| Time |
Nick |
Message |
| 00:09 |
hmmmm |
what the f.... |
| 00:09 |
hmmmm |
RBA quit because people didn't want music directly in the core |
| 00:10 |
hmmmm |
and he blames jthread for making his code slow |
| 00:10 |
hmmmm |
is this serious |
| 00:10 |
RealBadAngel |
hmmmm, you may merge this later on |
| 00:10 |
RealBadAngel |
and for the music i waited more than a year |
| 00:11 |
hmmmm |
main menu music - sure. i have no problem with that, i think it's a neat feature |
| 00:11 |
hmmmm |
but that's a per-game thing |
| 00:11 |
RealBadAngel |
there were also a code for that |
| 00:11 |
hmmmm |
i think what you want is to develop a *game* with less lua and more hard coded effects |
| 00:11 |
RealBadAngel |
even c55 found it |
| 00:11 |
hmmmm |
yes i saw it |
| 00:11 |
RealBadAngel |
it was never merged |
| 00:12 |
hmmmm |
probably because it says WIP on it |
| 00:12 |
RealBadAngel |
lol |
| 00:12 |
hmmmm |
if you're telling us it's still a work in progress... |
| 00:12 |
hmmmm |
then why would we merge that |
| 00:12 |
RealBadAngel |
i was begging for it many times |
| 00:12 |
RealBadAngel |
like i should lol |
| 00:12 |
hmmmm |
also i read that it breaks something |
| 00:12 |
RealBadAngel |
lemme find a link |
| 00:13 |
hmmmm |
why not just fix that, make it into a finished product, then merge it |
| 00:13 |
hmmmm |
then merge the main menu music into minetest_game |
| 00:13 |
hmmmm |
not the core... |
| 00:13 |
RealBadAngel |
https://gitlab.unix-experience.fr/epixel/epixel |
| 00:14 |
RealBadAngel |
point of the score was the engine having own |
| 00:14 |
hmmmm |
RealBadAngel, honestly, I think if you are looking to go into the direction of a monolithic, non-modular game with things hard coded, freeminer or minetest-classic is more your thing |
| 00:14 |
RealBadAngel |
and each game overrriding it at will |
| 00:14 |
* Taoki |
believes that a patch to allow per-game main menu music should be merged... urgently. Also that 1.5 MB is NOT a big file size... at all. Also that RBA should stay and not leave the project :) |
| 00:14 |
Taoki |
Head still hurts after this morning heh |
| 00:15 |
RealBadAngel |
i asked legendary comoser of PC demoscene to use his tunes |
| 00:15 |
RealBadAngel |
he said ok |
| 00:15 |
hmmmm |
yes, I read all of what you said last night |
| 00:15 |
hmmmm |
and I can't help but repeat the same counter arguments |
| 00:15 |
RealBadAngel |
but you just refused it |
| 00:15 |
hmmmm |
because they are indeed valid, imo |
| 00:15 |
RealBadAngel |
this case |
| 00:15 |
hmmmm |
i didn't even need to tell you no |
| 00:15 |
hmmmm |
other people did |
| 00:15 |
RealBadAngel |
i felt offended too |
| 00:16 |
RealBadAngel |
because im a demoscene coder too |
| 00:16 |
hmmmm |
feelings are a personal matter... this is minetest |
| 00:16 |
Taoki |
I think we all need better communication to avoid such situations, for one thing. |
| 00:16 |
RealBadAngel |
you have spit into my face |
| 00:16 |
hmmmm |
the communication is perfect |
| 00:16 |
hmmmm |
RBA just wants and gets mad if he does not get |
| 00:16 |
RealBadAngel |
thats why im leaving |
| 00:16 |
hmmmm |
RealBadAngel |
| 00:16 |
RealBadAngel |
i was patient for a year |
| 00:17 |
hmmmm |
I don't have qualms about you leaving, but I would suggest you contribute to minetest classic instead |
| 00:17 |
VanessaE |
hmmmm: there is one of those arguments from last night that is NOT valid: that 1.5 MB worth of extra content is somehow excessive. This is 2015, most people have good enough bandwidth to handle that (it's what, 10 seconds for the slowest of slow DSL's?) |
| 00:17 |
Taoki |
hmmmm: Not really. Devs argue sometimes, the atmosphere here hasn't always been that great. I can now see how that risks ruining the project really. |
