Time  Nick           Message
00:18 MinetestBot    kabou: Feb-07 08:18 UTC <sfan5> I know that 2.1.0-beta3 is fine, we also use it on Windows. but Arch straight up switched to the git version
00:20 kabou          sfan5 sfan5 I built libluajit 2.1.0-beta3+git from latest commit
00:20 kabou          then I rebuilt minetest 5.6-dev also latest commit
00:21 kabou          no problems building, no compiler errors
00:21 kabou          and it also runs just fine
00:22 kabou          so I'm inclined to wonder if #12054 is not due to unrelated breakage in arch linux or rollerozxa 's tool chain
00:22 ShadowBot      https://github.com/minetest/minetest/issues/12054 -- Minetest does not compile with LuaJIT v2.1 dev branch
00:23 kabou          perhaps he uses an unorthodox compiler?
00:24 kabou          I built the above with gcc 10.2.1
07:01 erlehmann      what would be a better way of shading a HUD element in low light conditions than applying a texture modifier to the texture? https://mister-muffin.de/p/BtQS.jpg
07:01 erlehmann      (the HUD element in question is the mcl_maps map – you may have noticed that i am improving mcl_maps recently)
09:35 MTDiscord      <GreenXenith> I dont see anything wrong with a texture modifier
09:42 MTDiscord      <luatic> erlehmann: if you really want it to adhere to the environment lighting, you might want to make it a first person attachment instead of a HUD element
09:43 erlehmann      luatic wel
09:43 erlehmann      l
09:44 erlehmann      luatic but can i still draw the player marker on it like i do now?
09:44 erlehmann      player location marker is a hud element
09:44 erlehmann      i am not sure how well attachment location can be controlled
09:44 erlehmann      i'll stay with the texture modifier though
09:45 erlehmann      bc i might want to zoom the map later
09:45 erlehmann      and i know how to scale up HUD elements, but not how to zoom into an attachment ^^
09:47 MTDiscord      <luatic> you would have to draw the marker using a texture modifier xD
09:47 MTDiscord      <luatic> so the best course of action is presumably indeed just going with a texture modifier
11:07 MTDiscord      <fatalerror420> It can be controlled... how well it can be controlled tho? Well its up for debate
11:07 MTDiscord      <fatalerror420> It can be controlled precisely, but very stupidly
11:12 erlehmann      luatic i am already drawing the banner map markers with a texture modifier
16:07 independent56  Is there a mod which will automatically remove mapblocks which have not been modified? My map.sqlite keeps on getting bigger, and most of it is areas explored by noobs.
16:18 MTDiscord      <Jonathon> See buckaroos tool, not sure if its db agnostic or not
16:21 BuckarooBanzai Independent56: i have a working nodejs tool for postgres maps and i considered redoing it in golang for sqlite and postgres
16:21 independent56  nice
16:22 BuckarooBanzai I'll ping you up if i need testers for the new version :D
16:22 independent56  I probably won't use it in production, but i can make a local copy of my world and try it.
16:23 independent56  Not until it gets eneough certification.
16:23 independent56  Oh also, minor flaw: https://buckaroobanzay.github.io/mtinfo/#/ Greengrocer's Apostrophie's.
16:23 independent56  Or maybe thats just the way of pluralising an acronym. I would use ABMs.
16:24 BuckarooBanzai But it works with actual nodename checks, not modified time: basically a nodename whitelist, everything else gets pruned around it. Also respects area protections in the current version
16:24 Pexin          I had an instructor once who grilled us on the difference between an acronym and a simple abbreviation
16:25 BuckarooBanzai English isn't my first language, will fix it when i work on that next time
16:25 independent56  makes sense
16:26 BuckarooBanzai Pr's welcome btw ;)
16:26 Pexin          laser (light amplification something something) was technically not an acronym, but "became" a word so retroactively is an acronym. or something.
