Minetest logo

IRC log for #minetest-dev, 2016-10-30

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

All times shown according to UTC.

Time Nick Message
00:08 proller joined #minetest-dev
00:18 MrIbby joined #minetest-dev
01:13 Tmanyo joined #minetest-dev
01:22 Taoki joined #minetest-dev
01:26 xunto joined #minetest-dev
01:42 Zeno` joined #minetest-dev
01:52 AcidNinjaFWHR joined #minetest-dev
02:13 Miner_48er joined #minetest-dev
02:30 garywhite joined #minetest-dev
02:46 nnnn20430 joined #minetest-dev
03:17 hiradur joined #minetest-dev
03:30 paramat is #1984 good as it is? it's so simple shall we just merge it?
03:30 ShadowBot https://github.com/minetest/minetest/issues/1984 -- Rename testsounds/ directory to sounds/ by Calinou
04:22 paramat left #minetest-dev
04:27 Zeno` I think you can make that decision
04:27 Zeno` :)
04:28 Zeno` why is it currently called testsounds/ anyway?
04:37 Zeno` merged
04:52 Zeno` ShadowNinja's made a good point, but it's not /really/ related to 1984 because the current behaviour doesn't do that anyway
05:42 Hijiri #4685
05:42 ShadowBot https://github.com/minetest/minetest/issues/4685 -- Add control information to player interacts by raymoo
05:47 Zeno` hmm
05:48 Zeno` does it break anything?
05:52 Zeno` The code looks good, I'm just not familiar with the networking side of things
06:21 AcidNinjaFWHR joined #minetest-dev
07:00 Hunterz joined #minetest-dev
07:03 CWz joined #minetest-dev
07:09 Hijiri Zeno`: well, I tested it and thingsseemed fine just walking around and touching things and such
07:09 Hijiri and it seems to fix the problem it's meant to
07:10 hmmmm good work johnson
07:10 hmmmm let's ship it
07:12 Hijiri seems to work with old client -> new server
07:17 hmmmm but... what happens when you connect a new client to an old server?
07:17 Zeno` that's what I was worried about
07:18 hmmmm https://blog.codinghorror.com/content/images/uploads/2007/03/6a0120a85dcdae970b0128776ff992970c-pi.png
07:22 hmmmm bah i wish there was info about doing protocol/serialization version bumps on the dev wiki
07:22 hmmmm was going to get that link and paste it here had it existed
07:31 Hijiri Zeno`: I'll try it
07:31 Hijiri Zeno`: looks like it's working
07:54 Krock joined #minetest-dev
07:59 Zeno` Hijiri, I'll test it as well. I'm cooking dinner atm though
08:00 Zeno` Hijiri, are you the author?
08:03 blaze joined #minetest-dev
08:05 Foghrye4 joined #minetest-dev
08:07 Foghrye4 Hello. May i ask your assistance? in ParticleSpawner::step(float dtime, ClientEnvironment* env) dtime is a milliseconds value or seconds value?
08:11 Foghrye4 Ok, it seconds, nevermind.
08:27 Hijiri Zeno`: yes, I'm raymoo
08:39 red-001 joined #minetest-dev
08:40 Zeno` Hijiri, I'll test now
08:53 CWz joined #minetest-dev
09:03 Hijiri Zeno`: Ok, thanks
09:07 Zeno` seems to be fine
09:07 Zeno` old client, new server -> OK. new client, old server -> OK
09:10 Zeno` I'm inclined to +1 in but I don't know enough about the network code
09:11 Zeno` valgrind related checks with the patch are fine as well
09:12 jin_xi joined #minetest-dev
09:12 jomat joined #minetest-dev
09:59 proller__ joined #minetest-dev
10:01 Thomas-S joined #minetest-dev
10:07 hmmmm joined #minetest-dev
10:12 Fixer joined #minetest-dev
10:36 FirePowi joined #minetest-dev
11:11 AcidNinjaFWHR joined #minetest-dev
12:02 srifqi joined #minetest-dev
12:56 proller joined #minetest-dev
13:12 nrzkt joined #minetest-dev
13:23 proller__ joined #minetest-dev
13:31 nrzkt hi guys, will merge 4612 in ~15min
13:31 Krock #4612
13:31 ShadowBot https://github.com/minetest/minetest/issues/4612 -- Player class cleanup / PlayerSAO - LocalPlayer ownership by nerzhul
13:32 nrzkt thanks Krockbot :D
13:32 Krock np
13:32 turtleman joined #minetest-dev
13:36 Taoki joined #minetest-dev
13:38 Taoki joined #minetest-dev
13:41 srifqi joined #minetest-dev
13:47 Karazhan joined #minetest-dev
13:49 proller joined #minetest-dev
13:50 est31 nrzkt: fine with me
13:50 nrzkt est31, good :)
13:53 nrzkt merged thanks all
13:58 Foghrye4 #4705 So, what are options we got? I could remove Irrlicht mentioning from Lua api doc and in time when someone decide to switch to another engine just rewrite video.pack_texture_blend_func to match ne engine. what you think?
