Minetest logo

IRC log for #minetest, 2015-03-31

| Channels | #minetest index | Today | | Google Search | Plaintext

All times shown according to UTC.

Time Nick Message
00:00 Enke joined #minetest
00:03 * Jordach wonders how /clearobjects work
00:04 DMackey joined #minetest
00:08 Warr1024 joined #minetest
00:17 YvesLevier Jordach: Clients when logging get a warning about cart.png not in textures.
00:17 YvesLevier Even holding shift - carts dont vanish
00:17 YvesLevier im mystified
00:20 Jordach joined #minetest
00:22 midnight joined #minetest
00:35 Sokomine Jordach: depends :-) if it was your mapgen.....
00:35 Jordach cram enough entities into a small enough space below the suspicious limit
00:36 Jordach and it'll crash
00:36 Jordach under OpenGL
00:39 Sokomine quite possibly, yes
00:40 Sokomine as far as falling entities go...i wish snow wouldn't destroy plants. i havn't found a way yet to prevent that. the falling from builtin removes plants. i'll probably have to copy and modify the code for my moresnow mod
00:40 Sokomine right now, it does cause a cascade similar to what you observed if a snowball is dropped on the wheat fields around a snow-covered village. at least it's a relatively small 2d-area, so there's no danger of a crash
00:41 Jordach Sokomine, ~4x~5x~10 area
00:44 Viper168 joined #minetest
00:48 Yepoleb joined #minetest
00:48 Pilcrow can falling nodes fall through "ignore"? because mine seem to be able to, but it may be some mod that makes them do that. all I know is that when sand or similar is updated while sitting on a chunk that hasn't been generated, it'll start to fall into the abyss....  :/
00:48 Yepoleb joined #minetest
00:50 Pilcrow ^ I'm experimenting with programming some sort of quarry, in a world composed entirely of falling nodes. the whole world collapses when the quarry digs to the bottom of the loaded chunk...  :P
00:54 Warr1024 anyone want to test out PR #2579 and confirm that it fixes #2558, and doesn't introduce any new problems?
00:55 Pilcrow I'll do it Warr1024. I've been waiting for a fix to that...  :)
00:56 paramat joined #minetest
00:56 Warr1024 thanks paramat
00:57 Warr1024 er, thanks Pilcrow
00:57 Warr1024 why does my tab completion suck?
00:57 paramat you're welcome
00:57 Pilcrow lol
00:58 Warr1024 I'm also curious how the fix works with custom texture packs and skyboxes.
00:58 Warr1024 Unfortunately, my system is pretty ultra-low-end, so I can't test a lot of the fancier things.
00:59 Pilcrow yeah, I doubt I can test those things very well either; crappy little dual-core laptop (HP G-60)...  :P
01:00 Warr1024 I also managed to sneak in a slight improvement: wieldmeshes should now be based on low-res texture, but use the high-res texture.  Before, it was all high-res, resulting in overcomplex meshes.
01:00 paramat good grief, that voxel world size comparison article misses the importance of vertical height. i would rank a world size on it's smallest dimension not it's largest, which makes MT the second largest
01:01 paramat blockstory mixes voxels with hi-res photorealism and therefore looks awful
01:01 Zeno` joined #minetest
01:02 Warr1024 frankly I never like to travel more than a few km in any direction from spawn, anyway
01:02 Warr1024 might make more sense for a PvP free-for-all server, or one where people have WAY too much time to build huge...
01:02 paramat too much horizontal dick-waving
01:02 Warr1024 but for me, I could probably live happily within a 2km cube
01:02 paramat MC's 256m vertical height is pathetic
01:03 Pilcrow ^ well, paramat, I think vertical space would be more important if there were more 'dimensions' generated, but you're one of the only modders that uses all that vertical space, so far...  ;)
01:03 Pilcrow (you and like 2 others)
01:04 paramat MT's vetical height is why i'm here
01:04 paramat (vertical)
01:04 paramat that was number one consideration when i was choosing
01:04 Warr1024 I did find MC's 128m vertical limit rather stifling; the 256m extension was an improvement, but they added 128m of air, not stone, so you still can't dig all that deep.
01:05 Warr1024 paramat: what engines did you consider and reject due to the height thing?
01:07 paramat erm i think it was MC MT and terasology
01:08 paramat but lua modding was another decider
01:08 Pilcrow Paramat, my number one consideration was the ease of modding. And then there's the fact that minetest is free, lol. The vertical height is a great plus, and could be awesome, I'm just saying we need more realms (parallel mapgens?) for it to really shine...  :)
01:09 paramat yes, that's been my mission here
01:09 paramat =)
01:10 Pilcrow Warr1024: my laptop's finally done compiling. testing now.
01:11 Warr1024 awesome
01:13 paramat ah looks like terasology has cubic chunks and 'infinite height' since Oct 2014
01:14 Pilcrow Warr1024: Seems to solve that problem just fine (tried with a completely clean minetest.conf, fyi). As long as it doesn't break anything else, this is a great fix~  :)
01:15 Warr1024 thanks
01:15 Warr1024 Pilcrow: if you can think of any curveballs to throw at it, like stuff that I might not have thought of, try to break it :-)
01:17 Pilcrow well, I will try a higher-res texture pack at least. not sure where to get a custom skybox to test though, or I'd do that too.
01:18 Warr1024 yeah, I saw the feature in there, but I don't know of any server or mod that uses it.
01:20 Pilcrow I think there was an outer space skybox mod somewhere on the forums. heck if I can find it now though...  :P
01:27 Pilcrow looks like it causes no problems with HDX64 texture pack. http://i.imgur.com/xC6Epw2.png
01:28 Pilcrow ^ nothing wrong with player skins either. that's something so trivial I almost forgot to check...  :D
01:29 Pilcrow aha! Warr1024, I found Wuzzy's Galaxybox mod at https://forum.minetest.net/viewtopic.php?t=11173. trying that next.
01:35 Pilcrow Warr1024: skyboxes seem to work fine as well. incidentally, this mod is really pretty... I may use it for something! http://i.imgur.com/LTCVuQd.png
01:36 paramat left #minetest
01:40 Pilcrow irc is sleepy tonight...  :P
01:44 frecel joined #minetest
01:47 Warr1024 Pilcrow: thanks for testing.  You could probably add a comment with your test results to the github PR thread.  It'd be good to have a few people independently verify that it works.
01:48 Pilcrow ok, I'll do that. I'm a beginner at github; I keep forgetting you -can- leave comments...  :P
01:48 Freejack joined #minetest
01:53 Taoki joined #minetest
01:54 Pilcrow Warr1024: just curious, what prompted the original change to texture_min_size? I mean, I'm assuming it does something, but I can't figure out what...  :P
01:57 Warr1024 Pilcrow: you have to enable texture filters to see the effect
01:57 Warr1024 a lot of people run with those disabled, since they tended to blur everything.
01:57 Warr1024 but with this change, it helps with anti-aliasing within textures (for those of us with cheap vidcards).
01:58 Warr1024 a lot of cheap video cards, like my Intel GMA, implement anisotropic filtering with an internal bilinear filter
01:58 Warr1024 so we end up with blurring if we turn on ANYTHING other than JUST mipmaps.
01:59 Pilcrow ah, ok. I think I still have an older build to check out; I should enable the texture filtering in both and compare! :)
01:59 Warr1024 Yep, with texture_min_scale, you should be able to turn on all the trilinear+mip+aniso options and still see squarish pixels.
01:59 Warr1024 I can run with texture_min_size=64 without issues, other people can often go higher.
02:01 Pilcrow I have a reletively crappy video card as well, because it's old. Would I be able to see the difference with a GeForce 8200M?
02:03 Warr1024 Pilcrow: try setting texture_min_size = 1 and enabling trilinear+mip+aniso with low-res textures.  If everything looks blurry, then you should be able to crank up texture_min_size to reduce it.
02:06 rickmcfarley joined #minetest
02:10 enricom_ joined #minetest
02:12 Pilcrow aha, now I see! incidentally, my shaders are working. the last time I tried using them, minetest said my video card was unsupported and wouldn't load the shaders. that was a long while ago, but I wonder what changed... waving nodes are working, as well as bump mapping...  :D
02:13 Warr1024 wow, awesome
02:13 Warr1024 we need an FXAA shader.  Any good shader programmers in here?
02:13 Zeno` we need someone to debug shaders on Android as well...
02:14 Pilcrow ^ sorry, Zeno`, I'd help with that if I had <space> on my Shield, but it's full to the brim right now...  :P
02:16 Pilcrow not sure how much help I'd be anyway; no understanding of C/C++...
02:16 Warr1024 man, it's just painful compiling MT for android from scratch...
02:16 Zeno` http://dpaste.com/0RWX9Z9
02:16 Warr1024 Pilcrow: IIRC shaders are programmed in their own language...
02:17 Warr1024 Zeno`: what are we looking at there?
02:17 Warr1024 possible source of the shader issue?
02:17 Zeno` Some others may be due to npot textures: http://dpaste.com/15XRZ9S
02:17 Zeno` Warr1024, probably yeah
02:17 Zeno` issues shown in logcat
02:18 Zeno` I dunno what those storage ones are actually
02:18 Warr1024 the apk apparently has to unpack the native stuff into separate files on storage
02:18 Warr1024 i.e. it doesn't run out of the APK.
02:19 Zeno` well ignore them
02:19 Zeno` the others are probably at least related to the shaders not working though
02:19 Zeno` (the others in the first post)
02:19 Zeno` s/post/link
02:19 Warr1024 texture handling is weird on android
02:19 Warr1024 frankly I think a software RTT is the way to go to get rid of a lot of the issues we have with GUI's too.
02:20 Warr1024 I had to enable inventory_image_hack, which got things working ok, though looking a little crappy
02:20 Warr1024 since it does this weird non-aspect-preserved scaling of the mesh it renders
02:20 Warr1024 no freakin' idea why
02:21 Warr1024 haven't looked at kahrl's branch much yet, though.
02:21 Zeno` hmm
02:21 Warr1024 I was sort of thinking of just doing it with a cheap "reference" type renderer, i.e. performance-naive but correctly-working and reasonably accurate, for starters.
02:21 Warr1024 RTT also tends to play non-nice with GUI Scaling.
02:22 Warr1024 er, filtered GUI scaling, that is.
02:23 Zeno` Pilcrow, you just need to delete all the other stuff on your shield :D
02:23 Zeno` Warr1024, going back to shaders... have they ever worked?
02:23 Zeno` I suppose that did otherwise the checkbox would have been removed
02:24 Pilcrow ^ Warr1024, is inventory_image_hack the one that basically turns off inventorycube in some devices where inventory textures just look like random garbage? I know that used to happen in the Nvidia Shield, too. Annoyed me quite badly...
02:24 Zeno` (I'm talking about Android only of course)
02:24 Warr1024 Zeno`: I've had them work for me on desktop, though I couldn't afford the perf cost.  IIRC I've never seen them work on droid, though I might not have tried, since I barely get playable framerates as is...
02:24 Warr1024 Pilcrow: inventory_image_hack renders the inventory cubes to the center of the screen and then uses a "screenshot" command to pluck the image back into the game.
02:25 Warr1024 Oddly enough, when I DON'T have inventory_image_hack on, RTT kind of works, but still renders to the bottom-left corner of the screen.
02:25 Warr1024 also, textures rendered via RTT can't be scraped back into images on my device, defeating gui_scaling_filter and ^[inventorycube.
02:26 Warr1024 with software RTT, we wouldn't have to worry about a lot of RTT hacks we're using right now.
02:26 Zeno` hmm
02:26 Warr1024 We can even afford a moderately *slow* RTT, since textures are cached after rendering once.
02:27 Pilcrow ah. like I said, it's been a while since I've used mt on droid. 0.4.10, actually. But yeah, I remember that textures were rendering as garbage, and we basically had to flatten them at the time, making them look panel-ish, rather than cubic.
02:31 Warr1024 Man, I just wanna finish polishing up client-side visuals.  Not even anything really fancy like shaders; just a little texture filtering.
02:31 Warr1024 Just seems like, since I opened that can of worms, I've exposed a lot of issues.
02:31 Warr1024 though, tbh, I've learned a lot about texture and image handling in MT.
02:32 Pilcrow the thing that dissuaded me from mt in android was an incompatability with nvidia's button mapper; when my joystick was mapped to the middle of the screen, it would work for the most part, but randomly think I was tapping the corners, which would open menues or send my character spinning to look straight up, etc...
02:32 Warr1024 oof, you've got one of those joysticks
02:32 Warr1024 seems like there should be a simple way to filter those
02:32 Warr1024 like the way a "despeckle" filter works on images
02:33 Pilcrow the joystick works fine in _every_ other game though. just minetest where it has that problem...
