Minetest logo

IRC log for #minetest-dev, 2017-03-23

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

All times shown according to UTC.

Time Nick Message
00:20 Fixer paramat: wuzzy used desert sand to turn regular grass into dry shrub, why can't we still keep this (conversion on desert sand)?
00:20 VanessaE junglegrass mod used to do that.
00:22 Fixer paramat: you don't place any green grass on desert sand anyway, so grass to shrub can be enabled for just this case
00:23 Fixer it seems
00:34 rubenwardy joined #minetest-dev
00:51 VanessaE 2017-03-23 01:50:45: WARNING[Server]: ServerEnv: Trying to store id = 1046 statically but block (-7,0,44) already contains 14096 objects. Forcing delete.
00:51 VanessaE oh really. :P
00:54 paramat Fixer ok good point, we should consider this
00:55 turtleman joined #minetest-dev
03:31 ssieb joined #minetest-dev
03:37 froike joined #minetest-dev
03:37 DI3HARD139 joined #minetest-dev
03:57 Puka joined #minetest-dev
04:24 ^v joined #minetest-dev
05:05 ^v joined #minetest-dev
05:07 mlucena_ joined #minetest-dev
06:06 Hunterz joined #minetest-dev
06:11 Icedream joined #minetest-dev
06:12 vifino- joined #minetest-dev
06:12 ekem joined #minetest-dev
06:13 AnotherBrick joined #minetest-dev
06:18 ^v joined #minetest-dev
06:43 nerzhul joined #minetest-dev
07:04 Karazhan joined #minetest-dev
07:08 Vadtec joined #minetest-dev
07:09 rdococ joined #minetest-dev
07:10 AnotherBrick joined #minetest-dev
08:19 kilbith joined #minetest-dev
08:32 nrzkt joined #minetest-dev
09:18 TheReaperKing joined #minetest-dev
09:48 Karazhan joined #minetest-dev
09:50 Darcidride joined #minetest-dev
09:55 Karazhan joined #minetest-dev
10:30 glut32 joined #minetest-dev
10:49 YuGiOhJCJ joined #minetest-dev
11:27 lisac joined #minetest-dev
11:38 proller joined #minetest-dev
12:05 Fixer joined #minetest-dev
13:00 proller joined #minetest-dev
13:29 proller joined #minetest-dev
13:53 octacian joined #minetest-dev
13:53 octacian joined #minetest-dev
14:02 Puka joined #minetest-dev
14:13 XtremeHacker joined #minetest-dev
14:36 celeron55 so can i revert the sneak fix to see what people think then?
14:37 celeron55 or, well, i can, but does someone have something to say that would make me not do it?
14:37 celeron55 this https://github.com/minetest/minetest/commi​t/14aa990cdd682b00904c4a84968a6c0e3c31716d
14:38 celeron55 it seems to be causing quite a lot of problems to be considered a fix
14:40 nrzkt celeron55: what are the problems ?
14:41 celeron55 players and server owners seem to act really passive-aggressively and not say anything out loud but there is this issue https://github.com/minetest/minetest/issues/5438
14:41 celeron55 and i even heard people are considering forking MT based on this PR, which sounds ridiculous but is telling of something i guess
14:42 celeron55 it's a too considerable gameplay change for many it seems
14:42 kilbith forking MT for sneaking-ascent, how smart...
14:43 nrzkt it's like blockmen fork :s
14:43 celeron55 well it's a bit similar to me not wanting to use firefox after they get rid of XUL
14:43 lisac Can't someone make it optional?
14:43 celeron55 i don't blame them, everything can be a feature
14:43 lisac so the server admin can enable/disable it?
14:43 celeron55 that would work
14:44 celeron55 but it has to happen; not just talk
14:44 Shara Server owner controlling it would be ideal, but the change itself is to the client.
14:44 celeron55 i'll revert the commit in 1 hour unless someone actually promises to do something
14:47 Taoki joined #minetest-dev
14:47 nrzkt celeron55: okay
14:48 celeron55 and it can be added back once it is optional or somehow otherwise acceptable to those that like the old behavior
15:12 YuGiOhJCJ joined #minetest-dev
15:14 Fixer celeron55: if you want to revert really hard, please fix damage evasion and nocliping while sneaking :trollface:
15:15 celeron55 i'm sure they'll get fixed eventually without the other problems being added
15:16 Fixer and i'm not
15:16 Fixer sfans fixed them but that killed off that sneak glitch, are there other ways?
15:16 nrzkt Fixer: see #4981
15:16 ShadowBot https://github.com/minetest/minetest/issues/4981 -- [WIP] Damage cheat fix: server side by nerzhul
15:17 nrzkt if you can help to fix the last point, we won. No more client side declared damaged
15:22 Fixer nrzkt: how?
