| Time |
Nick |
Message |
| 00:33 |
kaeza |
is this intentional? https://github.com/minetest/minetest_game/blob/master/mods/default/torch.lua#L60 |
| 00:33 |
kaeza |
sofar, ^ |
| 00:34 |
kaeza |
the actual issue is that walking "on" a torch plays the wood sound |
| 00:37 |
|
betterthanyou710 joined #minetest-dev |
| 00:50 |
|
twoelk left #minetest-dev |
| 01:26 |
|
AntumDeluge joined #minetest-dev |
| 01:50 |
|
Lunatrius joined #minetest-dev |
| 02:34 |
sofar |
kaeza: well torches are wood, aren't they? |
| 02:34 |
sofar |
and walking sounds are horrendous :) |
| 02:35 |
sofar |
I mean, buggy |
| 02:35 |
sofar |
walk in 1 deep water and you hear sand sounds and not slushing water |
| 02:36 |
sofar |
kaeza: torches are `walkable = false` though, so how can you walk on them? |
| 02:37 |
kaeza |
sofar, place a dirt node, put a torch on one of its sides, hop onto the node, then walk on the border just above the torch |
| 02:37 |
sofar |
right |
| 02:37 |
kaeza |
the game thinks you are walking on the torch :I |
| 02:37 |
sofar |
not a bug in torches, though |
| 02:38 |
sofar |
again, walking sounds |
| 02:38 |
kaeza |
yeah |
| 02:38 |
sofar |
should just move them to CSM |
| 02:38 |
sofar |
then we can do far smarter footstep sounds |
| 03:22 |
|
paramat joined #minetest-dev |
| 03:23 |
paramat |
#5388 is ready can anyone review/approve? |
| 03:23 |
ShadowBot |
https://github.com/minetest/minetest/issues/5388 -- Map gen limit: Rewrite by paramat |
| 04:13 |
|
ssieb joined #minetest-dev |
| 04:20 |
|
Hunterz joined #minetest-dev |
| 04:56 |
|
paramat joined #minetest-dev |
| 05:37 |
paramat |
sofar i assume it's ok to merge game#1641 on just your approval? i will merge game#1624 game#1645 game#1646 soon also |
| 05:38 |
ShadowBot |
https://github.com/minetest/minetest_game/issues/1641 -- Fix crash when using unconfigured default:key by octacian |
| 05:38 |
ShadowBot |
https://github.com/minetest/minetest_game/issues/1624 -- Biomes: Add and auto-select mgv7 floatland biomes by paramat |
| 05:38 |
ShadowBot |
https://github.com/minetest/minetest_game/issues/1645 -- Junglegrass: Prevent seeding of dirt_with_grass by paramat |
| 05:38 |
ShadowBot |
https://github.com/minetest/minetest_game/issues/1646 -- Leafdecay: Change Aspen radius back to 3 by paramat |
| 05:38 |
sofar |
yeah, that's ok |
| 05:38 |
sofar |
edge case, but well tested |
| 05:38 |
paramat |
will do then |
| 05:47 |
paramat |
merging those 4 in a few mins |
| 05:54 |
paramat |
merging .. |
| 05:59 |
paramat |
complete |
| 06:01 |
|
Hunterz joined #minetest-dev |
| 06:12 |
kaeza |
if you're going to ask for reviews at least give them some weight |
| 06:12 |
kaeza |
and maybe some time before merging |
| 06:21 |
sofar |
https://www.youtube.com/watch?v=5LdLz8but8E |
| 06:21 |
sofar |
wool mod using colored nodes & colored itemstacks |
| 06:21 |
sofar |
wip (testing - PoC) |
| 06:34 |
paramat |
weight ? The map gen limit PR is really heavy maaan. those 4 were already approved or were trivial |
| 06:43 |
|
nerzhul joined #minetest-dev |
| 06:47 |
nerzhul |
merging #5401 |
| 06:47 |
ShadowBot |
https://github.com/minetest/minetest/issues/5401 -- [CSM] Add core.get_timeofday & core.get_day_count env calls by nerzhul |
| 06:57 |
nerzhul |
#5405 merged too |
| 06:57 |
ShadowBot |
https://github.com/minetest/minetest/issues/5405 -- [CSM] Fix minimap problems by nerzhul |
| 08:05 |
nerzhul |
can someone take a look at #5406, it's needed for future CSM interactions with client |
| 08:05 |
ShadowBot |
https://github.com/minetest/minetest/issues/5406 -- Reduce memory & function cost of Game class functions by nerzhul |
| 08:29 |
|
proller joined #minetest-dev |
| 08:35 |
|
kilbith joined #minetest-dev |
| 08:38 |
|
proller joined #minetest-dev |
| 08:47 |
|
proller joined #minetest-dev |
