Time |
Nick |
Message |
01:03 |
|
twoelk left #minetest-project |
01:27 |
|
benrob0329 joined #minetest-project |
03:18 |
|
eidy joined #minetest-project |
03:19 |
|
benrob0329 joined #minetest-project |
03:28 |
|
benrob0329 joined #minetest-project |
03:39 |
benrob0329 |
That is the cat5tv chanel? |
03:45 |
benrob0329 |
*what |
03:46 |
paramat |
oh heh |
03:46 |
paramat |
https://www.youtube.com/channel/UCIcUvKUXCySOjSFCFgZ3otQ/videos |
03:47 |
benrob0329 |
I mean irc chaanel |
03:47 |
benrob0329 |
*Chanel |
04:01 |
|
eidy joined #minetest-project |
04:02 |
|
lordfingle joined #minetest-project |
04:12 |
|
est31 joined #minetest-project |
04:14 |
|
lumidify joined #minetest-project |
04:45 |
OldCoder |
|
04:45 |
OldCoder |
http://minetest.org/mtcompile.txt |
04:46 |
OldCoder |
This is a build script created just now for TommyTreasure ^ |
04:46 |
OldCoder |
|
04:46 |
OldCoder |
This version is untested but should work with minor tweaks. It is the one, largely, that I use. |
04:49 |
|
ssieb joined #minetest-project |
05:05 |
|
nerzhul joined #minetest-project |
05:23 |
|
Brackston joined #minetest-project |
05:42 |
|
lumidify joined #minetest-project |
06:14 |
|
passant joined #minetest-project |
07:14 |
|
proller joined #minetest-project |
07:33 |
|
nrzkt joined #minetest-project |
08:06 |
|
tenplus1 joined #minetest-project |
08:06 |
tenplus1 |
hi folks |
08:06 |
VanessaE |
hi |
08:06 |
tenplus1 |
hi VanessaE |
08:17 |
nrzkt |
hi |
08:17 |
tenplus1 |
hi nrz |
08:18 |
nrzkt |
tenplus1, VanessaE: i think i will need you later today (GMT timezone) |
08:18 |
tenplus1 |
whassup dude ? |
08:18 |
VanessaE |
? |
08:18 |
VanessaE |
what's up? |
08:18 |
VanessaE |
(or what will be) |
08:20 |
nrzkt |
i'm finishing the work on https://github.com/minetest/minetest/pull/5475 i will ask you to give me some player files with criterias on your live servers |
08:20 |
nrzkt |
not all, but an extract of very old and recent files to test my migration processus |
08:21 |
tenplus1 |
will check in a bit to see what it's all about :P |
08:21 |
nrzkt |
i finished the work on player saving/loading from/to database, now i'm working on automatic migration of player files |
08:21 |
VanessaE |
nrzkt: wait, so the plan is to ditch worlds/<foo>/players/<bar, baz, blah, ...> ? |
08:22 |
nrzkt |
exactly |
08:22 |
VanessaE |
:( |
08:22 |
tenplus1 |
willi tbe contained in the map or one huge file ? |
08:22 |
VanessaE |
that's gonna be a big problem for server admins unless you provide a good tool to read/edit a player's record. |
08:22 |
nrzkt |
many coredevs (including me) wants to change this non reliable behaviour |
08:23 |
nrzkt |
VanessaE: in fact as it uses relational DB (sql only) it's easy to edit :) |
08:23 |
VanessaE |
easy if you know databases. |
08:23 |
VanessaE |
most of us don't :P |
08:24 |
nrzkt |
tenplus1: player are in a dedicated database in sqlite3, in postgres it depend on which connection chain you set. You can use same as map, but i invite you to have a separated database |
08:24 |
tenplus1 |
if database becomes corrupt then ALL player files are a risk ? |
08:25 |
nrzkt |
for server owners needing performance i suggest to use postgresql backend, i will add a notice about the migration, because by default you will migrate to sqlite |
08:25 |
nrzkt |
database never corrupts if you never do shit yourself with filesystem :) |
08:26 |
tenplus1 |
just chceking :P |
08:26 |
VanessaE |
nrzkt: and mapblocks never get corrupted either ;) |
08:26 |
nrzkt |
VanessaE: yeah, it was my second argument: is your map corrupt ? map is heavy, please files is very tiny |
08:26 |
nrzkt |
VanessaE: also using database will permit server owners to plug a website over it for dynamic things |
08:27 |
nrzkt |
imagine a website connecting to your postgres and showing your inventory ? :) |
08:27 |
nrzkt |
also, i just migrate the player files, not the account file, this will be done in another PR |
08:27 |
VanessaE |
my point (and tenplus1's) I think is that if something gets trashed in the player database, minetest needs to handle it sanely |
08:27 |
nrzkt |
there is not player removal from core engine, if you remove it yourself you should know what you are doing |
08:28 |
nrzkt |
for sqlite backend, just copy the sqlite file |
08:28 |
nrzkt |
for a backup, like your map |
08:28 |
nrzkt |
for pg: pg_dump <database> > database.sql |
08:30 |
nrzkt |
i will document th process heavily because it's an important change for server owners (less important for local games) |
08:30 |
VanessaE |
well let's just suppose this player database gets damaged (nevermind how). suppose the corruption only exists in two players' records, such that they're broken, unreadable, whatever. what will MT do? |
08:30 |
nrzkt |
it's impossible to have records broken in a database |
08:30 |
tenplus1 |
hypothetically, what IF it happens? |
08:30 |
VanessaE |
don't split hairs. |
08:30 |
nrzkt |
restore your backup, like every system |
08:31 |
VanessaE |
what happens if two players' inventories are corrupted and the admin can't spot it right away? |
08:31 |
VanessaE |
what will MT do when it loads the database? |
08:31 |
nrzkt |
also, just for a note, i have heavy write postgresql databases at work (> 1To disk space) and no corruption happens, never |
08:32 |
VanessaE |
you're still avoiding the question... :) |
08:32 |
nrzkt |
MT load the data in database, if db is corrupted, either MT do'nt start because sqlite3 drive rsee it, or postgresql server doesn't start because it's corrupted |
08:33 |
nrzkt |
and, honestly, if a relational DB is corrupted, it's because your underlayer disk is corrupted, then damages are very important and you should have done backups, it's not possible to not do backups, it's your responsiblity |
08:33 |
VanessaE |
ok, so one broken player record, the whole database is trashed and MT can't load it. |
08:33 |
nrzkt |
no |
08:33 |
nrzkt |
you misunderstand |
08:34 |
nrzkt |
i said, a record cannot be broken by MT, it can only be broken because your server is dying |
08:34 |
VanessaE |
or because some write process crashed. |
08:34 |
nrzkt |
database are more intelligent than MT coredevs |
08:35 |
tenplus1 |
currently having all player files seperately means easier fixing by deleting the broken player files, the new all-in-one database would need to be updated on a daily basis just incase so it's available for recovery, sometimes more than once in a day for all new players |
08:35 |
|
benrob0329 joined #minetest-project |
08:35 |
nrzkt |
write process crash on MT means you have a bug in your sqlite driver |
08:35 |
tenplus1 |
hi ben |
08:35 |
nrzkt |
for postgresql a database crash only mean the last change is lost and you use the previous record version |
08:35 |
VanessaE |
if a world's map.sqlite can on occasion return corrupt mapblocks through no fault of the hardware or sqlite3, then so can a player database. |
08:35 |
VanessaE |
that's all I'm saying. |
08:36 |
VanessaE |
in the case of a map.sqlite, you can at least configure MT to ignore errors (though I forget what it actually does - empty mapblock?). the same should be true of the player database. |
08:36 |
nrzkt |
for DB corruption i want to say, just use a strong database, use postgresql, sqlite3 is lightweight but has some cons, if you don't enable sqlite WAL you are dead i ncase of recent corruption |
08:37 |
VanessaE |
WAL? |
08:37 |
nrzkt |
write ahead logs, it's the write logging for the database, containing the differential changes before including them to the real database storage |
