| Time |
Nick |
Message |
| 00:16 |
|
SpeedProg joined #minetest-dev |
| 00:24 |
Sudi |
celeron55_: hey here is an example of what the event system looks like over the network |
| 00:25 |
Sudi |
http://ideone.com/7M8UeO |
| 00:31 |
Sudi |
celeron55_: ? |
| 00:33 |
Sudi |
hmm is he gone? |
| 00:33 |
VanessaE |
it should be right around 1:30am his time I think |
| 00:34 |
VanessaE |
make that 2:34am |
| 00:36 |
Sudi |
i see |
| 02:30 |
|
dannydark left #minetest-dev |
| 05:17 |
|
VanessaE joined #minetest-dev |
| 05:23 |
|
VanessaE[L] joined #minetest-dev |
| 05:36 |
hmmmm |
yeow |
| 05:36 |
hmmmm |
i just got minetest to use 9gb of ram |
| 05:36 |
hmmmm |
and then it crashed |
| 05:36 |
hmmmm |
no idea what on earth happened |
| 06:02 |
|
rsiska joined #minetest-dev |
| 06:21 |
|
SitDog joined #minetest-dev |
| 06:23 |
SitDog |
Upgraded to a real chat client :) |
| 06:48 |
|
SpeedProg joined #minetest-dev |
| 08:33 |
|
SitDog_ joined #minetest-dev |
| 08:37 |
|
SitDog_ joined #minetest-dev |
| 08:48 |
|
Calinou joined #minetest-dev |
| 09:03 |
|
SitDog__ joined #minetest-dev |
| 09:05 |
|
SitDog__ joined #minetest-dev |
| 10:25 |
|
darkrose joined #minetest-dev |
| 10:25 |
|
darkrose joined #minetest-dev |
| 10:35 |
|
kotolegokot joined #minetest-dev |
| 10:35 |
|
kotolegokot joined #minetest-dev |
| 11:07 |
|
Sudi joined #minetest-dev |
| 11:19 |
|
SpeedProg joined #minetest-dev |
| 14:00 |
Sudi |
hey celeron55_ this is the event system i talked abou in action over the network |
| 14:00 |
Sudi |
http://ideone.com/CwUhBO |
| 14:04 |
celeron55_ |
yes, and...? |
| 14:05 |
Sudi |
you wanted to know how it would handle events over the network |
| 14:05 |
Sudi |
thats how |
| 14:06 |
Sudi |
you can of course also bind class methods to events |
| 14:13 |
celeron55_ |
doesn't look too special; just a regular C++ event system in a lazy "Doer" naming style |
| 14:18 |
Sudi |
true |
| 14:18 |
Sudi |
but it automaticly dispatches the events to the correct receivers |
| 14:18 |
Sudi |
on a nother host |
| 14:20 |
celeron55_ |
which is, like, 20 lines of code to implement? 8) |
| 14:21 |
celeron55_ |
i see it does use that funky templated way of making a class know the type of itself |
| 14:21 |
Sudi |
then i must have done something wrong |
| 14:21 |
Sudi |
its faster then dynamic cast |
| 14:22 |
celeron55_ |
what does the implementation code look like that calls serialize() and throws it through the tubes? |
| 14:22 |
Sudi |
bc that is seriously more than 20lines of code :P |
| 14:24 |
Sudi |
celeron55_: on the lowest level http://ideone.com/w52iKj |
| 14:25 |
celeron55_ |
that makes me want to see AppFrame::actions::Message_ |
| 14:25 |
Sudi |
haha |
| 14:26 |
celeron55_ |
you're like a human ctags; almost as good but way slower 8) |
| 14:27 |
Sudi |
thats all pretty old code |
| 14:27 |
Sudi |
have to find it |
| 14:27 |
celeron55_ |
why have you created it in the first place |
| 14:28 |
Sudi |
i am using it in my personal projects |
| 14:28 |
Sudi |
here |
| 14:28 |
Sudi |
http://ideone.com/b53o9q |
| 14:28 |
Sudi |
the custom id feature is not used by the system |
| 14:29 |
Sudi |
that was for consistency over savegames |
| 14:54 |
|
doserj joined #minetest-dev |
| 16:00 |
|
Calinou joined #minetest-dev |
| 16:06 |
|
serengeor joined #minetest-dev |
| 16:22 |
|
PilzAdam joined #minetest-dev |
| 16:41 |
|
rsiska joined #minetest-dev |
| 16:43 |
|
hmmmm joined #minetest-dev |