| 00:17 |
hmmmm |
it's probably a better fit of a project for your own style |
| 00:17 |
VanessaE |
and it's not as though *clients* have to download that extra data from a server. they'll already have it as part of their base install. |
| 00:17 |
RealBadAngel |
wait a second |
| 00:17 |
RealBadAngel |
whos AV maintainer? |
| 00:18 |
RealBadAngel |
theres just one to blame for |
| 00:18 |
hmmmm |
VanessaE: okay, point well made, but explain why this cannot be part of minetest_game instead |
| 00:19 |
VanessaE |
hmmmm: because the idea here is for between-games music, i.e. when nothing at all is selected at the main menu. |
| 00:19 |
RealBadAngel |
because each game can carry own tunes |
| 00:19 |
VanessaE |
or when waiting to connect to a server for example. |
| 00:19 |
RealBadAngel |
im the only dumb person here? |
| 00:19 |
hmmmm |
so why can't there be a mechanism to fall back onto another game's bundled music |
| 00:19 |
RealBadAngel |
rotfl |
| 00:19 |
hmmmm |
plenty of things rely on minetest_game already |
| 00:20 |
VanessaE |
hmmmm: I was going to suggest exactly that, actually. |
| 00:20 |
hmmmm |
right |
| 00:20 |
VanessaE |
however, |
| 00:20 |
VanessaE |
some kind of minimal tune for the main menu should still be provided. |
| 00:20 |
RealBadAngel |
anwyay, hmmmm youre free to merge it |
| 00:20 |
VanessaE |
(I mean so that the "fallback" doesn't become "the main" tune) |
| 00:20 |
hmmmm |
eh |
| 00:20 |
RealBadAngel |
you do have my aprroval and Peter Hajba, the composer |
| 00:21 |
hmmmm |
I'm not totally convinced on the need for music to begin with |
| 00:21 |
RealBadAngel |
sure, turn on tv |
| 00:21 |
hmmmm |
if this feature were added, I personally wouldn't ever use it |
| 00:21 |
RealBadAngel |
are you deaf? |
| 00:21 |
VanessaE |
hmmmm: the need is there as a player perception issue. The vast majority of games have at least some kind of basic music running in their main menu. |
| 00:22 |
Taoki |
hmmmm: Some default main menu music for minetest_game might be okay. Personally I'm interested for support for other games to do it, but really MTG could too. |
| 00:22 |
VanessaE |
and to the average user, Minetest is a game, not an engine. |
| 00:22 |
hmmmm |
so why can't we have some kind of "tap" sound when a user clicks on a button |
| 00:22 |
hmmmm |
so they know it's working |
| 00:22 |
Taoki |
Especially becaiuse it's local. The music file doesn't need to be networked, which causes most of the slowness. |
| 00:22 |
RealBadAngel |
its a video game production |
| 00:22 |
Taoki |
That would be good too! |
| 00:22 |
RealBadAngel |
aviable on just a one click |
| 00:22 |
VanessaE |
hmmmm: because that only gives the perception of feedback, not "first impression: this is a good game". |
| 00:22 |
RealBadAngel |
tap even |
| 00:23 |
RealBadAngel |
its fucking 21st century |
| 00:23 |
VanessaE |
that's really the point main-menu-no-adventure-selected-yet music |
| 00:23 |
RealBadAngel |
and its deaf production |
| 00:23 |
hmmmm |
VanessaE: so how about openarena? |
| 00:23 |
VanessaE |
give the player a good first impression of the game they've just launched. |
| 00:23 |
VanessaE |
hmmmm: touché |
| 00:23 |
RealBadAngel |
you guys are just nuts to not see and understand it |
| 00:23 |
VanessaE |
nevertheless, even OA has music once you're in a game |
| 00:23 |
VanessaE |
ominous though it may be |
| 00:24 |
hmmmm |
yeah and servers can already do that today |
| 00:24 |
VanessaE |
yes, but if a server does it, it has to send all that data to every client. |
| 00:24 |
RealBadAngel |
anyway, thats not my problem anymore |
| 00:24 |
VanessaE |
