IRC log for #minetest, 2015-06-06

Time	Nick	Message
00:01		jordan4ibanez joined #minetest
00:03	rjs232323	yes
00:03	rjs232323	My naming scheme are weird, I
00:03	rjs232323	I plan to correct it!
00:03	rjs232323	oh, I think I see why!
00:04	rjs232323	Lol I feel silly. when you mentioned that, I found out something wrong with it.
00:04	rjs232323	it's supposed to be pos after all!
00:05	rjs232323	well, it works perfect now
00:05	rjs232323	I feel stupid :C
00:15		Jordach joined #minetest
00:23		RealBadAngel joined #minetest
00:24		Wayward_Tab joined #minetest
01:01		Darcidride joined #minetest
01:08		Viper168 joined #minetest
01:13		rickmcfarley joined #minetest
01:14		clever joined #minetest
02:12		Ep1cMaN joined #minetest
02:13		Fritigern joined #minetest
02:25		LedInfrared joined #minetest
02:29		Pilcrow joined #minetest
02:30	Pilcrow	Hello again all! How are you doing tonight? :)
02:34	exio4	hi Pilcrow! long time no see!
02:35	Pilcrow	exio4: yeah, it's been a while. I've been away for a few days.
02:35	Pilcrow	anything new going on in the world of minetest? :D
02:36	exio4	I don't know!
02:37	exio4	just learnt about a ready-to-use really cool docker image
02:37	exio4	https://mail.haskell.org/pipermail/haskell-cafe/2015-June/119906.html
02:38	exio4	I just installed docker for trying it :P
02:40	exio4	I am going to sleep now, tho
02:40	exio4	good night Pilcrow :P
02:41	Pilcrow	aww, alright. sleep well, exio4.
02:43	exio4	it's too warm over here :(
02:44	exio4	25~ C, 26~ feels like, 80% humidity
02:46	est31	where do you live
02:47	exio4	Argentina
02:49		ndhelp joined #minetest
02:49	ndhelp	hi
02:49	Pilcrow	yeah, I have problems with heat, so I feel for you. anything over 21C (70F) is pretty uncomfortable for me...
02:49	ndhelp	my computer crashed
02:49	ndhelp	trying to resume my game
02:49	ndhelp	minetest gave mi this message
02:49	ndhelp	ServerEnvironment::loadMeta0:EnvArgsEnd not found
02:49	Pilcrow	and high humidity is always uncomfortable.
02:50	ndhelp	hhhhhhhhhhhhhhhheeeeeeeeeeeeeeeeeeellllllllllllllllllllllllllllllá¹•pppppppppppppppppppppp
02:51	ndhelp	hhhhhhhhhhhhhhhheeeeeeeeeeeeeeeeeeellllllllllllllllllllllllllllllá¹•pppppppppppppppppppppp
02:51	Pilcrow	ndhelp: calm down... can you please find your debug.txt and paste the last like 10 lines to http://dpaste.com/
02:52	ndhelp	ServerEnvironment::loadMeta0:EnvArgsEnd not found
02:52	ndhelp	done
02:53	ndhelp	http://dpaste.com/1DMZ1F3
02:53	est31	I guess that txt file is empty again
02:53	Pilcrow	which file, est31?
02:54	est31	envargs I think inside world folder
02:54	technomancy	Pilcrow: what up
02:54	est31	nope, env_meta.txt
02:54	Pilcrow	hmm. I don't know; it's not an issue I've ever encountered.
02:55	est31	yea the usual thing to do is delete that file
02:55	est31	most likely cause: a previous crash
02:55	Pilcrow	ah
02:55	Pilcrow	btw, hello technomancy
02:58	technomancy	Pilcrow: been kinda busy recently, but I think my next step with Calandria is to try to get the save/load logic of the server nodes' filesystem working better. it's still pretty shaky.
03:02	Pilcrow	technomancy: Yeah, that seems sensible. Also, does filesystem saving/loading work in the cli lua interpreter? looks like there's a couple of places where minetest.deserialize is used, so I'm guessing not.
03:05	technomancy	Pilcrow: correct; it's only implemented as part of the server node ATM
03:06	technomancy	at some point we'll probably want a save/load functionality that can fall back on an alternate serialization mechanism or something
03:06	technomancy	afaict minetest.serialize is just a prettyprinter
03:06	Pilcrow	est31: does that problem happen very often? I've had crashes quite a few times before, and I've never had that problem. I did have a problem long ago in like 0.4.5 where I lost my worldseed and it seriously screwed up the landscape though... :P
03:07	exio4	hai
03:07	exio4	I am still here, I don't know why
03:07	Pilcrow	lol exio4
03:07	technomancy	Pilcrow: did you see I replaced digiterm with digipad?
03:07	Pilcrow	exio4: I thought you were going to bed? just too hot and uncomfortable, I take it? :P
03:08	technomancy	simpler codebase, though I'm going to have to take some of the digiterm ideas and port them over to our digipad fork
03:08	exio4	no, it's too cold now
03:08	exio4	idk
03:08	Pilcrow	lol
03:08	technomancy	in particular, every digilines send/receive should be a table, not just a stirng, so we can include the name of the player and maintain more session data
03:08	technomancy	Pilcrow: https://github.com/minetest/minetest/issues/2743 is interesting too; perhaps if we switch to using an image button we can avoid the bug
03:12	Pilcrow	ah. yeah, an image button may work. I've never really tried to get a formspec to stay open when a player presses enter
03:14	technomancy	Pilcrow: digiterm "fixes" it by just redisplaying the formspec, but they screw it up somehow so that it becomes impossible to close without spamming the ESC key.
03:14	technomancy	so we don't want that
03:14	technomancy	gotta take off for a bit, will catch you later
03:14	Pilcrow	alright, see ya!
03:15	* Pilcrow	goes back to playing Pokemon, lol
03:19	exio4	\o
03:22	jordan4ibanez	I think people might like this new mod I'm brewing
03:22	jordan4ibanez	Juke boxes! \o/
03:23	Pilcrow	juke boxes? do they actually play music? :)
03:24		Xenoth joined #minetest
03:25	jordan4ibanez	Yes, and they're quite aesthetic, I think it'd add a nice feeling to the game for the player's that install the mod
03:25	Pilcrow	I think I need to work on getting some of my mods ready for release. I have so many that are just sitting here, mostly done, but I haven't shared them yet...
03:26	jordan4ibanez	Release them now! :P
03:26	Pilcrow	lol
03:27	Pilcrow	I have a few up on github right now, but so far I haven't posted any of them in the forums.
03:27	Pilcrow	I could probably release my hoverbot now...
03:30	jordan4ibanez	Hmmm
03:30	jordan4ibanez	Is there a way to get a particle spawner's position
03:32	Pilcrow	I couldn't say. particles are one thing I really haven't worked with at all. there /should/ be a way...
03:33		diemartin joined #minetest
03:51	Pilcrow	I take it exio4 is still here? at least, he hasn't signed out yet... :P
03:51	exio4	I don't sign out though
03:51	exio4	I am running my client on my vps
03:52	Pilcrow	heh, I didn't realize that. I'd imagine VanessaE does the same, as she's always signed in as well.
03:52	est31	unaffiliated/hacker????
03:53	Pilcrow	h4x0r?
03:53	est31	~whois exio4
03:53	ShadowBot	est31: exio4 (~exio4unaffiliated/hacker) has been on server sinisalo.freenode.net since 09:03 AM, April 25, 2015 (idle for 6 seconds). exio4 is on #minetest-dev and #minetest.
04:03	Pilcrow	~whois Pilcrow
04:03	ShadowBot	Pilcrow: Pilcrow (~Pilcrow97-90-230-132.dhcp.eucl.wi.charter.com) has been on server tepper.freenode.net since 04:29 AM, June 06, 2015 (idle for 1 second). pilcrow is on #minetest.
04:03	Pilcrow	^just wanted to know what it said for me, lol
04:07	exio4	yes, I am obviously a super h4x0r est31
04:07	exio4	since 09:03 AM, April 25, 2015
04:07	exio4	Pilcrow: that gives you an idea ^
04:07	Pilcrow	exio4 hacks in haskell. :P
04:08	exio4	ofc!
04:12	Pilcrow	I'm not dumb enough to try, but I wonder what would happen if you ran "dd if=/dev/zero of=/dev/sda" (assuming your os currently resides on sda)...
04:12	Xenoth	Pilcrow: Time to make a virtual machine!
04:12	Pilcrow	lol
04:13	Pilcrow	vm's don't really work very well on a 2.4Ghz dual-core laptop. and that's the srongest computer I own, lol.
04:16	Xenoth	2.53GHz quad here.
04:17	Xenoth	And they still don't run too well.
04:18	technomancy	Pilcrow: hoverbot is probably ready for release if you brush up the readme some more
04:18	technomancy	I really love the graphics for the hoverbot node
04:18	Pilcrow	I should really get a chromebook. I've been eying up those $250 ones with the quad-core nVidia Tegra K1 in them. I've got an nVidia Shield Portable (Android device) with the older Tegra 4, and it's a BEAST of a little machine... :P
04:19	technomancy	the Pixel ones look pretty fancy
04:19	technomancy	<3 400-nit screens for outdoor use
04:20	Pilcrow	technomancy: Yeah, I've always been horrible at documentation. You've probably noticed my hoverbot's code has almost no comments at all... What do you think should go into the readme?
04:23	exio4	I'd add high-level details on the structure
04:23	technomancy	Pilcrow: even though it feels redundant, a "how to install" section is good for newbies
04:24	exio4	oh, documentation for users not developers
04:24	exio4	nvm
04:24	technomancy	Pilcrow: screenshots of the drag-and-drop of instructions in action so people know what to expect
04:24	technomancy	Pilcrow: examples of what kinds of things you can program a hoverbot to do
04:24	technomancy	Pilcrow: known bugs section, future plans section
04:28	Pilcrow	the small readme I've given it does list install instructions and known bugs, but examples and future plans are good ideas to include...
04:30	technomancy	oh you're right; I was confusing your readme with that of Turtle
04:31	technomancy	IMO markdown readmes look nicer than .txt but that's your call =)
04:31	technomancy	screenshots would help a lot I think
04:32	Pilcrow	heh. markdowns annoy me a bit, but mostly because my laptop associates .md with megadrive game roms. I should probably fix that... :P
04:33	technomancy	haha
04:33		mccomase joined #minetest
04:33	technomancy	markdown is definitely a shitty format, but it's kind of a lingua franca, so I just roll with it
04:33	exio4	Pilcrow: megadrive?
04:33	exio4	you mean sega genesis!
04:34	technomancy	fwiw github supports README.markdown too I think
04:38	Pilcrow	eh, I'll probably eventually get around to fixing that file association, anyway.
04:38	Pilcrow	as idle curiosity though, I wonder if github supports other forms of markup. for example, a readme.tex? :P
04:39	technomancy	that might be asking too much, but there are lots of other formats that are supported that are nicer than markdown
04:39	bobomb	i suppose one could use position hashes to add and subtract positions?
04:39	technomancy	rst is supposed to be nice
04:41	technomancy	as is asciidoc
04:44	Pilcrow	interesting. I'm not really familiar with any of those, including TeX, but was just mentioning it for curiosity's sake... :P
04:44	technomancy	TeX is nice, but overkill for readmes
04:44	Pilcrow	lol
04:45	technomancy	I'm adding a bunch of comments to calandria's code
04:46	Pilcrow	probably. from what I've read, TeX seems to be sort of eqivalent in features to a .rtf document...
