Time  Nick        Message
02:31 paramat     now pushing #2476
02:31 ShadowBot   https://github.com/minetest/minetest/issues/2476 -- Mgv5/mgv7: Add desert temples if desert stone detected in mapchunk by paramat
02:43 paramat     done
02:52 paramat     good to see you here hmmmmm, you are much missed
03:08 hmmmm       hello paramat
03:08 hmmmm       so is there anybody who could reopen the PR
03:11 paramat     technically yeah ... i think nerzhul wants to know in detail what the problems are with his code
03:11 hmmmm       sure, but there's no policy saying that he's allowed to close a PR like that after a certain (short) amount of time
03:12 paramat     correct
03:12 hmmmm       i'm not sure about specific details, i was just reporting what i've heard from several other developers and users
03:12 hmmmm       i assumed they had already reported them and it seems like they did but nerzhul refused to acknowledge them
03:16 paramat     the PR asks a lot so people need to collect details on the problems and post them in the PR, i can re-open it to make that possible
03:17 paramat     BTW i'm completely neutral on all this
03:18 hmmmm       maybe the commits shouldn't have been cemented in
03:18 est31       commenting on closed bugs/prs is possible.
03:19 est31       see here :https://github.com/minetest/minetest/pull/2459
03:19 est31       I'll delete the comment in ~10 mins
03:19 paramat     AFAIK the changes are in a dev branch? so i guess if the work is bad it will not be merged
03:37 Zeno`       There's no dev branch; HEAD is the "dev branch"
03:38 Zeno`       There is a 0.5.0 dev branch which has the incompatible protocol changes in it
03:38 paramat     ah i see, thanks
03:39 Zeno`       ANyway I don't see a closed PR
03:41 paramat     ah cool
03:42 Zeno`       somebody must have re-opened it :)
03:46 Zeno`       I'm having trouble working out some bugs; e.g. apples not being eaten; and the tricky method of being able to give yourself (as a player) infinite health
03:47 hmmmm       I hope you guys consider what I've been saying all along and save minetest from turning into another GNOME 3
03:48 Zeno`       Also the "freeze"/reset-frame that I was talking in channel about the day before yesterday
03:49 paramat     we, and nerzhul, should take hmmmm's warnings very seriously considering his standing here
03:50 Zeno`       we probably should even disregarding them because I can't even (easily) bisect these
03:50 hmmmm       I would prefer for people to see it for what it is over blindly trusting me based on status/standing
03:51 Zeno`       the invincible hack is funny if it wasn't kind of serious. /me joins PvP server to run amok
03:52 Zeno`       And yes, I haven't opened an issue. I don't really want to publicize it before I have a fix :)
03:53 Zeno`       Personally I would branch from where the huge changes started, apply the changes in that new branch, and cherry pick them back to master one at a time but with more testing
03:54 Zeno`       Things happened very fast there for a while and considering how few active devs there are nothing was really tested I don't think. It was impossible to keep up with the pace
03:55 Zeno`       Also, considering the network changes. The first patch is fine. The others are simply to pave the way for the protocol breaking change if I understand correctly
03:55 Zeno`       So they should only be in that 0.5.0 dev branch at this time
03:55 Zeno`       Unless I'm missing something
03:55 Zeno`       But they are in master
03:56 Zeno`       Anyway, that's why I think the issue should remain open for discussion
03:57 hmmmm       from what I see, nerzhul has many stability issues fixed in the 0.5-dev branch but isn't backporting them to master
03:59 Zeno`       oh, is that where the fixes are happening. I was kind of wondering about that
03:59 * Zeno`     looks
04:00 hmmmm       one prominent fix is the memory leak stuff
04:01 Zeno`       umm which ones :)
04:01 Zeno`       there are quite a few leaks atm
04:01 hmmmm       I'm talking about Network rework: Packet fixes & cleanups c106df6
04:01 Zeno`       of course they are cleaned up properly before program exits so they're kind of hard to find :/
04:01 hmmmm       the one with the NetworkPacket allocations
04:02 hmmmm       I wasn't aware there were multiple
04:03 Zeno`       c106df68 isn't in master?
04:03 * Zeno`     looks again
04:03 hmmmm       I thought that was his so-called "3/4"
04:03 Zeno`       oh I see why it's not. Because it does many things (some breaking protocol)
04:03 hmmmm       maybe I'm mistaken
04:04 hmmmm       also lots of cases where a network packet is allocated with 0 length
04:04 hmmmm       i have no idea how this shit doesn't crash instantaneously
04:05 hmmmm       but I'm obviously wrong and have no proof
04:08 Zeno`       oh this is the one where pointers to local objects are passed to other functions
04:13 Zeno`       err
04:13 Zeno`       I can't connect the new client to the new server
04:13 Zeno`       protocol mismatch
04:13 Zeno`       wth
04:14 Zeno`       now it's connected. Weird
04:15 est31       Zeno`, from favourites?
04:15 est31       yes, I know it
04:15 Zeno`       you experienced that as well?
04:16 Zeno`       how did you know what I was saying? Are you psychic?
04:16 est31       no nrzkt but he reported it to me, as I did the checking
04:16 est31       guess :P
04:17 paramat     hmmmm if you still want to i'd like you to contribute a multi-mapgen underworld, i will merge it as long as the 'nether' word isn't used =) i was thinkng instead we could use a name of a Finnish or Nordic underworld
04:17 hmmmm       it depends if it sounds sufficiently cool
04:18 est31       Zeno`,  #2401
04:18 ShadowBot   https://github.com/minetest/minetest/issues/2401 -- Disable protocol version checking for favourite list by est31
04:19 est31       also explains how the bug happened the first place
04:19 VanessaE    paramat: "The abyss"
04:20 VanessaE    (just looked up a few synonyms for underworld)
04:20 Zeno`       Hey, I know. Call it underworld
04:22 hmmmm       less talk about naming things, more code
04:23 paramat     i have a thread somewhere ..