08:37 |
nrzkt |
it permits a point in time recovery if there is a problem with last logs |
08:38 |
VanessaE |
sounds like a journaled file system. |
08:38 |
nrzkt |
yeah it's like this |
08:38 |
VanessaE |
ok. |
08:38 |
VanessaE |
all of this aside, the need for tools to read/edit a player's record will be needed. |
08:39 |
nrzkt |
for postgresql, you have WAL too, but it's not needed because your postgresql storage pool grow in time and postgresql wil lhave dead rows corresponding to last rows version and will recycle them when they are long time dead |
08:39 |
tenplus1 |
having admin commands in-game would be handy for removing player files completely |
08:39 |
VanessaE |
(even if it consists of a few chatcommands like "/player delete <foo> inv_slot 4" or "/player whereis <foo>" or something |
08:39 |
VanessaE |
) |
08:40 |
nrzkt |
an inventory slot cannot be deleted, it can only be empty |
08:40 |
VanessaE |
well I will most definitely not mess with postgres. |
08:40 |
VanessaE |
nrzkt: s/delete/clear/ |
08:40 |
tenplus1 |
yeah... we have /whereis /reset_homepos and /reset_bedpos to help |
08:40 |
VanessaE |
I don't of course mean to remove the slot but merely to empty it |
08:41 |
nrzkt |
homepos is not in minetest core engine |
08:41 |
VanessaE |
anything you can do with a text editor on a player file now, you need to be able to do from within Minetest. |
08:41 |
nrzkt |
it can be interesting but not easy to do within minetest, in SQL it's easier because you do requests directly on data and have the view you need |
08:42 |
nrzkt |
select * from player where name='VanessaE' |
08:42 |
VanessaE |
why not easy? |
08:42 |
nrzkt |
it's not difficult |
08:42 |
tenplus1 |
this would make for a good mod |
08:42 |
nrzkt |
because we will need very huge amout of code for single SQL operations, SQL is a standard language permitting to query easily |
08:42 |
VanessaE |
does not Minetest know that player <foo> has item <bar> in slot 2, and was last at coords x,y,z, and had a,b,c,d, and e, in their crafting grid? |
08:43 |
nrzkt |
it only nows that if player is loaded, but never if player is not connected |
08:43 |
nrzkt |
knows* |
08:43 |
VanessaE |
then it needs to load the player's record from the database when requested. |
08:43 |
nrzkt |
it doesn't do this at this moment for player files |
08:43 |
VanessaE |
I know |
08:44 |
VanessaE |
but it doesn't NEED to either |
08:44 |
VanessaE |
but if you convert the player files into one big monolithic database, then you're gonna need to make it easy for admins to manage its content without resorting to sql commands. |
08:44 |
nrzkt |
i will not do this for this PR, it needs to be discussed with other devs, not only you to define a roadmap about edition, i tend to prefer an external tool |
08:44 |
nrzkt |
delete from player where name = 'VanessaE'; |
08:45 |
nrzkt |
bye bye all VanessaE data :) |
08:45 |
VanessaE |
that means chatcommands in Minetest, simple stuff like above. |
08:45 |
VanessaE |
nrzkt: my experience with other servers' admins is that you can't expect them to remember that |
08:45 |
VanessaE |
let alone more complex queries to change a record. |
08:45 |
nrzkt |
remember one thing, if we add an interface to a deletion command, a mod can delete all your players |
08:45 |
tenplus1 |
not if it requires SERVER priv |
08:45 |
tenplus1 |
which it should |
08:46 |
VanessaE |
a mod doesn't need privs |
08:46 |
nrzkt |
tenplus1: but if i have a rogue mod i installed which hide this call ? |
08:46 |
VanessaE |
he makes a good point, but to be fair, the same is true if mod security is disabled now |
08:46 |
VanessaE |
a mod would execute a 'rm' command and delete a player file |
08:46 |
nrzkt |
yes, but mod security will not prevent a database call :) |
08:47 |
nrzkt |
but yes i can add at first a lua call to remove a disconnected player |
08:47 |
nrzkt |
it's the minimum |
08:47 |
VanessaE |
well that's the responsibility of a server admin to at least know if a mod they install is dangerous. |
08:47 |
nrzkt |
VanessaE: yes ofc |
08:47 |
tenplus1 |
exactly |
08:48 |
nrzkt |
i think our future mod store should look at lua calls to shows the mod perms |
08:49 |
nrzkt |
okay i include a point in this PR for adding a lua call to remove players |
08:49 |
VanessaE |
and anyway, who says you have to open up raw database calls to Lua anyway? |
08:49 |
OldCoder |
So saith the Lord |
08:49 |
VanessaE |
I'm talking simple chat commands (or equivalent). |
08:49 |
tenplus1 |
lol |
08:49 |
nrzkt |
VanessaE: do we have builtin commands ? |
08:50 |
nrzkt |
or are all commands in mods ? |
08:50 |
VanessaE |
nrzkt: I don't know for sure. some I think |
08:50 |
tenplus1 |
for this particular playerfile change the commands should be hidden away in core |
08:50 |
VanessaE |
but we have little or nothing that deals with player records, but again, we don't need them yet. if players are in a database, we will need these things. |
08:51 |
nrzkt |
tenplus1: i can only set them in builtin to respect our model :) |
08:51 |
VanessaE |
and again, I am not suggesting Lua API calls either. just chat commands. |
08:52 |
VanessaE |
just simple stuff to dump a player's record, change something in it (like if they got their hands on an admin pick, I should be able to remove it), etc. |
08:53 |
VanessaE |
it doesn't have to be elaborate. it just has to replace the common reasons one might choose to edit a player file as they exist right now. |
08:53 |
tenplus1 |
+1 |
08:54 |
nrzkt |
dump a player record has no real sense o nyour screen |
08:54 |
VanessaE |
doesn't it? |
08:54 |
VanessaE |
I don't mean to dump the raw data |
08:54 |
VanessaE |
I mean just show me what I might see were this player's record still stored in a flat file. |
08:55 |
nrzkt |
i should look at coreside C++ commands but i don't think we have them, and i don't know if we have pure builtin possible calls, i don't think |
08:56 |
nrzkt |
okay a json styled representation can have some sense |
08:57 |
VanessaE |
not even that much. just some basic representation that looks like it could be a Lua table. |
08:57 |
VanessaE |
keep it simple. |
08:57 |
nrzkt |
it's similar output xD |
08:58 |
VanessaE |
ok |
08:59 |
VanessaE |
"player model"? |
08:59 |
VanessaE |
maybe you meant "player record" :P |
08:59 |
nrzkt |
the player representation |
09:00 |
nrzkt |
not player record is not the good term, as in database you have many records for a single player |
09:00 |
nrzkt |
(1 entry for player, 1 entry for each player inventory, 1 entry for each item in inventories, 1 entry for each player extra attribute) |
09:00 |
VanessaE |
don't call it a model, might be too easy to confuse with "the player model" as in the thing you see walking around in-game :) |
09:00 |
nrzkt |
also SQL permits to do nice things |
09:01 |
nrzkt |
imagine you want to kill all your players |
09:01 |
nrzkt |
UPDATE players set health = 0; |
09:01 |
nrzkt |
it only kills non connected players :p |
09:01 |
VanessaE |
back in the old days, a "record" was a collection of entries related to one thing. |
09:01 |
VanessaE |
a "player record" in that context would mean exactly what you described |
09:06 |
OldCoder |
In the old days, a "record" rotated at 45 rpm |
09:06 |
VanessaE |
anyway, some of us still think in terms of flat files :) |
09:06 |
VanessaE |
OldCoder: if that's the old days, you' |
09:06 |
VanessaE |
you're younger than you seem :P |
09:07 |
tenplus1 |
dunno how many corrupt player files I've deleted to fix an issue |