15:24 nrzkt just look at the issue, there is one mising thing to do
15:24 nrzkt you can help to do it, i missng dev time on minetest this week and next week, i can only do short PR
15:27 YuGiOhJCJ joined #minetest-dev
15:32 twoelk joined #minetest-dev
15:51 paramat joined #minetest-dev
15:52 paramat celeron55 please don't revert, the situation isn't as bad as you think
15:53 paramat give me 30 mins and i'll post links to show stuff is being sorted out, replacement mods are being made
15:54 sfan5 paramat it's not about replacement mods
15:54 sfan5 let him revert it
15:55 paramat it is about replacement mods, and the new bugs are minor and are probably fixable, unlike the old bugs
15:55 sofar the actual fuck
15:56 sofar putting the sneak ladder back in?
15:56 sofar why?
15:56 paramat the fixing of sneak bugs has a lot of support from devs and contributors, see the emoticons in first post of the PR thread
15:58 paramat #5327 please read the whole thread
15:58 ShadowBot https://github.com/minetest/minetest/issues/5327 -- Fix various problems with sneaking by sfan5
16:00 paramat see this comment from a previously angry player https://github.com/minetest/minetes​t/pull/5327#issuecomment-287840410
16:00 sfan5 ugh the whole local player::move code just seems to assume all nodes are full cubes
16:00 paramat the whole idea is that this is being moved from a bunch of serious engine bugs to being an intentional coded feature provided by mods
16:01 paramat i spent 2-3 days working with someone on a mod that can be used as a sneak ladder replacement https://github.com/Ezhh/handholds
16:02 paramat and intend to write a replacement for 2-node sneak-jump
16:02 sofar did I miss a constructive conversation between sfan5 and celeron55 where they both agree to revert or something?
16:02 sfan5 no you didn't and it isn't required
16:02 paramat no-one agreed to revert
16:03 sofar according to celeron55 he's going to revert unless someone fixes something
16:03 sofar and I don't understand what "fixes something" even means
16:03 sofar or what demands there even are
16:03 paramat i have made several PRs to make climbing faster and ladder/rail recipies much more generous to compensate
16:03 sfan5 fixes the problems people have with the pr
16:03 sfan5 namely: broken sneak ladder
16:04 sofar wow, just wow
16:04 paramat the new bugs need some time for us to work on and fix, and they are far less serious than the old bugs
16:04 sofar this is really about adding a bug back into the game?
16:05 sofar for what? just so that PC users have an advantage over mobile players?
16:05 paramat plus this is a dev version, stuff is expected to be suboptimal sometimes, we intend to sort the situation out before 0.4.16
16:05 sfan5 you misspelled feature
16:05 sfan5 ;)
16:06 sofar you are being way too modest
16:08 nrzkt wow many devs active for sneaking. What is the real end user problem ? Cheating to avoid damages or something else ?
16:10 paramat fixing the sneak bugs is enabling us to close about 7 issues, they were unfixable and tied-in to how sneak worked. the new method at least has a good chance of being fixable
16:10 sofar ffs just give all clients fly and noclip while you're at it...
16:10 paramat we can even add new engine code to allow mods to more closely replicate the old behaviour, if necessary
16:11 sofar paramat: to be fair, I don't see any of that happening
16:12 sofar and honestly, the new climbing holds are a much nicer addition to the game then some trivial sneak ladder
16:12 sofar both aestetically and mechanically
16:12 Calinou sneak ladders can be relatively annoying to build too, it doesn't bring much value
16:12 sofar above ground they look terrible
16:12 Calinou like, if you get *one* node wrong, you may have to rebuild all of it :)
16:12 benrob0329 joined #minetest-dev
16:13 sofar and they encourage lazy and ugly builds
16:13 sfan5 it does not matter how annoying it is when players like it
16:13 sofar I disagree, we're the ones who can make the decision
16:13 sofar we can make decisions on anything we want
16:14 sofar if we think the game will be better (if that's what we want) without or with it, totally valid choice
16:14 celeron55 the end user problem is the dropping of a mechanic they found really fun to use (and apparently also making sneaking feel worse, maybe)
16:14 Billre joined #minetest-dev
16:14 celeron55 it doesn't really matter if you found it fun to use; they did
16:15 sofar I didn't say it mattered how "I found it fun to use"
16:15 sofar I never said that
16:15 nrzkt joined #minetest-dev
16:15 sofar and I'm not basing my objections on that
16:15 rubenwardy joined #minetest-dev
16:15 sofar I'm basing my objections on my observations of the population of players using the feature
16:15 proller joined #minetest-dev
16:15 sofar not on my own game enjoyment
16:16 benrob0329 Having a game be fun is more important than having it be correct
16:16 Billre do I have voice?
16:16 sofar there is an argument for "taking out tedious tasks"
16:16 paramat in the end, devs know best, some players are overreacting or don't understand the whole situation, most players don't even know about sneak ladder and sneak-jump
16:16 sofar but sneak ladders are not tedious tasks IMHO
16:16 rubenwardy I agree that sneak ladders aren't good game design
16:17 sofar mining, now that is a tedious task
16:17 sofar let's remove that
16:17 sofar ?
