| Time |
Nick |
Message |
| 00:13 |
|
p_gimeno left #minetest |
| 00:21 |
|
cafee__ joined #minetest |
| 00:37 |
|
Fixer joined #minetest |
| 00:39 |
|
hecks joined #minetest |
| 00:40 |
hecks |
So, by fiddling with emerge settings, I finally managed to make singleplayer run somewhat reliably. |
| 00:40 |
entuland |
why, what was the issue with it? |
| 00:41 |
hecks |
Well first of all, there's the singleplayer problem; you run both the client and the server, and each is nontrivial. |
| 00:41 |
hecks |
Most of the server's workload is in emerge, the client's is in meshing. |
| 00:41 |
hecks |
and possibly rendering, if your viewing range is far enough |
| 00:42 |
hecks |
Now, default settings for emerge are very bad for this, and the behavior is quite counter-intuitive. |
| 00:42 |
rubenwardy |
you might as well consider singleplayer as 4 parts really |
| 00:42 |
rubenwardy |
meshgen <-> client <-> server <-> emerge |
| 00:42 |
entuland |
uhm... it must be that I've got a decent rig and pretty much always play singleplayer, so I can't really compare |
| 00:43 |
rubenwardy |
there may even be send threads, which makes it |
| 00:43 |
hecks |
Defaults may work for some, entuland |
| 00:43 |
rubenwardy |
meshgen <-> client <-> send/recv <-> send/rec <-> server <-> emerge |
| 00:43 |
hecks |
Yes, the thing is, client physics and rendering is negligible, and gameplay code (most of lua) on the server just as much |
| 00:44 |
hecks |
The big problem is though, minetest fails to prioritize that "light but critical" code while juggling it with the heavy voxel processing. |
| 00:44 |
hecks |
there's also liquids and abms but I'll dump that with "server voxel stuff" |
| 00:44 |
rubenwardy |
https://github.com/MT-CTF/capturetheflag/blob/master/build.sh |
| 00:44 |
rubenwardy |
oops |
| 00:45 |
hecks |
My earlier issues were that the server side gameplay code was stalling; often for ridiculous amounts of time |
| 00:45 |
hecks |
2 minutes without a decent globalstep were not unheard of! |
| 00:45 |
rubenwardy |
oh wow |
| 00:45 |
rubenwardy |
sounds like a faulty mod maybe |
| 00:46 |
rubenwardy |
or I/O |
| 00:46 |
hecks |
Your mom's faulty, I run my own subgame :) |
| 00:46 |
hecks |
and I/O was ruled out by running it all in ramdisks |
| 00:46 |
rubenwardy |
lol |
| 00:46 |
hecks |
The main culprit is emerge... default settings do not fit this "profile" well |
| 00:46 |
hecks |
My current "nice" emerge settings are: 1 thread, max queue 4 (!) and the rest set to 0 (auto I presume) |
| 00:47 |
hecks |
The engine generates the world in nice 4-mapblock bites and never blocks the rest of the game for too long, I haven't gotten almost any "lag detected" messages in my game since. |
| 00:48 |
hecks |
A tiny queue does not slow the mapgen, to the contrary. And using multiple emerge threads is a little counter-intuitive; most multi-CPU users will hike that hoping to speed up emerge. |
| 00:49 |
hecks |
But all that does is slightly increase the "noise sampling" part of map generation, at the cost of dumping bigger chunks of work onto the "joining" code! |
| 00:49 |
hecks |
slightly speed up, I meant |
| 00:50 |
hecks |
Hence 3-4 emerge threads with default queue settings will dump so much onto the main thread that the "lua game" will be starved of CPU cycles, despite needing next to none. |
| 00:50 |
entuland |
speaking of which... "/!\ You are using old player file backend. This backend is deprecated and will be removed in next release /!\" |
| 00:50 |
rubenwardy |
how is map loading? |
| 00:50 |
rubenwardy |
it won't be removed |
| 00:50 |
rubenwardy |
probably |
| 00:50 |
entuland |
I have noticed that message recently, but it disappeared in the last runs |
| 00:50 |
rubenwardy |
it's also been a few releases since that |
| 00:50 |
hecks |
Do you mean performance with my current settings? Basically the same it was with threads. |
| 00:51 |
entuland |
I did nothing (didn't update the game or anything) is that tied to mods / servers? |
| 00:51 |
hecks |
The only noticeable difference is that the map is loaded in those concise "strips" in the distance, as opposed to semi-random mapblocks. |
| 00:52 |
hecks |
Performance per vision range seems comparable as well, with the difference that it's now "only" stuttering, as opposed to filling up the queue and choking hard for over half a minute. |