| 17:07 |
|
hmmmm joined #minetest-dev |
| 17:28 |
|
sstrandberg joined #minetest-dev |
| 17:35 |
|
rubenwardy joined #minetest-dev |
| 17:53 |
|
Sudi joined #minetest-dev |
| 18:42 |
|
SpeedProg joined #minetest-dev |
| 18:52 |
|
SitDog joined #minetest-dev |
| 20:00 |
Taoki |
celeron55: I tested your 043_support branch. Compiled it and ran a local server with two clients connected to it. Then with the client, I connected to a 0.4.3 server and everything worked well. Didn't spend time on for long but I could see players moving aroung and nothing bad happened |
| 20:00 |
Taoki |
Hope that helps make up for the mistake with changing those ID's :P |
| 20:00 |
Taoki |
Again, let me know if you need something else tested with that, else I can delete the zip with your code and wait for it to go upstream |
| 20:16 |
|
rubenwardy left #minetest-dev |
| 20:30 |
|
SpeedProg1 joined #minetest-dev |
| 20:44 |
|
VanessaE[L] joined #minetest-dev |
| 21:13 |
celeron55 |
so, i created this thing: https://github.com/celeron55/minetest/blob/exp/src/util/template_serialize.h#L340 https://github.com/celeron55/minetest/blob/exp/src/util/template_serialize.cpp#L22 |
| 21:14 |
celeron55 |
which is used like this https://github.com/celeron55/minetest/blob/exp/src/nodedef.cpp#L336 |
| 21:14 |
celeron55 |
and i want comments from C++ programmers who understand it |
| 21:15 |
celeron55 |
it would be used for serializing stuff, to get better backward and forward compatibility and make taoki less prone to mistakes |
| 21:16 |
Taoki |
celeron55: Now that I know what I did wrong, I won't do things like mixing ID's again. |
| 21:16 |
Taoki |
But a consistent system for the packages (and well marked one) is always good, so I approve otherwise |
| 21:17 |
celeron55 |
i'm not really talking specifically about you, but about everybody who don't exactly know what they're doing |
| 21:19 |
celeron55 |
also that kind of a format would help those who know what they are doing too, because it's less error-prone |
| 21:20 |
celeron55 |
anwyay; i am interested if somebody here feels qualified enough to tell how that particular one would be bad, and what to possibly do instead |
| 21:22 |
Taoki |
BTW, I'm seeing another bug on VanessaE's server. Some torches are using the lua particles, but some smoke poofs are frozen in place and just sitting there |
| 21:22 |
Taoki |
Newly spaend particles seem to be good tho |
| 21:22 |
VanessaE |
I see that happen with all kinds of entities. |
| 21:22 |
VanessaE |
carts, smoke, slimes were doing it too. |
| 21:22 |
Taoki |
yeah. Might be a small problem with particles |
| 21:23 |
celeron55 |
when do they stop, and when do they start updating again? |
| 21:24 |
VanessaE |
I haven't seen a predictable pattern. |
| 21:24 |
Taoki |
I can't tell. I just see some currently frozen in place, but also new poes which work normally |
| 21:24 |
Taoki |
No stuck ones disappearing and no new ones getting stuck currently |
| 21:24 |
VanessaE |
/clearobjects still gets rid of them, but that kills off carts too. |
| 21:25 |
Taoki |
Did you just do a /clearobjects on the server? Since I still see them |
| 21:25 |
Taoki |
Unless that's a client feature |
| 21:25 |
VanessaE |
no. |
| 21:25 |
VanessaE |
I did not. |
| 21:25 |
Taoki |
ok. Can you try one now to be sure? If you're ok with carts disappearing |
| 21:25 |
VanessaE |
sure. |
| 21:26 |
VanessaE |
executed. |
| 21:26 |
Taoki |