| 09:01 |
|
Lunatrius joined #minetest-dev |
| 09:08 |
|
nrzkt joined #minetest-dev |
| 09:12 |
|
paramat joined #minetest-dev |
| 09:12 |
|
Puka joined #minetest-dev |
| 09:20 |
|
Lunatrius joined #minetest-dev |
| 09:40 |
|
nrzkt joined #minetest-dev |
| 09:47 |
|
rubenwardy joined #minetest-dev |
| 09:50 |
|
rubenwardy left #minetest-dev |
| 09:50 |
|
glut321 joined #minetest-dev |
| 10:13 |
|
proller joined #minetest-dev |
| 10:14 |
|
Lunatrius joined #minetest-dev |
| 11:03 |
|
Player_2 joined #minetest-dev |
| 11:16 |
|
proller joined #minetest-dev |
| 11:22 |
kilbith |
how about officially dropping support for OpenGL and OpenGL ES 1.x, they're representing a very low marketshare and it's deprecated |
| 11:22 |
kilbith |
hint to https://github.com/minetest/minetest/issues/5403#issuecomment-287296226 |
| 11:25 |
Calinou |
https://xkcd.com/1172/ |
| 11:25 |
nrzkt |
kilbith: i think it's irrlicht support nos us |
| 11:25 |
nrzkt |
not* |
| 11:25 |
Calinou |
it will break someone's workflow ;) |
| 11:25 |
nrzkt |
kilbith: about long time unused thing i really want to raise the minimum protocol, i don't see why we support minetest 0.4.9 |
| 11:26 |
kilbith |
I mean, if it breaks on one's device, we just have to tell him to buy a more recent hardware |
| 11:26 |
nrzkt |
raise to 0.4.13, have a user and forkers contract, we support maximum 2 years old protocol |
| 11:26 |
nrzkt |
a 0.4.9 player don't have breath xD |
| 11:38 |
|
xunto joined #minetest-dev |
| 11:44 |
|
red-001 joined #minetest-dev |
| 11:44 |
red-001 |
what's the oldest mt version any server on the serverlist uses? |
| 11:49 |
|
Fixer joined #minetest-dev |
| 12:00 |
VanessaE |
regarding LBMs, when `run_at_every_load = false` is so set, what exactly determines that it's been run? |
| 12:01 |
VanessaE |
is the LBM name stored in the map block or something? |
| 12:21 |
|
Remy__ joined #minetest-dev |
| 12:29 |
|
paramat joined #minetest-dev |
| 12:40 |
nrzkt |
red-001: the problem is more clients than server, but except one inactive server in 0.4.9 and 3 inactive servers in 0.4.12 224 servers are at least 0.4.13 |
| 12:59 |
|
lisac joined #minetest-dev |
| 13:14 |
red-001 |
maybe show_formspec should be moved to the new minetest.ui namespace? |
| 13:16 |
nrzkt |
red-001: minetest.ui contains references to objects not functions |
| 13:16 |
nrzkt |
but maybe we should do some namespace triage, don't know |
| 13:20 |
|
XtremeHacker joined #minetest-dev |
| 13:22 |
red-001 |
maybe game#1647 should be closed? |
| 13:22 |
ShadowBot |
https://github.com/minetest/minetest_game/issues/1647 -- Nyan Cat license |
| 13:23 |
red-001 |
nvm it just got closed by sfan |
| 13:51 |
|
numZero joined #minetest-dev |
| 14:15 |
|
juhdanad joined #minetest-dev |
| 14:19 |
|
juhdanad joined #minetest-dev |
| 14:55 |
|
YuGiOhJCJ joined #minetest-dev |
| 15:04 |
|
Remy__ joined #minetest-dev |
| 15:07 |
|
proller joined #minetest-dev |
| 15:18 |
|
Remy__ joined #minetest-dev |
| 15:21 |
red-001 |
huh looks like someone has copyrighted the nyancat and is willing to sue people |
| 15:21 |
red-001 |
also they have a trademark |
| 15:26 |
sfan5 |
nothing new there |
| 15:30 |
nrzkt |
sfan5: did you get some time for a short review ? |
| 15:30 |
VanessaE |
red-001: that's the reason I re-drew it. |
| 15:30 |
sfan5 |
nrzkt: of what? |
| 15:30 |
nrzkt |
#5113 |
| 15:30 |
ShadowBot |
https://github.com/minetest/minetest/issues/5113 -- [CSM] Give CSM access to core.colorize by red-001 |
| 15:31 |
nrzkt |
it's basicly code move from server lua builtin to common builtin for permit using it client side |
| 15:32 |
red-001 |
well I doubt minetest will be targeted, since that wouldn't make them any richer |
| 15:49 |
|
Taoki joined #minetest-dev |
| 15:53 |
|
glut32 joined #minetest-dev |
| 15:54 |
nrzkt |