| 00:53 |
hecks |
Lua on_generated calls throw extra problems into this, as they are single threaded. So the "dumping tons of data on the main thread at once" issue is magnified with those. |
| 00:53 |
hecks |
On the other hand, doing it one or a few mapblocks at a time lets the CPU allocate time more evenly. |
| 00:55 |
hecks |
So now I'm running this subgame at 60 frames server step (!) without a lick of issues. Pretty good. |
| 00:56 |
hecks |
But it shouldn't have taken me almost a year of fiddling with settings to find this out. Not to mention the computing knowledge required to figure this out. |
| 00:57 |
hecks |
Maybe the defaults need some reconsidering, and special considerations for singleplayer. |
| 01:01 |
hecks |
Even when running something as heavy as ffmpeg in the background, the server doesn't choke at all. |
| 01:04 |
|
NathanS21 joined #minetest |
| 01:06 |
hecks |
https://a.uguu.se/eSoRdvhEJyLO_smooth.webm |
| 01:07 |
hecks |
the lag message at the beginning is from alt tabbing, it's not actual lag |
| 01:08 |
entuland |
uh? that's minetest? |
| 01:08 |
hecks |
Yes it is, you can tell by the nodes ;^) |
| 01:08 |
entuland |
didn't know we had such fluently animated avatars :P |
| 01:08 |
hecks |
"we" don't |
| 01:09 |
entuland |
eheheheh |
| 01:10 |
hecks |
It also saddens me every time someone comments on the model and animation but doesn't notice the perfectly implemented action combat... |
| 01:10 |
hecks |
*that* was much harder, you know |
| 01:10 |
entuland |
oh well I can't comment on the action combat, I just seen the weapon swung once or twice |
| 01:10 |
hecks |
...with the player stepping forward, animated |
| 01:10 |
hecks |
Sorry, I'll record a better example :) |
| 01:11 |
hecks |
in a less eye-busting biome |
| 01:11 |
entuland |
well that was included in "such fluently animated avatars" :) |
| 01:11 |
hecks |
Animations are one thing, but out of the box, minetest only supports "jerk a pickaxe around while WASD-ing" |
| 01:11 |
hecks |
Moving actors using animation, this is harder |
| 01:12 |
entuland |
I take that this is stuff you're working on (and yes, that's impressively fluent in all the movements) |
| 01:12 |
hecks |
Yes, it's mine. |
| 01:12 |
hecks |
Mine Test. |
| 01:13 |
entuland |
:) |
| 01:13 |
entuland |
the effect on the hair when falling down (or simply after a jump) is very very nice as well |
| 01:13 |
|
Lia joined #minetest |
| 01:14 |
hecks |
It's just part of the animation. |
| 01:14 |
hecks |
It could be more dynamic IF entity attachment and bone position shenanigans weren't broken... |
| 01:15 |
entuland |
I take that you're "just" making it to work within minetest, but you didn't model / rig / animate any of that, correct? |
| 01:16 |
hecks |
Guess again. https://a.uguu.se/mo5rUPVy4UdD_mibi.png |
| 01:16 |
entuland |
oh nice |
| 01:17 |
entuland |
well, you mislead me brushing that off with "it's just part of the animation" - I took you were waiving credit about that! |
| 01:18 |
hecks |
I can't blame you, because buying assets and lacing them with your bad code is how you become an indie star these days. |
| 01:18 |
entuland |
great job so far, definitely impressive |
| 01:21 |
|
Cornelia joined #minetest |
| 01:21 |
entuland |
testing out your emerge settings |
| 01:22 |
hecks |
I imagine that for an online server, the optimal settings would be different because you *do* need to handle several different people's FOVs |
| 01:22 |
|
ANAND joined #minetest |
| 01:22 |
hecks |
however, "easy" generation like this is good enough for one player |
| 01:23 |
hecks |
I'm treating it as more of a proof that minetest isn't utterly, horribly, terribly borked like I assumed before today |
| 01:23 |
rubenwardy |
any chance of this being made into standalone mods? |
| 01:23 |
rubenwardy |
more as an example than actually useful |
| 01:23 |
hecks |
Standalone mods? |
| 01:23 |
hecks |
You mean, actually slicing out bits of my code for you human vermin to use? How dare you. |
| 01:24 |
rubenwardy |
lol |
| 01:24 |
hecks |
It wouldn't even work... you have no idea how much shit I piled up on top of minetest and lua |
| 01:24 |
rubenwardy |
will you be sharing the code for it anyway? |