16:17 rubenwardy the most upset players are the loudest
16:17 sofar someone tell me what the reason for adding sneak ladders was in the first place
16:18 sofar because without that argument on the table we're just going to make the same mistake again
16:18 paramat and yes the handholds mod is so good it makes sneak ladder seem crude and nasty
16:19 sofar something tells me the sneak ladder was an accident and nobody even knows anymore why it exists
16:19 paramat sneak ladder wasn't intentional, it was a side-effect
16:19 rubenwardy "the reason for adding sneak ladders" => was a ug
16:19 rubenwardy *bug
16:19 sofar one that allows players to go through solid nodes, too
16:19 paramat same with sneak-jump, the player was never meant to jump up 2 nodes
16:20 paramat and avoid fall damage and several other bugs
16:20 sfan5 small side node: sneak jump does not cause a jump height of 2, its less actually but enough to "sneak-clip" to a node y=2 above the original pos
16:20 celeron55 ok i'm going to wait until feature freeze of next stable to finally assess this
16:21 sofar you mean "we"
16:22 celeron55 well, my part of "we"
16:22 sofar I have no issues with re-evaluating decisions
16:22 sofar but let's do it before time runs out and people want to release
16:22 sofar and not force a decision by time
16:23 celeron55 should we set a date?
16:24 sofar we're talking about it now?
16:24 sofar are you in hibernation or something? :)
16:24 paramat setting a date isn't wise, release date has to be flexible
16:25 paramat but 0.4.16 release is intended for around june
16:26 TheReaperKing joined #minetest-dev
16:26 paramat sneak ladder and sneak jump are features that obviously need to be subgame things, not unavoidable engine bugs
16:26 celeron55 well i just don't think in a similar way to sofar i guess
16:27 celeron55 yes, and without frame-by-frame client-side physics you can't implement them
16:27 celeron55 unless the engine implements them and you can turn them on
16:27 paramat anyway, i do think we need to put a lot of effort into providing optional replacements
16:27 celeron55 or unless something like what i tried in my experiment was added
16:28 paramat yeah they can't be identically replicated, but that's npt necessary
16:28 paramat *not
16:29 paramat handholds mod takes the idea of holes in walls as handholds and improves it
16:29 sofar I can see this end up ending in a new "fly, fast, noclip" priv
16:29 sofar and I don't like that
16:29 sfan5 nah
16:29 sofar but that's the only thing that makes sense
16:29 sfan5 it makes way more sense to put this into the "sneak_glitch" phys. modifier
16:29 Shara paramat:  please stop using handholds as a sneakladder replacement idea. It's not
16:29 sofar many people will want to disable it
16:29 paramat i know
16:29 sofar many servers want to disable it, for all clients
16:29 paramat i mean 'it can be altered to act as a replacement'
16:29 sofar so it should be a priv
16:29 Shara I really appreciate you helped me with that mod, but you seem to have backtracked on considering it for MTG like you told me.
16:30 Shara Which is fine if so, but it was always intended to be a mod standing on its own merrits regardless of sneak ladderws
16:30 Shara ladders*
16:31 Shara We would need much better options than this mod to get back the mobility that is now missing.
16:31 sofar sfan5: is "sneak_glitch" a server packet to client?
16:31 sfan5 physics overrides are set by the server yet
16:31 sfan5 s/yet/yes/
16:31 paramat i'm undecided on inclusion, all i said was i wouldn't oppose it
16:31 sofar well that's something at least then
16:31 sofar you'd avoid the priv
16:31 paramat per-node climb speed would get a mod closer
16:32 Shara I'd love this if it was optional per server or priv based. I really would. Because I am not in any way against the server owner getting to decide how they want their server played and having a way to prevent abusive use of it to 'cheat'.
16:34 sfan5 oh also something i have noticed when working with the code:
16:35 sfan5 quite often the assumption is made that "air" is the only airlike node that ever exists
16:35 sfan5 so it checks for nodename == "air" instead of nodedef[nodename].type == "airlike"
16:35 sfan5 most likely nobody hit this yet so nobody complained
16:41 Shara Added an airlike node to a mod just this week...
16:45 Shara paramat: sorry, missed earlier message. Problem is, if no one opposed it for MTG, it would just sit there. Devs would actually need to want it included. So I'm left not knowing what to do when I'm told people wouldn't oppose.
16:48 paramat yes, the other mtg devs can give better feedback than me at the moment
16:50 Shara Would appreciate if any would then. I am quite happy to make changes as I said on the forum.