red-001: there is a debug stuff mentionned by sfan5 you added in last fix |
| 15:54 |
nrzkt |
can you remove it ? |
| 15:54 |
nrzkt |
except this debug stuff are you okay sfan5? |
| 15:55 |
glut32 |
Hi all, I have a bug on minetest running on mac osx. The "M" Key doesn't work with my french keyboard (AZERTY). Does someone know how to fix this bug? |
| 15:55 |
nrzkt |
irrlicht bug i think |
| 15:55 |
glut32 |
This bug is reported here : http://wiki.minetest.net/Controls |
| 15:55 |
glut32 |
I think also that it's in irrlicht |
| 15:56 |
nrzkt |
if you try irrlicht 1.9 you should see a different it fixes many keyboard problems |
| 15:56 |
sfan5 |
nrzkt: yes its fine with me otherwise |
| 15:56 |
nrzkt |
nice thanks |
| 15:56 |
glut32 |
when irrlicht 1.9 will be released ? |
| 15:57 |
nrzkt |
i don't know, it's in dev since years... lol |
| 15:57 |
glut32 |
ah ah |
| 15:57 |
glut32 |
it's like duken nukem |
| 15:57 |
glut32 |
*duke nukem |
| 15:57 |
glut32 |
lol |
| 15:57 |
|
Karazhan joined #minetest-dev |
| 15:58 |
nrzkt |
yeah or diablo3 |
| 15:59 |
|
Remy__ joined #minetest-dev |
| 16:04 |
|
Hunterz joined #minetest-dev |
| 16:09 |
red-001 |
done |
| 16:10 |
|
Remy__ left #minetest-dev |
| 16:12 |
|
numZero joined #minetest-dev |
| 16:17 |
|
Puka joined #minetest-dev |
| 16:29 |
|
Darcidride joined #minetest-dev |
| 16:29 |
|
blaze joined #minetest-dev |
| 16:30 |
|
Fritigern joined #minetest-dev |
| 16:47 |
|
Fritigern joined #minetest-dev |
| 17:29 |
|
Fritigern joined #minetest-dev |
| 17:50 |
|
Tmanyo joined #minetest-dev |
| 17:55 |
Fixer |
nrzkt: 4 servers with 0.4.12 (mostly empty, 2players), 13 with 0.4.13 (mostly empty, 7players), 77 with 0.4.14 (lots of empties and some heavy played ones, 232players), 114 with 0.4.15 (301 players) // from few weeks old data |
| 17:56 |
Fixer |
it is safe to say that currently 0.4.14to0.4.15dev are most used versions |
| 18:02 |
|
Fritigern joined #minetest-dev |
| 18:09 |
red-001 |
maybe raise the minimal protocol version to 0.4.12 then? |
| 18:11 |
|
nerzhul joined #minetest-dev |
| 18:18 |
|
Krock joined #minetest-dev |
| 18:18 |
|
Krock joined #minetest-dev |
| 18:20 |
nerzhul |
merging #5113 |
| 18:20 |
ShadowBot |
https://github.com/minetest/minetest/issues/5113 -- [CSM] Give CSM access to core.colorize by red-001 |
| 18:29 |
Fixer |
nerzhul: https://i.imgur.com/4N31sSs.png |
| 18:30 |
nerzhul |
yes :) |
| 18:31 |
nerzhul |
maybe 0.4.12 at least plus communicate with server admins |
| 18:31 |
Fixer |
may be a bit outdated, but should more or less represent the situation |
| 18:31 |
nerzhul |
i think we should cross with sfan5 serverlist, using only valid MT versions (not fork versions) |
| 18:31 |
Fixer |
55% of players are on 0.4.15+ |
| 18:32 |
nerzhul |
and this is nice, a lesser than 1 year old server |
| 18:32 |
Fixer |
42% are on 0.4.14(+dev) |
| 18:32 |
Fixer |
might be some slight errors in this |
| 18:32 |
nerzhul |
and no, not 55% of players, 55% player play on a vXXX server :) |
| 18:32 |
Fixer |
for 0.4.14 or 0.4.14dev |
| 18:33 |
sofar |
you can raise it to 0.4.14 from that data alone |
| 18:33 |
|
twoelk joined #minetest-dev |
| 18:33 |
sofar |
do you really care about 1%? |
| 18:33 |
sofar |
9 players |
| 18:33 |
nerzhul |
sofar, i think we should cross the server part and the client part |
| 18:33 |
nerzhul |
we need sfan5 report for the last 6 months |
| 18:33 |
nerzhul |
to have a credible version report |
| 18:34 |
nerzhul |
for CSM future usage we need to raise and have a contract about protocol maintenance |
| 18:34 |
Fixer |
seriously, why still on 0.4.12 and 0.4.13? it is too outdated, and 0.4.16 soon |
| 18:35 |
sofar |
well, if you do what others do |
| 18:35 |
sofar |
e.g. |
| 18:35 |
sofar |
one release every 6 months |
| 18:35 |
sofar |
maintain 4 versions (2 years) |
