Minetest logo

IRC log for #minetest-dev, 2015-12-29

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

All times shown according to UTC.

Time Nick Message
00:12 DogePony joined #minetest-dev
00:47 Warr1024_ joined #minetest-dev
00:52 sapier1 joined #minetest-dev
00:58 sapier1 left #minetest-dev
03:19 Player2 joined #minetest-dev
04:04 turtleman_ joined #minetest-dev
04:48 Soni joined #minetest-dev
04:51 Warr1024_ joined #minetest-dev
05:07 Terusthebird joined #minetest-dev
06:19 `PenguinDad` joined #minetest-dev
06:31 Hunterz joined #minetest-dev
07:29 Krock joined #minetest-dev
09:35 Darcidride joined #minetest-dev
09:35 Amaz joined #minetest-dev
09:38 * VanessaE pokes RealBadAngel
10:15 VanessaE celeron55: what's the word on your farmap code?
10:23 troller joined #minetest-dev
10:25 VanessaE (I note now that Fixer asked about this several days ago, but it bears repeating.  that code must not be allowed to die.)
10:25 RealBadAngel celeron55, you shouldnt close the PR when i agreed for WIP label and that i will look for better solution (faster)
10:27 RealBadAngel your point about using meshes is partially valid, there are cases that making a mesh runtime is way better than import stuff from outside
10:51 red-001 joined #minetest-dev
10:54 Calinou joined #minetest-dev
11:24 Zeno` joined #minetest-dev
11:36 red-001 #602  #667
11:37 VanessaE red-001: the bot doesn't work
11:37 red-001 https://github.com/minetest/minetest_game/pull/667
11:37 red-001 https://github.com/minetest/minetest_game/pull/602
11:37 red-001 I think all style issues are fixed
11:43 RealBadAngel hi Zeno`
11:44 Zeno` howdy :)
11:45 RealBadAngel whats going on? any news on ogre?
11:46 Zeno` nah, I kinda lost motivation :(
11:49 RealBadAngel did you get it to any working stage?
11:50 Zeno` it's "half" working
11:50 Zeno` there is still irrlicht fragments so both libs are used
11:51 Zeno` which is a bit insane
11:51 Zeno` (and the reason I never pushed)
11:52 RealBadAngel do you have it somwhere online?
11:56 Zeno` RealBadAngel, I'm on my laptop but I'll email you a link when I'm back on my main computer if you like
11:57 Zeno` it's not online atm
11:57 Zeno` kinda messy and hacky
11:57 Zeno` the types are the biggest problem
11:57 VanessaE messy and hacky is okay
11:58 VanessaE no project starts out perfectly polished.
12:01 RealBadAngel joined #minetest-dev
12:03 red-001 joined #minetest-dev
12:03 cmdskp joined #minetest-dev
12:44 Fixer joined #minetest-dev
12:49 DFeniks joined #minetest-dev
12:57 zupoman joined #minetest-dev
12:59 red-001 joined #minetest-dev
13:09 celeron55 joined #minetest-dev
13:15 red-001 https://github.com/minetest/minetest_game/pull/667
13:22 Amaz left #minetest-dev
13:22 Amaz joined #minetest-dev
13:26 cmdskp @red-001, for a TNT api, it would be nice to have the option of chain-reaction increasing blast radius if more TNT blocks are found (similar to the old Carbone game does, I believe)
13:26 celeron55 joined #minetest-dev
13:27 red-001 cmdskp
13:27 red-001 I have no idea how to code that
13:27 cmdskp Let me see if I can find the bit of Calinou's script that does it
13:29 cmdskp Inside the 'boom' function here: https://github.com/Calinou/carbone/blob/master/mods/tnt/init.lua
13:30 cmdskp It's rather nicely balanced, in that, the increase in radius decreases the bigger it gets, so it becomes rather wasteful of TNT eventually, but offers much fun stacking TNT for bigger booms!