04:46	Pilcrow	only much more open
04:47		Wayward_Tab joined #minetest
04:47	Pilcrow	oh, technomancy, this is the chromebook I was talking about earlier: http://www.amazon.com/Acer-Chromebook-CB5-311-T7NN-13-3-inch-NVIDIA/dp/B00MHX6V88
04:47	Pilcrow	Sure, Acer are pretty run-of-the-mill, but if I put linux on this thing it'd still be better than anything I've ever owned... :D
04:47	technomancy	I couldn't handle that screen resolution =\
04:48	technomancy	not bad for the price I guess though
04:48	Pilcrow	1366x768? I've never had anything bigger, tbh
04:48	technomancy	thing is you can always SSH into a nicer computer to get work done, but you can't SSH into a nicer screen or keyboard =)
04:48	exio4	once you go 1080p, you don't go back
04:49	exio4	the size matters you know
04:49	exio4	size and resolution
04:49	technomancy	I'm still running on a thinkpad from 2009 simply because all the "upgrades" to the X series offered no more than 768 vertical pixels =\
04:49	technomancy	I guess last year they finally fixed that, but then they did that insane move with bundling malware, so yeah... that's not going to happen
04:52	Pilcrow	oh, I take that back. I did have a desktop computer that could do like 1440x960 or something, but my monitor's physical size was too small for that resolution to be anything more than an eye-strain for me... :P
04:53	technomancy	it's really handy for minetest to be able to have an editor open side-by-side with the game
04:53	technomancy	I guess you could use an external display for that too
04:53	technomancy	speaking of which, is there any built-in way to reload mod code? having to restart the server just to have your changes take effect is pretty ghetto.
04:53	* Pilcrow	likes running everything maximized/fullscreen, regardless of resolution
04:55	Pilcrow	technomancy: I know there is/was a mod that can execute arbitrary lua commands from the chat window. not sure if the same principal could be applied to a mod as a whole though...
04:58	Pilcrow	ah, I found that mod: https://forum.minetest.net/viewtopic.php?t=9671
05:01	technomancy	cool
05:01	Pilcrow	I don't think there's an official way to reload a mod's code on the fly though.
05:01	technomancy	I guess dofile(minetest.get_modpath("orb") .. "/init.lua") would do it
05:02	exio4	as long as you don't register nodes, etc
05:02	technomancy	exio4: I see; that part is not idempotent?
05:02	exio4	at worst it'll be UB
05:02	exio4	:P
05:02	technomancy	I've been careful to separate out all MT-specific code from the pure-logic stuff, so I should be OK for the most part
05:02	technomancy	UB?
05:03	exio4	undefined behavior
05:03	Pilcrow	ah. yup. node registry is done on server startup. that's also why you can't register nodes inside a minetest.after()...
05:04	technomancy	good to know, thanks
05:04	technomancy	is there a way to wrap that somehow?
05:04	technomancy	if(not minetest.node_registered(...)) then
05:04	Pilcrow	should be
05:04	Pilcrow	just a sec
05:06	Pilcrow	hmm. I know I've used a check like that before. just gotta find it.
05:10	jordan4ibanez	Jukebox https://youtu.be/tZc1sVJ5ylM
05:12	Pilcrow	ah. yeah. technomancy, you should be able to wrap your node registries with this: if not minetest.registered_items["mod:item"] then minetest.register_node(BLAHBLAHBLAH) end
05:12	technomancy	nice; thanks
05:13	exio4	I'd define a small function
05:14	Pilcrow	I haven't quite used it for that though; usually I just check that the mod exists, using if minetest.get_modpath("modname") then...
05:14	exio4	local.register_node(x)
05:14	exio4	s/local/utils
05:14	technomancy	should register_node do that though?
05:14		LazyJ joined #minetest
05:14		LedInfrared joined #minetest
05:14	technomancy	maybe with some kind of warning to let you know the second call was ignored?
05:15	exio4	define it locally
05:17	technomancy	Pilcrow: commmmmmmmmmmments! https://github.com/technomancy/calandria/commit/09b76106003704754f50c91aee9a28d7bac88b69
05:17	technomancy	Pilcrow: hopefully that helps explain some of the weirder stuff
05:18	Pilcrow	oh! yeah that should help a lot! :D
05:19	technomancy	in particular the metatable stuff is a bit weird
05:19	technomancy	in lua 5.2 they added support for customizing iterators using metatables, but we don't have access to that =(
05:20	technomancy	Pilcrow: one fun thing would be to start experimenting with having two servers communicate over digilines
05:21	technomancy	both with other servers and with "peripherals", other node types that can receive digiline signals
05:21	technomancy	I think I've got the foundations working to support that (fifo nodes in /digi/$CHANNEL/in and out) but I haven't gotten around to trying to use it for anything yet
05:22	technomancy	gotta head off for a bit; catch you later
05:23	Pilcrow	wait just a minute please
05:23	Pilcrow	this should do what you wanted (but is completely untested): http://dpaste.com/1YSCMP3
05:23	technomancy	nice; thanks
05:24	technomancy	want to make a pull request?
05:24	Pilcrow	not tonight, I need to head off as well
05:24	technomancy	ok, sure
05:24	technomancy	if you like I can just add you so you can push directly
05:25	Pilcrow	heh, I'd be too afraid I'd ruin things. I'm not very familiar with git.
05:26	jordan4ibanez	Goodnight
05:26	technomancy	the great thing about git is you can always roll back
05:27	Pilcrow	eh, I'm more comfortable just doing pull requests. I'll probably put a PR together tomorrow, after I've actually tried that piece of code out, lol.
05:30	Pilcrow	anyways, I've got a busy day tomorrow, so I'm heading off to bed early. goodnight, all.
05:31	exio4	good night
05:47		Krock joined #minetest
05:47	Krock	moin
05:52		OldCoder joined #minetest
06:04		Thron joined #minetest
06:09		CWz joined #minetest
06:11		Hirato joined #minetest
06:16		werwerwer joined #minetest
06:39		nore joined #minetest
06:43		Telesight joined #minetest
06:58		fusion44 joined #minetest
07:02		Calinou joined #minetest
07:13		Viper168 joined #minetest
07:19		est31 joined #minetest
07:20		Viper168 joined #minetest
07:25	kaeza	Asdf
07:25	est31	ghjk
07:25	CWz	l;'\
07:25	est31	7
07:26	est31	http://xkcd.com/1530/
07:27	est31	hows your internal skeleton doing ? :)
07:31	Calinou	http://phoronix.com/scan.php?page=news_item&px=Intel-SKL-BXT-Firmware-Blobs
07:32		err404 joined #minetest
07:33	est31	how is the license of these blobs compatible with linux?
07:33	est31	someone with rights on the code should sue
07:33	est31	"no reverse engineering, decompilation, or disassembly of this software is permitted."
07:36	Calinou	Linus Torvalds just doesn't care
07:36	Calinou	the same way, proprietary blobs are surely a violation of GPLv2
07:36	Calinou	(Fedora developers say so too)
07:39		Cryterion joined #minetest
07:43		Sokomine_ joined #minetest
07:50		]DMackey[ joined #minetest
08:05		Yepoleb joined #minetest
08:06		RealBadAngel joined #minetest
08:08		jin_xi joined #minetest
08:24		Cryterion joined #minetest
08:26		mazal joined #minetest
08:28		s-l-teichmann joined #minetest
08:40		Consortium joined #minetest
08:40	Consortium	Hi JamesTait
08:41	Consortium	... anyone here?
08:46		jin_xi joined #minetest
08:49	Consortium	jin_xi: hi
08:51		H-H-H joined #minetest
08:53	Consortium	Anyone who works for the developer wiki?
08:56		alket joined #minetest
08:58	kaeza	"Don't ask to ask, just ask."
09:00	kaeza	Well, gone anyway
09:00	* kaeza	goes back to pkmn
09:02	* sfan5	throws an eevee at kaeza
09:08	kaeza	http://i.imgur.com/5U5lZbf.jpg
09:12		Consortium joined #minetest
09:12		diemartin joined #minetest
09:13	Consortium	Anyone here???
09:13	est31	Consortium, do you want to edit developer wiki pages?
09:15	Consortium	est31 yes
09:15	diemartin	Calinou, ^
09:15	est31	get an irc client and join #minetest-dev, there you can ask celeron55
09:16	Calinou	kaeza, â€œelegir un pokÃ©monâ€ hey, I can read that :D
09:16	Calinou	I have Spanish exams in 15 days
09:16	Consortium	est31: i see [17:16] == celeron55 is away: screen detached
09:16	Calinou	Consortium, I can make you the account, PM me your e-mail address
09:16	Calinou	you are not supposed to PM celeron55
09:16	diemartin	nice :)
09:16	Consortium	Oh
09:16	Calinou	never do that, if you don't want to get your ass kicked :P
09:20		Consortium left #minetest
09:21	Calinou	C# benchmark up: http://dl.hugo.pro/benchmarks/
09:21	Calinou	it's 2nd, between Go and Java
09:30		Tux[Qyou] joined #minetest
09:33		Megaf_ joined #minetest
09:48	Krock	Calinou, where's c(++)?
09:53	Calinou	haven't got around to them yet
09:53	Calinou	will make Ruby too
09:53	Calinou	and perhaps Perl/PHP
09:54	Krock	next level would be a database access/write time
09:58	Calinou	which database, SQLite? LevelDB?
09:58	Calinou	we need one that has bindings to many languages :P
10:01	kahrl	Calinou: doesn't that benchmark mostly measure stdio performance?
10:03	kahrl	so languages that buffer i/o a lot will win because there's less context switches and less requests to the terminal to display new stuff
10:03	Calinou	yep
10:03	Calinou	also I was told the Go program outputs to stderr, and not stdout
10:03	technomancy	Krock: haha
10:03	Calinou	will fix that soon
10:03		Jousway joined #minetest
10:03	Calinou	it's not a very serious benchmark, but you can see patterns appearing
10:03	technomancy	Calinou: pretty sure that DB comment was a joke
10:03	Krock	It was serious
10:04	Krock	I thought about an SQL database
10:04	technomancy	oh
10:04	technomancy	Krock: you mean across databases, or across languages for a given DB?
10:04	Krock	across languages
10:05	technomancy	I thought the joke was because users usually blame slowness on the language, but in actuality performance issues are often bottlenecked on the DB, so the language doesn't matter.
10:05	Calinou	Java is actually pretty fast, I'm proving it yet again :P
10:05	technomancy	like people saying Minecraft is slow because of Java when really it's just using sloppy algorithms
10:05	Calinou	although in this case, it's slightly surpassed by C# and Go
10:06	Krock	just testing calculation power is not a good way to say how good a language is
10:06	technomancy	Calinou: most of the edge java has over C# is in the GC, which isn't used in this problem at all but makes a huge difference in application performance
10:06	Calinou	so Java is faster than C#?
10:06	technomancy	...
10:06	technomancy	java is a language, not a program
10:07	technomancy	one language cannot be faster than another language; only programs can be faster than other programs
10:07	Calinou	no, you mean that in complex programs, Java is faster than C#?
10:07	technomancy	no
10:07	technomancy	I mean Java's GC is better than C#'s GC
10:07	technomancy	some complex programs will be faster on the JVM and others will be faster on the CLR
10:07	Krock	different languages can generate different program code which is different in its speed
10:07	technomancy	you can't make sweeping generalizations about this at all
10:08	Calinou	at programs that do the same task, you kind of can
10:08	technomancy	.NET does most of its optimizations up-front, while the JVM does more of them by analyzing what's actually happening at runtime
10:08	Calinou	look how slow was the Python minetestmapper
10:08	Calinou	until someone optimized it with numpy, it was slow as hell
10:08	technomancy	so Java programs start off slow, and then speed up as the JIT discovers the bottlenecks and stuff
10:08	Calinou	the numpy one had decent speed; but then someone made a C++ one which blowed it out of the water
10:09	technomancy	if a program consists mostly of operations the C# compiler is good at optimizing up front, then C# will probably be faster
10:10	Calinou	I noticed than building the C# program took quite a bit of time, yes
10:10	Calinou	more than 1 second which is a lot for a simple program like this
10:10	technomancy	up-front optimizations result in slower compile times but lower runtime memory usage
10:10	Calinou	however the resulting binary is very small, 3 KB
10:10	Calinou	the stripped Go binary is 340 KB, and the Rust one, 312 KB ( Í¡Â° ÍœÊ– Í¡Â°)
10:10	est31	far too large
10:11	est31	written in assembler its less than 1 KB I guess
10:11	Calinou	est31, tehcnically the smallest â€œbinaryâ€ is the JavaScript one, 416 bytes
10:11	Calinou	but it's not a binary :p
10:12		Telesight joined #minetest
10:13	technomancy	also the C# and JVM binaries are not a standalone executables
10:13	technomancy	they rely on an existing runtime, while go and rust bundle their runtimes with them
10:15	est31	meh thats not standalone either
10:15	est31	you need hardware to execute it
10:15	est31	and an OS
10:16	est31	ok perhaps you dont need an OK
10:16	est31	OS*
10:16	est31	I've seen somebody write a space invaders clone as a grub module
10:16	est31	entirely os-free c program!