16:52 paramat ok, because of the problem of easyness, i feel the tool recipe needs to be more expensive (diamonds)
16:52 Shara Well, I favour more difficulty there myself, and also lowering durability further
16:53 Shara I would want it as a ladder alternative, not replacement, so that balance is important
16:54 rubenwardy I like it, not sure about the balance of it vs ladders
16:56 Shara Thanks rubenwardy. If you have any suggestions for balance adjustments, please let me know. Thinking diamonds for recipe and drop to 250 uses at the moment.
16:56 rubenwardy yeah
16:57 rubenwardy I guess if the hand holds only work in natural rock it'll be fine
16:57 rubenwardy as ladders also work against walls etc
16:57 Shara It's currently stone, desert stone, sandstone and ice
16:57 rubenwardy that's good
16:58 paramat the 4 terrain nodes occuring in mapgen
16:58 Shara I'd only want to add other nodes if more were added to that mapgen group
16:58 Shara Anything crafted or soft should require ladders to climb up instead
16:59 paramat is sneak functionality disabled on climbable nodes? there may be issues with it being activated when 'moving down' on a climbable node, because it's the same key
17:00 paramat oh i see !climbing in the code now
17:01 paramat but it only affects a small block of code
17:01 twoelk joined #minetest-dev
17:01 Hunterz joined #minetest-dev
17:04 paramat perhaps that could be combined into the other bools that decide sneaking
17:11 twoelk what is actually the original reason sneaking exists? is that functionally still fully provided?
17:16 paramat because MC has it and players find it useful. and yes it is still provided
17:16 Shara In a rather limited and glitchy form, if the purpose is what I understand it to be.
17:21 paramat to be fair, the old behaviour was highly glitchy too
17:21 Shara Not disputing that, but at least I didn't fall off endges when I didn't expect to before
17:22 Shara edges*
17:22 Shara Snow and slabs (the smaller ones added by many mods) when placed on edges can now cause you to fall straight off even when holding shift
17:25 Fixer nrzkt: i can only help with testing and suggestions, Vanessa probably can enable this on her server for full experience for greater good
17:28 * VanessaE hides
17:31 DFeniks joined #minetest-dev
17:37 Fixer sofar: "what the reason for adding sneak ladders?" it was not intentional I think, and people just used, and exploited the bugs along it, I remember it since 2011
17:38 Fixer and ugly bugs they were
17:38 Fixer also, above ground sneak ladders are ugly, while we have cheap ladders and water elevators
17:42 Fixer most popular server have /spawn, you can dig down - place water, mine and then /spawn, or /home (and reuse), you can make waterboat elevator as I do, you can use mods like Shara's one
17:43 paramat i'll merge game#1662 later if no objections, fairly trivial as it roughly returns to original behaviour
17:43 ShadowBot https://github.com/minetest​/minetest_game/issues/1662 -- Flora spread: Re-enable dry shrub replacing flora in desert sand by paramat
17:43 benrob0329 joined #minetest-dev
17:44 Fixer i really wanted falling bugs to be fixed, and sfan5 did succeed here more or less, but sneak-ladder breaking was the result, and I;m personally fine with this
17:44 Fixer i would choose bugless falling over sneak-ladders any day
17:46 Fixer Calinou, "sneak ladders can be relatively annoying to build too", yep, thats why you can rarely find really deep sneak ladder, and if server has /spawn or /home... you just take water with you and don't bother with anything
17:46 Shara Fixer: Even those against this (or most of us I guess...) aren't against it because of the falling damage avoidance being removed.
17:46 paramat Krock sofar rubenwardy nerzhul #5301 seems good now, any approvals?
17:46 ShadowBot https://github.com/minetest/minetest/issues/5301 -- fine_pointed added by MarkuBu
17:47 paramat then we can get new slab placement in mtg
17:47 rubenwardy #5439 ?
17:47 ShadowBot https://github.com/minetest/minetest/issues/5439 -- Add pointed_thing_to_face_pos to misc_helpers and doc by MarkuBu
17:47 Fixer Shara: if any falling bugs get mirac-sly fixed and sneak ladders still working - i would not bother tbh
17:48 Fixer and that noclip while sneaking, and falling through slabs/stairs
17:48 Shara Not sure I understand what you mean.
17:48 Fixer thats the stuff from mc beta era, that was fixed ages ago
17:48 Fixer like 6 years ago, and we still have this crap unfixed
17:49 Fixer Shara: i don't care about sneak ladders, I care about falling bugs/noclip/slabs/damageevasion
17:49 Shara Either way, there's a lot of people unhappy with this. You might think spoliing many people's enjoyment is worth fixing falld amage, but I'd prefer to see a way to fix that without this.
17:50 Shara It's seems like there is that PR that, if finished, could do it... so maybe if that can happen then this reverted, almost everyone can be happy.
17:54 paramat well, sneak ladder and sneak jump are features that need to be decided per-subgame, they have no place being unavoidable in the engine
17:54 ssieb joined #minetest-dev
17:54 paramat which PR btw?
