think it has to be added for some things, but it is nontrivial to think of what it's responsibilities would be and what is worthwhile to include in it's api in the first place, AND how much of the actual client should be done with it then, as it would exist in any case 17:46 celeron55 and now i remember one important thing about scripting api compatibility 17:46 celeron55 it's the question of whether luabind would be used to ease the maintenance of the ever-growing api code 17:47 celeron55 if something like that would be done, then exact compatilibity can be surprisingly hard 17:49 celeron55 so much that the alternatives really are "compatibility or luabind/other" 17:54 PilzAdam do I miss anything obvious here: https://github.com/celeron55/minetest/pull/350 ? 18:05 celeron55 umm... just push it and we'll see 8) 19:32 celeron55 i think it's quite funky how the basic structure of the engine still practically exactly is what this one had http://wiki.minetest.com/wiki/File:Putkiksia.png 19:33 PilzAdam heh 19:38 celeron55 it just has a large bunch of optimizations, polish and a somewhat scalable content system bolted on top of it 19:39 celeron55 i absolutely hate thinking about making a minecraft clone from scratch, because while the core idea is very simple, you need to throw so much stuff at it before it becomes something people consider worth using 19:41 celeron55 on the other hand, i find it interesting to think of ways to make it as painless as possible 19:44 celeron55 but i really don't have the passion for making minecraft-like stuff anymore; the only reason i do stuff is because i sometimes feel like fixing bugs (it is always good programming practice) or want to learn and try a bit of something new 19:46 celeron55 as for the website and releases, it's kind of a different thing and the previous stuff doesn't apply to it much at all 19:47 PilzAdam so, if I understand you correctly, you dont want to make a good game anymore, you just see Minetest as a sandbox to code just for fun? 19:47 celeron55 no 19:48 PilzAdam what do you mean by "minecraft-like features"? 19:48 celeron55 of course i want to make good games, but those efforts tend to go elsewhere; i can't attempt to make a good game if i don't feel like using the end result much at all (which is the case for minetest) 19:49 celeron55 and coding minetest is not usually fun; it's more of a learning thing, and about doing good things to people 19:50 celeron55 PilzAdam: i don't see me having said "minecraft-like features" anywhere 19:51 celeron55 so my question is, what do you mean by that question? 8) 19:51 VanessaE that said, celeron55, hacking on it or modding for it IS fun for many of us 19:51 PilzAdam but i really don't have the passion for making minecraft-like stuff anymore; 19:52 celeron55 it connects strongly to the fact that i don't feel like playing mine* 19:53 celeron55 it applies to any first-person block-building boxel game 20:12 hmmmm so where did the name mineTEST come from 20:13 hmmmm testing something out? 20:24 celeron55 yep; that's what i do all time 20:24 celeron55 it's just one of my projects of trying something out, and this particular one got a bit out of hand 8) 22:18 Omnistudent Good evening 22:19 Omnistudent @SitDog_ Did you get any tips on the mac develpment?