| Time |
Nick |
Message |
| 00:31 |
|
EvergreenTree joined #minetest-hub |
| 00:38 |
|
Fixer joined #minetest-hub |
| 00:47 |
|
Fixer joined #minetest-hub |
| 00:53 |
Fixer |
more healthy offtopic |
| 00:54 |
Fixer |
disabled antivirus, delayed start to worst offenders, boot time improved from 160 to 50 sec |
| 01:00 |
red-001 |
#minetest-hub on-topic is [off] and offtopic is ontopic |
| 01:00 |
roben1430 |
Disabled Windows, removed worst offenders. Boot time improved from 3 minutes to 10 seconds |
| 01:09 |
Fixer |
.net is a bitch to start |
| 01:10 |
Fixer |
infamous ATI CCC, slow as fuck since 2003 (iirc) |
| 01:10 |
Fixer |
greenshot - advanced screenshoter, boot time like it is photoshop |
| 01:25 |
IhrFussel |
max_lag can max decrease by 50% per 5 minutes right? I think someone said that |
| 01:26 |
|
twoelk joined #minetest-hub |
| 01:33 |
red-001 |
oooh the bugs got catchy names |
| 01:52 |
|
EvergreenTree joined #minetest-hub |
| 01:53 |
|
roben1430 joined #minetest-hub |
| 02:02 |
|
sniper570 joined #minetest-hub |
| 02:08 |
|
twoelk left #minetest-hub |
| 02:11 |
|
roben1430 joined #minetest-hub |
| 02:14 |
|
paramat joined #minetest-hub |
| 02:27 |
|
EvergreenTree joined #minetest-hub |
| 02:28 |
|
roben1430 joined #minetest-hub |
| 02:59 |
|
roben1430 joined #minetest-hub |
| 03:05 |
* benrob0329 |
crashes hexchat with lua script |
| 03:05 |
benrob0329 |
And all of Xorg apparently |
| 03:26 |
|
Shara joined #minetest-hub |
| 03:26 |
shivajiva |
wb Shara |
| 03:30 |
red-001 |
hi Shara |
| 03:31 |
Shara |
Hi :) |
| 03:36 |
benrob0329 |
Hello Shara |
| 03:37 |
Shara |
I am so tired. But server was so broken. |
| 03:37 |
Shara |
And apparently I can't sleep when server is broken. |
| 04:02 |
|
EvergreenTree joined #minetest-hub |
| 04:08 |
shivajiva |
no excuse now :P |
| 04:09 |
shivajiva |
oops wrong window :) |
| 04:09 |
red-001 |
that explains the confusion |
| 04:10 |
shivajiva |
nothing can explain the confusion red lol |
| 04:53 |
|
paramat joined #minetest-hub |
| 05:54 |
* sofar |
goes to fix more shivajiva code |
| 05:58 |
* sofar |
rebases |
| 05:58 |
* sofar |
fixes conflicts |
| 06:02 |
* sofar |
luachecks |
| 06:03 |
|
roben1430 joined #minetest-hub |
| 07:07 |
|
CWz joined #minetest-hub |
| 07:07 |
|
nerzhul joined #minetest-hub |
| 09:29 |
|
nerzhul joined #minetest-hub |
| 10:58 |
Megaf |
Good morning |
| 11:06 |
|
IhrFussel joined #minetest-hub |
| 11:06 |
IhrFussel |
Hi guys |
| 11:14 |
Megaf |
morning Fussel |
| 11:22 |
Megaf |
Man, for days I've been trying to figure out how to make Linux dont use one VGA and use another instead |
| 11:23 |
Megaf |
My integrated GPU had it's outputs burnt, now I wanted to use a PCI VGA |
| 11:23 |
Megaf |
But I can't even make it send any output at all to the PCI VGA |
| 11:23 |
Megaf |
and by PCI I meant it, the old PCI standard |
| 11:25 |
IhrFussel |
So far I only see one entry where my avg_lag mod reports an average lag above 0.150 sec within hours ... very happy with this root server so far |
| 11:31 |
Megaf |
IhrFussel: what do you mean by root server? |
| 11:32 |
IhrFussel |
Root server can mean different things...in my case it's a VPS with dedicated CPU, so not really shared between all customers |
| 11:33 |
IhrFussel |
Or let's say I "own" certain cores of that CPU when I need them |
| 11:34 |
Megaf |
which one is your host? |
| 11:35 |
IhrFussel |
https://www.netcup.eu/vserver/ << I have the RS 2000 |
| 11:37 |
IhrFussel |
It can't get much cheaper than that considering it mentions dedicated cores |
| 11:39 |
Megaf |
That's great price |
| 11:41 |
IhrFussel |
In the support forum I saw people running CS:GO servers on it and they say the performance is outstanding |
| 11:43 |
IhrFussel |
So I thought if the product can handle CS:GO it sure as heck should be able to handle MT |
| 11:44 |
Megaf |
FPS shooters dont require much power |
| 11:44 |
IhrFussel |
But CS:GO runs with a MUCH higher tick rate (64 by default) |
| 11:46 |
IhrFussel |
Or was it 66? Something between 60-70 ... also the hoster *promises* 99.9% availability per year which is 364.876957 days |
| 11:47 |
|
Fixer joined #minetest-hub |
| 11:49 |
Megaf |
Morning Fixer |
| 11:51 |
IhrFussel |
