Minetest logo

IRC log for #minetest-dev, 2017-01-29

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

All times shown according to UTC.

Time Nick Message
00:10 STHGOM joined #minetest-dev
00:15 ShadowNinja BTW, TestPcgRandomRange seems to be throwing a floating point exception for me.  This is in a VM though.
00:20 ShadowNinja sfan5: You around?  I changed the thread sync PR a bit and found another issue that it solves if you'd like to look at it again.
00:22 sofar ShadowNinja: there's an issue open for that
00:22 sofar several people have that issue
00:22 sofar inclding me on my work distro (clearlinux)
00:23 sofar oh BTW
00:23 sofar my work is now testing minetest every day several times in an automated test, lol
00:23 sofar just because we ship it
00:23 sofar it's not testing git, though
00:27 ShadowBot joined #minetest-dev
00:28 ShadowNinja sofar: Thanks.  Do you mind saying what you're shipping Minetest in?
00:28 sofar my work distro (clearlinux)
00:28 sofar https://clearlinux.org/
00:29 ShadowNinja Oh, right, the Intel high-performance distro.
00:29 sofar that's the one :)
00:35 ShadowNinja Merging #5134 soonish unless there are objections...
00:35 ShadowBot https://github.com/minetest/minetest/issues/5134 -- Fix synchronization issue at thread start by ShadowNinja
00:47 octacian joined #minetest-dev
01:34 est31 joined #minetest-dev
01:34 est31 hi there, just want to relay this job offer http://kidscode.com/recruitment
01:34 est31 seems they also are searching for modders?
01:35 Tmanyo joined #minetest-dev
01:35 est31 dunno, don't speak french
01:35 est31 good bye
01:35 est31 left #minetest-dev
01:36 proller__ joined #minetest-dev
01:37 betterthanyou711 french?
02:09 sofar betterthanyou711: it's a language spoken in France
02:09 betterthanyou711 heh had no idea
02:10 betterthanyou711 i meant where french was added into the conversation
02:10 betterthanyou711 seemed without context
02:10 sofar the link he posted is in french
02:10 betterthanyou711 yeah i figured it out after a while
02:11 Miner_48er joined #minetest-dev
02:11 betterthanyou711 i have spoken french before
02:17 xerox123 joined #minetest-dev
02:36 VanessaE I think I found a bug in either the engine or _game, not sure which...  place a wallmounted node, with paramtype2 = "colorwallmounted", and it doesn't rotate to the correct orientation.
02:36 VanessaE (particularly, with drawtype = "signlike")
02:43 VanessaE this bit of code is sufficient to demonstrate it, http://pastebin.ubuntu.com/23884531/
02:55 VanessaE this bit of code serves ok as a workaround in my mod, http://pastebin.ubuntu.com/23884573/
02:57 proller__ joined #minetest-dev
03:12 sofar placement code probably checks the paramtyp2 first
03:13 sofar and doesn't know about it
03:13 sofar so, mtg bug essentially
03:21 Foz anyone care to look at https://github.com/minetest/minetest/issues/5135
03:22 Foz I think the error messages are missing the actual problem
04:35 Void7 joined #minetest-dev
04:38 proller joined #minetest-dev
05:14 paramat joined #minetest-dev
05:30 Zeno` sfan5: you looked at Darklands right? I just logged on and it took 38 seconds to teleport me from 0,0,0 to my position
06:20 Zeno` sofar, do you know much about these detached inventories?
06:21 sofar nope
06:26 Zeno` hmm. Thanks. I'll wait for sfan5
06:30 octacian joined #minetest-dev
06:31 Zeno` have a nice sleep octa?