04:25 paramat     https://forum.minetest.net/viewtopic.php?p=121427#p121427
04:26 hmmmm       it fits in with the rest of the celeron finnish theme
04:30 paramat     (and there's a few good ones in the first post too. i'm sure Nordic mythology would provide more good choices)
04:31 est31       nordic != finnish
04:32 est31       finnish are own group with hungarians
04:32 est31       nvm
04:34 paramat     sure, sorry, guess i mean Scandinavian
04:37 paramat     and 'nether' itself can be traced back to various cooler sounding words
04:42 hmmmm       i hope you realize 'nether' isn't as simple as teleporting the player to a deep underground coordinate and generating negative 3d noise with a couple random pools of lava
04:43 hmmmm       several client-side effects, notably color table adjustment, needs to be implemented first
04:48 paramat     i see
04:53 paramat     there have been many requests recently for above/below realms, i have floatlands covered and am considering adding these to v5/v7 as options, but i'm not so inspired to work on underworlds
04:55 VanessaE    the problem with making above/below worlds is shadown propagation
04:55 VanessaE    shadow*
05:01 paramat     yeah, i have tested that though, new mapchunks never cast shadows onto older
05:03 paramat     so i was only considering sparse scattered floatlands. also i have an idea for how to taper shadows during light calculations
05:09 hmmmm       keep in mind it has to work with Map::spreadLight and Map::unspreadLight
05:12 paramat     ah of course..
05:12 hmmmm       otherwise it'd be downright trivial
05:13 Zeno`       bisect is very annoying now that cmake has to be run between checks :(
05:14 hmmmm       what made that change necessary?  just curious
05:15 est31       ah that dying bug is a regression??
05:15 Zeno`       the not dying bug?
05:15 Zeno`       or the umm, dying one
05:15 Zeno`       I dunno... bisect is too time consuming with the cmake step added
05:16 est31       usually, you start bisects with commits you know that are good
05:17 est31       You can bisect inside full minetest history within 13 steps
05:18 Zeno`       umm ,yes
05:18 est31       but gtg. $est31 >> ~/bed
05:18 Zeno`       but there are changes between those commits that need cmake to be re-executed
05:19 est31       but you know its a regression?
05:19 est31       thats a start :)
05:19 Zeno`       no. I haven't gotten around to that one yet. Still wondering why c55 added 4d744cf87
05:21 est31       bye
05:21 est31       good luck
05:21 Zeno`       probably because on in-game formspecs, but...
05:21 Zeno`       cya
05:26 Zeno`       It's only half implemented :/
05:26 Zeno`       ugh
05:26 Zeno`       bbl
05:45 Zeno`       addSpiltPacket() hehe
09:09 nrzkt       ok guys, jenkins now generate a daily Android build: http://jenkins.unix-experience.fr/job/minetest-android-OfficialRepository-RollingRelease/
09:31 nrzkt       ok i will push a little thing which can fix the dead player move issue
09:33 neoascetic  > "dead player move" some kind of zombies mod? :D
09:41 nrzkt       VanessaE: https://github.com/minetest/minetest/commit/3067d782169a79ab3b13e8ba3d445b60d0d477c5 can you tes tthis patch for your players which are moving while dead ? :)
09:41 nrzkt       as it seems :p
10:09 nrzkt       also  there is a problem, breath is handled client side, it's not normal :)
10:10 neoascetic  nrzkt related issue https://github.com/minetest/minetest/issues/1894
10:11 neoascetic  I've even put a small bounty on it
10:14 nrzkt       you are right
10:16 nrzkt       i add it to 0.5 roadmap
10:17 nrzkt       the damage mustn't be handled by client, if server lag... it lags. Players must be in sync with server for these points
10:17 neoascetic  yeah
10:17 nrzkt       added to #2370
10:17 ShadowBot   https://github.com/minetest/minetest/issues/2370 -- [roadmap] Minetest 0.5.0 objectives
10:19 neoascetic  Btw I am going to release alternative to MMDB someday, I already have some code
10:19 nrzkt       MMDB ?
10:19 neoascetic  minetest mods database
10:20 nrzkt       okay. I could be great to have a better and userfriendly mod database. like apps.owncloud.org model, it's very cool model which integrate with owncloud instances for installing mods :)
10:20 neoascetic  Are there any ideas on main menu redesign?
10:21 nrzkt       I will push another modification to stop handling breath if player is dead. In those cases we don't need to change player breath. I also add player->isDead function
10:41 kilbith     neoascetic, perhaps something like that : https://lut.im/2sfkPpxD/8QKbkneJ  fewer, bigger, more convenient buttons on the side instead of the tabs
10:43 kilbith     which are often likely to be screwed with the tinker on scaling (going over the header)
10:43 neoascetic  Personally, I don't like these "tabs", no matter where they are placed
10:44 kilbith     or a MC-like menu, player-friendly
10:45 neoascetic  Better than MC-like, that's the goal :) Is there any good designed among us?
10:46 kilbith     http://fc06.deviantart.net/fs70/i/2012/227/3/c/minetest_gui_by_dakal_fr-d5b61yl.png
10:46 neoascetic  wow, that's much better, I like it!