Megaf, a FPS server has to calculate shots all the time plus collision... are you sure that's not demanding? |
| 11:51 |
Megaf |
Well, you need a high clock yes, but not a lot of memory |
| 11:51 |
Megaf |
nor disk space |
| 11:51 |
Megaf |
and you are not running ABMs |
| 11:53 |
IhrFussel |
True that...well the Xeon E5 "only" has 2.4 Ghz but I heard it can use turbo (not sure if my hoster supports that though) |
| 11:53 |
Megaf |
I hosteI some WOLFET servers back in the day |
| 11:54 |
Megaf |
and ArmyOps servers too |
| 11:54 |
Megaf |
hosted* |
| 11:54 |
Megaf |
Minetest is the most demaning game server I have ever hosted, seconded by OpenSim |
| 11:55 |
Megaf |
And Minetest is only demanding because of it's experimental design (inefficient, but improving) |
| 11:55 |
Megaf |
OpenSim runs 100s of scripts at the same time while managing 100s of players |
| 11:55 |
Megaf |
still using less resourses than my MT server with 2 or 3 people and a couple of ABMs... |
| 11:56 |
Megaf |
OpenSim is Mono/C# |
| 11:56 |
Megaf |
there's no excuse really... |
| 11:56 |
Megaf |
MT was very good in it's 0.3 version |
| 11:57 |
Megaf |
man, I miss my OSGrid region. Now that I don't have MT server running in my VPS anymore I might put my OSGrid region back online |
| 12:01 |
|
IhrFussel joined #minetest-hub |
| 12:02 |
IhrFussel |
Sorry in case I spammed the chat with my relogs ... IRC is a bit weird when it comes to reconnecting after the ISP decided to switch your IP |
| 12:03 |
Megaf |
did you read anything I wrote? |
| 12:03 |
IhrFussel |
I just called my hoster and they said the turbo mode of the CPU will be used "Minetest is the most demaning game server I have ever hosted, seconded by OpenSim" << last message I received |
| 12:08 |
Megaf |
IhrFussel: http://paste.debian.net/plain/1003588 |
| 12:09 |
Megaf |
you can just curl that |
| 12:12 |
IhrFussel |
Yes like I already said recently: MT can run great on very fast hardware ... but it's definitely not good enough for older CPUs/specs anymore which is a shame |
| 12:14 |
IhrFussel |
I mean you CAN play MT on a Celeron from 2003 ... with 20 FPS *after* you optimized everything in advanced settings |
| 12:17 |
IhrFussel |
Even an AMD Opteron from 2010 struggles if you run a MT server on it ... the Intel Xeon E5 my current root server uses is from 2016 AFAIK and even THAT cannot stay on consistent 150ms ticks |
| 12:19 |
IhrFussel |
We all know the main issue: Server thread only uses one single core of any machine |
| 12:24 |
sfan5 |
IhrFussel: have they ever hosted minecraft? :P |
| 12:25 |
sfan5 |
it's far more demanding |
| 12:28 |
IhrFussel |
Yes in the support forum customers also report performance when running MC ... they say mostly the TPS stay at 20 |
| 12:31 |
IhrFussel |
With 10 players max_lag now is 0.3 ... okay I admit my server/mods aren't that lightweight but there is still a lot of room for improvement |
| 12:31 |
IhrFussel |
engine-side* |
| 12:33 |
IhrFussel |
I think MT would perform best on a CPU that is clocked at 3+ Ghz per core |
| 12:39 |
Megaf |
MT is great for high clock CPUs yes |
| 12:39 |
Megaf |
all MT needs is 2 threads running at very high clock rates |
| 12:40 |
Megaf |
an AMD FX woul be great for MT |
| 12:41 |
Megaf |
https://www.newegg.com/Product/Product.aspx?Item=N82E16819113347 |
| 12:42 |
Megaf |
!title |
| 12:42 |
MinetestBot |
Megaf: AMD FX-9590 Vishera 8-Core 4.7 GHz Socket AM3+ 220W FD9590FHHKWOF Desktop Processor - Black Edition - Newegg.com |
| 12:42 |
Megaf |
Perfect CPU for Minetest ^ |
| 12:42 |
Megaf |
IhrFussel: ^ |
| 12:43 |
Megaf |
I think that's the highest clock x86 CPU we can buy |
| 12:43 |
sfan5 |
AMD's FX series has terrible single core perf so that's not actually a good cpu for mt |
| 12:43 |
IhrFussel |
4.7 Ghz?? I thought CPUs would potentially burn after 4 Ghz |
| 12:43 |
Megaf |
"8-core processor with 5.0 GHZ Max Turbo speed, and 4.7 GHz CPU base. Fully unlocked." |
| 12:43 |
Megaf |
this one can go to 5 GHz |
| 12:44 |
Megaf |
sfan5: at 5 GHz performance will be good for MT |
| 12:45 |
sfan5 |
no |
| 12:45 |
Megaf |
learned a new word today, dodeca. Means 12 |
| 12:45 |
Megaf |
dodeca core, 12 cores |
| 12:46 |
IhrFussel |