06:34 paramat my best guess at a fix #5138
06:34 ShadowBot https://github.com/minetest/minetest/issues/5138 -- Visual scale: Send sqrt(visual_scale) to pre-30 clients by paramat
06:35 sofar lol, hah, that would work
06:36 paramat high chance i did something silly
06:37 Zeno` you could use float sqrtf(float x);
06:37 Zeno` do avoid implicit double casts
06:38 Zeno` nvm
06:38 Zeno` it doesn't really matter
06:39 paramat ah sqrtf is better for floats? generally
06:39 Zeno` generally yeah
06:39 paramat thanks i didn't know
06:43 lumidify joined #minetest-dev
06:43 paramat although it seems we use C++98 and that can apparently use sqrt for floats
06:55 Zeno` ok *shrug*
06:56 Zeno` you can use sqrt for floats in C as well
06:56 Zeno` it just implicitly casts to double
06:56 Zeno` but *shrug* it doesn't really matter to me personally
06:57 Zeno` \o/ #5136
06:57 ShadowBot https://github.com/minetest/minetest/issues/5136 -- Add console height setting by Ezhh
07:29 Hunterz joined #minetest-dev
07:37 proller joined #minetest-dev
08:32 nrzkt joined #minetest-dev
08:34 Krock joined #minetest-dev
08:34 Krock joined #minetest-dev
08:35 lumidify joined #minetest-dev
08:41 Thomas-S joined #minetest-dev
08:56 Zeno` waiting for travis and then will merge #5139
08:57 ShadowBot https://github.com/minetest/minetest/issues/5139 -- Rename height to scale for openConsole() by Zeno-
08:57 nrzkt on gitlab there is a nice option: "merge when build succeed" i used it everytime :D
08:59 Krock nrzkt, maybe not the best idea when doing compability stuff for compilers that aren't tested
08:59 nrzkt you should know what you are doing :)
08:59 Zeno` oh, that'd be handy to have yeah!
08:59 Krock yeah
09:00 Zeno` especially for stuff as trivial as this
09:01 Krock Has anyone got time to review #4370 or at least give a feedback?
09:01 ShadowBot https://github.com/minetest/minetest/issues/4370 -- Builtin item: Tidy up the code by SmallJoker
09:01 Krock that pull is not dead. it's just resting.
09:02 Zeno` lol
09:02 Zeno` stop thing in python all the time :P
09:02 Krock :P
09:23 YuGiOhJCJ joined #minetest-dev
09:23 Zeno` Krock, seems fine to me
09:25 Zeno` I'm not confident enough with Lua to :+1: though :`(
09:26 Zeno` 5139 merged
09:43 fwhcat joined #minetest-dev
10:07 juhdanad joined #minetest-dev
10:36 juhdanad Why is the burning furnance not backface culled?
10:38 xerox123 joined #minetest-dev
10:47 red-001 updated #5133
10:47 red-001 nrzkt, ShadowNinja ^
10:47 ShadowBot https://github.com/minetest/minetest/issues/5133 -- [CSM] Block access to the `io` library by red-001
10:53 blaze joined #minetest-dev
10:57 lisac joined #minetest-dev
11:06 * red-001 is working on adding on_dignode
11:11 Fixer joined #minetest-dev
11:12 Player_2 joined #minetest-dev
11:38 proller joined #minetest-dev
11:45 juhdanad joined #minetest-dev
11:59 red-001 could someone review #5113 ?
11:59 ShadowBot https://github.com/minetest/minetest/issues/5113 -- [CSM] Add function to get the name of the player and functions to colour chat. by red-001
12:15 sapier joined #minetest-dev
12:38 proller joined #minetest-dev
12:55 red-001 #5140
12:55 ShadowBot https://github.com/minetest/minetest/issues/5140 -- [CSM] Add `on_dignode` callback by red-001
13:24 proller joined #minetest-dev
13:32 juhdanad joined #minetest-dev
13:42 Guest93432 joined #minetest-dev
14:15 Guest93432 joined #minetest-dev
14:31 lumidify joined #minetest-dev
14:36 numZero joined #minetest-dev
14:44 numZero_ joined #minetest-dev
14:46 xerox123 joined #minetest-dev
14:47 juhdanad joined #minetest-dev
15:11 xerox123 joined #minetest-dev
15:33 octacian joined #minetest-dev
15:33 octacian joined #minetest-dev
15:43 DFeniks joined #minetest-dev
15:52 VanessaE juhdanad, nore: either of you awake?
15:52 nore hm, I am
15:52 nore why?