10:46 kilbith     or : http://i.imgur.com/gf8hWAD.png
10:46 neoascetic  especially colors
10:48 neoascetic  second one is like some source-engine based games :)
10:48 kilbith     HL²-like menu
10:49 neoascetic  I also think it would be good to have some protocol support, minetest://some-server-ip for example - when user clicks on it, minetest launches and opens this server
10:52 neoascetic  Here is also idea for menu: for example, we may have playable menu, i. e. user is in some house built from boxes and may interact with some objects to call forms associated with them
10:53 kilbith     that sounds too utopistic - this way doesn't work as expected usually
10:54 neoascetic  why?
10:54 neoascetic  kilbith btw where did you find that? http://fc06.deviantart.net/fs70/i/2012/227/3/c/minetest_gui_by_dakal_fr-d5b61yl.png
10:55 neoascetic  I've tried to search on deviantart but didn't find this images
10:55 kilbith     old artifact left on the forum
10:56 neoascetic  What a pretty artifact!
11:04 neoascetic  I want this menu for minetest, really, even with these tabs and borders
11:10 kilbith     at work then ;)
11:11 kahrl       dunno, it seems quite unintuitive to me. 1. I can't really tell what the tab icons are supposed to mean. 2. The first tab shouldn't be an "Advanced"-style tab, it should be singleplayer. 3. Where do you select the world?
11:11 kilbith     agreed, that ought be more informative
11:15 kahrl       I guess it being an advanced style tab is due to the sketch being made in a time when the minetest menu was like that
11:16 neoascetic  Yes, icons should be more understandable, but I am wondering about colors and base idea
11:16 nrzkt       ok then, sorry for to be late, i push the fix for breath + player is dead now (was a bit busy, and the test was done late :p)
11:21 kahrl       I think on the artistic side it's fine (but I'm not an artist, if I try to make something it usually has horrible color schemes and layouts and such)
11:22 kahrl       and I can't tell in advance if a color scheme is going to be well received by people or not
11:22 kilbith     basically we need an artist, an ergonomist and a dev for the menu
12:40 srifqi      sfan5
12:42 srifqi      nrzkt
12:43 sfan5       hm?
12:44 srifqi      will this #1964 and #1900 merged?
12:44 ShadowBot   https://github.com/minetest/minetest/issues/1964 -- Update Indonesian Language by srifqi
12:44 ShadowBot   https://github.com/minetest/minetest/issues/1900 -- Graphic bugs, micro lags and Nvidia Optimus problems.
12:44 srifqi      oops, i mean #1990
12:44 ShadowBot   https://github.com/minetest/minetest/issues/1990 -- Add language option in settings menu by srifqi
12:48 srifqi      sfan5: ^^^^^
12:58 sfan5       srifqi: i don't know
12:58 srifqi      :(
13:26 nrzkt       srifqi: i will look at  1964 soon, i have a production to deliver at  work before
13:26 srifqi      nrzkt: thanks
14:01 VanessaE    nrzkt: the death loop bug still needs fixed
14:02 VanessaE    if the server has damage disabled but a player still manages to take damage (e.g. from a mod that disregards the setting), the player can't "finish" dying.
14:03 nrzkt       is this normal that player take damage from a mod if the damages are disabled ?
14:03 VanessaE    probably not.
14:03 VanessaE    but mods can directly set the player's health apparently
14:03 nrzkt       and if you said from a mod this is a good path
14:03 VanessaE    [off] for example, technic corium still damages players - this is how the bug was triggered.
14:04 nrzkt       then for the death loop bug you are saying it's there with fix, right ?
14:04 nrzkt       my patch doesn't fix that, it fix player move if player is dead
14:04 VanessaE    the workaround when this loop occurs is to re-enable damage on the server.
14:04 nrzkt       hot setting change i presume ?
14:05 VanessaE    restart the serevr, let the player sign on and they will then "finish" dieing and everything goes back to normal
14:05 VanessaE    no, seems to only work after a restart
14:05 VanessaE    the loop is is persistent across server restarts and player re-logins
14:05 VanessaE    so the only solution when it happens is to enable damage and restart the server.
14:05 nrzkt       a strange state is saved then
14:06 VanessaE    yeah
14:06 VanessaE    saved in the map itself apparently.
14:06 nrzkt       int ObjectRef::l_set_hp(lua_State *L) is the call
14:07 nrzkt       but we have: getServer(L)->SendPlayerHPOrDie(((PlayerSAO*)co)->getPeerID(), co->getHP() == 0);
14:10 nrzkt       in fact the function is set_hp, i could not disable the damage here, a do_damage call will be better to call to use the enable_damage parameter
14:11 nrzkt       can you precise what is the state of the player when he is stuck ? i need to find the precise state in the algorithm
14:11 VanessaE    player has zero hp, client keeps red-screen "you died" every so often.  that's about all I know
14:12 VanessaE    the player can move around (well they used to, but you patched that just earlier) and they cannot dig/pladce blocks.
14:12 VanessaE    place*
14:12 VanessaE    they can chat, and they look "dead" to everyone else
14:13 nrzkt       oh the client is notified
14:14 nrzkt       then i must search in other place
14:14 nrzkt       cannot dig/place is normal, because dig/place already has the fix i do on move :)
14:21 VanessaE    different topic:  why doesn't the F5 display identify unknown nodes?
14:22 VanessaE    (when I can punch them and see what they are in the server log)
14:22 srifqi      nice idea!
14:36 nrzkt       VanessaE: if i'm right, when player click on respawn he dies another time ?
14:36 VanessaE    yep
14:36 VanessaE    over and over without end.
14:36 VanessaE    and they're greeted with "you died" as soon as they sign in, also
14:37 nrzkt       at first respawn he is teleported to home or another respawn location before die ?
14:37 VanessaE    he is not teleported at all.
14:37 VanessaE    player seems to remain at the location they were at when they "died"
14:53 nrzkt       hmmmm, ok
14:53 nrzkt       and if they disconnect and reconnect, they are also died ?