I still see the broken smoke puffs |
| 21:26 |
VanessaE |
command completed. |
| 21:26 |
Taoki |
ah no, done |
| 21:26 |
VanessaE |
they disappeared for me. |
| 21:26 |
Taoki |
It cleared them. And funny thing... it also made Sokomine move again |
| 21:26 |
Taoki |
Jump to a far location |
| 21:26 |
VanessaE |
ok |
| 21:26 |
Taoki |
Now I see her moving for the first time in 15 minutes... I thought she was afk |
| 21:27 |
VanessaE |
teleported to 1000^3 |
| 21:28 |
celeron55 |
can any of you see any errors in the client log, or server log? |
| 21:29 |
VanessaE |
lemme check |
| 21:29 |
Taoki |
sorry, not having a console open for my client |
| 21:29 |
Taoki |
but saw nothing odd in chat |
| 21:29 |
VanessaE |
What am I looking for in the log? |
| 21:29 |
Taoki |
anyway, since that clearobjects all is good |
| 21:30 |
Taoki |
BTW. Is it also a bug that when I die and respawn, I'm only taken back to the spaen point sometimes? Half of the time I respawn in the same spot |
| 21:31 |
Taoki |
So if I die in lava, I respawn and re-die there a few times until I'm also teleported back |
| 21:31 |
VanessaE |
Taoki: old bug. |
| 21:31 |
Taoki |
ok |
| 21:31 |
Taoki |
ah, bad smoke puffs again |
| 21:32 |
Taoki |
haha! They have a bounding box and I can punch them |
| 21:32 |
Taoki |
they even flash when punched |
| 21:32 |
celeron55 |
"What am I looking for in the log?" <- you are looking at the case-insensitive string "error" |
| 21:32 |
Taoki |
killed the smoke buff by punching it many times :D |
| 21:32 |
Taoki |
*puff |
| 21:32 |
VanessaE |
celeron55: 16:10:42: ERROR[ServerThread]: ServerEnvironment::deactivateFarObjects(): id=39 |
| 21:32 |
VanessaE |
m_static_exists=true but static data doesn't actually exist in (-5,0,7) |
| 21:33 |
VanessaE |
and one about me overriding the gameid. that's it for "error" messages. |
| 21:33 |
celeron55 |
that's an old bug, doesn't matter |
| 21:33 |
celeron55 |
well, throw me the server address, i'll build master (once again...) and come see |
| 21:33 |
Taoki |
Smoke puff defeated! exp earned, 0.01 :P |
| 21:33 |
VanessaE |
ok |
| 21:33 |
VanessaE |
24.181.193.82 port 30000 |
| 21:34 |
VanessaE |
has a fair number of mods but you knew that. |
| 21:46 |
VanessaE |
celeron55: if you would like to investigate that player texture glitch, the mod I'm using to provide it is sdzen's player_textures, here: https://github.com/sdzen/player_textures |
| 21:46 |
VanessaE |
I do not know if failing to load a newly-added texture without a server restart is a bug in the engine or the mod. |
| 21:48 |
celeron55 |
it's not a glitch |
| 21:48 |
celeron55 |
it's minetest generating a placeholder texture on purpose when it doesn't have one |
| 21:48 |
VanessaE |
yes, but the file is there. |
| 21:48 |
celeron55 |
you didn't restart the server |
| 21:49 |
VanessaE |
yep I know |
| 21:49 |
celeron55 |
so you know exactly what happens; why do you call it a glitch |
| 21:49 |
VanessaE |
the reason I ask is, shouldn't the server just not bother looking for the file if it can't actually load it? |
| 21:49 |
celeron55 |
the server isn't looking for it |
| 21:49 |
celeron55 |
the client is |
| 21:49 |
VanessaE |
ahhhhh |
| 21:49 |
VanessaE |
gotchya |
| 21:53 |
celeron55 |
this load time is painful |
| 21:53 |
VanessaE |
yup. |
| 21:54 |
celeron55 |
i guess now this more like hanged completely |
| 21:54 |
VanessaE |
taoki and others seem to be doing ok save for lagging a bit |
| 21:54 |