Do you guys also have that problem with carts? Sometimes when you turn you are "stuck" for a few secs before you can drive further...I mean it's like the cart resets you a lot before it allows you to proceed |
| 12:46 |
IhrFussel |
Maybe I can make a GIF one sec |
| 12:47 |
Megaf |
ok, now talking truthly |
| 12:47 |
Megaf |
https://www.newegg.com/Product/Product.aspx?Item=N82E16819117792 |
| 12:47 |
Megaf |
!title |
| 12:47 |
MinetestBot |
Megaf: Intel Core i7-7740X Kaby Lake-X Quad-Core 4.3 GHz LGA 2066 112W BX80677I77740X Desktop Processor - Newegg.com |
| 12:47 |
Megaf |
that would run MT great |
| 12:47 |
Megaf |
great single core performance, not many threads, just enough to run MT at full speed |
| 12:48 |
Megaf |
sfan5: ^ |
| 12:48 |
Megaf |
IhrFussel: another 4+ GHz CPU |
| 12:48 |
sfan5 |
yes such a cpu would be better |
| 12:49 |
Megaf |
that CPU with two cores and HT disabled, running at 5 GHz with a real time kernel, so sweet |
| 12:49 |
sfan5 |
Megaf: an i3 with roughly the same price can deliver better single core perf than the FX https://www.cpubenchmark.net/compare.php?cmp[]=2014&cmp[]=2317 |
| 12:49 |
Megaf |
sfan5: I know, FX were terrible |
| 12:49 |
Megaf |
I always disliked their architecture. reason I still have a Phenom II X4 |
| 12:50 |
Megaf |
any modern high clock Intel CPU will be good for MT |
| 12:51 |
Megaf |
I'm quite found of i3s, they deliver good performance and are very cheap |
| 12:51 |
Megaf |
https://www.newegg.com/Product/Product.aspx?Item=N82E16819117772 |
| 12:51 |
Megaf |
!titel |
| 12:51 |
Megaf |
!title |
| 12:51 |
MinetestBot |
Megaf: Intel Core i3-7350K Kaby Lake Dual-Core 4.2 GHz LGA 1151 60W BX80677I37350K Desktop Processor Intel HD Graphics 630 - Newegg.com |
| 12:53 |
IhrFussel |
10 MB GIF warning > http://94.16.121.151/mt2.gif |
| 12:54 |
IhrFussel |
It's hard to capture it cause it only happens for a moment |
| 12:54 |
Megaf |
11 MB |
| 12:54 |
IhrFussel |
And it's not lag max_lag was 0.2 there |
| 12:55 |
IhrFussel |
It only happens at turns or when an obstacle appears ... it seems to fix itself faster when you turn the camera around |
| 12:59 |
IhrFussel |
Here I was able to capture it better and in MP4 http://94.16.121.151/mt3.mp4 |
| 13:00 |
IhrFussel |
Again max_lag is 0.2 |
| 13:03 |
IhrFussel |
Plus it doesn't happen with the boost_cart mod last I checked |
| 13:12 |
Fixer |
ah, """bestest""" european winter |
| 13:15 |
|
aerozoic joined #minetest-hub |
| 13:20 |
nerzhul |
IhrFussel, i have a 8350 at 4.2 Ghz it rocks :D |
| 13:20 |
nerzhul |
8 cores, MT compiled in 25sec |
| 13:25 |
|
Mr_Pardison joined #minetest-hub |
| 13:33 |
Megaf |
nerzhul: 25 seconds :O |
| 13:33 |
Megaf |
for minetestserver or minetest client? |
| 13:33 |
Megaf |
Hello Mr_Pardison |
| 13:33 |
Mr_Pardison |
o/ Megaf |
| 13:33 |
Megaf |
!server Megaf |
| 13:33 |
MinetestBot |
Megaf: Megaf Server V4.17 | mt.megaf.info:30003 | Clients: 0/10, 0/1 | Version: 0.4.17-Megaf / MegafXploreNext | Ping: 13ms |
| 13:34 |
Megaf |
sfan5: what do you think about including uptime in that output? ^ |
| 13:35 |
sfan5 |
sounds useful |
| 13:35 |
Megaf |
Could you do that for us? Please? |
| 13:39 |
|
EvergreenTree joined #minetest-hub |
| 13:39 |
nerzhul |
Megaf, both |
| 13:40 |
Megaf |
Wow, I can't beat that |
| 13:40 |
Megaf |
Maybe I can build minetestserver in 25 seconds |
| 13:40 |
Megaf |
both, no way |
| 13:40 |
Megaf |
only using ccache |
| 13:41 |
Mr_Pardison |
Communism: you have 2 cows. You take care of them, but that government takes all the milk. |
| 13:41 |
Mr_Pardison |
Capitalism: You have two cows. You sell one and buy a bull. You herd multiplies, and the economy grows. You sell them and retire on the income. |
| 13:41 |
Mr_Pardison |
Enron Capitalism: You have two cows. You sell three of them to your brother-in-law , who then sells you all four cows back, which gives you five cows in the annual report which gives you a tax exemption for eight cows. But you're not aware of it at the time. |
| 13:42 |
Megaf |
Brazil ^ |
| 13:42 |
Megaf |
People and politians do that all the time |
| 13:43 |
Megaf |
very efficient |
| 13:43 |
Mr_Pardison |
they use mark-to-market accounting? |
| 13:43 |
Megaf |
