Minetest logo

IRC log for #minetest-dev, 2022-06-05

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

All times shown according to UTC.

Time Nick Message
04:00 MTDiscord joined #minetest-dev
04:35 behalebabo joined #minetest-dev
05:36 calcul0n joined #minetest-dev
07:42 paradust who manages the homebrew build? I got an old Mac, confirmed the homebrew minetest is broken, but a fresh build at latest is fine.
09:03 sfan5 nobody
09:03 sfan5 as in, homebrew manages the builds themselves
09:03 Warr1024 joined #minetest-dev
09:04 sfan5 if you have the right hardware it'd be much appreciated if you could solve some of the macOS issues :)
09:34 Zughy[m] savilli: so it's just what the author suggested?
09:42 MTDiscord <savilli> yup
09:43 MTDiscord <savilli> #12406
09:43 ShadowBot https://github.com/minetest/minetest/issues/12406 -- Fix bsd iconv declaration by savilli
09:44 amicdict joined #minetest-dev
09:53 HuguesRoss joined #minetest-dev
09:54 Fixer joined #minetest-dev
10:29 proller joined #minetest-dev
12:55 HuguesRoss joined #minetest-dev
13:00 amicdict The assert message in `core.after` feels too vague. One of the variables could be invalid; but the assert message doesn't specify further
13:03 MTDiscord <luatic> assert(tonumber(after) and type(func) == "function", "Invalid minetest.after invocation")
13:03 MTDiscord <luatic> I'd say the message should be omitted entirely
13:03 MTDiscord <luatic> Unless we wanted to be particularly "beginner-friendly"
13:04 MTDiscord <luatic> Because an assertion points you towards the line where it was thrown
13:04 MTDiscord <luatic> And that line tells you which condition was not met
13:04 MTDiscord <luatic> Duplicating that in the error message would be convenient for modders but inconvenient for the engine codebase
13:08 sfan5 merging #12399, #12399, #12374 in 10m
13:08 ShadowBot https://github.com/minetest/minetest/issues/12399 -- Document itemstrings with metadata by appgurueu
13:08 ShadowBot https://github.com/minetest/minetest/issues/12399 -- Document itemstrings with metadata by appgurueu
13:08 ShadowBot https://github.com/minetest/minetest/issues/12374 -- Hide "Autosave Screen Size" on Android by rollerozxa
13:09 Zughy[m] merging 12399 twice, nice
13:09 amicdict I could try removing the message locally and see what it feels like
13:10 sfan5 s/12399/12406/
13:19 MTDiscord <luatic> sfan5: where merge
13:20 MTDiscord <luatic> there merge
13:20 sfan5 where what
13:20 MTDiscord <luatic> you're two minutes late ;)
13:36 amicdict The linenumber doesn't always give enough information though. What if a variable had changed during runtime to an invalid type? (Unless if it's possible to debug lua in runtime?)
14:05 amicdict I made a patch file for the change to the assert: https://ufile.io/bofrrh5l. Apply the patch with `cat <patch file> | git am`
14:22 erle amicdict, for future reference, this is a better pastebin https://mister-muffin.de/paste
14:39 Fixer_ joined #minetest-dev
14:47 amicdict erle, thanks!@ I didn't know about that site till now
14:48 amicdict Here's the patch on the alternate pastebin: https://mister-muffin.de/p/tWWW.diff
14:58 sfan5 is zero a valid content ID or do they start at one?
15:01 sfan5 and does anyone mind if I push this? http://sprunge.us/EwzCiX?diff
15:15 Krock id=0 is valid
15:15 Krock https://github.com/minetest/minetest/blob/master/src/nodedef.cpp#L1195-L1211
15:16 Krock it's even the first content ID that's registered after the builtin ones
15:16 Krock although you could reserve ID 0 by using m_next_id=1;
15:18 Krock either that or you reserve ID 0xFFFF which would otherwise be used for unknown nodes
15:24 behalebabo joined #minetest-dev
15:31 amicdict I wish irrlichtmt was included as a (git) submodule; it's the use case git was designed for.
15:32 Pexin I wish irrmt could get stripped the rest of the way and integrated into mt, but I guess nobody knows what happened to hecks
15:36 sfan5 Krock: i see, my fix makes sense then
15:36 sfan5 anything left on #12185 or can I merge it?
15:37 ShadowBot https://github.com/minetest/minetest/issues/12185 -- Add register dialog to separate login/register by rubenwardy
15:37 rubenwardy fine by me
15:37 sfan5 merging that + the diff I linked earlier in 5m then
15:42 Krock ๐Ÿ‘
15:53 Baytuch joined #minetest-dev
15:53 amicdict I tried out adding minetest_game.git as a submodule instead of cloning it. It actually works
15:54 sfan5 of course, there's no reason it wouldn't
15:56 amicdict Yep. I wonder why the READMe instructions still say to *clone* irrlichtmt then; git submodule works just as well.