13:58 ShadowBot https://github.com/minetest/minetest/issues/4705 -- Adding particle blend, glow and animation by Foghrye4
14:00 est31 I also don't like that it creates a new namespace for that
14:00 est31 it makes especially no sense as lua is ran on the server
14:00 est31 so yeah, better remove them
14:00 est31 (the irrlicht mentions)
14:01 Player_2 joined #minetest-dev
14:01 est31 and put it inside the minetest namespace
14:01 Foghrye4 est31, so "minetest.EBF_ZERO"?
14:01 est31 better make it minetest.ebf.zero
14:01 est31 lua has tables for this :)
14:01 Foghrye4 I could do that.
14:13 est31 also see my comments
14:13 Foghrye4 I've got it. In a process.
14:14 est31 note that in lua you can't implement the function in the way I pointed out
14:14 est31 thats why I suggested to remove it
14:14 est31 (the version of lua we use has no bitops)
14:31 Foghrye4 est31, about "use only tabs for indentation". I just follow existing style, so may be i just remove all indentation and make it look like a normal assignment unrelated to variable initialization?
14:36 est31 Foghrye4: I commented on that line because you have added one tab and then 4 spaces
14:37 est31 also, its really better written if each variable was declared and defined in the same line
14:37 est31 so u16  min_first_frame = 0
14:37 Foghrye4 Ok.
14:37 est31 and u16 max_first_frame = 0;
14:44 sfan5 so what about #4474 btw
14:44 ShadowBot https://github.com/minetest/minetest/issues/4474 -- Avoid sending a chatmessage that a player joined if in singleplayer mode by red-001
14:45 Foghrye4 Who dfck cares?
14:48 sfan5 small details add up
14:48 sfan5 nrzkt: it seems your PR adds quite a lot of warnings https://travis-ci.org/minetest/minetest/jobs/171771309#L710
14:49 nrzkt sfan5, oh i forgot to test with clang this time
14:49 nrzkt anoying i prefer it :p
14:50 nrzkt we need to add overload keyword in these functions
14:51 nrzkt let me finish the current PR thing i'm doing and i will fix that
14:51 sfan5 sure
14:59 blaze joined #minetest-dev
15:00 est31 or just avoid overload completely
15:00 est31 overloading is bad
15:01 nrzkt the problem here is playerSAO has a function with more parameters than its child
15:01 nrzkt i will try to find a good solution for this
15:05 nrzkt est31, then i propose to add setYawWithSending which will call setYaw (non overloaded) + do the required work for sending
15:05 nrzkt is that okay for you ?
15:05 est31 yes
15:06 est31 if you have to rename it at multiple unrelated places, please make a separate commit though
15:06 nrzkt yes it's not a problem
15:06 nrzkt i will do a PR for this in 5 mimns
15:10 est31 thanks
15:13 nrzkt sfan5, est31 #4706
15:13 ShadowBot https://github.com/minetest/minetest/issues/4706 -- Fix overloading problems mentioned by clang by nerzhul
15:14 sfan5 nrzkt: s/WithSending/AndSend/
15:14 nrzkt i added setHPRaw too for setHP, and LuaEntitySAO now has a little improvement in performance thanks to overload warning, we use const ref for position
15:14 nrzkt sfan5, okay, est31 did you agree too ?
15:14 sfan5 and it's "data" not "datas"
15:14 sfan5 fine otherwise
15:15 nrzkt i let datas ? rmlrml
15:16 nrzkt sfan5, typos fixed
15:17 sfan5 looks good
15:17 nrzkt i'm waiting for travis to do unittests and i will merge it
15:30 nrzkt https://travis-ci.org/minetest/minetest/jobs/171782875
15:30 nrzkt it's good, i merge it
15:37 proller__ joined #minetest-dev
16:09 proller__ joined #minetest-dev
16:36 red-001 is it possible to convert userdata to a table or string?
16:37 sfan5 no
16:37 sfan5 if you've come to the conclusion that it would solve your problem, then you're doing something very wrong
16:40 red-001 so what would be the correct way to save a copy of player inventory?
16:58 Foghrye4 joined #minetest-dev
16:59 Foghrye4 .
16:59 red-001 ?
17:00 Foghrye4 Thats accident. Don't mind me. I'm in my bed now ^_^
17:25 hmmmm what the hell
17:25 hmmmm nrzkt, could you stop making primitive argument types const
17:27 hmmmm what the hell did you even do to SAO anyway?  it's a 'cleanup' but I fail to see how it's any cleaner
17:28 hmmmm what the fuck actually
17:28 hmmmm stop
17:28 hmmmm STOP merging shit
17:28 hmmmm wow did any of this go through any review process at all?