which will drive server bandwidth counts up and make things slower for players. |
| 00:24 |
hmmmm |
see |
| 00:24 |
RealBadAngel |
jesus |
| 00:24 |
RealBadAngel |
nuts |
| 00:24 |
hmmmm |
I have nothing against music |
| 00:24 |
hmmmm |
but don't make it a basic part of the core |
| 00:24 |
hmmmm |
that's such a per-game thing |
| 00:24 |
* Taoki |
worries for who will manage the minimap and shaders code for Minetest now :( |
| 00:25 |
RealBadAngel |
just PUT THE FUCKIN FILE IN SOUNDS FOLDER |
| 00:25 |
VanessaE |
bandwidth for ME is a non-issue, I have like 10TB a month allowed on my servers, but imagine the average server owner who runs the game from, let's say, a cheap VPS that only gives them 50GB a month or something |
| 00:25 |
hmmmm |
honestly I am rpobably going to end up doing that, taoki |
| 00:25 |
Taoki |
ok |
| 00:25 |
hmmmm |
as a part of client side modding |
| 00:25 |
RealBadAngel |
server doesnt have to sent it |
| 00:25 |
hmmmm |
anyway |
| 00:26 |
hmmmm |
for a discussion such as this one with so many wildly varying opinions, i can't help but think that a poll isn't more appropriate |
| 00:26 |
VanessaE |
hmmmm: what RBA is trying to say with that outburst is that the feature is already supported in the core and has been for a long time now. All that is necessary to enable it, literally, is to supply the necessary ogg file. |
| 00:26 |
RealBadAngel |
if sound file is found locally its just used |
| 00:26 |
hmmmm |
VanessaE: celeron said that PR is broken somehow |
| 00:26 |
VanessaE |
define "somehow"? |
| 00:26 |
RealBadAngel |
R O T F L |
| 00:26 |
hmmmm |
i dunno i'm reading the logs |
| 00:26 |
VanessaE |
I think c55 is referring to another piece of code/PR? |
| 00:27 |
hmmmm |
he's referring to [WIP] Main menu music |
| 00:27 |
RealBadAngel |
i just provided media file |
| 00:27 |
hmmmm |
1651 |
| 00:27 |
RealBadAngel |
thats about EXTRA functionality |
| 00:27 |
hmmmm |
'altough it breaks some functionality in guiEngine.cpp which should be fixed before that is merged, if it will be' |
| 00:27 |
RealBadAngel |
and games able to have own tunes |
| 00:28 |
VanessaE |
so, here's my thinking: supply a little 60-120s main menu media file, let it play, and when the user clicks on a game, either cross-fade into whatever is supplied with the clicked-on game=, or just keep the tune playing if there is nothing at all. |
| 00:28 |
RealBadAngel |
jesus, this code is older than some of the dinosaurs |
| 00:28 |
hmmmm |
make a forum poll on it |
| 00:28 |
RealBadAngel |
i wont |
| 00:28 |
RealBadAngel |
im not mt dev anymore |
| 00:28 |
hmmmm |
look i get it |
| 00:29 |
hmmmm |
for somebody who's 42 you're acting awfully immature about this whole thing |
| 00:29 |
Taoki |
hmmmm: Sadly, I must agree there. Hope RBA won't hate me for saying it. |
| 00:29 |
RealBadAngel |
waiting a yr for a reaction is pretty long time |
| 00:30 |
VanessaE |
RBA may be overreacting a bit, but a year indeed is WAY too long to go for something this simple (relatively speaking) |
| 00:30 |
Taoki |
I'd rather not lie: The way RBA reacted is mostly what upset me today. Even if I think I actually agree with almost all his points. |
| 00:30 |
RealBadAngel |
i had a vision of what i was doing |
| 00:30 |
hmmmm |
maybe the reason why it wasn't merged for so long is because it's controversial |
| 00:30 |
RealBadAngel |
you have all pissed on me |
| 00:31 |
Taoki |
VanessaE: I think he should apologize at least for saying that everyone in the Minetest credits list is a lunatic. I'm also on there BTW... of course I don't mind that implying me, but it's just an idea and principle for me. |
| 00:31 |
hmmmm |
