| Time |
Nick |
Message |
| 00:15 |
|
Lesha_Vel joined #minetest |
| 00:24 |
|
frostsnow joined #minetest |
| 01:19 |
|
YuGiOhJCJ joined #minetest |
| 01:20 |
|
smk joined #minetest |
| 01:48 |
MTDiscord |
<RisingLeaf> I think there already is a tides mod |
| 02:39 |
|
Trifton joined #minetest |
| 02:42 |
|
Trifton joined #minetest |
| 02:45 |
|
Trifton joined #minetest |
| 02:48 |
|
Trifton joined #minetest |
| 02:54 |
|
FavoritoHJS joined #minetest |
| 03:39 |
|
lemonzest joined #minetest |
| 04:12 |
|
Zambyte` joined #minetest |
| 04:34 |
|
fluxionary joined #minetest |
| 05:00 |
|
MTDiscord joined #minetest |
| 08:01 |
|
TomTom joined #minetest |
| 08:14 |
|
calcul0n joined #minetest |
| 08:16 |
|
mrkubax10 joined #minetest |
| 08:28 |
|
dabbill_ joined #minetest |
| 08:31 |
|
olliy1or joined #minetest |
| 08:53 |
sfan5 |
paradust: no |
| 09:03 |
|
proller joined #minetest |
| 09:46 |
|
gxt joined #minetest |
| 09:53 |
|
proller joined #minetest |
| 10:37 |
|
proller joined #minetest |
| 11:44 |
|
gxt joined #minetest |
| 11:55 |
|
kabou joined #minetest |
| 12:25 |
|
ytterbium joined #minetest |
| 12:40 |
|
mrkubax10 left #minetest |
| 12:43 |
|
definitelya joined #minetest |
| 12:44 |
|
Trifton joined #minetest |
| 13:13 |
|
ytterbium1 joined #minetest |
| 13:13 |
|
dv^_^8 joined #minetest |
| 13:13 |
|
definitelya_ joined #minetest |
| 13:14 |
|
smk_ joined #minetest |
| 13:14 |
|
TomTom_ joined #minetest |
| 13:15 |
|
Trifton_ joined #minetest |
| 13:15 |
|
luk3yx joined #minetest |
| 13:15 |
|
dabbill- joined #minetest |
| 13:17 |
|
dzho_ joined #minetest |
| 13:17 |
|
search_social joined #minetest |
| 13:17 |
|
valhalla_ joined #minetest |
| 13:18 |
|
amfl2 joined #minetest |
| 13:22 |
|
jonadab joined #minetest |
| 13:47 |
|
definitelya__ joined #minetest |
| 13:51 |
|
mazes_80 joined #minetest |
| 14:56 |
|
mazes_80 joined #minetest |
| 15:06 |
|
mrkubax10 joined #minetest |
| 15:14 |
|
Fixer joined #minetest |
| 15:23 |
|
fluxionary joined #minetest |
| 15:27 |
|
FavoritoHJS joined #minetest |
| 15:30 |
|
proller joined #minetest |
| 15:36 |
|
frostsnow joined #minetest |
| 15:45 |
|
kilbith joined #minetest |
| 16:28 |
|
Oksanaa joined #minetest |
| 16:50 |
|
proller joined #minetest |
| 16:54 |
|
iamweasel joined #minetest |
| 17:08 |
kabou |
sfan5 have you seen my updates after your review of #13168 ? |
| 17:08 |
ShadowBot |
https://github.com/minetest/minetest/issues/13168 -- Improve the documentation for chat command definition in lua_api.txt. by kab0u |
| 17:09 |
sfan5 |
not yet |
| 17:09 |
kabou |
oh okay |
| 17:11 |
|
___nick___ joined #minetest |
| 17:15 |
|
kabou joined #minetest |
| 17:54 |
|
dunj3 joined #minetest |
| 17:59 |
|
lissobone joined #minetest |
| 17:59 |
lissobone |
Hello. |
| 18:02 |
|
ytterbium joined #minetest |
| 18:02 |
lissobone |
Hello too. |
| 18:03 |
MTDiscord |
<Bastrabun> Is there something wrong with the forum? Doesn't respond at all |
| 18:03 |
lissobone |
In what sense "respond"? |
| 18:03 |
ROllerozxa |
forum is just dying as usual |
| 18:03 |
lissobone |
Yeah. |
| 18:04 |
MTDiscord |
<Bastrabun> Repsond as in "I request a URL and nothing happens" |
| 18:06 |
ROllerozxa |
celeron55 "fixed" the server a week or so ago which has just made it worse somehow |
| 18:06 |
MTDiscord |
<Bastrabun> May I ask, who hosts the forum and what do we have to do to get a forum that does what it's meant to do? Is it a question of money, expertise, time? Do we need to an assassin to transfer ownership? |
| 18:06 |
lissobone |