10:18	est31	there you can grab it -----> http://www.erikyyy.de/invaders/invaders-1.0.0.tar.gz
10:21	technomancy	I guess go and rust are probably still going to use libc
10:25		Telesight joined #minetest
10:27		LittleJoe joined #minetest
10:52		nore joined #minetest
10:53		Amaz joined #minetest
11:07		blaaaaargh joined #minetest
11:10		Ipos joined #minetest
11:10		kaeza_ joined #minetest
11:21		Ipos joined #minetest
11:22		Ipos joined #minetest
11:24	crazyR	can anyone think of a reason why textures wont apply them selves to a mesh node but they will apply them selves to there nodebox equivalent
11:25	blaaaaargh	crazyR, code?
11:26	est	protected, no fun :(
11:26	kaeza	;)
11:28	crazyR	https://gist.github.com/crazyR14/bf83ec0248011e175251
11:28		Telesight joined #minetest
11:29	crazyR	pretty much the same as the original
11:29	crazyR	the mesh object was exported using nodebox editor
11:33	crazyR	the bit that gets confusing is that the mesh node looks exacly as it is suppose to. but without the textures applied to it
11:33		kilbith joined #minetest
11:34	kilbith	crazyR, you need to open your model in Blender and create its UV-map for the texture appliance
11:34	crazyR	thats odd, because when i created the node as 1 full node it worked the way it was suppose to. and that didnt go throught blender. but thanks il try that
11:34	kilbith	mesh tiling does not work at all the same way than nodebox
11:36	kilbith	the .obj itself indicates where to place the texture(s)
11:48		WSDguy2014 joined #minetest
11:56	Calinou	can someone explain me how to rebase commits? ie. how to solve merge conflicts?
11:56	Calinou	I know about â€œgit rebase masterâ€ but it tells me something like:
11:56	Calinou	http://paste.ubuntu.com/11602984/
11:56	nore	Calinou, open the file that has a rebase problem
11:57	nore	you will find "<<<<" and ">>>>" in it
11:57	nore	where git couldn't rebase properly
11:57	nore	then, fix the issue
11:58	nore	and finally, do git add <filename> and git rebase --continue when all files have been fixed
11:58		Haudegen joined #minetest
11:58	Calinou	I rebased it by hand, by skipping the patch
11:58	Calinou	I had to do it anyway due to river water's post effect colors
11:59	nore	that works too :)
12:09		ac_minetest joined #minetest
12:14	MinetestBot	[git] Calinou -> minetest/minetest_game: Improve water post effect color 750f957 http://git.io/vIswN (2015-06-06T14:13:01+02:00)
12:33		ecutruin joined #minetest
12:44	kilbith	!seen Zeno`
12:44	MinetestBot	kilbith: zeno` was last seen at 2014-12-22 23:03:51 UTC on #minetest
12:45		WSDguy2014 joined #minetest
12:45	est	!seen kilbith
12:45	MinetestBot	est: kilbith was last seen at 2014-12-23 17:10:00 UTC on #minetest
12:46	Krock	~seen Zeno`
12:46	ShadowBot	Krock: I saw Zeno` in #minetest 32 weeks, 2 days, 20 hours, 43 minutes, and 4 seconds ago saying * Zeno` is sorry for being angry but will not apologise for defending people's rights to do what is permitted under the LGPL and GPL and the spirit of those licenses
12:46	kilbith	broken bot
12:48	Amaz	s/bot/bots
12:53	MinetestBot	[git] LeMagnesium -> minetest/minetest_game: Improved flowers' registration system - Specific nodeboxes, - Color groups - All datas were put into a table then unpacked for add_simple_flower 01ad090 http://git.io/vIsD8 (2015-06-06T14:51:25+02:00)
12:53	MinetestBot	[git] quentinbdgmail.com -> minetest/minetest_game: Simplified flowers' registration - Created a new local function to register flowers 638add6 http://git.io/vIsD4 (2015-06-06T14:51:25+02:00)
12:57		ThronL joined #minetest
12:57	exio4	Calinou: the benchmark is still flawed
12:58	exio4	Calinou: also, you're counting the JVM `warmup`
12:59	exio4	you haven't ''benchmarked'' the haskell examples I posted yesterday :P
13:00		THR joined #minetest
13:00	exio4	that's also single-threaded
13:00	exio4	I could naively use 8 or more threads
13:02	est	yes, but to be fair other languages should be abled to do that too
13:03		Hirato_ joined #minetest
13:10		Xenoth joined #minetest
13:11		Xenoth joined #minetest
13:12	Calinou	I don't have enough knowledge to write a multi-threaded generator :/
13:13	Calinou	what would be a good first language to try that, Rust?
13:14		Haudegen joined #minetest
13:22		meldrian joined #minetest
13:22		NekoGloop joined #minetest
13:24		Viper168 joined #minetest
13:26	exio4	est: it's benchmarking IO though
13:26	exio4	not performance
13:26	exio4	Calinou: I would suggest Haskell ;)
13:26	exio4	est: most languages make parallelism a pain anyway
13:26	Calinou	can you give me your Haskell generators again? I no longer have the link
13:27	Calinou	Haskell is multithreaded by nature, right?
13:27	exio4	http://dpaste.com/0289NVX no idiomatic, http://dpaste.com/01BRM98 idiomatic
13:27	exio4	Calinou: not really
13:27	exio4	the default runtime only uses one~ core, even!
13:27	Calinou	which one is faster for you?
13:28	exio4	it's just that making the code parallel is easier than in most languages
13:28	exio4	Calinou: hm, the non-idiomatic version should be at first look, but if you compile the idiomatic version with -O2 I am pretty sure you will be removing the intermediate lists and getting a naive tail recursive loop as result
13:28	exio4	Calinou: try both, it may be different with your optimizations
13:29	exio4	I'd go with the idiomatic version
13:31	exio4	at worst, it shouldn't be that slower, list fusion does half the trick :)
13:31	exio4	(I would suggest install `parallel`, adding a few annotations, but that'd be using a dependency (which in the haskell world with hackage is pretty common, for pretty much anything non trivial))
13:32	exio4	installing*
13:33	exio4	also, if you are brave, you could use the inline-c package and write the prime checker in C, and just use Haskell as a middleware :P
13:35	Calinou	readability of your code is abysmal though :P
13:35	Calinou	is it like that for all Haskell code? ._.
13:35	exio4	Calinou: the idiomatic version is pretty clear
13:36	exio4	https://github.com/sseefried/open-epidemic-game/blob/master/src/GameM.hs#L50 this is some code I was reading yesterday, I liked it
13:36	exio4	(it's mostly boilerplate there, though)
13:37	exio4	https://github.com/sseefried/open-epidemic-game/blob/master/src/Game.hs this one is using all that fancyness
13:39	exio4	Calinou: as an example, I ran the idiomatic version of my Konsole, and I got 4247687
13:40		Wayward_Tab joined #minetest
13:40	exio4	simple things like switching out of the window changes lots of things, and try benchmarking it on the TTY and then in xterm
13:41	Calinou	will try
13:42	exio4	I ran that on my TTY, with a little tweak, doing sleep 1; and switching to X11 before letting it run, and now it is 4.7M
13:42	exio4	note, too, that this isn't top high-end hardware either
13:42	Calinou	exio4, in tty I get 219k with the Go program, then 213k
13:43	exio4	wait, are you _looking_ at the TTY?, I meant sleep 1; BENCHMARK and before letting the benchmark run, switch to other tty
13:43	exio4	and wait the 10 seconds
13:43	Calinou	248k in xfce4-terminal
13:43	Calinou	yes I was looking at the tty
13:44	exio4	don't, and you'll get different results
13:44	Calinou	245k when not looking
13:44	Calinou	trying againâ€¦
13:44	exio4	is that giving you an idea of it?
13:44	Calinou	250k
13:45	Calinou	it might be a tad faster
13:45	exio4	the benchmark is flawed by design, standard output isn't reliable
13:45	exio4	it's a virtual bottleneck you are adding for the sake of it
13:45	est	rol really funny, discussion on two channels right now is about the benchmarks :)
13:45	est	or more, dunno :)
13:45	exio4	I'd find a prime number that's in the 50 million range
13:46		SopaXT joined #minetest
13:46	exio4	and then see how long it takes to find it is a prime number
13:46	exio4	hm
13:46	THR	do MT have mem leakage.. some time my not so poor PC run into 3gig mem usage, then all go in LAgggg
13:46	exio4	more like in the 20M ^ 2 range
13:47		basxto joined #minetest
13:48	Calinou	est, successful troll ( Í¡Â° ÍœÊ– Í¡Â°)
13:48	Calinou	exio4, yeah, what can I do is look at how much time it takes to find a few large prime numbers
13:49	Calinou	eg. starting from 10 million
13:49	Calinou	I can just change nb variable for that
13:49	Calinou	that'll break all my scores though :(
13:49	THR	im not a troll, need help with mem leakage :\
13:49		blaise joined #minetest
13:49	exio4	you seem to care more about how good lucking the scores are more than how accurate they are
13:49	exio4	gg
13:50	blaise	there's a score in minetest?!
13:50	est	THR, the fact that minetest has high memory requirements in certain cases is known
13:51	est	try the following: if you have hd texture packs, disable them
13:52	exio4	blaise: Calinou's benchmark is flawed, he is trying to get a benchmark such that it is more good looking than how accurate it is
13:52	exio4	Calinou: you're also benchmarking compiler's quality
13:52	THR	but it happen just sometime, and after a while. Im building on one spot and then it happens to.
13:52	blaise	ah
13:52	exio4	Calinou: there's nothing avoiding a supercompiler :)
13:52	est	exio4, is benchmarking compiler bad?
13:52	exio4	est: yes, specially if you claim you are benchmarking programming languages
13:53		s-l-teichmann joined #minetest
13:53	exio4	because programming languages by themselves do nothing
13:53	exio4	they are just ideas
13:53	blaise	so like, are there any languages which compile and execute at execution?
13:53	exio4	as an example, .NET and Mono
13:53	est	yes exio4 thats why lua and luajit have two entrie
13:53	exio4	blaise: implementation of languages, that'd a be JIT
13:53	est	s
13:53	Calinou	I'm benchmarking implementations
13:53	est	^
13:53	exio4	est: but then why does the title says programming languages?
13:53	Calinou	I'll correct the page title to reflect that
13:53	blaise	I believe BASIC also does it as well, no?
13:53	exio4	:)
13:53	Calinou	I easily overlook this
13:54	THR	exio4, when you going to recode the Graphic engin.. that my Nvida can use it :)
13:54	s-l-teichmann	Calinou: Try this one for a concurrent Go version. http://pastebin.com/3sEbsKP6 Hacked it right down, Haven't tested it well, but should be a starting point.
13:54	exio4	it's also silly, benchmarking IO
13:55	exio4	s-l-teichmann: concurrent? why not parallel? ;)
13:55	blaise	some languages arnt compiled.. they're written in their compressed and compiled form..
13:55	blaise	like ASM
13:56	Calinou	impressive, s-l-teichmann
13:56	Calinou	however it does not return all primesâ€¦
13:56	exio4	is the order important?
13:57	s-l-teichmann	exio4: you know Pikes position on concurrency vs parallel ;-)
13:57		Consortium joined #minetest
13:57	Consortium	Hi
13:57	blaise	HI!
13:57	exio4	s-l-teichmann: Pikes? I was talking about Marlow's!
