| Time |
Nick |
Message |
| 00:02 |
Fixer |
paramat: it is worse-st NY ever, we should embrace it :] |
| 00:04 |
Fixer |
when you expect that NY mood with snow and stuff and you got shitty moisty +10*C shitty cloudy weather, climate change is a myth |
| 00:07 |
Shara |
It's warmer than you expect... and climate change is a myth? Not seeing your logic. |
| 00:14 |
sofar |
extreme weather changes are excepected as a result of global warming as demonstrated by weather models, including NOAA's weather model, the one Trump is in charge of. |
| 00:17 |
IhrFussel |
The majority of US people refuses to accept that global warming is a thing last I heard |
| 00:18 |
IhrFussel |
Or...not sure if it was majority or just a large part |
| 00:21 |
rubenwardy |
http://climatecommunication.yale.edu/visualizations-data/ycom-us-2016/?est=happening&type=value&geo=county |
| 00:25 |
IhrFussel |
That's what I meant actually...most Americans think the global warming is NOT mainly caused by humans which is total bollocks to think |
| 00:26 |
rubenwardy |
that's not true |
| 00:26 |
rubenwardy |
look at the scale |
| 00:26 |
rubenwardy |
52% think it's caused by humans |
| 00:26 |
rubenwardy |
34% think it's natural |
| 00:26 |
rubenwardy |
18% are unsure |
| 00:26 |
rubenwardy |
34% is rather a large percentage, but not "most" |
| 00:27 |
rubenwardy |
according to that source, at least |
| 00:27 |
rubenwardy |
people that do surveys based on climate change are more likely to care about it |
| 00:30 |
IhrFussel |
True, in Germany it's 60% believe it's (mainly) human-made an 16% don't |
| 00:31 |
IhrFussel |
Wait, 16% don't believe in it at all* |
| 00:35 |
Fixer |
Shara: it was sarcastic |
| 00:36 |
Shara |
When every other word is "shitty" it just sounds grouchy :) |
| 00:36 |
Fixer |
rubenwardy: it can be both, natural and man-made |
| 00:37 |
rubenwardy |
yes |
| 00:37 |
Fixer |
Shara: when you have rain with +10*C, seriously, send Santa to ISIS |
| 00:37 |
rubenwardy |
the question is about which one it is mostly |
| 00:37 |
Fixer |
yeah |
| 00:37 |
Fixer |
but it is happening |
| 00:37 |
Shara |
Fixer: snow is practically a myth here :) |
| 00:38 |
|
EvergreenTree joined #minetest-hub |
| 00:38 |
Fixer |
now it seems like not a big problem, but in far future exsessive heat may melt some arctic ice may cause unwanted ocean level rise |
| 00:39 |
Shara |
Well, some news would have you believe an ice age will magically come and buy us some more time... but that idea seems to do the rounds every so often |
| 00:40 |
Shara |
And of course, we're already in one |
| 00:41 |
shivajiva |
Trump could always nuke North Korea and cause a nuclear winter if he goes overboard, that would help :) |
| 00:45 |
shivajiva |
the petrodollar would rise to an all time high and the energy companies could rip us off with impunity |
| 00:46 |
shivajiva |
oh wait...that bit already happened |
| 00:48 |
Fixer |
ocean level rise is particularly nasty |
| 00:48 |
Fixer |
since so much population lives on low areas and in coastal areas |
| 00:49 |
Fixer |
shivajiva: useful oil will raise in price, very probably, due to scarcity, and its consumption will be hugely decreased |
| 00:49 |
Fixer |
problem is chemistry needs oil for a lot of stuff |
| 00:50 |
Fixer |
now you just stupiditly burning it |
| 00:50 |
Fixer |
there will be change in the future, and not everyone will like it, and not everyone will buy new iphone every year, it is not sustainable in current way |
| 00:52 |
Fixer |
it is pretty much: most of population does not care, business as usual, problems will be solved/mitigated when it is already fucked up too much, standard of living may decline (it is already declining in US) |
| 00:53 |
Fixer |
maybe fusion reactor will solve energy problems, and our big roads will be electrified with those electrobuses everywhere |
| 00:53 |
Fixer |
it is not ready yet, but they are getting closer |
| 00:54 |
Fixer |
wonder if world moves back to coal ;) |
| 00:54 |
Fixer |
when natural gas and oil is too expensive, business as usual goes horrible way instead of investing into fusion, cinical a bit |
| 00:55 |
Fixer |
but again, probably 80% of population thinks: i will be fine, future generations - good luck, you are on your own |
| 00:56 |
Fixer |
remember, GDP must flow, i mean grow |
| 01:02 |
sofar |
goddamn ukrainian VPNs |
| 01:02 |
sofar |
thousands of SASL logins on my dovecot |
| 01:03 |
sofar |
grep 91.200.12 /var/log/mail.log|wc -l |
| 01:03 |
sofar |
28563 |
| 02:04 |
|
jas_ left #minetest-hub |
| 02:12 |
|
EvergreenTree joined #minetest-hub |
| 03:36 |
|
EvergreenTree joined #minetest-hub |
| 04:06 |
|
skotos joined #minetest-hub |
| 04:38 |
|
Sokomine joined #minetest-hub |
| 04:50 |
|
redblade7 joined #minetest-hub |
| 04:53 |
Sokomine |
redblade7 asked me something related to server operations. i pointed to this channel as this is the right place for such things. hope that's ok? |
