Minetest logo

IRC log for #minetest-dev, 2018-04-13

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

All times shown according to UTC.

Time Nick Message
00:03 Shara Can #258 please be reopened? It was never fixed and, if you spent time speaking to new players you would realise it is not small.
00:03 ShadowBot https://github.com/minetest/minetest/issues/258 -- Make it clearer that the first login to a server is also a registration (Confirm password)
00:04 Shara Confirmation of password was only a suggested solution to this problem, but the way it's been implemented doesn't fix the problem at all.
00:10 Icedream joined #minetest-dev
00:15 paramat reopened
00:15 Shara Thank you
00:16 Shara I have had so many people complain about this... and those are just the players who eventually did manage to work it out and connect to my server. How many players we've lost because they didn't see how to register and gave up, I don't know.
00:17 b0bmq joined #minetest-dev
00:20 paramat yes. sorry for misunderstanding
00:20 Shara Adding a register button would be nice.
00:21 Shara Don't prevent registering the current way... so those who know about it can still quickly make acocunts when needed
00:21 Shara But have an actual button, that actually just goes through to fields that do the same thing
00:21 Shara At least that way those who get lost have a pointer showing them what to do
00:30 srifqi joined #minetest-dev
00:31 srifqi Can we have tabs inside "Multiplayer" tab? That way, we can add "Login" and "Register" tabs separately. The rest will be handled by the client.
00:32 srifqi If we can do that, we can drop the registration confirmation entirely.
00:35 JDCodeIt core.check_player_privs(name, ...) - if the mod passes a table with a function is_player() that returns false - the game dies (because then it checks if you passed a string = false). Is that a bug?
00:35 Shara srifqi: maybe, but not sure.
00:36 srifqi But, then, Android.
00:36 srifqi :")
00:36 Shara Heh
00:36 Shara As long as I can make accounts quickly for testing, I'm happy. The simplest thing would be to make the confirmation optional.
00:37 Shara Of course, that means the registration issue is still present
00:37 srifqi Something like "skip registration confirmation"?
00:37 Natechip_ joined #minetest-dev
00:37 Shara Should be a setting.
00:38 srifqi Which defaults to ... what?
00:38 srifqi I mean, the dialog is completely broken on Android.
00:38 Shara No way to fix it?
00:39 srifqi IDK, was trying. I've fixed the SIGSEGV error.
00:39 srifqi Just, it can't be used.
00:41 Shara I was going to see leave it on by default, since adding a line to minetest.conf is easy enough for those who get annoyed by it...
00:41 Shara say*
00:41 Shara But if it's totally broken on Android... I don't really know
00:41 srifqi The GUI doesn't respond to touch screen input. (._.
00:42 Shara Eww
00:44 twoelk left #minetest-dev
00:44 Shara I guess I'll stay amazed by people who manage to play from Android :)
00:45 srifqi You don't have Android device?
00:46 Shara I have an old tablet somewhere that could  maybe handle MT, but I don't find the thought of trying very appealing
00:55 proller joined #minetest-dev
01:02 paramat will merge game#2091 later
01:02 ShadowBot https://github.com/minetest/minetest_game/issues/2091 -- Add 'spawn' mod to spawn new players in suitable starting biomes by paramat
01:04 srifqi paramat: Can we use binoculars texture as zoom button? #5173
01:04 ShadowBot https://github.com/minetest/minetest/issues/5173 -- Android: Control for zoom?
01:07 srifqi (scaling 4x) :D
01:08 srifqi Ah, it is already used for "Range view" button.
01:09 srifqi left #minetest-dev
01:10 paramat i think android buttons should be just outlines with lots of transparent pixels, to not obscure the world behind them
01:11 paramat hmm maybe a magnifying glass for zoom?
01:14 paramat zoom wil have to be always on-screen so needs to not obscure much
01:15 paramat a magnifying glass is just a circle and a stick so might be good
01:15 paramat merging game#2091 in 5 mins
01:15 ShadowBot https://github.com/minetest/minetest_game/issues/2091 -- Add 'spawn' mod to spawn new players in suitable starting biomes by paramat
01:19 paramat there's a magnifying glass texture in the creative mod
01:20 paramat merging
01:23 paramat done
02:40 AndroBuilder_ joined #minetest-dev
03:19 JDCodeIt left #minetest-dev
03:47 UCUX1Steej joined #minetest-dev
04:06 crayon168 joined #minetest-dev
05:07 rdococ #7238
05:07 ShadowBot https://github.com/minetest/minetest/issues/7238 -- Add "on_metadata_inventory_shiftclick" callback
05:44 Fritigern joined #minetest-dev
07:26 longerstaff13-m joined #minetest-dev
07:36 Icedream joined #minetest-dev
08:03 Zigara left #minetest-dev
08:30 proller joined #minetest-dev
08:37 Icedream joined #minetest-dev
09:52 longerstaff13-m joined #minetest-dev
10:05 Fixer joined #minetest-dev
10:16 proller joined #minetest-dev
10:50 longerstaff13-m joined #minetest-dev
11:49 JDCodeIt joined #minetest-dev
11:54 longerstaff13-m joined #minetest-dev
11:55 JDCodeIt Anyone have any thoughts on my question about check_player_privs http://irc.minetest.net/minetest-dev/2018-04-13#i_5277897
11:57 JDCodeIt Seems that if the mod creates a "fake" player table for some reason, includes the function is_player that returns false = shutdown
11:57 JDCodeIt ...bad mod, you did bad things!
11:58 Wuzzy joined #minetest-dev