| 18:35 |
sofar |
so when 0.4.16 comes out, you remove 0.4.12 |
| 18:35 |
nerzhul |
sofar, it's what i suggested, 2 years contract |
| 18:36 |
nerzhul |
it permit to nuke some old unused code |
| 18:36 |
sofar |
and you send warnings to 0.4.13 that it's outdated |
| 18:36 |
sofar |
(if even possible) |
| 18:36 |
nerzhul |
and at a time, we should think about having a migration process on minetest databases permitting to convert very old mapblocks to new format (maybe external from minetest itself) permitting to evolve, but it's not the discussion there |
| 18:39 |
red-001 |
you could make new servers send chat messages imforming users they are using an old version |
| 18:39 |
nerzhul |
if i remember i implemented it in 0.4.14 or 0.4.13 |
| 18:39 |
red-001 |
server owners can't do this |
| 18:39 |
nerzhul |
or there is an open pr |
| 18:40 |
red-001 |
as minetest doesn't let lua know the protocol/mt-version of a client |
| 18:41 |
red-001 |
implemeted map converting? |
| 18:42 |
sfan5 |
red-001: it doesn't need to be done in lua |
| 18:42 |
Fixer |
current 0.4.12 and 0.4.13 servers from serverlist: https://i.imgur.com/FGmKub4.png They are mostly empty |
| 18:43 |
Fixer |
with just 3 players on... |
| 18:43 |
red-001 |
Fixer, the main issue isn't the number of servers |
| 18:44 |
Fixer |
number of players |
| 18:44 |
Fixer |
not servers |
| 18:44 |
red-001 |
but how many users still us old versions |
| 18:44 |
red-001 |
use* |
| 18:44 |
Fixer |
you mean clients |
| 18:44 |
red-001 |
yes |
| 18:44 |
Fixer |
will be interesting to know |
| 18:44 |
sofar |
tbh |
| 18:45 |
Fixer |
last time I seen graphs from sfan and someone else it was holyshit fragmented like with servers |
| 18:45 |
sofar |
if clients fetch the serverlist we can inform them |
| 18:45 |
red-001 |
you could try and get some server owners to log the mt-versions of users that connect |
| 18:45 |
sofar |
at a minimum the server list can include the "current version" and print a message in the UI somewhere |
| 18:45 |
red-001 |
oh is the mt-version in the user agent? |
| 18:45 |
Fixer |
red-001: rubenwardy has code for that |
| 18:45 |
sofar |
red-001: no, the client doesn't need to send any data |
| 18:46 |
sofar |
red-001: all that's needed is the serverlist send out the current stable client version |
| 18:46 |
sofar |
well um |
| 18:46 |
Fixer |
red-001: https://kitsunemimi.pw/tmp/serverlist_stats_2016-10-05.txt |
| 18:46 |
sofar |
that will help with *future* client releases, lol |
| 18:46 |
Fixer |
not sure if it is this |
| 18:46 |
Fixer |
should be |
| 18:47 |
red-001 |
I would be nice to have some sort of easy way to warn players that thier version of mt is outdated |
| 18:47 |
red-001 |
s/I/it |
| 18:47 |
Fixer |
ah yes, it is 50/50 mt vs multicraft |
| 18:47 |
Fixer |
and at the time most of folks used 0.4.13 it seems |
| 18:48 |
red-001 |
is that the one that spams a lot of ads> |
| 18:48 |
Fixer |
probably |
| 18:48 |
Fixer |
i don't use anything with ads and they are bloked everywhere including on googlespyos |
| 18:49 |
Fixer |
we need newer stats |
| 18:49 |
Fixer |
to see if 0.4.13 is safe to nuke |
| 18:49 |
Fixer |
or whatever you want |
| 18:50 |
red-001 |
I say keep 0.4.13 |
| 18:50 |
red-001 |
it's only two releases old |
| 18:50 |
Fixer |
0.4.13 was released on August 20, 2015 |
| 18:52 |
Fixer |
on that survery: 0.4.14 - 54%, 0.4.13 - 35% |
| 18:52 |
red-001 |
maybe someone could run another survey? |
| 18:52 |
sfan5 |
i have to do those surveys |
| 18:53 |
sfan5 |
because the access logs are obviously not public |
| 18:53 |
Fixer |
Mobile -> 2320046 92.51% |
| 18:53 |
Fixer |
kek |
| 18:53 |
* red-001 |
expects at least 95% |
| 18:54 |
Fixer |
93% will not miss sneak elevators |
| 18:54 |
Fixer |
93% uses calculator to play |
| 18:54 |
Fixer |
#pcmasterrace |