Sometimes it gets you in jail though, then you have to sell give away around 4 cows to buy freedom |
| 13:44 |
Megaf |
s/sell give away/give away |
| 13:50 |
Megaf |
*unplugs ethernet cable from switch |
| 13:50 |
Megaf |
*gets angry when ssh doesnt connect |
| 13:50 |
Megaf |
Me right now ^ |
| 13:59 |
|
Raven262 joined #minetest-hub |
| 14:03 |
Megaf |
Welcome back Raven262 |
| 14:04 |
Raven262 |
thx Megaf |
| 14:06 |
IhrFussel |
nerzhul, 8 cores are useless in MT |
| 14:07 |
nerzhul |
IhrFussel, for compiling it's useful, i don't run server on my PC |
| 14:07 |
IhrFussel |
It really only counts what *one* core can do |
| 14:08 |
IhrFussel |
Yes for compiling every core can be used, but those minutes don't matter much as server owner |
| 14:09 |
IhrFussel |
I wish we could at least split some tasks from the server thread into a new one |
| 14:09 |
benrob0329 |
Lua threads! |
| 14:09 |
Mr_Pardison |
sew! a needle pulling thread! |
| 14:12 |
IhrFussel |
How about splitting ABMs into another thread? Would that make sense? |
| 14:13 |
IhrFussel |
ABMs run pretty much independently right? |
| 14:13 |
Megaf |
we just have to bundle this with our Lua http://lualanes.github.io/lanes/ |
| 14:16 |
nerzhul |
no sense as the map is common and this will trigger many locks, it's not easy to split that |
| 14:18 |
IhrFussel |
nerzhul, isn't there a way to just trigger the ABM thread whenever the usual server thread would trigger the ABMs? |
| 14:20 |
IhrFussel |
And then the Server thread could do something non-critical until the ABM thread finished ... just a thought |
| 14:33 |
rubenwardy |
ABMs need to be on the server thread as Lua |
| 14:34 |
rubenwardy |
would be good to make Lua on a different thread, and communicate with a queue. Lots of things would need to be blocking however, namely the callbacks that affect inbuilt behaviour |
| 14:34 |
rubenwardy |
heh |
| 14:35 |
rubenwardy |
could maybe make our map storage multithreaded, achieve massive I/O |
| 14:48 |
|
rubenwardy left #minetest-hub |
| 14:50 |
|
benrob0329 joined #minetest-hub |
| 14:53 |
|
lumberJ joined #minetest-hub |
| 14:56 |
Megaf |
you can make Lua and SQLite very fast in MT |
| 14:56 |
Megaf |
problem is, if the server crashes you are screwed |
| 14:57 |
|
rubenwardy joined #minetest-hub |
| 14:57 |
Mr_Pardison |
*helically wrapped around an inclined plane |
| 14:57 |
Megaf |
there's some nice building options for Lua and SQLite that makes them totally async |
| 14:57 |
Megaf |
[12:56:34] <Megaf> you can make Lua and SQLite very fast in MT |
| 14:57 |
Megaf |
[12:56:45] <Megaf> problem is, if the server crashes you are screwed |
| 14:57 |
rubenwardy |
sqlite is still limited by disk I/O |
| 14:57 |
rubenwardy |
and you can't perform operations async |
| 14:58 |
rubenwardy |
with a highly distributed map backend, you could read/write millions of nodes in seconds |
| 14:58 |
rubenwardy |
source: I have a friend that did this for Minecraft |
| 14:58 |
Megaf |
rubenwardy: https://sqlite.org/asyncvfs.html |
| 14:59 |
rubenwardy |
? |
| 14:59 |
Megaf |
async module for sqlite? |
| 14:59 |
rubenwardy |
how would you use that to set all water to lava? |
| 14:59 |
rubenwardy |
that's what I meant by read/write |
| 15:01 |
rubenwardy |
I mean, asking for this is too much given that entities aren't even stored in a simple constantly bounded spatial partition |
| 15:01 |
rubenwardy |
they're stored in a list |
| 15:01 |
rubenwardy |
what is optimisation again? |
| 15:01 |
jas_ |
no clue |
| 15:01 |
jas_ |
what's wrong with lists? |
| 15:01 |
rubenwardy |
O(N^2) for collisions on entire set of loaded entities |
| 15:02 |
jas_ |
a lot of whooshing going on around here |
| 15:02 |
jas_ |
(seriously, heavy gusts!) |
| 15:02 |
rubenwardy |
basically every entity needs to compare itself against all other entities for collisions |
| 15:02 |
rubenwardy |
because they're all stored in one list |
| 15:03 |
jas_ |
ouch |
| 15:03 |
rubenwardy |
rather than a spartial partition, where they'd check all the entities in the same chunk, and then entities in neighbouring chunks if close enough to the border |
| 15:03 |
jas_ |
i think spatial, i think AreaStore, which i like |
| 15:03 |
jas_ |