15:59 sfan5 for me the primary motivation to avoid submodules is that you have to constantly push commits to update the revision the submodule "points" at
16:06 amicdict at what?
16:12 Pexin english is messy
16:16 amicdict oh
16:17 amicdict Couldn't you use `--recurse-submodules` (with `git push`)?
16:19 sfan5 huh?
16:21 Zughy[m] what should I do with #11791? Possible close as the guy hasn't provided any way to replicate it?
16:21 ShadowBot https://github.com/minetest/minetest/issues/11791 -- Shift-click in inventory not working
16:28 Zughy[m] also, Savilli, would you be so kind to check #10289 ? As it's related to FreeBSD (I guess you use it)
16:28 ShadowBot https://github.com/minetest/minetest/issues/10289 -- ld: error: undefined symbol: Json::Value::operator=(Json::Value&&) when compiling 5.3.0 on FreeBSD
16:32 rubenwardy Merging #12370 in 10
16:32 ShadowBot https://github.com/minetest/minetest/issues/12370 -- Android: Add support for sharing debug.txt by rubenwardy
16:32 rubenwardy and #12284 has been updated
16:32 ShadowBot https://github.com/minetest/minetest/issues/12284 -- [No Squash] Show dep errors in Select Mods modal by rubenwardy
16:32 rubenwardy I suppose [No Squash] should be [Some Squash]
16:33 MTDiscord <Jonathon> @savilli ^
16:35 MTDiscord <savilli> I'm sure I'm not the old person with FreeBSD server here, but okay, I'll check
16:35 rubenwardy nrz's avatar is literally the FreeBSD logo
16:35 definitelya joined #minetest-dev
16:35 MTDiscord <savilli> exactly
16:37 definitelya Uh, as of the latest commit Minetest crashes when I join worlds (I just rebuilt a clean profile, same thing).
16:39 rubenwardy CI integ test is failing too: https://github.com/minetest/minetest/runs/6745804645?check_suite_focus=true
16:39 definitelya rubenwardy: Thanks, I thought it was just my dumb system.
16:43 Krock oh my. getNodeNoCheck does not take nullptr values
16:43 sfan5 that is unexpected
16:44 Krock fix getNodeNoCheck instead
16:44 sfan5 but also I should push commits to a branch so CI can run first
16:45 sfan5 checking the call sites I see various code solving this by declaring a dummy boolean
16:45 sfan5 ...
16:45 sfan5 but so does getNodeNoEx so I guess that's inded
16:45 sfan5 intended*
16:48 definitelya BuckarooBanzai: I tried to repro the bug, but the same seed and position don't manifest it in new worlds. That's why I believe it to be an issue in the engine.
16:48 definitelya It always happens when schems cross a mapchunk, for example Y 47, and I think I maybe, MAYBE found a connection in #9627 (since it describes it happening with schematics too), or #1755.
16:48 ShadowBot https://github.com/minetest/minetest/issues/9627 -- Make Mapgen (mostly) work multithreaded again by sfan5
16:48 ShadowBot https://github.com/minetest/minetest/issues/1755 -- Mapgen V6 don't like altitude 47
16:49 sfan5 9627 just describes the bug
16:49 sfan5 rubenwardy: http://sprunge.us/kACrNM?diff
16:50 MTDiscord <savilli> commented #10289
16:50 ShadowBot https://github.com/minetest/minetest/issues/10289 -- ld: error: undefined symbol: Json::Value::operator=(Json::Value&&) when compiling 5.3.0 on FreeBSD
16:50 rubenwardy sfan5: sure
17:27 MTDiscord <ROllerozxa> damn, down to 999 issues now
17:30 Fixer joined #minetest-dev
17:36 rubenwardy It would be nice to skip the deprecation and start throwing errors for #12217 , there's been a few confused users recently and it's already an error log
17:36 ShadowBot https://github.com/minetest/minetest/issues/12217 -- Prevent loading a world with unresolved dependencies
17:59 rubenwardy I suppose with the select mods changes in my pr the user experience side of this becomes less pressing
18:00 Zughy[m] thank you savilli :)
18:02 Zughy[m] also, 997 issues, great job everyone ๐ŸŽ‰
18:02 rubenwardy ๐ŸŽ‰
18:09 MTDiscord <Niklp> ๐ŸŽ‰
18:21 Krock Zughy[m]: https://i.imgflip.com/3m29v9.jpg
18:27 Zughy[m] I can find them, but I can't review/merge PRs that fix them :P
18:39 Zughy[m] Also, for tonight's meeting, reminder that you could pick your favourite "Question" labelled issue and get rid of that too
18:39 Zughy[m] https://github.com/minetest/minetest/issues?q=is%3Aopen+is%3Aissue+label%3AQuestion
18:39 Zughy[m] The same works with "Discussion"
18:40 MTDiscord <luatic> wait the meeting is tonight
18:40 MTDiscord <luatic> how late
18:40 Zughy[m] my two cents, #11343 can be sorted quite fast
18:40 ShadowBot https://github.com/minetest/minetest/issues/11343 -- falling blocks don't stop falling at world border
18:41 MTDiscord <luatic> Zughy[m]: Well, how would you sort it out?