13:31 red-001 Ok I will try and add that as an opition
13:32 cmdskp Thank you :) I know some players on a server I play on who love it (as do I)
13:35 red-001 It seems that a lot of the code in boom was moved to destroy
13:35 cmdskp I see
13:35 red-001 and expload
13:35 cmdskp I've not compared it
13:35 red-001 I might add it as an alt expload function
13:36 cmdskp seems reasonable
13:37 red-001 the thing is voxel manipulater is now used
13:37 red-001 and this uses minetest remove node
13:37 red-001 with is a lot slower
13:39 cmdskp let me have a more detailed look at the API version...
13:39 cmdskp one moment
13:40 red-001 line 167
13:40 zat1 joined #minetest-dev
13:43 cmdskp I can see what you mean about performance - I can also see that it would be possible to do a count of TNT found (similar to line 192 but cid==c_TNT), then if any found perform a secondary explode for the increased radius, but skipping the inner radius already done
13:43 Amaz joined #minetest-dev
13:43 red-001 ok I will try that
13:47 PenguinDad joined #minetest-dev
13:50 cmdskp with the TNT found count, after the loop, you can do a while loop to calculate the secondary radius using Calinou's carbone's if's radius adds, thus saving performance inside the loop to just checking for TNT and adding to a count
13:51 red-001 what would be a good name for the option?
13:51 cmdskp it also ties in nicely with checking if a max radius option is set and not doing the secondary blast if it's <= radius
13:51 cmdskp max_radius would be a good option
13:52 red-001 ?
13:52 CraigyDavi joined #minetest-dev
13:52 red-001 isn't this suppose to increase the blast radius?
13:53 red-001 if it finds more tnt
13:53 cmdskp yes, but you'll do that in a second function loop for the 2nd blast radius size afterwards, if max_radius>original radius
13:53 red-001 ok
13:54 red-001 how to deal with multipile types of tnt?
13:54 cmdskp ah, good point!
13:54 cmdskp use a factor of their blast radius to increase the secondary radius
13:55 red-001 check the blast radius of each and calc an average?
13:55 cmdskp you could do that too
13:56 gregorycu joined #minetest-dev
13:56 gregorycu Hey guys
13:58 cmdskp hey
13:58 red-001 hi
14:00 * red-001 is thinking about logic
14:00 red-001 well this will need a table
14:07 red-001 https://github.com/minetest/minetest_game/pull/602
14:09 red-001 Also github for windows finally can handle branches
14:09 red-001 It still doesn't know how to rebase
14:10 cmdskp another way to refactor into two functions, primary blast counts TNT only inside it, secondary blast would then only destroy and ignore TNT blocks inside the greater radius
14:11 cmdskp (it'd destroy the outer radius TNT blocks, sorry, I meant ignore them as in they wouldn't add to the radius)
14:11 red-001 yeah that was my idea at first
14:11 gregorycu Doesn't sounds like refactoring to me
14:11 red-001 It sounds less confussing
14:12 red-001 olny call the first function if this is enabled
14:12 cmdskp I was discussing it with someone in a PM - I just copy+pasted to here as it was relevant :)
14:12 cmdskp yup!
14:12 red-001 otherwise just call the second one
14:12 gregorycu You talking about explosions that set off other TNT?
14:13 cmdskp yes, as an option for a TNT api
14:13 gregorycu Isn't it a very simple recursive algo?
14:13 red-001 well TNT already get set off
14:13 cmdskp recusion is much slower though
14:14 gregorycu Slower than what exactly?
14:14 cmdskp and more costly performance wise - which was a concern :)
14:14 red-001 this option makes the blast area bigger if there is more TNT
14:14 cmdskp doing two separate blasts, inner blast and secondary larger outer blast (without clearing the inner blast area already done)
14:15 gregorycu Ahh ok
14:15 gregorycu This isn't chained explosions
14:15 gregorycu More like amplified explosions?