14:54 VanessaE    yep
14:54 VanessaE    even if the server has been restarted (unless the admin turns damage back on, then the player is allowed to "finish" dieing)
14:55 nrzkt       yes. I'm looking at player login. the deathscreen is sent there
14:55 nrzkt       but this looping problem... i must investigate in AsyncRunStep
14:56 hmmmm       Hi nrzkt, from looking at the channel log it seems like there's an abundance of evidence of problems with your network rework
14:56 hmmmm       What were you saying before about how there's no evidence
14:57 nrzkt       please talk about facts, not ideas
14:57 nrzkt       point the code
14:57 hmmmm       nrzkt, did you not acknowledge this:
14:57 hmmmm       10:10 	neoascetic 	nrzkt related issue https://github.com/minetest/minetest/issues/1894
14:57 hmmmm       10:11 		I've even put a small bounty on it
14:57 hmmmm       10:14 	nrzkt 	you are right
14:57 hmmmm       but right after you said you'll fix it "in the 0.5 roadmap"
14:57 hmmmm       that's not acceptable, bugs need to be fixed right away in upstream
14:58 hmmmm       I noticed you did this same exact thing with a couple of other bugs that people pointed out to you
14:58 hmmmm       I am very concerned about this trend
14:58 nrzkt       hmmmm we are not talking about dev-0.5 and not the devel planning STOP.
14:59 nrzkt       we have an issue and i only route it to 0.5 because this issue cannot be applied to master.
14:59 hmmmm       nrzkt:  You can explain if you'd like, maybe I misunderstood
14:59 nrzkt       that's all.
14:59 nrzkt       please tell me what is your time. i'm GMT+1
14:59 hmmmm       nrzkt:  Why can't the fix be applied to the master branch?  Does the bug not exist in the master branch?
15:00 hmmmm       Again, maybe I'm misunderstanding the situation
15:00 nrzkt       one thing at each time i'm at work and don't try to confuse me or users.
15:00 nrzkt       if you want to fix bugs, please send PRs, else let us do the bug fixes.
15:00 VanessaE    uh...
15:00 hmmmm       nrzkt, who is "us"?
15:00 nrzkt       the core-devs
15:01 hmmmm       nrzkt, nobody else is touching your network changes
15:01 nrzkt       stop talking about network changes
15:01 hmmmm       nrzkt, why are you getting so defensive?
15:01 nrzkt       VanessaE tells it's present in 0.4-stable.
15:01 nrzkt       then please read all conversation instead of only some points
15:02 hmmmm       nrzkt, why not just explain it if I'm not understanding correctly
15:02 nrzkt       i explain you on the PR answer i do this morning (GMT+1), please read it :)
15:02 nrzkt       and before yesterday evening VanessaE didn't tell me that information because they haven't see it on stable-0.4, but now we all have the information.
15:03 hmmmm       nrzkt, if you expect your code to stay in the minetest codebase you need to promptly fix known bugs present in master and not in an upcoming development branch
15:03 VanessaE    how about this:  what's so special about the master branch that the client-side-only damage issue can't be fixed there?
15:03 nrzkt       bugs are not fixed in dev-0.5 because atm there only are network rework PR 3/4 and master commits
15:03 hmmmm       what?  that didn't even make sense
15:03 nrzkt       VanessaE because client are doing some things themselves ?
15:04 nrzkt       if we add server side handling for damages and we ignore the commands (it is possible), we need to look at client code for every version <= 0.4.12 to see if  the client doesn't do some gameplay issues when he know that he is dying or something else
15:05 VanessaE    why is this a problem?
15:05 nrzkt       but this is not the point here. This issue is more a design issue than a bugfix
15:05 nrzkt       because client can do uncontrollable fix when connected to servers with this patch ?
15:05 hmmmm       nrzkt, if the problem didn't exist before your commits then i'd say that qualifies as a bug
15:05 nrzkt       uncontrollable things*
15:06 hmmmm       fix your shit.
15:06 nrzkt       it's not my shit. And please be polite.
15:06 nrzkt       and give me good drivers for your Dell servers infrastructure please.
15:07 nrzkt       for my*
15:07 VanessaE    there used to be a bug where the player wouldn't "finish" dieing, but it didn't result in a loop like this.
15:07 nrzkt       VanessaE we are not talking about same thing
15:07 nrzkt       server side damage handling != death loop. It's not the same thing
15:07 VanessaE    nrzkt: I know.  I'm referring to the death loop now.
15:08 nrzkt       i'm working on death loop, the other fact is a feature and it's not needed atm and will not solve the issue properly for all clients :)
15:09 VanessaE    it doesn't have to solve the issue for all clients - only for the last version or two.  server owners can choose to block old versions anyway.
15:09 nrzkt       that's not the problem here
15:09 nrzkt       0.4 is designed for retro compat.
15:09 hmmmm       nrzkt, a design requirement of minetest is backwards compatibility
15:10 hmmmm       I don't really think it's a good idea to break compatibility with older versions and neither do a vast majority of other developers and users alike
15:10 hmmmm       nrzkt, you can't make huge unilateral decisions like that without a consensus
15:11 nrzkt       hmmmm, it's what i said here, then help to patch, or stop wasting my time with a useless conversation.
15:11 hmmmm       nrzkt, no, you need to stop patching shit and creating more work for the rest of us
15:11 nrzkt       i can fix the bug, then
15:11 nrzkt       bug not with your trolling ideas
15:11 hmmmm       nrzkt, need I remind you that you need approval from other developers to commit changes to upstream
15:11 nrzkt       in fact hmmmm i'm the network maintainer
15:12 hmmmm       saying "I'm going to commit X in 1 hour if nobody has any concerns"
15:12 hmmmm       is not enough
15:13 hmmmm       nrzkt, I think it would be a very smart idea to focus on fixing all existing bugs with your first two network changes as soon as possible, in the master branch
15:13 nrzkt       and we have a politic which said if a commit is little we can push. read the dev wiki and stop waisting my time then. Now i work on the bug , bye.