13:57	s-l-teichmann	Calinou: You can remove if prime > highest then you get them all (in abitrary order)
13:58	s-l-teichmann	exio4: Rob Pike
13:58	exio4	yes, I know
13:58	blaise	anyone tried this mod? https://forum.minetest.net/viewtopic.php?id=4587
13:58	est	blaise, it works
13:59	blaise	are there problems?
13:59	blaise	does it work with minetest-0.4.12 ?
13:59	est	yea some design problems
13:59		mecatninja left #minetest
13:59	exio4	s-l-teichmann: https://ghcmutterings.wordpress.com/2009/10/06/parallelism-concurrency/
13:59	est	for example, there is only an audio feedback if you extend the force field
14:00	exio4	s-l-teichmann: Go's approach to concurrency is pretty crappy though
14:00	blaise	:G
14:00	exio4	dat sharing of mutable data
14:01	exio4	`It is a statically-typed language with syntax loosely derived from that of C, adding garbage collection, type safety, some dynamic-typing capabilities`
14:01	exio4	am I the only that sees the oxymoron there?
14:03		eeew joined #minetest
14:05	exio4	dat unsafe concurrency http://en.wikipedia.org/wiki/Go_%28programming_language%29#Lack_of_race_condition_safety
14:06		Someguy123 joined #minetest
14:06	exio4	dat lack of first-class tuples, too
14:07	Consortium	Hello...
14:07	blaise	HI!
14:08		Ipos_ joined #minetest
14:08	blaise	HowAyaDoin?
14:09	* blaise	mines
14:09	Consortium	http://dev.minetest.net/Feature_and_mod_roadmap Doing this for the common good
14:12		ThronL joined #minetest
14:15		Cryterion joined #minetest
14:18	Consortium	blaise: did i do good?
14:18		AnotherBrick joined #minetest
14:21	Ritchie	Consortium: wtf is this? are you serious?
14:21	Ritchie	Consortium: do you want all this in minetest?
14:21	est	its more a brainstorm wishlist than a roadmap
14:21	Consortium	Ritchie: well, yeah... i mean if we call collectively gather as much idea as possible, its gonna grow fast
14:22	Consortium	est: I know, it needs a lot of cleaning to do
14:22	Ritchie	1) weather.. minetest is engine, not game, for game with weather write mod which does weather
14:22	Ritchie	2) textures.. there are many texture packs
14:23	Ritchie	3) there are many mods for this, for example money, economy or write some mod
14:24	Consortium	Ritchie: once i recieve a message like http://dev.minetest.net/Talk:Feature_and_mod_roadmap
14:25	Ritchie	4, 5, 8.. why complicate minetest with that mess? i can play Wurm online if i want it
14:25	Consortium	Ritchie: i was expecting a minecraft replacement
14:26	Ritchie	minetest is not minecraft clone and it hasn't a goal to be MC clone
14:26	exio4	about 5, I would agree only on reworking the health system to be more FPS-like
14:26	Ritchie	9) mesecons..
14:26	exio4	the problem would be breaking hella lot of mods
14:27	Consortium	Well how so, Ritchie ?
14:27	exio4	most things in that list are already mods, too
14:27	Ritchie	exio4: hunger mod.. why do a mess with this in MT engine?
14:27	Consortium	exio4: but is it possible to make mods collision free?
14:28	exio4	Ritchie: no, the health system, making it more FPS-like meant using base-100 health
14:28	exio4	Ritchie: and reworking the HUD to something less MCish
14:28	exio4	the MC approach seems pretty dumb to me, it's not fine-grained enough, and it disallow lots of fancy things like slow regen over time
14:29	exio4	base-100 would mean you could "take is a percent"
14:29	Consortium	Ritchie: i see that it has pickable options in a large library. Because MT can be used as a base for FPS, MMO/MOBA etc.
14:29	exio4	not really, it lacks the client-side prediction for a FPS
14:29	exio4	and without being designed as a MMO, it would suffer from a really lot of players :)
14:30	Calinou	I bet Consortium was going to do silly stuff on dev wiki :/
14:30	Calinou	this is not appropriate for mainspace; I'll move it to userspace
14:30	Consortium	Calinou: just a long list. Nothing more.
14:30	Ritchie	Consortium: MT can be used as a base for many games only if it doesn't contain weather and similar mess, this can be modded in a subgame
14:30	Consortium	Calinou: thanks
14:30	Calinou	also don't put your e-mail on wikip ages
14:31	exio4	that, there's reason the talk page is there
14:31	Calinou	use your user talk page for contact purposes
14:31	Consortium	K, where is the userspace link, Calinou
14:31	Calinou	http://dev.minetest.net/User:Consortium/Feature_and_mod_roadmap
14:32	Ritchie	exio4: i have no problem with the current state with health bar.. base-100 health can be modded, cann't?
14:32	Consortium	Calinou: could i ask a question?
14:32	exio4	Ritchie: no, it can't
14:33	Calinou	yes, don't ask to ask
14:33	exio4	Ritchie: what I was thinking was about copying the health management to a FPS like redeclipse
14:33	exio4	Ritchie: base-100 would mean you could "easily" tweak the health to have incremental +2.5% health as an example
14:33	Consortium	Why are mods not standardized in one single programming language or script type, Calinou
14:33	exio4	they are all written in Lua
14:34	exio4	that's a single programming language :)
14:34	Consortium	Okay. But i heard the mods might be written in Python or others, sometimes
14:34	exio4	what do you mean by `script type`, though?
14:34	exio4	Consortium: not really, anything like that is a nice hack
14:34	Calinou	all Minetest mods use Lua
14:35	Calinou	the IRC mods relies on an external .so library, which is written in C and Lua
14:35	Ritchie	exio4: so play redeclipse and don't take this to MT
14:35	Calinou	(Lua itself is written in C.)
14:35	Consortium	exio4: i heard from the grape vine that there is no regulation in mod scripting, some might use Python, Java etc...
14:35	Calinou	you can use languages that compile to Lua like MoonScript
14:35	Calinou	but currently, no one made a popular mod programmed that way
14:35	Calinou	maybe there's one small mod in MoonScript and that's about it
14:35	exio4	Ritchie: not really, Redeclipse is a FPS not a sandbox game :)
14:35	Consortium	Calinou: that was close.
14:36	Consortium	I think Minetest need a larger command prompt library e.g. teleport, factions etc.
14:37	Calinou	the standard library is large enough currently
14:39	Consortium	Calinou: Minetest is bad, right?
14:39	Consortium	Bad as in not "full enough"
14:40	Calinou	exio4, with your idomatic Haskell prime generator, I get â€œmain: Prelude.head: empty listâ€ after running the main binary (ghc main.hs)
14:41	Calinou	non-idomatic works
14:41	Calinou	I'll just use non-idomatic for now
14:41		Telesight joined #minetest
14:41	Calinou	it's incredibly fastâ€¦ and does not miss primes :o
14:41		jojoa1997 joined #minetest
14:41	exio4	Calinou: you have to give it -1 as a parameter
14:42	exio4	Calinou: wait, I think I know the problem
14:42	exio4	Calinou: are you going until sqrt n or until n-1?
14:43	Calinou	nb_test goes up to n-1
14:43	exio4	O(n^0.5) is .. different to O(n)
14:43	Consortium	exio4: what algorithm did you use?
14:43	Calinou	one sec, I'm putting the code up
14:43	exio4	it's just the naive approach
14:43	exio4	Calinou: wait, my code is doing the wrong thing
14:44	exio4	Calinou: it's going up to sqrt n, not n-1
14:44	Calinou	in non-idomatic?
14:44	exio4	both
14:44	Calinou	well that's good, it's fasterâ€¦
14:44	Calinou	I'll change other languages so that it works the same way
14:44		Haudegen joined #minetest
14:44	Calinou	but does it detect all primes?
14:45	exio4	yes'\
14:45	exio4	it's one of the most trivial optimizations :P
14:45		Cryterion joined #minetest
14:45	Calinou	thanks, I'll have to rerun all benchmarks :P
14:45	Consortium	exio4: how large can the numbers get? 2^32 or 2^64?
14:46	exio4	it's using machine integers
14:46	exio4	you could change the type, or remove it, and let the compiler use gmp (Integer, BigInt, `infinite` .. you know)
14:47	Consortium	exio4: so beyond BigInt?
14:47	Calinou	I don't need large numbers here, uint32 or equivalent is large enough here
14:47	exio4	uint32 numbers are too cheap to check though
14:47	Consortium	Calinou: do you teach Lua? I want to learn Lua
14:48	Calinou	yes, it'll be three fiddy
14:48	exio4	lol
14:48	Consortium	Damn
14:49	Consortium	Because i want to learn Lua ASAP (i know how loops and arithmatic works, so bring it on)
14:49	exio4	what's your background in programming
14:50	exio4	Lua is meant to have a short learning curve, given that you have experience on statement based imperative languages
14:50	est	https://www.youtube.com/watch?v=CvjcuiQE-cQ
14:50	Consortium	exio4: beginner in Java. a=+2, if else, ...
14:51	exio4	beginner in Java here means what? have you tried to learn Lua by yourself?
14:51	exio4	if you are stuck, you can ask for help
14:52	est	exio4 knows languages
14:52	exio4	I don't
14:52	Calinou	yesterday and today I learned Go, Rust, Java, C#, JavaScript, Ruby, â€¦
14:52	Calinou	;)
14:52	exio4	you didn't learn them
14:52	Calinou	was slightly sarcastic
14:52	exio4	you already knew the semantics of the constructs, you learnt new syntax
14:53	Calinou	yeah that's about it
14:53	Calinou	sometimes the line numbers are even corresponding
14:53	exio4	learning new syntax is dead easy
14:53	exio4	you don't need to learn anything, you just get used to it
14:53	exio4	concepts, though, those are a bitch!
14:53	Consortium	exio4: i know how to set variable, do maths and create loops. Not so good in array and command line
14:53	Calinou	I like how rustc throws meaningful errors, with colored output
14:54	Calinou	that's a good evolution over gcc
14:54	exio4	Consortium: you might want to take a course in coursera, go with code academy, [...]
14:54	exio4	Calinou: gcc 5.1 does that too!
14:54	exio4	anyway, most "meaningful errors" in Rust are cause it has lots of shiny things in its type system
14:55	exio4	:D right?
14:55	Calinou	I wonder why some people call Rust a replacement of C++, because it's slower, at least in single threaded operations
14:55	Consortium	exio4: links?
14:55	Calinou	it'll beat it for sure if you can multithread it, but you can't always multithread stuff
14:56	Calinou	not to mention multithreading reduces battery life
14:56	exio4	there's no static-typed imperative language with high order functions and garbage collection with non crappy syntax out there apparently :(
14:56	Calinou	what don't you like in Go and Rust?
14:56	exio4	Consortium: link to what? coursera? code academy? you can use your favorite web searcher
14:57	exio4	Calinou: I dislike the lacks of generics, and how concurrency is still unsafe, in Go
14:57	exio4	Rust, I heard it is pretty good and I liked the little bit I did
14:57	exio4	but I haven't used it
14:57	exio4	like, actively
14:57	Consortium	exio4: i know coursea but... which code academy?
14:58	exio4	Calinou: Rust a "C++ replacement" comes from the fact that Rust has static deterministic memory management, no garbage collection!
14:58	exio4	systems programing, you can write a garbage collector in Rust
14:58	exio4	!g codeacademy
14:58	MinetestBot	exio4: http://www.codecademy.com/
14:58	exio4	Consortium: ^
14:58	Calinou	exio4, changed my benchmark to start from 1000 billion
14:58	Calinou	I'll count the offset from 10 billion instead
14:58	Calinou	1000 billion*
14:58	Consortium	Coursea has no Lua
14:59	Calinou	Consortium, Ruby is probably the closest you can get because of its â€œendâ€ keywords and such
14:59	exio4	Consortium: in the code, there's `loop 2`, change that to loop 1000...
14:59	exio4	Calinou: ^
14:59		OldCoder joined #minetest
14:59	exio4	Consortium: sadly, learning the concepts will make learning Lua easier though :)
15:00	Calinou	exio4, that makes it much slower, for no benefit?
15:00	exio4	Calinou: what makes it slower?