| 18:54 |
Krock |
kids' flooding the stats! |
| 18:59 |
nerzhul |
please note 0.4.7 or older is sometimes rogue clients |
| 18:59 |
nerzhul |
for multicraft we are not the code base owners, it's not our problem to maintain compat but fork owner |
| 19:00 |
nerzhul |
he should see the communication and do sth |
| 19:00 |
sfan5 |
who should do something about what? |
| 19:00 |
nerzhul |
sfan5, can you do report from the last 6 months about version usage please ? |
| 19:00 |
sfan5 |
sure |
| 19:04 |
nerzhul |
ty |
| 19:09 |
red-001 |
Could I suggest changing the forum rules so that posting only complied bytecode for mods is not allowed? |
| 19:10 |
red-001 |
it can be a security issue |
| 19:11 |
Krock |
haven't seen such a mod yet but that's surely something that should be in the rules |
| 19:12 |
Fixer |
there is on |
| 19:12 |
Fixer |
one |
| 19:12 |
red-001 |
also the mod is likely to break if you use a diffrent mt/lua/arch |
| 19:12 |
Fixer |
anticheat by rnd |
| 19:12 |
red-001 |
^ |
| 19:12 |
Fixer |
has bytecode |
| 19:13 |
Fixer |
to hide code from cheaters |
| 19:13 |
Thomas-S |
https://forum.minetest.net/viewtopic.php?f=9&t=15424 |
| 19:13 |
Fixer |
maybe he published everything now |
| 19:13 |
Fixer |
and bytecode was MY idea |
| 19:13 |
Fixer |
btw |
| 19:13 |
red-001 |
security by obscurity |
| 19:13 |
Fixer |
he was not going to publist it, but I said he can use bytecode and at least make it available |
| 19:14 |
red-001 |
it's not even that much obscurity |
| 19:14 |
Fixer |
so he did |
| 19:14 |
red-001 |
you can easily decomplie bytecode or at least disassmble it |
| 19:14 |
Fixer |
i suggested googling obfuscator %) |
| 19:15 |
* Fixer |
runs away |
| 19:15 |
red-001 |
lol |
| 19:15 |
Krock |
ask kaeza to obfuscate it for you |
| 19:15 |
red-001 |
you can also to that with lua code to a certain degree |
| 19:17 |
Krock |
replacing tabs, LF/CR with spaces and variable names with numerals like _1, _2 and so on |
| 19:17 |
red-001 |
I mean normal complied lua has varible names in plaintext |
| 19:18 |
Fixer |
it was security via obscurity, yes |
| 19:19 |
Fixer |
i'm not sure if posting such mods is ok on forum though |
| 19:20 |
Fixer |
maybe someone will find it useful |
| 19:53 |
|
Taoki joined #minetest-dev |
| 19:56 |
|
Darcidride joined #minetest-dev |
| 20:15 |
|
Lunatrius joined #minetest-dev |
| 20:28 |
nerzhul |
merging #5384 |
| 20:28 |
ShadowBot |
https://github.com/minetest/minetest/issues/5384 -- avoid crashing when accessing mapgen early by zeuner |
| 20:32 |
|
DI3HARD139 joined #minetest-dev |
| 20:37 |
nerzhul |
Krock, sofar can you look at #5406 please ? |
| 20:37 |
ShadowBot |
https://github.com/minetest/minetest/issues/5406 -- Reduce memory & function cost of Game class functions by nerzhul |
| 20:38 |
Krock |
already looked at it and only found these two "this->" lines |
| 20:38 |
nerzhul |
it's normal |
| 20:38 |
nerzhul |
GH didn't notify me of your comment |
| 20:38 |
Krock |
Ctrl+F5 ;) |
| 20:39 |
|
Fixer joined #minetest-dev |
| 20:39 |
nerzhul |
or maybe it's not normal now, i needed it at a moment, but after rewriting another part, missing it, looking |
| 20:40 |
nerzhul |
exact, thanks for noiice |
| 20:40 |
|
juhdanad joined #minetest-dev |
| 20:41 |
Krock |
apart from that, lovely. I wonder how this mess even started :3 |
| 20:41 |
nerzhul |
i don't know, never see anybody working on it on 3 years :p |
| 20:42 |
nerzhul |
i think at first rundata was a local variable moved toclass at a point and code was never refactored |
| 20:42 |
nerzhul |
in fact it was added by ShadowNinja in Nov 2016 |
| 20:43 |
nerzhul |
it was local to Game::Run before |
| 20:43 |
nerzhul |
SN started to refactor it by moving to class, and do other optimisations to improve rendering, and my commit finish this variable refactoring as it seems :) |