14:15 cmdskp as most people naturally stack TNT together in a block in the centre, it copes with most normal use cases, while not being true chaining
14:15 red-001 yes
14:15 red-001 thats a good name
14:15 cmdskp indeed
14:16 gregorycu Who are you guys anyway?
14:16 * red-001 is learning NTFS junction point back magic
14:16 * cmdskp is just me
14:17 gregorycu Never seen you both before
14:17 red-001 I don't go on this irc often
14:17 cmdskp that'd be reasonable, as I only found Minetest a few months ago and only used chat recently
14:17 red-001 and I usally call myself red1
14:20 cmdskp true chaining could be a possible option in the secondary loop, by triggering a burn on it if found, rather than remove.  Perhaps with a limit as to how many will be triggered (a simple count decremented)
14:20 cmdskp This would also nicely stagger the explosions, reducing server impact at once
14:21 red-001 that is already done
14:21 red-001 when tnt normally explodes it lights other tnt
14:22 cmdskp cool :)  just the thought occurred in musing over it
14:24 gregorycu If you had say 6 blocks of connected TNT, you could take the midpoint of those blocks
14:24 gregorycu And make the radius some sort of function on the number of connected blocks
14:24 red-001 good idea
14:25 gregorycu Sorry, I mean connected TNT
14:26 red-001 what's the itemstring for locked chest and door?
14:27 cmdskp is there as group for locked items yet?
14:27 cmdskp *a
14:27 cmdskp It would be helpful in this case...
14:27 red-001 yeah
14:31 red-001 How do priv get defined?
14:32 celeron55 joined #minetest-dev
14:35 red-001 Is it a function?
14:35 cmdskp priv meaning privileges?
14:35 red-001 yes
14:36 cmdskp minetest.get_player_privs(name)?
14:37 cmdskp and minetest.check_player_privs(player_or_name, ...)
14:37 cmdskp that seems to be what the lua_api.txt defines
14:37 red-001 no as in is it possible to create a new privilege?
14:37 Calinou joined #minetest-dev
14:38 cmdskp minetest.register_privilege(name, definition) perhaps?
14:42 red-001 well it didn't crash
14:42 red-001 so maybe
14:44 red-001 no
14:44 red-001 :(
14:44 cmdskp Defining a group 'indestructible' would be suitable for locked doors/chests, as 'immortal' is symantically expected of living entities rather than inanimate blocks
14:45 cmdskp shame
14:45 celeron55 joined #minetest-dev
14:47 cmdskp In "int ObjectRef::l_set_physics_override(lua_State *L)", I notice a section for setting physics_override marked: // old, non-table format
14:47 cmdskp is this still necessary to add to when adding a new physics override setting?
14:47 red-001 nvm it works
14:48 red-001 I didn't give it a string
14:48 cmdskp ah
14:48 red-001 I make that mistake too much
14:49 cmdskp 'tis easy to do
14:50 red-001 can minetest be set to try and use the same port when in singleplayer?
14:52 cmdskp you mean like running two instances server/client on your own PC, using localhost as the server name?
14:53 cmdskp if you need to test client/server privileges, that's the easy way
14:53 cmdskp and no one else can join from outside
14:53 red-001 I'm running one as singleplayer and then connecting an other client to it
14:54 cmdskp you'll need to run one as a server with the server address set to: localhost
14:54 cmdskp rather than singleplayer
14:55 celeron55 joined #minetest-dev
14:56 red-001 well minetest prints out the server address when starts
14:56 red-001 but the port number changes every time
14:58 cmdskp I'd imagine that's just to allow it to act internally like a combined server/client
14:58 cmdskp but it won't let other clients connect to a singleplayer instance
14:58 red-001 I thing is it does
14:59 cmdskp mmm curious
14:59 red-001 it's usefull
14:59 red-001 but in 4.13 RC It is a bit strange
15:00 red-001 if the singleplayer client pause in the esc menu
15:01 cmdskp the whole engine stops I imagine?