| 06:41 |
|
lumberJ joined #minetest-hub |
| 07:59 |
|
sniper338 joined #minetest-hub |
| 08:00 |
|
nerzhul joined #minetest-hub |
| 09:07 |
|
Krock joined #minetest-hub |
| 10:56 |
|
DI3HARD139 joined #minetest-hub |
| 11:25 |
|
aerozoic joined #minetest-hub |
| 11:31 |
|
tenplus1 joined #minetest-hub |
| 11:32 |
tenplus1 |
hi folks |
| 11:32 |
tenplus1 |
Bonmeal mod updated to 0.7 - coral recipe for bonemeal added, api addition, can also grow papyrus and cactus now |
| 11:33 |
Raven262 |
Hi ten |
| 11:34 |
tenplus1 |
hi raven :) |
| 11:34 |
Raven262 |
I had an idea, for improving toolranks trough a new mod. |
| 11:34 |
tenplus1 |
ooh, do tell ? |
| 11:35 |
Raven262 |
What if the tool would count how many of one node type the user has dug? |
| 11:35 |
Raven262 |
For example, I can make my pickaxe count how many mese I've dug. |
| 11:35 |
tenplus1 |
would be a rather large list considering how many nodes exist in default and using mods |
| 11:36 |
Raven262 |
And thats why there is an other part of my idea. |
| 11:36 |
Raven262 |
This counting of nodes can only be achieved if the user attaches a special gem to the tool. |
| 11:36 |
Raven262 |
Which is already set to count a node type. |
| 11:37 |
Raven262 |
SO not all nodes can be counted, but the important ones such as ores can. |
| 11:37 |
Raven262 |
*So |
| 11:37 |
Raven262 |
And this also adds space for some fancy crafting recipes. |
| 11:37 |
Raven262 |
The gem info is stored in the metadata, changes no texture, just the description. |
| 11:37 |
Raven262 |
I might try to make this^ |
| 11:38 |
tenplus1 |
you could make it as an addition to existing toolranks mod |
| 11:38 |
Raven262 |
Nah, I'm more into making a gems mod with many fancy features such as this. |
| 11:38 |
tenplus1 |
once you craft a new tool it uses a different function to dig/count |
| 11:39 |
|
skotos joined #minetest-hub |
| 11:39 |
tenplus1 |
also remember that on use doesnt supply node name, so to do this you could have to use get_node each time you dig, and with 20+ players on server it'll lag somewhat |
| 11:39 |
tenplus1 |
just to check node names for special ores |
| 11:40 |
Raven262 |
Hmm, right. |
| 11:50 |
tenplus1 |
would be handy if the on_use function could supply node information as well, would save mods having to get it |
| 11:50 |
tenplus1 |
and speed things up :D |
| 11:55 |
|
xerox123- joined #minetest-hub |
| 11:55 |
tenplus1 |
hi xerox |
| 12:20 |
|
IhrFussel joined #minetest-hub |
| 12:20 |
tenplus1 |
hi fussel |
| 12:21 |
IhrFussel |
Ok I'm crazy..I bought the Intel Xeon dedi even though it's pretty expensive ... compiling took 45 secs which is very nice but single core performance is the important part |
| 12:21 |
IhrFussel |
Hello tenplus1 |
| 12:22 |
aerozoic |
hi tenplus1 |
| 12:22 |
tenplus1 |
o/ aerozoic |
| 12:30 |
|
xerox123_ joined #minetest-hub |
| 12:35 |
Shara |
Hi all |
| 12:35 |
|
xerox123_ joined #minetest-hub |
| 12:35 |
|
xerox123_ joined #minetest-hub |
| 12:40 |
IhrFussel |
Will cancel the dedi again, up to 1.3 sec lag when I fly around on the map ... that's not much better than my old dedi |
| 12:41 |
Shara |
Fussel, have you tested both all your mods on a new map, and without the mods you are using right now? |
| 12:41 |
tenplus1 |
o/ shara |
| 12:42 |
IhrFussel |
Shara, I'm testing an empty world now with just MTG and lag already spikes to 1+ sec at times ... that is a very bad sign |
| 12:43 |
Shara |
Initial mapgen can spike lag, though I'm not sure if it should go that high |
| 12:43 |
tenplus1 |
wait, did you say you have a single core ? |
| 12:44 |
Shara |
I ran RC on a single core for a long time without problems. |
| 12:44 |
Shara |
Two cores currently. |
| 12:44 |
IhrFussel |
tenplus1, the dedi has 4 cores but htop displays 8 (hyperthreading?) |
| 12:44 |
tenplus1 |
ah ok, have you set num_emerge_cores = 4 in minetest.conf to assist mapgen |
| 12:45 |
tenplus1 |
sorry, num_emerge_threads = 4 |
| 12:45 |
IhrFussel |
All those cores don't help much when MT still only uses mainly 1 core of any machine... I checked a snglecore benchmark for the Intel Xeon and compared it to my AMD Opteron and the benchmark said it would be twice as fast...I don't see it |
| 12:45 |
tenplus1 |
and: server_dedicated = true |
| 12:45 |
IhrFussel |
I set that to 3, doesn't help |
| 12:46 |
VanessaE |
IhrFussel: especially if you have *any* mods that hook into the mapgen |
| 12:46 |
tenplus1 |
true, voxelmanip will slow things down... also hi Vanessa |
| 12:46 |
VanessaE |
hi |
| 12:46 |
IhrFussel |
I have zero mods running outsie of MTG right now...let me check max_lag after just standing on the map for a few minutes |
| 12:46 |
VanessaE |
well I meant Lua will force it down to 1 thread |
| 12:47 |
IhrFussel |
0.7 ... garbage |
| 12:47 |
Shara |
If you have only MTG running on a new map and still don't like the performance, there is not much to say. |