09:07 |
OldCoder |
VanessaE, I am old enough to be nearly your father. tenplus1, I have a test to delete them on start if corrupted. |
09:07 |
OldCoder |
Saves a lot of trouble |
09:07 |
VanessaE |
... |
09:07 |
* OldCoder |
did have 78s |
09:07 |
OldCoder |
78s were cool |
09:07 |
VanessaE |
man, why do you always have to spoil the joke? |
09:08 |
* OldCoder |
wants his 78s back |
09:08 |
OldCoder |
Because I'm literal |
09:08 |
OldCoder |
It is what it is |
09:08 |
VanessaE |
well stop it. |
09:08 |
* OldCoder |
can't stop what he is |
09:08 |
OldCoder |
tenplus1, files do get corrupted |
09:08 |
OldCoder |
It's a weak spot |
09:08 |
* OldCoder |
had Robin Williams on LP. Probably 33 1/3. |
09:09 |
* OldCoder |
thinks there should be a MT _game that resembles Atari 2600. Good fit. |
09:09 |
* OldCoder |
will add some 78s to his jukebox/radio mod |
09:11 |
VanessaE |
guess I'll have to rewrite my "clear abandoned noob accounts" script. |
09:16 |
nrzkt |
tenplus1: corrupted files are due to minetest process itself being very rude and raw, a database is hopefully more secure :p |
09:16 |
tenplus1 |
the whole auth.txt is messy |
09:17 |
tenplus1 |
and playerfiles in the folder gather and mess with hdd/sdd |
09:18 |
VanessaE |
I understand the reasoning behind moving player data into a single database file, don't get me wrong. |
09:18 |
nrzkt |
tenplus1: when the players will be in database i will work after on auth in database but it's more complicated due to the lua interface |
09:18 |
VanessaE |
I'm just concerned about being able to manipulate it without resorting to sqlite queries that I'll never be able to remember. |
09:18 |
nrzkt |
auth database has another pro: you will share your accounts across your servers :) |
09:18 |
VanessaE |
sure, if you use postgres or redis or something |
09:19 |
nrzkt |
only postgresql & sqlite will be supported for this, redis is a caching system, it has no security, no transactions |
09:19 |
VanessaE |
(sqlite doesn't like to let two processes access one db file) |
09:19 |
nrzkt |
VanessaE: right |
09:19 |
nrzkt |
for sharing you will need postgresql, sqlite is for mono server |
09:24 |
VanessaE |
I...think I'll just let each server have its own databases, like it is now |
09:24 |
VanessaE |
this is already getting too complicated for me. |
09:25 |
tenplus1 |
yup |
09:26 |
|
yusf[m] joined #minetest-project |
09:28 |
* tenplus1 |
tests latest Midori browser |
09:44 |
|
proller joined #minetest-project |
09:59 |
tenplus1 |
https://www.youtube.com/watch?v=a_Wvlgop-bo lol |
10:28 |
tenplus1 |
may I ask a favour of everyone, could you maybe watch and if you like thumbs up this youtube video so it can get 150 likes and Xanadu will appear in the next one: https://www.youtube.com/watch?v=SaSYynSxAuY |
10:35 |
|
proller joined #minetest-project |
11:12 |
|
proller joined #minetest-project |
11:45 |
|
CWz joined #minetest-project |
11:54 |
|
lordfingle joined #minetest-project |
11:54 |
lordfingle |
yo |
11:55 |
tenplus1 |
hi fingle |
11:56 |
lordfingle |
I'm about to upgrade to your latest npc. Thought I'd login to chat for fun before I did so. :) |
11:57 |
lordfingle |
I like playerplus, but I think it probably needs to "remember" the last few nod_stand, nod_foot etc for my walkover event. That way my timer can be lazier |
11:57 |
lordfingle |
Hope that makes sense. |
11:57 |
lordfingle |
I'm trying to reduce the about of rapid globalsteps |
11:58 |
tenplus1 |
I hear you... playerplus checks every 1 seconds and updates global vars |
12:00 |
lordfingle |
yup. mobs_redo does as well. This stresses me out a bit, 'cos if there's too many mods around things start to get very slow |
12:01 |
tenplus1 |
the whole reason we need the engine to handle step checks for players (to reduce getnode's) |
12:02 |
lordfingle |
It would be nice for mobs to have their own thread |
12:03 |
tenplus1 |
if only :P |
12:03 |
lordfingle |
Have you experimented with coroutines at all? |
12:04 |
tenplus1 |
coroutines ???? |
12:04 |
lordfingle |
https://www.lua.org/pil/9.1.html |
12:04 |
lordfingle |
I saw a thread somewhere where some minetest.* calls work with them and some dont |
12:05 |
lordfingle |
https://github.com/minetest/minetest/issues/1709 |
12:05 |
lordfingle |
If everything was magical and wonderful, your docustom could be a coroutine |
12:06 |
tenplus1 |
looks interesting |
12:07 |
lordfingle |
It seems too good to be true, but might be worth investigating IDK |
12:08 |
tenplus1 |
would be nice to have a way to handle threads in mt... have mods/game run a little smoother |
12:09 |
lordfingle |
igor the npc....interesting |
12:09 |
tenplus1 |
heheh, made that one for halloween, has 8 different skind |
12:09 |
tenplus1 |
*skins |
12:09 |
tenplus1 |
many players made a castle and wanted an igor for down below to protect it |
12:10 |
lordfingle |
of course, no castle is complete without an Igor |
12:10 |
tenplus1 |
and if you right click with gold ingot he drops different items than npc |
12:12 |
lordfingle |
I've just realised that although I do have your npc mod, it has been the "peaceful npcs" (jojoa1997) mod that has been populating my world. Perhaps I'll switch it's engine over to mobs redo.... |
12:12 |
lordfingle |
https://github.com/jojoa1997/peaceful_npc |
12:13 |
lordfingle |
Looks like these dudes open doors. Do you open doors? |
12:14 |
tenplus1 |
nope |
12:15 |
tenplus1 |
I coudl add that tothe new pathfinding level2 functions |
12:15 |
tenplus1 |
can already break walls and build up to get to player |
12:15 |
lordfingle |
the code looks very simple (eg https://github.com/jojoa1997/peaceful_npc/blob/master/npc/npc_def.lua) if you're interested |
12:15 |
Calinou |
hi |
12:16 |
tenplus1 |
alreadythere :PPP |
12:16 |
tenplus1 |
hi cal |
12:16 |
lordfingle |
hi! |
12:16 |
lordfingle |
theres some nice models and textures there too |
12:16 |
|
lumidify joined #minetest-project |
12:16 |
tenplus1 |
I only have 2 npc textures (male/female), 8 igor textures and 3 trader textures... |
12:17 |
tenplus1 |
hi lumidify |
12:17 |
lumidify |
Hi tenplus1 |
12:18 |
lordfingle |
people have found peaceful npcs fun to look at, but they're pretty dumb |
12:18 |
lordfingle |
maybe it's time to merge ;) |
12:18 |
|
Jordach joined #minetest-project |
12:18 |
tenplus1 |
redo npc's wander around aimlessly, kill monsters nearby, follow players if owned and attack players who mess with them :D |
12:19 |
Calinou |
https://lut.im/Z9QuoLt4WY/PP6od59RPnhOaNIT.png |
12:19 |
lordfingle |
peaceful npcs are much the same but I havent seen them attack a player. It would be nice if they did |
12:19 |
Calinou |
I heard one of them will even talk to you about sneak glitches |
12:19 |
lordfingle |
lol |
12:20 |
tenplus1 |
which dont exist now... pah! |
12:20 |
lordfingle |
I don't think they know about beloved glitches |
12:20 |
lordfingle |
It's time we moved to implementing MT developers as NPCs |
12:21 |
tenplus1 |
latest daily build has errors in carts |
12:21 |
tenplus1 |
cart_entity.lua:96 unexpected symbol near 'and' |
12:21 |
tenplus1 |
????????????? |
12:21 |
lordfingle |
afk |
12:22 |
behalebabo |
a fix has been pushed, in 5fe831ed139d882b2838cbd75caa1f5a4d9674ad |
12:22 |
Jordach |
Calinou, christ that was dank |
12:22 |
Calinou |
dank, danker, dankest |
12:22 |
tenplus1 |
o.O no way!!!!!! I just checked, I cant sneak along ledges anymore..... dammit!!!!!!!!!!!!!! |
12:23 |
|
RobbieF joined #minetest-project |
12:23 |
tenplus1 |
hi robbie |
12:23 |
tenplus1 |
