Time Nick Message 00:06 paramat ok 02:12 p_gimeno marginally off topic, but does anyone know if editing a github message to include a @user, sends an email to that user? 02:14 VanessaE nope, don't think so. 02:14 p_gimeno hmm, then numberZero isn't notified 02:15 VanessaE when I want to make sure someone gets an update, I just write a separate reply, send it, then delete it and move its contents into the post I wanted it in in the first place 02:15 VanessaE that way they get tagged, and still end up on the right page. 02:15 VanessaE and no clutter, 02:16 p_gimeno I'm not sure if 8358 is an issue he needs to deal with 02:17 VanessaE *shrug* 07:57 nerzhul merging #8359 07:57 ShadowBot https://github.com/minetest/minetest/issues/8359 -- HPChange Reason: Fix push after free, and type being overwritten by rubenwardy 07:58 nerzhul and #8354 07:58 ShadowBot https://github.com/minetest/minetest/issues/8354 -- Fix cast from const by accessing string data directly by rubenwardy 11:16 nerzhul sfan5: when you get time i updated #7159 (now it has gcc 8 + clang 7 support) 11:16 ShadowBot https://github.com/minetest/minetest/issues/7159 -- Upgrade from clang 5.0 to 7.0 by nerzhul 11:16 nerzhul and it's green everywhere heh 11:38 powerolive hi pls support android4.0.4 12:53 p_gimeno what derives from ClientActiveObject besides GenericCAO? 12:53 nerzhul hi: too old phone for our code level if i remember 12:55 p_gimeno and TestCAO 12:58 p_gimeno in other words, could ClientActiveObject and GenericCAO be merged? 12:59 nerzhul i think it's not the case because GenericCAO is just an "interface" 12:59 nerzhul why do you want to merge them ? 13:00 p_gimeno it's forcing me to add methods that aren't really needed 13:00 p_gimeno hm, maybe particles could be spawned from a GenericCAO instead of a ClientActiveObject? 13:03 nerzhul i don't think because it's an interface 13:04 nerzhul want do you want to code ? 13:10 p_gimeno attached particle spawners obtain a ClientActiveObject but they work with the fields that are only in the GenericCAO 13:11 p_gimeno so I have to move these fields from GenericCAO to ClientAciveObject 13:12 p_gimeno it would help if I can assume that when the ClientActiveObject is attached, it MUST be a GenericCAO, and then cast it 13:12 p_gimeno by fields I meant methods, sorry 13:19 p_gimeno perhaps a dynamic cast is in order at that point 13:37 nerzhul yes cast from interface to the right type if it's the correct type in getType() and do what you want 13:37 nerzhul it's how we perform in SAO 15:54 nerzhul pushing a trivial unused attribute cleanup reported by clang 7 15:54 nerzhul https://pastebin.com/WDTbMYGs 15:54 nerzhul https://travis-ci.org/minetest/minetest/jobs/505136223#L1181 15:57 sofar yeah, looks fine 15:57 sofar did you see the gcc warnings? 15:58 sofar only 2 so worth fixing imho 16:02 nerzhul i didn't read them currently i'm at work heh 16:02 nerzhul and i was working on upgrading clang toolchain to v6 16:02 nerzhul v7* 16:23 JDCodeIt I was playing with the pathfinder function and wondered if anyone wants to discuss issue #6771. 16:23 ShadowBot https://github.com/minetest/minetest/issues/6771 -- pathfinding: unnecessary error 16:27 JDCodeIt return value of false from function Pathfinder::updateAllCosts is interpreted as an error - but is that so in all cases? 16:53 JDCodeIt Same comment for the function Pathfinder::updateCostHeuristic - for the A* methods 16:54 JDCodeIt Thing I don't get is why the function bails out with false if one of the directions checked is a worse cost than the last one tried. Seems if we progressed to the 3rd or 4th direction we could find a lower cost. 16:55 JDCodeIt Returning false when the sequence of checked directions results in increasing cost, seems wrong 18:01 Unarelith you finally released it haha, nice 20:12 rubenwardy 5.0.1, when? 20:22 sfan5 anything left in the milestone? 20:23 rubenwardy https://github.com/minetest/minetest/milestone/14 20:24 VanessaE hold up 20:24 VanessaE don't forget that dumbass texture alignment problem 20:26 VanessaE (fixing it enough to get 5.0.1 out the door would be as trivial as it gets) 20:28 rubenwardy is it true that it only happens on colorwallmounted? 