if it were brought up as a topic of discussion to me, personally, what i would do is add it (provided it didn't force music as part of the core) |
| 00:31 |
RealBadAngel |
controversial is video game without a music |
| 00:31 |
VanessaE |
Taoki: he's quoting the main website :) |
| 00:31 |
hmmmm |
since it seems to be an enhancement |
| 00:31 |
Taoki |
I uderstand RBA felt underconsidered too, however. |
| 00:31 |
Taoki |
Which is one reason why I said communication might have not been the best. |
| 00:31 |
hmmmm |
there's a clear difference between an added feature/enhancement and a stylistic choice |
| 00:31 |
hmmmm |
minetest_game is the place where those stylistic choices go into |
| 00:32 |
VanessaE |
hmmmm: the problem is that minetest_game is not always available |
| 00:32 |
RealBadAngel |
minetest game can have own music |
| 00:32 |
VanessaE |
(though in theory, it *should* be) |
| 00:32 |
RealBadAngel |
as each game |
| 00:32 |
RealBadAngel |
just update the pr. its simple lua |
| 00:33 |
Taoki |
Anyway, if there's a vote, I say include that music file. I don't see it doing any harm. |
| 00:33 |
hmmmm |
i'd rather take a closer look at what celeron is talking about before merging something that breaks other functionality |
| 00:33 |
RealBadAngel |
lol |
| 00:33 |
RealBadAngel |
he found another way to refuse |
| 00:33 |
hmmmm |
and not during a feature freeze either |
| 00:33 |
RealBadAngel |
im fed up with his shit |
| 00:34 |
VanessaE |
I don't advocate merging it during freeze. |
| 00:34 |
hmmmm |
so just to be clear: |
| 00:34 |
hmmmm |
i approve of pr 1651 if... |
| 00:34 |
hmmmm |
- the music file is removed |
| 00:35 |
hmmmm |
- the functionality is verified to be working |
| 00:35 |
hmmmm |
the critical mass of users or developers or whoever might like the music being right in the core |
| 00:35 |
hmmmm |
that's up to them |
| 00:35 |
hmmmm |
i'm not going to lie and suddenly say that i want it all of a sudden when i don't |
| 00:36 |
hmmmm |
it's the community's choice as a whole, though |
| 00:36 |
VanessaE |
hmmmm: so you're basically saying "let's make the code work first and decide on the style of its output, if any, later". |
| 00:36 |
hmmmm |
yes basically |
| 00:36 |
RealBadAngel |
im not going thiis way |
| 00:36 |
hmmmm |
RealBadAngel: well it doesn't matter, does it? because you're not a dev any longer |
| 00:37 |
RealBadAngel |
ofc |
| 00:37 |
RealBadAngel |
for the record |
| 00:37 |
VanessaE |
I can't say I disagree with that, with the proviso that if there's a clear consensus on a main menu tune or at least a style thereof, it be added in a separate commit later. |
| 00:38 |
RealBadAngel |
at august 18th, Peter Hajba member of Future Crew, alllowed minetest to use his tunes |
| 00:39 |
hmmmm |
yes we get it we get it |
| 00:39 |
RealBadAngel |
year was 2014 |
| 00:39 |
hmmmm |
can we move onto another topic already? |
| 00:39 |
RealBadAngel |
now i feel offended |
| 00:40 |
RealBadAngel |
and he will propably feel the same |
| 00:40 |
RealBadAngel |
if some1 will let him know |
| 00:41 |
RealBadAngel |
no, we cant move onto another topic |
| 00:41 |
RealBadAngel |
this one is already too old |
| 00:43 |
hmmmm |
it's not productive to stay on this topic. we already defined our action items. |
| 00:43 |
RealBadAngel |
yes |
| 00:43 |
RealBadAngel |
i left and said goodbye |
| 00:44 |
|
OldCoder joined #minetest-dev |
| 00:45 |
Taoki |
:/ |
| 00:45 |
RealBadAngel |
and minetest is not allowed to use "Mercury Rain" anymore |
| 00:45 |
RealBadAngel |
i get that approval and i am deciding which fork will get it |
| 00:46 |
RealBadAngel |
if you have pissed on a legend dont be suprised if some1 will piss on you |
| 00:51 |
RealBadAngel |