02:33 Warr1024 weird
02:33 Warr1024 I've got a galaxy tab, myself.  Touchscreen controls are usable enough, though my playstyle doesn't involve a lot of intense combat.
02:34 Warr1024 it's a bit underpowered, though.
02:34 Warr1024 I swear that the touchscreen stops responding when I'm running on battery, too, probably due to the massive CPU/GPU power drain...
02:35 Pilcrow it would be nice if minetest supported controller mapping, itself. In fact, a lot of work has already been done in that direction, what with the remappable keys and mouse buttons... maybe someday...
02:37 Warr1024 android build took me ~15.5 minutes
02:37 Zeno` what are you using to build on?
02:38 Warr1024 Core 2 Duo laptop
02:38 Warr1024 oooh, I successfully applied scaling to the touchscreen controls...
02:39 Warr1024 ...but apparently scaled them to the wrong size :-/
02:39 Zeno` first build for me took a while (building all the libs and stuff) but now subsequent builds not so long...
02:39 Pilcrow but yes, I've got the Shield Portable. Powerful little beast, but it has a terribly easy-to-smudge screen (so much so that I usually treat it as if it had no touch capabilities; I can't stand fingerprints, lol). If I could map the built-in controller to minetest, I'd probably use it more than my laptop!  :D
02:39 Zeno` You can install it with `adb install -r bin/Minetest-debug.apk`
02:39 Zeno` make[1]: Leaving directory '/home/crobbins/minetest/build/android'
02:39 Zeno` real    0m39.724s
02:39 Zeno` 40 seconds
02:40 Zeno` (that was using time make -j8)
02:40 Pilcrow j8? you have 8 cores?? wow...
02:40 Zeno` yeah
02:41 Warr1024 8 cores, or 8 threads?
02:41 Zeno` well, 4 cores hyperthreaded (i7)
02:41 Pilcrow ah
02:41 Warr1024 ah
02:41 Pilcrow lol
02:42 Warr1024 hm, I think it's because I left scaling enabled, but the images were stretched to their NPOT2 sizes.
02:42 Warr1024 btw, I've seen npot2 in the code, but what's it stand for?
02:42 Warr1024 next power of two 2?
02:43 frecel_ joined #minetest
02:44 Pilcrow I've got an odd thought: does minetest work in a chromebook (without installing a different os)?
02:45 Zeno` non power of two
02:45 Zeno` or not power of two
02:45 Zeno` one of the two :)
02:45 Warr1024 well, what it returns is NEXT power of two.
02:45 Zeno` really?
02:45 Warr1024 but wouldn't that either be "npot" or "npo2"?
02:45 Warr1024 not "npot2"
02:46 Zeno` is it in minetest code?
02:46 Zeno` if so it's probably a typo
02:46 Warr1024 yeah
02:46 Warr1024 it was all over client/tile.cpp
02:46 Warr1024 I migrated the function itself into util/numeric
02:46 Warr1024 er, not sure if that's been merged yet...
02:47 Zeno` err
02:47 Zeno` perhaps the comment tells you want it does :p
02:48 Zeno` @param size get next npot2 value
02:48 Zeno` in a strange kind of way
02:48 Zeno` lol
02:48 Zeno` I'd personally name it next_npot()
02:48 Zeno` grr
02:48 Zeno` I'd personally name it next_pot()
02:49 Zeno` oh, I hope you optimised that
02:49 Zeno` I seem to recall you talking about it?
02:49 Zeno` there are bit twiddling ways to do that that function does using a loop
02:49 Zeno` what that*
02:52 * Pilcrow is lost in all this dev-y talk, lol
02:52 Zeno` Warr1024, yeah it was you, and you did do it using bit shifts the fast way... good :)
02:52 Zeno` sorry Pilcrow... I forget which channel we're in sometimes
02:53 kahrl there's an is_power_of_two too
02:53 kahrl so maybe next_power_of_two
02:56 Zeno` yeah, even better
02:58 kahrl next_pot is something I'd expect in a tomb raider or zelda game ;)
03:01 Zeno` it's not really next power of two either
03:01 Zeno` it's more like      ceil_power_of_two(floor_power_of_two(n))
03:01 Zeno` e.g. npot2(32) returns 32
03:03 Zeno` well, ceil_power_of_two(n) is what it is
03:03 Zeno` ceil(32) is 32 *embarrassed*
03:04 Zeno` next_power_of_two() would imply that n==32 would return 64
03:04 kahrl hmm, just tested it, wouldn't it be more like floor_power_of_two?
03:05 Zeno` oh it's floor
03:06 * Zeno` puts it in a loop
03:07 Zeno` yeah it's floor
03:07 kahrl http://web.mit.edu/rust-lang_v0.11/d​oc/std/num/fn.next_power_of_two.html
03:08 kahrl they also call it next_power_of_two even though it leaves powers of two unchanged
03:08 rickmcfarley joined #minetest
03:08 Zeno` but this, as you noticed, isn't even next
03:08 Zeno` http://codepad.org/KT8dgCXq
03:09 kahrl of course
03:10 * Zeno` looks for Warr1024's optimisation to make sure it's floorpot
03:10 Zeno` lol
03:10 Zeno` floorpot... I like that
03:10 kahrl :D
03:16 Zeno` Warr1024, your optimisation is incorrect :(
03:16 Zeno` in #2536
03:16 Zeno` you've implemented ceil_pot
03:16 kahrl this is why stuff like this needs unit tests
03:17 Zeno` of course
03:17 Zeno` I think all util/ should have unit tests
03:17 kahrl well, it was tile.h before
03:17 Zeno` yeah I know
03:17 Zeno` but he hasn't changed it in tile.h
03:17 kahrl (I don't know why)
03:18 Zeno` It's in a different PR
03:18 Warr1024 Zeno`: wait, what optimization is incorrect?
03:18 Zeno` npot2
03:18 Warr1024 how's it incorrect?
03:19 Zeno` you've done it right, but the thing is the original npot2 is incorrectly named
03:19 Zeno` scroll up
03:19 Zeno` the original npot2 is actually floor_power_of_two()
03:19 Zeno` whereas in #2536 you've implemented ceil_power_of_two()
03:20 Warr1024 oh
03:20 kahrl "@param size get next npot2 value" is also a pretty weird comment
03:20 Warr1024 that doesn
03:20 Warr1024 't necessarily mean the optimization is incorrect
03:20 Warr1024 maybe I both optimized AND corrected the function
03:20 Zeno` kahrl, and also incorrect
03:20 kahrl yes
03:20 Warr1024 presumably, you wouldn't want to scale DOWN...
03:20 Zeno` Warr1024, maybe
03:21 Warr1024 If you had a 15px texture and you scaled it up to 16px, you'd double one row of pixels.  If you scaled it down, you'd have to throw away 7.
03:21 Warr1024 2**ceil(log2(n)) is the only way that makes sense to me.
03:22 Warr1024 even if I had to scale up from 33 to 64, I'd rather do that than throw away a potentially important detail to scale down to 32.
03:23 Warr1024 technically, for the Npot2 thing with textures, we should offer the NNAA option.
03:24 Warr1024 right now, it's doing plain NN scaling, and it'll ALWAYS be doing a non-integer scale.
03:24 Warr1024 (that is, if the scaling kicks in at all)
03:25 Warr1024 android incremental builds break if you have to kill the build partway through, or if a dep fails to build, btw.
03:25 Warr1024 at least, last time I checked, unless it's had some major surgery in the meantime.
03:26 kahrl intuitively I'd agree with you but there might be something weird about android that means it's better to scale down
03:26 kahrl so I'll leave that to the android experts
03:26 Warr1024 well, I'd guess that it might be to avoid hardware scalability issues
03:27 Warr1024 but in that case, you'd think it'd choose to go either up or down, depending on what's closest...
03:27 Warr1024 I sort of doubt that there are any real considerations for memory limits, though...
03:28 Warr1024 considering both the wide difference in capabilities of devices, and the wide difference in texture size among some mods/servers
03:28 kahrl I don't think it's memory, since how many textures are non-pot in a typical game?
03:28 kahrl 0, perhaps 1
03:28 Warr1024 no, I mean, memory could be one reason to scale down instead of up
03:28 Warr1024 realistically there are probably very few mods that include non-po2 textures anyway
03:29 Warr1024 if there ARE, though, I figure that any device ought to be able to handle scaling up instead of down...
03:30 Warr1024 well, if there was a reason to use floor instead of ceiling, it certainly wasn't documented anywere.
03:32 Miner_48er joined #minetest
03:32 Warr1024 everything seems to say "align to power of 2" but not WHICH power of 2.
03:32 kahrl probably best to ask sapier if you really want the answer
03:33 Warr1024 haven't seen sapier around.
03:33 kahrl ~seen sapier
03:33 ShadowBot kahrl: I saw sapier in #minetest 36 weeks, 5 days, 7 hours, 10 minutes, and 16 seconds ago saying "does anyone know martin_devil? I'd like to know how he managed to run an arm version on a mips device ;-)"
03:33 Warr1024 that doesn't sound too promising.
03:34 kahrl oh well, perhaps change it to ceil and see if anybody complains
03:37 Warr1024 yeah, that's sort of what I figure :-)
03:37 Warr1024 if it turns out there IS a reason for a separate floor function, it won't be hard to make one.
03:38 Warr1024 floorpo2(v) = (v & (v - 1)) ? ceilpo2(v) : (ceilpo2(v) >> 1), I guess.
03:38 Warr1024 though I'm pretty sure that's not the best way
03:40 Zeno` I just made a great one
03:40 Zeno` using the copy pasta method
03:41 Zeno` http://codepad.org/8sDBGHf6
03:41 Zeno` haha
03:41 Warr1024 seems a bit complicated for the job
03:42 Zeno` ref: graphics.stanford.edu/~seander/bithacks.html
03:42 Zeno` well, do better :P
03:42 Warr1024 yeah, that's where I stole mine
03:42 Zeno` just add a unit test :)
03:42 Warr1024 there's not a hell of a lot of difference between 12 instruction cycles and 8, let's say
03:42 Warr1024 considering that cache misses or branch pred fails are going to dominate runtime anyway...
03:43 Zeno` yeah it's a bit complicated because it's actually computing log base 2 (n)  and then bit shifting for the result
03:43 Zeno` so there is probably a better way
03:43 Warr1024 if we were doing this once per pixel and couldn't cache output, hell yeah, optimize.  Twice per texture that's going into a cache?  meh :-)
03:43 Zeno` yeah
03:44 Warr1024 well, it's easy to check for is power of 2
03:44 Warr1024 and we already have a function for ceil power of 2
03:44 Zeno` yep
03:44 Warr1024 floor power of 2 is ceilpowerof2 >> 1 if it's not already a power of 2, and same as ceil power of 2 if it is
03:44 Warr1024 hence my hack
03:44 Warr1024 which probably uses more instructions, but it's a bit more readable
03:44 Zeno` the branch on ARM is (usually) executed anyway and certainly would be for your solution I'd say
03:45 Warr1024 when you "optimize", there are a number of things you can optimize *for*, and sometimes readability is one of them...
03:45 Zeno` I did add "haha*" to my comment btw :P
03:45 Zeno` usually when I post code and add "haha" I mean (don't do this!)
03:46 Warr1024 ah
03:46 ndumont joined #minetest
03:46 Zeno` sorry, I need to be more serious
03:46 Warr1024 didn't know if the "haha" was related to where you got the code from
03:46 Warr1024 thought that was where I got mine :-)
03:46 Zeno` but no, yours is of course fine
03:47 Warr1024 I start to get wary whenever I see weird magic numbers and inline LUT's.
03:47 Warr1024 Once, in an assembly language class, I took advantage of the sparse memory handling of the emulator running my program to optimize something using a 4GB sparse LUT.
03:48 Warr1024 though the way they were measuring runtime was really naive: one "cycle" per executed instruction
03:48 Zeno` have you ever seen my code to check if a number is odd or even? Where I used mutually recursive is_even() and is_odd() functions calling each other with n-1 ?
03:48 Zeno` it's fantastic
03:48 Zeno` also my leap year function
03:48 Warr1024 oh, that's terrible :-)
03:48 Zeno` another great innovation
03:49 Warr1024 I don't even write functions to check a&1 or !(a&1)
03:49 Warr1024 leap year I imagine is a little more non-trivial...
03:49 Zeno` too obscure
03:50 Zeno` mine mutually recursive solution was so obvious it was elegence supreme
03:50 Warr1024 I'd hate to see you do fizzbuzz
03:50 Zeno` still trying to find my leap year
03:51 Zeno` found it
03:51 Zeno` http://codepad.org/RbX19xzo
03:51 Zeno` lol
03:51 Zeno` now THAT's optimisation
03:52 Zeno` pity it's so slow
03:52 Warr1024 I thought the Gregorian calendar was ratified in 1752?