| 01:24 |
entuland |
uhm... I have three different limits for emerge (absolute, disk, to generate) set to 256, 32, 32... you set them to 4, 0, 0 ? |
| 01:24 |
hecks |
It simply isn't extractable or compatible with minetest_game's way of doing things. |
| 01:24 |
rubenwardy |
I don't care about minetest game, really |
| 01:24 |
hecks |
As for sharing, I will gladly part with the source code and assets for a modest price of five hundred thousand US dollars. :) |
| 01:24 |
rubenwardy |
heh |
| 01:24 |
entuland |
eheheheh |
| 01:24 |
hecks |
Plus tax. |
| 01:25 |
rubenwardy |
I also don't care about the assets |
| 01:25 |
rubenwardy |
and I'll happily part with the source code for minetest for a modest price of 0$ |
| 01:25 |
hecks |
There's also the long way; Download a modern game engine, spend 5 years understanding how a game is actually supposed to work, reimplement it in lua... |
| 01:26 |
hecks |
The $500k isn't for mibi, it's internet's late payments on my other projects :) |
| 01:26 |
hecks |
If not, oh well, I'll finish mibi and then DBAN it away on stream while you watch. |
| 01:29 |
rubenwardy |
You could have just said "no, sorry. I'd like to release this as a commercial product some day" |
| 01:29 |
hecks |
I don't even want that, I just don't care. |
| 01:36 |
entuland |
weird... flying around the world I found a large dungeon made of cobblestone and mossy cobblestone _completely_ above the ground (barely touching the top of a hill) --- is it just me ignorant about how vanilla V7 works? |
| 01:36 |
entuland |
no mods that could possibly do that loaded (only my own mods which don't mess with map generation) |
| 01:36 |
hecks |
As long as there's one node connecting it to the ground, it'll "work". |
| 01:36 |
entuland |
oh |
| 01:36 |
hecks |
Yes, this does happen in vanilla. |
| 01:36 |
entuland |
so it's normal, thanks |
| 01:37 |
rubenwardy |
This is a feature apparently |
| 01:37 |
hecks |
It does make for interesting structures. |
| 01:37 |
hecks |
I assume it just picks one "entrance" node and plots whatever it wants from it. |
| 01:38 |
entuland |
yep, very nice |
| 01:38 |
hecks |
Changing the ruins node to something more interesting than cobble makes it look better. |
| 01:40 |
hecks |
Now I have a question, does the heatmap influence height in v7? I could swear I keep seeing the same kinds of features in the same biomes... |
| 01:42 |
entuland |
uhm... where would I be changing that setting, hecks? the interface I used to mess with emerge doesn't show any hit for "ruin" |
| 01:43 |
entuland |
and what node would you advise? |
| 01:43 |
hecks |
You don't, this is what mods can do. |
| 01:43 |
entuland |
ah |
| 01:43 |
entuland |
I thought there was some settings file not covered by that interface |
| 01:43 |
hecks |
And I advise original content :) https://a.uguu.se/LmJiags2rbpj_techruins.png |
| 01:43 |
ANAND |
There's even a setting for floating dungeons |
| 01:43 |
entuland |
eheheh |
| 01:44 |
ANAND |
Not sure why one would want that though :) |
| 01:45 |
entuland |
well, not everybody can be well-versed in creating assets (I suck at designing stuff) |
| 01:45 |
entuland |
:P |
| 01:45 |
hecks |
I'm pleased to say that optimizing emerge made dynamic lights much more viable. |
| 01:45 |
hecks |
Goodbye, torch pollution. |
| 01:58 |
|
Ruslan1 joined #minetest |
| 02:17 |
hecks |
Just observed something; even when trying to make the mapgen completely serial (1 0 0 1), it still dumps mapblocks in slices of 4 |
| 02:18 |
|
NathanS21 joined #minetest |
| 02:22 |
hecks |
emerge 4 0 0 1, mesher delay 1; the game now feels extremely snappy, even the input feels better - but map generation became very lazy |
| 02:22 |
hecks |
the mesher basically refuses to mesh blocks that aren't in my active range |
| 02:23 |
hecks |
Turning off the mesher delay brings things back to normal, maybe the mesher delay isn't working as intended? |
| 02:24 |
hecks |
It really feels like an all-or-nothing setting |
| 02:25 |
hecks |
A mesher delay of 50ms (the maximum) feels a little lazier |
| 02:27 |
hecks |
Trying mesher with a 512MB cache; it's choppy. |
| 02:28 |
hecks |
Disabled cache entirely, it's smooth again. |
| 02:35 |
hecks |
Finally, allocating an extra emerge thread did not make map generation much faster, but it sure added some stutter. |
