Time  Nick        Message
03:33 Zeno`       hi
03:33 Zeno`       bbl :)
08:12 nrzkt       what do you think about changing the unit tests ? They doesn't need to be there at each server launch. Unit tests are only for integration testing, using jenkins/travis, they permit to define if the programm pass the tests and then PR can be merged
08:14 est31       PR 3/4?
08:15 nrzkt       no, i think in general, unit tests aren't defined to be used on production. Imagine if libreoffice launch all its unit tests at its launch :)
08:16 nrzkt       i'll propose a patch today if i get time to disable unit tests and use a proper way to define them
08:16 est31       yes I agree
08:17 est31       in fact, unit tests shouldn't be even compiled into the binary
08:17 est31       who wants unit tests in production binaries
08:17 est31       they should only go for debug builds
08:17 est31       (dunno how they are currently)
08:17 nrzkt       ofc course and only run when we call the --enable-unittests argument
08:18 nrzkt       and this must return an exit code if all pass, not launch the program
08:18 est31       yes
08:19 nrzkt       in fact in release mode tests aren't launched (but compiled)
08:19 nrzkt       if think we could do better
08:25 nrzkt       i'll propose a patch soon
08:45 nrzkt       it will include unittest run for travis builds
08:45 nrzkt       est31: thanks for your patches, we have modified it to make release yesterday
08:56 est31       nrzkt, yea seen them
08:57 est31       couldnt fix them in time anymore
08:57 est31       but you both did it
08:58 est31       novatux could have attributed me, but for those few lines thats no problem :)
09:27 nrzkt       nore kahrl, could you review #2354 please ?
09:27 ShadowBot   https://github.com/minetest/minetest/issues/2354 -- Unit tests must be done only at integration process by nerzhul
09:32 nrzkt       est31: if you have comments on the 2354 don't hesitate :)
09:39 est31       nrzkt, perhaps not write "A must be done at ..." but "Do A at ..."?
09:39 est31       otherwise ok
09:40 est31       havent tested though
09:43 nrzkt       if only disable some core functions in release mode :)
11:30 nrzkt       okay :!
11:30 nrzkt       https://travis-ci.org/minetest/minetest/jobs/51358132
11:30 nrzkt       that works into the PR too :)
11:31 nrzkt       if a core-dev can review and validate ?
11:38 rubenwardy  How could I smooth the camera and be independant of FPS? #2325
11:38 ShadowBot   https://github.com/minetest/minetest/issues/2325 -- Add camera smoothing and cinematic mode (F8) by rubenwardy
11:38 nrzkt       first can you rebase using Zeno- modifications ?
11:38 rubenwardy  I am doing that
11:48 rubenwardy  #2325 now uses rangelim
11:48 ShadowBot   https://github.com/minetest/minetest/issues/2325 -- Add camera smoothing and cinematic mode (F8) by rubenwardy
12:17 sfan5       nrzkt: why remove unit tests from release builds? maybe a bug (that the unit tests would catch) only occurs in release builds.
12:19 est31       they add further weight to the binary
12:20 sfan5       omg 500 bytes more in 2015!
12:20 sfan5       if you're that low on disk space comment the code out
12:21 est31       I'm not the dude thats that conservative. ppl didnt like that I recommended to install git
12:21 est31       https://github.com/minetest/minetest/pull/2073
13:00 nrzkt       sfan5: because unit tests are util for integration, not production. They permit to find bug. If you don't have bug at integration in your unit tests you don't have to use it in production release. It's the model
13:01 nrzkt       release build = debug build without symbol. If you have a bug in release build it's on debug build, and if no problem in debug build, same in release build
13:38 est31       sfan5, you there?
13:40 shadowzone  So, why isn't celeron55 voiced?
13:40 nrzkt       about what?
13:41 est31       voiced in this irc channel
13:45 nrzkt       because he doesn't want ?
13:45 nrzkt       or he doesn't care ?
14:13 JakubVanek  hello, i found network bug. Minetest didn't crash, but on the console, I received this error message: http://pastebin.com/AgSRr91f
14:13 JakubVanek  i was just running over world
14:14 JakubVanek  i have minetest 0.4.12 windows 32-bit msvc
16:52 est31       kahrl, nore, you have time to review #2225? hmmm isn't "associated with minetest" anymore, as he sais.