03:53 Warr1024 so any date before then is technically not valid input.
03:53 Zeno` :/
03:54 Warr1024 heh, that's pretty funny, though
03:54 Warr1024 you implemented modulo arithmetic naively :-)
03:54 Zeno` of course
03:54 Zeno` they're cogs
03:54 Zeno` you cannot use modulo on cogs
03:54 Warr1024 cogulo
03:54 Zeno` lol
03:55 Warr1024 ok, I've got beautiful, anti-aliased android GUI buttons ... but all the textures are weirdly clipped.
03:55 Zeno` anyway Gregorian calendar is valid from 1582
03:55 Warr1024 hm, maybe it's a po2 problem
03:55 Zeno` err 1584
03:55 Warr1024 o rly?
03:55 Zeno` yes
03:55 Warr1024 hm, maybe I'm thinking of a SQL server bounds issue.
03:55 Zeno` just because Britain and America didn't start using it until 1752 doesn't mean it wasn't valid before then
03:56 Warr1024 hm, quite weird
03:56 Warr1024 you'd think that it could just as easily be extrapolated far into the past, anyway
03:56 Zeno` The cogs cannot lie!
03:56 Warr1024 i.e. any error introduced by confusing a Gregorian and a Julian date is probably not much worse than the inaccuracies of the Julian calendar itself...
03:57 Zeno` anyway, back to serious work
03:57 Zeno` (boring)
03:57 Warr1024 I bet I know the problem; the GUI buttons are probably centering the texture in the clip box, not upper-left-corner-ing it.
03:57 Warr1024 what kind of serious work.  MT, or day job?
03:58 Zeno` I'm out of work atm :(
03:58 Zeno` so MT
03:58 Zeno` profiling atm
04:07 Warr1024 aw friggin' yea
04:07 Warr1024 completely anti-aliased Android GUI, including the buttons
04:07 Shackra joined #minetest
04:13 Out`Of`Control joined #minetest
04:16 ]DMackey[ joined #minetest
04:18 Pilcrow Warr1024, your screenshots in https://github.com/minetest/minetest/pull/2536 look -so- nice! I want that...  :)
04:21 Warr1024 thanks!
04:21 Warr1024 I've had a lot of detractors saying that it makes the images too blurry
04:21 Warr1024 it's sort of a side effect of non-integer scaling to begin with
04:22 Warr1024 but anyway, I'm adding it as an option, not a mandate
04:23 Shackra joined #minetest
04:29 Pilcrow Warr1024, I wonder if there's a way to grow transparent areas slightly. If you could figure out a function that trims the edges of the NN-scaled texture and overlays it onto the AA one, you could get nice smooth edges without sacrificing detail (although there would still be random waviness within the texture)... I wonder if that would soothe the nay-sayers?  :P
04:29 Warr1024 Pilcrow: yeah, it sounds like the FXAA shader would get us something like that.  It might be a good compromise solution for some people, though I think it's a separate issue.
04:30 Warr1024 also, the anti-aliasing I'm doing also blends super-samples, if the original image is being downscaled
04:30 Warr1024 if it's being downscaled by a LOT, a pure nearest scaler could potentially lose a lot of detail that I preserve
04:30 Warr1024 so I'll probably continue to like my "accurate" method more than the "approximate" alternatives, for now...
04:31 Warr1024 though I think NNAA should be an option, and FXAA should be an option, and both, and neither.
04:31 Warr1024 though I could imagine some debate as to what the *default* settings should be...
04:32 Warr1024 I wonder if we should decide things like that via poll...?
04:33 Pilcrow good point about the downscaling. Isn't FXAA a hardware shader, though? If so, that's not so good for those who don't have support for it...
04:34 Warr1024 yeah, it's a hardware shader, but it's supposed to be a pretty efficient one.
04:34 Warr1024 my card supports shaders; I just turn them off because they're hella slow
04:34 Warr1024 I'd have to see how fXAA works...
04:34 Pilcrow lol
04:34 mazal joined #minetest
04:34 barrydk joined #minetest
04:35 Warr1024 it only has to operate on the 2d plane of the screen, after all the 3d stuff has been "flattened" onto it
04:35 Warr1024 so it might be fast
04:35 Warr1024 time to hold my nose and regression test with this fancy stuff turned off :-)
04:36 * Pilcrow doesn't know if I can enable fxaa in this Linux... I -think- I have the proprietary drivers, at least...
04:37 Warr1024 I think it can be done in software
04:37 Warr1024 i.e. it could be added to minetest independent of drivers
04:37 Pilcrow yeah, just not sure how to do it with X configs alone; I think I'd need to use nVidia Control Center, and I don't think I have it in this old thing...
04:37 Warr1024 I'd have filed a feature req for it, but I didn't want to become the "owner" of another issue :-)
04:39 Warr1024 I was sort of hoping HybridDog would take the initiative to request it :-)
04:40 fling joined #minetest
04:44 Pilcrow well. I actually do have a settings manager from nvidia, but there are no references to fxaa. all I have is an anti-aliasing slider that goes from 2x (2xMS) to 16x (8xMS, 8xCS). no other AA-related settings anywhere...  :P
04:45 Warr1024 man, you got some fancy hardware over there
04:45 Warr1024 I don't think mine supports any AA.
04:45 Warr1024 hence why all of my recent visual enhancement pr's have been software filters :-)
04:47 Warr1024 I'm running irssi inside tmux and I keep getting my irssi "change window" hotkeys mixed up with my tmux ones.
04:51 Pilcrow lol. Warr1024, yours is intel graphics, correct? Mine's and HP, about 8 years old, dual-core AMD Athalon CPU @2.1Ghz, but with nVidia GeForce 8200M graphics. Had to redo the thermal paste or I wouldn't be using it now, though; I've noticed *ALL* HPs in the last 10 years are sh*t for heat management...  :P
04:56 Pilcrow but I'm using a linux from like 2 years ago, not updated. 32-bit even though the laptop supports 64. the distro I put on this switched to 64-bit-only while I was ignoring this laptop due to heat issues. Finally took care of it and am using it again, but there's nothing to update officially since the 32-bit repos are gone, and I've got no more backup space to redo the os yet... Guess I'll have to suffer for now...  :P
04:59 Warr1024 buy one of those external HDD's
05:02 Pilcrow I've got a 1TB, but it's full. I'm between jobs. I will definitely buy one as soon as I can afford it, though...  :P
05:04 Chanku joined #minetest
05:09 Pilcrow I wish I could figure out how to use both PRs #2579 and #2536 in my local build. I am so noob when it comes to using github...
05:09 Warr1024 ha, I wrote a script that manages it for me
05:09 Warr1024 I keep my history in branch whatever, and it auto-squashes my commits into pr_whatever
05:10 Warr1024 so I can keep track of what I changed when and why
05:10 Warr1024 and it also builds a "bleedingedge" branch with all of my PR's stacked together.
05:10 Pilcrow lol, that's cool
05:11 Warr1024 https://github.com/Warr1024/m​inetest/commits/bleedingedge
05:12 SylvieLorxu joined #minetest
05:13 Pilcrow heh, you've piqued my curiosity; I'll try that...  ;)  But yeah, I signed up at github more than a year ago, but I've never actually used it for any of my projects, and have barely figured out how to do anything more than going back in time, commit-wise...  :P
05:14 Warr1024 I was more of a gitorious fan myself: free sotware hosting running on free software, instead of proprietary github.
05:14 MinetestBot [git] kwolekr -> minetest/minetest: GenElementManager: Pass opaque handles to Lua and rename to ObjDefManager ed10005 http://git.io/jMzv (2015-03-31T01:11:51-04:00)
05:14 Warr1024 but they've been swallowed up by gitlab now.
05:14 Warr1024 so I guess I'm on that now
05:15 Warr1024 but the nice thing about git is that you aren't stuck with one host.  Every clone of a repo you make is more-or-less the complete project history, which you can sync with any other hosting provider...
05:16 Warr1024 if you're used to a central system like CVS, SVN, TFS, or VSS, you have to unlearn all that stuff first, though.
05:16 Pilcrow well, I've personally never used any version control systems, so it's -all- foreign to me
05:17 Warr1024 I use them for EVERYTHING
05:17 Warr1024 even things I probably should't
05:18 Warr1024 for instance, what's the best way to take a nightly backup of a minetest server world?
05:18 Pilcrow heh, learning git seems almost as time-consuming as learning a programming language, tbh...  :P
05:18 Warr1024 I'm using git, and that means I can do a git gc and do delta compression, keeping a full history, with some delta compression benefits.
05:20 Pilcrow ah, didn't know there was delta support. I'd imagine modifying the existing files instead of downloading and overwriting them in whole saves a lot of time for nightlies...
05:20 Pilcrow or is that not what you meant?
05:21 Warr1024 it's not just that, but I want to keep a history going back all the way to the start of teh world.  Doing each as an independent tarball, and not taking advantage of similarities across time, would make that space-prohibitive.
05:21 Warr1024 my longest-running MC world, for instance, ran for about 4 years, with backups at least once a night.
05:21 Warr1024 one of these days I might try to get around to doing a time-lapse project or something with it
05:21 Warr1024 at least, I can go back in time for griefing repair and/or nostalgia, I guess.
05:22 Pilcrow ah. now I get it. yeah, that's a great use of space, as long as no in-between files get corrupted...
05:23 jin_xi joined #minetest
05:23 Pilcrow do you make backups of the backups?  B)
05:23 Warr1024 if they do, I can use git bisect to find out the point at which they got corrupted and roll back to it
05:23 Warr1024 yeah, git repos are naturally backed up if you clone them
05:23 Pilcrow oh. useful
05:23 Warr1024 so I push from my world backup into my git host, which gets fetched by my personal offsite machine
05:24 Warr1024 you can configure git to validate the objects it pulls too, so each fetch checks the consistency of data it's receiving
05:24 Warr1024 a commit ID is actually a cryptographic verification of the entire content of a commit, and its full history
05:26 Pilcrow so like an md5/sha256 sum?
05:28 Warr1024 sha1 to be specific
05:28 Warr1024 a bit dated, but not completely broken yet
05:28 Warr1024 as long as you use the long hash, not the short hash
05:29 Warr1024 er, at least, if you NEED cryptographic verification
05:29 Pilcrow I have no idea the differences between hash methods anyway, lol
05:31 Warr1024 MD5 is badly broken, I can't even remember if it's OK in HMAC form.  SHA1 is considered antiquated and people should be transitioning away.  SHA2 (256, 512, et al) are the current "standard."
05:32 Warr1024 I can't keep track of all the others, like Tiger, Whirlpool, RIPEMD, etc.
05:33 Warr1024 argh, screwed up my build for like the 3rd time in a row
05:33 Warr1024 switched branches with an unclean tree, forgot to wipe temp files, etc...
05:34 Shackra joined #minetest
05:34 Pilcrow hmm.
05:35 Warr1024 it's not a *real* sandbox game unless you can hack on the game source itself
05:37 Pilcrow wow. as soon as I opened that nvidia control panel, my colors went all wacky... such color! so saturation! very doge!  ...  ....   @_@
05:37 Warr1024 doge this.
05:40 Pilcrow ok, I fixed it I think. gamma for each color was set to 1.2 and I dropped it to 1.0...
05:43 Pilcrow anyone here familiar with KVirc? I've been using the webchat at freenode.net as I can't seem to get it working. I'm not very good with irc stuff...
05:44 Pilcrow ^ as in, I can't get KVirc to connect, so I've been using the webchat in my browser instead
05:46 WarrTab joined #minetest
05:47 Pilcrow_ joined #minetest
05:51 ndumont joined #minetest
05:53 Zeno` Warr1024, you have your world database in a git repository?
05:53 Warr1024 Zeno`: yep.
05:53 Warr1024 as long as you don't do a vacuum, sqlite databases will delta-compress okay.
05:54 Warr1024 I don't think leveldb or redis would as well
05:54 Warr1024 but leveldb is a lot less portable, and redis has zero security as implemented in MT, so those aren't really options for me.
05:55 Pilcrow_ there should be a pure text option for world db! lol, kidding.  :P
05:56 Pilcrow_ can you imagine the entire world in a giant world.txt file?  @_@
06:02 Warr1024 I'd actually kind of like the option for a very simple db with one file per mapblock...
06:02 Warr1024 it'd be murder on a filesystem, but with a decent filesystem, it'd actually perform very well with git compression.
06:03 Warr1024 that's assuming, of course, that we're doing safe writes.
06:03 Warr1024 I'd really like the opposite, too, actually: to have the option to store ALL data in the same key/value store, instead of splitting it between a DB and the filesystem.
06:05 Warr1024 ah, finally got this thing built and installed properly on android
06:05 Warr1024 anti-aliased gui buttons look a LOT better than the old ones did...
