Time  Nick        Message
00:27 rubenwardy  Hi! Does any one here have admin to the Minetest project on Weblate?
01:07 rubenwardy  I'd be very grateful to be given admin access to Minetest on weblate (or just my project if possible), so that I can update my app's translations
01:26 rubenwardy  just realised weblate uses a git repo, so I can just pull them in
15:57 Krock       merging game#1854 game#1857 and game#1858  in ~10 minutes
15:57 ShadowBot   https://github.com/minetest/minetest_game/issues/1854 -- Increase the maximum level of the diamond axe to 3 by elinor-s
15:57 ShadowBot   https://github.com/minetest/minetest_game/issues/1857 -- TNT: Only burn visually connected powder by SmallJoker
15:57 ShadowBot   https://github.com/minetest/minetest_game/issues/1858 -- default_clay_brick.png resize by TumeniNodes
16:46 sfan5       Krock: mind merging this trivial fix? https://github.com/minetest/minetest/issues/4810#issuecomment-304474260
16:47 Krock       sure. writing the commit right now to push
16:48 Krock       cleaning workspace before :3
16:51 sfan5       thanks
16:56 Krock       however, I wonder if it actually fixes the issue, as it's stored in a u32 variable again
16:57 sfan5       the problem is that max and min are s32
16:57 sfan5       u32 wouldn't be a problem
16:58 Krock       0x7fffffff - (-0x7fffffff - 1)
16:59 Krock       ah I see. the latter is the problem. Seems fine
17:04 Krock       oh. I forgot to set the author.. whatever, the source of this patch is known
17:09 Krock       Do we have a way to check whether the player position offset has been "corrected" or not?
17:12 Krock       merging #6179 in ~10 minutes
17:12 ShadowBot   https://github.com/minetest/minetest/issues/6179 -- Move the nametag back to the top of the player by TeTpaAka
18:00 ShadowNinja Thereis a meeting scheduled now.  Is there anything in particular that any devs would like to bring up for discussion?
18:01 sfan5       are we changing the versioning system for next release?
18:03 ShadowNinja I'm OK with that.  Would the -dev postfix also change in meaning from "development after this version" to "development towards this version" as specified by semver?
18:04 bigfoot547  That'd be nice
18:07 sfan5       well first the question is whether we are switching to semver or something custom?
18:07 sfan5       s/\?$//
18:46 Krock       we could continue our custom one, but trying to reach semver somewhen
18:46 Krock       starting with the -dev change
19:25 Wuzzy       I'm still confused about this “rebase needed” thing
19:25 Krock       git fetch upstream
19:25 Krock       git rebase
19:25 Wuzzy       why is it required so often
19:25 Krock       either you didn't rebase it correctly or the engine changes often in that region
19:25 Wuzzy       and why dont just the maintainers / "mergerers" do it?
19:26 Krock       especially lua_api.txt is changed often
19:26 Krock       because there will be conflicts to solve
19:26 Wuzzy       … and? those are impossible to solve by the maintainers?
19:27 Wuzzy       just curious. no shame
19:27 Wuzzy       i really havent gotten my head around all the development practises in MT, thats why
19:27 Krock       maintainers? I thought that would be the PR writers. The dev team tests and reviews PRs. If it's mergeable, it'll be added to the engine. Thing is, that most PRs aren't tested and can't be tested easily if a rebase is needed
19:29 Wuzzy       i mean the maintainers of MT of course. aka "core devs"
19:29 Wuzzy       by the way is this PR good now or did I just fail because I'm stupid?: https://github.com/minetest/minetest/pull/5795
19:29 Krock       (force) pushing to a PR branch is possible but looking at the amount of PRs it might not be worth rebasing a PR that never will be merged
19:30 Krock       hmm.. let's see
19:32 Wuzzy       what happens if I develop a non-trivial feature on a new branch branched from a freshly updated repo. development takes about 1-2 days, then I open PR. PR gets no or little attention for >30 days, then it gets 2 approvals. would it be likely that a rebase is needed at this point? and would the situation be different if the same PR  would have gotten the approvals instantly instead?
19:32 Krock       okay, one conflict in game.cpp
19:33 Krock       some PRs keep being mergeable for months, some don't. depending on the changed files and activity
19:34 Wuzzy       so the "rabase needed" is really beyond my control then
19:34 Krock       either it's controversial, nobody has an opinion on it or it's just forgotten
19:34 Wuzzy       so "rebase needed" actually just means "go resolve your own conflicts. shoo!"? xD
19:34 KaadmY      Wuzzy: rebase is only required if any changes in the PR's branch conflict with the upstream branch
19:34 Wuzzy       upstream branch?
19:34 Krock       rather "go resolve your conflicts caused by other people because nobody else is going to resolve them maybe"
19:34 Krock       but you got it
19:35 KaadmY      Yeah, in this case minetest/minetest/master and wuzzy2/minetest/testbranch or something
19:35 Wuzzy       Minetest has an uptream branch? since hen?
19:35 Wuzzy       when*
19:35 KaadmY      Well it's technically master
19:35 KaadmY      But I've heard the master repo+branch called upstream
19:35 Wuzzy       then why did you say upstream? :-(
19:35 Wuzzy       ok you mean master
19:36 KaadmY      Technically upstream fits better, but okay https://stackoverflow.com/questions/2739376/definition-of-downstream-and-upstream
19:36 KaadmY      Since master slightly implies it's not the main repo
19:36 Wuzzy       uhhh what?!
19:36 Wuzzy       master is just the name of a branch
19:37 Wuzzy       since when does "master" imply its not the main repo? O_O
19:37 Wuzzy       even slightly so
19:37 KaadmY      In the context of a forked software project, the 'upstream' is the source from which a particular project was forked. This might not be the absolute origin.