| 20:43 |
nerzhul |
it rebased + force push, can you re-look at this |
| 20:44 |
nerzhul |
please :)? |
| 21:15 |
nerzhul |
merging #5406 |
| 21:15 |
ShadowBot |
https://github.com/minetest/minetest/issues/5406 -- Reduce memory & function cost of Game class functions by nerzhul |
| 21:16 |
nerzhul |
thanks for your review Krock |
| 21:17 |
Krock |
np, it's easier to review these pulls than such that add new features :) |
| 21:17 |
|
rubenwardy joined #minetest-dev |
| 21:18 |
nerzhul |
this PR was a little bit anoying to read many lines are identical :p |
| 21:19 |
Krock |
indeed |
| 21:19 |
nerzhul |
sfan5, ping, does your processing finished ? |
| 21:19 |
nerzhul |
Krock, we should refactor the flag usage now, i need it for CSM, it contains the real flag to showing hiding minimap, the current model on CSM is not as perfect as using this flag |
| 21:21 |
|
proller joined #minetest-dev |
| 21:22 |
VanessaE |
sofar: another thing we desperately need: rollback_check works fine, responds nearly instantly usually, but rollback (that is, undoing actions) is dog slow these days |
| 21:23 |
sofar |
worth a bug report? |
| 21:23 |
sofar |
I've never used it |
| 21:23 |
nerzhul |
sofar, i tryied to used it 2 years ago, but only usage was disk usage, i removed it :p |
| 21:24 |
sfan5 |
nerzhul: i havent begun yet |
| 21:24 |
sofar |
seems with pgsql it shouldn't be so bad performance wise? |
| 21:24 |
VanessaE |
sofar: I think there's already a bug report for it |
| 21:26 |
nerzhul |
sofar, rollback only used sqlite no ? |
| 21:26 |
nerzhul |
i didn't looked at its requests, maybe a db index is missing |
| 21:26 |
VanessaE |
it's sqlite, yes. |
| 21:26 |
sofar |
I'm not sure where the slowdown would come from tbh |
| 21:26 |
VanessaE |
I assumed it had to do with the actual act of setting nodes etc. during the restore, not so much the database lookup |
| 21:27 |
nerzhul |
VanessaE, let me 5 mins to look at the code |
| 21:27 |
VanessaE |
ok |
| 21:27 |
nerzhul |
can you tell me which action is slow ? |
| 21:28 |
VanessaE |
/rollback <user> 999999 |
| 21:28 |
VanessaE |
(the user in question actually only had actions going back ~150000, I just use lots of 9's to kill off anything they built) |
| 21:29 |
nerzhul |
i think index are missing, i see there are indices on coordinates and id |
| 21:29 |
VanessaE |
sofar: get my note, btw? |
| 21:29 |
nerzhul |
there is missing indices |
| 21:29 |
nerzhul |
it's sure |
| 21:30 |
nerzhul |
some select conditions are not using indices |
| 21:30 |
Fixer |
we need rollback limit in days or whatever |
| 21:30 |
sofar |
UD protection checks? yeah. Although I'm sad I didn't paint a penis anywhere on your server as a PoC |
| 21:30 |
nerzhul |
action needs timestamp + actor index only |
| 21:30 |
Fixer |
sofar: drawing ***** is so 2011 |
| 21:30 |
nerzhul |
other cases seems covered |
| 21:30 |
sofar |
stars? I find them offensive |
| 21:31 |
nerzhul |
VanessaE, at a point, shutdown your server (you cannot add indice to your sqlite while db i poened by another process) |
| 21:31 |
VanessaE |
triggered! :) |
| 21:31 |
VanessaE |
nerzhul: every morning at 11am CET. |
| 21:31 |
nerzhul |
CREATE INDEX IF NOT EXISTS `action_ts_actor` on `action`(`timestamp`,`actor`); |
| 21:31 |
nerzhul |
this should help rollback actions based on timestamp & actor |
| 21:32 |
nerzhul |
also some actions are downloading the whole actor or node table, it can be very slow |
| 21:33 |
juhdanad |
Has it ever occured to you that Minetest disables a modpack (worldedit) when you join to the world, while other mods remain enabled? |
| 21:33 |
VanessaE |
juhdanad: sure, if its dependencies aren't met. some other wayu? |
| 21:33 |
VanessaE |
-i |
| 21:33 |
VanessaE |
-u* |
| 21:34 |
juhdanad |
I'll check... |