16:52 ShadowBot   https://github.com/minetest/minetest/issues/2225 -- Fix for getCraftRecipes implementation is very slow by gregorycu
16:52 shadowzone  I'll review it as well.
17:30 ShadowNinja est31: Reviewed.
17:56 est31       ahh here you are :)
17:57 est31       nrzkt, I'm preparing a patch to get the network breaking changes of #2308 into master
17:57 ShadowBot   https://github.com/minetest/minetest/issues/2308 -- Server should send other players inventories when its needed
17:58 est31       but wont work :(
18:02 est31       nrzkt, do you know what I've done wrong here: https://github.com/est31/minetest/commit/56978ec92a625f349a1923832f624dce9ae81787
18:11 est31       I guess the issue is that the serialized inventory "string" still can contain \0 characters, so I shouldn't send it that way?
18:33 nrzkt       est.0
18:33 est31       yes
18:33 est31       I'll have to do the length calculation manually?
18:34 est31       via std::string(pkt->getString(offset),(u16) pkt->getU8(length))?
18:34 nrzkt       do *pkt << name << s; no need to c_str(), it's already a string
18:35 nrzkt       you write a char* and read a std::string, maybe it's this :)
18:37 est31       yea thats it
18:40 est31       I dont think I will be abled to make full support soon for 3rd party client inventory setting, but I'll try to make the network breaking changes so that 3/4 can be merged?
18:43 nrzkt       what would you want to do wit this setting in fact ?
18:44 nrzkt       and you code isn't secure, here anyone can send the packet to other connected players
18:45 est31       just server can send that packet?
18:45 est31       its for bug #2308
18:45 ShadowBot   https://github.com/minetest/minetest/issues/2308 -- Server should send other players inventories when its needed
18:45 nrzkt       yes i don't think clients must have this possibility
18:45 nrzkt       if only LUA api is able to do this call, this can be good. In what usage you want to see that ?
18:46 est31       server will need to parse formspec, and when it sees that lua requests to show player a formspec containing another players inventory, then it will send that inventory
18:46 est31       otherwise not
18:47 nrzkt       but what the client will do with other player's inventory ? in fact you sent it, yes, but after ?
18:47 est31       display as formspec
18:48 nrzkt       okay, but this will be a RO formspec or RW ?
18:48 est31       RO
18:48 est31       but its for server admins
18:49 est31       dunno if PilzAdam needs RW formspec
18:49 est31       https://github.com/minetest/minetest/blob/0.4.12/src/client.cpp#L2439
18:49 est31       this is the client side code that reads the inventory
18:51 est31       I'll find out what the client tries to do when it wants to communicate an inventory change to the server
18:52 nrzkt       okay
18:52 nrzkt       i'll think we will let the PR 3/4 since the 28 february on github and i'll merge it to master on the 1st march
18:52 nrzkt       Calinou, can you provide us the changelog please ?
18:53 Calinou     OK
18:53 nrzkt       this delay permit to find all bugs, but protocol will permit discussions after the merge, and we could add the other PR's protocol break into master together
18:54 nrzkt       this will break the connection to old servers, then keep a stable-0.4 client on the side for connecting to your production servers :)
18:57 nrzkt       what do you mean Calinou by default walking speed increase ?
18:58 est31       nrzkt, can you replace "See InventoryAction in inventory.h" with See "InventoryAction in inventorymanager.h"?
18:59 Calinou     increase walking speed from 4 to 4.5 (or 4.75)
18:59 Calinou     currently it is too slow
18:59 Calinou     and change sneaking from 1.3 to 1.5 or so
18:59 nrzkt       where est31 ?
19:00 est31       in networkprotocol.h
19:00 est31       I can open PR if you want
19:00 nrzkt       no need
19:00 nrzkt       the problem is on master ?