According to my calculations and estimations (the same ones that approximated GNU GPLv4 release to be in 2026.1), the forum will have persisted on serving its respectable hackers by 2023-02-25T05:02+09:00 (KST). |
| 18:07 |
ROllerozxa |
well, it has a higher chance of responding relatively quickly now, but it's not particularly large of an improvement |
| 18:08 |
MTDiscord |
<MisterE> This is just expected behavior |
| 18:08 |
MTDiscord |
<Bastrabun> How can we make it that we see larger improvements? The current owner cannot be happy with the state it's in? |
| 18:08 |
lissobone |
Note: I am not responsible for any damages, nuclear holocausts and Pestilences caused by imprecise calculations. |
| 18:08 |
lissobone |
Larger improvements? |
| 18:09 |
DeepT |
its responsive enough at this second |
| 18:09 |
lissobone |
Hosting requires cash for maintenance and even more cash for hardware upgrades. |
| 18:09 |
lissobone |
Time is money. |
| 18:09 |
MTDiscord |
<Bastrabun> Ok, so its a question of money? |
| 18:09 |
DeepT |
most things are |
| 18:09 |
DeepT |
:P |
| 18:09 |
lissobone |
Not only money... |
| 18:10 |
lissobone |
Magnetic fields. |
| 18:10 |
MTDiscord |
<Bastrabun> No, this very specific problem: Can it be solved with money? |
| 18:10 |
ROllerozxa |
c55 already gets donations for covering server costs |
| 18:10 |
MTDiscord |
<MisterE> @celeron55 ... |
| 18:10 |
MTDiscord |
<Bastrabun> So money is available. What's the holdup then? Time? Expertise? |
| 18:11 |
lissobone |
Well, if each minetest player donates one dollar, c55 will be able to cover all hosting costs. |
| 18:11 |
ROllerozxa |
it's just something about the server that makes it godawful slow, I'm almost certain it doesn't have to do with the specs, the donation money should pay for something way better than the current state of it |
| 18:12 |
MTDiscord |
<Bastrabun> So it's a question of expertise? The current admin doesn't know how to fix it? |
| 18:13 |
lissobone |
I started wondering if it's possible to run such a server on GNU Hurd. |
| 18:13 |
lissobone |
Probably yes. |
| 18:13 |
lissobone |
Maybe the forum needs a special department? |
| 18:13 |
mrkubax10 |
run server on ReactOS |
| 18:13 |
lissobone |
It will be for server maintenance discussions. |
| 18:14 |
lissobone |
(Genious, isn't it?) |
| 18:14 |
lissobone |
You know what? |
| 18:14 |
lissobone |
You know what's funny? |
| 18:14 |
MTDiscord |
<ROllerozxa> https://cdn.discordapp.com/attachments/749727888659447960/1067507781965463702/image.png |
| 18:15 |
lissobone |
Tls? |
| 18:15 |
lissobone |
Increased [$!$$]? |
| 18:15 |
lissobone |
Is the server being DDOSed or something? |
| 18:16 |
lissobone |
Is there someone training an image processing neural network on it?? |
| 18:17 |
|
Talkless joined #minetest |
| 18:18 |
lissobone |
Maybe some brazilian blogger released a video again, causing these load spikes? |
| 18:18 |
lissobone |
Anyways, here's what I wanted to say regarding the funniest thing as of today. |
| 18:19 |
mrkubax10 |
blog post mentioned some youtuber recording video about Minetest |
| 18:19 |
lissobone |
It's re-e-e-ally funny how libera houses both #fsf and #windows. |
| 18:19 |
mrkubax10 |
lol |
| 18:20 |
lissobone |
And why is windows here? Isn't libera called "libera" for a reason? |
| 18:21 |
lissobone |
Blog post? So it's likely that minetest is experiencing yet another popularity spike, isn't it? |
| 18:21 |
rubenwardy |
not right now |
| 18:21 |
lissobone |
So what is it? |
| 18:21 |
ROllerozxa |
you can see the amount of users online at any given time on the forums, there's no spike of users right now |
| 18:21 |
lissobone |
Hm-m. |
| 18:21 |
lissobone |
Right. |
| 18:21 |