15:01 red-001 they becmoe frozen and phyics dosn't work on them
15:01 red-001 the other client can still place blocks
15:01 red-001 and lua stops running
15:01 cmdskp that's why running a localhost server behaves as expected =) and you can still play on the server instance too, as it's not set to headless mode
15:02 red-001 when singleplayer reusmes the blocks stay places
15:02 red-001 placed*
15:02 red-001 strange
15:02 cmdskp I imagine they're queued and resent/or collected and placed when the server unpauses
15:02 red-001 makes seens
15:03 red-001 sense*
15:05 gregorycu The actions are added the a queue on the server
15:05 red-001 nope still can't register privileges
15:05 gregorycu And processed once the game is unpaused
15:06 gregorycu Actually, probably added to an outgoing queue on the remote client
15:10 celeron55 joined #minetest-dev
15:17 Amaz joined #minetest-dev
15:22 behalebabo joined #minetest-dev
15:23 sapier joined #minetest-dev
15:23 sapier pushing https://github.com/minetest/minetest/pull/3479 in a few minutes as it's got two agreements and doesn't change anything if you don't use it
15:25 cmdskp that seems very useful
15:25 cmdskp I love the mini-me mesh item ^_^
15:29 sapier well that's just an example as I didn't want to create a new mesh for testing only ;-)
15:29 cmdskp yes, but it is a fun example to go with - along with the more standard ingot mesh starter ;)
15:29 sapier just waiting for the rebased pull to complete compile check then I'm gonna push it
15:29 Gael-de-Sailly joined #minetest-dev
15:30 cmdskp cool
15:30 sapier ingot isn't included
15:30 sapier ppl didn't like the hardcoded mesh so I removed it
15:31 sapier as it's not a big deal to build a mesh like that I let this do the games guys ;-)
15:31 Taoki joined #minetest-dev
15:31 celeron55 joined #minetest-dev
15:32 cmdskp "Ingots not included" - so like Yule toys...
15:32 red-001 https://github.com/minetest/minetest_game/pull/602
15:33 sapier Any graphical change in minetest is hard fighting and it's way more comfortable to let modders decide it by simple "usage"
15:33 cmdskp yep, I understand =)
15:34 red-001 I'm not sure did I put the registration of access(a new privilege) in the right place
15:34 sapier once everyone uses it decision is made ... and if not ... well shit happens ;-)
15:37 cmdskp "access" could be "access_locked" - to help explain what type of 'access' the privilege is?  It also reduces the chance of clashes with other privileges that might use 'access' for another reason?
15:40 red-001 well it has an explanation and access_locked would be longer then any other name for a privilege
15:42 celeron55 joined #minetest-dev
15:43 cmdskp seems like there's a history to 'access' for this too: https://github.com/minetest/minetest_game/issues/304
15:44 hmmmm joined #minetest-dev
15:47 red-001 I don't think you can convince the people that say it's useless by changing the name of the privilege
15:47 cmdskp no, I wasn't suggesting that - I just found out the history of a similar access feature, which backs up calling it as you intended
15:56 red-001 joined #minetest-dev
16:02 sapier https://github.com/minetest/minetest/pull/3467 pushing this one after rebase/style fixed pull build check is completed
16:04 celeron55 joined #minetest-dev
16:14 red-001 cmdskp
16:14 red-001 I think it would make most sense if TNT searches all the nodes directly connected to it to check if they are TNT
16:15 celeron55 joined #minetest-dev
16:16 cmdskp That would need to be limited to a certain range, otherwise people could build a long line of TNT or a maze of it for fun and set it off, causing a long chain of checks
16:17 red-001 yeah just check the blast radius
16:17 cmdskp A fill algorithm to find the nodes would be costly performance wise
16:17 red-001 true
16:18 kaadmy would it be faster to just check all nodes in the radius?
16:18 red-001 maybe check is there more then x% of TNT in an area?
16:18 kaadmy or do you want only connected tnt?
