Minetest logo

IRC log for #minetest-dev, 2020-10-21

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

All times shown according to UTC.

Time Nick Message
01:27 fluxflux joined #minetest-dev
04:47 sofar hmm, so, get_mapgen_object("heatmap") is broken with 5.3.0, but works with 5.2.0, bizarre
04:47 * sofar gives up and goes to bed
06:06 NetherEran joined #minetest-dev
06:30 Flitzpiepe joined #minetest-dev
08:00 ShadowNinja joined #minetest-dev
08:14 proller joined #minetest-dev
09:02 NetherEran joined #minetest-dev
09:41 Flitzpiepe joined #minetest-dev
10:17 proller joined #minetest-dev
10:25 cheapie joined #minetest-dev
10:28 calcul0n joined #minetest-dev
10:30 lisac joined #minetest-dev
10:31 Kray joined #minetest-dev
10:39 Kray joined #minetest-dev
10:48 proller joined #minetest-dev
11:10 Wuzzy joined #minetest-dev
11:34 proller joined #minetest-dev
12:04 Fixer joined #minetest-dev
12:18 proller joined #minetest-dev
12:37 NetherEran joined #minetest-dev
12:59 Flitzpiepe joined #minetest-dev
13:17 pmp-p joined #minetest-dev
13:20 calcul0n_ joined #minetest-dev
13:27 cimbakahn joined #minetest-dev
13:28 cimbakahn Hello!  Is anyone around?
13:28 cimbakahn Can anyone tell from these 6 screenshots if this is a singleplayer world or a public server?  https://imgur.com/a/bygc3n6
13:31 sfan5 wrong channel, this is development only
13:31 sfan5 the answer to your question is that an RTT (ping) of 1ms very strongly implies that those screenshots were made in singleplayer
13:38 cimbakahn I went to minetest-hub and it looked like my message wasn't getting through.  It read: #minetest-hub :Cannot send to nick/channel
13:39 cimbakahn I don't know what that means?
13:39 jonadab Could also be a local/private server.
13:39 cimbakahn Well, which minetest irc am i supposed to go to?
13:41 cimbakahn sfan5, Thank you!  I have only played singleplayer, and public servers from the minetest server list.
13:42 sfan5 #minetest is the right channel for general stuff
13:42 cimbakahn Atleast it has the seed and the coordinates, so maybe i can recreate the world.
13:42 cimbakahn Which one?  There is more that one minetest irc.
13:43 cimbakahn I don't see one that just reads minetest.
13:43 sfan5 on this irc network (freenode), just do /join #minetest
13:43 cimbakahn sfan5, Ok.  Thank you!
14:04 zughy[m] Code here says "itemstack or itemstring or table or nil", while documentation says just "item", without specifying anything. Which one should I keep? https://github.com/minetest/minetest/blob/master/src/script/lua_api/l_object.cpp#L344
14:14 sfan5 item
14:16 Lejo joined #minetest-dev
14:19 Lejo So I now waited a few days for answeres on this: (((((Another thing about the management thing of Issues/PRs (https://forum.minetest.net/viewtopic.php?t=25507) I think we should work more with tags, like adding more often "Supported by core devs" or if it's a bad idea nobody would merge if someone would create a PR just close the Issue. Contributors should clearly know if a Issue is likly to get merged when
14:19 Lejo creating a PR for it. I mean we don't want to waste their time. These PRs for "approved" Issues should be priorised. The second thing is about how to review PRs. I often wonder why PRs are half reviewed like one small mistake was just fixed and than nothing happens. You could use tags here as well to clearly show where what (idea, codestyle, working) needs to be reviewed. I understand that it's hard to handle so
14:19 Lejo much stuff but that's exactly the reason why we should order the mess.))))
14:20 Lejo I honestly don't care what you think about my ideas, I just want to try to make clear that there is currently a problem.
