Time  Nick      Message
11:43 sfan5     merging #11556, #11557 in 10m
11:43 ShadowBot https://github.com/minetest/minetest/issues/11556 -- Show message when changing block bounds by Wuzzy2
11:43 ShadowBot https://github.com/minetest/minetest/issues/11557 -- Allow lib/irrlichtmt to work for server builds (headers-only) by sfan5
11:55 sfan5     idk what other devs are up to but I just spent one or two hours reviewing PRs, if someone else did too I'm sure there'd be many we could merge
12:01 MTDiscord <Jonathon> Sfan5: have time for one more?
12:01 sfan5     which would that be
12:01 MTDiscord <Jonathon> https://github.com/minetest/minetest/pull/11130 Already has one approval
12:02 sfan5     hm dunno, lots of code there
12:03 MTDiscord <Jonathon> Thought i would at least ask.
14:41 sfan5     @josiah_wi mind opening that same cmake as a PR for minetest/minetest?
14:53 MTDiscord <wwar> @josiah_wi ^
15:03 MTDiscord <josiah_wi> Do you mean the version range?
15:05 sfan5     yes
15:09 MTDiscord <josiah_wi> I will PR it to Minetest then when I get back on my development PC.
17:02 Krock     > 0x00007ffff7d8d1e0 in irr::scene::CMeshCache::getMeshByName(irr::core::string<char, irr::core::irrAllocator<char> > const&) () from /data/Minetest/irrlicht/__output/lib/libIrrlichtMt.so.1.9
17:02 Krock     very cool thanks
17:36 Krock     writing up a commit message for the inventory list out-of-bounds fix
17:37 MTDiscord <josiah_wi> #11560 done
17:37 ShadowBot https://github.com/minetest/minetest/issues/11560 -- Set policies through version 3.9 to allow good IPO support by JosiahWI
17:38 Krock     however I wonder why every little check has a warning text.. it's sanity checks, and would normally not trigger
17:38 MTDiscord <Jordach> please explain what IPO does for the average user
17:39 sfan5     nothing because it's not enabled
17:39 sfan5     ;)
17:41 Krock     https://krock-works.uk.to/u/patches/0001-Inventory-Fix-rare-out-of-bounds-access.patch
17:41 Krock     will push that and #11548 in 10 minutes
17:41 ShadowBot https://github.com/minetest/minetest/issues/11548 -- HUD: Reject and warn on invalid stat types by SmallJoker
17:41 sfan5     I'd remove the /* ... */ comment
17:42 Krock     the entire thing? alright!
17:42 sfan5     yes
17:42 sfan5     the reason or circumstances this can normally happen is not relevant, the server has to check values anywaa
17:42 sfan5     s/a$/y/
17:43 Krock     yes, I'd even go further, removing the log messages of those checks. it is likely never triggered, and if - no-one cares
17:43 Krock     that however is better suited for a separate PR
17:44 sfan5     I don't disagree but consistency...
17:44 Krock     actually
17:44 Krock     there's another check on L339: if ((u16)to_i > list_to->getSize()) {
17:45 Krock     which is kinda flawed for "move anywhere" commands
17:46 sfan5     actually if the index is null based shouldn't that be >= ?
17:46 Krock     nvm
17:46 Krock     yes, that too
17:46 Krock     but negative numbers should not be allowed for to_i and from_i
17:48 Krock     to_i must be 0 or more to not fail in itemFits()
17:49 Krock     and to not cause another out-of-bounds error for 	ItemStack src_item = list_from->getItem(from_i);
17:49 sfan5     to_i = -1 is valid for move_somewhere = true I think
17:50 sfan5     so basically from_i should be checked before if (move_somewhere) and the existing to_i check should cover < 0 || >= size
17:51 Krock     Ctrl+F5. I updated the patch
17:52 Krock     moved below "move_somewhere" so that the index is only checked after
17:52 Krock     because "move_somewhere" does not depend on any of those negative values
17:53 Krock     does that look alright to you?
17:53 Krock     eh.. I could also open a PR at this point, to make sure there's no side-effects
17:53 sfan5     sure
17:55 Krock     anyway, will merge the HUD PR in a few minutes
17:57 MTDiscord <josiah_wi> IPO technically was supported before 3.9... only on intel processors IIRC.
17:58 MTDiscord <josiah_wi> Ah, it was the Intel compiler, on Linux.
18:03 Krock     *merging 1 PR*
18:04 Krock     what does IPO mean?
18:05 Desour    https://en.wikipedia.org/wiki/Interprocedural_optimization
18:05 Krock     thanks
18:05 sfan5     short for https://cmake.org/cmake/help/latest/prop_tgt/INTERPROCEDURAL_OPTIMIZATION.html
18:05 sfan5     compilers normally call this LTO (link-time optimization)
18:05 Krock     CMP0069. nice.
18:06 sfan5     in fact just about all compilers do IPO by default inside a compilation unit
18:06 MTDiscord <Jordach> can we please force contributors to explain acronyms
18:06 sfan5     what the CMake option refers to is across files, so LTO
18:13 MTDiscord <josiah_wi> I will take note of that and explain the acronym. Let me update the PR.