18:42 MTDiscord <luatic> Would you check whether the pos is just above map borders?
18:42 MTDiscord <luatic> TBH Minetest should use different nodes for world borders, but obv. that'll have to wait until 6.0
18:42 MTDiscord <luatic> currently ignore is used for both out-of-bounds and not-generated-yet
18:43 MTDiscord <luatic> (as well as not-loaded-yet)
19:13 Krock Meeting time, although a quick one from what I can tell https://dev.minetest.net/Meetings#2022-06-05
19:14 Krock ping HuguesRoss rubenwardy sfan5
19:14 HuguesRoss here
19:14 Krock also ping nrz
19:17 sfan5 here
19:18 Krock > make a 5.5.2 release due to security fixes?
19:19 Krock how many commits does this include? I've only seen one mentioned in the other channel
19:20 Krock ah. another one about metatables
19:20 HuguesRoss Looking back through the commit history I see one or two things that might be security-related, doesn't seem like a ton of fixes
19:21 HuguesRoss imo it probably depends on how close we think 5.6 actually is--are we likely to hit the july deadline?
19:22 Krock hmm. in a month from mow. 8 PRs and 2 issues
19:22 Pexin july doesn't mean the end of july?
19:22 HuguesRoss The date on github is July 3
19:22 HuguesRoss https://github.com/minetest/minetest/milestone/19
19:23 Krock dates are pretty much irrelevant anyway "it's done when it's done", yet July 3 as a goal
19:24 HuguesRoss Yeah, I know our dates aren't set in stone or anything. My thought is more: If we expect to finish in 4-6 weeks, them a patch release may not be worth it. But if we're 2 months or more out it's probably a good idea
19:26 Krock the commits do not seem to be that severe. annoying to have, still.
19:27 HuguesRoss It would probably be an easier decision if we had more folks to discuss. was this your question, or did another coredev pitch it?
19:28 rubenwardy One of them is a sandboxes escape in single player
19:28 Krock added by sfan5. we could discuss this later
19:28 HuguesRoss I think that's a good idea
19:28 sfan5 sorry ,sec
19:29 sfan5 https://github.com/minetest/minetest/security/advisories/GHSA-663q-pcjw-27cc
19:29 sfan5 can you see this
19:29 Krock 404
19:29 HuguesRoss 404 over here as well
19:29 sfan5 try now
19:29 Krock works
19:30 sfan5 this would be the reason
19:30 HuguesRoss Ah yeah, that seems bad
19:31 Krock huh? weren't there already checks for secure.* ??!
19:31 sfan5 that one isn't covered by them
19:31 HuguesRoss Do we know if this is in the wild, or just something we figured out?
19:31 Krock nvm apparently I cannot read
19:31 sfan5 or rather the setting existed even before sandboxing was added
19:32 sfan5 HuguesRoss: just discovered in the source, I think it's unlikely that someone would write Minetest malware but we can't take chances obviously
19:34 HuguesRoss Definitely needs to be fixed either way, I just figure that's a potential factor for deciding how soon we need the fix to be live
19:34 Krock nobody noticed it for years, and now it's suddenly a problem?
19:35 HuguesRoss Well, it was always a problem. Just not one we were aware of ;)
19:35 Krock if it's not known to anyone this isn't too big of a deal, also thanks to ContentDB which partially ensures code sanity
19:35 sfan5 we could pull in #12396 for 5.5.2 for good measure since that caused some ruckus
19:35 ShadowBot https://github.com/minetest/minetest/issues/12396 -- Sanitize player position and speed server-side by sfan5
19:35 sfan5 also translations which weren't updated in 5.5.1
19:36 sfan5 it's a bit of work to do a point release but justified here IMO
19:36 sfan5 ping rubenwardy, opinions?