VanessaE |
had to get sdzen un-stuck by forcing a respwn - I think he was "attached" to sokomine, who appears to be afk |
| 21:55 |
Taoki |
All I know is the server is laggy and slow for some reason. Might either be a slow connection to VanessaE or some mod causing bandwidth to hand |
| 21:55 |
Taoki |
**hang |
| 21:56 |
VanessaE |
CPU usage of the server is minimal (about 10% of the core it's running on), bandwidth is always good (30Mbps down, 3Mbps up, and I actually get those speeds on long, sustained transfers) |
| 21:57 |
VanessaE |
(well, 10 to 25% it seems, probably due to mods) |
| 21:57 |
celeron55 |
how many items does the server have? |
| 21:57 |
VanessaE |
lesse.. |
| 21:58 |
celeron55 |
1416 are visible in the creative inventory; i wonder how many there are that don't get put there |
| 21:58 |
VanessaE |
damm it what was that string to search for |
| 21:58 |
VanessaE |
dammit* |
| 21:58 |
celeron55 |
or, well, 14** |
| 21:59 |
VanessaE |
creative inventory size: 1395 |
| 22:00 |
celeron55 |
there are 1741 media files |
| 22:01 |
VanessaE |
sounds about right; most of the stuff in my mods have more than one texture per node |
| 22:01 |
celeron55 |
hmm, there are roughly 2000 items |
| 22:01 |
celeron55 |
quite exactly actually |
| 22:02 |
Sudi |
you should maybe think about texture maps |
| 22:02 |
Sudi |
bc texture switches are expensice operations |
| 22:02 |
Taoki |
nice one: If you right click as block like the chest to get a menu, if someone removes that block while its menu is still open for you, the emnu turns into a black background |
| 22:02 |
Taoki |
I mean only the background is left |
| 22:03 |
Taoki |
nticed that with the rail chest |
| 22:03 |
celeron55 |
holy shit this lags now |
| 22:03 |
celeron55 |
everybody is walking at a static position 8D |
| 22:25 |
VanessaE |
celeron55: yes, we see you chatting just fine |
| 22:26 |
celeron55 |
your server doesn't seem to be able to get anything through to this client anymore, except replies to minetest's ping packets |
| 22:27 |
celeron55 |
now it started to |
| 22:32 |
celeron55 |
the chat seems to have a lag of like 2 minutes 8D |
| 22:33 |
celeron55 |
or even more |
| 22:33 |
VanessaE |
ouch |
| 22:33 |
VanessaE |
I changed the server step as requested btw |
| 22:33 |
celeron55 |
did you say it in the chat? |
| 22:33 |
VanessaE |
yes |
| 22:33 |
celeron55 |
haven't seen that yet |
| 22:33 |
VanessaE |
"done." |
| 22:33 |
celeron55 |
last thing i saw was MK going brb |
| 22:34 |
VanessaE |
I did it about 2 minutes after you asked |
| 22:34 |
VanessaE |
17:30:41: ACTION[ServerThread]: CHAT: <MirceaKitsune> brb in 5 |
| 22:34 |
VanessaE |
17:30:52: ACTION[ServerThread]: CHAT: <celeron55> vanessaE: try setting dedicated_server_step to 0.2 |
| 22:34 |
VanessaE |
17:32:48: ACTION[ServerThread]: CHAT: <VanessaEzekowitz> celeron55: done. |
| 22:34 |
celeron55 |
ha, there it came |
| 22:36 |
celeron55 |
minetest has one funny feature: the size of the output buffer is infinite |
| 22:37 |
celeron55 |
of the network stack |
| 22:37 |
celeron55 |
it can throttle the output and just grow the buffer infinitely and make an infinite lag on a reasonable RTT 8) |
| 22:37 |
VanessaE |
I wonder if that's necessarily a good thing :-) |
| 22:37 |
celeron55 |
ah, now this works perfectly |
| 22:37 |
celeron55 |
the 10/s object position send interval was too much |
| 22:38 |
celeron55 |
ehm... now nothing comes through again |