| 12:47 |
IhrFussel |
I will NOT pay $35/month for that crap |
| 12:47 |
tenplus1 |
what OS on server ? |
| 12:48 |
IhrFussel |
Ubuntu 16.04 |
| 12:49 |
IhrFussel |
Intel(R) Xeon(R) CPU E3-1230 V2 @ 3.30GHz << NOT recommended for MT |
| 12:49 |
tenplus1 |
ah okie... have you tried adding 'noatime' to partitions in /etc/fstab to reduce writes to drive, that helps somewhat |
| 12:49 |
tenplus1 |
also setting vm.swappiness = 10 at end of etc/sysctl.conf helps too |
| 12:50 |
IhrFussel |
It's the CPU that spikes when I fly around...it spikes to 50-90% |
| 12:51 |
tenplus1 |
what gpu you using ? |
| 12:51 |
IhrFussel |
I cannot tolerate this when several players move on the map the lag will be just as bad as on my old dedi ... cancelling before they charge me |
| 12:51 |
IhrFussel |
Don't dedis always have integrated GPUs? |
| 12:51 |
Shara |
Once map is generated, it's rare to get many players generating new map all at once |
| 12:51 |
tenplus1 |
true |
| 12:52 |
Shara |
And your map is already well developed, so your concern should be if normal map loading, not new map generation, causes lag |
| 12:52 |
IhrFussel |
Shara, even after standing on the same spot for 5 minutes only reduces max_lag to 0.7 |
| 12:52 |
Shara |
With only MTG? |
| 12:52 |
IhrFussel |
Yes |
| 12:52 |
tenplus1 |
how many players connected ? |
| 12:52 |
IhrFussel |
Only me |
| 12:52 |
tenplus1 |
ssd ? |
| 12:53 |
|
xerox123_ joined #minetest-hub |
| 12:53 |
|
xerox123_ joined #minetest-hub |
| 12:53 |
IhrFussel |
No, the dedi got a 1 TB HDD but it's not the HDD ... there is nothing to load/save from/to HDD when only 1 player stands somewhere |
| 12:53 |
tenplus1 |
you;d be surprised what gets accessed in the background at times |
| 12:54 |
|
EvergreenTree joined #minetest-hub |
| 12:54 |
tenplus1 |
have you tried using leveldb map instead of sql ? |
| 12:54 |
tenplus1 |
hi evergreen |
| 12:55 |
Shara |
I never saw any gain in performance from leveldb |
| 12:55 |
tenplus1 |
wait, are you using Ubuntu with unity gui ? |
| 12:56 |
IhrFussel |
When I compare this dedi#s lag times to my old dedi's lag times I see only a minor difference...a difference that's not worth paying $35 for ... I will just cancel this machine and stay on the VPS for now (the VPS can manage 0.15 max_lag with up to 5 players at once) |
| 12:56 |
tenplus1 |
prolly easier for now |
| 12:56 |
IhrFussel |
A dedi is a machine without DE it's all command line |
| 12:56 |
tenplus1 |
just checking :PPP |
| 12:57 |
Shara |
I'm just trying to work out how my VPS with 160+ mods even manages now |
| 12:57 |
VanessaE |
remember that with a VPS you have to fight with other users for resources. |
| 12:58 |
Shara |
Yet I get basically no problems ever |
| 12:58 |
VanessaE |
same with my dedicated server. it performs as I would expect it to |
| 12:59 |
Shara |
24 players currently. I don't get how this runs so well with so many mods when Fussel has such issues with just MTG |
| 12:59 |
tenplus1 |
who are you getting dedi from fussel ?? |
| 13:01 |
IhrFussel |
VanessaE, I already have a dedi but its performance is pretty bad ... now someone here recommended me a VPS which worked fine the FIRST day (absolutely no lags) ... but then the nightmare started and I had hourly 30-60 secs lags (last night it was even 222 secs) and I decided to go for a more modern dedi but now I realized this new dedi isn't a lot better than my old one so I cancel it again |
| 13:01 |
* VanessaE |
shrugs |
| 13:02 |
tenplus1 |
do you guys do anything to optimize .conf for dedicated servers ? |
| 13:02 |
VanessaE |
mine (from SoYouStart, a division of OVH) works fine for me |
| 13:02 |
IhrFussel |
I'm not rich...very good dedis are too expensive for my budget ($100+) |
| 13:02 |
VanessaE |
tenplus1: not really, no. |
| 13:02 |
VanessaE |
mine costs $42/mo |
| 13:02 |
tenplus1 |
that's why we run a desktop :) ehehehe |
| 13:03 |
IhrFussel |
I already cancelled my contract, hope they won't charge me before |
| 13:03 |
VanessaE |
and yes, dedis usually have low-end onboard graphics/video, but minetestserver does not use the gpu. |
| 13:04 |
* tenplus1 |
gives fussel's server some lucozade |
| 13:05 |
Shara |
I thought I'd need a dedi given the way RC was growing, but tried upgrading to a slightly better VPS instead, and it turned out to be more than enough. |
| 13:06 |
VanessaE |
I guess in my case a dedi is needed since I run several MT instances |
| 13:06 |
tenplus1 |
how many you got now Vanessa ? |
| 13:06 |
Shara |
Yea, I'd need to switch if I wanted to start more servers. |
| 13:07 |
Shara |
sniper is hosting DarkLands as it is though, and TPS hosts my third servers that may or may not ever open publically. |
| 13:07 |
VanessaE |
tenplus1: my 5 plus like 4 or 5 that I host for others. |
| 13:07 |
IhrFussel |
Shara, you most likely just have very good luck that the other customers in the same cluster don't exceed the usage |