17:29 hmmmm no
17:29 hmmmm we rollback historty to two days ago or revert all the commits
17:29 hmmmm what the fuck
17:30 hmmmm minetest is not your personal playground
17:34 hmmmm "Added my own approval"  - wow, just wow
17:48 sfan5 hm i missed those const things
17:49 sfan5 it was reviewed by est31 and me
17:49 hmmmm i'm not really upset with that one
17:49 hmmmm just look at the recent commits from the past two days
17:50 hmmmm here's your typical nerzhul PR:  https://github.com/minetest/minetest/pull/4574
17:54 hmmmm okay this isn't as bad as i thought
17:54 Foghrye4_ joined #minetest-dev
17:54 hmmmm a lot of these are just unsquashed commits from a single PR that did get appropriate approval
17:54 hmmmm there are still a couple that didn't get proper approval though...
17:54 hmmmm and who the hell signs off on a commit titled "Cleanup"
17:56 hmmmm that's one step away from freeminer where every other commit is titled "fix"
17:57 nrzkt in fact for the cleanup commit it's a merge error, i simply missed and clic on rebase instead of squash. This was mentionned on IRC and i didn't revert the merged commits because i see the mistake later the same day
17:58 nrzkt for other problems you mentionned, just read the fucking IRC log (RTFIL) :)
17:59 hmmmm you CAN'T approve your own commit
17:59 kaeza joined #minetest-dev
17:59 hmmmm maybe you can on paper, but that just bumps the requirement up another core dev reviewer
17:59 nrzkt i can, as est31 tells you if you read the github log and the documentation. We need two coredev approvals. and i'm a coredev, like you
18:00 hmmmm the whole point of the peer review is to have a *second* set of eyes that aren't your own
18:00 nrzkt and that's done on all the PR i posted.
18:00 hmmmm and like i said
18:00 hmmmm people have used that same argument before
18:00 nrzkt trivial PRs too, sfan5 is the second set of eyes there :)
18:00 hmmmm it just increases the number of approvals required
18:00 hmmmm you have a very fucked definition of trivial
18:01 nrzkt and you are just a cow
18:01 hmmmm or maybe it's warped because you want to get away with shit that obviously nobody else would go for
18:02 nrzkt i think you vote for Trump no ?
18:03 hmmmm well, you heard it people, i'm a "cow" and a "trump voter"
18:03 nrzkt => you are a texan ? :p
18:03 hmmmm be sure to eat my cheese and drink my milk as i make america great again
18:03 hmmmm and what this has to do with PR approval, i have no idea
18:03 sfan5 h-how
18:04 sfan5 how did you move from MT development to trump in like three lines?
18:04 nrzkt just keep in mind we are all people, and humans, not gods :p then humans have mind and ideas, and some are related to some others :D
18:06 hmmmm we have high standards in theory to keep code quality high and problems low
18:06 kaeza o/
18:07 hmmmm but this keeps getting eroded by sidestepping the process for some perceived increase in convenience
18:07 nrzkt i don't know if our code quality is good why MT is so far behind MC in terms of performance
18:07 hmmmm because the two have nothing to do with eachother
18:08 nrzkt except these are exactly the same games with a very very close gameplay
18:08 sfan5 https://github.com/minetest/minetest/issues/4707#issue-186150025
18:08 nrzkt one you mod in Lua the other in java (or not, thanks to M$
18:08 sfan5 does anyone understand the content of that issue...?
18:08 hmmmm minetest has more effective occlusion culling
18:08 nrzkt sfan5, i think he wants to install and configure mods from a special menu
18:08 hmmmm that's
18:08 hmmmm it
18:09 nrzkt hmmmm, yes, but the light is not very good xD
18:09 hmmmm they also have the money to buy experts at GPU optimization
18:09 hmmmm how does that have to do with code quality
18:09 sfan5 hmmmm: considering the existence of OptiFine it looks like they don't do that
18:09 nrzkt notch has done a very good game alone before being bought by M$
18:09 hmmmm also by making this assertion
18:10 hmmmm you're implying that MT is a cesspit because it's not up to par in terms of GPU performance with MC, therefore anything goes
18:10 hmmmm just shit on top of the pile
18:10 nrzkt it's not just the GPU, we have a very high CPU usage in some parts of code
18:10 Darcidride joined #minetest-dev
18:10 nrzkt GPU is quite good, except the view range which cannot be very high
18:13 hmmmm sfan5, perhaps I was thinking of minecraft PE
18:13 sfan5 MCPE is very nicely optimized, yeah
18:13 hmmmm there was a post made by one of the developers that discusses how their cave occlusion algorithm works
18:13 hmmmm and it's very nifty
18:14 hmmmm that's the level of optimization minetest simply does not have, because our priorities are everywhere else but there and frankly i don't think the majority of minetest developers have the skill necessary to create such a thing
18:14 hmmmm sorry to say this
18:15 sfan5 it's just the truth
18:15 hmmmm well that's probably not a fair assessment
18:15 hmmmm it's like that at the vast majority of software shops
18:16 hmmmm you'll have like one or two super smart guys that work on the big algorithm-y things and the rest usually work on the more routine code
18:18 garywhite joined #minetest-dev
18:28 Darcidride joined #minetest-dev
18:31 hmmmmm joined #minetest-dev
18:34 Fixer we gonna build the wall, around hmmmmm, and let him develop for us
18:35 Krock ..and throw in some food frequently
18:36 Fixer i don't think MC team has a lot of devs, iirc it had quite a few people actually in Mojang
18:36 Krock but they were developers as full time job
18:36 Fixer ofc
18:36 Krock (or still are)
18:38 Krock the organization work is likely exponential to the amount of developers. No clue how they do it in M$
18:38 hmmmmmm joined #minetest-dev
18:39 garywhite You know that now MC has all of Microsoft Studios' devs at its disposal
18:39 Krock yeah, they can import bugs now easily
18:40 sfan5 importing bugs with python is really easy too
18:40 sfan5 you just:
18:40 sfan5 import bugs
18:40 sfan5 -done
18:40 Fixer I dunno... even with those high standards nobody fixed most annoying gameplay bugs...