15:00	exio4	starting from 1000B?
15:00	Consortium	exio4: learning Ruby?
15:00	exio4	yes, because it won't try the smaller primes
15:00	Calinou	exio4, how is loop 1000 going to start it from 1000 billion?
15:00	exio4	Calinou: no, that's why I used ...
15:01	exio4	Calinou: which "billion" are we using here?
15:01	exio4	1000^3 or 1000^4 ?
15:03	Calinou	10^12?
15:03	exio4	^3 then
15:03	exio4	(10^3)^4 :)
15:04	exio4	do loop (10^12) then
15:07	Calinou	meh
15:07	exio4	what?
15:08	Calinou	it doesn't work in all languages, and makes risk of overflowing higher
15:08	exio4	overflowing? if we're using 64bit, it shouldn't overflow
15:09	exio4	it's bigger than an int32 though
15:09	exio4	and you could benchmark the GMP bindings :)
15:09	s-l-teichmann	Calinou: For your gallery http://pastebin.com/ATc69bEF (all primes, correct order) and I've put also a 'goto' in it hoping that exio4 dies an early haskellish death. ;-)
15:09	Jordach	lmao
15:10	exio4	don't worry, I can use continuations!
15:10	exio4	s-l-teichmann: 1 ain't prime!
15:11	Calinou	fmt.Println("1")
15:11	Calinou	fmt.Println("2")
15:11	Calinou	hahaha
15:11	exio4	Calinou: I also hardcoded 2 as an exception :)
15:11	s-l-teichmann	Calinou: I've never said that it only issues primes. ;-)
15:12	Calinou	seems to print all primes, s-l-teichmann
15:12	Calinou	however how can I optimize it so that it stops checking at the square root of the number?
15:12	Calinou	your variable names are cryptic :s
15:13	s-l-teichmann	Calinou: feel free to rename them
15:14	Calinou	how to use square root in Go?
15:14	blaise	lol
15:14	Consortium	exio4: could you make a mega folder of all primes under 2^64, with each file having the same number of primes?
15:15	Consortium	>The most extreme challenge ever
15:15	s-l-teichmann	Calinou: math.Sqrt http://golang.org/pkg/math/#Sqrt
15:18		Player_2 joined #minetest
15:19		Tux[Qyou] joined #minetest
15:20		Tux[Qyou] joined #minetest
15:21	Calinou	s-l-teichmann, now with this optimization, your multithreaded one isn't any faster :/
15:22	Consortium	exio4
15:23	s-l-teichmann	Calinou: If I would take this benchmark serious that would be the part where i would take out the profiler.
15:24	s-l-teichmann	Calinou: gccgo may result in slightly better code as gc
15:27	Consortium	Calinou: what game is MT suppose to be? Is it a survival sim?
15:28	Calinou	no, not at all
15:28	Calinou	it is a sandbox game
15:29	Consortium	Calinou: but sandbox for what purpose? or is it multi-purpose?
15:30	VanessaE	Consortium: "Minetest" is just a sandbox game engine to run other stuff. "minetest_game" is a very basic subgame mostly good for modding. There are many other subgames as well, that do other stuff.
15:30	VanessaE	some are geared toward survival, some toward creativity, some do both..
15:31	Consortium	VanessaE: but the fanbase is not as big and polished as the Minecraft community...
15:31	VanessaE	so?
15:31	CWz	less whiny kids
15:31		Cryterion joined #minetest
15:32	Consortium	Minecraft has a modding standard (forge etc.), and also has a good mod tracking system. I see that Minetest can be better
15:32	VanessaE	Minetest has an actual modding API - minecraft's modding is non-sanctioned reverse-engineered stuff.
15:33	VanessaE	and we have a mod tracking system, if you mean a "store"
15:33	VanessaE	(it just isn't well-managed and most modders don't like it)
15:33	technomancy	minecraft's mod tracking system is a mess
15:33	Consortium	VanessaE: okay. Could the mods avoid collisions and "glitch cause by incompatibility"?
15:33	technomancy	I had a really hard time figuring it out; failed and got too bored of it to continue twice before I actually got it to work
15:34	Consortium	technomancy: they have a whole website for it.
15:34	NekoGloop	curse is evil
15:34	Consortium	Curse IS Evil.
15:34	VanessaE	Consortium: such things generally don't happen on minetest to begin with, so such a thing has never been seen as a necessity
15:35	technomancy	Consortium: the fact that they need all piles and piles of documentation just to get something simple running is a bad sign
15:35	VanessaE	minetest mods don't naturally have a conflict concern, unless you consider two mods made for the same purpose and given the same basename. in that case it's obvious that you can only use one or the other of those - but that sort of conflict won't extend to the other mods you are using at the same time.
15:35	technomancy	I found it easier to write my own mod in minetest than use someone else's mod in minecraft
15:35	NekoGloop	minetest's modding API is a lot more.... ah, limited. It's simpler to understand and create simple things, sure, like a few blocks or whatever.
15:36		Telesight joined #minetest
15:36	VanessaE	NekoGloop: you haven't looked at the API recently :)
15:36	NekoGloop	technomancy: if you're referring to installing Forge, the shenanigans with post-1.6 installation have been simplified into an installer
15:36	Consortium	NekoGloop: It needs a bigger base library so the will be more modding opportunities
15:37	technomancy	NekoGloop: luckily I found minetest and now don't really care any more =)
15:37	NekoGloop	what's honestly changed in the modding API :V
15:37	technomancy	but good to know, I guess
15:37	VanessaE	NekoGloop: more than I can actually remember.
15:37	VanessaE	Consortium: what do you mean a bigger "base library"?
15:37	VanessaE	Consortium: look at the modding API: https://github.com/minetest/minetest/blob/master/doc/lua_api.txt
15:38	VanessaE	^^^^^ and that's just the baseline stuff in minetest engine
15:38	VanessaE	any mod can add their own API calls as well
15:38	VanessaE	(look at, for example, the API in plants_lib, https://github.com/VanessaE/plantlife_modpack/blob/master/plants_lib/API.txt )
15:38	Consortium	VanessaE: what ever command prompt Minecraft has, Minetest has to make it better. For example: teleport, currency, factions, PMs
15:38	VanessaE	minetest has PMs, teleport, and there are various currency mods.
15:39	VanessaE	I think there are a couple of factions mods too
15:39	technomancy	great to have you around to tell everyone what to do though; that should help solve a lot of problems. =)
15:40	NekoGloop	the problem with seperating the modding API into lua that hooks into C++ rather than just having the mods -be- the same base programming language is that you can only do what there are hooks for; meanwhile minecraft mods are written in java, so the mods have free range to do anything java can do, not just whatever the modding API allows (Forge is mainly a compatibility layer between mods and a mod lo
15:40	NekoGloop	ader, rather than itself a layer between minecraft's code and the mods' code, though it does help to serve that function.)
15:41	technomancy	NekoGloop: you still have to stick with a published API if you want it to continue to work after an upgrade
15:41	VanessaE	yes, that's true. it also leads to better control over what breaks.
15:41	technomancy	that's the whole point of having a clearly-defined interface
15:41	VanessaE	ninja'd.
15:41	technomancy	knowing what's going to stick around vs what's an implementation detail... pretty important in the long run, even if it feels like a pain up-front
15:41	technomancy	VanessaE: it's the dvorak
15:41	Consortium	technomancy: imagine having a billboard full of ideas, and everyone can choose which to make it happen. The two problems are A. nobody noticing and B. two guys fighting to take "street cred" by making the same idea happen.
15:42	technomancy	=)
15:42		hmmmm joined #minetest
15:42	VanessaE	technomancy: and my bad hands finding two keys for every one I push ;)
15:42	NekoGloop	Consortium: literally only two of those commands you stated were in MC are from MC itself; the other two are mods.
15:42	technomancy	Consortium: talk is cheap; let's see some code =)
15:43	VanessaE	Consortium: we have an issue tracker for that.
15:43	VanessaE	https://github.com/minetest/minetest/pulls and https://github.com/minetest/minetest/issues
15:43	blaise	Consortium: what do you think of this? http://www.subgenius.com/
15:44	Consortium	technomancy: ideas and code to make the ideas happen are separate issue
15:44		Zymotic joined #minetest
15:44	VanessaE	Consortium: the reason minetest isn't as polished or as popular is one simple thing: Minecraft has first-mover advantage,.
15:44	Consortium	blaise: WTF is that?
15:44	NekoGloop	technomancy: the issue with mods breaking after version upgrades in MC is mainly because mojang is dumb and re-obfuscates the java code of MC every version update.
15:45	blaise	Consortium: I'm not entirely sure..
15:45	VanessaE	(no, that doesn't mean it had to come out first -- Infiniminer predates MC -- merely reach the market first)
15:45	Consortium	VanessaE: i know. I am saying "Does Minetest have a Catch-up mechanism to beat Minecraft earlier"?
15:46	blaise	yes,
15:46	blaise	with massive rubber chickens..
15:46	VanessaE	I suppose it does, in the form of good coders.
15:46	VanessaE	both for the engine and for mods.
15:46	* blaise	gasps at the discovery of his empty coffee cup
15:47	VanessaE	but the question of where "catch up" will lead to is as yet unanswered.
15:47	VanessaE	as Minetest is not meant to be a MC clone, by definition it doesn't necessarily need to have the same feature set
15:48	blaise	I own a copy of MC, I chose to play MT over it becuase it runs better and has better graphics, in large thanks to VanessaE and several others in here
15:48	VanessaE	:)
15:49	Consortium	VanessaE: is it possible to gather all FPS lovers in one region, MMO/MOBA in the next room, SIms in the third, and generate as much ideas as possible? Minecraft is already a clone to TF2, Pokemon, Mirror's Edge, Assasin's Creed... Minecraft and Minetest are both sandbox engine. The only difference is that Minecraft is slightly more survival oriented
15:49	NekoGloop	I choose to play MC because the variety of mods gives me a lot to do in the world.
15:49	VanessaE	I got into minetest because I wanted to play around a bit with something equivalent to minecraft's redstone. now three years later, I code mods for it instead because it's more productive.
15:49	blaise	a lot of people consider MT to be vastly ahead of MC in several ways
15:50	NekoGloop	MT is innately ahead of MC in the region of game performance because it was coded competently :V and MT's (numerous) developers actually try to make the game run better.
15:50	blaise	;)
15:51	* blaise	scampers off to refill his coffee mug
15:51	VanessaE	blaise: grab me a cappuccino? :)
15:51	sd1001	Coffee mugs are always too empty
15:51	sd1001	I want a neverending cofe mug
15:51	NekoGloop	however, the style of gameplay that MC lends itself to (survival, with a little bit of engineering and some magic thrown in) is just something I simply find more enjoyable
15:52		RealBadAngel joined #minetest
15:52	Consortium	VanessaE: do you skype or gmail? I really want to live the fast track of coding.
15:52	RealBadAngel	hi
15:52	NekoGloop	hiya RBA
15:52	RealBadAngel	https://imgrush.com/IP7Naqzcb93f.png
15:53	Consortium	RealBadBoy? Oh eah
15:53	NekoGloop	I see you've been busy
15:53	Consortium	*yeah
15:53	RealBadAngel	i think im done with new parallax and auto generation of heightmaps
15:53	kilbith	merge merge merge
15:53	RealBadAngel	as you can see it works with any texture pack
15:55	RealBadAngel	https://imgrush.com/h_F-0fk8FC8Z.png
15:55	Amaz	That is awesome!
15:56	Consortium	That Minecraft skin is awesome
15:57	Calinou	Consortium, TF2 is a clone of Quake ( Í¡Â° ÍœÊ– Í¡Â°)
15:57	NekoGloop	one of the main issues I have with minetest's mods is that, to my knowledge, there's not really a lot of variety.
15:57	Calinou	https://www.youtube.com/watch?v=W1ZtBCpo0eU
15:57	VanessaE	hey RBA
15:57	NekoGloop	Consortium: I'm really beginning to believe you have no idea what you're talking about ^^"
15:57	Consortium	Calinou: Its all a clone-ception
15:57	VanessaE	Consortium: gmail yes, but not skype.