15:52 VanessaE I have a feature request :)
15:52 nore so what is it? :)
15:53 VanessaE regarding hardware node coloring, could something be added to allow for a non-colorized overlay imageon top of the colored texture?
15:53 VanessaE (I can fake it with a meshnode, but that's not ideal)
15:54 Zeno` VanessaE, you want us to do complicated things?
15:54 sapier :-) I wonder how many levels of colored and uncolored overlays might be usefull :-)
15:54 VanessaE think colorized-via-param2 bricks, with always-grey mortar overlaid onto it
15:54 VanessaE sapier: I'd be happy with just one (per tiles={} texture/material)
15:54 Zeno` will merge #5136 shortly
15:54 ShadowBot https://github.com/minetest/minetest/issues/5136 -- Add console height setting by Ezhh
15:55 sapier Yeah what I was trying to say was I believe this could be difficult to implement in a way which ain't a extremely special handling for your case
15:55 nore VanessaE: I remember asking about it, and juhdanad said it wasn't easy
15:55 VanessaE hm, ok
15:56 sapier but
15:56 VanessaE I use the meshnode hack in unifiedbricks and blox mods
15:56 nore I had that in mind for the sides of the grass as well
15:56 sapier are texture overlays affected of hardware coloring?
15:56 VanessaE (over the past several days I've converted a bunch of mods to use param2)
15:56 nore sapier: yes, it's the whole resulting texture that is coloured
15:57 sapier hmm thinking about it a second time ... I believe yes as "hardware" most likely is done at last step
15:57 nore exactly :)
15:57 sapier so the only way making overlays not beeing affected would be subtracting the later applied hw overlay from them ... guess that'd result in exactly different speedup then wanted
15:58 VanessaE why not a second material?
15:58 VanessaE I mean, if I can do it with a meshnode, surely the engine could do it more efficiently.
15:58 sapier do we support multiple materials per node by now?
15:58 nore well, juhdanad said it would be possible with two polygons and GL_EQUAL thing
15:58 VanessaE yes
15:58 * nore looks for that comment
15:59 VanessaE what I did was make a cube slightly smaller than 1m, and another exactly 1m, centered on the first.  the larger one holds the overlay image, the smaller is the base texture
15:59 sapier that's a workaround not two materials ;-)
15:59 nore https://github.com/minetest/minetest/pull/4986#issuecomment-271855741
15:59 VanessaE it means more polys per node, and there is a z-fighting issue at large distances (depending on view angle), but that's irrlicht's fault
15:59 VanessaE sapier: actually it's two materials.
16:00 sapier no it's just two nodes with one material ;-)
16:00 VanessaE not with a mesh node.  each cube is textured by one of the two tiles={} entries I'm using - each tiles={} entry is a material when you're using meshes
16:01 sapier yes but there's only one tiles entry per mesh so one material
16:01 VanessaE no
16:01 VanessaE each texture in tiles={} becomes a material in a mesh node.
16:01 VanessaE so one node can have up to 6 materials entries that way
16:01 VanessaE it's a handy feature RBA came up with
16:01 sapier ok true but still only one applied per polygon
16:01 VanessaE yes
16:02 VanessaE that's why I bury one poly slightly behind the other :)
16:02 nore VanessaE: but you get z-fighting at larger distances, no?
16:03 VanessaE yes, like if I'm say 50-60 nodes away, but by that time, the textures start to get noisy anyway, so it's not noticable.
16:03 VanessaE sapier: I figured the engine could do the same thing, but without the actual offset (just hard-code the Z-order layering or however it works)
16:03 sapier well larger distances cause numeric issues thus that's reasonable and "by design"
16:03 nore well, see the comment I linked
16:04 sapier yes but you'd have to specify the texture order for eache polygon (in corner case)
16:04 nore it is exactly telling that it can be done, using two polys with same vertices and something about GL_LEQUAL in the depth buffer
16:04 nore (I don't really know what this means however)
16:05 sapier that's no wichcraft of course but it'd make texturing way more difficult
16:05 sapier hmm I think it's spelled witch
16:05 VanessaE nore: oh yes, the comment.  that's what I was talking about
16:07 Fixer there is also z-fighting in liquid situations
16:07 VanessaE that's a different issue
16:08 VanessaE btw, the z-fighting is just an accuracy issue.  if I increase the distance between the two layers, it increases the distance from the node before I start to see z-fighting again (but it comes with a cost, visible gaps between the nodes, at close range.)