18:40 garywhite Coming soon, Windows viruses that can attack Minecraft on any operating system, from Microsoft Studios
18:41 sfan5 garywhite: *except Linux
18:42 Fixer why? you attack win from everywhere anyway
18:42 garywhite My joke is that it attacks
18:42 garywhite MC
18:42 garywhite not the OS itself
18:42 garywhite and then uses MC to hack stuff and cause DDoS
18:44 hmmmmm joined #minetest-dev
18:44 Fixer minetest has worse performance than minecraft on my pc, but bigger problem is rendering... blinking stuff, translucency, stuttering, ghost lighting blahblah, add to this simple but big gameplay bugs, like sneaking up trough solid block or avoiding fall damage, or resetting fall damage at long fall, or taking your items after death from bones,
18:44 Fixer water reflow
18:44 Fixer is not fixed since its birth
18:47 Fixer hey, what do you think about switching rendering engine? i know a guy who has like 90 days of free time, is it possible to do in 90 days?
18:48 Hijiri probably
18:49 Fixer is it worth it? will devs oppose? or accept?
18:49 Foghrye4_ Unless you talking about yourself, its looks hardly plausible that this guy will work on it 90 days straight for free.
18:50 Hijiri I would guess the biggest part of the time would be getting to know the rendering engine though, not changing minetest over
18:51 Hijiri Fixer: main issue I would see, assuming that your guy will actually do the thing, is that any other dev working on graphics stuff will have to learn the new engine
18:52 Hijiri But that might not be such a bad thing, if it's easier to use than irrlicht
18:52 Fixer i don't think there are devs working on graphics, celeron does not, and RBA died
18:52 Calinou Irrlicht is fine, we just aren't using it well
18:52 Hijiri does anybody do anything touching irrlicht?
18:53 Calinou (note that we're kind of using it for an unintended use already)
18:53 Foghrye4_ Irrlicht is open source. What prevent us from fixing engine bugs in engine development?
18:54 Hijiri if they're not accepting patches then we would need to fork it, and make it harder to distribute
18:54 Calinou Foghrye4_: modifying Irrlicht requires us to bundle it in the source
18:54 Calinou and I don't think we need to modify it
18:57 Fixer whats wrong with texans?
18:58 blaze joined #minetest-dev
19:23 est31 we can modify irrlicht source
19:23 est31 but obviously needs to be upstreamed first .)
19:30 sfan5 Calinou: it doesn't, see: https://github.com/minetest/minetest/tree/master/build/android/patches
19:30 sfan5 obviously that is only for small fixes and not whole new features
19:30 est31 yeah
19:30 Fixer irrlicht is broken?
19:30 est31 also, its only done for the android buidl
19:30 est31 build*
19:31 est31 and it would still be great if those changes would be upstreamed
19:31 est31 that remembers me.... someone needs to port us to newer irrlicht
19:31 est31 we still use age old irrlicht
19:32 garywhite What about java like MC?
19:32 est31 from 2015 or something
19:32 Fixer age old for android?
19:33 Calinou garywhite: uh, we're not going to rewrite Minetest in Java…
19:33 est31 ^
19:33 est31 MC is more performant despite java not because of it
19:34 est31 (in the areas it is more performant than mt)
19:36 est31 so the latest commit on the ogl es branch is r5336
19:36 est31 and we use commit no 5122
19:36 est31 from 2015-08-29
19:36 est31 Fixer: ^
19:40 sfan5 est31: there's some problems with that
19:41 sfan5 for example the texturehack patches don't apply to new irrlicht
19:41 sfan5 the files no longer exist
19:41 sfan5 and it seems the code was refactored and not just moved
19:41 est31 nice.