16:18 red-001 only connected
16:19 red-001 I'm not go at using vm
16:19 red-001 good*
16:19 kaadmy could find all tnt nodes, then check for connectivity
16:19 red-001 thats a good idea
16:20 est31 joined #minetest-dev
16:20 cmdskp seems a lot of extra processing - any benefits to checking connectivity versus just counting them in the blaast radius?
16:21 red-001 makes more irl sense?
16:21 red-001 I guess
16:22 red-001 not that minetest makes a lot of sense punching trees and all
16:22 est31 sapier, what are the two agreements you talk about?
16:22 est31 in terms of this pr https://github.com/minetest/minetest/pull/3479
16:22 est31 I only see one agreement in github: yours
16:22 cmdskp performance with TNT was a concern, naturally and if people don't know any different and get similar results in most cases...games are all about slight-of-hand compromises for performance =)
16:23 cmdskp still, you could test and see what the results are like speed-wise
16:25 celeron55 joined #minetest-dev
16:35 celeron55 joined #minetest-dev
16:38 red-001 can  obj:punch(puncher) be nil?
16:39 red-001 or is a puncher always needed?
16:40 cmdskp the problem is whether all mods cope with nil in "on_punch"
16:41 red-001 do any mods use it?
16:41 cmdskp yes, mobs
16:41 cmdskp or at least mobs_redo does that I know of
16:41 hmmmmm joined #minetest-dev
16:42 red-001 well all should
16:45 nolsen joined #minetest-dev
16:46 celeron55 joined #minetest-dev
16:46 cmdskp someone could argue, how can something be 'punched' without a 'puncher'?
16:46 red-001 no they don't deal with it
16:46 cmdskp symantically, it's unexpected behaviour
16:46 cmdskp not surprised
16:46 cmdskp I had the same problem a few weeks ago
16:48 cmdskp regretably, I didn't find a satisfying solution to offer
16:54 red-001 there really should be a on_damage callback
16:59 Darcidride_ joined #minetest-dev
17:01 younishd joined #minetest-dev
17:02 rubenwardy joined #minetest-dev
17:08 celeron55 joined #minetest-dev
17:18 celeron55 joined #minetest-dev
17:27 rubenwardy joined #minetest-dev
17:28 celeron55 joined #minetest-dev
18:03 celeron55 joined #minetest-dev
18:04 hmmmmm joined #minetest-dev
18:04 Darcidride joined #minetest-dev
18:05 Robby joined #minetest-dev
18:10 nolsen joined #minetest-dev
18:16 celeron55 joined #minetest-dev
18:22 cmdskp joined #minetest-dev
18:26 cmdskp Slippery nodes (with physics override 'slip' factor - e.g. could allow boots to grip) https://github.com/minetest/minetest/compare/master...cmdskp:master
18:27 cmdskp and some test values for default game: https://github.com/minetest/minetest_game/compare/master...cmdskp:master
18:29 cmdskp I need assistance on how to handle checking the protocol version in content_cao.cpp
18:54 sapier est31 no need to do so I already did it as I obviouly misinterpreted those agreements in there and I don't have any interest in fighting for this one
18:59 sofar cmdskp: looks simple enough, why the negative slippery value for jungletree?