some of the best mt parkour had ledges to sneak across.... |
12:25 |
behalebabo |
tenplus1: are you testing with the sneak glitch enabled? |
12:26 |
behalebabo |
hmm, should be on by default still anyway |
12:27 |
tenplus1 |
default world with no mods running |
12:27 |
tenplus1 |
shoudl be enabled |
12:30 |
|
benrob0329 joined #minetest-project |
12:30 |
tenplus1 |
hi ben |
12:38 |
* Jordach |
feels like nerfing food but in a good way |
12:39 |
tenplus1 |
lol |
12:39 |
Jordach |
food can still stack |
12:39 |
Jordach |
but foods that require cutlery and such will leave dirty equivilents that need to be cleaned before they can be used again |
12:40 |
Jordach |
these dirty items don't stack at all |
12:41 |
Jordach |
their craftable state stacks |
12:46 |
RobbieF |
benrob0329 - thanks for the note!! Updating now. |
12:46 |
RobbieF |
TPS-Scifi is now up to date. |
12:46 |
tenplus1 |
turns out pathfinding level2 mobs/npc's dont need to open doors, they simple break them down :P |
12:47 |
|
NathanS21 joined #minetest-project |
12:47 |
Jordach |
tenplus1, ayy lmao |
12:49 |
|
proller joined #minetest-project |
13:00 |
|
lumidify joined #minetest-project |
13:02 |
|
Amaz joined #minetest-project |
13:09 |
|
troller joined #minetest-project |
13:22 |
|
troller joined #minetest-project |
13:43 |
|
xtremehacker_ joined #minetest-project |
13:43 |
|
octacian joined #minetest-project |
13:48 |
|
twoelk joined #minetest-project |
13:53 |
|
twoelk left #minetest-project |
13:58 |
|
Amaz joined #minetest-project |
14:08 |
benrob0329 |
Hello |
14:08 |
benrob0329 |
(I leave my phone connected tobi |
14:09 |
benrob0329 |
*to irc) |
14:11 |
|
twoelk joined #minetest-project |
14:15 |
|
Guest84138 joined #minetest-project |
14:41 |
|
sniper338 joined #minetest-project |
14:42 |
|
tenplus1 left #minetest-project |
14:50 |
|
benrob0329 joined #minetest-project |
14:50 |
benrob0329 |
https://vid.me/ffD3 |
14:54 |
nerzhul |
in lua can we do continue keyword ? |
14:54 |
|
Fixer joined #minetest-project |
14:56 |
benrob0329 |
I believe so |
15:10 |
Fixer |
hi |
15:13 |
|
est31 joined #minetest-project |
15:18 |
|
RobbieF joined #minetest-project |
15:19 |
|
Brackston joined #minetest-project |
15:25 |
Fixer |
no tidyman when you need it |
15:40 |
Calinou |
https://www.reddit.com/user/xNotch |
15:40 |
Calinou |
so many awards |
15:42 |
|
troller joined #minetest-project |
15:45 |
Fixer |
Calinou: such much awards |
15:48 |
Calinou |
also, he basically gets lifetime free reddit gold |
15:48 |
Calinou |
as if he wasn't rich enough :( |
15:50 |
Calinou |
:O |
15:50 |
Calinou |
if you have tons of points on a single comment, it shows in "k" units |
15:50 |
Calinou |
https://www.reddit.com/r/OutOfTheLoop/comments/4p5p22/whats_up_with_notch_on_twitter/d4io9q9/ |
15:50 |
Calinou |
17.2k points |
15:50 |
Calinou |
that's the first time I've seen this |
15:51 |
Calinou |
(also, wow, he basically got 6.67 times my lifetime comment karma acquired in 5.5 years, in a single comment) |
16:06 |
|
tenplus1 joined #minetest-project |
16:07 |
tenplus1 |
hi folks |
16:09 |
twoelk |
o/ |
16:19 |
benrob0329 |
Calinou: great words will go a long ways. |
16:19 |
Calinou |
benrob0329: well, sometimes he says stuff like "wat" and gets 9 points |
16:19 |
Calinou |
I can't get away with that :) |
16:19 |
tenplus1 |
lol |
16:20 |
Calinou |
hi tenplus1 :) |
16:20 |
tenplus1 |
hi al |
16:20 |
tenplus1 |
er... hi Cal :P |
16:20 |
* tenplus1 |
thinks he needs a new keyboard soon |
16:20 |
Calinou |
tenplus1: https://hastebin.com/hurixipube.xml some backlog |
16:21 |
* benrob0329 |
looks at comment |
16:21 |
benrob0329 |
Oh, its notch |
16:21 |
benrob0329 |
that's why |
16:21 |
|
troller joined #minetest-project |
16:22 |
tenplus1 |
lol, mansplaining :P |
16:27 |
benrob0329 |
Aka Notch gets upvotes because people go "ohmygosh its notcchhmineceft mustuovoye eaahh" |
16:27 |
tenplus1 |
celebrity status |
16:28 |
benrob0329 |
I mean, notch made one game that's well known |
16:29 |
benrob0329 |
And suddenly he's more of a celebrity than most actors.. |
16:29 |
tenplus1 |
mc is good, but not that good |
16:30 |
benrob0329 |
I don't think even David Tennant could compete |
16:30 |
tenplus1 |
I'd choose Tenant over Notch anyday |
16:30 |
benrob0329 |
Same |
16:31 |
red-004[IRC] |
well it looks like brexit finnaly happened |
16:31 |
tenplus1 |
got there in the end :P |
16:31 |
benrob0329 |
what's that? |
16:32 |
tenplus1 |
britain(UK) leaving the EU (european union) |
16:32 |
red-004[IRC] |
they finnaly invoked article 50 |
16:32 |
benrob0329 |
Ah, that thing |
16:32 |
benrob0329 |
Feel any different? |
16:33 |
tenplus1 |
nah, things going on as usual |
16:33 |
red-004[IRC] |
you have to wait for about 2 years now for something to happen |
16:33 |
benrob0329 |
sounds like every "big thing " |
16:33 |
benrob0329 |
Its all this hype and then |
16:33 |
benrob0329 |
Nothing happens |
16:34 |
red-004[IRC] |
well it's not a short term thing |
16:34 |
tenplus1 |
asshole companies are using brexit as an excuse to hike up prices of their products |
16:34 |
* benrob0329 |
remebers 2013 well |
16:35 |
benrob0329 |
*2012 |
16:35 |
red-004[IRC] |
what happened in 2012? |
16:35 |
benrob0329 |
The world was supposed to end |
16:35 |
tenplus1 |
lolol |
16:35 |
benrob0329 |
Or at least that's what crazy people thought |
16:35 |
red-004[IRC] |
that was pretty funny |
16:36 |
benrob0329 |
And then it was actualy going to be *14 or somethin |
16:36 |
benrob0329 |
'14 |
16:36 |
tenplus1 |
and 2000 was bad for padanoia as well, I had to fix all of the computers in the comany I worked at :) |
16:37 |
benrob0329 |
Now we have to obvoid 2038 |
16:37 |
red-004[IRC] |
lets wait for 2038 now |
16:37 |
benrob0329 |
Lets not wait |
16:37 |
benrob0329 |
Lets just fix the issue |
16:38 |
tenplus1 |
hopefully by that time most pc's will be using the 64-bit date formats |
16:38 |
tenplus1 |
that'll do us for a few trillion years :D |
16:39 |
benrob0329 |
If we're still using the unix timestamp at that point, that'd be incredible |
16:39 |
tenplus1 |
think we'll still be using xorg ? or will we migrate to wayland/mir :PPPPPPPPehehe |
16:40 |
benrob0329 |
I want a UTS watch though |
16:40 |
red-004[IRC] |
and windows 10 |
16:41 |
red-004[IRC] |
or will mircosoft give up on their current idea |
16:41 |
benrob0329 |
no, I rea |
16:41 |
benrob0329 |
lly don't want windows 10 |
16:41 |
benrob0329 |
Dang mobile keyboards |
16:41 |
tenplus1 |
google will prolly have their own Os by that time :D |
16:41 |
benrob0329 |
ChromeOS |
16:42 |
red-004[IRC] |
CarOS |
16:42 |
red-004[IRC] |
android |
16:42 |
benrob0329 |
BrainOS |
16:42 |
benrob0329 |
aIOS |
16:43 |
garywhite |
MS SucksOS |
16:43 |
tenplus1 |
ehehe |
16:44 |
benrob0329 |
Break the WindOS |
16:52 |
|
nerzhul joined #minetest-project |
16:52 |
tenplus1 |
wb nrz |
16:53 |
nerzhul |
ty |
17:01 |
|
twoelk|2 joined #minetest-project |
17:02 |
|
Krock joined #minetest-project |
17:02 |
|
Krock joined #minetest-project |
17:02 |
Krock |
o/ tenplus1 |
17:02 |
Krock |
HA |
17:03 |
tenplus1 |
hi Krock :) |
17:03 |
|
Grandolf joined #minetest-project |
17:03 |
tenplus1 |
was checking out "Ghost in the Shell" trailers :D |
17:04 |
tenplus1 |
hi Grandolf |
17:05 |
Grandolf |
hi tenplus1 |
17:05 |
Krock |
tenplus1, hehe. must do that aswell. it looks promising |
17:06 |
tenplus1 |
1995 anime was amazing, and this does look great :) might just go to theatre for this one |
17:08 |
nore |
^ did someone here see it already? |
17:08 |
tenplus1 |