that was very fast, iirc, so much so that areas was converted to use it |
| 15:03 |
jas_ |
is it related somehow? |
| 15:03 |
rubenwardy |
yeah, AreaStore is an API for spatial partitioning |
| 15:04 |
jas_ |
too bad about the entities, i suspect the server gets slower the longer it's played, and the more entities are spawned -- if one is in a peculiar position, it could cause errors |
| 15:06 |
jas_ |
oh i see your comment about AreaStore now, on this gh issue |
| 15:09 |
Fixer |
quassel on start reads all my fonts, 2 second wasted, why? |
| 15:10 |
jas_ |
because |
| 15:11 |
benrob0329 |
Because Hexchat is better :P |
| 15:11 |
jas_ |
weeeeeeeeeeeeeeechat |
| 15:11 |
benrob0329 |
Actually I'm pondering an IRC client, you could do p2p VoIP using DCC :P |
| 15:12 |
benrob0329 |
That is, if DCC worked in any modern clients anyways |
| 15:12 |
jas_ |
https://www.reddit.com/r/Maine/comments/7o3vku/hannafids_last_night/ |
| 15:13 |
|
sniper570 joined #minetest-hub |
| 15:14 |
benrob0329 |
jas_: weechat is good, but I got tired of having to constantly retype all autojoin channels just to change one |
| 15:14 |
jas_ |
? |
| 15:14 |
jas_ |
i dunno about that, there's a script |
| 15:14 |
jas_ |
plus u can use iset too, and alt-k it |
| 15:14 |
jas_ |
but there's a script |
| 15:14 |
jas_ |
just type /script, i gotta rescue my pizza |
| 15:15 |
jas_ |
and search join or autojoin, i think the script is called autojoin |
| 15:15 |
jas_ |
and don't forget go, that's great |
| 15:15 |
* benrob0329 |
should write some lua for Hexchat |
| 15:15 |
benrob0329 |
I tried last night for an [off] script but ended up crashing everything lol |
| 15:15 |
jas_ |
there's a lua plugin for weechat |
| 15:16 |
benrob0329 |
Ik |
| 15:16 |
jas_ |
but the best plugin, for me, is the relay plugin |
| 15:16 |
jas_ |
allowing me to access the box from an irc relay, or a weechat relay |
| 15:18 |
|
nerzhul joined #minetest-hub |
| 15:19 |
jas_ |
well i guess the irc plugin is the most important |
| 15:23 |
|
Krock joined #minetest-hub |
| 15:25 |
jas_ |
i burnt it a little bit |
| 15:26 |
jas_ |
https://weechat.org/scripts/source/autojoin.py.html/ |
| 15:27 |
jas_ |
but using /iset and alt+k works too, so u dont have to type a lot |
| 15:34 |
|
sniper570 joined #minetest-hub |
| 15:48 |
benrob0329 |
o/ Krock, Wb sniper570 |
| 15:48 |
Krock |
o/ benrob0329 |
| 16:14 |
sniper570 |
\o benrob0329 |
| 16:19 |
Fixer |
/o\ |
| 16:19 |
Mr_Pardison |
\o/ |
| 16:30 |
|
IhrFussel joined #minetest-hub |
| 16:42 |
IhrFussel |
My cars still suddenly stop at a certain speed... could it be caused by the send optimize distance?? I set that to 3 ... it is definitely not the HDD or RAM or CPU |
| 16:42 |
Krock |
do they also stop when running into unloaded areas? |
| 16:43 |
IhrFussel |
That never happens, the map loads too fast |
| 16:43 |
IhrFussel |
One sec will make a MP4 |
| 16:46 |
IhrFussel |
Krock, http://94.16.121.151/mt4.mp4 |
| 16:47 |
IhrFussel |
Window may be a bit small |
| 16:49 |
Krock |
that's indeed strange. Either it's the collision code that does weird stuff or the velocity is reset in the mod somehow due wrong calculations |
| 16:51 |
IhrFussel |
It worked in 0.4.16 fine ... now the server is 0.4.17 and it broke I guess |
| 16:52 |
IhrFussel |
But maybe it really is the block optimize send distance...I removed it and will see if it's fixed after restart |
| 16:56 |
rubenwardy |
how does one get the kernel update to protect against meltdown? |
| 16:56 |
rubenwardy |
can't see anything special looking in apt upgrade |
| 16:57 |
sofar |
once I know, I'll let everyone know too |
| 16:57 |
sofar |
lol |
| 16:57 |
IhrFussel |
You just have to enable PTI (Page Table Isolation) AFAIK |
| 16:58 |
IhrFussel |
If anyone got time, please check if the movement code here maybe has a bug https://pastebin.com/s4ETkJdS |
| 17:01 |
IhrFussel |
rubenwardy, https://en.wikipedia.org/wiki/Kernel_page-table_isolation ... as for OS update it will just be a regular security update by Debian/Ubuntu etc |
| 17:02 |
IhrFussel |
The bad part is that they are not done with the patch yet (at least Ubuntu isn't) |
| 17:04 |
IhrFussel |