demoscene was born propably before you hmmm |
| 00:52 |
RealBadAngel |
im with all those guys, including Skaven |
| 00:52 |
RealBadAngel |
so, begone |
| 00:54 |
VanessaE |
O_o |
| 00:56 |
* Taoki |
sighs |
| 00:59 |
|
dzho joined #minetest-dev |
| 01:05 |
hmmmm |
lol |
| 01:05 |
hmmmm |
he makes it sound like a demoscene group is the mafia |
| 01:07 |
|
Hijiri joined #minetest-dev |
| 01:08 |
Taoki |
This is being taken way, way too personally... |
| 01:09 |
RealBadAngel |
mafia? |
| 01:09 |
RealBadAngel |
hehhe, maybe |
| 01:10 |
RealBadAngel |
folks are knowing each others |
| 01:10 |
RealBadAngel |
going on meetings |
| 01:11 |
RealBadAngel |
so you know the other coders for like 20yrs or more |
| 01:12 |
RealBadAngel |
you can ask for everything and excuse everting |
| 01:21 |
hmmmm |
well, we already explained our position. basically RBA wants us to unilaterally agree that music will be add it, and add it, or else he's going to hang around here and taunt us until we do |
| 01:21 |
hmmmm |
i already explained the route that must be taken in order to see if it will be added or not |
| 01:21 |
hmmmm |
there's nothing more to say about this really |
| 01:22 |
hmmmm |
s/will be add it/will be added/ whoops |
| 01:23 |
RealBadAngel |
hmmmm, you dont get it |
| 01:23 |
RealBadAngel |
im not "here" anymore to taunt you |
| 01:24 |
RealBadAngel |
i am just kind to let u know whats going on |
| 01:25 |
RealBadAngel |
if there will be some problems, or patches needed minetest can get them |
| 01:25 |
RealBadAngel |
also i can be of help |
| 01:27 |
Taoki |
RealBadAngel: I'm mostly concerned about minimap and shader code. They're your work, and it might be harder for others to maintain them. If you can't develop patches for them in core Minetest any more, this might harm the project a lot. |
| 01:27 |
Taoki |
I think and hope that's not something you wish to do. |
| 01:28 |
RealBadAngel |
as i said, mt can use my patches |
| 01:29 |
Taoki |
Can you still create pull requests? I mean if both core Minetest and your fork are on Github, it should hopefully be easy to exchange wanted updates on each end. |
| 01:29 |
RealBadAngel |
theres a shitload of work for shaders upcoming |
| 01:29 |
RealBadAngel |
our fork is on git compatible site |
| 01:29 |
hmmmm |
our? |
| 01:30 |
RealBadAngel |
so you can exchange patches |
| 01:30 |
Taoki |
Ok. As long as they make it in core minetest as well, that's good. |
| 01:30 |
RealBadAngel |
hmmm just woke up ;) |
| 01:30 |
RealBadAngel |
yes, our |
| 01:30 |
Taoki |
Perhaps trying to do with your own fork might be a good idea, if you feel it gives you more freedom. |
| 01:30 |
Taoki |
Ultimately, it and Minetest will probably exchange most features and changes. |
| 01:31 |
hmmmm |
who else is on board with your fork, RBA? |
| 01:31 |
RealBadAngel |
c55? lol |
| 01:32 |
RealBadAngel |
just jokin |
| 01:35 |
RealBadAngel |
hmmmm, if they feel like in mood to let you know they propably will |
| 01:36 |
Taoki |
I just hope the fork has a well set direction and purpose. I... do question at this moment whether you want to fork as a way of taking revenge on c55 and other developers. |
| 01:37 |
Taoki |
Which of course would be your right all as much... just probably not as helpful. |
| 01:37 |
Taoki |
Realistically speaking, things would be a lot better if everyone could continue down one project... but I guess that matters little now. |
| 01:40 |
RealBadAngel |
i had my own vision |
| 01:41 |
Taoki |
ok |
| 01:41 |
Taoki |
Need to go now. Later, see everyone tomorrow |
| 01:41 |
RealBadAngel |
just got tired pissing on my head constantl y |
| 01:42 |
RealBadAngel |
cya |
| 01:42 |
RealBadAngel |