15:58	VanessaE	RealBadAngel: looks good. what algorithm did you end up going with for the autogen?
15:58	RealBadAngel	relief mapping
15:58		blais3 joined #minetest
16:00	VanessaE	ah
16:00		blaise joined #minetest
16:07		est31 joined #minetest
16:07	RealBadAngel	relief mapping is quite expensive but effects are great
16:07	est31	does minetest suffer from this bug too? https://bugs.launchpad.net/ubuntu/+bug/1
16:07		rjs232323 joined #minetest
16:07	RealBadAngel	way better than parallax mapping
16:08		GeHa joined #minetest
16:09	rjs232323	hi, parallax mapping? Thats something I wouldn't know!
16:11	Consortium	VanessaE
16:11	VanessaE	hm?
16:12		jojoa1997\|Linux joined #minetest
16:15	RealBadAngel	rjs232323, we already have parallax mapping in the engine. now we just going to have better one together with autogeneration of heightmaps
16:16	rjs232323	like better in term of looks, performance, or ability?
16:16	RealBadAngel	in fact its going to be something in between parallax mapping and relief mapping
16:16	RealBadAngel	all of them
16:17	RealBadAngel	for fast processing we can use parallax with slope information and just one iteration
16:17	RealBadAngel	for better quality we can choose relief mapping (15 iterations per pixel)
16:17	s-l-teichmann	Calinou: Try this one http://pastebin.com/XKCB4q0j for me the sqrt early stop helps alot, but maybe my code is wrong
16:18	RealBadAngel	if we dont have any normalmaps we can turn on autogeneration, thats most slow one - about 30 iterations per pixel
16:19		Freejack joined #minetest
16:24	Calinou	s-l-teichmann, not faster :/
16:24	Calinou	I'll stick to single thread benchmark for now
16:24	Consortium	VanessaE: gonna go
16:29	VanessaE	er...ok
16:32	Krock	srsly? http://i.imgur.com/X0Ggq21.png
16:35	s-l-teichmann	Calinou: That's interesting ... 494987 w/o sqrt 28794949 w/ sqrt on my i5 ... maybe the i/o buffering on your machine is more dominant. Have you tried to run it bit longer than the 10 seconds. as the primes get sparser the i/o effects should become less important.
16:35		Kenney joined #minetest
16:38		blais3 joined #minetest
16:38	Calinou	http://dl.hugo.pro/benchmarks/
16:38	Calinou	new comparisons up
16:39	Calinou	(with the sqrt)
16:39	Calinou	LuaJIT is now the winner :)
16:43	Kenney	woohoo!
16:47	s-l-teichmann	Calinou: I get 4531091 with your Go version and 28794949 with my mult-thread one which is over 6 times faster. I really dont understand why the multi-threaded version performance so badly at your side.
16:49	exio4	I am surprised
16:49	exio4	Calinou: I get 4+ million with the haskell version, which GHC version are you using?
16:51	Calinou	The Glorious Glasgow Haskell Compilation System, version 7.6.3
16:51	exio4	also, you are counting the warmup of the JVM, pretty sure it could be faster :)
16:51	Calinou	(Debian 8 64-bit)
16:51	exio4	Calinou: ah, 7.6
16:51	exio4	Calinou: may you try with -fllvm?
16:51	Calinou	the JVM warmup is only really noticeable with JRuby
16:52	Calinou	main.hs:11:7: parse error on input `loop'
16:52	Calinou	if I use -fllvm
16:52	Calinou	nvm, was my fault
16:52	Calinou	does Haskell require tabs?
16:53	Calinou	does Haskell require tabs?
16:53	Calinou	oops
16:53	exio4	Calinou: no, not tabs, it uses the layout rule though
16:53	exio4	Calinou: you can use spaces, which is the idiomatic way in Haskell
16:54	exio4	tab size is hardcoded to 8 in the compiler (if you use tabs, make sure to use it with that size)
16:54	Kenney	Waddaya guys think of my UI/HUD suggestions? https://forum.minetest.net/viewtopic.php?f=3&t=12494
16:55	Krock	why do we need a left AND a right texture for the edges?
16:55	s-l-teichmann	Calinou: maybe you put all the version numbers of the compiler/interpreters/runtimes and compile instrutions into the overview page as well?
16:56	Calinou	please no pixel font by default, Kenney
16:56	Calinou	it sucks
16:56	Calinou	this is not 1996 anymore
16:56	Kenney	No idea Krock, that's just the way it's been done for years hah!
16:56	Calinou	else we might also default to 8 bit colors
16:56	Calinou	and .pcx textures
16:56	Kenney	Pixel fonts would fit in with the style of the game, you know, pixelart and voxels.
16:56	Kenney	where a TTF totally clashes with all the other visuals on the screen
16:56	Krock	god, use real fonts if there's already a good font support
16:56	est31	game creators should be abled to chose
16:57	est31	when you connect to a server, the buttons should be the server's theme
16:57	Kenney	That's true though, there should always be the option to switch. But if you'd want Minetest to look like a fairly decent game, work on keeping the style fluent
16:57	est31	when you are in singleplayer or inside the client, the buttons should be the current selected game's ones
16:57	Calinou	bitmap fonts are plain obsolete, too
16:57	Kenney	No bitmap fonts
16:57	Calinou	with rare exceptions (very stylized numbers)
16:57	Kenney	TTF, but bitmap
16:57	Calinou	that makes no sense :P
16:57	est31	minetest is a retro game Kenney is right with that
16:57	Calinou	you mean a pixelated TTF
16:58	Kenney	http://www.kenney.nl/assets/kenney-fonts
16:58	Kenney	Yeah
16:58	Kenney	similar to these fonts I've made earlier
16:58	Calinou	also the big issue of pixelated conf is internationalization
16:58	Krock	ttf's are not supposed to be edged..
16:58	Calinou	Russian/Greek or worse, CJK support, is nonexistent sometimes
16:58	Calinou	pixelated font*
16:58	Kenney	Not sure what you mean by that Krock
16:59	Kenney	but yeah Calinou, that's very true - there could/should be a fallback
16:59	Krock	Kenney, well, you could just write a 2-color bitmap with minimal font size and then upscale. TTFs should be used for round/more complicated things
16:59	Krock	^ IMO
17:00	Kenney	TTF's are used because of the ease, not to limit them to only rounded fonts
17:00	Calinou	we can use GNU Unifontâ€¦ that's what Minecraft did
17:00	Kenney	Bitmap fonts should be used to add certain effects or symbols to fonts, like an outline
17:00	Calinou	it's a tad thin though
17:03	Kenney	I'll probably be forking Voxus for the sole reason of changing the menu, launcher and HUD.
17:03	Kenney	As a designer, Minetest's default is absolutely dreadful
17:04	Krock	VanessaE, oops. reported a wrong post.
17:04	VanessaE	oh well
17:04	blaise	was "item_tweaks" ever merged with default?
17:04	Calinou	Kenney, we can use this week's trendy font: Roboto
17:04	Calinou	that's what I already do :)
17:05	Krock	blaise, only partwise. the items get deleted after a while
17:05	Kenney	Yes, do. Add text reflections, make sure the HUD displays health by adding blood stains and hey why not just get rid of the hotbar and create a Windows style dialog
17:05	Kenney	btw I do use Roboto in my design work, it's an awesome font
17:06	blaise	interesting..
17:06	Calinou	Kenney, me too :P
17:06	Calinou	Roboto Mono's my coding/terminal font, Roboto is my system and website fontâ€¦
17:06	Krock	Kenney, using textures for buttons and other GUI elements is generally a good idea but just not that pixel art.
17:06	Kenney	I'm more of a Robot Slab type of guy though
17:06	Calinou	no italic variant :(
17:07	Kenney	Oh well Krock that was just a sample; it could be anything.
17:07	blaise	Krock: you know if it conflicts with default now? or did they remove the halflife of node code from tweak now that it's implemented in default?
17:07	Calinou	Krock, textures for buttons don't scale well
17:07	Calinou	not good for high DPI screens and downscaled screens
17:07	Kenney	That's why you use 9-slice, so it scales well
17:07	Krock	blaise, ask god, maybe he knows more than me.
17:07	Krock	Calinou, depends on the scale method
17:08	Krock	well, everything has its limits
17:08	blaise	Krock: what is this 'god' you speak of?
17:08	VanessaE	http://www.pickafont.com/fonts/Mister%20Pixel%2016%20Pt%20%20%20Regular.html
17:08	VanessaE	this font looks interesting.
17:09	s-l-teichmann	Calinou: Which Go version do you use? gccgo or gc? 1.4.x or below? This info may help me understand the bad performance of the multi-threaded version. Another question is, does the performande change if you redirect the output to a file?
17:09	Kenney	http://www.dafont.com/pixelmix.font
17:09	Kenney	this is the one I've used in my sample, but it's for personal use only
17:09	Krock	blaise, the dude in the infinite space. However, the best way to find an answer to test it ;)
17:09	Kenney	I'd probably create a custom font for Minetest
17:09	Krock	*is to
17:09	blaise	nice
17:09	blaise	I plan on it
17:10	blaise	I'm sad to see that mt-firearms mod was abandoned..
17:10	blaise	:G
17:10	Krock	I personally Tahoma, it's similar to Minetest's font but I'm more familiar with that one
17:10	Krock	+use
17:11	Krock	Kenney, the bold variant is very similar to Nokia's font http://www.dafont.com/nokia-cellphone.font
17:12	Kenney	Ah yeah it is
17:13	VanessaE	http://www.kreativekorp.com/software/fonts/apple2.shtml -- the BSW font found here (about half-way down the page) is one I "grew up with" ---> http://www.kreativekorp.com/swdownload/fonts/retro/berkeliumii.zip
17:13	blaise	glad to see riesenpilz is still alive and well
17:14	Kenney	drawtype='face_player' would be a godsend when creating good looking trees
17:14		LazyJ joined #minetest
17:15		jordan4ibanez joined #minetest
17:15	Kenney	http://puu.sh/ieKQZ/10b5611795.jpg
17:15	Kenney	pretty close to how I want it now
17:16	jordan4ibanez	That looks beautiful Kenney
17:16	rjs232323	pwetty
17:16	RealBadAngel	https://imgrush.com/qplk9nMDX-m4.png
17:17	Kenney	nice RBA!
17:17	sfan5	Kenney: !
17:17	RealBadAngel	it looks like parallax works quite nice with even 16px and filtering on
17:17	jordan4ibanez	Oh wow
17:17	sfan5	Kenney: on your patreon you say something about LUA, there is not LUA acronym though; it's called Lua
17:18	blaise	man, sumpf is running an error
17:18	blaise	:(
17:18	Kenney	Oh woops, will fix
17:18	blaise	I knew I should have left everything alone..
17:18	blaise	D:
17:18	blaise	12:16:25: ERROR[main]: ...ib/minetest/.minetest/mods/sumpf/jungletree/init.lua:70: attempt to call field 'copy' (a nil value)
17:18	* blaise	whimpers
17:19	Calinou	s-l-teichmann, go 1.3.3
17:21		eeew joined #minetest
17:21	s-l-teichmann	Calinou: Pretty old .. Debian stable I guess. you should try 1.4.2 ... but 1.3.3. is not old enough to explain the bad performance.
17:22	blaise	lol
17:22	Calinou	Debian 8 indeed
17:22	* blaise	uses funtoo
17:22	Calinou	stretch has 1.4.2
17:23	hmmmm	I tried to install funtoo but it wasn't fun
17:24	blaise	hmmmm: what about gentoo ?
17:24	hmmmm	that's really not fun
17:24	sfan5	* blaise uses funtoo
17:24	blaise	hehehe
17:24	Kenney	chances of item spawning when cutting down tree trunks; http://puu.sh/ieLl1/61699b48ac.png
17:24	* sfan5	wonders whether blaise does -funroll-loops
17:24	blaise	sfan5: I do not.
17:24	hmmmm	fun fruit rollups
17:24	Kenney	seeds and fiber will be received when breaking leaves
17:25	s-l-teichmann	Calinou: I will download 1.3.3 and check it.