Regarding Ubuntu they will update this page as soon as a patch has been released https://wiki.ubuntu.com/SecurityTeam/KnowledgeBase/SpectreAndMeltdown |
| 17:06 |
IhrFussel |
"The original coordinated disclosure date was planned for January 9 and we have been driving toward that date to release fixes. Due to the early disclosure, we are trying to accelerate the release, but we don't yet have an earlier ETA when the updates will be released." |
| 17:07 |
IhrFussel |
2017 Nov 09: the Ubuntu Security team is notified by Intel under NDA << They had almost 2 MONTHS to fix this and aren't done yet |
| 17:07 |
|
Darcidride_ joined #minetest-hub |
| 17:08 |
Mr_Pardison |
this affects all OS's. |
| 17:09 |
IhrFussel |
Ubuntu specifies Spectre as "High" while they specify Meltdown as "Critical" ... IMO both are critical |
| 17:09 |
* Mr_Pardison |
agrees |
| 17:09 |
benrob0329 |
Ubuntu, the OS to rule no one but everyone seems to think they do. |
| 17:17 |
IhrFussel |
benrob0329, that date apparently was the CRD for everyone if this article is right "there was a coordinated release date of January 9, 2018, agreed upon by essentially every operating system, hardware, and cloud vendor in the world." |
| 17:17 |
Mr-Pardison |
makes sense. |
| 17:18 |
Mr-Pardison |
release the patch at the same time so that it couldn't be exploited on a system that didn't have the patch. |
| 17:19 |
IhrFussel |
I wonder if they would have kept it secret if the public wouldn't know about it until Jan 9 |
| 17:20 |
benrob0329 |
Probably |
| 17:21 |
Mr-Pardison |
Fedora alerted me just now that important OS updates are available. |
| 17:23 |
sofar |
are you kidding me? |
| 17:23 |
sofar |
"<benrob0329> Probably" |
| 17:23 |
sofar |
seriously? |
| 17:24 |
sofar |
"there was a coordinated release date of January 9, 2018, agreed upon by essentially every operating system, hardware, and cloud vendor in the world." |
| 17:24 |
sofar |
I mean, wat? |
| 17:24 |
sofar |
how do you? |
| 17:24 |
sofar |
even? |
| 17:25 |
sofar |
http://s2.quickmeme.com/img/f7/f71beaed2b2a0d18a081fe28448c68ec43cc052c8c1539addfdaeede78e59800.jpg |
| 17:25 |
benrob0329 |
Lol, I didn't say that keeping is secret until it was fixed was a bad thing |
| 17:27 |
benrob0329 |
IhrFussel: [...] until Jan 9 |
| 17:27 |
IhrFussel |
I think there is a very good chance that no vendor/manufacturer etc would've said ANYTHING about the vulnerabilities if it wouldn't have been gone public |
| 17:27 |
Megaf |
off, I shoud attach some wings to my two servers |
| 17:27 |
Megaf |
they sound like they are about to take off |
| 17:27 |
rubenwardy |
lol, offtopic by talking about minetest |
| 17:27 |
rubenwardy |
gtfu |
| 17:29 |
IhrFussel |
I mean we get security patches all the time and never really know what it was about (well you can read about many things that get patched but you don't see always all details) |
| 17:34 |
|
benrob0329 joined #minetest-hub |
| 18:21 |
|
nerzhul joined #minetest-hub |
| 18:47 |
IhrFussel |
I tried to compile 0.4.17 client and when I run it I get "2018-01-05 19:45:36: ERROR[AsyncWorker-0]: Execution of async base environment failed: Failed to load and run script from /usr/local/share/minetest/builtin/init.lua:" |
| 18:48 |
IhrFussel |
Wait maybe I know what the problem is |
| 19:07 |
IhrFussel |
Ok it's fixed ... but now I can confirm that the cars also stop in 0.4.17 client ... It could be something server side |
| 19:07 |
Fixer |
*,,,,,,,* |
| 19:09 |
IhrFussel |
Maybe a PR that was backported |
| 19:09 |
Fixer |
sofar: btw not everyone will receive new update on Windows, since some AV shitware goes into BSOD, so it checks if you have outdated one and disables update, had to manually install it |
| 19:09 |
Fixer |
security quality update one for Win7 |
| 19:09 |
sofar |
I don't windows |
| 19:10 |
Fixer |
about that meltdown/spectre vulnderability |
| 19:10 |
Fixer |
i decided to disable that Avast for now, to just feel if there is improvement in responsiveness |
| 19:10 |
Fixer |
win7 is slow as hell on HDD |
| 19:11 |
Fixer |
running ATI Control Centre and Greenshot apps written in .NET 2 and 4 literally rape my HDD |
| 19:11 |
Fixer |
during boot |
| 19:12 |
Fixer |
used delayed start on a lot of stuff, like Shype and torrents, boots in 50 sec now, year ago it was 300 sec |