15:13 hmmmm       we can't have blatant breakages like that
15:13 hmmmm       or else the breaking commits may need to be removed
15:14 hmmmm       nrzkt, if you feel like I'm wasting your time you can easily stop responding to me
15:14 nrzkt       Point the bugs instead of telling pure shit. We are devs, proove your talk with a line of code or a patch.
15:14 Wayward_One why was double-clicking to exit the in-game menu disabled?
15:14 hmmmm       it was disabled for android only
15:15 Wayward_One apparently not
15:15 hmmmm       iirc there was some kind of android issue with a menu popping up or losing focus or something on double click
15:15 hmmmm       ahh, that's a problem
15:15 nrzkt       no
15:15 nrzkt       it was disabled for non android only.
15:15 nrzkt       hmmmm please read the patchs. https://github.com/minetest/minetest/commit/e74b8da54e4443773404f2642916fd2f2b7f625e
15:16 hmmmm       maybe so, haven't been around for a while.
15:16 nrzkt       then stop telling false things and read the commits properly.
15:17 hmmmm       nrzkt, I am not saying anything false
15:17 nrzkt       you are saying ideas. Not facts. When you will tell me facts i will listen to you.
15:18 hmmmm       nrzkt, I don't understand why the burden of proof is on everybody else not only that a bug exists, but that we need to fix your own defective code as well
15:18 nrzkt       stop it
15:18 hmmmm       nrzkt, I think you're a little too over-confident in your code
15:18 nrzkt       or revert to minetest-0.2 many bugs will disappear, it seems it's you mean to fix bugs.
15:18 nrzkt       i'm confident with it because it was inspired by my experience for a 2K client program ?
15:19 hmmmm       I am more than willing to believe a problem exists if somebody says so and I fix it ASAP
15:19 nrzkt       then fix it
15:19 nrzkt       isntead of reverting commits
15:19 nrzkt       and stop whine
15:19 VanessaE    s/I fix/they fix/
15:20 Zeno`       nrzkt, why do you keep writing "jenkins build please" in comments?
15:21 nrzkt       to verify the build for this commit to be sure before merge because android failed
15:21 Zeno`       oh, I see
15:21 Zeno`       I thought it must have caused jenkins to build or something
15:22 nrzkt       the builds are triggered automaticly but sometimes android build fail, i think there is a problem in the makefile to fix (it's not a memory issue there)
15:23 hmmmm       why should we fix your broken commits
15:23 hmmmm       I don't get it
15:24 nrzkt       my commits aren't broken ,else many and many users will complain on github. But it's not the case. Hopefully :D
15:24 hmmmm       fwiw your changes didn't even make any substantial improvements.  all you've done is shift code around into different subdirectories and changes a lot of instances of std::list to std::vector, guaranteeing maximum merge conflicts
15:24 nrzkt       and if you don't xwant to fix commits, and you are not constructive, you are not welcome to minetest-dev.
15:24 VanessaE    "hopefully".  nrzkt remember what I said earlier about the 1-in-10 (or 1-in-100) rule...
15:24 hmmmm       nrzkt, you aren't kicking me out of here
15:25 hmmmm       so please do fuck off
15:25 Zeno`       none of this is constructive
15:25 nrzkt       please be polite, it's in our rules :)
15:26 hmmmm       I'd love to contribute to minetest
15:26 hmmmm       but if the project is going to be a train wreck, it'd be a waste of my time
15:27 nrzkt       then, as OpenBSD developpers (which are better than proprietary devs for compat code and quality) said: speak less, code more.
15:27 nrzkt       if you waste my time and the project waste your time, go ahear
15:27 nrzkt       go ahead*
15:27 Zeno`       oh ffs
15:27 Zeno`       I want hmmmm back as a core dev
15:28 Zeno`       surely there is some way to work this out
15:28 hmmmm       I really don't want to revert commits
15:28 hmmmm       I'll openly admit that PR was a publicity stunt.  I actually want you to include fixes to the master branch instead of 0.5-dev
15:29 nrzkt       i don't know what to fix because the bug report said: "THERE WAS A BUG FIX IT MEN". okay no information, it's like end users in fact, not constructive.
15:29 hmmmm       and I want you to stop being an asshat in general, breaking compatibility and disregarding the existing code
15:29 hmmmm       no, we don't need to "go fast and break things"
15:30 hmmmm       that's the reason why I removed myself as a core developer
15:30 nrzkt       hmmmm stop saying bad things. The network code needs a break . It's too old the design is not good and many packets are broken and cannot be fixed without a break. I agree for compat; but we absolutely need a rework NOW. not in 5 years because the code will become spaghettis. NOW.
15:30 hmmmm       you seemingly took over and started changing not only the code, but the culture and the development methodology
15:30 hmmmm       I don't want to work in an environment like that
15:31 hmmmm       besides, who the hell ARE you to just step in and start screwing everything up?  I don't know how you didn't encounter more resistance
15:31 nrzkt       i don't want to change the culture stop saying shit. I only said we need to break the network one time. Not every day, not every week not every year. ONE TIME. UNDERSTAND ?