| 02:36 |
hecks |
rubenwardy if you're still around; would this be worth documenting? |
| 02:38 |
hecks |
that is, a warning for the num_emerge_threads setting, that it may negatively affect performance in singleplayer and/or with lua mapgens |
| 02:47 |
|
androidxl joined #minetest |
| 02:57 |
|
paramat joined #minetest |
| 02:59 |
paramat |
hecks "does the heatmap influence height in v7?" no, not in any mapgen either |
| 02:59 |
hecks |
okay then |
| 03:00 |
paramat |
entuland the setting 'protruding dungeons' can remove the er, protruding dungeons |
| 03:00 |
hecks |
Could've sworn I've seen more cliffs and huge pits in hot biomes but if it's bollocks, it's bollocks |
| 03:01 |
paramat |
it's bollocks |
| 03:01 |
entuland |
in the normal GUI settings page, paramat? |
| 03:02 |
hecks |
search "Projecting dungeons" |
| 03:03 |
paramat |
https://github.com/minetest/minetest/blob/9854340c0b21d01a94ba7ef526c775e59e37b898/minetest.conf.example#L1656 |
| 03:03 |
paramat |
advanced settings or a line added to .conf |
| 03:04 |
paramat |
yes, 'projecting' not 'protruding' |
| 03:05 |
paramat |
set to false all dungeons will be sliced off at terrain surface level |
| 03:05 |
paramat |
instead of 'most of them' |
| 03:07 |
paramat |
setting might be 5.0.0 only |
| 03:09 |
hecks |
That is correct, 0.4.16 does not show any dungeon related settings (or even dung related) |
| 03:09 |
paramat |
"a warning for the num_emerge_threads setting" aha i was about to rewrite that anyway |
| 03:10 |
paramat |
as its currently too promising |
| 03:10 |
paramat |
could you open an issue with advice on new wording? |
| 03:13 |
paramat |
see https://github.com/minetest/minetest/issues/4879#issue-194766354 what 'Rogier says' |
| 03:15 |
|
puzzlecube1 joined #minetest |
| 03:23 |
paramat |
an issue isn't essential if you don't have time, i can rewrite the docs based on what you've written here in IRC |
| 03:33 |
|
MinetestSam joined #minetest |
| 03:40 |
hecks |
ctf full of cheaters as usual |
| 03:46 |
hecks |
rubenwardy, fluffy_puppy is hacking on your CTF |
| 03:57 |
|
illwieckz joined #minetest |
| 04:27 |
Ruslan1 |
Who is hacker |
| 04:31 |
Ruslan1 |
hecks: who did hack on ctf |
| 04:31 |
hecks |
fluffy_puppy was |
| 04:33 |
Ruslan1 |
hecks: how I stop people who did hack in server |
| 04:34 |
hecks |
Burn their house, beat their mother |
| 04:34 |
Ruslan1 |
Nice one |
| 04:55 |
|
jluc joined #minetest |
| 05:05 |
|
blaise joined #minetest |
| 05:12 |
rdococ |
I hope you meant beat her in a game of CTF |
| 05:13 |
ANAND |
hecks: Using /report is a more efficient alternative ;) |
| 05:13 |
hecks |
there is one? |
| 05:13 |
ANAND |
Uhh... yes |
| 05:14 |
ANAND |
/report name reason |
| 05:14 |
ANAND |
Even moderators on IRC are notified |
| 05:21 |
hecks |
oh no it crashed |
| 05:21 |
hecks |
maaaaaaaaan |
| 05:59 |
rdococ |
Sneak peek...! https://i.imgur.com/6US1a0J.png |
| 06:19 |
Xiong |
Bad technic chest formspec: https://github.com/minetest-mods/technic/issues/478 |
| 06:34 |
|
YuGiOhJCJ joined #minetest |
| 06:54 |
|
paramat joined #minetest |
| 07:23 |
|
SanskritFritz joined #minetest |
| 07:29 |
|
blaise joined #minetest |
| 07:42 |
|
CWz joined #minetest |
| 08:02 |
hecks |
pr0 ctf strats https://a.uguu.se/5fcXJINkhCxC_screenshot_20190106_074706.png |
| 08:03 |
hecks |
This will become meta in a couple days just like bridges on the ice map |
| 08:06 |
|
MinetestSam joined #minetest |
| 08:08 |
ANAND |
Haha, I saw that as well |
| 08:10 |
hecks |
haven't lost a game with this yet I think |
| 08:12 |
ANAND |
It's very weak though, and can be destroyed from a distance using guns |
| 08:12 |
hecks |
Oh sure, I'm talking about the pit |
| 08:12 |
hecks |
The glass was for style because the match was going on forever |
| 08:13 |
hecks |
Then again even having no walls would be good; this strat is about vision |
| 08:33 |
MinetestBot |
[git] translations minetest.net -> minetest/minetest: Run updatepo.sh 3a9fe2b https://git.io/fhstB (2019-01-06T08:33:11Z) |
| 08:33 |
MinetestBot |
[git] translationsminetest.net -> minetest/minetest: Update translations from Weblate 893b5d5 https://git.io/fhstR (2019-01-06T08:32:00Z) |
| 08:56 |
|
cautiouspotato joined #minetest |
| 09:01 |
|