19:00 est31       yes
19:01 est31       and your branch too
19:01 est31       guess you push this to master and then rebase
19:01 est31       so that it doesnt get forgotten, and on rebase, it gets back in due to error
19:01 nrzkt       i will rebase my pr after, i'm not on linux, i'm playing on windows (booo, but far cry 4 only on windows)
19:03 nrzkt       commited
19:03 Calinou     I am working on changelog
19:04 rubenwardy  #2325
19:04 ShadowBot   https://github.com/minetest/minetest/issues/2325 -- Add camera smoothing and cinematic mode (F8) by rubenwardy
19:15 Calinou     the changelog is up!
19:16 Calinou     feel free to edit if something big is missing
19:17 rubenwardy  Is that it? Did you go back to 0.4.11 release and then look through all commits since?
19:18 Calinou     yes
19:18 Calinou     I didn't note very small changes
19:20 rubenwardy  Do older changelogs list "very small changes"
19:23 nore        Add modname convention checking (Novatux) <-- actually, est31 did the most work on it
19:23 nore        I just made the changes that needed to be done so we could release...
19:24 nore        so you should attribute that to him
19:24 VanessaE    also I don't see mention of that "blocks full of garbage" fix either.
19:24 rubenwardy  fixing
19:25 kilbith     furnace re-write (and many other commits by PilzAdam) are not included though
19:25 rubenwardy  Yeah, you missed a lot
19:26 nore        VanessaE: wasn't it introduced after 0.4.11?
19:26 kilbith     rubenwardy, the network rework is not in the release
19:26 VanessaE    I think it made it into 0.4.11 actually didn't it?
19:27 rubenwardy  https://github.com/minetest/minetest/commits/stable-0.4
19:27 rubenwardy  Read from there
19:28 VanessaE    no matter, it's not important now.
19:29 kilbith     i double-checked, either the patch 1/4 or 2/4 are not included
19:34 rubenwardy  Corrected
19:34 rubenwardy  And I'm bored now. I got to https://github.com/minetest/minetest/commits/stable-0.4?page=2#
19:34 kilbith     also, mgv7 has been worked
19:35 VanessaE    rubenwardy: http://xkcd.com/859/
19:36 rubenwardy  lol
19:36 VanessaE    (last line of "bug fixes")
19:36 rubenwardy  fixed
19:51 sfan5       we have a problem with the git version: -- *** Detected git version 0.4.11-241-ge357577 ***
19:51 sfan5       the 0.4.12 tag is not "part" of the master branch
19:52 sfan5       all builds that were build with git version detection will say 0.4.11
19:52 est31       sfan5, do rm CPack* && rm -r CMakeFiles
19:52 est31       or make clean clone
19:52 sfan5       did that
19:52 sfan5       also did that
19:54 est31       there is *** Will build version 0.4.12-dev *** AND there is *** Detected git version 0.4.11-240-g627c747 ***
19:54 est31       lol
19:55 sfan5       the 0.4.12 tag does not describe something in the history of master -> git describe picks 0.4.11
19:55 est31       yea :/
20:00 est31       time to manually generate it?
20:07 est31       tuesday, at backup time
20:08 est31       oh sorry wrong #channel
20:09 est31       sfan5, you working on sth? if not I could try
20:25 nrzkt       thanks for the changekig
20:27 nrzkt       you have added some things which are not in 0.4.11:  Ignored old entities from 0.3 (Novatux)
21:12 fz72        the best solution for the missing tag on master would be to create a new dev branch based on master and overwrite master with stable0.4 and then merge the new dev branch to master
21:19 est31       thats ugly
21:19 est31       see #2357
21:19 ShadowBot   https://github.com/minetest/minetest/issues/2357 -- Make git version discovery work on master by est31
21:23 fz72        but this is another version as in stable0.4, or am I wrong?
21:24 est31       its the same diff, just another hash
21:28 fz72        ok
21:29 ShadowNinja est31: Your solution is even uglier.  I like fz72's idea better.
21:31 est31       It would require changing history at every commit
21:32 est31       s/commit/release/
21:33 est31       stuff is out, do with it what you want. if you modify & commit it pls @attribute me in the commit msg I'm off
22:13 nrzkt       no fz72 it's ugly
22:19 VanessaE    no offense to fz72, but that solution sounds really ugly... sounds an awful lot like rewriting history, and it's WAY too late for that.