16:09 sapier it's not an accuracy issue it's a numeric issue
16:09 VanessaE oh.
16:09 VanessaE well not a big deal, like I said the workaround works well enough :)
16:10 sapier it's what beginners do wrong when using float values for calculations ;-)
16:10 Zeno` merged
16:14 juhdanad I am awake now!
16:15 VanessaE you missed the whole conversation :)
16:16 juhdanad Sorry, I was playing on your creative server...
16:17 blaze joined #minetest-dev
16:17 VanessaE no worries :)
16:19 Fixer here comes dat smooth lighting
16:19 sapier hmmm is there any way to make grass grow below a fence?
16:20 red-001 I though it does grow?
16:20 Fixer i want to rise this again, I don't wanna see another minetest console window in Windows, it trashes my taskbar, i want it hidden by default (and enabled via options)
16:20 red-001 well it's not possible to do that without major hacks
16:20 sapier any windows developer here?
16:20 juhdanad VanessaE: I was thinking about texture overlays. It is possible without z-fighting (provided that the vertices are the same). Just I have to code it.
16:21 red-001 windows doesn't really allow it
16:21 red-001 even msvc uses a hack for it
16:21 VanessaE juhdanad: I look forward to it. :)   I have two mods that use the above hack, and at least one more that would benefit from an overlay.
16:22 juhdanad But there will be limitations: if you use a translucent overlay, the base image must be opaque to prevent translucency bugs.
16:22 VanessaE that's reasonable
16:22 Fixer i remember with 0.4.13 it was showing console window only ONCE, and next startups it was hidden somehow
16:25 sapier talking about overlays :-) what about #5045?
16:26 ShadowBot https://github.com/minetest/minetest/issues/5045 -- Add "multiply" texture modifier which uses multiplication method for colorizing textures by sapier
16:26 sapier I hate my green faced sheep when using the colorize modifier ;-)
16:27 sapier and I don't really wanna add a texture per mob for each possible color
16:28 sapier btw is anyone working at a more sophisticated collision handling?
16:29 Fixer have not heared of this
16:29 sapier sad
16:29 Fixer i remember this https://github.com/minetest/minetest/pull/3810 if relevant
16:31 sapier no I was more thinking about collision handling with rotating collision boxes right now if you have any asymetric mob it's quite silly
16:35 VanessaE juhdanad: oh and by the way, the total idiocy of that guy on my server is just par for the course :-/
16:39 juhdanad VanessaE: how should overlays be implemented? An "overlay" drawtype? Or one more tile field to nodedef?
16:42 red-001 what happened with  #5005?
16:42 ShadowBot https://github.com/minetest/minetest/issues/5005 -- Recursive refining mapgen API prototype by MillersMan
16:42 juhdanad I think it is pending.
16:43 VanessaE juhdanad: tiles = { foo, bar, ..., { name = "baseimage.png", overlay = "overlay.png", color = 0xffffffff } }
16:43 VanessaE or well, you get the idea.
16:43 juhdanad But MillersMan seemed very enthusiastic. I'm sure he will upload a new commit sooner or later...
16:43 sapier imho the grass growing is broken
16:43 juhdanad VanessaE: And is color for the base image or for the overlay?
16:44 juhdanad (because the streets mod could colorize the overlays, then)
16:48 VanessaE it's for the overlay
16:48 VanessaE (so I suppose the overlay and the color should be in a table too
16:49 VanessaE tiles = { foo, bar, ..., { { name = "baseimage.png", overlay = { name = "overlay.png", color = 0xffffffff } } }   or so
16:51 juhdanad What about an array-like definition? tiles = { foo, bar, ..., {{ name = "baseimage.png" }, { name = "overlay.png", color = 0xffffffff } } }
16:52 juhdanad (but only the first two elements are read)
16:53 VanessaE well I'm okay with any way you'd do it, as long as it's painfully obvious in the code what element belongs to where :)
16:53 VanessaE (I have trouble parsing a table when it's compacted into one line :-) )
16:55 juhdanad tiles={up=down=front={base={name="a.png", color="red"}, overlay={name="b.png", color="white}}, back=left=right="c.png"}
16:56 VanessaE that seems good
16:57 kaeza joined #minetest-dev
17:02 VanessaE am I to assume you want to put the literal words up, down, etc in there?