hecks joined #minetest |
| 09:24 |
MinetestBot |
[git] rubenwardy -> minetest/minetest: Deprecate modpack.txt and use modpack.conf instead (#7892) 70bf343 https://git.io/fhsqX (2019-01-06T09:23:35Z) |
| 09:26 |
MinetestBot |
[git] SmallJoker -> minetest/minetest: Fix various bugs (Anticheat, Lua helpers) (#8013) a122ba0 https://git.io/fhsq1 (2019-01-06T09:24:44Z) |
| 09:32 |
|
Tux[Qyou] joined #minetest |
| 09:32 |
|
Krock joined #minetest |
| 10:46 |
|
FFleder joined #minetest |
| 10:47 |
FFleder |
Good morning, is there a List with Bricks/ Walls containing the strength of the Wall against Pickaxe / Intruders? |
| 10:56 |
Krock |
obsidian |
| 10:56 |
Krock |
I don't think so |
| 10:59 |
oiaohm |
FFleder: fairly much once you have a steel pick anything is mineable other than bedrock. This is why protection mods exist. |
| 10:59 |
FFleder |
Obsidian Brick? I'll test this, thanks |
| 10:59 |
oiaohm |
FFleder: steel pick will mine Obsidian |
| 11:00 |
FFleder |
I try to fight against the Server lag - Some Guys break into our Area through protected Walls. Glas is a bad Idea :-) |
| 11:01 |
FFleder |
Actually i replaced the Walls with Planks - this help a lot |
| 11:01 |
oiaohm |
FFleder: https://forum.minetest.net/viewtopic.php?f=53&t=17815&sid=c492cf85cd1d6b92aa29d10a2167a203 If server is using a protection mod. |
| 11:01 |
oiaohm |
Players without permission cannot even burn or damage their way in. |
| 11:01 |
oiaohm |
Material does not matter with protection mods. |
| 11:02 |
FFleder |
But the Server lag seems to be my major Problem - they break in faster than the Server realize it's protected |
| 11:03 |
Calinou |
you can use unbreakable nodes from Map Tools to make this more difficult |
| 11:03 |
Calinou |
in this case, only a cheater could bypass it |
| 11:03 |
oiaohm |
FFleder: read that mod. Players on servers with that mod require a client side mod. If they break into somewhere it can be instant kick and inventorary delete and damage reversed. |
| 11:04 |
FFleder |
Yep, I'm on the page |
| 11:04 |
FFleder |
thx |
| 11:05 |
FFleder |
describes exactly my Problem |
| 11:10 |
|
Void7 joined #minetest |
| 11:22 |
|
kaeptmblaubaer joined #minetest |
| 11:26 |
|
Markow joined #minetest |
| 11:32 |
|
ensonic joined #minetest |
| 11:46 |
|
riff-IRC joined #minetest |
| 11:47 |
|
YuGiOhJCJ joined #minetest |
| 11:58 |
|
cafee__ joined #minetest |
| 12:04 |
|
Fixer joined #minetest |
| 12:06 |
|
sonicpp joined #minetest |
| 12:11 |
|
cafee__ joined #minetest |
| 12:15 |
|
cafee_ joined #minetest |
| 12:25 |
|
illwieckz joined #minetest |
| 13:02 |
|
cafee__ joined #minetest |
| 13:19 |
|
cafee__ joined #minetest |
| 13:27 |
MinetestBot |
[git] rubenwardy -> minetest/minetest: Content store: Fix update button not working f5de187 https://git.io/fhsG5 (2019-01-06T13:18:40Z) |
| 13:34 |
|
Krock joined #minetest |
| 13:38 |
|
Unarelith joined #minetest |
| 13:43 |
|
Gael-de-Sailly joined #minetest |
| 14:17 |
|
SanskritFritz joined #minetest |
| 14:23 |
|
cafee_ joined #minetest |
| 14:32 |
|
Wuzzy joined #minetest |
| 14:37 |
|
SanskritFritz_ joined #minetest |
| 15:24 |
|
entuland joined #minetest |
| 15:30 |
|
IhrFussel joined #minetest |
| 15:31 |
|
cafee__ joined #minetest |
| 15:32 |
IhrFussel |
I think some of you don't realize how extremely EASY it is to accidentally dig a protected zone while you stand near/in front of it and don't realize (cause your HUD shows no protection a node in front) ... you are crazy talking about kicking/deleting inventory |
| 15:33 |
IhrFussel |
I don't think many would want to play on such a server... an alternative would be what tenplus1 does in his protector_redo mod ... turning the player around when trying to dig a protected node...that gets annoying REAL fast |
| 15:40 |
FFleder |
:-) N1ce Idea |
| 15:41 |
Krock |
IhrFussel: then revoke interact for 20s if they violate 3 nodes within like 6 seconds |
| 15:42 |
IhrFussel |
That can work too but requires more complex code...likely a globalstep |
| 15:43 |
IhrFussel |
I once had it on my server so that it HURTS you to dig a protected node... but then the server spawn was FULL of bones within a short time |
| 15:43 |
IhrFussel |
So that is no solution |
| 15:45 |
IhrFussel |