19:41 est31 (note the sarcasm)
19:42 sfan5 well it's our fault for patching irrlicht with those hacks
19:42 est31 yes
19:42 est31 also our fault for not documenting for what those hacks were for
19:43 sfan5 well android has far worse problems
19:43 sfan5 the UI, the Makefile
19:43 est31 The makefile is the common way how you build stuff on android with ndk
19:43 est31 I came across many open source projects using ndk due to fdroid and everywhere makefile is used
19:44 est31 or something else
19:44 sfan5 no i mean Makefile
19:44 sfan5 not Android.mk
19:44 est31 libreoffice uses automake
19:44 sfan5 Android.mk is fine
19:44 sfan5 the Makefile with the dependency downloading, building, installing is horrible
19:44 est31 godot engine uses its scons system
19:44 est31 pep uses some combination of makefile, sh scripts and ugly hacks that I don't remember
19:45 est31 very conveniently, they use bashisms and are prefixed with /bin/sh
19:45 est31 so on debian it breaks because of dash
19:46 est31 Android.mk doesn't help you with downloading deps and compiling them
19:46 est31 unless you
19:46 est31 1. track all your deps in git or in submodules
19:46 est31 2. list every single file in the file
19:47 est31 and 3. manage to somehow emulate all the custom params passed to the compiler that makes stuff work
19:47 est31 but I agree that the UI is a problem
19:48 sfan5 est31: i'm not saying that the Makefile doesn't work
19:48 sfan5 but it looks horrible
19:48 sfan5 and its code is horrible
19:50 garywhite Hey guys do you know what license MT is under?
19:50 sfan5 have you tried reading the README
19:51 est31 sfan5: any ideas with what the makefile can be replaces
19:51 sfan5 a shell script
19:52 sfan5 the makefile uses only very few features of GNU make over a plain shell script
19:52 sfan5 also sometimes incorrectly
19:52 sfan5 with a shell script you can have functions that do stuff
19:52 sfan5 instead of duplicating e.g. the timestamp code 10 times
19:52 garywhite right, forgot the README file
19:52 est31 yes
19:52 est31 but you have no dependency management
19:53 est31 there is lots of logic hidden in that
19:53 sfan5 not really
19:53 est31 plus parallelism
19:53 est31 you can do make -j whatever and it manages everything for you
19:53 sfan5 last time i tried that it broke something
19:54 sfan5 here's some dependency management for you https://github.com/mpv-android/buildscripts/blob/master/buildall.sh
19:54 sfan5 see line 10 and line 45
19:55 est31 mhh
19:56 est31 well, if you make a PR with a shell script and it looks nicer than before and it supports parallelism AND IT WORKS, I'm all for adding it.
19:56 sfan5 nah that's way too much work
19:56 sfan5 maybe in a few weeks :P
19:56 sfan5 also "it looks nicer than before" obviously isn't hard to achieve
20:00 garywhite It appears to be GPL, so question: If I were to put together a set of software that is used for development, would it be legal under GPL to distribute MT client under said SDK?
20:03 est31 1. its not GPL
20:04 est31 2. ask a lawyer, that's the best way to get a definite answer
20:04 est31 (except fighting over it in a high court)
20:04 * Fixer .oO(0.0.0 spawn bug)
20:17 garywhite 1. I'm scrapping it
20:22 rubenwardy joined #minetest-dev
20:41 nrzkt joined #minetest-dev
20:47 hmmmmm nrzkt, https://github.com/minetest/minetest/pull/4706/files#diff-96348a5fec2608cad0492ea0352240f6R53
20:48 hmmmmm i really, really wish you'd give anybody a chance to review your commits before merging them
20:48 hmmmmm or better yet, follow the fucking rules and not merge things without peer review
20:51 nrzkt => http://irc.minetest.net/minetest-dev/2016-10-30#i_4733812
20:52 hmmmmm i didn't get a chance to even look at it
20:52 hmmmmm i would've pointed out the grave error of making those unittests do even less than they do right now
20:53 hmmmmm first of all those unit tests are shit
20:53 hmmmmm they are the bare minimum of what constitutes a unit test i think
20:53 nrzkt hmmmmm, just do it yourself or answer my questions when you talk on PR
20:53 hmmmmm and then you nerf them even more by changing setHP with send_data = true to not even sending
20:53 hmmmmm so now you're testing a short-circuited version of the function that does nothing but set a variable equal to something else
20:54 hmmmmm now we don't even know if the real setHPAndSend instantly crashes
20:54 nrzkt in fact it crashed because you need a server an environnment etc etc
20:54 hmmmmm i'm not around all the time, and you tend to try merging things as quickly as possible
20:54 hmmmmm then maybe you should fix the function huh
20:54 nrzkt and the short version is not for unit tests this function is used elsewhere in the code
20:55 nrzkt there are some code parts which doesn't need to send the datas
20:55 hmmmmm i'm willing to bet $5 you just broke HP sending again for the literally 4th time with this "cleanup"
20:55 nrzkt lol, i don't take $ this is shit money, just give me euros or bitcoins :D
20:55 hmmmmm is there any unit test that verifies that hp sending actually works after you changed it?
20:56 nrzkt in fact nothing changed, but you didn't read my pr
20:56 hmmmmm no, you claim behavior hasn't changed
20:56 hmmmmm but the code sure changed
20:56 nrzkt no
20:56 nrzkt only attributes changes their ownership.
20:56 hmmmmm and the WHOLE FREAKING PURPOSE of unit tests are to verify by automatic testing that your code changes did not modify the expected behavior
20:56 nrzkt all other code parts has only been adapted to the scope of SAO instead of the fucking memleak RemotePlayer
20:57 nrzkt do it youself if' you're angry instead of complaining. I added unit tests covering my part.