17:54 Shara Hold on, let me check the log from earlier
17:55 Fixer paramat: https://github.com/minetest/minetest/issues/4981
17:55 Fixer paramat: +1 about per subgame basisc
17:56 Shara Fixer was quicker :)
17:56 Shara And would just be nice if server owner could decide.
17:57 paramat a subgame needs to be able to decide what is climbable, the bug made any alternating pattern climbable. likewise the sneak jump made 2 node high jumps possible but how high you can jump should be decided by physics settings and per-player physics override
17:57 Shara paramat: this isn't about default player jump height though
17:57 paramat i'm not against sneak ladder/jump being possible, but they need to be intentional and optional
17:58 Shara Optional is what I'd like.
17:59 paramat hm well that PR addresses damage bug only, only one small part of the issue. everybody won't be happy because most devs will not be
18:02 paramat so, reversion is not good because the old code forced all related bugs, it was not possible to separate desirable parts from the bugs or make them optional, but new code may be able to
18:02 Shara paramat: problem is, with this now merged, no dev with the ability seems to want to do that :)
18:03 paramat not sure, sfan5 ?
18:04 paramat #4959 is simple and needed, can anyone review?
18:04 ShadowBot https://github.com/minetest/minetest/issues/4959 -- Fix retrieving blocks in 3rd front camera view. by lhofhansl
18:04 paramat but anyway, mods can already provide replacements, we mau only need minor new engine code such as per-node climb speed
18:05 paramat *may
18:05 Shara per node climb speed would be nice anyway.
18:05 paramat much more important is fixing the bugs of the new code
18:05 Shara But I don't see how mods can do most of this, (or I'd be making the mods)
18:06 paramat you just have :]
18:06 Shara I don't know how to extend the distance of the edge that you can sneak, or how to enable the double jumping to climb the edge of a node.
18:07 Shara Or how to switch back from this new way of controlling sneak either...
18:07 Shara It's a lot of things.
18:07 paramat i'll attempt the 2nd. the 1st may not be possible but is minor
18:08 Shara Second is the main one for me. Would appreciate it if you could do a mod for that.
18:08 paramat exact replication is not possible but is not necessary
18:08 Shara Happy to help with it if I can as well.
18:09 Fixer https://github.com/minetest/minetest/pull/5270 can this be considered for merging? i think it has two approvals
18:10 Krock joined #minetest-dev
18:18 paramat oh yeah
18:18 paramat devs need to make sure the 'approvals' labels are set correctly
18:19 paramat as those show up in the thread list
18:19 paramat that can be merged
18:24 Krock ack
18:33 proller joined #minetest-dev
18:38 Krock paramat, what do you mean with "How can pitch replace yaw?"?
18:38 paramat :] what do you mean by the comment?
18:39 paramat 3rd person is a complete reversal of view
18:40 paramat yaw + 180deg pitch to -pitch
18:40 paramat seems good
18:41 paramat sorry i mean 3rd person front view
18:41 Krock but (180 * 100) % 360 is 0
18:41 Krock so, no actual change
18:42 paramat nope
18:42 Krock as long pitch is changed, the whole packet is sent anyway
18:42 paramat er, that actually means 180deg, the * 100 is just encoding
18:42 Krock oh!
18:42 Krock well then it's alright
18:43 paramat some values are * 100 for some reason for packets
18:43 paramat great
18:44 nerzhul joined #minetest-dev
18:44 Krock but previously this factor 100 wasn't in there
18:45 paramat see client.cpp line 936 onwards
18:45 paramat yaw was previously * 100
18:46 paramat so when adding extra yaw the * 100 is correct
18:46 paramat because it's addition not multiplication
18:47 Krock no, it's not. The calculating part is done within "writePlayerPos"
18:48 Krock there's where it multiplies by 100, not before.
18:48 paramat sorry wrong line number, code has changed
18:48 paramat https://github.com/minetest/minetes​t/blob/master/src/client.cpp#L1023
18:48 Krock nvm, I found it already :)
18:49 paramat so yaw * 100, then if 3rd person add another 180 * 100
18:49 Krock it should add just 180 instead of 18000
18:50 Krock if this should be a 1/2 turn
18:50 paramat nah, because result is (yaw + 180) * 100 which is correct encoding
18:50 paramat adding 180 would add 1.8 deg
18:52 Krock oh, we're talking past eachother
18:53 paramat oh you're talking about sendPlayerPos
18:53 Krock check out lines 946 to 949 in the pull's diff. there it's correct, but only there
18:53 Krock yes, right
18:53 paramat sorry, will look again
18:54 Krock hehe. I already wondered how this mess could even start :3
18:54 paramat hm you may be right here
18:55 paramat depends on whether myplayer->getYaw() returns * 100
18:55 Krock no, checked that before. it's not encoded
18:55 paramat ok sorry
18:56 Krock no problem. I'm glad we solved this little conversation problem :)
18:58 nerzhul sfan5 are you around please ?