| 22:38 |
VanessaE |
engine was choking off it's ...haha |
| 22:38 |
Taoki |
The chat in MT is dead, is it the lag? |
| 22:38 |
VanessaE |
70+ KB/sec outgoing now, which is a damn sight better than before btw |
| 22:38 |
VanessaE |
well it was, now ~18 |
| 22:39 |
celeron55 |
that much outgoing and nothing coming to me? |
| 22:39 |
VanessaE |
yup |
| 22:39 |
celeron55 |
that's a nice connection you have there |
| 22:39 |
VanessaE |
that's UDP for ya :) |
| 22:39 |
Taoki |
no, chat isn't dead, everyone was just silent :P |
| 22:40 |
celeron55 |
i hereby declare this server completely useless :P |
| 22:40 |
VanessaE |
I hereby declare your network code crappy :D |
| 22:40 |
celeron55 |
fix it then |
| 22:41 |
celeron55 |
or re-evaluate it's crappiness! |
| 22:41 |
VanessaE |
wish I could :( |
| 22:41 |
celeron55 |
does your server host anything else than minetest |
| 22:41 |
VanessaE |
no. |
| 22:41 |
celeron55 |
because i'd like to try pulling something via tcp from it |
| 22:41 |
VanessaE |
fwiw, everyone else is doing okay, chat is active, 5-6 seconds for people to answer, etc. |
| 22:42 |
VanessaE |
how would I go about making such available? |
| 22:42 |
VanessaE |
(I rarely run servers so I'm not well versed in that stuff. Not since 10 years now) |
| 22:42 |
VanessaE |
maybe scp a file from my machine to you? |
| 22:43 |
Taoki |
funny thing: Chat is VERY fast and nearly instant. But everything else is lagging horribly |
| 22:46 |
VanessaE |
celeron55: see msg. |
| 22:46 |
VanessaE |
Taoki: weird. |
| 22:47 |
Taoki |
anyway, is anything still being tested? If not I think I'll log off from the server |
| 22:47 |
Taoki |
mmaybe try relogging once more if i catch anything special |
| 22:47 |
VanessaE |
not actively, per se. |
| 22:48 |
Taoki |
ok |
| 22:48 |
VanessaE |
heh, I was just about to restart the server. |
| 22:49 |
celeron55 |
seems to pull >200kB/s |
| 22:50 |
Taoki |
it's ok, I'm out |
| 22:50 |
VanessaE |
that's about right |
| 22:50 |
celeron55 |
even 300 |
| 22:50 |
VanessaE |
I can sustain 380 kB/sec via rsync to more local servers |
| 22:51 |
VanessaE |
well, local-ish |
| 22:51 |
VanessaE |
(somewhere in the USA anyways) |
| 22:52 |
celeron55 |
god i hate dolphin when it does this... 80% CPU usage for no reason, and it does not have session save/restore |
| 22:53 |
celeron55 |
and i have 13 tabs open |
| 22:55 |
celeron55 |
while true; do killall kio_thumbnail; sleep 0.1; |
| 22:55 |
celeron55 |
nice fix |
| 22:55 |
celeron55 |
+done |
| 22:55 |
celeron55 |
any higher sleep and it will make that pop up for an infinite time, using all it can get |
| 22:56 |
celeron55 |
or, well, pops it up for such a time that it has time to actually use CPU 8) |
| 22:57 |
celeron55 |
(and would just suck it until the end of the world) |
| 22:57 |
celeron55 |
at least minetest isn't this bad |
| 22:58 |
VanessaE |
heh |
| 22:58 |
celeron55 |
it a) saves sessions, b) nothing else even matters after that, goddamnit |
| 22:58 |
VanessaE |
lol |
| 22:59 |
celeron55 |
in software development, there are two options |
| 22:59 |
celeron55 |
the first one is to not make bugs |
| 23:00 |
celeron55 |
the second one is to make the software painless to restart |
| 23:00 |
celeron55 |
best ones do both |
| 23:00 |
|
serengeor joined #minetest-dev |
| 23:00 |
celeron55 |
dolphin does neither |
| 23:03 |
|
VanessaE[L] joined #minetest-dev |
| 23:04 |
|
Doc22 joined #minetest-dev |
