Minetest logo

IRC log for #minetest-dev, 2018-01-26

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

All times shown according to UTC.

Time Nick Message
00:31 EvergreenTree joined #minetest-dev
01:21 twoelk joined #minetest-dev
01:22 twoelk left #minetest-dev
01:31 ThomasMonroe joined #minetest-dev
02:18 ThomasMonroe joined #minetest-dev
02:51 EvergreenTree joined #minetest-dev
04:49 Darcidride_ joined #minetest-dev
05:27 ssieb joined #minetest-dev
06:27 nerzhul joined #minetest-dev
07:18 sofar joined #minetest-dev
07:33 Dumbeldor joined #minetest-dev
10:32 Jordach joined #minetest-dev
10:56 nerzhul joined #minetest-dev
11:20 Krock joined #minetest-dev
11:58 Fixer joined #minetest-dev
13:13 Jordach joined #minetest-dev
13:33 rubenwardy I
13:33 rubenwardy urgh
13:33 rubenwardy so, I'm placing a massive schematic for CTF map
13:33 rubenwardy I need to emerge first otherwise it doesn't place correctly
13:34 rubenwardy however, this causes the mapblock to be sent multiple times to players as the block changes due to emerge being async
13:35 rubenwardy would you think it's better to:    1) make schematic placing emerge blocks reliably     2) add emerge_area_blocking     3) use an LVM and custom map placer, as LVMs will emerge the block
13:35 rubenwardy hmm
13:36 sfan5 2) makes no sense, emerge_area has a callback
13:36 sfan5 unless there's some reason you can't use that
13:36 rubenwardy it sends the block multiple times
13:37 rubenwardy as it emerges and then sends the block to client
13:37 rubenwardy after the area is emerged I then place the schematic, which causes sends again
13:37 sfan5 that's indeed suboptimal
13:38 rubenwardy the issue I'm facing is the map being all grey for clients, I believe it's due to excess writing like this (packet monitor is fine, consistent 5pps)
13:38 rubenwardy I think it's still trying to send old blocks far away before teleport
13:38 rubenwardy I'm probably talking shit tbh
13:41 rubenwardy if you want to see the effect,   ctf.rubenwardy.com 30001
13:49 ThomasMonroe joined #minetest-dev
13:55 RobbieF joined #minetest-dev
13:55 RobbieF left #minetest-dev
14:04 EvergreenTree joined #minetest-dev
14:45 rubenwardy update: the bug only happens with the map index != 0, ie: the map origin isn't at 0,0,0
14:46 rubenwardy so there's something odd with my code somewhere
14:46 rubenwardy I also can't reproduce it on my local machine, with same minetest version and settings
14:46 rubenwardy -> -hub I guess
14:47 rubenwardy oh, also, dedicated server + client. So only thing that's different is it's a remote server
14:57 nerzhul merging #6940 in ~ 5 mins
14:57 ShadowBot https://github.com/minetest/minetest/issues/6940 -- Fix issues with earlier CSM HUD commit by red-001
15:04 Krock #6948 seems to be ready too, nerzhul
15:04 ShadowBot https://github.com/minetest/minetest/issues/6948 -- Add Android drivers to the video_driver drop-down menu by Wayward1
15:04 Krock but the first approval is .. missing?
15:06 nerzhul I don't really like this pr, it will show OGLES on desktops
15:06 nerzhul but yes it's trivial
15:24 Wayward_One all other drivers are visible on android though
15:25 red-001 and directx is shown on linux which is kinda annoying
15:36 paramat joined #minetest-dev
15:37 ThomasMonroe hey paramat
15:37 ThomasMonroe I redid the fov setting to use a callback
15:37 paramat Krock and MTG devs, game#2016 is updated and ready, makes the ABM 3 times less intensive plus other improvements
15:37 paramat ok
15:37 ShadowBot https://github.com/minetest​/minetest_game/issues/2016 -- Flower spread ABM: Optimise by paramat