14:21 sfan5 I think all of the devs know
14:21 Lejo It's your decison what to do about it, nobody will or can stop you. I just hope that you understand that a lot of people are currently not happy with the current setup.
14:21 Lejo So will you change anything?
14:25 Lejo I mean you don't have to exactly state how, when, what you might change. I just want to know if there is any will to fix the current situation?
14:27 sfan5 only two(?) devs have given their opinion so I can't tell if or how your suggestions can be implemented
14:28 sfan5 I'm sure there is will to fix the problems plaguing development, but if the solution boils down to "you need more time to deal with issues/review PRs" then it's not something that can "just" be done
14:35 Lejo Sure and It's also a great thing that you spend your time contributing for no money! You just need to keep in mind that others want too but can't because of your bussyness. Don't understand me wrong that is totaly ok. You don't have to do anything. There should just be someone helping you. So we need on one side more reviewers (don't have to be core devs) and a clear order how to review. I'm sure my quick ideas
14:35 Lejo aren't good or nearly perfect. I just haven't seen any will on your side for discussion brainstorming or anything to fix that. I'm sure you have seen the topcis on the forum talking about what to change. Were you there? I only saw rubenwardy and v-rob... In the end it's up to you and I would be happy if a change happens sooner than later.
14:39 sfan5 I have read the topics you are referring to
14:41 Fixer joined #minetest-dev
14:42 Lejo So what do you think about it? Got any ideas what to change?
14:45 sfan5 The problems raised by the topic(s) obviously exist and should be solved for Minetest to become a (more) successful project, but I don't have any great idea how to fix it
14:46 sfan5 more reviewers and/or more coredevs would be nice but I wonder where those people would come from
14:49 nerzhul_ it's not only PR, but real, develop-able roadmap. As we miss people with shader skills it's hard to increase the user facing part too
14:51 zughy[m] If I were (any) of you, I'd start from the roadmap, I'd choose a couple common points to follow and accept the idea you can't manage everything. Yes, you're absolutely right, PRs and issues rain down like hell and you're human, guys. Active core devs are, what? 5-6?
14:52 Lejo If we can't increase the amount of people which would surly be possible at least a bit we can only make reviewing/developing more efficient. Some of my Ideas on this are: Make codestyle automatic with tools like uncrustify, make a clear queue what to review, kind of approved Ideas in Issues (PRs doing that will be approved).
14:52 sfan5 more like 3-4
14:52 zughy[m] This is a fresh piece from KDE devs. I think it can help you understanding what to do: https://pointieststick.com/2020/10/21/inside-kde-leadership-and-long-term-planning/
14:52 MTDiscord <j​osiah_wi> I could easily review the coding style / formatting parts of PRs.
14:52 MTDiscord <o​neplustwo> anarchy?
14:53 MTDiscord <o​neplustwo> looks like somebody needs to be liberated from uncivilization
14:54 sfan5 Lejo: I think the last point of those suggestions would have the most impact and personally like the suggestion, but until we have discussed it with all/most devs it won't happen
14:54 sfan5 zughy[m]: by "accepting that you can't manage everything" do you mean contributions which don't match that current roadmap should be rejected? if not, how does that solve the issue?
14:57 Fixer_ joined #minetest-dev
14:57 Lejo Great! I can only advice you to take a look on nextclouds(https://github.com/nextcloud/server) way to handle that. They use tags and bot for that.
14:58 zughy[m] What people are asking in those topic is basically to know where this is going, and to be an active part of it. Aside the whole frustration, I think it's a fair question, and we can't answer that even if we wanted to
14:59 Fixer_ joined #minetest-dev
14:59 sfan5 we're not talking about the same things then
14:59 sfan5 I don't personally consider lack of a single roadmap a bad thing, but let's leave that aside
14:59 sfan5 the much bigger problem with development is that we don't have near enough time to deal with all issues and PRs
15:00 sfan5 ^ that's what I had in mind with my questions
15:02 Lejo why dont we make a queue + a guide how to review and let it the community do? I'm sure there are a lot of experienced contributors capable of that. Not every reviewer needs to be a core dev.