16:26 Pexin          english is a puke mess   :]
16:32 independent56  I hate english too
16:32 independent56  https://buckaroobanzay.github.io/mtinfo/#/mods There is an emopty bullet point at the top
16:32 independent56  I hate english too
16:32 independent56  Sorry, i had a small keyboard error
16:33 independent56  i'm forced to speak it because of the world and because i am a native englishman
17:00 ghoti          I regularly get server crashes because of an unhandled postgresql error in multicraft. Is this something I can reasonably ask about here?
17:01 rubenwardy     we can't provide support for forks, as their code is quite different
17:01 BuckarooBanzai ghoti: i'm curious, why multicraft and not minetest?
17:01 ghoti          BuckarooBanzai: better support for mobile clients, I'm told.
17:02 BuckarooBanzai Ah, for the old ones, ok :)
17:03 ghoti          rubenwardy: would you happen to know if the postgres schema has changed in the predominant fork? I was under the impression that this was an area where the fork had not diverged much..
17:03 ghoti          At least, until recently. (And I'm running an older version.)
17:03 rubenwardy     no idea
17:03 rubenwardy     we have changed our code though
17:04 BuckarooBanzai The schema hasn't really changed fyi
17:04 rubenwardy     In terms of a server, multicraft gets your support for v5 and v0.4, but doesn't improve mobile clients in any other way
17:04 ghoti          Ah. So, 'duplicate key value violates unique constraint "auth_name_key"' is not something I should see in minetestserver? :)
17:05 BuckarooBanzai Heh, thats actually a race condition in the engine :) haven't seen that in a while...
17:05 rubenwardy     supporting v0.4 is problematic as it greatly limits the features you can use
17:06 rubenwardy     https://github.com/minetest/minetest/issues/11877
17:06 ghoti          BuckarooBanzai: I have had 74 crashes due to this since the bgeginning of the year.
17:08 ghoti          rubenwardy: thanks for the link, it is good to know I am not alone! I have also not been able to reproduce it .. yet it keeps happening.
18:02 BuckarooBanzai ghoti: this could be monkey-patched with an upsert instead of the insert. The root problem still persists though :/
18:16 independent56  https://imgur.com/a/Olhf1Gs
18:16 independent56  How do i make the face of this wall less bare?
18:16 sfan5          hang some paintings 8)
18:18 MTDiscord      <Warr1024> vary the materials, add some cutouts/inclusions, add another layer with some embellishments...
18:19 independent56  Can i have an example of another layer?
18:19 ghoti          why not learn to love the moiré?
18:30 MinetestBot    02[git] 04appgurueu -> 03minetest/minetest: Send HUD flags only if changed 13b9ee29a https://github.com/minetest/minetest/commit/b9ee29a9456a66c3670b2a1389878e0896395f58 (152022-02-08T18:28:32Z)
18:31 independent56  How does worldedit work on unloaded mapblocks? Does it try to load them, or build theblocks when the mapgen is finished, or does it place the blocks directly?
18:32 sfan5          when worldedit uses a voxel manip the engine will try to load blocks
18:32 MinetestBot    02[git] 04Gael-de-Sailly -> 03minetest/minetest: Noise params serialization fixup 130dd8e8c https://github.com/minetest/minetest/commit/0dd8e8c2427f12a8a7d080132e9c588c66baa41b (152022-02-08T18:30:49Z)
18:32 MinetestBot    02[git] 04sfan5 -> 03minetest/minetest: Update MinGW used by CI 13ce199d6 https://github.com/minetest/minetest/commit/ce199d6f9e65d6ba51cbf2e1f948586ddc617317 (152022-02-08T18:31:24Z)
18:33 MinetestBot    02[git] 04sfan5 -> 03minetest/minetest: Remove awful Mingw32 workarounds 13ba6fbc4 https://github.com/minetest/minetest/commit/ba6fbc417ecb812345c1747f42b6606dfc8e1d5b (152022-02-08T18:31:24Z)
18:34 ghoti          BuckarooBanzai: sure, ON CONFLICT would surely help; it would be great for this to be the default, but then we'd have to *require* pg 9.5+ rather than just recommending it.