lissobone |
I missed that. |
| 18:22 |
mrkubax10 |
ah it's from November |
| 18:22 |
MTDiscord |
<MisterE> Blog post for January, next weekish, I sincerely hope |
| 18:22 |
MTDiscord |
<Bastrabun> So MT finally gets attention and we're delivering a "no they cant" picture? Oh my 😄 C'mon, what do we have to do to get a working forum? |
| 18:22 |
rubenwardy |
it's a matter of time and working out the problem |
| 18:23 |
lissobone |
Perhaps certain proprietary software components are sucking out the server's capabilities? |
| 18:23 |
lissobone |
(I am not being serious right now.) |
| 18:23 |
lissobone |
(Or am I?) |
| 18:24 |
MTDiscord |
<Bastrabun> Ok, so there is a technical problem and it needs time to be worked on. Is there something we can contribute? We certainly don't need the admin have to come up with a solution all by themselves? |
| 18:25 |
lissobone |
We can't really directly influence the server's computations, but we can discuss that publicly with celeron. |
| 18:27 |
lissobone |
It's really great how contentdb and minetest.net have no nonfree JS. |
| 18:27 |
lissobone |
LibreJS stays quiet and content. |
| 18:28 |
lissobone |
My inner Stallman remains satisfied. |
| 18:28 |
mrkubax10 |
lol |
| 18:30 |
lissobone |
Is your Stallman satisfied? |
| 18:32 |
lissobone |
Mine is not, as there's always nonfree software to fight with. |
| 18:41 |
mrkubax10 |
well nvidia drivers unfortunately |
| 18:41 |
mrkubax10 |
:/ |
| 18:41 |
mrkubax10 |
nouveau is unusable for me |
| 18:48 |
|
orwell96 joined #minetest |
| 18:56 |
|
kamdard_ joined #minetest |
| 18:56 |
celeron55 |
it seems to respond roughly as it did after the recent fix |
| 18:58 |
|
Talkless joined #minetest |
| 18:58 |
celeron55 |
has the problem already went away? |
| 18:59 |
celeron55 |
i can see large amounts of sql queries have been occurring for maybe 40 minutes nw |
| 18:59 |
celeron55 |
now* |
| 19:00 |
|
orwell96_ joined #minetest |
| 19:03 |
MTDiscord |
<Bastrabun> It feels a tiny bit faster than when I first noticed. It's not "fast" though. Is there anything we can do to help you get it up to speed? |
| 19:03 |
celeron55 |
seems like the cause is two simultaneous uncommon crawlers... one of which is doing what seems like really stupid things. lol, it's fetching a wiki main page over and over again in some sort of redirect loop |
| 19:04 |
celeron55 |
i'll tweak some things, this is a good test for some configuration improvements |
| 19:05 |
celeron55 |
the other one is crawling the wiki probably as fast as it can |
| 19:07 |
celeron55 |
new nginx rate limit configs going in |
| 19:12 |
|
sys4 joined #minetest |
| 19:19 |
celeron55 |
i probably need to upgrade the server to something faster. not because of the amount of legit users, but because of crawlers |
| 19:21 |
MTDiscord |
<Bastrabun> Locking them out is no option, I guess? Or telling them via robots.txt how to behave and throwing those who don't behave out? |
| 19:21 |
celeron55 |
it can currently handle about 2x page loads (meaning pages that result in mysql requests) compared to normal traffic |
| 19:21 |
celeron55 |
that needs manual work every time |
| 19:22 |
celeron55 |
it's not viable. would be an absolute waste of my time certainly |
| 19:22 |
MTDiscord |
<Bastrabun> We can't be the only ones who want to do this. How to others do that? |
| 19:22 |
|
Thelie joined #minetest |
| 19:22 |
celeron55 |
well, 2x is too little margin, that's certain |
| 19:23 |
celeron55 |
i'd feel more comfortable with 5x |
| 19:25 |
celeron55 |
cloudflare would probably help with some things, but it depends on the exact behavior of the crawler. i tend to not think it would generally solve the issue |