17:02 juhdanad Well, that would be good too, but out of scope...
17:02 VanessaE oh ok
17:03 VanessaE it's probably not needed since the texture order is hard-coded anyway
17:03 VanessaE s/texture/tile/
17:04 juhdanad Well... I caused a compiler warning with my PR that fixed compiler warnings. I feel so ashamed...
17:05 VanessaE heh
17:05 Void7 joined #minetest-dev
17:12 sapier :-)
17:14 juhdanad VanessaE: if you dig a node with overlays, how should the node partivles look like?
17:14 juhdanad *particles
17:16 VanessaE oh, just whatever's easiest
17:28 est31 joined #minetest-dev
17:50 numZero_ joined #minetest-dev
17:55 numZero joined #minetest-dev
17:56 YuGiOhJCJ joined #minetest-dev
18:35 red-001 #5142
18:35 ShadowBot https://github.com/minetest/minetest/issues/5142 -- [CSM] Add `on_punchnode` callback by red-001
18:47 nrzkt great job red-001
18:47 nrzkt just a code style fix to do
18:47 nrzkt thanks for your involvement
18:51 celeron55 wait what
18:51 red-001 fixed
18:51 celeron55 there's no player argument even while the callback is apparently solely triggered by player interaction
18:52 celeron55 how can that make sense from a mod's standpoint
18:52 red-001 it's client-sided and only runs if it's done by the local player
18:52 celeron55 why does it say "a player" then
18:53 celeron55 it should say "the local player"
18:53 red-001 yeah sorry about that
18:54 fwhcat joined #minetest-dev
18:56 red-001 fixed it
19:05 celeron55 red-001: i meant doc/client_lua_api.txt
19:07 Warr1024 joined #minetest-dev
19:10 Miner_48er joined #minetest-dev
19:11 fwhcat joined #minetest-dev
19:12 red-001 celeron55, fixed that, along with a few similar issues
19:14 sofar red-001: can't client-side see all player node punches then?
19:14 sofar at least the ones nearby
19:15 red-001 I though the server just send an updated copy of the mapblock>
19:15 red-001 ?
19:15 sfan5 sofar: nope
19:16 sofar so we can't do e.g. digging particles for other players?
19:16 sfan5 not currently
19:16 sofar aw, damn, that sucks
19:19 celeron55 can't a server-side mod just send that info?
19:20 celeron55 we have modding on both sides; one could reasonably hope that's possible
19:30 red-001 so is the next mt version going to be 0.4.16 or 5.0 or 0.5?
19:30 sapier for the time being I'd expect 0.4.16
19:31 red-001 I remember a few days ago there as talk of 5.0
19:32 red-001 was*
19:32 sapier for what reason? what happened to 1.0 2.0 3.0 4.0?
19:33 red-001 not sure...
19:33 red-001 think it's because of the 0.5 hype
19:33 sapier we're building csm right now imho we're far from releasing anything that could be called "stable" ;-)
19:46 Thomas-S joined #minetest-dev
19:51 Thomas-S joined #minetest-dev
19:52 silwol joined #minetest-dev
19:53 silwol I just got the hint that I should tell in this room what I just said in #minetest:
19:54 silwol I want to shout out a big thank you to all the devs of the engine, game and the mods.
19:54 silwol Having quite a lot of fun with my 5- and 8-year old sons on weekends playing minetest.
19:54 silwol I tried one of the first minetest versions several years ago, and it couldn't run longer than a minute before it crashed back then. It is really amazing how the game evolved in the meantime.