18:35 ghoti          independent56: I was under the impression that asking worldedit to act on a block caused it to be loaded. (I have not checked.)
18:35 MinetestBot    02[git] 04appgurueu -> 03minetest/minetest: Increase max objects per block defaults (#12055) 13ad1da99 https://github.com/minetest/minetest/commit/ad1da994b2b9d660c41f8ba784ff830aa2693d3b (152022-02-08T18:33:10Z)
18:35 independent56  I am thinking that, based on mapserver renders, worldedit places the blocks directly without waiting for stuff to load around it.
18:36 BuckarooBanzai ghoti: i suggest using a latest postgres version, there were a few performance- and index/space-improvements since <=9.5 :/
18:37 BuckarooBanzai <=9.x is EOL anyway...
18:38 BuckarooBanzai independent56: WE uses vmanip to place the blocks, if the area isn't generated it will stay that way, you may have to emerge the area beforehand
18:38 independent56  Does that mean that it places after or before the blocks are loaded?
18:38 sfan5          so has nobody found out how duplicate users entries make their way into the db?
18:38 sfan5          are the entries actually different?
18:39 ghoti          BuckarooBanzai: not worried for myself, I have 11.14 on one server, 12.7 on another. Personally, I think the minimum supported version of postgres should be no less than what is supported by the postgres project, which I believe is currently 10.
18:39 BuckarooBanzai sfan5: the entries are the same, not sure what causes is initially, but it was something with the network-thread afaik
18:39 sfan5          do we not use transactions?
18:39 BuckarooBanzai i don't think so
18:39 ghoti          sfan5: no, the question is WJU do we not use transactions?
18:39 ghoti          I have wondered this for a while.
18:40 ghoti          I think it is because the pg stuff was added to replace functionality of sqlite3, which does not have transactions.
18:40 BuckarooBanzai oh wait, i think there _is_ a transaction in the map saving process...
18:40 ghoti          s/JU/HY/
18:40 sfan5          sqlite3 has transactions
18:40 sfan5          or it would not be functioning database
18:41 sfan5          BuckarooBanzai: MT's db stuff pretty much always uses transaction so that'd be a big oversight if the pg auth database does not
18:41 ghoti          Hmm. I was pretty sure sqlite3 did not have 'BEGIN' and 'COMMIT', but it seems they do. My mistake.
18:43 BuckarooBanzai still, if you insert the same data twice (in the auth case) a transaction won't help you
18:43 ghoti          No, for that one needs error checking. Or more resilient SQL. :)
18:45 BuckarooBanzai old issue with an sqlite auth-conflict error: https://github.com/minetest/minetest/issues/9498 it's been some time since i've debugged this but i think there was a network-thread that kicked off the auth-insert, which races with the sync (globalstep) insert afaik
18:46 Pexin          !title
18:46 MinetestBot    No title found.
18:49 sfan5          "UNIQUE constraint failed in sqlite3-auth database · Issue #9498 · minetest/minetest"
18:49 ShadowBot      https://github.com/minetest/minetest/issues/9498 -- UNIQUE constraint failed in sqlite3-auth database
18:49 sfan5          should probably fix that some day
18:51 BuckarooBanzai no rush, there are a few workarounds and it looks like only a handful of server-operators have that problem :)
18:51 sfan5          the !title command I meant
18:52 BuckarooBanzai ah :P
18:57 ghoti          BuckarooBanzai: any idea what it is that distinguishes those of us to tickle that race condition? Is there a quick workaround that does *not* require compiling?
18:59 BuckarooBanzai not sure, sorry, if you find it out please share though
18:59 BuckarooBanzai the workaround does require re-compiling :)
19:04 sfan5          if I am reading the code right this issue is exclusive to servers that have a default password
19:04 MTDiscord      <Jonathon> ive seen it in servers that dont
19:05 MTDiscord      <Jonathon> and fairly sure https://github.com/minetest/minetest/issues/11877 server yourland doesnt have a default password
19:06 mrkubax10      Could someone look at minetest_game #2926?