20:57 hmmmmm you didn't
20:57 hmmmmm you changed a lot more
20:57 nrzkt less you talk more you can code it.
20:57 nrzkt in fact no, but in fact you didn't know the code i change and all it does and why i have done all of this
20:58 hmmmmm i'm not working cleanup to fix your code changes
20:58 hmmmmm don't you think i've done enough of that already?
20:58 hmmmmm how long did it take to track down and fix all the bugs and security holes from your "better" network code?
20:58 nrzkt no i think you are just a main who like complain about code review
20:58 hmmmmm because this is the problem with minetest
20:58 hmmmmm we have hit n' run developers who sure like to add or change a whole lot of code
20:59 nrzkt which security issue ? the security issue that we store password in memory client side ? the security issue that our Lua sandbox is not a full sandbox ?
20:59 hmmmmm but ask them to support their shit or at least add tests so we can know if it breaks anything, nope
20:59 hmmmmm because that's not fun is it
20:59 nrzkt and you're not fun, and you don't contribute since 6 months.
20:59 hmmmmm no, how about the blatant buffer overflows from user input
20:59 nrzkt lol, just read the client side code then
20:59 hmmmmm so anyway
21:00 hmmmmm this is what we'd expect is par for the course with non-core-dev members who just contribute occasionally via pull request
21:00 hmmmmm but you're a CORE DEV
21:00 nrzkt if you are complaining about bugs, it seems you have the solution, then propose a PR to fix it.
21:00 hmmmmm gee i sure would if you didn't merge the PR within 5 minutes of opening it
21:00 hmmmmm and again
21:00 hmmmmm i have my own set of things to work on
21:00 nrzkt it will be more constructive than loosing my time on this channel to talk with a complaining people which doesn't want to code a fix he knows
21:01 hmmmmm it seems to me like you want to be a cowboy coder and then have a bunch of dudes/ranch hands clean up your mess
21:01 hmmmmm no
21:01 hmmmmm that's not how it works
21:01 hmmmmm if you make the mess you clean your own mess up
21:01 hmmmmm but it's not too easy to do that now that it's already been merged, now, is it
21:01 hmmmmm which is why i propose reverting your shit commits until you do things the proper way
21:01 nrzkt i'm not a cowboy coder, this fix takes me 4-5 hours to code to properly  test everything works as intended.
21:01 nrzkt no
21:02 nrzkt reverting things you don't try to understand is just little boy attitude.
21:02 hmmmmm then convince me that the code is safe and it works by having good unit tests for it
21:03 nrzkt i don't try to convince you, i don't care.
21:03 hmmmmm the onus should be higher to add new code or change new code than it is to prove existing code works
21:03 nrzkt sfan5 est31 and sofar reviewed it
21:03 hmmmmm really
21:03 nrzkt just change you glasses men
21:03 hmmmmm i see sfan5 look over it for less than 3 minutes before saying "looks good!"
21:03 hmmmmm on that very IRC log link you psted
21:03 nrzkt i'm not in sfan5 head and you neither
21:03 hmmmmm s/psted/posted/
21:04 hmmmmm hey nrzkt
21:04 hmmmmm remember the last time you made a "trivial" change?
21:04 sfan5 am i supposed to review every single changed byte for a minute or what
21:04 hmmmmm i was inactive and you went ahead and took the initative to change all the instances of std::list to std::vector
21:04 nrzkt hmmmmm, remember the last time you do a really great contribution PR in MT ?
21:04 hmmmmm and in the process of doing so breaking tons of shit
21:04 hmmmmm thanks for the crashes nrzkt
21:05 hmmmmm you're really helping optimize minetest
21:05 nrzkt hmmmmm, and thanks for you shit in early days of MT too
21:05 hmmmmm my shit never crashed and broke things as much as yours did
21:05 hmmmmm damn i should've been keeping track of what your contribution-to-bug-rate ratio is
21:05 nrzkt if we have release and feature freeze it's fix potential bugs not seen by reviews.
21:06 nrzkt hmmmmm, i don't care, i don't care about you, if you just have a fucking life of criticism yes, go with it, but i don't care, really, you didn't interest me
21:06 hmmmmm feature freeze is a totally dumb, bunk concept only around because the development model sucks
21:06 nrzkt hmmmmm, yes but it's out process, accepted by everybody
21:06 hmmmmm i tried to fix the development model but it keeps getting eroded by fast & loose devs like yourself
21:06 hmmmmm i dunno about you
21:06 nrzkt and you are NOT the project owner. everybody accept our process of devel.