19:01 Shara paramat: recipe and durability now adjusted for handholds.
19:02 paramat ok
19:06 Shara Not sure if there is anything else I can adjust, so down to MTG devs now.
19:46 octacian_ joined #minetest-dev
19:58 DI3HARD139 joined #minetest-dev
20:02 numzero1 joined #minetest-dev
20:07 red-001 nerzhul, you can check off "Add a function to generate minetest color codes" from #5394
20:07 ShadowBot https://github.com/minetest/minetest/issues/5394 -- [CSM] Second Roadmap
20:07 nerzhul done
20:11 numzero joined #minetest-dev
20:17 benrob0329-phone joined #minetest-dev
20:17 numzero1 joined #minetest-dev
20:18 nerzhul thanks for the tip red-001
20:19 nerzhul i hope at some point my clang-format PR will be merged, then i will work to a blacklist for it, permitting to block CI when codestyle is wrong in files not in blacklist, permitting coredev to focus less on codestyle
20:30 paramat joined #minetest-dev
20:33 nerzhul sofar, can you tell us if you approve #5436 ?
20:33 ShadowBot https://github.com/minetest/minetest/issues/5436 -- [CSM] Change command prefix to "." and add "help" command. by red-001
20:33 nerzhul it seems all smalljoker points are fixed
20:35 OldCoder Hm. I just woke up. Will the devs accept nerzhul solution with additional work or not?
20:42 sofar nerzhul: I didn't test the code or verify it works
20:42 sofar nerzhul: but the '.' I approve
20:45 nerzhul sofar, if you get some time, don't hesitate to review, it's mainly a code move from server side + little additions client side
20:45 nore hmm, why "." and not some other character, like "\"?
20:45 nerzhul \ is a pain to type
20:45 nore ";" then?
20:45 nore or even ","
20:46 nerzhul and . is common in many games, like World of Warcraft for example
20:46 nerzhul .toto is better than ,toto no ?
20:46 sofar '.' is already used widely
20:46 nore I mean, I use "." as the beginning of messages much more that "," or "/"
20:46 sofar it's also one key off /
20:46 est31 left #minetest-dev
20:46 nore sofar: "," is quite close as well :)
20:47 sofar twice as far on my keyboard
20:47 nerzhul nore, in azerty yes, not qwerty
20:47 nore hmm, maybe provide a way to use "." at the beginning of messages then?
20:47 nerzhul if not a command, it's sent to server
20:47 sofar /say .hello.
20:48 nerzhul red-001, am i right ?
20:48 nore nerzhul: no, it's also quite close, I have a qwerty keyboard
20:48 nore ah, in that case it's ok for me
20:48 nore as long as you don't call a command "..." :)
20:49 nerzhul this should be stupid, but client side and just annoy client :p
20:49 VanessaE I write "..." by itself pretty frequently
20:50 nerzhul type " ..." :p
20:50 sofar .password oops this ended up in global chat
20:50 VanessaE imho if you need a client-side command prefix, \ makes the most sense
20:50 VanessaE and users seem to reach for that key when they want / anyway half the time.
20:51 sofar \ is gonna drive me crazy
20:51 nore sofar: why so?
20:51 sofar it's the default escape character
20:51 sofar do\ you\ know\ how\ often\ I\ use\ it?
20:51 VanessaE I assume never, in in-game chat :P
20:52 VanessaE (at least not as an escape, anyway :) )
20:52 nore sofar: s/\\\ // :p
20:52 nerzhul yes \ is for escape, and as sofar said many games are using . and it works without crying
20:58 OldCoder I will be making a statement here with C55's permission
20:58 OldCoder It will be about 5 paragraphs. I can simply link it but as it includes a significant announcement I ask for permission to state it explicitly as well.
20:59 sofar if it's significant, feel free to verbatim post it in here
21:01 OldCoder Will do subsequent to review by C55 and others in the next half hour
21:01 OldCoder To be clear, the text will be quoting C55 and I have his approval
21:02 sfan5 a statement about what
21:02 sfan5 if its about the sneak thing, you can stop before you even begin #5443
21:02 ShadowBot https://github.com/minetest/minetest/issues/5443 -- Bring back sneak elevators/ladders by sfan5
21:03 OldCoder sfan5, your own remarks about forks were a little derisive. But, go on.
21:04 Puka joined #minetest-dev
21:04 sfan5 my remark about outdated mess was not about forks
21:04 OldCoder Very well
21:04 OldCoder I'd like to make a key point that C55 and I agree on. It can wait while your own issue and/or PR is discussed
21:05 OldCoder Shall I yield the floor to you for that, sfan5 ?
21:05 sfan5 say what you want to say
21:05 OldCoder It will depend on the sneak resolution though the key point is the same regardless
21:05 OldCoder I'll hold it for a day if you have a proposal
21:06 VanessaE OldCoder: dude.  out with it.