15:41 paramat please could an OP 'voice' or whatever Shara in this channel so she appears next to the other devs in the userlist? sfan5
15:45 Krock paramat, whereas the code looks fine I'm not sure whether we should place multiple plants at once
16:33 pgimeno_ ThomasMonroe: re #6957 it's not on topic there but I realized that the Del.Favorite button could be iconized and placed together with the refresh button maybe, because its current placement is weird, it doesn't belong to the connect settings form but to the list. I have a candidate icon if interested.
16:33 ShadowBot https://github.com/minetest/minetest/issues/6957 -- Add a refresh button to the serverlist by ThomasMonroe314
16:41 * celeron55 secret op(erations)
16:45 pgimeno_ http://www.formauri.es/personal/p​gimeno/temp/Star-crossed-out.png
16:48 paramat thanks
16:55 EvergreenTree joined #minetest-dev
17:27 H4mlet joined #minetest-dev
17:36 Darcidride joined #minetest-dev
17:38 Gael-de-Sailly joined #minetest-dev
17:48 ThomasMonroe joined #minetest-dev
18:52 ssieb joined #minetest-dev
19:15 YuGiOhJCJ joined #minetest-dev
19:51 ThomasMonroe paramat: sure! I could do that.
19:56 pgimeno_ ThomasMonroe: did you mean pgimeno_?
19:57 pgimeno_ I guess I should submit it as an enhancement proposal
19:59 MillersMan joined #minetest-dev
20:04 ThomasMonroe oops yes sorry
20:05 ThomasMonroe I have colored nicks on and both of your nick are green for me
20:05 ThomasMonroe nicks*
20:05 pgimeno_ dang, I changed my nick to the _ version because someone else complained about the same thing :)
20:06 ThomasMonroe lol
20:29 Krock Trivial bugfix for 4f688d5, as in reported in #6972  https://pastebin.com/raw/kXTmVpts (tested, works fine again)
20:29 ShadowBot https://github.com/minetest/minetest/issues/6972 -- HUD Fails since 4f688d5
20:29 Krock pushing in 10 minutes if there are no objections
20:32 nerzhul joined #minetest-dev
20:36 antims joined #minetest-dev
20:41 Krock pushing...
20:43 paramat joined #minetest-dev
20:48 ThomasMonroe paramat I added fast, fly, and noclip because of a convo I had with aerozoic in which he said:
20:48 ThomasMonroe [08:15:56] <aerozoic> I've had to explain fast/fly/noclip to so many people on touch screens and they can never find it.
20:49 ThomasMonroe also, out of curiosity, but do phones have fn keys?
20:50 sfan5 where would a phone have fn keys?
20:50 ThomasMonroe idk, how do you enable/disable fog in-game on the phone then?
20:50 Krock ..touch on buttons?
20:52 ThomasMonroe they have equivalents of fn keys on their touch keyboards?
20:53 Krock do you mean F* keys by any chance, as fn is used for a 2nd function of these keys for laptops?
20:53 ThomasMonroe yes
20:53 Krock ah man, you broke the question chain
20:54 ThomasMonroe lol
20:54 ThomasMonroe sorry :P
20:54 ThomasMonroe was I supposed to kee it going?
20:54 ThomasMonroe :P
20:54 ThomasMonroe keep*
20:54 Krock I'm quite sure they also have access to the Advanced Settings formspec in the main menu, if there's no button in-game
20:55 ThomasMonroe so a setting in-game would not be a terrible idea?
20:56 Krock too trivial. that setting is changed once on set-up and done.
20:56 ThomasMonroe true
21:07 paramat ah hmm
21:08 paramat does android have controls for fast, fly, noclip?
21:08 ThomasMonroe I think they have fast and fly
21:09 ThomasMonroe not sure about noclip
21:10 ThomasMonroe because I had a friend(Abramowitz971) who couldn't use noclip bcs he was on a tablet, but idk if that got updated or not
21:11 paramat if android is missing some controls we would need to think about how to add them, i think your PR should wait for that discussion, for now i suggest only concerning yourself with desktop MT