19:00 sofar cmdskp: I'd have dirt remain the default value though, instead make wet soil slippery instead
19:01 cmdskp The negative value was a test to see the effect of increasing acceleration, it's quite amusing
19:03 kaadmy i want to see bunnyhopping in MT :)
19:05 cmdskp dirt should be?  but dirt_with_snow has slippery
19:08 cmdskp With small negative slippery values you can propel yourself at fast speeds, counter-intuitively, big negative values offer slower acceleration increase
19:10 cmdskp I'll let someone else consider upwards movement - my focus is sideways and next, slopes ;)
19:12 sofar dirt should not be slippery
19:12 sofar with snow yes
19:12 cmdskp It's not
19:12 sofar \wet soil yes
19:12 sofar ok
19:13 Calinou snow probably shouldn't be slippery either
19:13 Calinou only ice
19:13 cmdskp It's very slight, as in, you slow to a stop slightly slower
19:14 turtleman_ joined #minetest-dev
19:16 cmdskp But does make snow slightly risky footing, as you can go over the very edge, when near full speed (and not sneaking)
19:17 cmdskp The slippery values though as just for test, I'm happy to not include them - they were for people to test the feature easier :)
19:17 cmdskp *as=are
19:19 red-001 joined #minetest-dev
19:34 alket joined #minetest-dev
19:48 Robert_Zenz joined #minetest-dev
19:48 sofar I would certainly enable slipperyness if this gets merged
19:49 sofar let me know if you feel it needs test feedback, I certainly like the idea, and the implementation seems trivial
19:59 cmdskp It's always nice to get feedback :)
20:00 cmdskp I just need some help on how to code a test for the network protocol version and then I can offer it as a pull
20:01 neoascetic joined #minetest-dev
20:01 neoascetic Hi everyone
20:02 neoascetic Will #3143 and #2485 be merged someday?
20:02 neoascetic A lot of PRs here for ages...
20:03 red-001 the bot is dead
20:04 neoascetic https://github.com/minetest/minetest/pull/3143
20:04 neoascetic https://github.com/minetest/minetest/pull/3143
20:04 red-001 https://github.com/minetest/minetest/pull/2485
20:04 neoascetic whoops https://github.com/minetest/minetest/pull/2485
20:04 neoascetic yeah
20:05 red-001 most pr not from collaborators are like that
20:05 red-001 PR's*
20:06 neoascetic And this is bad for community
20:06 red-001 and all open source projects have it
20:07 red-001 even Wikipedia has it
20:07 red-001 in a diffrent way
20:08 neoascetic Okay, is there a way to solve this? Maybe some bot that pings on too old PRs?
20:08 neoascetic Or do a global PRs/issues review once per month?
20:09 neoascetic https://github.com/minetest/minetest/pull/817
20:09 red-001 yeah as if the sort of community that this happens in will agree to a global review once a month
20:09 neoascetic the oldest open PR, hi from 2013, with some life within
20:10 red-001 and an ancient meme
20:10 red-001 https://github.com/minetest/minetest_game/pull/667 Is that ok? I think all style issues are fixed
20:11 red-001 @shadowninja
20:11 red-001 nvm he is offline
20:20 Drangue joined #minetest-dev
20:40 Hunterz joined #minetest-dev
20:51 neoascetic joined #minetest-dev
21:20 CraigyDavi joined #minetest-dev
21:31 GhostDoge joined #minetest-dev
21:35 red-001 does cmake care what ever a / or \ is used?
21:36 red-001 whatever*
21:38 Krock it converted them all to / in cmake-gui here
21:39 Krock but it should be able to handle it
21:39 red-001 ok
21:41 red-001 can cmake have variables in variables?
21:42 Krock when you use a variable to define another, the current value will be taken
21:42 red-001 thats ok for me
21:43 red-001 so do you just do .8/bin/ZlibDllRelease/zlibwapi.lib
21:43 red-001 ZLIB_LIBRARIES:FILEPATH=$MYPATH/zlib-1.2.8/bin/ZlibDllRelease/zlibwapi.lib
21:43 red-001 and define MYPATH somewhere else?
21:57 Krock CMake variables needs brackets around it, like this: ${MYPATH}
22:00 red-001 ok
22:01 red-001 well I already fixed the problem with a powershell script
22:06 red-001 what type of variable does MYPATH have to be?
22:06 red-001 static?