20:29 VanessaE that's what I first saw it with, but there are enough other ways to use nodeboxes that I don't know for sure. 20:30 rubenwardy well, if colorwallmounted was added after global textures, then maybe it wrongly applies to it 20:30 rubenwardy the setting confuses me though 20:30 VanessaE negative - just the opposite. 20:30 VanessaE colorwallmounted came first. 20:30 rubenwardy right, interesting 20:30 VanessaE this global thing is new in 5.0.0 20:42 paramat we have no idea if #8358 is trivial to fix, waiting on numberzero for an opinion. i'm not sure it's a blocker as it only affects a small minority of nodes 20:42 ShadowBot https://github.com/minetest/minetest/issues/8358 -- Nodebox facedir no longer rotates textures to follow it. 20:44 VanessaE "small minority" of nodes used in large quantities... 20:44 VanessaE if it broke the look of default:dirt_with_grass, that's just a single node. is that "a small quantity"? 20:44 paramat it may only be colorwallmounted nodeboxes that use overlay tiles, which is what these glowlights are 20:44 VanessaE yes. 20:44 VanessaE and people use the shit outta them. 20:45 VanessaE (can't say I expected that, either) 20:45 VanessaE (though they do look better than mese lamps :) ) 20:45 paramat 5.0.1 will still have many bugs, only critical bugs should delay it 20:46 rubenwardy I've emailed numberZero 20:46 rubenwardy I don't understand the point of the setting really 20:50 paramat dirt with grass is used much more than glowlights, one node from one mod 20:50 VanessaE paramat: my point stands. 20:52 paramat no, it's an irrational point to compare to dirt with grass 20:52 VanessaE you're the one who said "a small number"./ 20:52 VanessaE "small minority" rather. 20:54 rubenwardy is dirt_with_grass even rotatable? 20:54 paramat i suggest we release 5.0.1 at the weekend, if someone finds a fix for glowlights before then it can be fixed, but i don't agree to this delaying 5.0.1 if all else is ready 20:54 VanessaE rubenwardy: not by default :P 20:54 VanessaE paramat: NO NO NO 20:54 VanessaE G*D damn it, no. 20:54 rubenwardy what's the problem with glowlights? 20:55 VanessaE stop with this "we much rush it out the door" shit 20:55 VanessaE rubenwardy: see paramat's link, 8358 20:55 paramat yes 1 node from an optional mod is a small minority of nodes, and you have easy workarounds for this problem too 20:55 VanessaE paramat: first, three node. second, there IS no easy workaround. 20:56 VanessaE to my knowledge 20:56 VanessaE nodes* 20:57 paramat vanessa you're always trying to unreasonably delay releases with non-urgent bugs 20:57 VanessaE and you're always coming up with *excuses* that take more time to discuss than the fix would take 20:58 paramat how do you know the fix is trivial and quick? 20:59 VanessaE one line, change the default to disabled? 20:59 VanessaE how much more trivial does it need to be? 20:59 rubenwardy does that break global textures completely? 21:01 VanessaE rubenwardy: not sure on that, but as I understand it, the scaling is a separate feature from alignment.. 21:01 VanessaE so if alignment is forced back to node-oriented, would not the non-rotating nodes just end up looking the same as before? 21:02 paramat ah that may be a good idea 21:03 paramat it would disable all world aligned textures though, but if it's broken ... 21:03 VanessaE QED. 21:04 paramat 'world_aligned_mode = disable' have you tested that? 21:05 VanessaE I have not. 21:05 VanessaE because I don't understand all of what it can affect. 21:05 paramat sorry i didn't realise that was your suggested fix 21:05 VanessaE all I know is it's hardcoded to *enabled* when no setting exists 21:08 paramat please test that set to 'disabled' to see if it fixes your bug 21:09 paramat world-aligned textures is new in 5.0.0 so disabling them is not much of a loss until 5.1.0 when a proper fix will be ready 21:09 VanessaE which was what I was suggesting./ 21:10 VanessaE I'll check it now. 21:10 VanessaE nope. 21:10 VanessaE doesn't fix it. 21:12 red-045 So does setting the alignment on a block to "node" fix it? 21:13 paramat hmm 21:14 red-045 from reading the code yesterday it should default to old style alignment 21:14 VanessaE that makes fixing it not only a little less trivial than before, it also makes it more important, because it can't be worked-around without presumably altering mods. 21:15 red-045 so I'm thinking it could just be some small bug in how it does the alignment that doesn't match up quite with how it used to work 21:16 paramat since the setting doesn't fix the issue i don't agree with delaying 5.0.1, the mod will have to be changed to use a workaoround, after all the feature is 21:17 paramat ... very new 21:17 VanessaE no. 21:19 red-045 Well I'm just asking cause if changing the node def fixes it, it will be easier for someone other than the numberzero to fix it 21:19 red-045 if it does* 21:19 VanessaE 5.0.1 has already been delayed. 