| 13:08 |
Shara |
Fussel, hard to say. I have a history of crazy luck when it comes to this kind of thing, but other people have done fine with VPS as well. |
| 13:08 |
IhrFussel |
I'm not sure what kind of script makes all customers lag for 200+ secs...it must be some stupid backup script or so |
| 13:08 |
VanessaE |
looks like there are only 7 total right now, at least that are announcing to the master list |
| 13:09 |
tenplus1 |
fussel: sounds silly but have you tried disabling rollback and anticheat ? |
| 13:09 |
IhrFussel |
That's not it...when the VPS lags a long time I see the load going up up to 3-5 |
| 13:10 |
Shara |
I seem to get the full resources I should on my VPS. |
| 13:10 |
VanessaE |
9 total running on the server. guess a couple of them just don't announce. |
| 13:10 |
IhrFussel |
The lags mainly happen every full hour ... it must be some expensive cron or whatever |
| 13:11 |
tenplus1 |
fussel: I know that may not be the initial problem, but the tips I'm giving you were used to bring Xanadu's lag right down and it definitely helped |
| 13:12 |
IhrFussel |
Rollback is enabled and to disable anticheat allows Android users to dig nodes from far away (300 nodes or so) |
| 13:13 |
IhrFussel |
0.4.16 or an earlier Android version introduced that interact bug...it only happens when in 3rd person view |
| 13:13 |
tenplus1 |
*shakes fist at android bugs* |
| 13:14 |
|
Jordach joined #minetest-hub |
| 13:14 |
tenplus1 |
hi Jordach |
| 13:15 |
Shara |
Well, I literally can't run anticheat on my servers without a load of other things being spoiled for players... not a fun dilema |
| 13:15 |
IhrFussel |
tenplus1, fixed in 0.5.0 https://github.com/minetest/minetest/pull/6187 |
| 13:16 |
tenplus1 |
w00t... cant wait for 0.5 release... |
| 13:16 |
IhrFussel |
Wait it's part of 0.4.17 |
| 13:17 |
tenplus1 |
even better |
| 13:17 |
VanessaE |
is 0.4.17 coming out soon? |
| 13:18 |
IhrFussel |
Is game.cpp client side? |
| 13:18 |
VanessaE |
server |
| 13:19 |
IhrFussel |
Ok then it should be fixed in 0.4.17 servers no matter which version the client has |
| 13:20 |
|
EvergreenTree joined #minetest-hub |
| 13:21 |
IhrFussel |
VanessaE, I was asking last night whether or not 0.4.17 would be an "official" release in the play store and someone only said likely just 0.5.0 |
| 13:22 |
tenplus1 |
in which case I want 0.5.0 more... |
| 13:22 |
IhrFussel |
Not sure about other places like websites ... but seems like "0.4.17" is really just meant for server bugfixes |
| 13:23 |
VanessaE |
well, idc about play stores or whatever |
| 13:24 |
sfan5 |
IhrFussel: that's the plan yes |
| 13:24 |
VanessaE |
since I run a real PC. |
| 13:24 |
sfan5 |
i mean 0.4.17 is a good idea for clients too, but it won't be released on the play store as 0.5.x is the real successor |
| 13:24 |
tenplus1 |
:P |
| 13:24 |
Shara |
You can't enforce people to use an updated client unless compatibility breaks anyway |
| 13:25 |
Shara |
0.4.17 is important for servers far more than clients |
| 13:25 |
tenplus1 |
yeah, will be using that when it's out to fix many issues |
| 13:26 |
lisac |
hey tenplus1 |
| 13:26 |
lisac |
when is .5 coming out? |
| 13:26 |
IhrFussel |
tenplus1, you can already use it when you compile backport-0.4 branch |
| 13:26 |
IhrFussel |
You mean 0.5.0? |
| 13:26 |
lisac |
yeah, that's .5. |
| 13:27 |
IhrFussel |
0.5.0 has no release date yet ... last I heard was Feb/Mar 2018 |
| 13:27 |
Shara |
Better to get it right than rush release |
| 13:28 |
IhrFussel |
"You mean 0.5.0" was directed at tenplus1 though sorry xP |
| 13:28 |
tenplus1 |
I d ont compile my builds, I get 'em from launchpad on the ubuntu/mint unstable build |
| 13:29 |
IhrFussel |
Well let's hope that the PPAs will get updated with 0.4.17 then |
| 13:29 |
tenplus1 |
bloody hope so |
| 13:35 |
* tenplus1 |
is off to nom chili... laters all |
| 13:35 |
|
tenplus1 left #minetest-hub |
| 13:47 |
IhrFussel |
There are IRC channels where you can get a VPS/dedi? Doesn't sound very reliable |
| 13:50 |
Jordach |
^# |
| 13:53 |
|
EvergreenTree joined #minetest-hub |
| 13:58 |
|
CWz joined #minetest-hub |
| 13:58 |
VanessaE |
sfan5: maybe you remember: there was a PR, since merged into the 0.5.0 codebase, to fix dig drops not inheriting the param2 color of the node that was dug...but for the life of me, I can't find it. |
| 13:59 |
sfan5 |
hm |
| 14:00 |
VanessaE |
I just want to request that it be backported to 0.4.x |
| 14:01 |
VanessaE |
it's the last piece of the puzzle needed so that unified dyes can support colored itemstacks |
| 14:01 |
sfan5 |
this one? https://github.com/minetest/minetest/commit/71b02d626f4a9629e0f282adef908eedc8ed252a |
| 14:02 |
VanessaE |
no, not that.. all I can remember, and I'm not sure of this much now, was that the issue number was some assortment of 6's, 4's 0's and maybe 2's :) |
| 14:03 |