17:25	VanessaE	sfan5: you forgot -fracing-stripes and -ftype-r-sticker ;)
17:25	sfan5	Kenney: is that last thing a crossaint?
17:25	Kenney	...well, close, it's a maggot
17:25	sfan5	VanessaE: something something http://fun.irq.dk/funroll-loops.org/
17:25	Kenney	or well grub actually
17:28	blaise	ttps://gcc.gnu.org/onlinedocs/gcc-3.4.4/gcc/Optimize-Options.html
17:28	sfan5	>gcc-3.4.4
17:28	blaise	http://dev.minetest.net/User:Consortium/Feature_and_mod_roadmap
17:29	est31	Calinou, thats ^ a better place for the page
17:35		eeew joined #minetest
17:36	blaise	hrmmm
17:36	blaise	who wrote sumpf anyway
17:36	est31	sumpf?
17:37	blaise	HybridDog,
17:37	blaise	whoever that is
17:37	blaise	https://github.com/HybridDog/sumpf
17:37	Krock	( = w_laenger on IRC and some servers)
17:37	blaise	you think it's possible that it was encorperated into moretrees ?
17:38	* blaise	looks
17:38	blaise	nope
17:39	blaise	I think it's not a problem with sumpf so much, as a problem with jungletree
17:39	blaise	https://github.com/bas080/jungletree
17:40	blaise	:\
17:40	s-l-teichmann	Calinou: I've rebuild it with 1.3.3 and getting 20048089 for mine . 4565669 for yours ... which is over 4 times slower. So mystery still exists.
17:41	s-l-teichmann	Calinou: Don't you mind to do a test build with Go 1.4.2 from https://storage.googleapis.com/golang/go1.4.2.linux-amd64.tar.gz ?
17:43	Calinou	or I can upgrade to stretch :)
17:47		Shackra joined #minetest
17:47	blaise	I guess I gotta disable sumpf till jungletree's gets fixed
17:48	blaise	what bug reporting system does minetest use? jira ?
17:48	blaise	bugzilla ?
17:48	Krock	It uses stringstream
17:48	blaise	and, is there a minetest mods bug reporting site that is seperate ?
17:49	Krock	report bugs in the forum topic of the mod
17:49	est31	every mod has their own bug reporting site
17:49	est31	basically
17:49	est31	either on github or on the topic
17:49	Krock	or notabug.org
17:49	blaise	interesting
17:50	est31	or bitbucket.com
17:50	Krock	*.org
17:50	blaise	lmao
17:51	Krock	btw, any interest in a new forum game about "bad fairy"? "I would like to have an ice cream" -> "You get one but it's molten. I want a car"...
17:51	blaise	XD
17:52	blaise	12:52:00: ERROR[main]: ...t/mods/minetest-mod-mesecons/mesecons_torch/init.lua:79: attempt to perform arithmetic on field 'LIGHT_MAX' (a nil value)
17:52	blaise	POOP!
17:52		WSDguy2014 joined #minetest
17:52	blaise	now that I synced all the mods on my server, they're all b0rked! :((((
17:53	blaise	GAH DAMNAH!!
17:53	* blaise	facepalms
17:53	Calinou	s-l-teichmann, upgrading to stretch (9) right now :P
17:54	blaise	I should have left well enough alone.....
17:54	blaise	but I wanted the laitest mods....
17:55	WSDguy2014	Hello guys, i think are mese lamp are added in the game :\|
17:55	blaise	would anyone here be willing to help me with a new ebuild for minetest on gentoo/funtoo ?
17:55	est31	ebuild?
17:55	blaise	I believe the ebuild can remain the same, but there needs to be patches
17:56	blaise	est31: ebuilds are scripts that portage uses to compile packages for gentoo/funtoo from source
17:56	WSDguy2014	in newest version don't have mese lamp
17:56	blaise	WSDguy2014: then where does that code come from?
17:56	blaise	because that's a recent git pull
17:57		Jordach joined #minetest
17:57	blaise	Jordach: howdy
17:58	WSDguy2014	i want mese lamp in the game
17:58	WSDguy2014	not a mod about mesecons
17:58	blaise	lol
17:59	blaise	>---light_source = default.LIGHT_MAX-5,
17:59	blaise	so that's line 79
17:59	blaise	is that line not compatable with minetest-0.4.10 ?
18:00	WSDguy2014	im playing in newest version 0.4.12 and i cannot see mese lamp in the game
18:01	blaise	it's definitly in the mod
18:01	blaise	they're not going to put mod stuff into the default game like that
18:01	blaise	they would have to add the entire mesecons modpack to the default
18:01	blaise	and there's obviously problems with it..
18:02	blaise	where is Jeija when ya need him
18:02	blaise	:G
18:02	Krock	blaise, default.LIGHT_MAX = default.LIGHT_MAX or LIGHT_MAX
18:04	s-l-teichmann	Calinou: I've lock in remotely to an older i7 at work and repeated the test. Here I get 5289217 for your and 41764007 for mine! On this architecture mine is slower!
18:04	blaise	so.. I need to s/default.LIGHT_MAX/LIGHT_MAX/ ?
18:04	Krock	blaise, define default.LIGHT_MAX
18:05	blaise	how would I do that?
18:05	Krock	default.LIGHT_MAX = default.LIGHT_MAX or LIGHT_MAX
18:05	blaise	do I just add that somewhere in the lua ?
18:06	Krock	best place is somewhere in the upper part of the init.lua file
18:06	blaise	I havn't been able to make sense of lua syntax yet
18:06	Krock	(of either mesecons or default itself)
18:06	blaise	so I put "default.LIGHT_MAX = default.LIGHT_MAX or LIGHT_MAX" on a line all by it'self at the top ?
18:07	blaise	Krock: what about in settings.lua ?
18:07	Krock	games\the active game of your world\mods\default\init.lua a line below default = {}
18:09	blaise	hrmm
18:09	blaise	I don't have my structure set up that way
18:09	Krock	then replace \\ with /
18:09	Krock	how else would you use a subgame?
18:10	blaise	the server is ran under the user minetest.. and everything is in ~/.minetest/ aside from the default game
18:10	blaise	in /usr/share/games/minetest/games/minetest_game/
18:11	Krock	that's what I've meant
18:12	blaise	in here /usr/share/games/minetest/games/minetest_game/mods/default
18:12	blaise	with init.lua
18:12	blaise	k
18:12	WSDguy2014	Hello guys, i found about mese lamp in the game, not mods. But theres no Mese lamp in latest version in the game but mese lamp are exist from version 0.4.12-dev ?
18:12	WSDguy2014	I found about mese lamp in the game, not mods. But theres no Mese lamp in latest version in the game but mese lamp are exist from version 0.4.12-dev ?
18:12	blaise	I'll pop it under the GUI related stuff?
18:12	WSDguy2014	sorry about spam again
18:13	Krock	there's a mese lamp in the mesecons mod
18:13	WSDguy2014	i will take screen shot soon...
18:13	blaise	Krock: he wants mesecons_lamp in his default mod
18:13	blaise	lmap
18:13	blaise	s/p/o/
18:14	blaise	WSDguy2014: Krock just told me how to modify the default game
18:14	Krock	is lmao btween lol and rofl?
18:14	blaise	laughing madly and openly
18:14	blaise	or laughing my arse off
18:15	s-l-teichmann	Calinou: When I use "sleep 60" instead of 10 I get 164777777 for mine and 19176193 for yours which is more than 8 times faster. The old i7 has 8 cores.
18:15	blaise	but not rolling on the floor, not yet..
18:15	blaise	so yeah, I suppose it's somewhere in the middle.. XD
18:16	Krock	(cuz laughing off (should be funny (really)) )
18:16	s-l-teichmann	Calinou: So the i/o dominates for lower primes. Mind to run a 60 sec test on your machine?
18:17	blaise	"IT FELL THE FSCK OFF!!!"
18:17	blaise	k, so now I have moar issues with mesecons mod
18:17	blaise	hehe
18:17	WSDguy2014	guys, i can't upload image from imgur.com, i registreded to WSDguy2014 and uploading image takes forever :(
18:17	blaise	13:16:05: ERROR[main]: ...s/minetest-mod-mesecons/mesecons_movestones/init.lua:117: attempt to call field 'copy' (a nil value)
18:17	blaise	I feel like this might could be a lost cause..
18:18	blaise	is mesecons abandoned ?
18:18	Krock	your minetest version is.
18:18	jordan4ibanez	New jukebox
18:18	jordan4ibanez	https://youtu.be/6geusXzeTgE
18:19	Calinou	s-l-teichmann, I'm uploading new benchmark results after finding an optimization
18:19	Krock	blaise, get the latest builtin directory from HEAD to fix that problem, the other soualtion is to get minetest 0.4.12
18:19	Calinou	you can find the results here: http://dl.hugo.pro/benchmarks/
18:19	Calinou	Go is the winner nowâ€¦
18:19	Krock	jordan4ibanez, too large notes for my taste
18:19	Calinou	also C++ was added
18:19	Calinou	Krock, ^
18:19	Krock	\o/
18:20	Calinou	it's not the fastest, surprisingly
18:20	blaise	Krock: I have it compiled in my user dir on my workstation.. the problem is there isn't an ebuild or patches for my linux distro for minetest-0.4.12
18:20	jordan4ibanez	Time to make them BIGGER
18:20	WSDguy2014	im playing in version 0.4.12, must suposed to be see mese lamp block in game version called 0.4.12-dev dev - is Development version
18:20	s-l-teichmann	Calinou: A good multi-threading strategy would be to increase the number of workers only if targeting for higher primes.
18:20	jordan4ibanez	Lol jk, I think you may be right
18:20	blaise	Krock: and while I have used diff to create patches, I'm not fermiliar with it enough to do so on a project which I didn't author..
18:20	Krock	blaise, lovely. I'm using an 14 year old OS and never had any bigger problems with updating :P
18:21	blaise	Krock: which one?
18:21	Krock	Windoze 5.1 aka XP
18:22	blaise	zoiks!
18:22	blaise	you're in Kahootz with the diabla!
18:22	blaise	Krock: https://packages.gentoo.org/package/games-action/minetest
18:23	Krock	I'm in... eh, what?
18:23	s-l-teichmann	Calinou: Can you confirm my results for the 60 seconds test?
18:24	Krock	Calinou, main.cpp line 21: why do you define nb_test = 2 a 2nd time?
18:25	blaise	Krock: you're working with the devil
18:25	blaise	lol
18:26	Calinou	s-l-teichmann, the multithreaded one?
18:26	Calinou	I'm not running Go 1.4.2 yet
18:28	s-l-teichmann	Calinou: 1.3.3 should be fine (1.4.2 performs better). The current thesis is that longer runs favor the multi-threaded version (even with 1.3.3)
18:28	Calinou	can you give me the latest multithreaded version code?
18:29	rjs232323	hello guys, I have a question. I am playing with nodebox and I am trying to determine what would be the best way for nodebox to be as dynamic (the shapes of the nodebox keeps changing base on its nearby nodes)
18:29		Haudegen joined #minetest
18:30		Taurith joined #minetest
18:30	rjs232323	Would I need to register each nodebox for shape it makes or is there a way to update single block's node size? I'm using nodebox with defined fixed numbers
18:30	est31	rjs232323, the technic mod has the cleanest solution: defining one node for every possible combination
18:30	est31	with nice for loops :)
18:30	rjs232323	ah?
18:30	rjs232323	looking into that
18:31	s-l-teichmann	Calinou: http://pastebin.com/XKCB4q0j
18:32	est31	this file especially: https://github.com/minetest-technic/technic/blob/master/technic/machines/register/cables.lua
18:33	WSDguy2014	how do find version 0.4.12-dev, help me
18:33	rjs232323	interesting
18:33	rjs232323	so it loops all that and regist er it
18:33	rjs232323	I dont suppose we could elimate that step and just edit single block's directly eh?