15:03 zughy[m] Right now it looks like one of those old families where the father breaks his own back because of working without asking help to anyone, almost like it was a weakness. But I think people would understand if the father said "I can't handle all of this", and they'd actually be supportive. Those were my 2 cents
15:04 Fixer joined #minetest-dev
15:06 sfan5 hm
15:06 sfan5 other people can already review PRs to help the author correct stuff before coredevs have a look
15:07 sfan5 if you make reviewers an official thing and count their approval I'd be wondering why those people don't just become real coredevs
15:07 rubenwardy I support automated code styles. I use clang format on my own projects, running it on all changed files before committing
15:08 sfan5 also regarding that I wonder if community members have the required expertise and view on e.g. api design, backwards compatibility and code structure to review "correctly"
15:08 rubenwardy I have made PRs to attempt to make it mandatory for Minetest, but there's a bunch of edge cases
15:08 rubenwardy I think that having 95% of the code correctly styled justifies the edge cases
15:08 rubenwardy Currently, we have something like 5% correctly styled but at least a few edge cases are ok
15:08 sfan5 (of course those things can be written down in a guide (as suggested), but the amount of knowledge that can be transferred is limited; some things just need someone with experience)
15:09 Lejo you might know the code style and can test if a PR works but don't have a overview over the whole project (like me) But wouldn't such people be able to help?
15:10 sfan5 help yes and that's already possible now
15:11 sfan5 but to really make a difference you'd want reviewers whose approval you can trust so that coredevs can just blindly merge a PR that has two reviewer approval but has never been reviewed by a coredev
15:11 sfan5 or was that not the direction your suggestion is supposed to go in?
15:13 Lejo Yep that's what I thought. Of course the core devs aprooval is still needed (maybe reduce to 1). I thing we should mobilize the community and tell them that there is work to do.
15:14 Lejo Maybe there should also be auto merge bots + blocker depending on approvals + checks
15:16 Lejo Another problem I currently see is that PRs are half reviewed or not approved and merged after review
15:16 Fixer joined #minetest-dev
15:17 sfan5 that's an issue with priotisation and I'm guilty of that too
15:17 sfan5 (probably because I randomly review a few recent/eye-catching PRs whenever I feel like it)
15:20 Lejo That's a thing which could be done using tags or maybe also using bots. There are multiple things which could be improved using them. Stuff like: Inform about rebasing/stale issues
15:21 rubenwardy I personally use this when prioritising reviews: https://github.com/rubenwardy/minetest/projects/1
15:22 MTDiscord <L​one_Wolf> If there was an official review queue I'd probably drop a few reviews in if I had the time
15:24 Lejo I agree such a queue should be more official
15:27 MTDiscord <L​one_Wolf> I could review without an official queue but I lean towards reviewing the easier PRs that I want the most, and those aren't always the ones the devs have their eyes on (Not making a poke there, there are a lot of good PRs)
15:33 Lejo1 joined #minetest-dev
15:48 NetherEran joined #minetest-dev
15:53 MTDiscord <J​onathon> i know im a bit late to the party, but > more reviewers and/or more coredevs would be nice but I wonder where those people would come from yet literally you had this a few days ago 🤷‍♂️  > Well okay, I'll be a core dev then > I'm still going to make my java Crafter game as well, but I'll be happy to modify the engine to not be so insane
15:54 MTDiscord <J​onathon> * yet literally had this a few days ago
15:57 rubenwardy oil_boi only has 2 commits in the repo, I don't think it's a good idea to grant privs before obtaining experience
15:58 MTDiscord <a​ppguru> ^
15:58 MTDiscord <a​ppguru> I also don't think he has much C++ / Irrlicht experience
15:58 nerzhul_ yep a coredev is a major contributor
16:00 MTDiscord <J​onathon> >  but I wonder where those people would come from so now its both ways, want people, but don't want people?