sfan5 |
but that's the exact commit implementing saving of the color to item meta |
| 14:04 |
VanessaE |
nono, there was another, regarding auto-inheritance |
| 14:04 |
sfan5 |
https://github.com/minetest/minetest/pull/6604 ? |
| 14:05 |
sfan5 |
that used to have auto-inheritance but was changed to use a callback for flexibility reasons |
| 14:05 |
VanessaE |
no, it was by juhdanad I'm pretty sure |
| 14:07 |
sfan5 |
how about this https://github.com/minetest/minetest/commit/0fcaf9fb1b61caaf8ed78a5d3005b1d9d1c43b57 |
| 14:07 |
sfan5 |
if that's not it then i don't know what you mean |
| 14:07 |
IhrFussel |
I just found a root server with 8 dedicated cores and 120 GB SSD for $25/month ... the CPU seems to be not much better than my Opteron though > Intel® Xeon® E5-2680V4 |
| 14:08 |
sfan5 |
a xeon will definitely be better than some opteron from 2009 |
| 14:08 |
* VanessaE |
keeps searching |
| 14:08 |
IhrFussel |
I just had a Xeon and the max_lag was almost as high as on my Opteron |
| 14:09 |
IhrFussel |
You need to check singlecore performance and it's not a lot better |
| 14:12 |
sfan5 |
lol what |
| 14:12 |
sfan5 |
IhrFussel: https://www.cpubenchmark.net/compare.php?cmp[]=2779&cmp[]=1406 |
| 14:13 |
VanessaE |
sfan5: actually, I think 71b02d62 *is* the one I had in mind |
| 14:13 |
IhrFussel |
sfan5, how did max_lag stay at ~ 0.6 then? |
| 14:14 |
sfan5 |
no idea, there's many more factors to that than just the CPU |
| 14:14 |
IhrFussel |
Your benchmark doesn't compare singlecore performance.... look at this http://cpuboss.com/cpus/Intel-E5-2680V4-vs-AMD-Opteron-1381 |
| 14:14 |
VanessaE |
hah! https://github.com/minetest/minetest/pull/6042 |
| 14:15 |
sfan5 |
IhrFussel: it does actually, look above the red numbers: 1782 (xeon) vs 915 (opteron) |
| 14:15 |
sfan5 |
not sure what site cpuboss.com is but it doesn't seem to be any good |
| 14:17 |
VanessaE |
sfan5: ^^ that was the issue number. I'd like to request that it be backported (not necessarily by you, just whoever). being all Lua, I Imagine it's not too hard? |
| 14:18 |
sfan5 |
added it to the TODO |
| 14:18 |
sfan5 |
lua or not it's only "hard" when merge conficts occurr |
| 14:18 |
VanessaE |
thanks :) |
| 14:18 |
VanessaE |
at least I was right about the selection of digits in the issue number ;) |
| 14:19 |
VanessaE |
wow, that's some list. |
| 14:21 |
IhrFussel |
sfan5, I had nothing else running, the RAM is DDR3 and the HDD doesn't matter since I didn't move during the time frame and max_lag *still* was at ~ 0.7 ... that screams weak single core power at me |
| 14:21 |
sfan5 |
was that on a vps? |
| 14:21 |
IhrFussel |
No, dedi |
| 14:21 |
sfan5 |
no idea then |
| 14:21 |
sfan5 |
on paper the Xeon is definitely better |
| 14:22 |
IhrFussel |
htop showed minetestserver @ 50-90% CPU usage whenever I was flying around...sounds a bit much |
| 14:25 |
Shara |
Hmmm... 28/30 players causes my server to drop right down the list |
| 14:25 |
Shara |
Yes my survival at 18/22 is quite high |
| 14:26 |
* Shara |
continues to find the server list ranking weird |
| 14:26 |
sfan5 |
the algorithm penalizes highly loaded servers |
| 14:26 |
IhrFussel |
sfan5, E3-1230v2 was the CPU I had on the test dedi |
| 14:26 |
IhrFussel |
E5 seems to be a bit higher in single core power |
| 14:27 |
Shara |
sfan5: You are really making it so servers that don't allow high numbers of players can never be at the top anymore then? |
| 14:27 |
sfan5 |
i don't understand your question |
| 14:27 |
sfan5 |
also it wasn't me who designed this algo |
| 14:28 |
Shara |
I don't object to not being at the top, but I do find the situation really silly |
| 14:28 |
Shara |
Hometown certainly never scores as high as it seems it should, to give another example |
| 14:29 |
sfan5 |
improvements welcome https://github.com/minetest/master-server/blob/master/server.py#L303L344 |
| 14:30 |
Shara |
Hmm, might think about it. |
| 14:30 |
Shara |
It's no criticism of you either way. |
| 14:42 |
Krock |
rubenwardy, you could add one or multiple "safety" slots to the player's inventories which require a certain score in the running game in order to be used |
| 14:43 |
Shara |
So a person respawns, who can't be killed for some seconds, with a good weapon ready to use? :P |
| 14:55 |
Krock |
not saying that they can use it for the first seconds. i.e. not in the "main" inventory |
| 15:14 |
rubenwardy |
IhrFussel, that Xeon should be more than enough |
| 15:15 |
rubenwardy |
how are you reading the CPU percentage? |
| 15:15 |
IhrFussel |
rubenwardy, E5 will be likely enough, E3 definitely wasn't |
| 15:15 |
rubenwardy |
usually % is terms of cores, so 400% is complete usage |
| 15:15 |
IhrFussel |
You know that MT runs on one core |
| 15:15 |
rubenwardy |
emerge threads |
| 15:15 |
rubenwardy |
and networking |
| 15:15 |
rubenwardy |
but you have some very heavy mods |
| 15:15 |
IhrFussel |