| 19:25 |
celeron55 |
robots.txt... need to take a look, not sure what's there at the moment |
| 19:25 |
celeron55 |
pretty sure these ones wouldn't follow it even if it was there |
| 19:26 |
rubenwardy |
ContentDB handled 5x page loads increase with no issue sduring the mexican spike, interested to see how it scales beyond that |
| 19:26 |
rubenwardy |
mexican wave |
| 19:26 |
ROllerozxa |
just block crawlers if they don't respect crawl-delay and aren't important search indexers (google, bing, yandex, DDG, whatever...) |
| 19:26 |
|
___nick___ joined #minetest |
| 19:26 |
celeron55 |
https://forum.minetest.net/robots.txt says Crawl-delay: 10, which seems good |
| 19:26 |
celeron55 |
the wikis are missing robots.txt. need to add it |
| 19:26 |
lissobone |
Crawlers... |
| 19:27 |
celeron55 |
ROllerozxa: well if there was a magic button i could press that would do it, of course i would press it |
| 19:27 |
celeron55 |
or i could throw money at the problem (in a way that doesn't incur recurring costs that cannot be backed out of) |
| 19:27 |
celeron55 |
but that's not how it works |
| 19:29 |
|
___nick___ joined #minetest |
| 19:29 |
rubenwardy |
what's the specs of the machine you use? |
| 19:31 |
Krock |
https://i.redd.it/arkj47j7nrn51.jpg |
| 19:32 |
rubenwardy |
lmao |
| 19:33 |
rubenwardy |
I was thinking more https://global.discourse-cdn.com/eveonline/original/3X/6/3/63db795497646f2b913665554a383ae50621fed7.jpeg |
| 19:33 |
ROllerozxa |
you can return 444 at the nginx web server level on an UA fragment that matches that of an aggressive crawler |
| 19:33 |
ROllerozxa |
e.g: `if ($http_user_agent ~* (AhrefsBot|SomeOtherAggressiveBot)) { return 444; }` |
| 19:34 |
ROllerozxa |
(444 is an internal nginx error code that just drops the connection without a response) |
| 19:35 |
celeron55 |
it's a (believe it or not) a dedicated server, hosted in france, with an intel atom C2338, 4 GB of RAM and a SanDisk X400 2.5 7MM 128GB SSD |
| 19:37 |
Krock |
surprisingly close to what I posted |
| 19:37 |
rubenwardy |
I used to use a dedicated server (Kimsufi) but at the cheap end you don't get much for your money. I now use a virtual dedicated (VPS with dedicated CPU) - ContentDB uses a netcup RS 2000 G9.5, 6-core 3.35GHz, 16GB RAM for about £16pm |
| 19:37 |
celeron55 |
it was the cheapest way for me to get away from VPSes. I hate VPSes when I need reliable processing power and especially reliable I/O performance. this one does it. not massively fast, but nobody can throttle it in any way. very predictable |
| 19:38 |
DeepT |
celeron55, i just want to say i blame you entirely for the hours I have wasted playing minetest |
| 19:38 |
DeepT |
thanks for a great game :D |
| 19:38 |
celeron55 |
on any VPS I've always run into issues with someone else on the host ruining my hosting experience. on this, it's always been either my fault, or a client doing something with my services |
| 19:40 |
celeron55 |
i won't accept any hosting offer unless i have a dedicated CPU, dedicated RAM and a dedicated disk |
| 19:41 |
celeron55 |
it's the only way to actually know how much of each you have |
| 19:41 |
celeron55 |
and the only way to make sure you actually have it |
| 19:41 |
Krock |
fixed allocations are also possible in regular VMs/VPS' |
| 19:42 |
rubenwardy |
yet the forums are currently being throttled |
| 19:42 |
celeron55 |
that's because of reasons i can measure and understand, and not because some bullshit i/o limits done by the host without any transparency |
| 19:43 |
celeron55 |
