Time Nick Message 02:38 MTDiscord Is luau breaking compatability with the c api a big deal? Luau is security oreinted, maybe there is a jit version. 03:54 technomancy luau breaks TCO ಠ_ಠ 05:06 MTDiscord idk what TCO is 05:06 MTDiscord total cost of ownership is what google said 05:53 MTDiscord tail call optimization 05:54 MTDiscord in lua, return f(...) is a tail call, it just jumps straight into f; you can have an infinite loop using function calls and it won't stack overflow 10:03 pgimeno cheapie: i'm curious if you agree with the criticism in this article https://dmitry.gr/?r=06.%20Thoughts&proj=12.%20RV 10:03 pgimeno Nice article. It makes some very valid points, others not so valid. Didn't know RISC-V was such a mess. I don't think RVController needs to care though. 10:04 cheapie I don't really have time to read the whole thing before I leave for work in a few minutes, but it seems about the same as most of these 10:06 cheapie I've only dealt with a handful of ISAs so far (at least in terms of assembly programming), and at least for me RISC-V is about tied with 68k for being the least-bad to program for, but I'm also not a hardware designer so I can't really speak for that angle at all 10:08 cheapie x86 was indeed a nightmare to program for when I tried :P 10:16 pgimeno the main point about optionality which the author makes throughout the article is weak: you don't code something "for RISC-V" generically and try to autodetect features; you instead require a minimum set of extensions, like the requirement of RVA23 for some Linux distros, and base your code on them 10:18 pgimeno is RVController RVA23 compliant, by the way? I don't know what that entails and the extensions it requires 10:18 cheapie Not even close :P 10:48 cheapie pgimeno: FWIW I don't remember exactly what the requirements are for RVA23, but I do remember it requires 64-bit, floating-point, and vector features (which RVController doesn't have), among other things 10:48 cheapie I think it comes out to something like RV64GCVBH (plus some Z stuff) whereas RVController is RV32IMACB 10:49 pgimeno makes sense, also user/supervisor mode separation I guess 10:49 cheapie yes 10:49 cheapie Although RVController does have some parts of the privileged architecture already (just the M-mode parts so far), and I'd like to implement U-mode and S-mode at some point 10:50 cheapie found it https://docs.riscv.org/reference/rva23/v1.0/rva23-profiles.html 10:50 pgimeno oh great, thank you 10:53 cheapie Out of what it lists for RVA23U64, RVController supports: M, A, C, B, Zicsr, Zicntr, Ziccif, Ziccrse, Ziccamoa, Zicclsm, Za64rs, Zihintpause, Zkt, Zicond, Zimop, Zcmop, Zcb, Zawrs, Zicfilp 10:54 cheapie And for RVA23S64, it supports: Zifencei (and that's it) 10:55 pgimeno uh, single precision IEEE semantics is pretty hard to implement in Lua 10:55 cheapie For the "development options", Zabha, Zacas, and Zama16b 10:55 cheapie Yeah, which is why I have no floating-point at all :P 10:56 cheapie My end goal is to maybe eventually boot Linux on this thing, Linux does not need an FPU 12:53 MinetestBot 02[git] 04red-001 -> 03luanti-org/luanti: Clear pointer from userdata during GC (#17472) 137c2d1ef https://github.com/luanti-org/luanti/commit/7c2d1efd63f9bf6b845a697f7b7af9904b5f4561 (152026-08-20T12:52:12Z) 12:56 MinetestBot 02[git] 04sfan5 -> 03luanti-org/luanti: Fix use-after-free caused by core.clear_objects() 13954fb6c https://github.com/luanti-org/luanti/commit/954fb6cb08cef6c8dee85263bbdffeb3bbc1e20b (152026-08-20T12:55:10Z) 12:56 MinetestBot 02[git] 04sfan5 -> 03luanti-org/luanti: Fix use-after-free with metadata deletion 13a768328 https://github.com/luanti-org/luanti/commit/a76832838acf7869453d338c3efff93c0a877932 (152026-08-20T12:55:10Z) 12:56 MinetestBot 02[git] 04sfan5 -> 03luanti-org/luanti: Don't allow client-local CAO to be deleted 1302cc57f https://github.com/luanti-org/luanti/commit/02cc57f2542497a39e68ca198ea9f26c8a99141a (152026-08-20T12:55:10Z) 14:46 crazylad 5.17.0 release day! 14:47 MTDiscord yayayyayaya 16:40 MinetestBot 02[git] 04veselsem@users.noreply.hosted.weblate.org -> 03luanti-org/luanti: Translated using Weblate (Slovenian) 