19:54 sfan5 it's mostly thanks to all the contributors
19:56 rubenwardy joined #minetest-dev
20:53 Fixer remember that low fps in technic areas? Please adopt! https://github.com/minetest/minetest/pull/3848 (Send only changed node metadata to clients instead of whole mapblock)
21:12 Fixer https://github.com/minetest/minetest/issues/4923 interesting
21:24 nrzkt sapier, the fact was CSM is a very good feature for a major bump, in minetest 0.4 -> 0.5, but 0 is really useless, and may be removed at a moment, but i vote at least for 0.5 and why note remove the 0
21:34 shivajiva joined #minetest-dev
21:37 CuriousLoon joined #minetest-dev
21:37 CuriousLoon left #minetest-dev
21:52 rubenwardy nrzkt: I peronally like the idea of going full 1984, and pretending the 0. never existed in the first place - ie: 4.15 not 0.4.15.  Makes it semvar, basically
21:52 rubenwardy almost
21:54 red-001 thats an interesting idea
21:55 nrzkt full 1984 ? except george orwell i don't understand what are you talking about :)
21:55 Fixer rename minetest?
21:55 Fixer this is as ridiculous name as qtpfsgui
21:57 red-001 nrzkt, I think a major plot point in 1984 was changing history
21:58 red-001 so remove the 0. from the version and pretend it was never there
21:59 red-001 so 0.4.10 will be known as 4.10 from that point onwards
21:59 Fixer nah
22:00 Fixer i still think ver 1.0 should be more or less complete game
22:06 sapier I'm not a friend of alternative facts
22:06 sapier and I don't see any reason in changing versioning style just to be "modern" ... quite stupid for a game in pixel retro style ;-)
22:09 Calinou retro pixel style doesn't mean it should be unfriendly
22:09 Calinou it's not like you need to install DOSBox to run Minetest already :)
22:10 sapier good point we need a dos port
22:10 sapier way more important then versioning  ;-)
22:11 juhdanad joined #minetest-dev
22:12 kaeza well, there's MesaGL for DOS...
22:13 sapier irrlicht contains software rendering so we wouldn't even need mesa
22:13 nrzkt i hope at a moment this year or next year we will switch to CPP11 or CPP14 as debian 7 will be deprecated and vista is not supported anymore
22:13 sapier sadly there doesn't seem to be a dos port of irrlicht
22:16 Calinou nrzkt: Minetest still supports Windows XP...
22:16 Calinou and no plans to drop it
22:16 nrzkt micrsoft drop it, firefox drop it, no need to support it anymore
22:16 red-001 ^
22:16 sapier imho we should switch to reactOS
22:16 nrzkt google drop it, java drop it, there is nothing more on it
22:17 red-001 I'm not sure if mt would run in reactos
22:17 red-001 graphics drives
22:17 sapier I'm still running winME ;-)
22:17 sapier ok ok in a rarely used virtual machine for running a freelancer dedicated server ... which I do about 1 a decade
22:17 red-001 and software rendering is completely broken and should be removed imo
22:18 sapier is it?
22:18 red-001 try it last time I did everything was red
22:18 red-001 and plants where broken
22:18 sapier of course same happens if you enable shaders
22:18 sapier on any platform which doesn't support them
22:20 kaeza software rendering may be useful in some cases. I get better performance with burningsvideo than with opengl inside a VM :P
22:20 sapier it's for sure not as good as d3d or opengl but "completely broken" is wrong for sure
22:20 red-001 well it should auto disable shaders
22:21 sapier kaeza: last time I tried my win7 vm it did run direct3d in native speed ;-P
22:21 Calinou VirtualBox 3D isn't very good, both in standards compliance and performance
22:21 sapier well reason was most likely I did pass a graphics card directly to the vm so it was native speed ;-)
22:21 Calinou if you use vmWare it's a whole different story
22:21 red-001 well if mt works in reactos someone should add it to their software store
22:22 sapier kvm pcie passthrough ...everything else is just fooling around ;-)
22:22 VanessaE any chance we can get another release soon?