15:32 hmmmm       nzkrt:  You can trivially add a compatibility layer for 0.4.x networking
15:32 nrzkt       no
15:32 hmmmm       Why not
15:32 nrzkt       because the session layer is also reworked
15:32 hmmmm       oh, right, because you have this idea that you must switch to TCP
15:33 nrzkt       and we must rework authentication
15:33 hmmmm       for a so-called "network professional", you certainly don't know how to update an existing protocol in a low impact manner
15:34 nrzkt       lol
15:34 nrzkt       in fact when the client init is broken. You don't have choice
15:35 nrzkt       i know how to keep a proper compatiblity code, but spaghettis code is not a compat code, it's spagghetis mamma mia
15:35 nrzkt       now let me work on the bug pointed by VanessaE.
15:36 Zeno`       I'd like to say something from the perspective of both a dev and also a person who runs a server
15:37 nrzkt       :)
15:37 hmmmm       I will admit the client init is broken, but there's nothing wrong with maintaining the current workaround
15:37 Zeno`       I won't be upgrading my server to the new protocol even if it's substantially better. Why? Because I'd lose (probably) 99.9% of my users. A compatibility layer would be good but I stated this quite a long time ago.
15:37 hmmmm       if you want to break shit and work with an unstable codebase, perhaps Freeminer is more your thing
15:38 nrzkt       we don't talk about instability hmmmm.
15:38 hmmmm       Zeno`:  doesn't matter, nrzkt has his mind made up already
15:38 nrzkt       we are talking about working code.
15:38 nrzkt       with the compat we cannot evolve the protocol to add new engine functions handled by client
15:38 VanessaE    I would lose most of my users as well if I were to upgrade to the 0.5 protocol.
15:39 nrzkt       some packets are missing for some lua needs
15:39 Zeno`       Personally I think that this can be done in a backwards compatible manner
15:39 Zeno`       But maybe I'm missing something fundamental
15:40 nrzkt       in fact, advertise users can update their clients
15:40 Zeno`       nrzkt, how?
15:40 Jordach     we actually need (somehow) auto updating clients
15:40 nrzkt       with the Denied packet
15:41 nrzkt       you connect, we deny you instanly with the proper message
15:41 Zeno`       most of my users, unfortunately, use unofficial clients
15:41 nrzkt       forks are a pain...
15:41 nrzkt       you talked about android ?
15:41 Zeno`       well, yes. That's a perfect example
15:41 VanessaE    nrzkt: many users use unofficial clients *that haven't been updated by their vendors since 0.4.10*
15:42 VanessaE    that's the important thing here
15:42 nrzkt       If you agree i will look at the releasing process and publish a stable-0.4 build to google play store this week or next week
15:42 nrzkt       i have a google play store developper account
15:42 Zeno`       That's one approach. What is wrong with a compatibility layer though?
15:43 nrzkt       many of the minetest apps are using ads or a little charge, i will give the APK free (on play store a free app cannot be non free after the publication)
15:43 nrzkt       i studied the layers. And it will be very very very very very difficult to maintain some parts of the protocol compatible
15:43 Zeno`       hold on. Minetest is an engine
15:43 Zeno`       does irrlicht impose these restrictions?
15:44 nrzkt       it's not the session layer, this is the easy part. It's the layer 5
15:44 Zeno`       if someone wants to make money by using the engine as it's intended to be used then good luck to them
15:45 Zeno`       either it's an engine or an app. At the moment it's an engine
15:45 Jordach     Zeno`, nrzkt means that he will publish a non ad supported official android build of minetest on the Play Store
15:45 nrzkt       and if i found a workaround, but it's not sure, some packets are really broken
15:45 Zeno`       Jordach, which is great :)
15:45 nrzkt       and Zeno`, on playstore it's not only an engine it's an app, because minetest_game will be included :)
15:46 Zeno`       nrzkt, what did you mean by (on play store a free app cannot be non free after the publication) though?
15:46 nrzkt       google doesn't allow a free app to become non free (money)
15:46 Jordach     Zeno`, if the app is free the author can't change the price at any time - they'd have to remove it then reupload
15:46 VanessaE    nrzkt: yes they do.
15:46 VanessaE    nrzkt: in-app purchases.
15:46 Zeno`       what?
15:46 Zeno`       so the license has changed?
15:47 nrzkt       i develop android apps since 1 year and it was the rules :p
15:47 Jordach     Zeno`, it'd force the non=paying users to buy an app they use possibly daily
15:47 Zeno`       minetest is not an app
15:47 Zeno`       it's an engine
15:47 nrzkt       in-app require a privilege shown to users, if we have a program with only network privilege it's good :p
15:48 Zeno`       if you're going to change the license and the meaning of the license I think celeron55 needs to be involved AND all the people who have ever contributed
15:48 Zeno`       I have no plan to make an android paid/ad version, but this is silly
15:49 nrzkt       which licence ?
15:49 Zeno`       LGPL
15:49 nrzkt       but why change it ?
15:49 Zeno`       you're saying people are not allowed to charge money for minetest
15:49 nrzkt       i'm talking about money not about licence
15:49 nrzkt       i don't say that
15:50 nrzkt       i said we will publish a free software with free price :p
15:50 Zeno`       so what does "on play store a free app cannot be non free after the publication" mean?
15:50 Zeno`       that violates the LGPL
15:50 nrzkt       Jordach answers to you
15:51 Jordach     Zeno`, the author who publishes the app can't change the price at any given moment
15:51 Calinou     we need someone who is OK with giving $25 to Google :P
15:51 nrzkt       a free price app cannot become payable :p
15:51 VanessaE    free/gratis, not free/libre, Zeno`
15:51 nrzkt       i have the account Calinou
15:51 Zeno`       free/libre means free to charge
15:51 VanessaE    Calinou: I'd have given the $25 myself if someone would have committed to a stable, ongoing build before.