21:11 ThomasMonroe ok
21:12 red-001 ty Krock
21:13 red-001 I really need to read the spec for auto
21:14 paramat hopefully soon we will have people working on android and improving it, they can tackle controls then
21:15 Krock that's why I prefer the long version instead of replacing everything with auto :3
21:16 ThomasMonroe ok, fair enough :D
21:16 red-001 in fairness it worked with the pointer
21:16 Krock ThomasMonroe, but instead.. mouse sensitivity? we also have an issue for that one
21:16 Wayward_One noclip, fast, fly, f5, and f7 are all available on android by pressing the gear icon on the right of the screen.
21:17 red-001 and references should be little more then syntax suger for pointers
21:17 ThomasMonroe I was thinking about that, Krock, but its more of a one time setting
21:20 Wayward_One and btw ThomasMonroe, judging by the three numbers in your friend's name, it seems quite likely they're using a clone, possibly on iOS, most of which charge for extra privs
21:22 ThomasMonroe Wayward_One, possibly
21:22 ThomasMonroe but he didn't have a problem with any other privs
21:24 Wayward_One ah.
21:24 Shara If it's not the official version, all bets are off anyway.
21:24 Wayward_One yep
21:24 ThomasMonroe touche Shara
21:25 ThomasMonroe I don't remember what he used though
21:26 paramat thanks Wayward_One so those certainly are not justified
21:27 Wayward_One np
21:33 ThomasMonroe joined #minetest-dev
21:34 paramat ugh we're over 100 PRs
21:34 ThomasMonroe lol I was getting ready to say something about that
21:39 rubenwardy celeron55, I adopted one of you PRs. When does the issue actually happen?
21:39 rubenwardy #6950
21:39 ShadowBot https://github.com/minetest/minetest/issues/6950 -- Check that Lua's v3f inputs are not NaN by rubenwardy
21:39 rubenwardy can't seem to make a NaN float in Lua
21:39 rubenwardy even with div by 0
21:42 pgimeno_ when I launch Lua and print(0/0) I get -nan
21:43 rubenwardy reproduced
21:43 rubenwardy but then   local a = 0/0
21:43 rubenwardy a == nil
21:43 rubenwardy ahhh
21:43 rubenwardy that's why
21:43 pgimeno_ you mean a==nan?
21:43 rubenwardy you can't store directly
21:44 rubenwardy you have to check immediately
21:44 rubenwardy oh wait no
21:44 celeron55 umm... well, it was obviously something
21:44 rubenwardy in the interpretter local is for each line
21:44 pgimeno_ also keep in mind that nan is always ~= nan
21:44 celeron55 maybe get one from C++?
21:44 celeron55 or convert an object into a number or something
21:44 rubenwardy ah, I was trying 1/0 instead of 0/0, which gives you inf instead of nan
21:45 celeron55 i think it was discussed in IRC so logs will tell
21:45 rubenwardy oh true
21:45 celeron55 on IRC*
21:45 rubenwardy also,    !(a==a)  won't work with fast-math
21:45 rubenwardy so the check only works in debug, as I believe we use -03 to build which includes -fast-math
21:45 pgimeno_ > nan = 0/0 print(nan==nan, nan~=nan)
21:45 pgimeno_ falsetrue
21:46 pgimeno_ I think the last one is wrong, but that's me
21:59 celeron55 http://irc.minetest.net/minetest-dev/2015-11-01
22:01 celeron55 the NaN discussion is there
22:02 celeron55 so basically, it's unknown where the NaNs came from
22:02 celeron55 and the PR is for making it possible to see where they come from
22:02 Fixer joined #minetest-dev
22:02 celeron55 well that's what i'm able to decipher anyway
22:20 Megaf_ joined #minetest-dev
23:10 Fixer joined #minetest-dev
23:10 BakerPrime joined #minetest-dev
23:48 Foz joined #minetest-dev
23:59 Gael-de-Sailly joined #minetest-dev

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