I don't care about mapgen, I want much power for my mods |
| 15:16 |
rubenwardy |
well, the CPU is enough for Minetest, just not for you |
| 15:16 |
IhrFussel |
No, I had it run just a test world with MTG and max_lag never went below 0.5 |
| 15:17 |
rubenwardy |
also "it's not the hdd" -> try dummy backend to verify this |
| 15:17 |
rubenwardy |
I mean you can't now you've cancelled |
| 15:17 |
IhrFussel |
It doesn't matter now, I just realized how greedy the my current hoster is and chose a product from another smaller one now...I pay $15 per month for 4 cores+ 12 GB RAM + 120 GB SSD |
| 15:17 |
|
compunerd joined #minetest-hub |
| 15:18 |
rubenwardy |
nice |
| 15:18 |
IhrFussel |
I wonder how they can make it that cheap but I'll trust them |
| 15:19 |
IhrFussel |
It's a root server though not "real" dedi... so it's virtual but dedicated |
| 15:23 |
IhrFussel |
The only problem I got is that they sent me a contract which I need to print and send back to them...and I got no printer :/ |
| 15:25 |
|
Fixer joined #minetest-hub |
| 15:31 |
* Jordach |
dies laughing at rubenwardy's email |
| 15:39 |
|
celeron55 joined #minetest-hub |
| 15:42 |
|
xerox123 joined #minetest-hub |
| 15:51 |
|
aerozoic joined #minetest-hub |
| 15:55 |
red-001 |
looks like minetesthosting went down |
| 15:57 |
rubenwardy |
yup |
| 15:58 |
rubenwardy |
and the average quality of servers has thus increased |
| 15:58 |
rubenwardy |
just need CTF to go down now, for the quality to raise further! |
| 16:00 |
red-001 |
lol |
| 16:01 |
rubenwardy |
<CTF_Bot> *** minecraft joined the game |
| 16:01 |
rubenwardy |
shieeeettt |
| 16:02 |
red-001 |
the minetest.com page sure is old |
| 16:03 |
rubenwardy |
yup |
| 16:03 |
rubenwardy |
and the owner is AWOL |
| 16:03 |
Sokomine |
VanessaE: i'm on 0.4.16x and digged blocks inherit their color. what is keeping me from getting further with circular saw support for the colored blocks is that i have to set up color matching (translation from 256 colors for full block to 8 colors for stair shape) manually. it would help a lot if i'd be able to read the palette directly and be able to find out which colors are the same |
| 16:04 |
Sokomine |
and not just in one mod but in a more general lib |
| 16:07 |
VanessaE |
Sokomine: digged blocks with a custom drop (e.g. stone -> cobble) don't inherit. |
| 16:12 |
CWz |
VanessaE, are you able to compile the backport branch on your server? |
| 16:14 |
Shara |
Now RC is #1 on the list with only 12 players? And DL at 21/22 is second? While other servers have 30+ players and come much lower... I don't understand |
| 16:15 |
VanessaE |
CWz: I'm waiting for a few things to go in first |
| 16:15 |
CWz |
might need to update your version of gcc |
| 16:15 |
rubenwardy |
Shara, CTF just restarted ;) |
| 16:16 |
Shara |
rubenwardy: It's just about how senseless it is |
| 16:16 |
Krock |
Shara, not only the user count matters. much depends on the player names (no 3/4 digit endings) or the age |
| 16:17 |
rubenwardy |
DL has a lot of users with 3 numbers |
| 16:17 |
Shara |
When RC was at 28/30 earlier, it fell way outside the top ten, but at 12/35 it can hit #1. Yet DL just 1 below cap with a lower maximum can hit 2nd? |
| 16:17 |
rubenwardy |
RC doesn't |
| 16:17 |
Shara |
Which should bring it down. |
| 16:17 |
Shara |
It at close to cap DL ranks high while RC at close to cap just drops right down |
| 16:18 |
Shara |
Yet at* |
| 16:21 |
Fixer |
rubenwardy: high quality players (c) |
| 16:23 |
red-001 |
(TM) (R) |
| 16:25 |
|
EvergreenTree joined #minetest-hub |
| 16:25 |
red-001 |
wtf? |
| 16:26 |
|
bigfoot547 joined #minetest-hub |
| 16:27 |
Sokomine |
VanessaE: aah. yes, i remember. you mentionned that. however...cobble...is bad in most ways. just returning the colored stone would be best... |
| 16:28 |
red-001 |
bigfoot547, what was that about? |
| 16:29 |
red-001 |
ah I see linode got k-lined |
| 17:07 |
|
red-001 joined #minetest-hub |
| 17:11 |
|
twahm joined #minetest-hub |
| 17:15 |
|
EvergreenTree joined #minetest-hub |
| 17:16 |
Fixer |
COBOL |
| 17:18 |
IhrFussel |
How can I imagine a root server with dedicated cores? A root server is also a form of VPS no? |
| 17:20 |
Calinou |
don't let your root servers be dreams |
| 17:22 |
|
red-001 joined #minetest-hub |
| 17:23 |
IhrFussel |
Calinou, the description says "Our root servers combine the advantages of virtual and dedicated servers. Despite virtualization the root servers provide guaranteed Intel® XEON® Broadwell® CPU cores and fast DDR4 main memory (RAM)." |
| 17:23 |
Calinou |
doesn't that mean you get a full machine, but intended for virtualization? |
| 17:24 |
IhrFussel |
I chose the 2nd https://www.netcup.eu/vserver/#features |
| 17:28 |
IhrFussel |
Maybe their virtualization software allows to split full cores between customers?? IDK |
| 17:31 |
red-001 |
IhrFussel, that sounds like marketing mumbo jumbo |
| 17:31 |