Not to mention if you have lots of auth data on your server and many players need to lose/receive their interact...then your server could lag quite a bit |
| 15:46 |
IhrFussel |
Updating privs always triggers a auth save AFAIK ... it hurts especially if you still use the old files auth |
| 15:46 |
Krock |
then let bones appear starting from 10m around spawn |
| 15:48 |
IhrFussel |
10m is not enough... and there are several travelnets at my spawn...so each of the target locations might be full of bones after some time... no thanks |
| 15:48 |
|
scr267 joined #minetest |
| 15:49 |
IhrFussel |
tenplus1's idea adds almost no additional server load...just (one?) packet that tells the client to move the camera |
| 15:50 |
Krock |
well, for bone hunters it's very good to have it like this |
| 15:51 |
Krock |
free stuff |
| 15:55 |
MinetestBot |
[git] rubenwardy -> minetest/minetest: Content store: Disable more details dialog for now (#8060) 6169eca https://git.io/fhsWP (2019-01-06T15:54:22Z) |
| 16:15 |
|
Hawk777 joined #minetest |
| 16:22 |
MinetestBot |
[git] DS-Minetest -> minetest/minetest: Fix wrong code comment (#8061) 07c1c72 https://git.io/fhsl1 (2019-01-06T16:21:04Z) |
| 16:24 |
|
cafee_ joined #minetest |
| 16:50 |
|
red-001 joined #minetest |
| 16:50 |
|
cafee__ joined #minetest |
| 16:51 |
|
entuland_ joined #minetest |
| 17:02 |
|
Lia joined #minetest |
| 17:06 |
|
Cornelia joined #minetest |
| 17:12 |
|
cafee__ joined #minetest |
| 17:22 |
|
MinetestSam joined #minetest |
| 17:43 |
|
cafee__ joined #minetest |
| 18:16 |
|
huesal joined #minetest |
| 18:24 |
|
cafee__ joined #minetest |
| 18:27 |
|
FreeFull joined #minetest |
| 18:46 |
|
puzzlecube1 joined #minetest |
| 18:53 |
|
Ruslan1 joined #minetest |
| 19:05 |
|
Lia joined #minetest |
| 19:16 |
|
Cornelia joined #minetest |
| 19:16 |
|
Markow joined #minetest |
| 19:34 |
|
Fixer joined #minetest |
| 19:42 |
|
arsdragonfly joined #minetest |
| 19:44 |
|
ensonic joined #minetest |
| 20:00 |
|
turtleman joined #minetest |
| 20:05 |
|
scr267 joined #minetest |
| 20:34 |
|
dabbill joined #minetest |
| 20:36 |
|
Gael-de-Sailly joined #minetest |
| 20:44 |
|
puzzlecube joined #minetest |
| 20:45 |
|
cafee__ joined #minetest |
| 20:51 |
|
xSmurf joined #minetest |
| 20:58 |
|
scr267 joined #minetest |
| 21:01 |
|
entuland joined #minetest |
| 21:02 |
|
cafee__ joined #minetest |
| 21:06 |
|
cafee__ joined #minetest |
| 21:07 |
|
hecks joined #minetest |
| 21:09 |
hecks |
I started a singleplayer instance. Now I'm trying to connect to 127.0.0.1 in another window, but it's not working. |
| 21:09 |
sfan5 |
did you start it from the "singleplayer" tab? |
| 21:10 |
hecks |
Yes, does that disable all connectivity? |
| 21:10 |
sfan5 |
well actually there's a checkbox "Host server" |
| 21:10 |
sfan5 |
you need to check that |
| 21:10 |
|
cafee__ joined #minetest |
| 21:10 |
hecks |
Okay, and how do I prevent randos from joining? |
| 21:10 |
sfan5 |
also minetest randomizes the port every time IIRC |
| 21:10 |
sfan5 |
simple, you don't tell them your server address |
| 21:11 |
hecks |
It won't broadcast to master by default? |
| 21:11 |
sfan5 |
it won't, no |
| 21:11 |
hecks |
Thanks. |
| 21:12 |
hecks |
inb4 sfan is just trying to trick me into hosting mibi; I'll set a password just in case >:) |
| 21:12 |
|
riff-IRC joined #minetest |
| 21:12 |
hecks |
Oh I see now, there's an "Announce server" box for that |
| 21:12 |
sfan5 |
do you not have a firewall? |
| 21:12 |
hecks |
Who doesn't? |
| 21:13 |
sfan5 |
the point is that you can't announce a server without forwarding the port anyway |
| 21:13 |
hecks |
Well then, I couldn't host even if I wanted to, not on this mobile peasant connection |
| 21:14 |
rubenwardy |
I doubt it would be playable over a network anyway |
| 21:17 |
IhrFussel |
Looks like his idea (preloading X entries) exceeds the time the engine allows for the auth handler to setup |
| 21:17 |
IhrFussel |
Oops wrong channel |
| 21:17 |
hecks |
https://a.uguu.se/4lCrxgQ5rnde_mp.png |
| 21:17 |
sfan5 |
don't they get cold? |
| 21:18 |
hecks |
They're hot enough~ |
| 21:18 |
hecks |
But todo: implement freezing to death |
| 21:18 |
|
MinetestBot joined #minetest |
| 21:19 |
|
cafee__ joined #minetest |
| 21:19 |
hecks |
Why would this not be playable over a network? |
| 21:27 |
hecks |