19:39 Krock the only reason why Lua still enforces the F1000 bounds is due to object storage which cannot be changed
19:39 Krock so adding clamp functions does seem to be a reasonable soludion
19:40 Krock anything outside of those bounds also seems to be a mod error
19:42 Krock testing the sanitize PR so that you have another good reason for 5.5.2 ;)
19:55 sfan5 "are we likely to hit the july deadline?" <- good question btw, I'm not so sure
19:58 sfan5 but at this point we should be reasonably sure that the stuff in the 5.6 milestone is everything we want in the release
19:58 sfan5 maybe we can discuss that too
19:59 Pexin I was trying not to ask, but how significant does something need to be to be on the 5.6 roadmap? #11939
19:59 ShadowBot https://github.com/minetest/minetest/issues/11939 -- Restore and enhance bouncy behavior by pecksin
20:00 Krock I'd add that. the current bouncy code is shit
20:01 Krock question is whether there are objections for adding it to the milestone
20:03 sfan5 I have not been motivated to look into that PR so you could say I'm neutral
20:08 Krock meanwhile I'll just go ahead closing #12355. no support for it, and possible close label too.
20:08 ShadowBot https://github.com/minetest/minetest/issues/12355 -- Raise default windowed resolution to 1280x720 (720p) by rollerozxa
20:09 HuguesRoss go for it
20:10 HuguesRoss By the way, opinions on #12405? Could be helpful
20:10 ShadowBot https://github.com/minetest/minetest/issues/12405 -- Bug report template: ask contributors to provide a code snippet by Zughy
20:10 sfan5 sounds useful
20:11 Krock trivial change; OK to me
20:11 HuguesRoss I'll approve and merge then, I like it as well
20:11 Krock go ahead.
20:12 HuguesRoss Done
20:26 sfan5 okay so
20:27 sfan5 https://github.com/minetest/minetest/pulls?q=is%3Aopen+is%3Apr+label%3A%22Roadmap%3A+Needs+approval%22 anything to discuss here?
20:28 Krock I don't even know where to begin with...
20:30 Krock #10587 wasn't this jump issue fixed in a recent PR?
20:30 ShadowBot https://github.com/minetest/minetest/issues/10587 -- Only count entity as "on ground" if they are colliding on the Y axis. by nathanfranke
20:31 sfan5 dont think so
20:33 sfan5 in principle the change makes sense but I think we may want to reject it because it changes movement too much
20:33 Pexin one comment seems to suggest that this behavior should be configurable by the individual game?
20:34 Krock more physics inconsistencies among games? I'd rather not want that.
20:36 Krock #11705
20:36 ShadowBot https://github.com/minetest/minetest/issues/11705 -- Limit stepheight smoothing to the stepheight and stop smoothing during jumps by TurkeyMcMac
20:36 Krock I thought this one fixed most of the concerns
20:37 Krock looking at it, that's wrong. it does not change the jump behaviour
20:39 Pexin fwiw, my non-core view is that voxels are an abstraction, and it makes sense to climb a hill smoothly
20:41 sfan5 as for #11855 or #12353 I think it should be done but it needs to be somehow optional
20:41 ShadowBot https://github.com/minetest/minetest/issues/11855 -- Correct gravity calculation by EliasFleckenstein03
20:41 ShadowBot https://github.com/minetest/minetest/issues/12353 -- Fix acceleration by appgurueu
20:41 sfan5 or we'd have to wait until 6.0
20:42 x2048 joined #minetest-dev
20:43 Krock I doubt there's a convenient and sane way to make this compatible
20:43 Krock 6.0 seems to be the best solution IMO
20:43 sfan5 I'll give concept approval to #12388, looks sane to me
20:43 ShadowBot https://github.com/minetest/minetest/issues/12388 -- Extend bone override capabilities by appgurueu
20:44 sfan5 last thing: #12168 is also in the meeting list
20:44 ShadowBot https://github.com/minetest/minetest/issues/12168 -- Make `minetest.write_json` use "[]" instead of "null" for empty tables by Desour
20:46 sfan5 since neither is more correct I don't see how this PR improves the state of affairs
20:48 x2048 It improves that roundtrip on an empty table returns an empty table.
20:49 x2048 If you only use lua, is does not matter which way it is in json
20:49 sfan5 but then there's no reason to involve json
20:52 x2048 I'd suggest at most a parameter then to write_json to write 'null', '[]' or '{}' for empty tables, if that is important to the caller.
20:53 x2048 * to add a parameter to write_json
20:58 sfan5 sounds okay
21:00 x2048 joined #minetest-dev
21:01 x2048 I'll add ย comment to the issue then, unless there are objections from other coredevs
21:11 x2048 done
21:24 x2048 joined #minetest-dev
21:35 appguru joined #minetest-dev
21:41 appguru joined #minetest-dev
21:48 kaeza joined #minetest-dev
21:56 Pexin is #11973 still a thing?
21:56 ShadowBot https://github.com/minetest/minetest/issues/11973 -- WIP: Performance improvements for entity collision by JosiahWI
22:35 panwolfram joined #minetest-dev

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