06:05 Pilcrow_ heh, you've been working on that for a while now... lots of problems, I take it?
06:06 Warr1024 yeah, apparently on the device side
06:06 Warr1024 also, looks like that jump button is actually a good bit lower res than the others; maybe a fresh texture is warranted?
06:08 Pilcrow_ I was never a fan of that interface anyway. the arrows are arranged badly, and the jump button is just... odd-looking.
06:10 Warr1024 I think it's supposed to be a trampoline :-)
06:11 Warr1024 the one thing that bothers me is that the tap-and-drag for moving the viewport around is counter-intuitive if you're coming from a touchscreen background.
06:12 Warr1024 i.e. you expect to drag your finger to the right across the scene, and move the *scene* to the right, i.e. turn your character to the left.
06:13 Warr1024 ideally when you touch the screen and start a drag, it'd calculate the corresponding 3D vector, and rotate your view so that vector stays under your finger.
06:13 Warr1024 subject to limits like pitch +/- 90 degress.
06:14 Zeno` I dunno... my world
06:14 Pilcrow_ it'd be nice to have more android customization capabilities. an option to invert drag direction, some kind of interface skin support... we could have button-texture packs! maybe also a way to move them around...  :D
06:14 Zeno` I dunno... my world's database if 14GB
06:14 Zeno` is
06:14 Zeno` grr I cannot type today
06:14 Warr1024 Zeno`: yeah, that's beyond the ouch limit for git as far as I'm concerned
06:14 Warr1024 the largest git repos I've worked with have been around 12GB, and gc'ing them gets to be a nightmare.
06:15 Warr1024 I'd also like if we had tap-to-look on tablet.
06:15 Warr1024 er, android, sry
06:15 Warr1024 i.e. tap a sign to read its infotext
06:15 Warr1024 we have "free touch target" but it doesn't actually work that way unless you either drag, start a dig, or "right-click"
06:18 Pilcrow_ personally... I'd like to go the other way, against 'mobile' standards; I want a virtual trackball mouse or joystick in the bottom right corner and a virtual joystick instead of arrows in the left. to heck with dragging the whole screen...  :P
06:18 Warr1024 dragging the whole screen might be the only option on a small-screen phone
06:19 Warr1024 it'd be nice to have a choice
06:19 Warr1024 I mean, we have invert_mouse on desktop
06:20 Pilcrow_ yeah, choices are always nice with non-trivial things like that. personally, I don't like the current situation as it forces you to obscure your view with your own hands...
06:21 Warr1024 heh, accelerometer options?
06:22 Pilcrow_ heh. I'm sure -someone- likes using the accelerometer...... but it's not me!  :P
06:23 Warr1024 I'm trying to think of how it'd even be feasible
06:23 Warr1024 I guess you could sit in a swivel desk chair and use orientation for lookdir :-)
06:24 Warr1024 it might be worth implementing just to see people look ridiculous trying to do it.
06:24 Pilcrow_ tilt left/right to rotate screen, tilt forward/back to look up/down?  haha
06:25 Warr1024 it'd be a pain in the ass, though, if you want to change the "neutral" tilt, like if you're sitting in a recliner and want to lean back.
06:25 MinetestBot [git] Zeno- -> minetest/minetest: Change format of screenshot names 862d4ea http://git.io/jMMn (2015-03-31T16:24:25+10:00)
06:25 Zeno` you have a recliner?
06:25 Warr1024 yes
06:25 Zeno` man, you have it easy! In my day we sat on a wooden bench; and we liked it
06:25 Warr1024 ha
06:25 Warr1024 I've got one of those too
06:25 Zeno` :)
06:26 Warr1024 For me, the major benefit of the tablet is the ability to play while recumbent.
06:26 Warr1024 too awkward with a laptop
06:26 Zeno` depends
06:26 Zeno` if your laptop is suspended from the ceiling it's not too bad
06:27 Zeno` or you have a dentist chair with a nice sturdy arm/bench
06:27 Pilcrow_ Warr1024, isn't using accelerometer controls -always- a pain in the ass? who needs a configurable neutral tilt anyways?  B)  haha
06:27 Zeno` I suppose that'd look pretty dumb in your living room though
06:28 Warr1024 Zeno`: it'd be pretty convenient if you need to have dental work done anyway, though.
06:28 Zeno` true
06:29 Pilcrow_ Zeno`, everyone fears the dentist; you'll never have company again if you have a dentist chair in the middle of your living room...  :P
06:30 Zeno` Pilcrow_, you make it sound as though that's a bad thing
06:31 Pilcrow_ on the contrary; there'd be no one to think it looks dumb!  :D
06:40 Viper168 joined #minetest
06:46 meldrian joined #minetest
06:49 Pilcrow_ wow, Warr1024. that gui_scaling_filter makes first startup _much_ slower. once the textures have been cached, the next startup is fine, but on the first startup it took so long for minetest to open that I actually thought I froze my computer...  @_@
06:49 Viper168 joined #minetest
06:50 Warr1024 hm, that's a new one to me
06:50 Warr1024 you using a high-res texture pack?
06:50 hmmmm well, if it's really that problematic you *can* just disable it
06:50 hmmmm :p
06:51 Warr1024 oh, and are you using something like unified_inventory that'll force you to scale like a hundred different images at once when you first open the inventory?
06:51 Warr1024 also, I could really only test using gcc on debian, so I don't know if there are compiler optimization differences.
06:51 Pilcrow_ nope, default. took maybe 20 seconds for minetest to even appear. I was sat staring at the bin folder wondering if I should ctrl-alt-F1 and log into a fullscreen terminal to do a killall minetest... and then it finally popped up...
06:51 LazyJ VanessaE, if you are up and about, the new towel rod mesh in HomeDecor creates a duplicate, brown image 3 blocks to the left.
06:52 Warr1024 in my own testing, I've noticed it being just barely perceptible, if at all, on both my older desktop and my samsung galaxy tab
06:52 Warr1024 oh, wait, I know what you're talking about
06:52 Warr1024 you've got either a texture pack or some custom game that has a wallpaper
06:52 Warr1024 yeah, at the moment, I'm scaling those too
06:53 Warr1024 I suppose if it gets to be a problem, specifically, we can add an option to disable wallpaper scaling
06:53 Pilcrow_ ahhh, that's it then. yep. my own custom game has a wallpaper at the moment.
06:53 Warr1024 or if there are any hardcore image handling experts who want to take a crack at optimizing my code, it might be faster locking the image buffer inetad of using getpixel/setpixel...
06:54 Warr1024 slow I did notice; 20 seconds I did not.  I think I was using HDX-64 (the most my machine can handle) and it was maybe a second or 2
06:55 Pilcrow_ might it be possible to check the size of the image in pixels before performing the scaling? then if one if the dimentions exceeds, say, 512, don't use the scaling on it?
06:56 Warr1024 I don't really want to add a magic number like 512 to the thing anywhere
06:56 Warr1024 and wallpapers should be the only thing we should really have to worry about
06:56 Warr1024 and possibly the title image
06:57 Warr1024 also, ultra-high-res images were actually one of the things that people specifically wanted this for, from what I saw in the original Issue that led to this.
06:57 MinetestBot [git] Warr1024 -> minetest/minetest: Move texture_min_size even further down the pipe. Now, textures are JIT-upscaled using an image transformation, right at the time they're added to a mesh or particle; images used in 2D elements are left unscaled. This should fix any remaining issues with HUD elements. db32e6c http://git.io/jMHo (2015-03-31T16:56:33+10:00)
06:57 alket joined #minetest
06:57 Warr1024 nice
06:58 Akagi201 joined #minetest
06:58 Pilcrow_ that seemed fast. the fix was good though. so I suppose it deserves a speedy merge.
07:00 Warr1024 thanks
07:00 Warr1024 should close #2558 too, then?
07:01 Pilcrow_ yeah. do I need to do that myself, as I'm the one who opened it?
07:01 Warr1024 oh, hm, I dunno
07:02 Warr1024 either you or a contributer could probably do it?
07:02 Pilcrow_ I think I can. I just hit the close issue button under the comments, right?
07:03 Warr1024 sounds right to me.
07:03 Pilcrow_ ok. looks like that's closed now.  :)
07:03 nrzkt joined #minetest
07:05 Pilcrow_ abouth the magic number thing, you may be right, but I've seen a few people who've included screenshots in their mods. will this scale them too?
07:10 Warr1024 it might, if they're used on a formspec
07:11 Warr1024 to be clear, the change doesn't add/remove scaling, it just affects the precision with which scaling is done
07:12 Pilcrow_ might need to add some kind of configurable 'magic number' to minetest.conf then, just as an option...  :/
07:13 Warr1024 hm, ideally there'd be some way to pick a threshold based on number of pixels involved, which is roughly what I expect the runtime to be
07:13 Pilcrow_ no, I know it doesn't add/remove scaling. I worded that badly is all. I meant will it be scaled with _your_ method...
07:14 Warr1024 oh, wait, you're on 32-bit, aren't you?
07:14 Pilcrow_ yes
07:14 Warr1024 I keep forgetting how much of an impact that really has on performnce.
07:14 Warr1024 the scaling is all done by the CPU, so 32-bit hurts in this area
07:15 Warr1024 not only are 32-bit proc slower, but the compilers won't be keeping up with optimizations for them...
07:15 Pilcrow_ oh? well I think my computer might've just been having a brain fart; I removed the cache folder, etc. and started minetest again. it loaded straight away, no 20-second delay... how odd.
07:16 Pilcrow_ actually my cpu _is_ 64-bit, but my os is 32. no idea why I thought that was a good idea at the time.... I must've been dumb.  :P
07:16 Warr1024 hm, maybe your computer was busy with something else?
07:16 Warr1024 I ran a 32-bit OS on 64-bit hardware once
07:17 Warr1024 there was a time when a lot of software was not 64-bit clean
07:17 Warr1024 a lot of things used to run more reliably on a 32-bit little-endian intel-compat processor, probably because that was the only thing it was tested on...
07:17 Warr1024 I switched when I sensed that was no longer the case.
07:18 Pilcrow_ ehrm, you're right; looks like my os was indexing the new files, lol. that would explain the slowdown...  :/
07:19 Warr1024 argh indexing services
07:19 MinetestBot [git] nerzhul -> minetest/minetest: Typo in getOutgoingSequenceNumber => successfull to successful 7c12172 http://git.io/jMFk (2015-03-31T09:18:11+02:00)
07:20 Pilcrow_ yes, I know about those 32- vs 64-bit problems, but this was only installed a couple year ago; 64-bit software at the time was fine, for the most part...  :P
07:22 Pilcrow_ and yes, indexing services. this thing uses the old, slow nepomuk shit that came with earlier KDE4 builds... I really need a new backup drive so I can do an upgrade...  @_@
07:22 Pilcrow_ anyways, it's 2am here and I need sleep
07:22 Warr1024 yeah, I'ma check out too...
07:23 Pilcrow_ nice talking with you, Warr1024.
07:23 Warr1024 you too
07:25 zero-ghost joined #minetest
07:32 MinetestBot [git] nerzhul -> minetest/minetest: RunCommandQueues: little optimization ab77bf9 http://git.io/jMN8 (2015-03-31T09:29:33+02:00)
07:43 Calinou joined #minetest
07:50 fishyWET joined #minetest
07:52 Akagi201_ joined #minetest
08:01 Yepoleb_ joined #minetest
08:14 LazyJ joined #minetest
08:22 aheinecke joined #minetest
08:31 JamesTait joined #minetest
08:32 JamesTait Good morning all; happy World Backup Day!  :-D
08:38 roniz joined #minetest
08:44 err404 joined #minetest
09:14 Haudegen joined #minetest
09:21 MinetestBot [git] nerzhul -> minetest/minetest: Connection::Receive(): receive Network Packet instead of SharedBuffer<u8>. 1fe4256 http://git.io/jDRS (2015-03-31T11:01:08+02:00)
09:35 enricom joined #minetest
09:47 enricom joined #minetest
09:57 SylvieLorxu joined #minetest
10:32 Haudegen joined #minetest
10:48 OldCoder joined #minetest
10:51 ClaudeRoy joined #minetest
10:53 Zeno` joined #minetest
10:54 Zeno` yay
10:54 Zeno` https://www.youtube.com/watch?v=3iNxpIhjLws
10:54 Zeno` da river runs to you
10:57 Zeno` through desert dust
10:57 Zeno` and you still owe me something... more than this
10:57 Zeno` say you're coming back :(
10:57 Zeno` baby, love, wait for me
10:58 Zeno` say you want me back
10:58 xenkey no -.-
10:59 * Zeno` bursts into tears
10:59 Hirato joined #minetest
11:01 Zeno` waaaaaaaaah
11:03 xenkey Im sorry :c
11:04 raffahacks joined #minetest
11:04 raffahacks left #minetest
11:05 rubenwardy joined #minetest
11:18 rubenwardy Hi all!