hi nore :) |
17:08 |
nore |
(I don't know when it is scheduled to be released, lol) |
17:08 |
nore |
hi tenplus1 :) |
17:08 |
tenplus1 |
am sure it's today ?! |
17:08 |
nore |
oh, ok :) |
17:09 |
nore |
then my question still holds :) |
17:09 |
tenplus1 |
https://www.youtube.com/watch?v=hu-sZ3ra2Os |
17:09 |
* nore |
only saw the anime as well |
17:11 |
nore |
tenplus1: I'm wondering, are you French? |
17:11 |
* nore |
noticed tenplus1 writes spaces before question marks |
17:11 |
tenplus1 |
scottish |
17:13 |
nore |
hmmm |
17:13 |
nore |
I thought it was only while writing french that we added spaces before question marks :D |
17:14 |
benrob0329 |
10+1 I never would've guessed |
17:15 |
tenplus1 |
https://www.youtube.com/watch?v=vZsG7WJVZv8 <--- +100 |
17:19 |
Krock |
can confirm that the 1995 anime was good. sadly the newer GitS: Arise episodes weren't too great |
17:19 |
Krock |
hello twoelk |
17:21 |
twoelk |
o/ |
17:22 |
|
kaeza joined #minetest-project |
17:22 |
* twoelk |
is in office ircstealth mode |
17:23 |
tenplus1 |
hi kaeza |
17:23 |
kaeza |
o/ tenplus1 |
17:31 |
Fixer |
hi tenplus1 |
17:31 |
Fixer |
tenplus1: your stairs redo seems a bit outdated, does not have that efficient slabs placing and no place slab on slab to convert to full block |
17:32 |
tenplus1 |
yeah, I kept rotation simple on purpose :D |
17:33 |
tenplus1 |
and players decorate with slab on slab so i disabled full block conversion on purpose |
17:33 |
tenplus1 |
when fine_pointing is implemented I'll redo it again |
17:33 |
Fixer |
tenplus1: it sucks, i miss efficient slab placing, need to rotate every damn slab |
17:33 |
tenplus1 |
lol |
17:35 |
tenplus1 |
using base/side/top placement and using shift lets you place in a variety of angles already... |
17:36 |
sofar |
fixer: did we ever merge that slab placement changes in mtg? |
18:00 |
|
Shara joined #minetest-project |
18:00 |
tenplus1 |
hi Shara |
18:01 |
benrob0329 |
Moreblocks slabs should be default slabs if possible imo |
18:01 |
tenplus1 |
??? |
18:02 |
benrob0329 |
they're not currently |
18:03 |
tenplus1 |
which slabs ? moreblocks has a few |
18:03 |
benrob0329 |
1/2 |
18:04 |
benrob0329 |
same for fulk stairs |
18:04 |
sofar |
send a patch |
18:04 |
benrob0329 |
*full |
18:05 |
benrob0329 |
Doesnt moreblocks use looping to define each node? |
18:06 |
tenplus1 |
yeah, the stairsplus mod that's included with it loops through all solid blocks and makes them into all kinds of stairs.... I dont like it :P |
18:08 |
benrob0329 |
Perhaps an if statement could be added to check if node if from default, if so skip half slab and ful stair def |
18:08 |
benrob0329 |
*full |
18:12 |
tenplus1 |
nope, dont like it... so slow registering all the stairs :)) would rather have control over which nodes I make into stairs |
18:15 |
Fixer |
sofar: same to same slab works nice, another PR for different slabs is not merget yet i think |
18:16 |
sofar |
"so slow registering all the stairs" |
18:16 |
sofar |
man, do you even performance? |
18:16 |
sofar |
seriously, what, you've profiled it? |
18:16 |
tenplus1 |
my humble net-top feels the lag and slowness of any mod :D |
18:17 |
sofar |
if (performance_claim) then |
18:17 |
Fixer |
sofar: he uses crap-crap-top |
18:17 |
sofar |
get_hard_numbers_from_claimant() |
18:17 |
sofar |
end |
18:17 |
tenplus1 |
ehehe |
18:17 |
sofar |
tenplus1: show me your performance numbers |
18:17 |
sofar |
especially |
18:17 |
sofar |
please, show me the numbers AFTER the startup has completed |
18:17 |
tenplus1 |
put it this way, if I see "initializing nodes" during connect then there's an issue :P |
18:18 |
sofar |
that's bullshit |
18:18 |
sofar |
registration is already done when you see that |
18:18 |
sofar |
"initializing nodes" is just the client receiving the node defs |
18:18 |
sofar |
which means it's already done registering nodes |
18:19 |
sofar |
which means you just made a nonsense performance claim |
18:19 |
tenplus1 |
I watch it countiny up the percentage dude... |
18:19 |
tenplus1 |
*counting |
18:19 |
sofar |
I'm not attacking you, I'm attacking your nonsense claim |
18:19 |
sofar |
when you see the percentage bar, you are seeing network packets with node definitions |
18:19 |
sofar |
you're not seeing the nodes getting registered |
18:19 |
benrob0329 |
here's sonething related to GitS https://youtu.be/hYViV3nFlDA |
18:20 |
sofar |
registration of all those nodes is like, microseconds of time |
18:20 |
benrob0329 |
I thought it wad an interesting video at least |
18:20 |
tenplus1 |
then why does it make me wait while it counts up the % ? |
18:20 |
tenplus1 |
to me that's slow |
18:20 |
Fixer |
registrations are microseconds, i agree |
18:20 |
sofar |
tenplus1: go ahead and print os.time before and after node registration lines |
18:21 |
sofar |
tenplus1: go, do it, teach yourself something |
18:21 |
nerzhul |
VanessaE, are you around ? |
18:21 |
Fixer |
maybe media loading affects it more |
18:21 |
tenplus1 |
could be fixer |
18:21 |
nerzhul |
and/or tenplus1 |
18:21 |
sofar |
textures are loaded |
18:21 |
sofar |
sounds |
18:21 |
tenplus1 |
it's registering around 20 stair types for each node... |
18:21 |
sofar |
and it needs to construct the client copy of the node defs |
18:21 |
benrob0329 |
10+ Im pretty sure mobs redo is way worse on proformance than stairs+ |
18:22 |
|
kaeza joined #minetest-project |
18:22 |
sofar |
benrob0329: performance claim detected, please provide data and method |
18:22 |
benrob0329 |
*Performance |
18:22 |
tenplus1 |
redo only registers slab;stair;corner;invcorner;slope |
18:22 |
tenplus1 |
wb kaeza |
18:22 |
benrob0329 |
No no no, I didn't make a performance claim |
18:23 |
|
diemartin joined #minetest-project |
18:23 |
twoelk |
wouldn't it be usefull if a dev-server with connection to dev-irc existed that had public performance logs? |
18:23 |
sofar |
<benrob0329> 10+ Im pretty sure mobs redo is way worse on proformance than stairs+ |
18:23 |
benrob0329 |
I made a porformance claim :P |
18:23 |
sofar |
guilty |
18:23 |
sofar |
well, it could have been about aestetics :) |
18:26 |
nerzhul |
who have one of serverlist big server and is around, and have access to player files on disk please , |
18:26 |
benrob0329 |
My data is called perceived lag, my method is called broken eyeballs |
18:26 |
benrob0329 |
Plugins: glasses |
18:31 |
tenplus1 |
stairsplus time: 1.3585 stairs redo time: 0.1459 |
18:31 |
red-004[IRC] |
how do I build a release version again? |
18:31 |
* VanessaE |
wanders in |
18:31 |
tenplus1 |
hi Vanessa |
18:32 |
VanessaE |
hi |
18:32 |
red-004[IRC] |
hi |
18:32 |
nerzhul |
VanessaE, nice :) |
18:32 |
nerzhul |
can you help me VanessaE ? |
18:32 |
VanessaE |
possibly |
18:33 |
twoelk |
carefull VanessaE, he's after your collected data :-P |
18:33 |
nerzhul |
no :) |
18:34 |
|
troller joined #minetest-project |
18:34 |
nerzhul |
VanessaE, can you duplicate a server world dir into a temp world dir |
18:34 |
nerzhul |
a huge player files directory |
18:34 |
VanessaE |
I have backups already :) |
18:34 |
nerzhul |
and run a minetest master server + https://patch-diff.githubusercontent.com/raw/minetest/minetest/pull/5475.patch patch |
18:34 |
VanessaE |
(my servers do that every night) |
18:34 |
nerzhul |
nice, can you use one of your huge backups to test the migration processus please ? |
18:35 |
VanessaE |
however, perhaps I should just send a zip of them to you, I'm not inclined to fiddle with my MT build |
18:35 |
nerzhul |