Well well well https://a.uguu.se/eaCQdwvmMmFE_holdthephone.gif |
| 21:28 |
hecks |
I'm simulating an RTT of 80ms for both clients, and yet the animations are synced up |
| 21:32 |
|
peaceguy joined #minetest |
| 21:32 |
peaceguy |
to |
| 21:33 |
peaceguy |
hi all |
| 21:33 |
hecks |
peace man |
| 21:34 |
peaceguy |
peace |
| 21:34 |
peaceguy |
:) |
| 21:35 |
hecks |
anyway, I thought animation timing would be the nail in the coffin but it's pretty solid apparently |
| 21:35 |
hecks |
rollback would've been better of course but this is the next best thing |
| 21:36 |
sfan5 |
somewhat relevant https://i.4cdn.org/g/1546798847931.jpg |
| 21:41 |
hecks |
Mai Waifu |
| 21:41 |
hecks |
peace and anime tiddies https://a.uguu.se/1YtVEcdcg1cc_tiddies.gif |
| 21:41 |
hecks |
https://a.uguu.se/Qs8P5TJkdqGt_2.gif |
| 21:41 |
hecks |
|
| 21:43 |
hecks |
That doesn't really look like Lua to me though, unless the cup size is meant to represent the memory footprint. |
| 21:45 |
hecks |
Players attached to entities "vanishing" is a client side problem, study shows. |
| 21:46 |
hecks |
Often, instead of displaying the player at the entity's position, the player will be displayed at 0, 0, 0 |
| 21:47 |
hecks |
In other news, the monster spawning code still has some issues, as we're dealing with an overpopulation of jellies https://a.uguu.se/dKfUX7pN0svN_jelly.png |
| 21:50 |
IhrFussel |
Huh? sounds to me like active_block_range is too low to display the entity together with the player anymore -> entity (and player) unload |
| 21:51 |
IhrFussel |
And then the name tag is indeed showing up at 0,0,0 I think |
| 21:53 |
hecks |
This happens even when they're next to one another :( |
| 21:54 |
hecks |
yay dynlights https://a.uguu.se/Z2Rq6D7Df5Gw_dynlight.gif |
| 21:56 |
hecks |
Uh oh, I know this bug https://a.uguu.se/tmjJv06VAdop_oops.png |
| 21:56 |
IhrFussel |
But you set the entity to visible right? |
| 21:57 |
hecks |
Yes, it's a dummy one pixel transparent sprite if I recall right |
| 22:00 |
IhrFussel |
I never saw that happen on my server... players drive cars/carts/horses/boats but nobody ever reported such a thing |
| 22:04 |
hecks |
Maybe it's another artifact of placing the player's feet at local 0,0,0 |
| 22:04 |
hecks |
Maybe the game tries to cull entities that are inside nodes? |
| 22:05 |
hecks |
Another problem; animation of one player is stuck until I tab out to the other one, and perform some action on them... |
| 22:05 |
hecks |
Jury's still out on whose fault this is |
| 22:11 |
|
Swedneck joined #minetest |
| 22:14 |
|
scr267 joined #minetest |
| 22:15 |
hecks |
IhrFussel I think the issue is that carts and other vehicles are persistent entities |
| 22:15 |
hecks |
while I create my dummies on the spot every time the player is rooted, which happens often |
| 22:34 |
|
huesal joined #minetest |
| 22:48 |
|
puzzlecube joined #minetest |
| 22:49 |
|
bas080 joined #minetest |
| 22:50 |
|
huesal joined #minetest |
| 23:00 |
|
sonicpp joined #minetest |
| 23:05 |
|
paramat joined #minetest |
| 23:05 |
Swedneck |
Hi all, i just tried minetest on mobile and i'm quite impressed! |
| 23:07 |
entuland |
neat |
| 23:07 |
paramat |
MT 5.0.0-dev or current playstore version? |
| 23:08 |
IhrFussel |
Swedneck, you mean 0.4.17.1 or 5.0.0-dev? |
| 23:08 |
IhrFussel |
xD |
| 23:08 |
Swedneck |
current F-Droid version |
| 23:08 |
entuland |
F-Droid? |
| 23:08 |
paramat |
prob 0.4.17.1 then |
| 23:09 |
paramat |
the app is going to get far better in a few weeks when we release MT 5.0.0 |
| 23:09 |
Swedneck |
entuland: https://f-droid.org/en/packages/net.minetest.minetest/ |
| 23:09 |
IhrFussel |
Version 0.4.17.21 (21) - Added on 2019-01-04 |
| 23:09 |
Swedneck |
Basically play store for FOSS apps |
| 23:10 |
paramat |
yeah app '0.4.17.21' is MT 0.4.17.1 |
| 23:11 |
entuland |
oh nice |
| 23:12 |
entuland |
just yesterday my youtube app stopped working because of play store going unsupported on my galaxy tab 2 10.1 - weird |
| 23:12 |
Swedneck |
i'm also wondering if there is an easy way to get something more.. Interesting? than the base game |
| 23:12 |
Swedneck |
i know there are other games and mods, but it's a bit overwhelming at first |
| 23:12 |
entuland |
yep, getting a grip on the vanilla one may be good at first |
| 23:13 |