15:51 Zeno`       and that is what GPL/LGPL means
15:51 VanessaE    Zeno`: I know.  that's the stipulation.
15:51 Zeno`       it always has
15:51 sfan5       <Calinou> we need someone who is OK with giving $25 to Google :P
15:52 VanessaE    no, nonono
15:52 sfan5       celeron has a google play dev account
15:52 VanessaE    (L)GPL is about free/libre as in freedom.  free/gratis means free of cost/charge.
15:52 neoascetic  someone please merge this stuff: https://github.com/minetest/minetest/pull/2471
15:52 nrzkt       Australian !
15:52 nrzkt       neoascetic: no we are talking about money !
15:52 neoascetic  Heh
15:53 Zeno`       yes, I am talking about free/libre which is what the GPL and LGPL promote
15:53 nrzkt       stupid english for free. In france we have gratuit for money and libre for licences :p
15:53 VanessaE    Zeno`: right.  from the standpoint of free/libre, yeah you can't change the license.  BUT from the standpoint of free/gratis, you can't change what's set on the Play Store after the app is uploaded.
15:53 VanessaE    once it's free/gratis, it's stuck that way until you replace it with a new app.
15:53 VanessaE    (in-app purchases aside)
15:54 Zeno`       If I want to make a Linux distro called BlueHat and want to charge a price for you to get my version then that's fine. Correct?
15:54 nrzkt       if you app is payable, you can change the prices, but if it was free/money or become free/money it could not become payable :)
15:54 VanessaE    yep
15:54 VanessaE    you can do that, Zeno`.
15:54 VanessaE    provided RedHat doesn't sue you for trademark reasons :P
15:54 Calinou     are we dicussing about making the official app paid? :/
15:54 nrzkt       no
15:54 Zeno`       well, I'll call it BlueShorts then
15:54 Zeno`       lol
15:54 neoascetic  what about then?
15:54 nrzkt       not paid
15:55 VanessaE    Calinou: no, just the semantics of putting an app on the store and what can happen to it after it's up and available for download.
15:55 nrzkt       a FREE/money FREE/licenced app :p
15:55 VanessaE    Zeno`: in that case, you're probably...ahem... FREE and clear ;)
15:55 Zeno`       If you look on the mt forums I've always said this
15:55 neoascetic  i. e. free as a bear and free as a speak
15:55 VanessaE    neoascetic: "Free as in beer, and free as in speech"
15:56 nrzkt       i want to break free !
15:56 neoascetic  yeah
15:56 Zeno`       I'm not charging $ and don't plan to, but... people misunderstand the lgpl and gpl I think
15:56 VanessaE    I'm pretty sure bears consider themselves free no matter what we say :P
15:56 T4im        fsf has actually a faq entry on paying for gpl/lgpl stuff, and they say its generally ok and wanted by them, as long as the source is still distributed ;)
15:56 Calinou     people misunderstand free software licenses all the time
15:56 Zeno`       T4im, exactly
15:56 Calinou     T4im, this: http://gnu.org/philosophy/selling.html
15:56 T4im        yea
15:56 Calinou     you can charge $1, $100, $1,000,000
15:56 VanessaE    Zeno`: indeed so - GPL or LGPL or AGPL doesn't matter - you can charge for software under these licenses as long as the license is otherwise fulfilled (e.g. source code is available easily, etc etc etc)
15:56 nrzkt       you can sell free software, it's not a problem. But you must redistribute your modifications
15:57 Calinou     actually, if you're a third party, you must release source code even if you didn't modify it
15:57 * Calinou   learned this the hard way
15:57 nrzkt       BSD licence, and the licence problem disappear <3
15:57 nrzkt       Zeno` #1964 ok?
15:58 ShadowBot   https://github.com/minetest/minetest/issues/1964 -- Update Indonesian Language by srifqi
15:58 Calinou     BSD license, and the proprietary forks become impossible to remove…
15:58 Calinou     except if you shut them up through the forum or such (using guidelines)
15:58 nrzkt       right, but it depend of your needs :p if you don't care about that BSD licence is the better
15:59 srifqi      Zeno`: ^^^^^
16:02 Zeno`       Calinou, not if it's dynamically linked
16:03 Zeno`       For example, I can choose to use the LGPL version of Qt to make an application and never release any source code if I don't modify Qt and I dynamically link it
16:03 neoascetic  http://www.wtfpl.net/wp-content/uploads/2012/12/wtfpl-strip.jpg
16:03 Zeno`       I think we;ve gone a bit off track though hehe
16:05 Zeno`       I doubt anyone is going to pay for a $1,000,000 version of the mt client though
16:05 neoascetic  Only if you release it on iOS probably
16:07 Zeno`       lol
16:07 Zeno`       1964 is fine
16:07 Zeno`       it's a language that I cannot read :)
16:08 Zeno`       but google translate seems to show it adds nothing nefarious
16:09 nrzkt       pushed
16:09 Zeno`       what is iOS?
16:10 neoascetic  Iphone
16:10 sfan5       >Iphone
16:10 sfan5       it hurts
16:10 Zeno`       oh, those things
16:10 sfan5       it's iPhone
16:10 neoascetic  it doesn't matter
16:10 sfan5       it does
16:25 srifqi      Zeno`: How about this? #2480 I don't know how to resolve.
16:25 ShadowBot   https://github.com/minetest/minetest/issues/2480 -- Add mg spflags in dlg create world by srifqi
16:28 srifqi      nrzkt: ^^
16:33 VanessaE    srifqi: eeek, rebase that
16:33 VanessaE    also, a screenshot would be a good idea
16:33 srifqi      VanessaE: how to rebase that "complicated" commit?
16:33 srifqi      or, should i re-fork minetest?