22:22 VanessaE (just to make some of my users happy :P )
22:23 sapier last time I didn't manage to start reactos within a kvm virtual machine
22:23 sapier well without csm maybe vanessae
22:23 nrzkt VanessaE, absolutely not
22:23 VanessaE sapier: I was thinking some kind of interim release, might be logically equivalent to 0.4.15.1 or so :)
22:23 VanessaE nrzkt:  :(
22:24 nrzkt except if we want to do a maintenance release before CSM merge
22:24 red-001 sapier, it works in virtualbox
22:24 sapier nrzkt: I'd suggest that way
22:24 nrzkt but i hope CSM will trigger the 0.5
22:24 sapier red-001: I didn't try for I while
22:24 nrzkt because it's really a new huge part of mt
22:24 red-001 nrzkt, whats stoping the colour in chat stuff from being merged to csm?
22:24 nrzkt second approval i think red-001 ?
22:24 VanessaE nrzkt: that's what I'm asking for - a maintenance release or so
22:25 red-001 It's working correctly
22:25 red-001 at least the features it add are
22:25 nrzkt i don't know if the current master could be a maintenance release, there are maintenance but also new features :)
22:25 nrzkt red-001, can you provide me the PR another time ?
22:25 red-001 the link to it?
22:25 sapier I'd consider current master to be a regular release
22:25 nrzkt yes
22:26 nrzkt i think CSM will be ready for merge into master in ~1 month
22:26 red-001 #5113
22:26 nrzkt then we should do the maintenance release in 1 month i think, last of 0.4.X series
22:26 ShadowBot https://github.com/minetest/minetest/issues/5113 -- [CSM] Add function to get the name of the player and functions to colour chat. by red-001
22:26 nrzkt ty
22:26 rubenwardy you could always do a release before CSM is merged, then use the next dev to develop CSM nicely
22:26 VanessaE nrzkt: do one now
22:27 rubenwardy overwise it would be month until the next one
22:27 VanessaE then do another just before CSM goes in
22:27 nrzkt VanessaE, absolutely not, we have feature freeze workflow, announce etc
22:27 red-001 so a new release after just 3/2 months?
22:27 VanessaE :(
22:27 sapier nrzkt: do you have any additional change requests for #5045?
22:27 ShadowBot https://github.com/minetest/minetest/issues/5045 -- Add "multiply" texture modifier which uses multiplication method for colorizing textures by sapier
22:27 sapier same question to sfan5 ;-)
22:27 nrzkt rubenwardy, the problem is not working on CSM in a separate branch, but some features are needed server side too, like the storage API
22:27 red-001 Question can servers check the protocol version of a client from lua?
22:28 VanessaE nrzkt: I'm saying short-circuit the normal freeze, just like some folks do who release their own dev builds
22:28 nrzkt yes
22:28 nrzkt red-001, ^
22:28 red-001 huh what's the function called then?
22:28 red-001 I could never find it
22:28 sapier nrzkt: why not merge the base stuff to master instead of keeping it on csm only?
22:29 nrzkt VanessaE, it's not possible, this cycle release include new features, we should keep the regular workflow and MT doesnh't have many hands to keep a patchfix branch outside of master
22:29 sapier "base stuff" meaning the common parts not beeing csm specific
22:29 VanessaE :-/
22:29 red-001 VanessaE, why not ask your users to use daily builds>
22:29 nrzkt sapier, personnaly i'm experimenting it client side only, but it's just my own tests
22:29 red-001 *?
22:29 sapier I see
22:29 VanessaE red-001: because some of them think dev builds are dangerous or just stubborn
22:30 rubenwardy the whole point of releases is that they're stable
22:30 red-001 I'm not sure if thats a good reason to rush a new release
22:30 VanessaE no one said to rush :P
22:31 nrzkt yes, no need to rush, we have a protocol bump in master, too
22:31 sapier I'm quite sure it's never a good idea to rush a release ;-)
22:31 nrzkt it's another reason to keep the feature freeze as an RC
22:31 red-001 There are moderators on servers right now that can't get access to even 0.4.15
22:32 nrzkt sapier, except a little code style issue it's good to me
22:33 sapier additional issues? I thought I fixed them :-)
22:33 nrzkt but VanessaE i agree we should have shorter releases cycles, 7 months is too big, 3 months is better
22:33 nrzkt a release on middle march is a good option
22:33 nrzkt last was 22/12
22:33 red-001 nrzkt, for  #5133 I'm not sure if it should be the job of a fairly small PR to improve the code style of a fairly large file
22:33 ShadowBot https://github.com/minetest/minetest/issues/5133 -- [CSM] Block access to the `io` library by red-001
22:34 nrzkt red-001, you talk about shadowninja's request to factorize initialize functions ?