19:37 Krock       turns out only a variable name changed
19:37 Wuzzy       i was only talking about the brach, which is master
19:38 Krock       pushing the rebased branch now
19:38 Wuzzy       wait, are YOU going to resolve the conflict for me? thanks
19:39 Krock       you basically begged for it
19:39 Wuzzy       :D
19:40 Wuzzy       lol
19:40 Wuzzy       what the hell does the "Adoption needed" label mean?
19:40 Krock       the author left and the PR needs a takeover
19:41 Krock       means, the PR has potential
19:41 Wuzzy       thats funny
19:41 Wuzzy       I am still here
19:41 Wuzzy       yet this happened:
19:41 Wuzzy       https://github.com/minetest/minetest/pull/4703
19:42 * Krock     looks at paramat
19:42 Wuzzy       do you agree on the key name “Special”?
19:42 Wuzzy       aka "Use" (the E key)
19:42 * Krock     tries to guess what he meant by adding that label
19:42 Wuzzy       I always thought the name “Use” is very vERY misleading
19:42 Wuzzy       from the end of discussion it seems Special is the kay to go, but I just ask again jsut to be sure
19:43 Krock       special is no better than use IMO
19:43 Wuzzy       i disagree
19:43 Wuzzy       “Use” is misleading because it rarely has something to do with use
19:43 Krock       yes sure you do. but you asked for my opinion :P
19:43 Wuzzy       I think the rightclick is the REAL “use” key
19:44 Wuzzy       because you do almost everything with it: USE chests, USE doors, USE item frames, etc.
19:44 Wuzzy       I agree, the name “Special” is kinda meh but at least not _wrong_
19:45 Krock       Changing the name is fine for me. But the players must know what's best
19:45 Wuzzy       and definitely better than “Aux1” (What the actual fuck? Who makes this stuff up?)
19:49 Wuzzy       Can some1 plz look at this one?: https://github.com/minetest/minetest/pull/6170
19:55 Krock       !tell nerzhul PR 6170: wgettext can't be moved. The strings to translate are fetched with a script, searching for wgettext("[A-z]+")  (or similar)
19:55 ShadowBot   Krock: O.K.
19:55 Wuzzy       Krock: I think nerzhul knows that already ;)
19:56 Wuzzy       "valid anoying point :("
19:56 Krock       oh right. I just seen the comments below
19:58 Wuzzy       I just saw this in dev.minetest.net: “To prevent problems with the development like before Minetest Game is maintained by only six people (sfan5, nore, ShadowNinja, paramat, sofar, rubenwardy).”
19:59 Wuzzy       what were the problems with Minetest Game “before”?
19:59 Wuzzy       i can only recall that Minetest NeXt got merged and the developer team was thinned out but I don't really know the rationale
20:01 sfan5       it's probably supposed to prevent mtg not moving at all
20:01 sfan5       aka "too many cooks ruin the soup"
20:04 Krock       No background information from my side. I wasn't active in this channel back to that time
20:05 Wuzzy       funny. paramat often complains you "lack dev time"
20:08 sfan5       i think it's more about constant disagreements and the resulting "we can't ever add/change stuff"
20:17 Wuzzy       like we have now? ;-)
20:28 KaadmY      Could we have an item callback for a globalstep on specific inventory items?
20:29 KaadmY      Ie. on_held_timer = function(player, slot_number)
20:29 KaadmY      That updates every second for the itemstack that's current held on a player
20:29 kaeza       isn't get_wielded_item() enough for that?
20:29 KaadmY      kaeza: that still requires a globalstep that iterates players
20:30 kaeza       and?
20:30 KaadmY      Having a native or builtin method would be a lot easier
20:35 Krock       that's way too specific. you basically want player timers instead of node timers
20:35 KaadmY      Good point
20:36 KaadmY      Is there currently a better way of having a per-second update function for all connected players?
20:36 KaadmY      Instead of a globalstep
20:36 kaeza       no
20:36 Krock       minetest.after is no answer there, as it bases on globalstep as well. So: no.
20:36 KaadmY      Yeah
20:37 kaeza       KaadmY, iterating every player does not take much time anyway
20:37 KaadmY      Yeah
20:38 kaeza       worst case would be 100 players
20:38 KaadmY      It's more of a cleanliness thing than a performance thing
20:38 Krock       or 999 players if you use it on freeminer ;)
20:38 KaadmY      Eh 32 players is plenty 98% of the time
20:39 kaeza       realistically, I don't think any server had more than 40 players
20:39 kaeza       :P
20:39 KaadmY      I think I've seen 40-50, not more than 50 though
20:39 KaadmY      Are there logs of the maximum amount of players online?
20:39 Krock       realistically, I don't think a modded FM server could handle 999 players at all
20:39 KaadmY      Just a 999 loop in Lua is like 0.05 seconds :D
20:40 Krock       0.05 seconds times the functions you want to execute plus all block updates due placement/node removal
20:40 KaadmY      Yeah
20:41 KaadmY      Actually having a low max player count might emphasize less popular servers
20:41 KaadmY      Wonder how well that will work
20:45 kaeza       are ItemStacks passed by reference or copied in inv:get_stack(), etc?
20:47 Krock       hmm.. those passed by iterators for pairs(inv:get_list("main")) are copies
20:47 Krock       so I suppose these in get_stack are copies too
20:48 Krock       yes, the C++ code also does a few copies
20:51 kaeza       indeed they are :/
20:53 rubenwardy  kaeza, ItemStacks are a C++ object which hold a copy of a C++ ItemStack
20:53 rubenwardy  so yes, copy :P
20:53 rubenwardy  as compared to players, which hold a points to a player
20:54 kaeza       I'm also wondering why nobody thought of implementing __eq for them :P