i tested on some files, but on my server i'm using postgresql DB since 2 years and i havent them anymore :) |
18:35 |
nerzhul |
VanessaE, if you want, send me in private |
18:37 |
nerzhul |
just player files, i don't want auth |
18:37 |
nerzhul |
i don't need map either :) |
18:37 |
sofar |
tenplus1: just timed all of `default`'s initialization time (ALL node registrations, crafts, items) |
18:37 |
sofar |
tenplus1: takes 0.016s |
18:38 |
sofar |
tenplus1: do it yourself - edit init.lua and print os.clock() above and below all the dofile() lines |
18:38 |
|
benrob0329 joined #minetest-project |
18:38 |
tenplus1 |
I did, and already posted the times above when registering nodes ^^^^ |
18:38 |
tenplus1 |
stairsplus time: 1.3585 stairs redo time: 0.1459 |
18:38 |
Fixer |
homedecor:plasma_lamp |
18:38 |
VanessaE |
nerzhul: the zip in /msg goes with https://daconcepts.com/vanessa/hobbies/minetest/worlds/Building_World.tar.gz if you need the actual world too. |
18:38 |
benrob0329 |
Dangit, mobile irc is a pain |
18:39 |
* benrob0329 |
watches latest cat5 episode |
18:39 |
sofar |
tenplus1: 1sec startup time isn't going to make any client connect slower |
18:39 |
nerzhul |
VanessaE, no i don't care about world being playable, i only want to have correct data in database :) |
18:39 |
nerzhul |
with various freshness, it's the most important |
18:39 |
sofar |
the only people who take the 1sec penalty is you when you restart the server |
18:39 |
Fixer |
his players run minetest on Commodore 64 |
18:39 |
tenplus1 |
:PPP |
18:39 |
sofar |
irrelevant |
18:40 |
VanessaE |
hey now, don't knock the c64 |
18:40 |
sofar |
even on a c64 client, you would not EVER take the INIT penalty |
18:40 |
sofar |
the client simply doesn't RUN that lua code |
18:40 |
tenplus1 |
that was just default nodes tho... IF it had all nodes available on server it would take even longer |
18:40 |
sofar |
the node registration is NOT done on the client, period |
18:41 |
tenplus1 |
did I say it was ? |
18:41 |
Jordach |
afternoon, sofar |
18:41 |
sofar |
the time to connect to a server is related to the *amount* of nodes, not the *way they are registered in lua* |
18:41 |
benrob0329 |
Hi Jordach |
18:42 |
tenplus1 |
hi Jordach |
18:42 |
sofar |
<tenplus1> put it this way, if I see "initializing nodes" during connect then there's an issue :P |
18:42 |
benrob0329 |
Then get gigabit internet :P |
18:42 |
sofar |
no |
18:42 |
sofar |
use less nodes on the server |
18:43 |
sofar |
each node is a network packet I think |
18:43 |
sofar |
so if you have 500 nodes defined, 500 packets |
18:43 |
benrob0329 |
Oh, I had an idea the other day |
18:43 |
VanessaE |
sofar: what, seriously? |
18:43 |
benrob0329 |
what if nodes had states? |
18:43 |
VanessaE |
O_o |
18:43 |
benrob0329 |
Like, on or off |
18:44 |
twoelk |
I'm not sure you are completely correct with the amount sofar, as that might be influenced by clever coding in Lua iirc |
18:44 |
benrob0329 |
Or on fire |
18:44 |
Fixer |
sofar: and 1 packet size is how much bytes? |
18:44 |
benrob0329 |
ect rather than changing the node all the time |
18:44 |
sofar |
depends on the size |
18:44 |
sofar |
most minetest packets are 150-300 bytes |
18:45 |
sofar |
benrob0329: then if you change the node state, all nodes on the map would change at the same time |
18:45 |
benrob0329 |
Nit necessarily |
18:45 |
sofar |
all nodes on the map share nodedef |
18:45 |
benrob0329 |
If the state was per node it wouldn't |
18:45 |
sofar |
(of the same type) |
18:45 |
sofar |
there already is per-node state |
18:45 |
sofar |
1) param1/param2 |
18:46 |
sofar |
2) node metadata |
18:46 |
benrob0329 |
I mean being able to define extra states |
18:46 |
benrob0329 |
Like, the client woukd display the node based on the state |
18:46 |
sofar |
nodemeta |
18:46 |
sofar |
meta set nodedef, actually |
18:47 |
benrob0329 |
but without swapping the node |
18:47 |
benrob0329 |
Like, if I wanted to have a lever |
18:47 |
benrob0329 |
I would set up two states |
18:47 |
sofar |
https://github.com/minetest/minetest/pull/1118 |
18:47 |
benrob0329 |
One would display an on mesh, the other an off mesh |
18:48 |
nerzhul |
nice 1 minute to migrate 5254 players from files to sqlite |
18:48 |
sofar |
nerzhul: nice |
18:48 |
benrob0329 |
But as to which mesh, that'd be decided by the client |
18:48 |
sofar |
uh |
18:48 |
sofar |
nerzhul: 1 minutes? |
18:48 |
nerzhul |
i'm logging with one player file |
18:48 |
sofar |
nerzhul: that's all player attributes and inventories? |
18:48 |
nerzhul |
yes, maybe it's due to dumping all named on screen :p but i don't think |
18:48 |
sofar |
benrob0329: that's what that PR would allow |
18:49 |
nerzhul |
inventory is good, i don't have mod it's unk but it works :p |
18:49 |
sofar |
change texture, change model, change visual |
18:49 |
twoelk |
nerzhul: can I use 3rd party tools to view the data or is it all in zipped blobs? |
18:49 |
sofar |
zip is open standard |
18:49 |
sofar |
so, easily viewed in external programs |
18:49 |
nerzhul |
it's not a zip it's a database |
18:50 |
nerzhul |
all is relational, no serialization |
18:50 |
benrob0329 |
sofar then I eagerly await it :-) |
18:50 |
sofar |
yes but is it lz/zlib compressed? |
18:50 |
nerzhul |
no |
18:50 |
nerzhul |
why should i compress a field containing a number ? :p |
18:50 |
twoelk |
ever tried to study the map-database? it's pretty much a pita |
18:50 |
nerzhul |
it's a pure relational db |
18:50 |
nerzhul |
it's not map database, for map it's normal to serialize |
18:50 |
nerzhul |
mapblocks are blob data |
18:50 |
nerzhul |
inventory and players are relational data |
18:51 |
sofar |
nerzhul: agreed compression isn't that useful |
18:51 |
nerzhul |
54s sofar for this test on migration reducing screen output :) |
18:51 |
nerzhul |
i have a SSD but everyfile is in my computer disk cache |
18:51 |
sofar |
well, it's migration |
18:51 |
sofar |
so one time problem |
18:51 |
nerzhul |
ofc |
18:51 |
sofar |
no use optimizing it much more |
18:51 |
sofar |
5000 per minute isn't that bad |
18:51 |
nerzhul |
sofar, the main problem i have, how we communicate with players before merge ? |
18:51 |
nerzhul |
no :) |
18:52 |
sofar |
maybe print progress every 100 migrated? |
18:53 |
nerzhul |
i print progress for each permitting to see failure with some points on screen, it'snot very important and the output is verbose i'ts important |
18:53 |
nerzhul |
44012 inventories |
18:53 |
nerzhul |
255308 items |
18:53 |
sofar |
afk a bit |
18:53 |
nerzhul |
and only 4 metadata VanessaE boo |
18:53 |
nerzhul |
:p |
18:53 |
nerzhul |
you don't use player metadata :p |
18:53 |
* benrob0329 |
is watching cat5 |
18:54 |
twoelk |
so I could use for example this to edit entries https://www.etl-tools.com/database-browser/overview.html |
18:54 |
benrob0329 |
"Please gives us 5 satures review if you feeling everything good" |
18:54 |
VanessaE |
nerzhul: I do not use it in any of my mods, that I am aware of. |
18:54 |
Fixer |
*,....,* |
18:54 |
nerzhul |
on windows yes |
18:55 |
nerzhul |
on unix i just use sqlite3 <file> and lets go |
18:55 |
Fixer |
twoelk: or this http://sqlitebrowser.org/ |
18:55 |
nerzhul |
and for postgresql => pgadmin3 |
18:55 |
twoelk |
Fixer: yeah, got that one too ;-P |
18:55 |
nerzhul |
i use psql myself, but pgadmin3 to edit your pg database is a professional tool used by hundred of postgresql users around world |
18:56 |
nerzhul |
erf we don't have function to move files in core, i should have a portable function :p |
18:57 |
nerzhul |
oh we have rename, it seems it's good |