18:34	rjs232323	its just for the looks
18:34	rjs232323	I could use that, thanks est31
18:35	blaise	https://bugs.gentoo.org/show_bug.cgi?id=551386
18:35	blaise	poof, done
18:35	blaise	XD
18:38	blaise	hrmm
18:38		foo_ joined #minetest
18:38	blaise	when exactly was minetest-0.4.12 released?
18:38	blaise	where can I find a timestamp? or a changelog?
18:39	* blaise	looks at github
18:39		Halamix2 joined #minetest
18:39	blaise	ah, yes
18:39	blaise	sfan5 released this on Feb 18 Â· 424 commits to master since this release
18:41	WSDguy2014	how about version 0.4.12-dev, have mese lamp, help me
18:42	Krock	Calinou, is this an acceptable c++ code? http://pastebin.com/3S6CdkQK
18:43	Calinou	I already have the C++ port
18:43	Calinou	it runs just fine
18:43	Krock	Would that one run better?
18:44	Calinou	I don't know :/
18:44	Calinou	try the original and new one on your PC
18:44	* Calinou	is tired as hell with all that benchmarking
18:44		LedInfrared joined #minetest
18:44	Krock	dunno how to timeout in batch
18:45	Calinou	just compare for approximately 10 seconds, Ctrl + C when appropriate :P
18:45	Calinou	poor man's timeout
18:45	Krock	my hero.
18:47	est31	Calinou, heres the c one https://gist.github.com/anonymous/ebc64d49abc5f1ee4e81
18:50	Krock	Okay, my code seems to be a little faster, I would need a 1 minute long check to make sure it's actuall faster :/
18:50	Krock	est31, typedef enum { false, true } bool; ?
18:50	Halamix2	or use VM with linux\]
18:51	est31	oh even nicer Krock
18:51	Krock	found it somewhere in stackoverflow
18:51	est31	its 1:1 what the c code does
18:52	Krock	now let's do it with brainfk
18:53	Calinou	I won't accept esoteric languages, probably
18:53	est31	sorry meant to write: "its 1:1 what the js code does"
18:54	Halamix2	(I kow someone who may be able to write it in QBasic :) )
18:54	Halamix2	*know
18:55	Halamix2	(and I may be able to port it to Pascal)
18:59		edakiri joined #minetest
19:00	edakiri	I tried minetest from debian package and it SegFaulted. I'm using the mesa software rasterizer. Maybe that is related.
19:00		Calinou joined #minetest
19:01		Calinou joined #minetest
19:02		rubenwardy joined #minetest
19:05	MinetestBot	[git] HybridDog -> minetest/minetest: Decrease minetest.after globalstep lag 8383a61 http://git.io/vIZI6 (2015-06-06T21:00:38+02:00)
19:05	Calinou	go 1.4 is much slower on my single-threaded Go program :/
19:05	Calinou	6978k only
19:08		Vadtec joined #minetest
19:11	s-l-teichmann	Calinou: Removing the floating point operation from the inner loop may help. And you should replace println() with fmt.Println & Co.
19:14	edakiri	I'm surprised people voluntarily program in Go.
19:15	blaise	so, I think I may have a working minetest-0.4.12.ebuild
19:15	blaise	:)
19:15	blaise	I'm doing my dependancy check and ebuild manifest now..
19:16	blaise	also syncing portage ..
19:16	blaise	I did submit a bug report for minetest ebuild to be bumped in portage...
19:18		Vadtec joined #minetest
19:22	exio4	edakiri: I am not
19:22	exio4	edakiri: people also write Java, voluntarily
19:23	Calinou	C++ isn't all that great, you know
19:23	edakiri	exio4: I do. It's the linking of Java that allows good support in IDEs.
19:23	exio4	edakiri: well, I meant Java because they think the language is great
19:23		Jordach joined #minetest
19:24	exio4	not because the tooling is great, which is the only great thing Java got
19:24	Calinou	Java has a garbage collector, and is only slightly slower than C++
19:24	exio4	Calinou: who said anything about C++?
19:24	edakiri	I don't think the language is great, the linkage is better than many others and the tool support is good.
19:24	Calinou	it's definitely a step up IMO
19:24	exio4	Calinou: how do you know it's "slightly slower"?
19:25	Calinou	https://en.wikipedia.org/wiki/Jake2
19:25	Calinou	http://bytonic.de/html/jake2.html
19:25	Calinou	http://bytonic.de/html/benchmarks.html
19:25	est31	have you benchmarked c yet Calinou?
19:25	edakiri	Also the Java collections library is better than what I otherwise know. Better than C++ STL.
19:25	Calinou	nope, but C looks mostly obsolete, except for embedded work
19:26	Calinou	I could do it for the sake of it as it's still used
19:26	exio4	edakiri: well, I dislike how Java approaches pretty much everything when it comes to libraries, but that's just liking or not
19:27	Megaf_	Hi
19:27	exio4	hi Megaf_
19:27	exio4	Calinou's benchmark is flawed anyway
19:27	exio4	we need garbage-collector straining benchmarks
19:27	Megaf_	I wonder why my IRC client keeps renaming me to Megaf_
19:27	Calinou	exio4, I am going to wrap my answer in a class constructor, let me take 3 minutes. :-)
19:29	exio4	gc perf, integer perf, floating point perf
19:29	exio4	idiomatic vs non-idiomatic code
19:30	exio4	there are lots of non-idiomatic shortcuts that can be used for optimizing performance in common cases
19:30	Calinou	but you don't always have dozens of hours to do that
19:31	exio4	you should also benchmark libraries :)
19:32	Calinou	feel free to write that in 3+ languages :P
19:32	Calinou	it quickly gets impossible to maintain
19:34	exio4	why do you say that?
19:34	exio4	also, benchmarks shouldn't get tweaked that much
19:34	exio4	anyway enchmarks are also useless
19:34	exio4	because they don't show what the language is capable of doing
19:35	Calinou	let's rewrite Linux to GDScript then
19:35	exio4	the best parts of a language can't be benchmarked automagically with a machine
19:36		Viper168 joined #minetest
19:36		Jordach_ joined #minetest
19:37	exio4	how do you benchmark "how easy to mantain" a program that way?
19:40		neoascetic joined #minetest
19:41	neoascetic	Hi all. Trying to migrate my world to redis and get this error: "Migration to redis is not supported". I have compiled server with hiredis support. What to do?
19:43	est31	is that the precise error message?
19:44	neoascetic	Sorry, I was wrong. I thought that ENABLE_REDIS is set to true by default. Going to recompile it then
19:44		blaise_ joined #minetest
19:45		blaise joined #minetest
19:45	blaise	hrmm
19:45	blaise	lmao
19:47		FR^2 joined #minetest
19:58	neoascetic	Hmm. Now I have "Set redis_address and redis_hash in world.mt to use the redis backend" message (however, I have them in the file).
19:58	neoascetic	Maybe incorrect format? Now I have redis_address = 172.17.0.5 and redis_hash = world
20:01		Trustable joined #minetest
20:04		LittleJoe joined #minetest
20:04		Calinou joined #minetest
20:07	blaise	so far so good with minetest-0.4.12.ebuild and minetest_game-0.4.12.ebuild
20:07	blaise	:D
20:10	Halamix2	Calinou: http://codepaste.net/puaxqe
20:11	Calinou	is that Pascal?
20:11	Halamix2	yup
20:11	Calinou	quite obsolete, but I'll add it when I have time :P
20:11	Halamix2	it SHOULD work (I don't write in pascal but I've been learning it in school)
20:11	Halamix2	:)
20:21		blais3 joined #minetest
20:26		YvesLevier joined #minetest
20:30		Ritchie joined #minetest
20:32		s-l-teichmann joined #minetest
20:34	blaise	the ebuild works
20:34	blaise	perfectly
20:34	blaise	yay
20:34	blaise	I did it!
20:34	blaise	took me almost a week!
20:34	blaise	HEH
20:45	blaise	sure is quiet in here
20:50		Amaz joined #minetest
20:52	edakiri	blais3: tried exherbo?
20:52	edakiri	blaise: ?
20:53	edakiri	or I should write #exherbo
20:57		LedInfrared joined #minetest
21:03	blaise	edakiri: what's that?
21:06	edakiri	blaise: it is a distribution using Paludis, which should have been the new package manager for Gentoo if the old Gentoo initiator hadn't been dismissed from MS and returned to the project.
21:07	blaise	edakiri: are you talking about Daniel?
21:07	edakiri	I forget his name.
21:07	blaise	edakiri: he forked and called it funtoo
21:07	blaise	he also forked portage.. ditched rsync and uses git
21:08	blaise	edakiri: he's in #funtoo with me and a bunch of other maintainers...
21:08	blaise	I've been around since it was called Enoch
21:08	exio4	funtoo looks pretty cool
21:08	exio4	I haven't used funtoo though
21:08	exio4	only gentoo
21:09	blaise	Drobbins didn't file the paperwork in time to prevent losing control over the gentoo name.. it had nothing to do with being dismissed from microsoft, nore did he work there.
21:09	blaise	so he created funtoo after some of the other gentoo dev's jumped on his lazyness like a pack of hungry rabid wolves.
21:10	edakiri	There was a core Gentoo developer who.... Maybe I'm confused and thinking of one from Red Hat.
21:11	blaise	paludis wasn't created by the founder of gentoo.. it was a different gentoo developer that decided to go in a different direction than the other developers as far as portage goes..
21:11	edakiri	No, I think it was Gentoo.
21:11	edakiri	I didn't mean that Paludis was created by a founder of Gentoo.
21:11	blaise	I believe paludis was an attempt to code a portage alternative in C
21:11	edakiri	C++
21:11	blaise	yeah
21:11	blaise	I played with it when it first came out.. I wasn't really impressed with it at the time, although I was very interested in a C++ written alternative..
21:12	blaise	I'm sure the guy has made some serious advancements though..
21:13	blaise	I gotta run to the store, I'll brb
21:13	blaise	:D
21:15	edakiri	Before Exherbo, there was a small poll of people who had tried both portage and paludis on Gentoo. The majority answered that they prefered paludis. That's why I'm estimating that the choice to remain with portage was more political than technical.
21:16	edakiri	Though it seems you tried it also and didn't prefer it at that time. So I'll count that also.
21:18	edakiri	Checked the tidings. Daniel Robbins did work at MS briefly.
21:21	edakiri	I quit using Gentoo because they were exclusive, didn't want participation. Exherbo is the opposite.
21:22	edakiri	At least the SPArc team of Gentoo was that way.
21:25	edakiri	Drat. I didn't realize he disconnected his client.
21:39		hmmmm joined #minetest
21:41		est31 joined #minetest
21:42		Haudegen joined #minetest
21:51		blaise joined #minetest
21:57		s-l-teichmann joined #minetest
21:57	* blaise	is back
22:07	blaise	is this broken? 16:58:35: ERROR[CurlFetchThread]: servers.minetest.net/announce not found (Timeout was reached) (response code 0)
22:12		s-l-teichmann left #minetest
22:14	edakiri	Before Exherbo, there was a small poll of people who had tried both portage and paludis on Gentoo. The majority answered that they prefered paludis. That's why I'm estimating that the choice to remain with portage was more political than technical.
22:15	edakiri	I quit using Gentoo because they (at least the SPArc team) were exclusive, didn't want participation. Exherbo expects participation.
22:27		meldrian joined #minetest
22:27	blaise	!up midnightsystems.net
22:27	MinetestBot	midnightsystems.net:30000 seems to be down
22:28	blaise	poop
22:44		LandMine joined #minetest
22:44	LandMine	Oldcoder
22:44	LandMine	are you there?
22:46	LandMine	VanessaE are you here?
22:52	blaise	anyone know the range on the Quarry?
22:52	blaise	radius limitation that is?
22:54	LandMine	NekoGloop hi long tim e no see
23:09		Thron joined #minetest
23:16		alket joined #minetest
23:30		Player_2 joined #minetest
23:31		Wayward_Tab joined #minetest
23:35		Haudegen joined #minetest
23:40		RealBadAngel joined #minetest
23:50		sythe joined #minetest
23:51	blaise	sup?

← Previous day | Channels | #minetest index | Today | Next day → | Google Search | Plaintext