16:35 VanessaE    *shrug*
16:38 srifqi      okay, I'll ask another "active"-dev
16:38 srifqi      maybe
16:38 nrzkt       erf you merge...
16:39 nrzkt       then. Create another branch from master, cherry-pick the commits you do to that branch, then reset the PR branch to the commit before your first PR commit and rebase the PR branch with your newly cherry picked branch
16:42 srifqi      1)merge to where?
16:42 srifqi      2)what do you mean cherry-pick?
16:42 nrzkt       you merge master into your PR , you musn't do this
16:42 nrzkt       cherry pick is a git subcommand
16:43 srifqi      wait, explain like explain to a noob
16:43 nrzkt       difficult :p
16:44 nrzkt       create a new branch like if you want to do a new PR. then in this branch play every commit you do in your PR branch (it's the cherry pick)
16:45 srifqi      play?
16:45 nrzkt       apply
16:45 srifqi      wait, please,
16:45 srifqi      *opening GTW
16:46 srifqi      Github for Windows, XD
16:48 nrzkt       http://minetest-ftp.unix-experience.fr/Android/Minetest-release-unsigned.apk
16:48 nrzkt       if somes want to test this build :)
16:48 nrzkt       if some wants*
16:48 neoascetic  srifqi checkout this branch. reset --hard to master. apply this patch (https://github.com/minetest/minetest/pull/2480.diff) and commit/force-push
16:49 nrzkt       it's a solution but can have error because he merges master branch
16:49 srifqi      so?
16:50 srifqi      new branch from my `master`
16:50 srifqi      then? how to cherry-pick?
16:51 nrzkt       man git ? :)
16:51 srifqi      okay, wait
16:52 nrzkt       the apk doesn't install on my android device... strange
16:53 srifqi      nrzkt: allow non-market install?
16:53 nrzkt       already done because i need it for my devel
16:53 nrzkt       the phone start installation and said: unable to install. Strange
16:54 srifqi      has it packed rightly?
16:54 nrzkt       i don't know, it's done by Makefile, i'm not at home then i cannot review it properly
16:54 nrzkt       i will try to install it with TWRP to be sure
16:56 nrzkt       and also with PM from command line
16:58 srifqi      nrzkt: after cherry-pick done, then? commit it?
16:58 nrzkt       commit to your PR branch with force option
16:58 nrzkt       push*
16:59 srifqi      wait, it's in another branch, how to push to another branch?
16:59 nrzkt       go to your PR branch
16:59 nrzkt       do a git log, and find the first commit from master branch before all your PR commits
17:00 nrzkt       get the SHA trace and type: git reset --hard <shatrace>
17:00 nrzkt       next, git rebase <your new branch>
17:00 nrzkt       or you can also force push to the remote repository from your new branch if you doesn't want to get the loca lbranch
17:01 srifqi      1)git checkout my-PR?
17:01 srifqi      2)git log?
17:01 srifqi      3)git reset --hard <sha>?
17:01 srifqi      4) git rebase my-newBRANCH?
17:02 srifqi      5)how to force push to remote?
17:02 srifqi      nrzkt: see above
17:03 nrzkt       git push origin <pr branch> --force
17:04 nrzkt       i must go, the AndroidManifest.xml is corrupted in the build it's why this doesn't install.
17:04 nrzkt       see you
17:04 srifqi      see you
17:48 srifqi      nrzkt, is it already fixed?
17:48 srifqi      oh, he is still outside
17:48 srifqi      *quit
17:55 srifqi      okay Krock thanks
19:58 rubenwardy  argh. Travis build has been changed. My PRs now fail :(
20:30 neoascetic  ShadowNinja
20:30 neoascetic  I wonder why we need all these lines: https://github.com/minetest/minetest/blob/master/CMakeLists.txt#L147-L154
20:31 neoascetic  Couldn't we simply copy entire "games" directory, could we?
20:31 VanessaE    there might be "testing" stuff in there that shouldn't be in the final install
20:31 VanessaE    say, a modified "minimal" game or so
20:32 VanessaE    (as in modified and given a different name)
20:32 neoascetic  minimal game going to be copied anyway
20:32 neoascetic  ok, got it
20:33 ShadowNinja neoascetic: No, you don't want to copy, eg, ".git".
20:34 neoascetic  there is EXCLUDE keyword in cmake's install
20:35 neoascetic  I just want include a few games in the games and want CPack build package with these games within
20:35 neoascetic  Currently I need to modify CMakeList by my hands. Is there better way to do this?
21:18 neoascetic  There is such logic for android: https://github.com/minetest/minetest/blob/master/build/android/Makefile#L654-L672
21:19 neoascetic  it would be nice to have something similar
21:58 est31       ~tell nrzkt if you publish minetest on google play store, you should perhaps ask for inclusion into f-droid app store too
21:58 ShadowBot   est31: O.K.
21:59 est31       alot of people chose to not install play store, as it needs to be a system app, and other problems with it (like proprietary)
22:01 Calinou     F-Droid inclusion is not easy to do
22:02 Calinou     it has its own build system and such
22:02 Calinou     just distribute direct APK download
22:02 Calinou     (for now)
22:04 est31       Calinou, yes, but direct apk downloads are dangerous -- they are likely to become stale
22:04 est31       with f-droid you can update with one click/touch
22:04 Calinou     yes, but we're likely to run into issues
22:05 Calinou     perhaps because we're using Irrlicht/C++ for instance
22:05 est31       they also have option of manually uploading apps
22:05 est31       I think firefox isn't built automatically for example
22:05 est31       and thats a very central app
22:06 est31       so if they trust their devs in that important point, then non-relevant apps are ok too
22:06 est31       just have to motivate an f-droid dev to build minetest for us