21:19 VanessaE it. can. wait. 21:20 VanessaE you guys were in such an all-fired hurry to get 5.0.0 out the door that this inventory bug (whatever it amounts to) snuck in, along with this texture bug. 21:22 rubenwardy 5.0.0 was in development for 2 years 21:22 p_gimeno no, there were just not enough testers, the inventory bug has been around for a long time 21:22 VanessaE yes...so? 21:22 rubenwardy with another year, we probably wouldn't have found these bugs 21:22 VanessaE who discovered the inventory bug? 21:23 paramat 5.0.0 wasn't hurried, the bug wasn't our fault 21:23 red-045 what's the inventory bug? 21:23 rubenwardy red-045: krock introduced a limit on the maximum size of a detached inventory 21:23 p_gimeno rubenwardy: #8325 21:23 ShadowBot https://github.com/minetest/minetest/issues/8325 -- Large inventory lists are not sent correctly 21:23 rubenwardy by switching a package from just appending data, to writing it with a u16 21:23 p_gimeno err @red-045 ^ 21:24 rubenwardy here's the best summary, red-045: https://github.com/minetest/minetest/pull/8331#issuecomment-470288892 21:24 paramat this bug wasn't a result of 5.0.0 'hurry' either, 5.0.0 was delayed and put back many times for more testing 21:25 VanessaE you DO realize why I asked who found it, right? 21:25 red-045 ugh that's a nasty bug 21:25 rubenwardy VanessaE: a user reported it - ShadMOrdre 21:26 red-045 so I guess you can't fix it for 5.0.0 21:26 rubenwardy red-045: it's already fixed in 5.0.1-dev 21:26 red-045 yeah but people will have to update 21:26 red-045 since it's a network thing 21:26 VanessaE ok then. I asked because I wanted to make sure that the finder isn't someone who should have ready access to dev builds. 21:27 rubenwardy red-045: my fix is backwards compatible 21:27 paramat anyway i expect numberzero will pop up before weekend so a fix may happen 21:27 rubenwardy old clients will still crash, but that's the same with any other bug 21:28 rubenwardy nerzhul: when will you push 5.x from the Android beta channel? 21:28 rubenwardy with 5.0.1? 21:29 VanessaE paramat: you're missing the point. I'm not worried about how long it takes to get it fixed. I'm concerned about making *a release* without it. 21:29 VanessaE s/it/a fix/ 21:31 VanessaE users are already hesitant enough to update. even moreso for a 0.4.x -> 5.0.x change... and now 5.0.1 is on its way, and you want me to tell people "5.0.2" (or whatever it would be)? 21:33 paramat vanessa, i know, but it's just a visual glitch, not critical. unless we're tough on this nothing will ever get released 21:34 jas_ That feature was merged a 21:34 VanessaE the inventory bug isn't critical either. someone here already said it can be worked-around, and so far no one on my servers has triggered it (and complained)............... 21:34 rubenwardy that's because you use UI, which probably paginates 21:35 red-045 is it a hard crash or just the client failing to read the inventory and disconnecting itself? 21:35 VanessaE rubenwardy: indeed I do. 21:35 rubenwardy red-045: only the client crashes, and it shows an error message and disconnects 21:35 red-045 ah 21:36 red-045 Crash made me thing of like a core dump or something like that 21:36 red-045 think* 21:37 paramat the inventory bug is critical, as proved by the panic over it, it prevents gameplay 21:38 red-045 I feel there is going to be a bunch of people asking what that error means in the near future 21:44 VanessaE red-045: yes, and I think more will be bugging me about the broken lights (and whatever else that I can't think of now). 21:45 VanessaE oh well, whatever. 21:53 nerzhul rubenwardy 21:53 rubenwardy nerzhul: 22:51 nerzhul oh sorry i missed to answer; yes i wait for 5.0.1