| 21:34 |
juhdanad |
No, it has no dependencies. |
| 21:36 |
VanessaE |
nerzhul: how does that work exactly? (I so rarely use sqlite for anything, I don't know how to use it on the rollback log) |
| 21:36 |
nerzhul |
sqlite <file> |
| 21:36 |
nerzhul |
run query |
| 21:36 |
nerzhul |
close |
| 21:36 |
nerzhul |
start your server |
| 21:36 |
VanessaE |
hm, ok |
| 21:37 |
nerzhul |
index can be long to create depending on table size |
| 21:37 |
VanessaE |
ok, I'll do that in a while, next time I need to reboot |
| 21:38 |
* VanessaE |
records this conversation in a text editor |
| 21:39 |
VanessaE |
sofar: heh, wiseass :) |
| 21:39 |
VanessaE |
(re: protection check) |
| 21:41 |
nerzhul |
i don't guarantee a miracle indices are use when context permits it, but i see a missing indice here, if it works better for you we will add it to our scheme |
| 21:42 |
VanessaE |
ok |
| 21:53 |
sofar |
VanessaE: I thank you for the compliments to my rear |
| 21:54 |
VanessaE |
heh |
| 22:02 |
|
Warr1024 joined #minetest-dev |
| 22:44 |
sfan5 |
some madman is running an android build in nacl Minetest/0.4.15-bc2861e2-Android (nacl/31 nacl)" |
| 22:44 |
|
proller joined #minetest-dev |
| 22:45 |
VanessaE |
sofar: btw, you might want to chime in on #1530. |
| 22:45 |
ShadowBot |
https://github.com/minetest/minetest/issues/1530 -- Add minetest.register_on_time_change by HybridDog |
| 22:45 |
nerzhul |
what is that sfan5 |
| 22:45 |
VanessaE |
er |
| 22:45 |
VanessaE |
game#1530 |
| 22:45 |
ShadowBot |
https://github.com/minetest/minetest_game/issues/1530 -- Add paint API by juhdanad |
| 22:45 |
nerzhul |
for me nacl is Na Cl :p |
| 22:45 |
sfan5 |
https://developer.chrome.com/native-client/nacl-and-pnacl nerzhul |
| 22:45 |
VanessaE |
bbl |
| 22:45 |
nerzhul |
oh |
| 22:46 |
sfan5 |
also i think use on desktop (as opposed to mobile) has risen |
| 22:57 |
|
proller joined #minetest-dev |
| 23:15 |
sfan5 |
nerzhul: https://kitsunemimi.pw/tmp/serverlist_stats_2017-03-17.txt |
| 23:18 |
nerzhul |
sfan5, thanks |
| 23:18 |
|
kaeza joined #minetest-dev |
| 23:19 |
nerzhul |
2.5% 0.4.11 hmmm |
| 23:19 |
nerzhul |
more forks now as it seems |
| 23:20 |
nerzhul |
then we should update our minimum protocol to 0.4.11 at least and drop previous support |
| 23:20 |
Calinou |
almost 9% of IPv6, not bad |
| 23:20 |
nerzhul |
newest Linux version: 4.9.9-rt6-1-rt |
| 23:21 |
nerzhul |
this is wrong i'm using 4.10.3 xD |
| 23:21 |
Calinou |
MultiCraft, 60% of all requests :( |
| 23:21 |
nerzhul |
0.4.11 client are ios ? |
| 23:21 |
sfan5 |
nerzhul: sort -n doesn't sort versions correctly ¯\_(ツ)_/¯ |
| 23:22 |
nerzhul |
tail -n 2623000 /usr/local/nginx/logs/serverlist.log strange it's not a date, normal ? :) |
| 23:22 |
sfan5 |
Calinou: better than 60% of requests from a majorly outdated fork that was basically abandoned by its author |
| 23:22 |
Fixer |
had strange fps and minimap related problem, updating again |
| 23:22 |
sfan5 |
yes i tune the line count manually to match the 10 day span |
| 23:22 |
nerzhul |
sfan5, nice |
| 23:22 |
nerzhul |
okay then 10 days instead of 6 months, ^^ |
| 23:23 |
nerzhul |
maybe 1 month could be better then ? possible to extend it ? |
| 23:23 |
sfan5 |
yes but i don't feel like redoing that stats right now |
| 23:23 |
nerzhul |
you cannot do a script calculating all ? :) |
| 23:23 |
nerzhul |
i will open an issue to udpate at least our compat to 0.4.11 |
| 23:23 |
sfan5 |
nerzhul: https://xkcd.com/1205/ |
| 23:24 |
Fixer |
Windows -> 63410 03.71% |
| 23:24 |
Fixer |
O_O |
| 23:24 |
nerzhul |
Fixer, i think windows users use MC only |
| 23:24 |
Fixer |
probably me+krock+few guys make up this 3.71% |
| 23:24 |
|
proller joined #minetest-dev |
| 23:25 |
Fixer |
jordach |
| 23:25 |
red-001 |
I'm sometimes part of the 4% |
| 23:26 |
Fixer |