21:07 hmmmmm but i want a minetest that can be released at any commit
21:07 nrzkt no
21:07 nrzkt it's impossible because we are a game not a fucking hmmmmm playground to test things for his CV
21:07 hmmmmm what have i contributed for my "CV"
21:07 hmmmmm it's not impossible
21:08 hmmmmm you're just using the 'game' excuse to break things without consequence
21:08 nrzkt it seems you thought you are the warchief of minetest, in fact you are just a coredev like others, but with a very long absence this year
21:08 nrzkt lol
21:08 hmmmmm i guess it's not much of a discouragement if your only consequence is me getting angry at times and yelling in text on an IRC channel
21:08 nrzkt if somethings broke, it could be repaired, if the review has lost the change problem, and nobody die about this, except you
21:08 hmmmmm nrzkt, is there something wrong with the substance of my argument?
21:09 nrzkt hmmmmm, yes, you talk
21:09 hmmmmm but,
21:09 hmmmmm is there something wrong with the substance of my argument?
21:09 hmmmmm "stfu and code" is not a valid response
21:09 nrzkt and i don't care about hmmmmm which only wants to be the chief and didn't contribute really to PR except complaining without helping PR owners to improve REALLY their PR
21:09 hmmmmm okay?  stfuing and coding is no doubt the main cause of our quality problems
21:09 nrzkt there is no rule for a valid response in fact, then this is the response i give you.
21:10 hmmmmm you need to code less, talk more
21:10 hmmmmm add more tests for the code you do write
21:10 hmmmmm make the tests actually test things
21:10 nrzkt and you need to code more and stfu :)
21:10 hmmmmm "code more and stfu" is a cancerous attitude
21:10 nrzkt and i added tests for the code i do, that's the only part , then stop taht immediately because you didn't read carefully the PR and you are complaining
21:11 nrzkt everybody die because of a cancer these days
21:11 hmmmmm that's not true
21:11 nrzkt except if you are black against police in america
21:11 hmmmmm there are tons of successful software projects in the world
21:11 hmmmmm and they all share something in common
21:11 nrzkt and minetest isn't and will not be because there is a huge concurrence MC.
21:11 hmmmmm the members don't have a bunch of cowboys saying "stfu and code or else it's worthless!"
21:12 hmmmmm "less discussion, less talk, more codezorz!"
21:12 hmmmmm "oh shit we broke something because we didn't talk it over"
21:12 hmmmmm "that's okay we have another commit"
21:12 nrzkt in fact hmmmmm you think everybody should share your mind, but no the world doesn't turn around you, sorry
21:12 hmmmmm the whole while you've been arguing that i lack authority, that america sucks and is fat, and categorically denying my claims
21:12 nrzkt minetest works and move, and you are not following and blocking interesting changes.
21:12 hmmmmm at no point have you actually raised a counterpoint to my concerns
21:13 nrzkt i cannot do the main huge PR which fix everything in one part because it's not really reviewable :)
21:13 nrzkt there are no points, we are not in a debate in fact
21:13 hmmmmm you're just making yourself look like an idiot
21:14 nrzkt i don't care about what you said.
21:14 nrzkt it's not because you tell this that somebody agrees
21:14 hmmmmm your whole attitude is:  "gotta move fast, make lots of changes, generate lots of code" but have you ever had any moment of introspection and wondered how effective all that work is?
21:14 nrzkt or everybody
21:15 nrzkt i didn't said lots of code, i said, i contribute, i do a change, i add tsts to verify my changes before merging, it's review by 3 coredevs, + me and merged, and now there is one coredev which just cry cry cry because he didn't read the PR.
21:15 nrzkt if you are so suspicious about my PRs you are totally free to test them, it's free software, in fact it's a better review than just reading the code
21:16 red-001 joined #minetest-dev
21:16 nrzkt but convince you ? no, sorry, i will not do this anymore, it's like convice a stonebrick
21:16 hmmmmm on the other hand, i don't understand why you feel entitled to have your changes merged into the main project
21:17 hmmmmm you need to convince everybody that the changes you're making provide a concrete improvement to the project and that, to the BEST of your efforts, you've determined that it won't cause any ill effects
21:17 hmmmmm and others need to agree with that
21:17 nrzkt no, not everybody in fact, just 2 coredevs :)
21:17 hmmmmm :)
21:18 hmmmmm the only reason why that restriction is there is because we can't reasonably expect everybody to be online all the time
21:18 nrzkt and in fact, some little refactors are needed into MT to do better things, and just read the player loading / saving code,  + the remote player pseudo memleak issue (yes minetest never free this memory area when player disconnects)
21:18 hmmmmm in spirit, everybody needs to come to the consensus
21:18 nrzkt stop that shit.