11:21 Trustable joined #minetest
11:23 Jousway joined #minetest
11:28 MinetestBot [git] Zeno- -> minetest/minetest: Fix use of unitialized variable in gettext.cpp d1d5618 http://git.io/jyTu (2015-03-31T21:26:34+10:00)
11:30 Tablet_Two joined #minetest
11:31 xenkey thx Zeno` :)
11:33 Wuzzy joined #minetest
11:33 Zeno` :3
11:34 * Zeno` panics
11:36 alket joined #minetest
11:49 Akagi201 joined #minetest
11:50 Akagi201_ joined #minetest
11:53 proller joined #minetest
11:56 rubenwardy joined #minetest
12:08 Jordach joined #minetest
12:12 proller rubenwardy, inventory fixed!
12:12 proller it was missed merge from mt
12:12 rubenwardy Yay! What was the problem?
12:12 rubenwardy ah
12:12 rubenwardy Kool
12:19 Wuzzy Since Minetest the subgame is now called “Minetest Game”, who will draw the new text logo? ;)
12:20 rubenwardy minetest_game -> Minetest Game isn't much of a change
12:23 turtleman_ joined #minetest
12:25 Wuzzy “minetest_game” was always a kinda unofficial name
12:25 rubenwardy http://a.pomf.se/adnygw.png
12:25 Wuzzy Also, the textlogo clearly says “MINETEST”, not “minetest_game” :P
12:25 rubenwardy Isn't that the engine, though?
12:25 Wuzzy What's your point?
12:26 rubenwardy That screenshot was unrelated
12:26 Wuzzy I am talking about Minetest the subgame, not Minetest the engine.
12:27 Wuzzy the subgame has been renamed to “Minetest Game” → https://github.com/minetes​t/minetest_game/issues/437
12:27 rubenwardy But the text logo is from the engine?
12:28 Wuzzy I am talking about menu/header.png in the subgame folder
12:28 rubenwardy oh
12:28 Wuzzy I always thought that the name sharing of the subgame and the engine (same logo, same “official” name) was ugly
12:29 kattsmisk joined #minetest
12:29 Wuzzy and I am happy that at least the devs have come to a conclusion to rename it
12:32 Zeno` yep, we're close. It should be named "Creme Brulee" any day now
12:33 Wuzzy o_O
12:33 Wuzzy Where did you pick up THAT name? LOL
12:33 Wuzzy This is the weirdest suggestion I head so far.
12:34 Calinou let's call it “The Game Wuzzy Wanted To Rename”
12:34 Zeno` https://www.youtube.com/watch?v=7Bbfrn400zs
12:34 Calinou or “Go Develop Your Business Java/C# Application Elsewhere” if mgv7 is used
12:34 Zeno` ^--- Awesome
12:34 Zeno` Vooooodoooooooo lady
12:35 Zeno` you do, voo doo
12:36 Zeno` lol
12:37 Zeno` hmm... no applause
12:38 Zeno` when you do vooo dooo
12:43 Zeno` <Zeno`> Vooooodoooooooo lady
12:43 Zeno` oop
12:43 Zeno` https://www.youtube.com/watch?v=YOtpgz4L5d8
12:43 Zeno` lol
12:45 eeew joined #minetest
12:50 Zeno` nobody "gets" my sense of humour :(
12:50 * Zeno` cries again
12:50 rubenwardy 12 messages in a row...
12:51 Calinou S A Y I N G   M A C H I N E
12:58 Zeno` lol
12:58 Zeno` Calinou, what's the real problem with Coverity?
12:59 rubenwardy !g coverity
12:59 MinetestBot rubenwardy: https://www.coverity.com/
12:59 Calinou didn't I highlight it enough?
12:59 Calinou maybe I can edit my post and surround it with **
12:59 Zeno` Calinou, not really
12:59 Zeno` what is the problem with it being proprietary but free to use with open source?
13:00 Calinou it's the same problem as GitHub, Transifex, Slack and so many other random startups
13:00 Calinou it's hiring people to work gratis for them
13:00 Zeno` minetest lately has not been very open source (free software) friendly either
13:00 Calinou we have
13:00 Calinou with a few exceptions sadly (MMDB)
13:00 Zeno` ummm, no
13:00 Viper168 joined #minetest
13:01 Calinou sending DMCAs isn't anything wrong
13:01 Zeno` minetest doesn't appear to like forks
13:01 Zeno` that is not in the spirit of free software
13:01 Calinou I'm fine with forks, it's not my problem if random developers dislike them
13:02 Zeno` there is nothing wrong with VALID dmca's you mean
13:02 Zeno` but anyway, we're not allowed to use proprietary tools?
13:02 rubenwardy Use proprietary tools for yourself, but don't force others to
13:03 Zeno` it's not being forced on anyone
13:03 Calinou it's ethically wrong
13:03 rubenwardy Use MS Visual Studio or Coverity if you want to
13:03 Calinou you're using secret algorithms
13:03 Zeno` huh
13:03 Calinou (also, most projects using Coverity end up being obsessed with it.)
13:03 Calinou (it's a waste of time)
13:04 Zeno` most, if not all, static analysis tools are a waste of time
13:04 rubenwardy Calinou, it does not take any licensing away from Minetest, and you don't have to use it
13:04 Zeno` doesn't mean they shouldn't be used
13:04 Zeno` github is unethical?
13:04 Calinou yep
13:04 Zeno` err...
13:04 Calinou TPW is a troll
13:04 Zeno` why?
13:05 Zeno` I mean why is github not ethical
13:05 Calinou proprietary back-end, shitty staff, …
13:05 Calinou it perfectly deserves being called ShitHub :)
13:05 Zeno` I don't see what that has to do with ethics :/
13:06 Zeno` How could github change to make itself "ethical"?
13:06 rubenwardy I don't see closed source as unethical
13:06 Calinou release their back-end, let people self-host
13:06 Calinou it's as simple as that
13:06 Zeno` How would they make money?
13:06 * Calinou facehooves
13:06 Calinou 1. free software doesn't mean you don't make money
13:07 Calinou 2. developers should lower their expectations of income
13:07 Calinou 3. GitHub is not profitable as it is right now
13:07 Zeno` Who would pay for github's hosting?
13:07 Calinou people who can't be arsed to host it themselves
13:07 Zeno` err...
13:07 Zeno` and who will support those people?
13:08 Calinou employees?
13:08 Zeno` how do you pay the employees?
13:08 Calinou by selling support contracts?
13:08 Calinou for self-hosted or official instance
13:08 Zeno` which is basically what github does
13:08 Calinou no, they currently sell private repos
13:08 Calinou not the same thing ;)
13:08 SopaXorzTaker joined #minetest
13:09 Calinou [by the way, I believe a good GitHub replacement would be a non-profit, but that's another story]
13:09 Zeno` so private repos are the thing that's wrong?
13:09 Calinou no
13:09 Zeno` yeah and that alternative went broke, right?
13:09 Calinou Gitorious was a for-profit
13:09 Calinou a bad one
13:09 Zeno` that went broke
13:10 Karlton Zeno`: There are already free software projects including games that are commercial, so it is pointless to argue the notion that you can't make money with freesoftware.
13:10 Zeno` Karlton, I am not arguing that... Calinou is
13:10 Zeno` I suggest that it's perfectly fine to make money from free software
13:10 rubenwardy You make less money though.
13:11 Zeno` there are bills to pay, infrastructure to pay for, employees to hire, etc, etc, etc
13:11 Calinou I won't tell people lies by telling them they can make more money, that's for sure
13:11 Calinou learn to live with less money :P
13:12 Zeno` I just don't *get* this obsession with being angry at people making money who then provide us things for free
13:13 rubenwardy Companies need money to survive, more money the better
13:13 Zeno` rubenwardy, exactly
13:13 Calinou more money doesn't mean life will be better
13:13 Calinou nor that companies will survive better
13:13 Zeno` and if they let us use the product for free, even bettter
13:13 Calinou (because they can spend even more than what they earned, thus ruining the whole thing)
13:13 Calinou companies letting us use something for free is often a trap
13:13 Calinou (when it isn't, it means the company has no business model, so let's profit from that)
13:14 rubenwardy Calinou, sounds like you have more of a problem with capitalism than with closed source
13:14 Calinou capitalism isn't my problem
13:14 Calinou [that said, it doesn't mean capitalism doesn't have excesses in it]
13:18 Zeno` :/
13:18 Zeno` so github is not allowed to make money to improve their site
13:18 Zeno` makes sense (NOT)
13:19 Calinou they are allowed to
13:19 Calinou a license that disallows commercial usage isn't free
13:19 Zeno` what license is that?
13:19 Calinou CC BY-NC-SA is the most prominent example of that
13:19 Calinou (but it isn't a software license)
13:19 Zeno` yes
13:20 Zeno` GPL isn't free either
13:20 Calinou it is
13:20 Calinou anyone saying otherwise is engaging in political revisionism
13:20 Calinou no less than that
13:20 Zeno` no it's not because it doesn't allow me the freedom to make changes and not release the changes
13:20 Zeno` that's not freedom
13:20 Calinou 1. http://www.gnu.org/licenses/license-list#GNUGPL
13:20 Calinou 2. http://opensource.org/licenses/gpl-license
13:21 Zeno` BSD is probably the only really free license
13:21 Zeno` (and ones similar)
13:21 Calinou BSD is crap, better use MIT/Expat or Apache 2.
13:21 Calinou (it's confusing)
13:21 Karlton Zeno`: You can make all the chagnes you want without releasing them
13:21 Calinou also, yes, that's an huge GPL misunderstanding
13:21 Calinou I didn't see it
13:21 Karlton you just can't share them without releasing the code
13:22 Zeno` Karlton, I know
13:22 Calinou people making such a misunderstanding should not be allowed to complain about it
13:22 Calinou read the whole text of the GPL then complain.
13:22 Calinou it takes about 20 minutes
13:22 Zeno` Karlton, so long as you don't give other people the binary :P
13:22 Zeno` once you do that your freedom is lost
13:22 Calinou the GPL isn't about sharing changes, it's about passing on freedoms
13:23 Zeno` Calinou, not really because it doesn't give me the freedom to make changes, distribute that, and not make the changes in the source code available upon request
13:24 Zeno` I don't mind the GPL btw
13:24 Zeno` I'm just saying that it's not truly "freedom"
13:24 Zeno` But avoiding software that allows us to use it for free/gratis is just dumb
13:25 Zeno` Let them make their money; no skin off my back. We can use it, so who cares?
13:25 Karlton The unethical freedoms we don't care about, much like the freedom to murder or rob someone
13:26 Calinou freedom and anarchy are very different concepts
13:26 Karlton basicly not having the freedom to take away freedom is a good thing
13:26 Zeno` Karlton, I don't think you can compare murder to using proprietary software to help develop an open source project
13:27 rubenwardy I use proprietary software because it works better than the FLOSS alternatives
13:27 rubenwardy Simple as
13:27 Karlton Zeno`: both are unethical, but clearly not on the same level
13:27 Zeno` God forbis if one of our devs uses Windows as their platform; should we ban their PRs?
13:27 Calinou it's their own computing, not ours
13:27 rubenwardy Exactly
13:28 Zeno` So we should not use github
13:28 Zeno` what's the alternative?
13:29 Calinou self-hosting Gogs/GitBucket/Kallithea/…
13:29 Zeno` Personally I think github is fine
13:30 Calinou or using NotABug (Gogs instance)
13:30 Calinou or using git.framasoft.org (GitLab CE instance)
13:30 rubenwardy NotABug does not allow PRs
13:30 Zeno` yeah and they will go out of business soon enough
13:30 Calinou yes, will be eventually fixed
13:30 Calinou NAB is a non-profit, so is git.framasoft.org
13:30 Calinou they can't go out of business
13:30 Zeno` just like your precious gitorius
13:31 rubenwardy Yes they can go out of business, through lack of funding (see Mediacru.sh) or owners giving up
13:31 Zeno` SOMEONE has to pay for the infrastructure
13:31 Jordach hey
13:31 Jordach my own server is non-profit
13:31 Haudegen joined #minetest
13:31 Jordach i pay for it out of my own pocket ;P
13:31 Calinou then self-host, if you don't trust them
13:32 Calinou lots of projects do this already
13:32 Zeno` Calinou, but... who pays for self-hosting?