16:02 rubenwardy they usually come from active contributors
16:07 rubenwardy I'm hoping that if I ask numberZero enough times if they want to be a dev, they'll say yes
16:07 rubenwardy https://media.giphy.com/media/GYeLcrWi5DMqs/source.gif
16:16 pgimeno "A wise maintainer listens to the input of others and changes his or her mind when presented with reasonable alternative perspectives, but ultimately that person makes the call. If you don’t like the call, tough. It’s their project, not yours." - wish there was that for MT
16:19 zughy[m] how about choosing a couple macro-argoments from the roadmap and any new PR which is not related to that or it's not a bugfix, gets closed?
16:21 zughy[m] for instance, I've been trying to clean up code because I remember ruben proposed to get rid of rot code, so I tried to stick with it. Yes, there are other things I'd like to do but I'm not doing them because I know they're not #1 priority
16:21 ShadowBot zughy[m]: Error: That URL appears to have no HTML title within the first 4KB.
16:48 rubenwardy #10535
16:48 ShadowBot rubenwardy: Error: That URL appears to have no HTML title within the first 4KB.
16:48 pgimeno https://github.com/minetest/minetest/pull/10535 -- Add roadmap by rubenwardy
16:52 zughy[m] yeah pgimeno  but the "benevolent dictator" here is basically non existent. I mean, core devs are the ones choosing 99% of time
16:53 zughy[m] so we're more into the "elders" situation. Which is good, as it's not a small project
16:53 zughy[m] * yeah pgimeno  but the "benevolent dictator" here is basically non existent. I mean, core devs are the ones choosing 99% of the time
16:54 pgimeno hence why I said "wish there was that for MT"
16:55 MTDiscord <G​enshin> MT is a community project, there is no dictator
16:59 rubenwardy it's more like a constitutional monarchy with an oligraphy
16:59 MTDiscord <G​enshin> more like a democracy if you ask me
16:59 rubenwardy celeron55 delegates to us, but ultimately retains control
17:00 rubenwardy it's not a democracy as we are unelected
17:00 MTDiscord <G​enshin> but we elect what goes on the game engine
17:01 MTDiscord <G​enshin> fact is we can't please everybody, so we have to vote on what makes sense or what's possible to implement
17:07 MTDiscord <B​enrob0329> MT is not a democracy, nor has it ever been one. It was c55's per project, which was then handed to other developers who became interesting in continuing the work.
17:08 MTDiscord <G​enshin> then justify the elections on implementing certain PRs then
17:08 Fixer joined #minetest-dev
17:08 MTDiscord <G​enshin> explain why it isn't ademocracy
17:09 zughy[m] no Genshin, if there is someone who can decide for everyone (in this case celeron) is not a democracy
17:09 MTDiscord <B​enrob0329> MT being a game engine only came about with 0.4, and the hierarchy of management has yet to move from those who originally were making a game, but now are making an engine for games.
17:10 pgimeno when I said that, my point was that I believe it would be beneficial for the project if there was a unique person with a clear vision, deciding what gets in and what stays out
17:10 MTDiscord <G​enshin> Okay, then explain why c55 isn't participating modt of the time then? He's leaving it to others to solve it.
17:10 MTDiscord <B​enrob0329> I literally just said he passed the torch...
17:14 MTDiscord <G​enshin> If that's not a democracy, then i don't know what the bloody hell this system is. I just hope things get solved. carry on
17:14 MTDiscord <β​ασιλεύς Αλέξιος VI Αποκιος> I think I put it a year or so ago that C55 is more of the Gaurdian of the copyright to keep MT opensource
17:14 Sokomine joined #minetest-dev
17:14 MTDiscord <β​ασιλεύς Αλέξιος VI Αποκιος> /g/ I think it is more of an oligarchy
17:15 MTDiscord <G​enshin> THat never works in a community based project
17:18 celeron55 i wonder what it would take to get the automated code style issues solved
17:20 celeron55 personally i have a way to automatically format code in a high enough standard to suit me, but technically it's a bit hacky as it involves running uncrustify and afterwards a custom formatting python script that reworks indentation and a few other small things uncrustify doesn't do properly
17:29 NetherEran joined #minetest-dev
17:34 MTDiscord <j​osiah_wi> I might submit a PR for clang format this weekend.