so at b52f300 i had strange fps problem, i've used minimap for some time, off/on, map was loaded in memory most of the time, noticed that fps even while looking at sky was just 65... that was odd |
| 23:27 |
Fixer |
it feels like minimap was enabled but not shown |
| 23:27 |
Fixer |
updating to newest, will see how it works later |
| 23:31 |
Fixer |
i see players from necrodistros |
| 23:32 |
Fixer |
0.4.13 as top for nonmobile, most of them are linux guys, ridiculously old |
| 23:32 |
Fixer |
NOT FINE |
| 23:34 |
red-001 |
intersting there are less moblie users this time |
| 23:38 |
nerzhul |
sfan5, rubenwardy sofar Zeno` please comment your opinion about bumping minimum version in #5411, server owner comments are welcome |
| 23:39 |
ShadowBot |
https://github.com/minetest/minetest/issues/5411 -- Update server min protocol version to v24 by nerzhul |
| 23:39 |
sfan5 |
seems fine to me especially since this will only really take effect when .16 is released |
| 23:39 |
sfan5 |
for the 0.4.11 min. |
| 23:40 |
nerzhul |
can you add your comment on issue, ti's better for tracking |
| 23:40 |
nerzhul |
thanks :) |
| 23:40 |
sfan5 |
also swap_node is irrelevant because server-side, no? |
| 23:40 |
nerzhul |
when we decide which version, i will drop code parts related to this |
| 23:40 |
nerzhul |
sfan5, it's in the protocol changelog |
| 23:41 |
nerzhul |
i don't know how it's used exactly |
| 23:41 |
nerzhul |
proto v22 |
| 23:49 |
kaeza |
nerzhul, would you consider making the client-side API as identical as possible to the server-side API? |
| 23:50 |
kaeza |
I can envision distributing a mod as both client- and server- side, and only enable features depending on which side we are on |
| 23:50 |
nerzhul |
kaeza, not in term of functionalities because each other have its owner specs, but when a call can be same between client and server it's done |
| 23:50 |
rubenwardy |
reusing code is good |
| 23:51 |
nerzhul |
client will not have all server specs and same for server, for example client don't have http API (for security reasons), and server don't have minimap placement api :) |
| 23:51 |
nerzhul |
rubenwardy, for maintenance yes :) |
| 23:51 |
kaeza |
I know what you mean, but it makes sense to reuse the server-side code even if some arguments are dummies |
| 23:52 |
nerzhul |
generally you have, either a similar class or the same class but a InitializeClient which only reuse some functions calls and make it more generic (example ServerEnv => Env) |
| 23:52 |
kaeza |
e.g. some callbacks are missing playerref arguments. maybe it could always be the local player? |
| 23:52 |
nerzhul |
with the current progress i don't know what can be possible, but i know when more calls will be added craftking books will be purely client side |
| 23:53 |
nerzhul |
we will have a special minetest.localplayer object i think for localplayer |
| 23:53 |
nerzhul |
minimap is minetest.ui.minimap |
| 23:53 |
nerzhul |
in ui we can set our ui functions maybe, but ui objects |
| 23:54 |
nerzhul |
i don't know if i will have time for this, but i will try to start the base for our localplayer object client side, and try to use as close api as server |
| 23:55 |
|
YuGiOhJCJ joined #minetest-dev |
| 23:57 |
Fixer |
ok, i'm panicing npw |
| 23:57 |
Fixer |
something is wrong with minetest |
| 23:58 |
Fixer |
after visiting certain area with minimap on I have instand fps drop ??? |
| 23:58 |
Fixer |
looking at sky - 60fps (it was 95 or so) just before this |
| 23:58 |
Fixer |
not sleeping on f5 graph for some reason |
| 23:58 |
Fixer |
o_o |
| 23:59 |
nerzhul |
Fixer, minimap is not very optimal, maybe it's due to recent somebody minimap "optimizations" which are not |
| 23:59 |
Fixer |
nerzhul: i disabled minimap, still 60-65 fps while looking at sky, smth is wrong now |