Swedneck |
Try newpipe from F-Droid :D |
| 23:14 |
Swedneck |
well i've already played minecraft, i think i can figure out the base game :P |
| 23:15 |
Swedneck |
Could minecraft modpacks be compared to minetest games? |
| 23:16 |
entuland |
there are some interesting differences |
| 23:16 |
paramat |
the basic MT game 'minetest game' is very basic, it's not really a 'game' |
| 23:16 |
entuland |
one I like a lot is the height of the map - minetest ones are pretty tall and deep :P |
| 23:16 |
|
huesal joined #minetest |
| 23:16 |
paramat |
MTG is more a base to add mods to |
| 23:17 |
Swedneck |
i've noticed, i spawned in what i guess is a mangrove forest but with mountains, quite a few times |
| 23:18 |
Swedneck |
Maybe MTG + some mods i think look cool? |
| 23:18 |
paramat |
MT 5.0.0 will have a new improved movement controls |
| 23:19 |
paramat |
*have new |
| 23:19 |
Swedneck |
nice |
| 23:19 |
Swedneck |
what are the main differences compared to current? |
| 23:20 |
Swedneck |
one of the things that impressed me was actually the movement controls, but maybe that's just the same as minecraft PE and i just have nothing to compare to :P |
| 23:21 |
|
huesal__ joined #minetest |
| 23:25 |
paramat |
see https://github.com/minetest/minetest/pull/7126 |
| 23:27 |
Swedneck |
thanks |
| 23:29 |
Swedneck |
oh shit the lower middle button is for jumping |
| 23:29 |
hecks |
Oh no. |
| 23:29 |
Swedneck |
somehow i ended up just ignoring it |
| 23:29 |
hecks |
<paramat> the app is going to get far better in a few weeks when we release MT 5.0.0 |
| 23:30 |
hecks |
I realize what this means and I don't like it. |
| 23:30 |
Swedneck |
hmm, maybe i interpreted that wrong, the lower middle button walks backwards.. |
| 23:31 |
IhrFussel |
Those arrow buttons will be gone in 5.0.0 |
| 23:31 |
paramat |
what does it mean? |
| 23:32 |
IhrFussel |
Swedneck, I hope you like virtual joysticks as much as buttons |
| 23:32 |
hecks |
It means breaking changes will be harder to get through no matter how necessary. |
| 23:32 |
IhrFussel |
For movement |
| 23:32 |
Swedneck |
isn't what we have now basically a virtual joystick? |
| 23:33 |
paramat |
ahh :) |
| 23:33 |
hecks |
I can already see you're rubbing your hands, paramat |
| 23:33 |
IhrFussel |
https://github.com/minetest/minetest/pull/7126 |
| 23:34 |
IhrFussel |
^ the screenshot shows the new controls |
| 23:34 |
paramat |
if there'an essential network breaking change needed it should be discussed now because it needs doing now |
| 23:34 |
Swedneck |
i've already been linked that :P |
| 23:35 |
IhrFussel |
Oh oops |
| 23:35 |
paramat |
the new joystick will take the first touch point as the joystick 'centre', easier to use |
| 23:35 |
Swedneck |
ah |
| 23:35 |
paramat |
and easier to change direction by sliding your finger around |
| 23:36 |
Swedneck |
nice |
| 23:36 |
IhrFussel |
Almost every mobile game works like that... left screen side to move around and right screen side to look around |
| 23:36 |
IhrFussel |
At least every single MMORPG I tried |
| 23:36 |
|
LMD joined #minetest |
| 23:37 |
|
cddepppp256 joined #minetest |
| 23:37 |
Swedneck |
i don't play many mobile games :P |
| 23:38 |
Swedneck |
really only The Room series and that tower jumping game on F-Droid |
| 23:39 |
Swedneck |
to install a game i just drop the .zip in ~/.minetest/games right? |
| 23:41 |
IhrFussel |
Swedneck, in 5.0.0 there will be an ingame content "store" (everything will be free there so store is not the right word) |
| 23:41 |
IhrFussel |
The official name is Content DB |
| 23:42 |
Swedneck |
oh god yes |
| 23:42 |
IhrFussel |
Basically browse through mods, download and install with 1 click |
| 23:42 |
Swedneck |
so like FTB launcher? |
| 23:44 |
IhrFussel |
One sec, making screenshots |
| 23:46 |
|
Ruslan1 joined #minetest |
| 23:50 |
IhrFussel |
Swedneck, here https://nofile.io/f/d3nHWVGDGDe/cdb_preview.mp4 |
| 23:50 |
MinetestBot |
[git] paramat -> minetest/minetest_game: Update small pines in schematic_tables.txt a2c9523 https://git.io/fhsKA (2019-01-06T23:48:57Z) |
| 23:52 |
IhrFussel |
Something better than a screenshot :P |
| 23:52 |
Swedneck |
and you can get both games and mods? |
| 23:53 |
IhrFussel |
Games, mods and texture packs |
| 23:55 |
Swedneck |
niiiiiice |