17:35 sfan5 for a clang format configuration or ..?
17:35 sfan5 because we already have a clang format config and CI step that runs it
17:35 MTDiscord <j​osiah_wi> Clang format config.
17:36 MTDiscord <j​osiah_wi> I looked at it and it doesn't look like it agrees with the style guide.
17:36 MTDiscord <j​osiah_wi> Probably why it reports falso positives so much.
17:37 sfan5 improvements to it are definitely welcome
17:41 rubenwardy this is my current attempt https://github.com/rubenwardy/minetest/tree/clang-format
17:43 rubenwardy and here's another attempt to make it less aggressive https://github.com/rubenwardy/minetest/pull/new/clang-format-simple
17:44 homthack joined #minetest-dev
17:45 Krock til I can open PRs for other people's branches
17:45 Krock "Documentation has changed since you last contributed 10 days ago. Take a look before submitting a pull request:"
17:46 Krock thank you GitHub
17:46 rubenwardy lol
18:24 Lejo joined #minetest-dev
18:27 absurb joined #minetest-dev
18:29 MTDiscord <j​osiah_wi> Rubenwardy, you've done exactly what I was planning to do, but isn't the GPL 2.0 License a problem?
18:30 rubenwardy I don't think it counts, as a configuration file
18:30 rubenwardy could perhaps remove the kernel bits
18:30 MTDiscord <j​osiah_wi> I was told it wouldn't work to copy Linux' .clang-format directly because of the License, but you're probably more familiar with the rules.
18:59 rubenwardy perhaps best to avoid that then
19:00 rubenwardy this is the PR for that branch where I make clang-format required: #10029
19:00 ShadowBot rubenwardy: Error: That URL raised <Connection timed out.>
19:00 pgimeno https://github.com/minetest/minetest/pull/10029 -- Change clang-format configuration, remove all original files from whitelist by rubenwardy
19:00 rubenwardy I don't really have the time or energy to go through and fix all the edge cases
19:05 rubenwardy draft guidelines for git triagers (ie: Wuzzy / Calinou):  https://dev.minetest.net/Git_Guidelines#Issue_and_Pull-Request_Management
19:14 Calinou looks good to me :)
19:17 proller joined #minetest-dev
19:18 GreenXenith joined #minetest-dev
19:22 rubenwardy What is minetest-services on GitHub?
19:23 rubenwardy nerzhul_ ^
19:29 rubenwardy merging trivial PR #9478 in 10
19:29 pgimeno https://github.com/minetest/minetest/pull/9478 -- Fix Media... 0% on loading screen by MoNTE48
19:29 ShadowBot rubenwardy: Error: That URL appears to have no HTML title within the first 4KB.
19:31 rubenwardy and #10477
19:31 ShadowBot rubenwardy: Error: That URL appears to have no HTML title within the first 4KB.
19:31 pgimeno https://github.com/minetest/minetest/pull/10477 -- update jsoncpp to 1.9.4 by Zughy
19:32 ShadowNinja Hmmm, looks like ShadowBot needs fixing...
19:33 Krock hello ShadowNinja \o/
19:33 rubenwardy oh hi
19:34 ShadowNinja Hey Krock.
19:38 ShadowNinja I've got a mapblock v29 RFC patchset that I've been working on a bit that I should finish at some point...
19:38 rubenwardy what does that contain?