21:20 nrzkt you think you are the rules, it's just that, and you invent things that doesn't exist anywhere and nobody has written nor discussed. The coredev approval/review is there to verify the code quality & the feature quality before merging. don't invent philosophical reasons
21:20 hmmmmm i don't personally have much spare time to contribute to minetest these days.  but i do look at the commit page and the IRC channel every so often.  when I see something that indicates to me that the project is going off in the wrong direction, I can't help but have an opinion on it
21:20 nrzkt i remember one time you added a rule without discussing with everybody last year, nobody answers but the rule was added, just lol
21:20 hmmmmm okay what was that
21:21 hmmmmm "don't invent philosophical reasons"
21:21 hmmmmm nobody invented them
21:21 hmmmmm and again, you're not entitled to have your random PR merged just because
21:21 hmmmmm it has to actually do something to benefit the project
21:21 hmmmmm if you want your own playground, make a fork
21:21 nrzkt in fact i didn't do random PR i haven't coded the random part of sending PR yet
21:22 nrzkt i already have a fork in fact
21:22 hmmmmm why not go contribute to freeminer instead
21:22 nrzkt and i'm backporting some things into MT these days
21:22 hmmmmm they share your philosophy of software development
21:22 hmmmmm they're "much better" in your opinion, i'd say
21:22 nrzkt because freeminer hasn't the things i added to my fork and hasn't the philosophy
21:22 hmmmmm lol
21:23 hmmmmm your philosophy is to fuck up the codebase as quickly as possible without regard to anybody else
21:23 nrzkt hmmmmm, i can say same thing about you => you think because you are in a great group you know all on all and you know how the project should go
21:23 hmmmmm sounds just about right
21:24 nrzkt and no, i'm trying to make the SAO model better than before, instead of duplicating attributes and keeping in memory attributes which are not part of a player session but a playersao
21:25 hmmmmm but what about all the gaps in test coverage
21:25 nrzkt just read the fucking server side player code, it's just ugly, and this refactor PR permits to remove some duplicated codes and let the SAO own its attributes, not the remoteplayer which is a .... Session + SAO + Misc player attributes
21:25 hmmmmm yeah i can see what your PR is supposed to do
21:25 hmmmmm i get that
21:25 hmmmmm you can stop repeating yourself now
21:25 nrzkt lol
21:25 hmmmmm christ
21:25 nrzkt in fact i think i should repeat another time because it seems you don't listen what i said
21:25 hmmmmm i raised the concern about no unit tests in that PR
21:26 hmmmmm in response to that
21:26 nrzkt WTF MAN
21:26 hmmmmm you code up this pathetic thing that tests two functions out of everything else
21:26 nrzkt you quoted the unit test which cover it 1 hour ago
21:26 nrzkt THE PR ONLY COVER THIS TEST
21:26 nrzkt THE PR ONLY COVER THIS TEST
21:26 nrzkt THE PR ONLY COVER THIS TEST
21:26 nrzkt THE PR ONLY COVER THIS TEST
21:26 nrzkt THE PR ONLY COVER THIS TEST
21:26 nrzkt THE PR ONLY COVER THIS TEST
21:26 hmmmmm not his happened a couple days ago
21:26 nrzkt you READ NOW ?
21:26 hmmmmm no, this*
21:26 nrzkt THE PR ONLY COVER THIS TEST
21:26 hmmmmm shut the hell up, i'm NOT talking about the exact commit i linked
21:27 hmmmmm so you code up this pathetic thing you call a unit test
21:27 nrzkt then be explicit or useful in your life man
21:27 sofar guys, stop it.
21:27 hmmmmm merge it within less than 5 minutes of opening up the PR
21:27 hmmmmm no feedback whatsoever
21:27 hmmmmm and then you think that somehow satisfies my concerns
21:27 hmmmmm "oh great now i get to push the shiny merge button"
21:27 nrzkt i didn't satisfy you, i hope you have a wife for this.
21:28 hmmmmm whatever
21:28 hmmmmm i don't care
21:28 nrzkt me too, then stop talk.
21:28 hmmmmm i'm just saying, the bet is still on the table
21:28 hmmmmm $5 says there's a new HP sending bug, caused by nerzhul's most recent commits
21:28 hmmmmm AGAIN
21:28 hmmmmm this already happened 3 times before
21:28 hmmmmm you're so freaking careless
21:29 nrzkt in fact i learnt a thing in the life: if you want to be happy, just do some things yourself when you are angry about a thing.
21:29 nrzkt and another tme, you didn't read the PR man.
21:29 nrzkt or have a decent IDE many
21:29 nrzkt maybe*
21:57 Fixer https://i.imgur.com/CamGhHf.jpg <<< now this is good performance
21:59 Fixer https://i.imgur.com/5ju734l.jpg another one
21:59 Fixer optifine was used, and fps hit the cap...
22:00 Fixer in mt it will not only render partially but also probably have like 4 fps...
22:00 * Fixer dives out
22:04 Darcidride joined #minetest-dev
22:05 STHGOM joined #minetest-dev
22:05 STHGOM joined #minetest-dev
22:10 garywhite Has anyone heard of SliTaz?
22:11 garywhite it's a Linux distro that seems small enough that I can run in VirtualBox, and it claims it can run packages from Debian
22:14 red-001 no
22:16 kaeza joined #minetest-dev
22:16 sofar offtopic?
22:16 red-001 wow 35mb
22:34 sofar paramat: I can merge game@1347 with the whitespace fixed if you agree
22:34 sofar oh, I see rubenwardy also approved
22:35 * sofar merges
22:41 rubenwardy joined #minetest-dev
23:32 red-001 joined #minetest-dev
23:39 Tmanyo joined #minetest-dev
23:44 Tmanyo joined #minetest-dev

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