11:58 sfan5 you're not calling the function correctly, how is that a bug?
12:01 longerstaff13-m joined #minetest-dev
12:03 Fritigern joined #minetest-dev
12:04 Krock joined #minetest-dev
12:21 JDCodeIt OK, any idea why any is_player() would need to return false?  Seems strange in itself.
12:28 sfan5 "need to"?
12:31 JDCodeIt The API has this: `player_or_name`: Either a Player object or the name of a player - so passing fake player is a no-no?
12:31 sfan5 1) why would you even pass a fake player when you can pass a name?
12:32 JDCodeIt Dunno - pipeworks seems to do it...
12:32 sfan5 2) a fake player would have is_player() -> true, what you're passing is something that resembles an ObjectRef but not a player
12:33 JDCodeIt So internally it's OK because the engine will always return a player object that has this function returning true?
12:34 sfan5 "real" player objects will always have is_player() -> true
12:36 JDCodeIt Defined as a function for future flexibility? Seems like a table entry is_player=true would have also worked in that context.
12:37 JDCodeIt I've seem this come up in mods where there are non-player things interacting with the world - like who authorized the mesecon to give a signal to the door?
12:38 sfan5 player objects are userdata which can't have "properties" like that
12:41 JDCodeIt OK, understood. New topic then...
12:41 JDCodeIt Client.cpp - This class Client has 6 private member variables that are pointers - but the destructor doesn't do anything with these - should it be calling delete on them?
12:42 sfan5 it has more than 6
12:42 sfan5 there's probably no code to delete these because the Client class does not own these
12:42 JDCodeIt Right, I just didn't list the ones in question - but generally, yes or no - delete should be taken care of in the destructor?
12:43 JDCodeIt OK, for Irrlicht owned items - the destructor should iterate through drop()?
12:49 sfan5 if the class in question owns these objects, yes
12:54 JDCodeIt OK, checking some - Game::createClient makes a camera with new() hands that pointer to Client, then forgets to clean up on either shutdown() or ~Game()
12:56 sfan5 8th line inside Game::~Game()
12:56 sfan5 delete camera;
12:57 JDCodeIt Bad eyes...
13:01 JDCodeIt I'll keep digging - I'm trying to solve some parts of #7235
13:01 ShadowBot https://github.com/minetest/minetest/issues/7235 -- Memory is not freed when exiting a world back to the main menu
13:01 sfan5 tip: valgrind
13:03 JDCodeIt Yes, got that - but the strange bit is that this memory is hanging around in the game-menu loop, not totally forgotten at program exit. For sure I will try to catch those smaller ones that are forgotten all the way to exit.
13:05 sfan5 you can break the loop early so valgrind will still be able to help
13:17 JDCodeIt I will try that. Thanks!
13:26 antims joined #minetest-dev
13:31 Lunatrius` joined #minetest-dev
13:36 proller joined #minetest-dev
13:42 DI3HARD139 joined #minetest-dev
13:52 Darcidride joined #minetest-dev
14:12 RichardTheTurd joined #minetest-dev
14:19 longerstaff13-m joined #minetest-dev
14:26 proller joined #minetest-dev
14:32 longerstaff13-m joined #minetest-dev
14:42 JDCodeIt Interesting result of valgrind by stopping after the_game goes out of scope. Lots of stuff locked up in the FontEngine created by the_game's activity, but not cleaned up on destruction
15:09 longerstaff13-m joined #minetest-dev
15:09 longerstaff13-m joined #minetest-dev
15:26 ssieb joined #minetest-dev
15:27 proller joined #minetest-dev
15:38 Gael-de-Sailly joined #minetest-dev
16:08 JDCodeIt valgrind also did not report the origin of some of these, as if they became orphaned at some point - it's another global variable to track down who owns it and who is modifying it: g_fontengine
16:46 Darcidride_ joined #minetest-dev
16:47 longerstaff13-m joined #minetest-dev
17:08 proller joined #minetest-dev
17:15 Lunatrius joined #minetest-dev
17:17 Beton joined #minetest-dev
17:20 Foz joined #minetest-dev
17:40 Beton_ joined #minetest-dev
17:53 Krock joined #minetest-dev
18:27 longerstaff13-m joined #minetest-dev
18:52 Taoki joined #minetest-dev
18:54 JDCodeIt g_fontengine checks out OK. Assumption the only one instance of ClientLauncher will exist.
19:05 paramat joined #minetest-dev
19:08 paramat don't forget 0.4.17 .. i did :)
19:11 paramat game#2108
19:11 ShadowBot https://github.com/minetest/minetest_game/issues/2108 -- Biomes: Remove some warm and humid biomes at altitude by paramat
19:20 paramat i bumped a lot of old issues which have little core dev comment or interest, please could core devs add their comments?
19:31 proller joined #minetest-dev
19:35 paramat in other words, decide if they're worth keeping open
20:17 fwhcat joined #minetest-dev
20:18 paramat I've checked the PRs included in MTG backport 0.4, seem ok, and updated the TODO list of PRs to add to it https://github.com/minetest/minetest_game/pull/1966#issuecomment-362662323
21:24 Fixer joined #minetest-dev
21:31 proller joined #minetest-dev
21:34 Foz joined #minetest-dev
21:35 Raven262 joined #minetest-dev
21:38 proller joined #minetest-dev
21:39 DI3HARD139 joined #minetest-dev
22:31 Gael-de-Sailly joined #minetest-dev
23:13 paramat #7234 updated and ready
23:13 ShadowBot https://github.com/minetest/minetest/issues/7234 -- FOV: Raise lower limit to avoid zoom-loading distant world by paramat
23:39 paramat joined #minetest-dev

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