or because someone else messing around on the host ruining my day |
| 19:43 |
celeron55 |
anyway. i'm up for upgrading this, as long as it's a dedicated machine for cheap |
| 19:46 |
celeron55 |
it needs to be cost effective, because i don't want to beg for money. but i neither will pay it out of my pocket |
| 19:46 |
celeron55 |
and i don't want to have to downgrade it due to lack of funds, that would be the worst |
| 19:48 |
DeepT |
im asking my canadian ISP administrating friend if he has something beefier for cheap |
| 19:49 |
celeron55 |
this box costs 12.64 €/month + 24% VAT |
| 19:49 |
DeepT |
good to know, that way i can define "cheap" |
| 19:49 |
celeron55 |
it needs to be a well known big host. i don't want to switch machines because the host is going bankrupt or whatever either |
| 19:50 |
DeepT |
ah nvm then |
| 19:50 |
rubenwardy |
your requirements are basically guaranteeing poor performance |
| 19:50 |
celeron55 |
and it needs to be in europe. i won't transfer user data out of europe |
| 19:51 |
rubenwardy |
Well, depends how well known. The big name hosts are all about cloud and cost more. Much better value with less known hosts - like netcup or contabo |
| 19:51 |
rubenwardy |
the 1000€ is a pretty sizable buffer |
| 19:52 |
celeron55 |
i've been half seriously considering hosting it myself too. i have 100M fiber, solar + battery backup for power, and i could hide behind cloudflare |
| 19:53 |
celeron55 |
and could upgrade the fiber to multiple times faster given the upgrade would be covered as hosting expenses |
| 19:53 |
celeron55 |
would need to get a machine for it though. don't have anything suitable at hand |
| 19:54 |
|
mrkubax10 left #minetest |
| 19:55 |
ROllerozxa |
looking at netcup, there's the RS 1000 G9.5 right below the server contentdb is on for 11.76€/mo... root server means dedicated right? |
| 19:55 |
rubenwardy |
yes |
| 19:55 |
rubenwardy |
virtual dedicated* |
| 19:55 |
ROllerozxa |
8GB RAM, AMD EPYC 7702 with 4 dedicated cores, 160GB SSD... should be enough I assume |
| 19:55 |
celeron55 |
what you guys definitely are right about is that this server is at its limits and it needs upgrading one way or another |
| 19:57 |
rubenwardy |
there's also https://www.tuxfamily.org/ |
| 19:57 |
rubenwardy |
hosts Godot infrastructure |
| 19:59 |
celeron55 |
bandwidth is what i don't need a lot of. it's averaging at around 700 kbit/s and peaks at just under 2 Mbit/s. that's, what, 230 GB/month? |
| 19:59 |
celeron55 |
of course more peak bandwidth is always useful |
| 19:59 |
rubenwardy |
I don't think you'll have any trouble raising money for hosting though. Without any fundraising campaigns, my patreon etc got to $57. Just need to be transparent about what you're using it for. You |
| 20:00 |
rubenwardy |
ContentDB currently uses about 2TB per month. Bandwidth is a limiting factor - netcup offers 120TB |
| 20:01 |
rubenwardy |
Ideally, I'd like to have a org + opencollective, but alas |
| 20:02 |
celeron55 |
well, let's make a deal |
| 20:02 |
celeron55 |
i'll get the netcup RS like contentdb |
| 20:02 |
celeron55 |
and then nobody complains for 5 years? |
| 20:03 |
celeron55 |
i don't trust it, but if the users do, then that's fine to me |
| 20:04 |
celeron55 |
based on the specs it should be about 10x faster compared to the current machine. assuming the virtualization doesn't ruin it |
| 20:05 |
celeron55 |
the cpu that is. the disk is anyone's guess as nothing is specified |
| 20:05 |
rubenwardy |
d if=/dev/zero of=./testfile bs=1G count=10 oflag=direct |
| 20:05 |
rubenwardy |
10737418240 bytes (11 GB, 10 GiB) copied, 27.9403 s, 384 MB/s |