18:57 |
Jordach |
nerzhul, >mt engine doesn't have any useful file handling at all |
18:57 |
Jordach |
nice game engine m8 |
18:58 |
* Jordach |
sips lipton |
18:58 |
Jordach |
but that's none of my business |
18:58 |
nerzhul |
Jordach, what do you mean by useful file handling ? the API ? |
18:58 |
Fixer |
lipton pepe |
18:58 |
Jordach |
nerzhul, i meant as part of core doing something for core |
18:59 |
nerzhul |
the API is portable, but i think we should finish ShadowNinja part and rename fs:: namespace functions to stdc++17 norm for non c++17 builds :) |
19:00 |
benrob0329 |
theres a cpp '17 standard? |
19:03 |
tenplus1 |
I wonder how different it is from standard C |
19:04 |
Krock |
^ * C++ |
19:05 |
nerzhul |
benrob0329, yes, there are many new interesting features, like filesystem api portability for example |
19:05 |
nerzhul |
it's the last huge portability not covered by C++11 |
19:05 |
|
paramat joined #minetest-project |
19:05 |
tenplus1 |
hi paramat |
19:05 |
nerzhul |
C++11 has standard mutexes and threads, the only remaining part was filesystem access |
19:06 |
nerzhul |
but there are other interesting features in C++17, but we should use C++11 at first in MT, it can improve a little bit core engine performance if we use constexpr correctly |
19:06 |
nerzhul |
we already use unordered containers in MT and it imrpove ~15% performance on our map containers usage |
19:06 |
tenplus1 |
nice |
19:07 |
tenplus1 |
every little helps :D |
19:07 |
Krock |
Krock quit: https://img-9gag-fun.9cache.com/photo/aBW6rYZ_700b.jpg |
19:07 |
tenplus1 |
lolol |
19:09 |
benrob0329 |
There was an issue on Urho3D that I saw one time, someone was saying that the patch needed non c++11 support as well |
19:09 |
benrob0329 |
And I was like, why is c++11 still an issue? |
19:09 |
benrob0329 |
Its 6 friggen years old |
19:09 |
tenplus1 |
you would think all the issues would have been sorted by now |
19:10 |
benrob0329 |
Its like, what are you using? GCC 3? |
19:12 |
Fixer |
a.out format |
19:14 |
|
garywhite joined #minetest-project |
19:14 |
benrob0329 |
a.out is the unix equivalent of a 16 bit dos executable |
19:14 |
tenplus1 |
wb gary |
19:14 |
garywhite |
thanks |
19:17 |
red-004[IRC] |
requiring code to work with c++98 is like requiring support for windows xp |
19:18 |
tenplus1 |
linux for example should have repo's compiled for the latest versions of the os and code |
19:19 |
tenplus1 |
am still surprised to see software depending on python2 for example |
19:20 |
benrob0329 |
like blender... |
19:20 |
Fixer |
tenplus1: gimp still gtk2, gtk3 is ---- :trollface: nowhere to runq |
19:21 |
* benrob0329 |
hates the fact that blender uses python2 |
19:21 |
tenplus1 |
python3 was released 2008 so you'd think software using 2.x would be updated |
19:21 |
benrob0329 |
But hey, Tcl/Tk is the future! |
19:21 |
benrob0329 |
Ya know? |
19:21 |
tenplus1 |
whassat ? |
19:22 |
Fixer |
tenplus1: nah |
19:22 |
benrob0329 |
tenplus1: Exactly |
19:22 |
Fixer |
tenplus1: 95% of popular packages already support Py3 |
19:22 |
tenplus1 |
good good, 5% to go... |
19:24 |
Fixer |
tenplus1: http://py3readiness.org/ |
19:27 |
tenplus1 |
mpv and samba are newer releases and you'd think they would be on 3 by now |
19:27 |
|
nanovad joined #minetest-project |
19:27 |
tenplus1 |
hi nano |
19:29 |
garywhite |
It takes time to write new versions of software. After all, it takes Microsoft 4-5 years to throw rubbish together for Windows |
19:30 |
OldCoder |
<benrob0329> But hey, Tcl/Tk is the future! |
19:30 |
OldCoder |
Before Java, Tcl/Tk was a primary cross-platform GUI SDK |
19:31 |
benrob0329 |
Its still around, but barely |
19:31 |
OldCoder |
People, dozens of projects, had struggled to get their toolkits to the top |
19:31 |
OldCoder |
Right |
19:31 |
OldCoder |
Tcl/Tk, a bit simple and blocky (like MT) was arguably the most successful |
19:31 |
OldCoder |
Tcl is an interesting language. But, no, the combination is no longer the future. |
19:32 |
OldCoder |
http://oldcoder.org/general/beworld/browse-beworld.html |
19:32 |
OldCoder |
^ This is what Tcl looks like |
19:32 |
tenplus1 |
what larger programs use tcl? |
19:32 |
OldCoder |
That one |
19:32 |
OldCoder |
It is the world's largest... |
19:32 |
nerzhul |
benrob0329, you are inexact, C++11 is fully supported only since GCC 5.1 |
19:32 |
OldCoder |
single-script Tcl game (even excluding data) |
19:32 |
OldCoder |
About 100 printed pages |
19:32 |
nerzhul |
and clang 3.6 |
19:33 |
OldCoder |
tenplus1, wish to see *my* pre-MT mobs? :-) |
19:33 |
tenplus1 |
okies |
19:33 |
OldCoder |
http://oldcoder.org/general/beworld/beworld-main.jpg |
19:33 |
OldCoder |
Behold the complexity and wonder of Tcl Mobs |
19:33 |
OldCoder |
:P |
19:33 |
tenplus1 |
ehehe cute |
19:33 |
OldCoder |
The pig oinks, the car honks, and the cow moos! |
19:33 |
OldCoder |
O_O |
19:34 |
OldCoder |
Basically, the setup is Nethack and those are GUI Nethack characters |
19:34 |
OldCoder |
In Tcl; excuse the remarks, but it isn't a bad language |
19:34 |
tenplus1 |
kewl |
19:34 |
benrob0329 |
OGL in Tcl?? |
19:34 |
benrob0329 |
Dang |
19:34 |
OldCoder |
All languages are good. Except C++ and, well, Whitespace and its friends. |
19:35 |
benrob0329 |
I'm legitimately impressed |
19:35 |
OldCoder |
Object Oriented Tcl actually |
19:35 |
OldCoder |
Read the source before commenting |
19:35 |
OldCoder |
Take nothing on faith |
19:35 |
* OldCoder |
says this knowing the code is good |
19:35 |
* tenplus1 |
pulls out Cobal |
19:35 |
OldCoder |
http://oldcoder.org/general/beworld/beworld-caspak.jpg |
19:35 |
* benrob0329 |
pulls out Pascal |
19:35 |
OldCoder |
COBOL * |
19:35 |
tenplus1 |
oopsie |
19:35 |
tenplus1 |
pascal was good, I made my first gui lib on that one |
19:36 |
OldCoder |
Pascal was a good learning language |
19:36 |
OldCoder |
It was primarily Java that killed it |
19:36 |
OldCoder |
If you had Java, which was Super Pascal Plus Classes, whither Pascal? |
19:36 |
OldCoder |
<Pascal> Time to retire |
19:37 |
benrob0329 |
Modern pascal is still a pretty cool toolchain |
19:37 |
OldCoder |
I have a few programs that use it |
19:37 |
benrob0329 |
With Lazarus n everything |
19:37 |
tenplus1 |
Delphi :P |
19:37 |
OldCoder |
Right |
19:37 |
OldCoder |
There is a ZIP program based on Lazarus |
19:37 |
OldCoder |
MT could have gone a different direction... C++ plus Lua is pretty good, but... |
19:38 |
benrob0329 |
Its funny, java was the big thing not too long ago |
19:38 |
OldCoder |
What if a MT domain language... |
19:38 |
OldCoder |
Hm? benrob0329, it was considered the future |
19:38 |
benrob0329 |
But now its just faiding away |
19:38 |
OldCoder |
As in buy our Java books |
19:38 |
OldCoder |
Or die |
19:38 |
OldCoder |
Ruby had its moment, too |
19:38 |
nerzhul |
VanessaE, i added core.remove_player call to remove them from database :) |
19:38 |
OldCoder |
benrob0329, well Oracle has mishandled it |
19:38 |
nerzhul |
and /remove_player <name> call |
19:38 |
tenplus1 |
good one nrz |
19:38 |
nerzhul |
i just need to finish the call |
19:38 |
paramat |
sofar new slab placement in mtg is waiting for engine #5485 (fine pointed) i agree having a slab auto-rotate to vertical when placed against a wall is bad, usually it is wanted horizontal, so the new slab code is good |
19:38 |
benrob0329 |
Slowly dying, very slowly |
19:39 |
nerzhul |
to look if player is connected (it's not very important atm) and handle the exit code |
19:39 |
|
ssieb joined #minetest-project |
19:39 |
VanessaE |