| 19:12 |
Fixer |
AV alone adds entire 1 minute to boot time, it is ridiculous |
| 19:13 |
Fixer |
quassel went to delayed mode too, it is painfully slow to start, amazingly |
| 19:14 |
Krock |
> literally rape |
| 19:14 |
Fixer |
4 sec to read a quassel database... it is not like 100gb in size, and it is not 386 cpu, WTF |
| 19:15 |
Krock |
there can't be too much physical damage to a HDD by read/write cycles |
| 19:15 |
Fixer |
it is just 256mb, lol |
| 19:15 |
Fixer |
god damned |
| 19:15 |
Krock |
symlink it to a ramdisk :P |
| 19:16 |
Fixer |
it also wastes TWO seconds on reading all my /FONTS, WHY? |
| 19:16 |
Fixer |
i only use 3 fonts at most |
| 19:16 |
Fixer |
in quassel |
| 19:17 |
Mr-Pardison |
g2g cya in a bit. |
| 19:17 |
Fixer |
https://i.imgur.com/dx2stIu.png DANGEROUS |
| 19:18 |
Krock |
use it. it's there for a reason |
| 19:22 |
|
Mr_Pardison joined #minetest-hub |
| 19:36 |
|
compunerd joined #minetest-hub |
| 19:37 |
Fixer |
Krock: you like to live dangerously |
| 19:39 |
Mr_Pardison |
Fixer: life is more fun that way :D |
| 19:39 |
Krock |
not really. I just find it quite amusing that they documented such a parameter if people should better not use it |
| 19:40 |
Fixer |
"16 dec 2014 - If you're using HDMI output with an Nvidia graphics card, you likely aren't getting full and accurate color range, specifically at the black and white ends of the spectrum." |
| 19:40 |
Fixer |
o__________O |
| 19:40 |
Fixer |
patch available |
| 19:40 |
IhrFussel |
Stupid German media "Browsers are receiving patches against Meltdown & Spectre" That will fix NOTHING |
| 19:41 |
Fixer |
it is like back to 90s |
| 19:41 |
Krock |
but without gifs or 56k modems |
| 19:41 |
Fixer |
https://pcmonitors.info/articles/correcting-hdmi-colour-on-nvidia-and-amd-gpus/ o.O |
| 19:44 |
sofar |
IhrFussel: fixing browsers is critical |
| 19:44 |
sofar |
since that is where you run untrusted code |
| 19:45 |
red-001 |
huh I wonder how this affects luaJIT |
| 19:45 |
IhrFussel |
sofar, the average joe is dumber than just visiting stranger websites |
| 19:45 |
sofar |
oh sure don't patch your browser, you'll be better of for sure (said nobody ever) |
| 19:46 |
sofar |
the dumber they are the more serious you should be about them updating their browser |
| 19:46 |
IhrFussel |
There are many ways for those exploits to work... outside of browsers |
| 19:48 |
sofar |
I guess you don't understand security defense strategies |
| 19:48 |
sofar |
"Defense in depth" |
| 19:50 |
IhrFussel |
I never said that patching browsers is wrong ... but if the German media reports it like THAT, some people might think that they are safe then |
| 19:51 |
sofar |
most consumers will get OS patches automatically through Microsoft/Apple |
| 19:51 |
IhrFussel |
Nobody can fix Spectre 100% right now |
| 19:51 |
Fixer |
Keep your OS and programs up to date (except those that degrade over time), use adblocker, noscript, remove flash/java, done |
| 19:51 |
sofar |
reminding them to patch their browser where they run random code from the internet as they surf german porn sites seems like a good idea |
| 19:51 |
IhrFussel |
The fixes are mainly just for Meltdown |
| 19:52 |
sofar |
"oh sure just postpone patching then" |
| 19:53 |
sofar |
if anything it's amazing the press is trying to make sense of it all and tell users what they can do, right? |
| 19:53 |
sofar |
I mean, it's almost like... they're doing what they should do |
| 19:54 |
Mr_Pardison |
it's like one of those "FINALLY!" moments. |
| 19:54 |
Fixer |
sofar: lol, yes |
| 19:56 |
sofar |
here's an image of germans after they get instructions on how to patch their computer: https://i.imgur.com/nb0AsTE.jpg |
| 19:56 |
Krock |
that's how native germans look like |
| 19:57 |
sofar |
:) |
| 19:57 |
sofar |
wait |
| 19:57 |
Mr-Pardison |
>.< |
| 19:57 |
sofar |
are you implying that there's "positive" germans? |
| 19:57 |
sofar |
as a dutch person, I was not aware |
| 19:58 |
red-001 |
can't get viruses if you burn your computer can you? |
| 19:58 |
sofar |
just cancer |
| 19:58 |
Krock |
no, that's bacteria that gets killed by heat |
| 20:01 |
Fixer |
colorized |
| 20:02 |