13:32 srifqi joined #minetest
13:32 xenkey You, the host
13:32 xenkey s/,/+/g
13:32 Zeno` so it's not free
13:32 Calinou we already do
13:32 Calinou it's not free as in price, but it's free as in freedom
13:33 Zeno` it's not free as in freedom because I can stop hosting any time I want
13:33 Zeno` you are bound to me
13:33 Calinou
13:33 Zeno` (I am using me as an example)
13:33 Calinou please don't abusively stretch stuff like that
13:33 Zeno` (e.g. if I hosted the project)
13:34 Zeno` it's not stretching
13:34 Zeno` someone has to pay
13:34 Karlton http://savannah.gnu.org/ is basicly a free version of github
13:34 Calinou Karlton, manual approval sucks, and the interface sucks
13:34 Calinou no one wants that in 2015
13:34 Calinou it can't be called a free version of GitHub
13:35 Zeno` I think it's best just to let github keep paying
13:35 Zeno` and they can raise funds to support the free hosting however they like
13:35 Calinou they already raise funds through private risk stuff
13:35 Calinou they're barely profitable, as said before :P
13:35 Calinou they too can go down
13:35 Zeno` of course
13:35 nrzkt you can use gitlab instead of github
13:36 Zeno` but they're not doing anything ethically wrong by allowing private repos
13:36 srifqi where should I ask about modding-related?
13:36 Zeno` gitlab costs money to host a project doesn't it?
13:36 nrzkt no
13:36 Calinou nrzkt, GitLab is partially proprietary due to the proprietary Enterprise Edition; I wouldn't recommend it
13:36 Calinou gitlab.com runs the EE
13:36 nrzkt you can host your instance
13:36 Calinou thus, it can't be recommended
13:36 Calinou host your own CE instance instead
13:36 Calinou or use git.framasoft.org
13:36 nrzkt my enterprise uses gitlab for its own problems
13:36 Calinou which is basically the same thing
13:36 Calinou it even lets you use private repos
13:36 nrzkt projects*
13:37 nrzkt git.framasoft.org is a gitlab
13:37 rubenwardy srifqi, come to #minetest-mods, you walked into a flamewar
13:37 Calinou *successful troll*
13:38 nrzkt gitlab is a pure shit ruby build which is not a good open source project because of their build which doesn't run well on many systems :p
13:38 Zeno` next people will be telling me not to use google because they make money from ads :/
13:39 rom1504 Zeno`: that's what framasoft is saying (see up there)
13:39 Zeno` Calinou, do you use google?
13:39 Karlton Zeno`: nobody cares about making money
13:39 Jordach Calinou uses duckduckgo
13:39 Zeno` Karlton, I know!
13:39 Zeno` Karlton, you are missing my point!
13:39 rom1504 gitlab is more ethical than github, maybe it's not perfect but still, it's better
13:40 Zeno` just because a company makes money does not mean we should avoid them on some questionable ethical ground
13:40 rubenwardy I used duckduckgo for a month, the only useful thing are the bangs like !yt search term. The actual search engine isn't as good as google
13:40 Karlton Zeno`: Who is saying that?
13:40 Zeno` I don't care if google make a billion $ a day; I will keep using them because their service to me is free
13:41 Zeno` (free as in gratis)
13:41 Calinou then you are not a free software supporter
13:41 Zeno` Calinou, and neither are you
13:41 zero-ghost its not free
13:41 zero-ghost you are the product
13:41 Calinou I happen to be one
13:41 zero-ghost the money comes from selling you
13:41 Zeno` zero-ghost, I understand that
13:41 rom1504 (but afaik gitlab hasn't solve some problems : how to I make a merge request on gitlab using my account on an other gitlab, or using github)
13:42 srifqi Zeno`: what language is "gratis"?
13:42 srifqi Netherlands?
13:42 Zeno` but avoiding using tools just because the owners of the tools might make money is DUMB
13:42 zero-ghost thats not why people avoid things like google
13:42 Zeno` did you know that the FSF makes... *gasp* money?
13:42 zero-ghost its HOW they make that money
13:42 zero-ghost and how evil those actions are
13:43 nrzkt rom1504 gitlab should use a system like a federation to manage the PR and the authentication ? like owncloud for example or XMPP
13:43 Calinou the FSF does make fundraisers, but they're a non-profit
13:43 Karlton Google openly spies on them, I avoid all their products even their free stuff.
13:43 Zeno` the FSF even accepts donations
13:43 Calinou they create ethical jobs, where employees use free software only
13:43 Calinou accepting donations is fine
13:43 Calinou selling stuff is, too
13:43 Calinou srifqi, gratis is Spanish
13:43 zero-ghost i love giving my money to people who deserve it
13:43 Calinou so is libre (but libre also is French)
13:43 zero-ghost google doesnt deserve any money they might make off of my existence
13:43 Zeno` anyone can declare their company as non-profit
13:44 Calinou calling the FSF evil is inconsistent when you think Google is fine
13:44 srifqi Calinou: oh, because it is also means "free" in Indonesian
13:44 Zeno` I didn't call them evil
13:44 Zeno` I said they make money
13:44 zero-ghost comparing the FSF to google
13:44 Karlton Zeno`: the issue isn't money, it is how you make it
13:44 zero-ghost you might as well have said something equally as insane
13:44 rubenwardy Hate me, but I'd love to work at Google.
13:45 zero-ghost watch i can play twist-the-logic game awesomely
13:45 Zeno` Karlton, and offering people developing open source projects to use your product for free is wrong?
13:45 zero-ghost your wife is a whore cause all girls cost money
13:45 zero-ghost boom
13:46 rubenwardy For me, open source etc is about sharing and collaborating, and hacking your software to make it fit your purposes
13:46 Karlton Zeno`: Only if it encourages the unethical practice
13:47 Zeno` Karlton, using a static analysis tool won't encourage anything (the output from nearly all of those tools is next to useless anyway)
13:49 Zeno` llvm and gcc cover almost all situations already
13:49 Zeno` but we probably shouldn't use llvm
13:50 Zeno` i.e. clang
13:50 Zeno` it's apple and evil
13:50 nrzkt clang is BSD licenced, it's good :D and it's not only apple, hopefully. GPL bleuarg :D
13:51 Karlton Only issue I have with clang is that you can make it ship with non-free addons
13:52 Zeno` nrzkt, but apple pays employees to work on it
13:52 Zeno` and IBM
13:52 Zeno` surely it cannot be ethical to use
13:52 Zeno` Apple might even have secret optimisations in their official version!
13:53 Zeno` It'd surprise me if they didn't, actually
13:54 Karlton Apple is hands down the most unethical company, so that is why people don't trust them
13:54 Karlton when it comes to software
13:55 srifqi is there any Apple's product that it is free?
13:58 Haudegen joined #minetest
13:59 Zeno` srifqi, well they contribute a lot to clang
13:59 Zeno` clang/llvm
13:59 Jousway joined #minetest
13:59 srifqi do they?
14:01 swaaws joined #minetest
14:03 Karlton srifqi: They do, and they basicly made it because you cannot extend GCC with non-free code
14:04 srifqi so, GCC's extension is non-free code?
14:05 Karlton GCC is GPLv3 so any extension has to be free whereas Clang is BSD so they can use it with non-free stuff
14:14 Tux[Qyou] joined #minetest
14:16 Zeno` joined #minetest
14:18 shadowzone joined #minetest
14:20 Sokomine i'm wondering how to best present a list of all the nodes present in a schematic so that players can select replacements for non-existing ones. a list might be practical, but an inventory-style view could be more intuitive
14:22 shadowzone joined #minetest
14:26 DMackey joined #minetest
14:26 someguy_irc joined #minetest
14:28 turtleman_ joined #minetest
14:44 xenkey So I'm making a whole city with the darkage mod
14:44 xenkey Figured a little house would merge my modern city and old city nicely
14:48 Sokomine the darkage mod is a very nice one. if you're finished, don't forget to at least post a screenshot!
14:50 xenkey Oh it's a 300x290 town
14:51 xenkey I don't think I'll be around by the time I can post a scrot :)
14:51 xenkey Sokomine: Have you build towns with it before? Perhaps you could help me since mine looks like someone ran over a carehome with a bulldozer.
14:52 proller joined #minetest
14:52 Sokomine xenkey: i usually don't build entire towns. it takes me enough time to finish lone houses :-)
14:52 Sokomine hmm. then you ought to spend more time on one house instead of making an entire town
14:52 xenkey I'm just making one house but it's meant to be a link between old and new
14:53 Sokomine ah. so make it a very nice one then :-)
14:53 xenkey Basically a gateway so it's not too rough a transition
14:53 Sokomine it's good that you think about those transitions. many people do not care about where they build which kind of building
14:54 xenkey I know that's true. The place I'm building it is now flat because I had to worldedit out many people's houses because they were unprotected and well... awful.
14:58 jin_xi joined #minetest
15:04 twoelk joined #minetest
15:05 Haudegen joined #minetest
15:05 hmmmm joined #minetest
15:09 SopaXorzTaker joined #minetest
15:12 swaaws-bnc joined #minetest
15:13 swaaws-bnc left #minetest
15:13 turtleman_ joined #minetest
15:16 Sokomine awful houses happen a lot. the classic is a box made out of cobble - complete with a steel door, a chest (not a locked one!) and a furnace inside. windows or a floor are considered excessive luxory and left away
15:16 proller joined #minetest
15:16 xenkey Yes exactly
15:16 twoelk xenkey, maybe add a park or small plaza or use anything else that is not like anything of the part you want to transition between. A lot of real life city planning uses that.
15:16 xenkey And all were about 5 blocks from the spawn borders
15:16 xenkey twoelk: You mean a third separate part?
15:17 Sokomine hi twoelk
15:17 ac|2 joined #minetest
15:17 twoelk yeah and hi :)
15:18 twoelk has the storm been through at your place yet Sokomine?
15:18 Sokomine twoelk: it's very stormy over here. hard to tell if the worst is over. there's still a lot going on
15:18 * twoelk is watching the gales bending the trees in front of his window
15:19 proller twoelk, road map online again
15:19 Sokomine same here
15:19 proller twoelk, also try "walk to infinity"
15:19 twoelk oh, hi proller, seems you have been reading the logs :)
15:20 fling joined #minetest
15:20 Sokomine proller: ah, fine. though a real building server would be a lot nicer...
15:20 proller just implemented again minetest compat mode in freeminer
15:20 twoelk nice to hear
15:20 proller Sokomine, i like survival and not boring exploration maps
15:21 Sokomine oh! a compat mode sounds very nice
15:22 Tablet_One joined #minetest
15:22 Sokomine proller: hmmm. that's not exactly mt's strength. though there has been some progress on mobs. it's quite impressive how far rnd got on his server where fighting mobs is the main objective. for some time, i thought that wouldn't be possible with this game. but it seems to work now
15:22 twoelk the problem though, proller, is that your maps somehow tend to evolve to pillar forests
15:22 proller its only near spawn
15:23 proller ~500m and everything ok
15:23 Sokomine spawn is the card you show to new players - how they'll think your entire server is. and a pillar-like spawn is pretty disgusting
15:23 proller Sokomine, mobs is very big problem and need to be solved in c++
15:24 proller give initial stuff make easier progress on very hard maps
15:24 Sokomine proller: fire up mt and check rnd's lab server :-) you're probably going to like it. the mobs there - although sadly almost only unfriendly ones - work pretty well
15:25 proller will try later
15:25 Sokomine no, maintaining a server requires a bit more than handing out material to players
15:26 Sokomine i don't like rnd's attitude in that regard much either. for me, vanessae maintains her servers best. best service one could expect :-) a truely safe, inviting place to build
15:28 proller buld server != survival server
15:29 proller Losing is fun!
15:29 Sokomine twoelk: i'm currently trying to create a newer version of my build chest for placing stored buildings. i'm not sure yet how best to present the list of used nodes so that replacements can be selected for nodes that are not installed (or where just another one is desired). list view would be the easiest and probably most practical - as long as people know the node names...
15:30 Sokomine proller: a server can be both - build and survie. though if there's a very good builder, nobody might complain if such a player would be switched to almost-creative :-)
15:35 alket joined #minetest
15:36 twoelk Sokomine, that nodes listing sounds very interesting. Might be usefull for a lot of things
15:40 Nytram joined #minetest
15:48 Sokomine the listing of nodes that make up a worldedit or mts file is not a problem. it just requires players to know about nodenames. maybe i ought to offer two views of the same thing: once as a list of nodenames, where the desired replacement can be inserted as another nodename, and one version where the nodes are shown inventory-style like. for the later one, i don't have a good idea yet as to how to show replacements
15:49 Sokomine the main advantage of such a build chest is that you don't have to worry about facedir though and that it'll rotate the building accordingly
15:50 twoelk wouldnT you need a database of all nodes and their orientations for that?
15:51 twoelk maybe make the list consider the nodegroup and suggest replacements according to that
15:51 twoelk so that a bush should not be replaced by a bench
15:52 twoelk or rather suggestions to avoid that
15:52 Sokomine that's too tricky i'm afraid. i can and do that within the villages themshelves - a lot gets replaced there according to installed mods - but it can't be truely universal
15:52 twoelk this could make the list of choice a little smaller
15:53 uroboro joined #minetest
15:54 xenkey Do you play any songs when you build?