19:06 ShadowBot      https://github.com/minetest/minetest/issues/2926 -- Flying dungeons
19:06 mrkubax10      ah it should be without space
19:11 sfan5          how does that default password stuff work anyway
19:11 sfan5          what do players have set as password when they register?
19:13 erlehmann      thanks to whoever implemented the “opaque leaves” setting. for some forests, it can be the difference between 25fps and unplayable.
19:32 sfan5          BuckarooBanzai: what would help is more logging to find out which createAuth call it was
19:32 sfan5          since there's no stacktrace it will be one of the two from C++
19:33 sfan5          both of those cannot be triggered but there's one that could potentially be triggered twice to cause that error
19:33 sfan5          but that'd be a client doing that on purpose, this can't happen accidentally
19:44 independent56  Finally, my server has attracted two regulars
19:44 independent56  yay!
19:45 independent56  There seem to be three types of player: Those who do not speak my langauge, those who are noobs, and those who just want to play on a server and socialize
19:46 independent56  And the rgulars are always the last
20:12 mmuller        independent56: out of curiosity, how long were you running it before you developed regulars?
20:12 independent56  I began at 21-06, and at 21-01 began the first regulars
20:12 independent56  Is this slow or average development?
20:12 independent56  *22-01
20:13 mmuller        oh I have no idea :-) yours is the first datapoint I've heard.
20:13 mmuller        but thank you for sharing.
20:14 independent56  So overall, that's around 6 months
20:21 MTDiscord      <Warr1024> That sounds like about my experience running a server.  Anecdotally I've heard that "a few months" is about how long to expect, and what evidence I've seen backs that up.
20:38 independent56  Nice
20:39 independent56  I can't wait until more players come until my server is on the level of others
20:39 independent56  Since i've built an average of one town a month, players keep on having more to come back to
20:41 independent56  Is doing frequent small talk with regulars the best way to keep retention?
20:42 definitelya    Depends if they like small- talk, I suppose?
20:46 independent56  Hmm
20:50 independent56  How do i shorten sapling growth time on servers with time which follows real-world time?
20:50 independent56  I'm getting reports from aplayer on my server
21:24 MTDiscord      <j45> The best way to get players is just to have something unique and fun about your server(s)
21:25 MTDiscord      <j45> Im just pointing out the obvious tho lmoa
21:25 MTDiscord      <j45> *lmao
21:46 Pexin          dailies + multiple abstracted layers of ingame currencies supported by nonrefundable microtransactions
21:47 Pexin          a proven winner
21:56 MTDiscord      <luatic> a disgusting winner
22:07 erlehmann      non-fungible nodes
22:07 erlehmann      (can't be placed or given to other players)
22:40 MinetestBot    02[git] 04sfan5 -> 03minetest/minetestmapper: Fix overflowing multiplication leading to apparent hang 13ba4d2c8 https://github.com/minetest/minetestmapper/commit/ba4d2c8c9e48dae84674c79cd9e445e0d07783f8 (152022-02-08T22:35:23Z)
22:40 MinetestBot    02[git] 04sfan5 -> 03minetest/minetestmapper: Inherit custom exceptions from std::exception 137da64ea https://github.com/minetest/minetestmapper/commit/7da64ea1af50eae5329769cb875b55145f8bcfe3 (152022-02-08T22:39:03Z)
22:45 MinetestBot    02[git] 04sfan5 -> 03minetest/minetestmapper: Fix overflowing multiplication leading to apparent hang 138e9805c https://github.com/minetest/minetestmapper/commit/8e9805c3fffd3ca83c98856894da41b7cfcdd2d3 (152022-02-08T22:43:20Z)
22:45 MinetestBot    02[git] 04sfan5 -> 03minetest/minetestmapper: Inherit custom exceptions from std::exception 132e35331 https://github.com/minetest/minetestmapper/commit/2e353312b5ce27d85a28dd4c8ff9e561cdc6d1ec (152022-02-08T22:43:20Z)