nerzhul: that's a start. |
19:39 |
OldCoder |
benrob0329, this is probably off-topic if others are discussing MT. But, I'll say that Java has quite an ecosystem. |
19:40 |
paramat |
hi tenplus1 could you describe what you mean with 'can't sneak across a ledge' here https://github.com/minetest/minetest/issues/5438 so we can try to fix it? |
19:40 |
benrob0329 |
But for all that ecosystem it isn't doing very well atm, but maybe thays just cause I'm out of the java dev loop |
19:41 |
benrob0329 |
As far as widely used apps on the desktop |
19:41 |
paramat |
there was a fix for sneaking on slabs last night |
19:42 |
paramat |
tenplus1 have you opened an issue about mob staticdata? |
19:42 |
tenplus1 |
I rewrote those fucntons, is ok... using 0.4.15 dev stuff |
19:43 |
tenplus1 |
before with sneak you could walk across a ledge, a gap in a wall, but with sneak changes you cant anymore |
19:43 |
benrob0329 |
Ninja sneak is dank sneak |
19:43 |
paramat |
please describe or screenshot 'gap in a wall' |
19:44 |
paramat |
so there is no mob staticdata bug in MT? it was your own code that was buggy? |
19:44 |
* twoelk |
wonders wether npc dat could be stuffed into an external database to have persistant npcs with a history and memory |
19:45 |
paramat |
do you mean this? "you can't sneak walk from one node to another it the space is only one node high" |
19:46 |
tenplus1 |
mob staticdata worked on 0.4.15 stable then all of a sudden it didnt on latest devs... I rewrote functions and it started working again |
19:46 |
paramat |
ok please can you help us fix official MT by opening an issue? |
19:47 |
paramat |
maybe just a link to your own code? |
19:47 |
* sofar |
peeks back in (missing backlog) |
19:47 |
benrob0329 |
We had a talk about old progamming languages, amd sneal features came up again |
19:47 |
benrob0329 |
*sneak |
19:49 |
paramat |
tenplus1 the functions you rewrote, were they in your mods or part of MT? |
19:50 |
paramat |
we need to know if there is a bug in MT |
19:51 |
Fixer |
fps drop bug reproduced, good-good |
19:51 |
tenplus1 |
if I come across it again I'll post an issue |
19:51 |
Fixer |
7a4878 |
19:53 |
Fixer |
interesting |
19:53 |
paramat |
so there is no bug in MT? which functions did you rewrite? |
19:54 |
paramat |
tenplus1 please be clear and answer the questions |
19:56 |
tenplus1 |
https://github.com/minetest/minetest/issues/5438 pic added |
19:56 |
paramat |
thanks for the comment |
19:56 |
nerzhul |
okay player to database now only required one remaining point, migrate from sqlite to pg and invert |
19:56 |
nerzhul |
that's nice |
19:57 |
nerzhul |
and some communication before merge i think |
19:59 |
OldCoder |
Communication is good |
20:00 |
Fixer |
nerzhul: finally database, no more thousands of player files, some popular servers had like 20 000 accounts, and that was damn slow to backup |
20:00 |
Fixer |
some people will dislike yet again |
20:01 |
Fixer |
2baa191 time |
20:01 |
paramat |
tenplus1 does current official MT have a bug with mob staticdata? |
20:01 |
nerzhul |
Yeah, i backported my fork changes which are in production sice a long time :p |
20:01 |
|
tenplus1 left #minetest-project |
20:02 |
Fixer |
nerzhul: there is one more thing that annoys me, it is debatable, but /cache folder with thousands of files |
20:03 |
Fixer |
my cache has 33578 files, size 254 629 kbytes |
20:04 |
benrob0329 |
it'd be cool if mods could kick a server to anothet server |
20:04 |
benrob0329 |
*kick a player |
20:04 |
|
electrodude512 joined #minetest-project |
20:04 |
paramat |
good grief =/ |
20:05 |
paramat |
he makes a huge fuss about a bug then dodges any questions about it |
20:05 |
nerzhul |
Fixer, the cache contain a sha1 of each file downloaded before |
20:05 |
nerzhul |
it's not very optimized yes |
20:06 |
Fixer |
nerzhul: yep, it is just pain in the ass to move or copy since moving 30 000 takes ages %) |
20:06 |
Fixer |
if it is one 250 mb file - copy/move is like 3 seconds, if 30 000 small files - 3 minutes %) |
20:06 |
Fixer |
but thats ok |
20:06 |
Fixer |
people rarely move it |
20:07 |
nerzhul |
just drop your cache :p |
20:08 |
twoelk |
I often move the cache when I update mt, just as fovorite-server file and other data |
20:08 |
twoelk |
helps a lot on slow internet |
20:15 |
benrob0329 |
Anyone have any ideas for keeping some files synced between desktop and laptop? |
20:15 |
benrob0329 |
Both would be running Linux |
20:27 |
sofar |
rsync or git |
20:29 |
sofar |
there's plenty other alternatives, of course |
20:30 |
nerzhul |
nextcloud client ? |
20:34 |
benrob0329 |
I want to sync things like my .minetest folder |
20:34 |
benrob0329 |
but rsync would probably work |
20:34 |
RobbieF |
i gotz a better suggestion |
20:35 |
RobbieF |
(not that I'm a linux guru or anything) |
20:35 |
paramat |
~tell tenplus1 ok i assume there is no mob staticdata bug in MT. when you wrote 'i rewrote functions' i needed to check whether you were rewriting parts of builtin or a subgame, in which case work would be needed in MT or MTG |
20:35 |
ShadowBot |
paramat: O.K. |
20:35 |
benrob0329 |
Hi RobbieF!) |
20:35 |
RobbieF |
Hi :) |
20:35 |
RobbieF |
This is the answer: https://www.cis.upenn.edu/~bcpierce/unison/ |
20:38 |
benrob0329 |
"Sasha what are you up to?" 'I'm fidgeting!' |
20:38 |
benrob0329 |
Me: XD |
20:40 |
* benrob0329 |
bookmarks that link |
20:42 |
OldCoder |
<benrob0329> Anyone have any ideas for keeping some files synced between desktop and laptop? |
20:42 |
OldCoder |
Not familiar with unison. Some admins and developers just use rsync. |
20:43 |
OldCoder |
Hm, moment |
20:43 |
OldCoder |
https://github.com/OldCoder/rsyncit |
20:43 |
OldCoder |
benrob0329, ^ that is a useful rsync wrapper |
20:44 |
OldCoder |
In Perl, the language of the Gods :-) |
20:44 |
benrob0329 |
XD |
20:44 |
OldCoder |
https://imgs.xkcd.com/comics/lisp.jpg |
20:45 |
red-004[IRC] |
the alt text is great |
20:45 |
benrob0329 |
Oh so that's how it only took God 6 days |
20:45 |
benrob0329 |
He just hacked together a dirty perl script |
20:48 |
OldCoder |
Yep |
20:49 |
Shara |
Hello all |
20:53 |
OldCoder |
o/ |
20:55 |
Fixer |
hi |
20:56 |
benrob0329 |
I kinda want to hop on an anarky server and just lag it the heck out with a mesecons clock XD |
21:00 |
Shara |
Hi OldCoder, hi Fixer |
21:01 |
OldCoder |
Shara, moin |
21:36 |
Fixer |
current mood: 281. 猫 ã‚· Corp. - SECRETS OF MY AGE [with ãƒãƒ¼ãƒžãƒ³Roman] [3:07] |
22:28 |
Fixer |
reproduced the bug again :( |
22:30 |
Fixer |
2baa191 |
22:53 |
|
troller joined #minetest-project |
23:00 |
|
DI3HARD139 joined #minetest-project |
23:06 |
|
Nathans21 joined #minetest-project |
23:08 |
octacian |
I keep getting an error that says "table index is nil" |
23:11 |
octacian |
I have local meta_contexts = {} then I try to set meta_contexts[name] to something, and I get the above error |
23:12 |
behalebabo |
what's name? |
23:12 |
behalebabo |
that'll be the nil value |
23:12 |
octacian |
You think? |
23:12 |
octacian |
I'll take a look. It's supposed to be the player name. |
23:12 |
behalebabo |
it's the only thing you're indexing there, unless the right-hand of the = also indexes |
23:13 |
octacian |
it was nil >_< |
23:13 |
octacian |
Didn't realize it as the issue actually goes quite a ways back in my code |
23:18 |
OldCoder |
It's often advisable to each pieces for nil as an expression is constructed |
23:18 |
OldCoder |
Either returning or logging an error message as soon as nil is detected |
23:59 |
|
benrob0329 joined #minetest-project |