21:06 sofar don't do that, you can't have an open discussion if people withold their voice for some reason
21:06 OldCoder It needs to be reworded if there is a proposal on the table
21:06 OldCoder Very well
21:06 OldCoder C55 and others need to review it
21:06 OldCoder Excuse me while I attend to that
21:07 celeron55 this is kind of silly; even i don't know what the announcement is
21:07 OldCoder celeron55, PM to review it
21:08 OldCoder And I covered part of this in our PM about 5 hours ago. Have not had much sleep BTW.
21:08 * OldCoder notes that Old Coders need to snooze now and then
21:10 celeron55 well, i'd recommend sleeping over anything MT-related but that's just me, lol
21:11 OldCoder Hm. Well, I try to balance things. MT and sleep are both important.
21:11 * OldCoder is editing
21:15 sofar "do not pay attention to the man behind the curtains"
21:19 OldCoder Hm?
21:22 celeron55 is this about the fork or what?
21:22 celeron55 the fork by the group of people that really like sneak ladders
21:25 OldCoder celeron55, you will be linked in PM very shortly
21:25 OldCoder Wording this carefully because the goal is unify and not disrupt
21:25 VanessaE maybe mt_game should have a LBM that detects sneak elevators and replaces them with stacks of cobble and ladders :)
21:26 Fixer someone forks mt over a bug?
21:27 paramat well a fork is unnecessary, this is a dev version after all and we are putting effort into keeping people happy
21:27 paramat using a dev version there is a risk of stuff being a bit broken sometimes
21:34 OldCoder Fixer, fork or spoon is about a much more basic issue
21:34 OldCoder And it isn't easy to find the right words sometimes. Working on it.
21:35 OldCoder This is the issue that, in the end, "killed" Maciek. And we've lost others the same way.
21:35 OldCoder I've observed this for 5 years. BTW This Spring is my 5th anniversary.
21:37 red-001 <nerzhul> if not a command, it's sent to server No the client captures all client-sided commands, I could add a way to escape a command so it's sent to the server
21:38 red-001 e.g. you can use .. instead of . to send a message that starts with .
21:38 red-001 /s/can/could
21:42 red-001 if someone has time could they review #5444?
21:42 ShadowBot https://github.com/minetest/minetest/issues/5444 -- Add multiline support to colorize. by red-001
21:47 paramat well, another fork could be a good thing :]
21:47 celeron55 the most interesting thing about forks is finding out what name they ended up using
21:48 celeron55 so far there are minetest forks with interesting names and boring names
21:48 red-001 minetest-test
21:48 nerzhul the collisionmove porting to server side is annoying, or i should rewrite it in a simpler way ... mmm
21:53 paramat minepass
21:53 paramat btw freeminer has had no dev for 3 months
21:54 OldCoder freeminer is dead
21:54 OldCoder nerzhul, see IRC PM
21:54 celeron55 i guess they're happy they protected their work by changing the license so that nobody can reuse it
21:55 nerzhul already looked for just not commented
21:57 Puka joined #minetest-dev
21:58 paramat meyencandy? shinyminey?
21:59 red-001 we should document that american and not british english should be used for api functions since that seems to be the current precedent
21:59 nore OldCoder: what is the problem, actually?
21:59 OldCoder nore, PM thanks
22:00 sfan5 paramat: >I support the intention of this PR, need some time to review as i do not yet understand the code enough.
22:00 sfan5 if it detects a sneak ladders by its structure it sets m_sneak_ladder_detected=true
22:01 proller joined #minetest-dev
22:01 sfan5 and in the control handling code it allows you to jump if you press sneak and there's a sneak ladder there
22:01 sfan5 (even if you normally weren't allowed to jump)
22:01 sfan5 that's it
22:01 Warr1024 minetest > mineter > minet
22:02 sfan5 oh and it also limits y speed to >0 to prevent you from falling down again afterwards
22:03 proller joined #minetest-dev
22:09 OldCoder sfan5, see PM
22:10 sfan5 seen it
22:11 OldCoder Comment there if you wish. Press Reload for added comments by C55 at bottom.