22:08 Krock anyway, why do you need that? Using cmake-gui is much easier
22:09 red-001 I'm making a file that already has all the paths in it
22:09 red-001 if I move the location of the folder I build minetest in
22:10 red-001 I just have to change one variable
22:10 Player2 joined #minetest-dev
22:12 Krock ah.. portability
22:13 red-001 and a reason to learn more about cmake and powershell
22:23 Fixer !tell paramat permanent fire has no sound, is this right>
22:23 ShadowBot Fixer: Error: Spurious ">".  You may want to quote your arguments with double quotes in order to prevent extra brackets from being evaluated as nested commands.
22:24 Fixer !tell paramat permanent fire has no sound, is this right?
22:24 ShadowBot Fixer: O.K.
22:38 alket joined #minetest-dev
22:41 Robby joined #minetest-dev
22:56 sapier https://github.com/minetest/minetest/pull/3469 ... obviously has three agreements ... hope I'm not missinterpreting it again ... pushing in a few minutes
23:08 red-001 sapier can you take a look at my pull requests?
23:08 red-001 https://github.com/minetest/minetest_game/pull/667
23:08 red-001 https://github.com/minetest/minetest_game/pull/602
23:08 sapier I usually don't have a lot to do with minetest_game
23:08 red-001 ok
23:09 red-001 who should I ask?
23:09 sapier I can have a look at it but the "arts" department isn't mine
23:09 red-001 well it's not art
23:10 sapier well I count "game type" to arts ... not exactly correct yes
23:11 Robert_Zenz left #minetest-dev
23:14 RealBadAngel red-001, ask paramat, hes an artist here, he knows everything bout it and got best taste
23:14 est31 joined #minetest-dev
23:33 Robby joined #minetest-dev
23:43 sapier can someone explain to me what an AreaStore does do and how to use it? documentation is kindof useless ;-)
23:43 est31 joined #minetest-dev
23:45 est31 sapier, AreaStore stores areas efficiently
23:45 est31 you can add areas to it, and remove them
23:45 sofar easy way to save/restore areas?
23:45 est31 yes.
23:45 sapier no more?
23:45 est31 lookup is optimized
23:45 est31 you can pass a point and get the list of areas it is in
23:46 est31 and do the same for an area too
23:46 sapier true but I don't actually see a lot of uses except of maybe backup and rollback
23:46 est31 that's where the optimisation is
23:46 cmdskp joined #minetest-dev
23:46 est31 protection mod for example
23:46 est31 SN's areas mod used alot of computation
23:46 sapier I see ... very special thing
23:46 est31 every server step, it gets the position of every player, then checks which areas the player is in
23:47 est31 for that, it traversed all areas, previously
23:47 est31 yes, currently the use cases are bit limited.
23:47 sapier I was hoping this one could be used for some sort of mapgen or map checks in async lua threads but if you can only add and remove it in total I don't think it can be used that way
23:47 est31 but later on we can use it for splitting up the map lock.
23:48 sapier well it's certainly usefull in that area but I don't see the direct connection to it
23:50 est31 well if you want to split up the map lock, e.g. where code can request lock on a specific area of mapblocks, you can either implement it manually, with a large list of all mapblocks and who owns them, but AreaStore can use R* trees, so if you add the library its faster
23:53 sapier Actually I did think more about some sort of transaction system for map but I didn't think this to the end by now ... especially as I never did dig to the deep in map
23:54 est31 wanting to push this fix https://github.com/est31/minetest/commit/a142e4f4b2256e7c4a16d14aecde6e04be33457c
23:54 cmdskp who would be best to ask about the network protocol checks?
23:54 est31 if nobody objects, and there is no agreement, I'll push it in 10 minutes
23:54 est31 clearly trivial
23:54 est31 cmdskp, me
23:54 sapier go ahead
23:55 cmdskp ty est31, I need to find a way to check which protocol version in content_cao.cpp, marked TODO here: https://github.com/minetest/minetest/compare/master...cmdskp:master
23:56 cmdskp I don't see an easy way to get at the protocol version and I'm unsure of the path to take through the program structure
23:57 cmdskp basically, I've added a float into the physics_override for 'slip' factor on slippery nodes and need to only read it when on the latest protocol

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