| 20:05 |
rubenwardy |
not the best not the worst |
| 20:05 |
rubenwardy |
although maybe not the highest demand right now |
| 20:06 |
rubenwardy |
ContentDB is still slower than I'd like, but I blame that on my shitty SQL/ORM usage |
| 20:07 |
celeron55 |
i can guarantee shitty SQL/ORM usage for the forum and wikis |
| 20:07 |
celeron55 |
i get that for free |
| 20:08 |
|
Fixer joined #minetest |
| 20:08 |
celeron55 |
looks like the disk is about 10x faster too. well, that varies, possibly a lot, but looks like there's some room for that |
| 20:13 |
ROllerozxa |
hooray, new server time |
| 20:13 |
ROllerozxa |
if it's enough for contentdb, then surely it can handle the forums and wiki |
| 20:15 |
celeron55 |
you're putting too much faith on phpbb. but phpbb's legacy tech has a good side: however bad i talk about it it's fairly lightweight, because it was designed at a time when computing resources were way more expensive than today |
| 20:16 |
rubenwardy |
I would get the next server up for CDB, but the price jump is pretty huge |
| 20:17 |
celeron55 |
well technically the best solution would be to get the next server up, and host contentdb and the forum and the wikis there, but that's probably putting too many eggs in the same basket |
| 20:17 |
celeron55 |
i'm wondering whether just getting the machine from the same host is a bad idea |
| 20:18 |
celeron55 |
it's nice to have only one service go down when an entire host goes down |
| 20:19 |
|
Fixer_ joined #minetest |
| 20:19 |
|
kabou joined #minetest |
| 20:20 |
rubenwardy |
it's kind of funny how different that is to the cloud approach, where if one host goes down the service might just be slower |
| 20:21 |
rubenwardy |
not that I'd want the cloud approach - the cloud only makes sense if the cost of going offline is more than the cost of being in the cloud |
| 20:22 |
celeron55 |
yeah it's fun but too expensive unless every client is making you money |
| 20:22 |
MTDiscord |
<luatic> celeron55: how is the server so slow when so little bandwidth is being used? |
| 20:22 |
MTDiscord |
<luatic> what is happening under the hood? |
| 20:22 |
MTDiscord |
<luatic> have you tried some profiling? upgrading the phpBB, PHP, MySQL? |
| 20:24 |
MTDiscord |
<luatic> wild guess: MySQL is probably doing full table scans whenever someone is searching something |
| 20:25 |
MTDiscord |
<luatic> ContentDB uses an index (IIRC it's a Postgres extension or something there) to speed searches up |
| 20:25 |
celeron55 |
the hardware can only handle a certain rate of mysql (mariadb) queries. if you didn't look it up, the CPU is about half as fast as a raspberry pi 4 B |
| 20:26 |
MTDiscord |
<luatic> oof |
| 20:26 |
MTDiscord |
<luatic> still I diagnose the stack with algorithmic problems ;) |
| 20:27 |
celeron55 |
you're welcome to diagnose it, but i won't give you the data :-) and i don't have time for diagnosing it. it's just phpbb and mediawiki, with mostly default settings and the recommended setup |
| 20:27 |
celeron55 |
it seems to me they're performing as expected on the hardware |
| 20:28 |
MTDiscord |
<luatic> MySQL seems to have support for fulltext indexes: https://dev.mysql.com/doc/refman/8.0/en/fulltext-search.html |
| 20:28 |
MTDiscord |
<luatic> Not sure if phpBB uses this feature though |
| 20:28 |
celeron55 |
it doesn't matter what mysql has. i'm not making the hosted software myself |
| 20:29 |
celeron55 |
phpbb has its own search speed-up system as far as i know |
| 20:29 |
celeron55 |
i know that, because sometimes it ends up doing silly heavy operations when you're doing moderator or admin stuff |