22:38 paramat ok thanks
22:40 nerzhul i need an expert in movements to help me to finish #4981
22:40 ShadowBot https://github.com/minetest/minetest/issues/4981 -- [WIP] Damage cheat fix: server side by nerzhul
22:41 nerzhul i ported the collision checks server side when falling, this permits to disable TOSERVER_DAMAGE & prevent cheat, but the ported code, whereas it's nearly same as client doesn't work well everytime
22:41 octacian__ joined #minetest-dev
22:41 nerzhul sometimes collisions are not detected and sometimes multiple times
22:41 nerzhul if a coredev can help me to finish this we, we can nuke clients which doesn't take damages due to hacked clients
22:42 nerzhul the client algorithm is there: https://github.com/minetest/minetest/pull/4981/​files#diff-f78a0d3142f422b5d1fc990f616d75b5R212
22:42 nerzhul and server https://github.com/minetest/minetest/pull/4981/​files#diff-da1e28445fc732755d2b64683bd94ee4R972
22:43 celeron55 i'm going to guess it's a sync issue between the client and the server
22:43 celeron55 and the code works otherwise fine
22:44 nerzhul celeron55, yeah i think it's a sync problem too, i am locally and have thos problems, with internet between client and server this could be a little bit more complicated
22:44 nerzhul celeron55, i think it's due to speed attribute sync
22:45 nerzhul i have a death respawn damages, i think it's due to speed > 0 on respawn
22:49 nerzhul okay resetting speed when respawning fixes it
22:50 paramat about the sneak document: arrogant comments by 1 or 2 devs (by me for example) should not and do not represent all devs. my experience is that generally devs do listen to players and server admins a lot, and this is indeed happening on -project, effort is being made especially on the sneak issue
22:50 DI3HARD139 joined #minetest-dev
22:54 nerzhul celeron55, i don't really know how to fix the last remaining desync to prevent this non detected collision
22:56 celeron55 is it due to the client telling the server to skip the collision?
22:56 celeron55 like, saying the player is standing on ground with speed=0 before the server saw the player collided with the ground
22:57 celeron55 a similar thing can make a double collision
22:57 nerzhul my test are a little bit more simpler, i'm in fly, i disable it and fall, sometimes i don't take damages
22:57 nerzhul yes for the double collision i can make a timer check to prevent this to happen
22:58 nerzhul 0.5 sec timer collision check disable you prevent laggy situation with double collisions
22:58 nerzhul my main problem are not detected collisions...
22:59 celeron55 when the client sends a packet with a new speed value, compare it with the old speed value
22:59 celeron55 if there's a sudden stop, that's a collision 8)
23:00 nerzhul oh i see what you mean, client set speed 0,0,0 and server set this to zero whereas we are in movement
23:01 rubenwardy joined #minetest-dev
23:01 celeron55 there are probably lots of details that need to be taken into account in that case and it'll never be perfect
23:01 nerzhul yes, but if it detect 99.5% cases instead of 75% could be better :p
23:02 nerzhul in fact lag will benefit to collision check because will notify speed = 0 slowly
23:02 octacian_ joined #minetest-dev
23:04 celeron55 lag will make you take fall damage in cases where you barely manage to avoid it but the server doesn't know the fine details
23:04 nerzhul also yes... it will not be easy... hmmm
23:04 celeron55 that'll surely be annoying
23:04 celeron55 like barely landing in water
23:04 nerzhul how MC works, i don't think they let player send the damage like us
23:05 octacian joined #minetest-dev
23:06 rubenwardy MC also uses TCP for sockets, so they're not a good example about good networking
23:06 rubenwardy _however_ clientside prediction, server side reconciliation is good
23:06 celeron55 i think some fast fps games run their physics in a fixed frame rate and then the server can exactly check the client's actions by running the exact same simulation delayed by the maximum amount of lag
23:06 nerzhul TCP is not bad for reliable packets, but movements are not a good TCP packet
23:07 nerzhul celeron55, we send position every 0.1sec
23:07 nerzhul maybe i should do the collision check on same interval
23:07 nerzhul it should be more accurate
23:07 celeron55 that's far from exact
23:07 celeron55 i don't think MT can ever have anything exact related to this though
23:08 celeron55 nothing in MT is designed for exactness here
23:08 nerzhul oh i don't remember server sent the recommended_send_interval
23:08 nerzhul let me try to use this interval to see the difference
23:10 nerzhul same effect
23:11 nerzhul or maybe i should increase the speed from 1 to 5 % virtually server side
23:14 nerzhul celeron55, maybe have a direction of player recording a vector between two server steps + compensate with a speed 5% factor can compensate
23:15 nerzhul if i set speed * 1.1 i see less non collision cases, but it's far from being perfect
23:21 nerzhul celeron55, the send interval is in fact sync with server step... then we send the positions less often than step due to lag, this is the lag cause
23:31 Tmanyo joined #minetest-dev
23:43 Taoki joined #minetest-dev
23:49 OldCoder
23:49 OldCoder http://minetest.org/sneak.html
23:49 OldCoder If you have already seen this, press Reload. A number of comments by others have been added.
23:49 OldCoder http://minetest.org/sneak.html
23:49 OldCoder
23:50 OldCoder celeron55, rubenwardy, nore, sofar, others, thank you for your comments
23:51 OldCoder Zeno`, Fixer, Robby, twoelk, DI3HARD139, paramat, Billre, Calinou, others, feel free to weigh in
23:51 OldCoder
23:51 OldCoder http://minetest.org/sneak.html
23:51 OldCoder AFK for a while in a bit
23:51 OldCoder

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