Fixer |
sofar: updating helps, it is just long legacy of WinCLF__K ecosystem with painful OS updating and absense of good update mechanism for programs itself |
| 20:02 |
Fixer |
Nix systems are very good in this aspect |
| 20:02 |
Fixer |
win ecosystem is minefield too |
| 20:03 |
Fixer |
while they push store for now, i dislike the way they do it |
| 20:03 |
Megaf |
That chat... ˆ^ˆ^ˆ^ˆ^ˆ^ |
| 20:20 |
|
EvergreenTree joined #minetest-hub |
| 20:39 |
Calinou |
https://forum.minetest.net/viewtopic.php?f=3&t=19208 |
| 20:39 |
Calinou |
nice |
| 20:39 |
red-001 |
yeah |
| 20:39 |
Calinou |
I just reminded myself that Minetest's birthdate is October 10th 2010 |
| 20:40 |
Calinou |
…which is the same day on which My Little Pony: Friendship is Magic first aired, by the way |
| 20:40 |
Calinou |
(:D) |
| 20:40 |
Calinou |
or Ubuntu 10.10's release |
| 20:41 |
red-001 |
wait those are 3 diffirent things? |
| 20:49 |
benrob0329 |
Calinou: noice |
| 20:50 |
rubenwardy |
ew |
| 20:50 |
benrob0329 |
Ew? |
| 20:52 |
red-001 |
ew |
| 20:55 |
Calinou |
yes, Ubuntu 10.10 was a meh release |
| 20:55 |
Krock |
Calinou, also minetest had a nyan cat with rainbow nodes. There's a pony with a rainbow tail. Illuminati confirmed |
| 20:57 |
Calinou |
illumineighti! |
| 21:01 |
Mr-Pardison |
g2g cya tomorrow peeps. |
| 21:02 |
|
Mr-Pardison left #minetest-hub |
| 21:52 |
IhrFussel |
Is MT safe from a sudden shutdown? My hoster says they will now shutdown all systems for the CPU bugfixes but in waves, so I cannot know when my machine will be shutdown |
| 21:53 |
Calinou |
[ 0.000000] Kernel/User page tables isolation: enabled |
| 21:53 |
IhrFussel |
I mean is the map.sqlite safe then? |
| 21:53 |
Calinou |
I booted the patched kernel from Fedora today |
| 21:54 |
Calinou |
IhrFussel: SQLite doesn't like very much brutal shutdowns |
| 21:54 |
red-001 |
^ |
| 21:54 |
Calinou |
so it's best to gracefully shut down your server before |
| 21:54 |
Calinou |
(/shutdown or Ctrl+C once) |
| 21:54 |
Calinou |
(or stopping the systemd service if you used that) |
| 21:54 |
IhrFussel |
Not brutal, my hoster says they will send a regular shutdown signal which waits 300 secs for all processes to finish |
| 21:55 |
IhrFussel |
Or to clean up |
| 21:57 |
|
Fixer joined #minetest-hub |
| 22:16 |
IhrFussel |
Calinou, I read that SQLite is *designed* to recover from any unexpected situations (OS crash, power failure etc) |
| 22:18 |
Calinou |
IhrFussel: the best it can, yes |
| 22:19 |
Calinou |
but it's not perfect at doing this, like any program |
| 22:19 |
rubenwardy |
it'll probably be fine as long as you haven't changed sqlite_synchronous |
| 22:20 |
IhrFussel |
I have no such setting in my conf |
| 22:20 |
Calinou |
it's set to 2 by default (safest, but slowest) |
| 22:20 |
Calinou |
0 is the least safe, but can be much faster (sometimes 50 times faster) |
| 22:20 |
Calinou |
you should use that one only if you have regular backups |
| 22:21 |
Calinou |
(but you should use PostgreSQL anyway…) |
| 22:30 |
|
paramat joined #minetest-hub |
| 22:34 |
|
benrob0329 joined #minetest-hub |
| 23:09 |
red-001 |
it's just a frontend to lua loadstring/pcall combo |
| 23:32 |
Megaf |
IhrFussel: I think no server has crashed more than mine in its 5 or 6 years of life |
| 23:32 |
Megaf |
to this day, no map corruption |
| 23:33 |
Megaf |
And I did change sqlite syncronous options to less safe options |
| 23:34 |
Megaf |
well, everything else got currupted and I had to rebuild several times... |
| 23:34 |
Megaf |
auth, players, homes, landmarks, world.mt, and so on... |
| 23:35 |
Megaf |
but map.sqlite was always rock solid |
| 23:36 |
Megaf |
The most often crash I had lately was OOM, seconded by indexing nil values, then maybe seg faults and no more disk space |
| 23:36 |
Megaf |
all of them rather frustrating |
| 23:55 |
|
Fixer joined #minetest-hub |
| 23:57 |
Fixer |
reading about that meltdown/spectre thing, pretty neat stuff |
| 23:57 |
Fixer |
the more you read - more creepy it gets |
| 23:58 |
Fixer |
IhrFussel: you should check minetest db performance before/after that KPTI patch |
| 23:58 |
Fixer |
https://github.com/HarsaroopDhillon/Meltdown who wants to try this? :D |