133ab5dc0 https://github.com/luanti-org/luanti/commit/3ab5dc0c2e566e2ec8dc60c98f6c98048778e9cd (152026-08-20T16:10:36Z) 16:40 MinetestBot 02[git] 04niohIguess -> 03luanti-org/luanti: Translated using Weblate (French) 13eac17bf https://github.com/luanti-org/luanti/commit/eac17bfc2bfb064503bbd925e25fe22ea3498fdb (152026-08-20T16:10:36Z) 16:40 MinetestBot 02[git] 04dainispolis123@gmail.com -> 03luanti-org/luanti: Translated using Weblate (Latvian) 133378c41 https://github.com/luanti-org/luanti/commit/3378c419eb222a2ac6610341fcd1eff7c04e3f7b (152026-08-20T16:10:36Z) 16:40 MinetestBot 02[git] 04BreadT -> 03luanti-org/luanti: Translated using Weblate (Japanese) 136f2ec0c https://github.com/luanti-org/luanti/commit/6f2ec0c6561d3b3fa88fcea09ad8b9f0869cae50 (152026-08-20T16:10:36Z) 16:40 MinetestBot 02[git] (10 newer commits not shown) 16:42 MinetestBot 02[git] 04AbbyRead -> 03luanti-org/luanti: Fix a typo in profiling.md (#17469) 137aacf72 https://github.com/luanti-org/luanti/commit/7aacf72303a15c4168aa7c67cd378aafdf1f0f75 (152026-08-20T16:40:47Z) 17:01 MinetestBot 02[git] 04SmallJoker -> 03luanti-org/luanti: Settings-related fixes 13db7d81e https://github.com/luanti-org/luanti/commit/db7d81edb25ebc0ae95ef48da9cdeb1bdaa03c5c (152026-08-20T16:58:33Z) 17:10 MinetestBot 02[git] 04sfan5 -> 03luanti-org/luanti: Bump version to 5.17.0 13c0e6812 https://github.com/luanti-org/luanti/commit/c0e6812b1a4260bb25a1f606f70f55f4962bb97d (152026-08-20T17:08:34Z) 17:12 crazylad FIVE POINT SEVENTEEN 17:19 * cheapie rushes to be the first one to join VE-C with 5.17 17:20 Krock cheapie: you can change the version string to be the first to join with 5.18.0-dev 17:20 * user333_ compiles it on his linux phone 17:20 crazylad lol 17:20 cheapie Krock: If I wanted to edit the version string I'd make it 6.21-dev or something 17:20 Krock or to 4.2.0 just to confuse the statistics 17:23 * cheapie updates VE-C itself to 5.17, for good measure 17:23 user333_ i need to update my servers too.. 17:23 user333_ compiling on my RPi1 (again) will be fun 17:27 user333_ https://user333.sweeux.org/tmp/DSC03991.JPG 17:27 crazylad dang that's so cool 17:30 user333_ it essentially just runs alpine linux with a *near mainline* kernel, so it can run a lot of stuff 17:31 user333_ much better than a locked down ecosystem by a large tech company 17:36 cheapie Krock: New idea, change the version string to claim to be Minecraft, /really/ confuse the stats :P 17:37 MinetestBot 02[git] 04sfan5 -> 03luanti-org/luanti: Continue with 5.18.0-dev 1331ec2b6 https://github.com/luanti-org/luanti/commit/31ec2b68dfa37d101ff29127783500fe35cb1a2c (152026-08-20T17:08:36Z) 17:38 user333_ i think i actually did set my version string to 1.21.4 or something like that at one point 17:39 user333_ https://user333.sweeux.org/tmp/DSC03992.JPG (zoom in to read it) 21:08 MinetestBot 02[git] 04red-001 -> 03luanti-org/luanti: Cache vcpkg builds (#17470) 138d68881 https://github.com/luanti-org/luanti/commit/8d68881c8974911c55201139d6d78f616119b220 (152026-08-20T21:06:15Z) 23:01 Sheriff_U3 Since when did minetest have a material field in core.register_node? 23:02 Sheriff_U3 Note the fact I said minetest and not Luanti. 23:23 [MatrxMT]_ material? I've not heard of that one 23:24 Sheriff_U3 Well I'm working a mod from 2009 that has a material field defined in register_node 23:25 Sheriff_U3 Looks like it was used in place of drawtype 23:30 cheapie Sheriff_U3: 2009? 0.4 didn't exist until late 2011 and the project as a whole isn't too much older than that 23:31 Sheriff_U3 opps I meant 2011... 23:31 Sheriff_U3 https://forum.luanti.org/viewtopic.php?t=568 23:34 cheapie I found where those digprop functions came from, at least: https://github.com/luanti-org/luanti/blob/0.4.dev-20111209-1/data/builtin.lua#L341 23:36 cheapie And where that data is then read (including what fields it has): https://github.com/luanti-org/luanti/blob/0.4.dev-20111209-1/src/scriptapi.cpp#L1112-L1124 23:37 Sheriff_U3 I just need to figure out how to convert the mod to current standards... 23:38 cheapie It looks like you'd probably want to drop the material field and replace it with the usual dig groups