IhrFussel |
red-001, the VPS description mentions "Only the dedicated CPU cores, the satisfaction guarantee and the very high guaranteed average annual minimum availability of at least 99.9% are not included in the VPS system." |
| 17:33 |
IhrFussel |
I don't think companies in Germany are allowed to mention "dedicated" where it#s not |
| 17:34 |
IhrFussel |
The root server's CPU is the same across all plans "Intel® Xeon® E5-2680V4" ... the VPS CPU is the same across all plans "Intel® vCore" |
| 17:35 |
red-001 |
so I guess they pin one cpu core to the vm and call it dedicated? |
| 17:41 |
IhrFussel |
I'll see what I get... in case they really lied to me and I have inconsistent performance I can just cancel it again I guess |
| 17:44 |
IhrFussel |
I found another hoster with root servers mentioning "guaranteed" performance: Full virtualization of the root server provides you guaranteed resources in terms of CPU and RAM. |
| 17:45 |
red-001 |
if you think it's false advertisement you could report them to a consumer protection watchdog |
| 18:02 |
Fixer |
c o m p i l e t i m e |
| 18:08 |
Jordach |
IhrFussel: scaleway does hand out dedis like candy |
| 18:08 |
Jordach |
jordach.net runs it |
| 18:09 |
IhrFussel |
I feel more comfortable with a German hoster, sorry :/ But is what netcup advertizes possible? To map full cores to certain customers? |
| 18:09 |
Jordach |
core affinity |
| 18:10 |
Jordach |
cores can be locked to KVM processes |
| 18:10 |
Jordach |
but if one needs more load, the affinity generally stops working until the load dies down |
| 18:10 |
IhrFussel |
So will I have guaranteed CPU power at any time? |
| 18:11 |
Jordach |
yes |
| 18:11 |
Jordach |
you will have close to 100% power at any given moment |
| 18:11 |
Jordach |
per "core" |
| 18:11 |
IhrFussel |
So other customers cannot make me lag for 30+ secs? |
| 18:11 |
Jordach |
only in certain circumstances |
| 18:12 |
Fixer |
lol, cheap hosting sucks these days |
| 18:12 |
Jordach |
by certain, i mean the 1 or 2 other users of that machine |
| 18:12 |
Calinou |
Fixer: not necessarily |
| 18:12 |
IhrFussel |
But they say I get more than "vCores" |
| 18:13 |
Fixer |
the worst was LazyJ server hoster, constant max_lag 6-8 with just one player and even his ssh lagged, what a pos, he switched it to better one |
| 18:13 |
Krock |
virtual cores? |
| 18:13 |
Jordach |
with the jetpack and my server generating *new* mapblocks i can't get max_lag over 0.5 |
| 18:14 |
IhrFussel |
vCores are only mentioned in the vServer (VPS) section not in the root server section |
| 18:17 |
Jordach |
root server may be actually a true dedi |
| 18:17 |
Jordach |
but not in the sense of a real dedi |
| 18:18 |
Jordach |
aka it's a dedi that gets reformatted constantly |
| 18:18 |
Jordach |
why waste old hardware when you can make cheap re-usable dedis from it |
| 18:19 |
Jordach |
my current dedi is SSDs top to bottom |
| 18:19 |
IhrFussel |
I really don't know...I just see "4 dedicated cores" and therefore expect full 4 cores to be mine and only mine |
| 18:20 |
Jordach |
it's just a VM with less users at any given time essentially |
| 18:27 |
|
red-001_ joined #minetest-hub |
| 18:29 |
IhrFussel |
Jordach, they offer a site that shows the differences between vps and root https://www.netcup.eu/vserver/vergleich-root-server-vps.php |
| 18:30 |
Jordach |
IhrFussel: they're VPS |
| 18:30 |
Jordach |
Despite virtualization the root servers |
| 18:30 |
Jordach |
just with more affinity over other processes |
| 18:33 |
IhrFussel |
It says "these cores will not be shared with other systems" isn't that a lie then? |
| 18:33 |
Jordach |
cores can be locked with affinity |
| 18:33 |
Jordach |
except in certain cases your spare power will compensate for a while |
| 18:47 |
|
redblade7 joined #minetest-hub |
| 18:55 |
benrob0329 |
Idea for end-portal type effect: use an ignore type block that shows the skybox, and change the skybox |
| 18:58 |
benrob0329 |
Or we could ask for a skybox drawtype |
| 18:59 |
Jordach |
pls gib borderless item_image_buttons |
| 19:19 |
benrob0329 |
Can you change player buoyancy? |
| 19:19 |
Calinou |
yes, but it doesn't work very well in practice, I tried |
| 19:19 |
Calinou |
there's one or two minetest.conf settings for that |
| 19:19 |
Calinou |
also, liquid viscosity (per-node) |
| 19:20 |
benrob0329 |
Hmm |
| 19:20 |
benrob0329 |
Boats need to be redone |
| 19:20 |
benrob0329 |
Boat elevators might actually be viable if boats themselves didn't suck (no offence, they just do) |
| 19:25 |
benrob0329 |
Calinou: per player buoyancy? |
| 19:57 |
Fixer |
btw, boats and carts in mc beta suck a lot |
| 20:13 |
Sokomine |
Jordach: power consumption of old hardware may be much higher than new hw |
| 20:14 |
Jordach |
you still have to *sell* it |
| 20:15 |
Krock |
Sokomine, also consider production and recycling energy |
| 20:16 |
Sokomine |