15:55 Sokomine no. i don't play music
15:56 Jordach joined #minetest
15:57 Calinou http://cloudkickermusic.com/ | http://pharaohsandkings.bandcamp.com/releases
15:57 Calinou both of these
15:57 twoelk I have a piano in my unfinished manor on VenessaE's creative server but sadly I cannot play it.
15:58 xenkey :c
15:58 xenkey Shame
15:59 turtleman_ joined #minetest
15:59 * twoelk does like to fool a little around on his colecction of flutes though - when he's alone - and no one is near - all alone - best deep in the forest
16:00 Robby and underground, in a soundproof bunker
16:01 twoelk nah bad accustic there. best was up in the Asir mountains, that was eeirie but lovely
16:13 Ataron joined #minetest
16:13 Ataron_ joined #minetest
16:18 phantombeta joined #minetest
16:20 ElectronLibre joined #minetest
16:23 turtleman_ joined #minetest
16:26 ungali joined #minetest
16:26 * Calinou is working on a server
16:26 Calinou will run paragenv7, More Blocks, Carbone Mobs…
16:27 Sokomine hmm. you had one in the past, calinou, didn't you? it continued to exist for some time as one of oldcoders worlds
16:28 Calinou yes
16:28 Calinou this is a reboot, new world
16:28 Calinou will have areas too
16:28 Calinou I'm trying to keep it lightweight on mods
16:29 twoelk not using carbone?
16:29 Calinou nope
16:29 Calinou will have increased movement speeds (not as fast as Carbone, but faster than minetest_game) for better gameplay
16:30 Sokomine hmm. how long will it last?
16:30 Calinou until I'm fed up with griefing, I hope I won't be :)
16:31 Calinou it's a survival server btw
16:31 Calinou stealing from unlocked chests is allowed, too
16:31 Calinou so is PvP
16:32 twoelk hope you do keep some limits in place for fire and lava though
16:32 Sokomine ah. so a server of no intrest to me. unless pvp requires mutual prior agreement?
16:34 Calinou the PvP stuff may be changed in the future, depending on what players say
16:34 twoelk PvP can be fun if it is well balanced but most servers can not keep up the balance for a long while. they tend to tilt
16:34 Warr1024 joined #minetest
16:35 Calinou will have 3d armor so that you spend your iron on it :P
16:35 Calinou and to defend against mobs of course
16:35 twoelk btw there are some interesting analogies to power balance and development in reallife colonies that are pretty close to things that happen on came servers
16:36 turtleman__ joined #minetest
16:38 Haudegen joined #minetest
16:39 shadowzone joined #minetest
16:41 Calinou it apears in list, horray!
16:41 Calinou !server Calinou
16:41 MinetestBot Calinou: Calinou [FR] | 149.91.81.111 | Clients: 0/32, 0/0 | Version: 0.4.12-dev / minetest | Ping: 11ms
16:41 Calinou shutting it down, need to tweak config a bit
16:41 twoelk considering how gameplay evolved lately on casimir's ocean experiment and some developments on the just-test server I think a server should be carefull not to let stealing recources from other players become the best and easiest way to advance
16:43 Calinou well then…
16:46 DFeniks joined #minetest
16:59 Tablet_One joined #minetest
17:00 Robert_Zenz joined #minetest
17:02 Chanku joined #minetest
17:03 Calinou Sokomine, PvP is now only allowed in emergencies
17:03 * Sokomine nods to twoelk
17:03 Calinou ie. it's still enabled, you can fight back if you're attacked
17:04 Sokomine calinou: hmm. make it the same as on most other servers - ok if both parties agree to it beforehand. that's afaik the best solution. that way, players who really want to fight each others can do so - but others that don't want to can play as well
17:04 Calinou yes, you can also agree to do it
17:05 twoelk maybe make people with armor slower, so that you have a chance to flee :D
17:05 Sokomine no, that's a rule you've got to decide on :-) and which you have to tell your players. "emergency" might be a lot of diffrent things to players
17:06 Sokomine but armor's also good for fighting hostile mobs...
17:06 Calinou well, I'll note it on forum topic
17:07 Xenoth joined #minetest
17:10 Calinou Sokomine, can you connect to mt.nerdpol.ovh
17:10 Calinou to see if domain works
17:10 Calinou port 30000
17:18 WSDguy2014 joined #minetest
17:19 Krock joined #minetest
17:21 roniz joined #minetest
17:25 Jordach !up mt.nerdpol.ovh
17:25 MinetestBot mt.nerdpol.ovh:30000 is up (11ms)
17:25 Jordach Calinou, ^
17:25 martin____ joined #minetest
17:25 xenkey .ovh
17:25 xenkey thats a thing?
17:26 Calinou yeah
17:26 Calinou got it from nsupdate.info
17:32 Wuzzy xenkey: ICANN has invented a broad number of crazy TLDs some time ago. .ovh is a so-called brand-level TLD. Here's a list: https://en.wikipedia.org/wiki/Li​st_of_Internet_top-level_domains
17:33 Wuzzy There is even .foo, .bar, .xyz
17:33 xenkey crazy
17:33 xenkey what about .tar.gz
17:34 Wuzzy no
17:34 Wuzzy that's not a tld
17:34 Wuzzy TLD*
17:34 Wuzzy and .gz does not exist
17:35 xenkey shame
17:35 ]DMackey[ joined #minetest
17:36 Sokomine calinou: seems to work
17:36 Wuzzy same for me, I pinged it
17:39 Wuzzy Calinou: What other TLDs can I possibily get with nsupdate.info?
17:42 Calinou only two public domains right now
17:42 Calinou *.nerdpol.ovh and *.nsupdate.info
17:42 Calinou nothing shorter :/
17:51 sfan5 xenkey: i think there is .zip
18:04 turtleman_ joined #minetest
18:07 ndumont joined #minetest
18:09 Krock Lol. .wtf is a TLD too
18:17 Miner_48er joined #minetest
18:19 WSDguy2014 left #minetest
18:25 Haudegen joined #minetest
18:27 rickmcfarley joined #minetest
18:27 turtleman_ joined #minetest
18:31 Krock Calinou, str_replace("Stealing in unlocked chests", "Stealing from unlocked chests")?
18:32 Calinou ok
18:42 fusion44 joined #minetest
18:44 Calinou !server Calinou
18:44 MinetestBot Calinou: Calinou [FR] | 149.91.81.111 | Clients: 12/32, 4/12 | Version: 0.4.12-dev / minetest | Ping: 11ms
18:44 Jordach !server Jordach
18:44 MinetestBot Jordach: Jordach's Public Server | jordach.net:30001 | Clients: 1/15, 1/2 | Version: 0.4.11-dev / minetest | Ping: 8ms
18:44 sfan5 !server ping:least
18:44 MinetestBot sfan5: RND | server.neoascetic.me | Clients: 1/15, 1/3 | Version: 0.4.10 / minetest | Ping: 1ms
18:44 sfan5 a .me domain
18:44 sfan5 those are damn expensive
18:46 exio4 are they?
18:47 Calinou !server Calinou
18:47 MinetestBot Calinou: Calinou [FR] | 149.91.81.111 | Clients: 12/32, 4/12 | Version: 0.4.12-dev / minetest | Ping: 11ms
18:48 DMackey joined #minetest
18:53 est31 joined #minetest
19:00 sfan5 exio4: if i understand the price chart right, €17,31/yr
19:00 sfan5 (if you take >1 year)
19:00 sfan5 (otherwise it's €8,29)
19:00 sfan5 (https://www.namecheap.com/​domains/registration.aspx)
19:02 Nitori joined #minetest
19:11 rickmcfarley joined #minetest
19:12 Viper168 joined #minetest
19:12 DMackey joined #minetest
19:13 iqualfragile joined #minetest
19:16 turtleman_ joined #minetest
19:17 ]DMackey[ joined #minetest
19:21 DMackey joined #minetest
19:31 Calinou !server Calinou
19:31 MinetestBot Calinou: Calinou [FR] | 149.91.81.111 | Clients: 15/32, 6/17 | Version: 0.4.12-dev / minetest | Ping: 11ms
19:31 Calinou 15 players!
19:36 exio4 that was fast
19:37 est31 !server Trepca
19:37 MinetestBot est31: No results
19:38 est31 !server TREPCA
19:38 MinetestBot est31: No results
19:38 * twoelk decides, like a good sheep, to follow the herd
19:41 est31 ouch
19:41 est31 d0wnt1me
19:42 FR^2 joined #minetest
19:42 Calinou https://lut.im/Y2Spn59E/Hw0QrnMu
19:42 Calinou server map
19:43 est31 you should try https://github.com/est31/leaftest
19:46 msantana joined #minetest
19:48 paxcoder joined #minetest
19:49 paxcoder I've built the mandatory layers and the customary concrete layer around a reactor core, loaded it with uranium rods (1 in each slot), planted the switching station on top of the HV cable, and connected a HV battery. It says the core is "idle". How come?
19:49 paxcoder est31
19:50 proller joined #minetest
19:53 Viper168 joined #minetest
19:54 paxcoder VanessaE, any idea?
19:55 VanessaE because there needs to be more
19:55 VanessaE the core must be surrounded with water, that surrounded with *stainless* steel,
19:56 VanessaE then blast resistant concrete around that
19:56 VanessaE and the regular concrete on the outside layer, optionally.
19:57 paxcoder oh stainless?
19:57 VanessaE mmhmm
19:57 paxcoder is that new?
19:57 VanessaE nope
19:57 paxcoder :/
19:57 VanessaE been that way for a while :)
19:57 paxcoder but it was iron before?
19:57 rickmcfarley joined #minetest
19:57 paxcoder *normal
19:57 paxcoder that's it then, thanks.
19:58 VanessaE core in the center, 3x3x3 water around it (minus the bottom center, for the cable).  5x5x5 stainless steel blocks, 7x7x7 blast concrete, 9x9x9 regular concrete
19:58 paxcoder i have a hole top center
19:58 paxcoder hope that's ok
19:58 VanessaE in all cases, center bottom has a HV cable.  top center can have a hole for access to the core
19:58 VanessaE careful of the radiation :)
19:59 paxcoder but it'll run?
19:59 paxcoder i mean with a top hole
19:59 est31 yea
19:59 paxcoder k
20:00 ClaudeRoy joined #minetest
20:00 rickmcfarley https://m.youtube.com/watch?v=KfBwygvEJNA
20:00 rickmcfarley Dammit sorry for mobile
20:01 rickmcfarley No more than 2 holes and they can be anywhere
20:01 rickmcfarley Unless it changed
20:01 VanessaE yep
20:01 VanessaE and those two holes kinda need to be in the bottom and the top
20:02 rickmcfarley Oh nm
20:02 VanessaE (well I suppose they could be in the sides, but that makes accessing the core a little harder if you have too short of reach in your chosen game)
20:02 rickmcfarley Not if it's all under water
20:04 Elinvention joined #minetest
20:05 VanessaE if the hole is in the side, you won't have room to walk in a bit to reach the core, if your reach length is too short
20:05 VanessaE but from the top, you can climb in (you can't walk into a 1-node-tall hole, but you can climb down into a 1-node-wide hole)
20:10 rickmcfarley I see, I never go in unless it's on the top or bottom
20:10 paxcoder i can reach the reactor with just the hole in the customary layer
20:10 rickmcfarley Oh and the reach was usually far enough
20:12 rickmcfarley I never thought of that. Thanks VanessaE
20:16 VanessaE paxcoder: I was thinking of perhaps some custom game where your normal reach range is really REALLY short
20:18 VanessaE or if someone turns up the radiation of the active core to where you have to stand further away to safely access it
20:18 VanessaE (I forget if radiation in technic decays with distance)
20:18 est31 is that bad?
20:19 VanessaE ?
20:19 est31 its the same effect like the light of a torch, that also gets less the farther you stand away
20:19 est31 common physical principle
20:19 VanessaE ok, thought so
20:19 VanessaE been a while since the original discussion :)
20:20 iqualfragile_ joined #minetest
20:20 rickmcfarley Lol I've also died a lot due to radiation
20:21 VanessaE I *still* think the reactor code should be tweaked to allow for obsidian glass at the centers of the sides, as viewports
20:21 paxcoder VanessaE, why?
20:21 VanessaE why what?