| 20:30 |
celeron55 |
related to the search index |
| 20:31 |
MTDiscord |
<luatic> phpBB does seem to have code related to using FULLTEXT indexes of MySQL |
| 20:32 |
celeron55 |
could be it was an older version and they've since updated it, or whatever. in any case, it kind of doesn't matter |
| 20:32 |
MTDiscord |
<ROllerozxa> hopefully the minetest forums actually uses "fulltext mysql" instead of the "fulltext native" indexer |
| 20:33 |
celeron55 |
ah so that's what the phpbb options are called |
| 20:35 |
celeron55 |
lol, looks like phpbb native is selected |
| 20:39 |
MTDiscord |
<luatic> that thing frankly is a mess |
| 20:39 |
celeron55 |
looks like from scaleway (the current provider) a 5x faster machine would cost about 27€/month including VAT |
| 20:39 |
celeron55 |
that's a bit steep |
| 20:40 |
celeron55 |
about 10x faster would be about 37€ |
| 20:40 |
rubenwardy |
that's about what I was paying on linode |
| 20:41 |
celeron55 |
that would have very solid performance but it's also about 3x more than a virtual dedicated, or similar |
| 20:41 |
rubenwardy |
also, 1000€ would last 3 years at 27€ a month |
| 20:41 |
rubenwardy |
if you want to think of it like that |
| 20:42 |
celeron55 |
that's exactly how i think of it. it's my "no worries" buffer |
| 20:43 |
celeron55 |
well, we could even do a vote somewhere |
| 20:43 |
celeron55 |
it's probably a waste of money from the donor's standpoint |
| 20:43 |
celeron55 |
i wouldn't like it if my money was spent on that |
| 20:47 |
celeron55 |
i definitely feel like i have the obligation to spend the money in a way that ensures the websites have a high chance of being available a good while. that means hosting must be painless and reliable, but also cost effective |
| 20:55 |
celeron55 |
i wonder what's involved in changing phpbb's search backend other than choosing a new one from the dropdown menu... |
| 20:57 |
celeron55 |
i'll yolo it |
| 21:00 |
celeron55 |
it creates the index now. it might be barely usable while it's doing it. it's going to index about half a million words and about 8 million word-to-post relations, according to the phpbb native cache statistics |
| 21:17 |
fluxionary |
on a test server running 5.7.0-dev w/ this commit https://github.com/minetest/minetest/commit/b8aaad4f1ef2e4b22cb7d47de9cb54068ccbbe18, we're suddenly seeing crashes due to the fact that `minetest.get_objects_inside_radius` is suddenly returning objects where `obj:is_player() == false and obj:get_luaentity() == nil`. while it's clear that we need to fix the lua code which assumes always one or the other, i'm really curious as to what might have |
| 21:17 |
fluxionary |
changed since 5.6.1 which would result in such objects? |
| 21:48 |
DeepT |
i never understand why people will join a server, not even move or do anything then quit |
| 21:52 |
MTDiscord |
<Jonathon> fyi @x2048 |
| 21:54 |
|
FavoritoHJS joined #minetest |
| 22:01 |
|
proller joined #minetest |
| 22:02 |
|
Fixer joined #minetest |
| 22:07 |
|
Leopold joined #minetest |
| 22:10 |
fluxionary |
DeepT: possibly because they've got low power machines and nothing loads |
| 22:26 |
|
proller joined #minetest |
| 22:26 |
sfan5 |
fluxionary: can you bisect? |
| 22:27 |
fluxionary |
sfan5, i don't have a way of causing it, though i'm working on that |
| 22:27 |
fluxionary |
it's caused 2 crashes, and i've got no further details on what the relevant object actually was |
| 22:28 |
|
Flabb joined #minetest |
| 22:57 |
DeepT |
mobs_sky for mobs_redo, anyone know how to use gull eggs? |
| 23:25 |
|
kabou42 joined #minetest |