19:41 ShadowNinja I've got an issue on GitHub that lists a bunch of improvements to the format.  One new one that isn't on there is using zstd.  That sound give a significant performance boost, and maybe a good compression boost, especially if I add dictionaries, but it also adds a new dependency, so there's a tradeoff.
19:44 rubenwardy nice
19:44 rubenwardy merging #10494 in 10
19:44 ShadowBot rubenwardy: Error: That URL appears to have no HTML title within the first 4KB.
19:44 pgimeno https://github.com/minetest/minetest/pull/10494 -- Add documentation to builtin core.run_callbacks by Desour
19:52 ShadowBot joined #minetest-dev
19:54 nore slight question for the core devs: is there anyone besides me who thinks an implementation of #5155 should be merged? we have talked about it for a long time, but there has never been any decision...
19:54 ShadowBot nore: Error: That URL appears to have no HTML title within the first 4KB.
19:54 pgimeno https://github.com/minetest/minetest/issues/5155 -- Change light banks meaning
19:55 rubenwardy yes, I think it is worth doing that - it makes certain algorithms easier
19:55 rubenwardy but the compat will be painful
19:56 rubenwardy you can derive daytime and nighttime from sunlight and artificial, but you can't do the inverse. You need daytime and nighttime to have colored artificial light
19:56 rubenwardy well, I suppose you have have 3 channels for both day and night
19:56 sfan5 will it? are light banks guaranteed in lua api?
19:56 rubenwardy I don't think param1 is guaranteed
19:56 sfan5 other than that you "just" need a flag or mapblock version to do the recalculation
19:56 rubenwardy yeah, and network compat to recalc
19:57 rubenwardy actually, maybe not that difficult
19:57 sfan5 uh og yea
19:57 rubenwardy because you can derive daytime and nighttime from sunlight and artificial, the network compat shouldn't be too hard
19:57 sfan5 what do new clients do that receive the old values? calculate themselves?
19:57 rubenwardy well, it might be expensive
19:57 rubenwardy I suppose
19:57 rubenwardy good point
19:57 nore Yeah, the problem is old->new migration
19:58 nore The clients could use an approximation for rendering on new versions when the server is old
19:58 sfan5 sounds good
19:59 rubenwardy GreenXenith actually worked on this
19:59 nore (the approximation that is currently used when trying to compute sunlight from light banks)
20:00 nore Less sure about colored light, there are quite a lot of people in favor of it, but it increases mapblock size quite a bit (probably?)
20:00 MTDiscord <G​reenXenith> Not a whole lot
20:00 MTDiscord <G​reenXenith> I believe ANAND did extensive calculations on how much a u16 would increase map size
20:00 nore I think juhdanad had a working implementation of natural/artificial light
20:01 MTDiscord <G​reenXenith> As do I
20:02 ShadowNinja ShadowBot: issue_msg https://github.com/minetest/minetest/issues/10494
20:02 ShadowBot https://github.com/minetest/minetest/issues/10494 -- Add documentation to builtin core.run_callbacks by Desour
20:02 ShadowNinja I've disabled the messageparser for now so that you don't get duplicat messages from pgimeno.
20:03 nore (I think it's #5687)
20:03 pgimeno https://github.com/minetest/minetest/pull/5687 -- Artificial bank by juhdanad
20:05 nore ShadowNinja: I guess you can enable it again, I think pgimeno's messages were only a temp fix until ShadowBot was fixed
20:16 pgimeno indeed
20:16 pgimeno if it's fixed already I'll disable it
20:17 pgimeno done
20:18 pgimeno @ShadowNinja ^
20:20 ShadowNinja Thanks, I've re-enabled it on ShadowBot.
20:50 rubenwardy Krock: comments on #10152 ?
20:50 ShadowBot https://github.com/minetest/minetest/issues/10152 -- Semi-transparent background for nametags by Zughy
21:30 Seirdy joined #minetest-dev
22:51 proller joined #minetest-dev
23:05 Foz joined #minetest-dev
23:20 homthack joined #minetest-dev

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