20:22 paxcoder why were you thinking about a custom game where your normal reach range is really REALLY short (this was not copy pasted)
20:22 err404 joined #minetest
20:22 VanessaE OH, I wasn't thinking about making one, per se, but rather if someone should happen to
20:22 VanessaE i.e. just thinking ahead
20:24 rickmcfarley I think a radiation suit would be cool
20:24 jojoa1997 joined #minetest
20:24 est31 there are none in RL either
20:26 est31 those suits that protect you from radiation at the doctor do help, but not against radiation inside a nuclear reactor
20:27 paxcoder VanessaE, if you need ideas as to what to program from minetest, i've got one
20:27 VanessaE what's your idea?>
20:27 paxcoder a mod thoguh, not a minigame
20:28 VanessaE *looks at homedecor's TODO lists*
20:28 rickmcfarley Power suit?
20:28 paxcoder a "turtlebot" mode, kinda like simplerobots, but with logo as the programming language (except modified for 3D), and with external memory media to store programs on
20:28 VanessaE been done before.
20:29 VanessaE jin_xi: *poke*
20:29 VanessaE there's already a built-in language in minetest for the purpose
20:30 VanessaE L-systems
20:30 paxcoder does that L stand for LOGO?
20:30 VanessaE used by the spawn_tree() call, but can build any shape you want, using up to I think four nodes.
20:30 ]DMackey[ joined #minetest
20:30 paxcoder o.O
20:30 VanessaE naw, it's some guy's name, whoever it was who came up with it originally years ago, but to a computer it's basically 3d turtle graphics
20:31 VanessaE (it originated as a way to model how plants branch off and grow)
20:31 VanessaE anyway j_x here made a mod that lets you play around with the L-systems language in realtime
20:32 paxcoder i'll look it up. right now i'm busy digging for chromium
20:32 VanessaE here, have some ;-)  http://en.wikipedia.org/wiki/Chromium
20:32 VanessaE ok I'll crawl back into my hole now :P
20:32 * VanessaE <-- wise ass
20:35 OldCoder joined #minetest
20:41 Wuzzy paxcoder: No, it stands for “Lindenmayer”. https://en.wikipedia.org/wiki/L-system
20:42 Calinou !server Calinou
20:42 MinetestBot Calinou: Calinou [FR] | 149.91.81.111 | Clients: 9/32, 8/17 | Version: 0.4.12-dev / minetest | Ping: 11ms
20:42 TenPlus1 joined #minetest
20:42 jin_xi someone made a better one
20:42 jin_xi so dont use mine if its still around
20:43 paxcoder Wuzzy, interesting. VanessaE, but that looks nothing like Logo
20:43 VanessaE paxcoder: I know, but my point is it's the same basic idea
20:43 paxcoder these look to be fractals, ie generation not drawing
20:43 TenPlus1 hi folks, quick question for those in the know...  Stairs mod works fine but if I copy the stairs mod and place it in local mods and enable it I get errors...  "attempt to concatenate local 'subname' (a table value)"...   why ? it's the same mod with the same code
20:44 Wuzzy paxcoder: Yes, those ARE fractals. In Minetest, the L-system trees can also be fractals, but of course with finite iterations.
20:45 VanessaE paxcoder: LOGO is "repeat 3[left 10: forward 5: right 5: forward 30]" and so on..
20:45 VanessaE Lsystems does the same
20:46 Wuzzy yeah, but the “L” still stands for “Lindenmayer” and not “LOGO”. :P
20:46 Wuzzy btw there was a mod somewhere which uses some LOGO-like language
20:46 Wuzzy I guess it was called “turtle graphics” or something
20:46 Wuzzy no engine support of course
20:46 VanessaE only real difference between the two is you can't define variables, you can only have up to four "subroutines", and your left/right/pitch-up/etc turn commands are limited to one angle
20:47 jin_xi fact is that l system as a whole consists of the retarded axiom and stuff fractal 'code' and a simple turtle graphics turtle which draws the resulting mess
20:47 VanessaE (which is a hell of a big difference :)
20:47 ElectronLibre left #minetest
20:47 paxcoder VanessaE, i don't know lassie, sounds pretty difeferent to me
20:47 Wuzzy VanessaE: This difference is actually only a limit in Minetest and not inherent to L-systems
20:47 VanessaE Wuzzy: right
20:48 * paxcoder thinks Wuzzy's argument is the drawing DSL equivalent of the Turing completeness argument
20:48 Wuzzy umm, no
20:48 Wuzzy I don't care honestly. :D
20:48 fz72 joined #minetest
20:48 rickmcfarley LOGO was cool. I didn't think of it as programming in elementary school
20:49 Wuzzy Btw, does anyone in here made a cool L-system tree for Minetest and want to show it off?
20:49 Wuzzy I want to collect examples for the wiki.
20:49 VanessaE you mean besides moretrees? :)
20:50 paxcoder That's weird when people explicitly say they don't care. Not sure what to think.
20:50 Wuzzy I still do not understand why you did not reply to my thread, VanessaE. Maybe too lazy?
20:50 VanessaE did I miss the thread?
20:50 * paxcoder lols at that poke
20:51 VanessaE link please
20:51 Wuzzy let's find out
20:51 VanessaE I don't read the forums regularly anymore, just the few topics I am subscribed to
20:52 Wuzzy https://forum.minetest.net/vi​ewtopic.php?f=47&amp;t=11106
20:52 * TenPlus1 prefers schematics
20:53 Wuzzy Understandable, but schematics can't have randomness AFAIK.
20:53 VanessaE I feel kinda funny posting about the moretrees mod here...  I thought everyone already knew about it.
20:53 TenPlus1 paramat's new v7mapgen has schematic use with randomness in use
20:54 VanessaE besides, RBA did most of the models
20:54 TenPlus1 https://github.com/paramat/biomesdev
20:55 Wuzzy Well, I would be happy if you just post the treedefs. If you are too lazy, it's ok too.
20:56 iqualfragile joined #minetest
20:57 VanessaE there, posted the acacia model
20:59 Jordach Wuzzy, i found a few bugs with your mana mod
20:59 Jordach espcially one with hudbars
20:59 Wuzzy I listen.
21:00 Jordach Wuzzy, when hudbars is installed, it tries to remove mana's own text
21:00 Wuzzy Yay, thanks VanessaE.
21:00 Jordach which doesn't exist
21:00 Jordach and when players leave, causes a crash
21:00 VanessaE Wuzzy: you're welcome.  hope that's useful
21:00 Jordach if you have more than 200 mana, it resets it back to 200 on rejoin
21:01 Jordach http://paste.debian.net/164190/ <- fixed edition
21:01 Jordach (it also saves the mana.mt file on player leave so stats do keep)
21:02 TenPlus1 Question for devs:  why does stairs mod use abm's to rotate stairs and slabs when on_place = minetest.rotate_node works just as well ???
21:03 Wuzzy Jordach: Did you change the max mana? How did you manage to have more than 200 mana?
21:04 Jordach Wuzzy, i did mana.setmax(playername, mana.getmax(playername) + 25
21:04 Jordach ) *
21:04 Wuzzy hmm
21:04 Wuzzy weird
21:04 VanessaE TenPlus1: no idea.  maybe pre-6d-facedir conversions?
21:04 Jordach basically you set the max mana and other stats to NIL when they join
21:04 Jordach as well as setting player data to nil
21:05 Wuzzy OK, Jordach, I will look into it.
21:05 Jordach Wuzzy, that debian paste contains my fixes
21:07 TenPlus1 thx vanessa
21:09 kotaro joined #minetest
21:12 Wuzzy Jordach, I can't reproduce your bug.
21:13 Wuzzy I left the room with 225 mana, rejoined, still had 225 mana.
21:13 Wuzzy Did you play on singleplayer or on a multiplayer server?
21:14 torgdor joined #minetest
21:14 iqualfragile joined #minetest
21:23 frecel joined #minetest
21:36 shadowzone joined #minetest
21:38 Wuzzy Jordach: I just pushed a fix. Can you check for me if it works OK now? You can download a snapshot here: http://repo.or.cz/w/minetest_mana.git/snapshot/​1b3770dcdf2992e75684c96286410f442d8dd05f.tar.gz
21:39 turtleman_ joined #minetest
21:41 * est31 guesses Jordach knows git
21:42 SouL_ joined #minetest
21:42 H-H-H joined #minetest
21:47 ecutruin joined #minetest
22:05 phantombeta joined #minetest
22:06 err404 joined #minetest
22:24 Jordach joined #minetest
22:24 Tux[Qyou] joined #minetest
22:26 phantombeta joined #minetest
22:30 Hirato joined #minetest
22:40 Pilcrow joined #minetest
22:43 iqualfragile joined #minetest
22:45 Pilcrow I was thinking about mobs and had an interesting idea. current mob mods handle spawning by registering an ABM for the nodes a mob can spawn on. would it not be faster (i.e. less server work) to register a globalstep that iterates through connected players, generates a random number to check if it should spawn a mob, and then searches the nearby area for a place to spawn it?
22:47 Pilcrow or better yet, if it generates a single set of x/z coordinates, scans up/down about 5 nodes to find the surface, and then compares that surface node to a mob lookup table to see if there's one that can spawn there.
22:48 VanessaE nope, that would be slower
22:48 VanessaE because an ABM does almost the same thing
22:48 VanessaE abms only execute for the blocks near a given user
22:48 twoelk|2 joined #minetest
22:48 VanessaE globalstep executes continuously, as long as the server is running
22:55 Pilcrow well, I know it executes continuously, but if put on a timer of, say, 10 seconds (using dtime ofc), would it not be faster to center the calculations on a player versus having an ABM on dirt every 10 seconds? globalstep would have one random number per player to see <if> a mob should spawn nearby, versus 50+ random "chance" calculations (one per each dirt block in the area), followed by time of day calculations for each one that s
23:01 iqualfragile_ joined #minetest
23:02 Pilcrow ^ or am I overlooking something important, VanessaE?
23:03 VanessaE well that MIGHT be faster
23:03 VanessaE but the trick is looking for land to spawn on
23:04 VanessaE surely you'd want the 10-second interval to always be able to find a spot
23:04 VanessaE or at least to have a good chance
23:06 Warr1024 you can combine the strategies, too
23:06 Pilcrow like I said earlier though, I'm thinking it should calculate a single set of x/z coords and scan up/down by around 5 nodes to look for the surface. if the surface happens to be, say, gravel, and nothing is set to spawn on gravel, maybe generate a new set of coords? or maybe just end, and try again with a new set of coords on the next interval...
23:07 Warr1024 use the ABM to find good candidate spots, and write them into a list.  When the list is full, each new entry replaces and existing one
23:07 Warr1024 then you can use your globalstep timer to perform spawning from that list.
23:07 Warr1024 something like that would smooth out spikes if you needed to do spawning in bursts
23:09 Warr1024 tbh, I HAVE done the global timer thing before, and it's not so bad if the limiting factor is mainly a very expensive "eligible to spawn" check.
23:11 Pilcrow true, Warr1024. I was thinking of spawning a single mob per interval, but maybe spawning in bursts would be better. my other thought was, if the spawning algorithm is specifically focused on the players, mobs can potentially be programmed with no "player-searching" capabilities, as they could spawn tied to the player that they were generated for. this would prevent mobs rubber-banding between nearby players (unlike Simple Mobs)...
23:12 Warr1024 if you're just interested in adding mobs to add a bit of combat challenge, that's okay, I guess.
23:12 iqualfragile joined #minetest
23:12 Warr1024 I'd be more interested in mobs as a more complex gameplay mechanism, taking into account the limitations of their ability to observe their world
23:13 Warr1024 That'd be really hard to do scalably, though.
23:13 Warr1024 One other commonality I've seen in most mob frameworks is the use of entities.  I was actually more interested in the possibility of "staybs", i.e. creatures in node-space
23:16 Warr1024 cellular automata are already a very interesting gameplay mechanic
23:16 Warr1024 hm, guess I gotta go...
23:17 Warr1024 btw, Pilcrow, thanks for the upvote on that GUI anti-aliasing merge req.
23:17 Pilcrow yeah, the kind of complexity needed for environment awareness is beyond the scope of this idea, though; I'm thinking this would be nice for a lightweight hostile-mobs mod. the only reason I really I've ever want mobs in my minetest games is for the added challenge and danger of them.
23:18 Pilcrow I guess that comment was a bit late, lol
23:25 iqualfragile joined #minetest
23:37 kongque joined #minetest
23:39 iqualfragile_ joined #minetest
23:45 Enke joined #minetest
23:51 iqualfragile joined #minetest
23:52 Wayward_One joined #minetest
23:52 Player_2 joined #minetest
23:53 Pilcrow lol awesome name, Player_2. where's Player_1?  :P
23:55 Player_2 he is gone Pilcrow
23:56 Player_2 dead after glorious battle, or in search of glorious tacos, i do not know
23:57 Player_2 but none the less, he is not here
23:57 Pilcrow :P
23:59 Viper168 joined #minetest

| Channels | #minetest index | Today | | Google Search | Plaintext