22:35 red-001 No I mean your request to split misc_helpers
22:35 nrzkt sapier, can you review #5113 please ?
22:35 ShadowBot https://github.com/minetest/minetest/issues/5113 -- [CSM] Add function to get the name of the player and functions to colour chat. by red-001
22:35 red-001 checking INIT is done all over that file
22:35 nrzkt oh it's a current issue ? then let it, no problem, i didn't saw that
22:35 nrzkt it's very ugly :(
22:36 red-001 I would agree
22:36 red-001 I think another pr should be do to fix that
22:36 red-001 done*
22:37 nrzkt yeah, but it's not important in CSM
22:37 nrzkt great jobs on dig callbacks etc
22:39 red-001 I think we server <-> client communication and server being able to send mods to the client,we might be able to cut down on people gliching through protected nodes
22:39 red-001 I still need to add on_placenode
22:40 sapier nrzkt: done, imho only a small paranoia check is missing
22:40 rubenwardy going through protected nodes should really be solved server side
22:40 rubenwardy as hacked/cheaty clients could remove the mods/checks client side
22:40 red-001 I did say cut down
22:40 red-001 not remove
22:41 red-001 but at that point it's just noclip
22:43 nrzkt rubenwardy, area protection are fully serverside, i don't understand which issue you refer
22:43 nrzkt no_clip ?
22:44 red-001 nrzkt, you can gliche through protected areas by digging down and exploiting the lag
22:44 rubenwardy nrzkt, I'm talking about anti-noclip and was referring to <red-001>  ... people gliching through protected nodes
22:45 rubenwardy *was
22:45 nrzkt i don't understand how this can be possible as protection is server side, only no-clip check is not server side
22:45 nrzkt but with suffocation mod you can block no clip
22:46 rubenwardy with high latency, you can dig a block and walk through it before it comes back
22:46 rubenwardy also, relevant PR #3894
22:46 ShadowBot https://github.com/minetest/minetest/issues/3894 -- Add serverside noclip enforcement by est31
22:47 nrzkt rubenwardy, oh you mean the client side node dig prediction then
22:47 nrzkt just remove it from client and that goes away, it's not needed
22:49 sapier well it's two sides of same medal if you remove it user experience in high latency scenarios will be wors then today (whenever no protection is relevant)
22:50 rubenwardy it is needed for UX
22:50 sapier the only real solution would be add it to core
22:51 sapier protection would have to be a part of node properties
22:51 sapier but that wouldn't solve the issue
22:51 sapier user could still use a modified client
23:03 sapier1 joined #minetest-dev
23:03 Fixer rubenwardy: +1
23:04 sapier1 comments about #5073? I know there's not a lot of ppl using that stuff but that guy did really good work
23:04 ShadowBot https://github.com/minetest/minetest/issues/5073 -- Rebased Pull request #3620 by sapier
23:04 red-001 fixxed issues with #5113
23:04 ShadowBot https://github.com/minetest/minetest/issues/5113 -- [CSM] Add function to get the name of the player and functions to colour chat. by red-001
23:05 Fixer sapier1: that naming sucks :(
23:05 sapier1 :-) well blame german telecom but I guess I could reconnect in between
23:06 Fixer sapier1: please name 5073 in some meaningful way
23:06 sapier joined #minetest-dev
23:06 sapier ok should be back now
23:06 sapier I can't review 5073 myself as I did rebase it ... silly github :-)
23:08 Guest93432 joined #minetest-dev
23:09 red-001 lol
23:16 sapier left #minetest-dev

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