Krock: i don't think hosters count that way. disposal may be an issue. but old hw needs more energy -> more cooling -> expensive!. ok, german hosters move to france where energy is cheper, so...probably it's worth keeping old hw there |
| 20:29 |
|
octacian joined #minetest-hub |
| 20:55 |
Jordach |
https://i.imgur.com/73r3Eep.jpg |
| 20:56 |
sofar |
that 90s button shading |
| 20:56 |
Jordach |
but i like win 9x |
| 20:56 |
Krock |
that Alt+Printscreen instead of F12 |
| 20:56 |
Jordach |
two step operation to upload |
| 20:57 |
Krock |
save your screenies to a synced cloud directory? |
| 20:57 |
Jordach |
bitch, does it look like i have a server with sharex support....wait a second |
| 20:58 |
Krock |
bro, the second is over |
| 20:59 |
Jordach |
https://jordach.net/images/dtlzqvelvslapygsvl.png |
| 20:59 |
Krock |
sofar, irrlicht's buttons aren't that much better regarding the shading :P |
| 21:00 |
Krock |
topkek |
| 21:00 |
* sofar |
has 'printscreen' bound to script that uses `import` and automatically uploads to imgur, while making a log entry for posterity in ~/.imgur.log with the URL |
| 21:00 |
sofar |
$ wc -l ~/.imgur.log |
| 21:00 |
sofar |
394 /home/sofar/.imgur.log |
| 21:01 |
sofar |
https://i.imgur.com/oCJMpCi.png |
| 21:01 |
Sokomine |
Jordach: looks nice, yet...it doesn't have those nice practical buttons unified_inventory has. that copy-to-craft-grid, the serach button for receipes and the general setup are very useful |
| 21:02 |
Krock |
sofar, nice one :) |
| 21:02 |
sofar |
at one point I even had a local web script to pull all of the screenshots into a viewable page, lol |
| 21:03 |
sofar |
but the fact I can screenshot a region, a window or the desktop with one button and be done, saves a bunch of time |
| 21:03 |
sofar |
it even opens a browser window with the uploaded screenshot |
| 21:04 |
sofar |
https://gist.github.com/sofar/113c9be7256a37ef1d3e2c328ea5b322 |
| 21:18 |
octacian |
I just have a simple python script rubenwardy gave me to allow right-clicking on pictures and directly uploading them to a VPS. |
| 21:34 |
* Jordach |
sits back |
| 21:38 |
Calinou |
ShareX does this easily |
| 21:38 |
Calinou |
I should look into setting `incron` up for doing something like that |
| 21:38 |
Calinou |
optimizing using oxipng, then uploading via SFTP to my server |
| 21:39 |
Calinou |
(incron is like cron, but does things when files are added/modified/removed) |
| 21:39 |
Calinou |
or is there a better alternative to incron? |
| 21:39 |
Calinou |
(it's to be used on a client PC, so it should stay simple enough) |
| 21:46 |
Jordach |
Calinou: >2018 |
| 21:46 |
Jordach |
>PNG optimisation |
| 21:52 |
Calinou |
Jordach: it can bring 30%+ gains on screenshots |
| 21:52 |
Calinou |
usually above 10% |
| 21:52 |
Calinou |
so it's worth it, and takes a few seconds at most with oxipng (which is multi-threaded) |
| 21:53 |
Jordach |
>460 bytes per PNG in some cases |
| 21:53 |
red-001 |
maybe emacs has something, it's a pretty good OS, just missing a decent text editor |
| 21:59 |
benrob0329 |
Calinou: >sftp in 2017 |
| 21:59 |
benrob0329 |
Rsync master race :^) |
| 22:05 |
Krock |
me: https://i.imgur.com/CMfpZ5x.gif |
| 22:38 |
Jordach |
sofar: mind announcing you're streaming sometime |
| 22:40 |
xerox123 |
do people still need to be whitelisted for the server list? |
| 22:42 |
Calinou |
benrob0329: it doesn't need to be incremental, scp/sftp works fine for this |
| 22:43 |
Calinou |
it's fast and secure |
| 22:43 |
Calinou |
Jordach: nah, over 1 MB on many PNGs I've tried |
| 22:44 |
red-001_ |
sofar, nice mapgen |
| 22:52 |
benrob0329 |
Calinou: s/ftp is not fast |
| 22:53 |
benrob0329 |
Nor is it stable |
| 22:57 |
|
ThomasMonroe joined #minetest-hub |
| 23:04 |
Sokomine |
qbruell :welcomes the year 2018 and wishes all a happy new year |
| 23:04 |
* Sokomine |
welcomes the year 2018 and wishes all a happy new year |
| 23:05 |
ThomasMonroe |
happy new year to you too Sokomine |
| 23:05 |
* red-001_ |
guess $current_year is going to be the year of the Linux desktop |
| 23:06 |
|
EvergreenTree joined #minetest-hub |
| 23:06 |
red-001_ |
happy new year to you too ThomasMonroe and Sokomine |
| 23:06 |
ThomasMonroe |
thanks red-001_ |
| 23:08 |
Calinou |
happy new year! |
| 23:09 |
benrob0329 |
Its still 2017 here |
| 23:09 |
benrob0329 |
Hello people of the future, how is it? |
| 23:10 |
ThomasMonroe |
same here for me |
| 23:11 |
IhrFussel |
Happy New Year! |
| 23:13 |
|
Raven262 joined #minetest-hub |
| 23:13 |
lisac |
Happy new year |
| 23:13 |
lisac |
aaand good night |
| 23:14 |
ThomasMonroe |
night lisac |
| 23:16 |
Sokomine |
benrob0329: so far quite good. bit too bright outside to enjoy any firework; also too warm for this time of year |
| 23:16 |
Sokomine |
benrob0329: you can come over to 2018 once your time zone catches up. mt still runs in 2018 :-) |
| 23:27 |
|
sys4 joined #minetest-hub |