i also have lotsa PRs to push tomorrow |
| 01:43 |
RealBadAngel |
cya |
| 02:57 |
|
harrison joined #minetest-dev |
| 03:52 |
|
kaeza joined #minetest-dev |
| 04:16 |
|
Lunatrius joined #minetest-dev |
| 04:28 |
|
est31 joined #minetest-dev |
| 04:28 |
est31 |
how could ever a version of minetest be released with mmdb? |
| 04:36 |
kaeza |
I just noticed that the toggling the "double tap space to toggle fly" checkbox in the in-game key config dialog is ignored |
| 04:37 |
kaeza |
togglign "use = descend" works, and both also work from the main menu |
| 04:38 |
hmmmm |
blatantly ignored settings like that is a blocker for sure |
| 04:39 |
hmmmm |
anyway who takes care of the website |
| 04:39 |
hmmmm |
is it sfan or somebody else?? |
| 04:39 |
est31 |
I think celeron55 hosts that domain. |
| 04:39 |
hmmmm |
i don't think he does |
| 04:40 |
est31 |
kaeza, can you file a bug at github? |
| 04:40 |
hmmmm |
last i checked it was xyz and he transferred it to somebody else |
| 04:40 |
est31 |
hmmmm, https://forum.minetest.net/viewtopic.php?f=18&t=12611 |
| 04:40 |
kaeza |
est31, sure |
| 04:40 |
hmmmm |
ahh |
| 04:40 |
hmmmm |
okay this is good |
| 04:41 |
hmmmm |
erm est that post is from the year 2011 |
| 04:42 |
hmmmm |
if celeron did really host the forums, we'd be okay because he's been semi-active lately |
| 04:42 |
hmmmm |
whereas sfan, not so much |
| 04:42 |
hmmmm |
so the mmdb by definition relies on an outside source |
| 04:42 |
est31 |
Forum is now hosted by celeron55 (report issues here) |
| 04:42 |
est31 |
by celeron55 » Sat Jun 13, 2015 10:39 |
| 04:43 |
hmmmm |
we can't REMOVE mmdb, if anything maybe hide the tab if the server cannot be contacted |
| 04:43 |
hmmmm |
oh whoops.. |
| 04:43 |
hmmmm |
I read the "Joined" date |
| 04:47 |
est31 |
there is a difference between network problems, and total not-hosting of the page. |
| 04:47 |
est31 |
but yeah, perhaps better to hide the button "online mod repository" |
| 04:49 |
hmmmm |
or why not just have text that says |
| 04:49 |
hmmmm |
"Whoops! It seems like the MMDB is currently down. Sit tight!" or something to that effect |
| 04:49 |
hmmmm |
better than a friggin "AsyncWorkerError: HTTP request returned error code 404" |
| 04:50 |
* est31 |
remembers those vimeo "aww fiddlesticks" you got in the past if you didn't have flash, nor h264 html5 support |
| 04:50 |
est31 |
you still get them today |
| 04:50 |
est31 |
hmmmm, yes, if there is an error shown in the gui,, its ok |
| 04:51 |
hmmmm |
it's trendy to make a goofy 404 error page instead of the serious business "not found" |
| 04:51 |
est31 |
most people just dont see the debug.txt |
| 04:51 |
hmmmm |
on the web interface to the thing I work on for work we have a sad puppy for internal server errors |
| 04:51 |
hmmmm |
lol |
| 04:52 |
est31 |
zaroo bugs found :) |
| 04:52 |
hmmmm |
so just code that up really quick and then I have to fix up delete_area |
| 04:52 |
hmmmm |
then we should be okay to release |
| 04:53 |
* hmmmm |
crosses fingers |
| 04:53 |
hmmmm |
I think we should release like 1 day after we're done with all the blockers instead of waiting out the full 2 extended weeks |
| 04:57 |
est31 |
we should wait, for the case 3011 is indeed an engine fault |
| 04:58 |
hmmmm |
you know what, I don't think it is |
| 04:59 |
hmmmm |
I think what could be happening is that an errant mod goes crazy and accidentally deletes members from the global namespace |
| 04:59 |
est31 |
also, we have to wait for mtgame to make a release tag. |
| 05:00 |
hmmmm |
then some kind of runtime error happens soon after (of course) and then debug.traceback() has a runtime error when a) debug can't be accessed, b). traceback can't be accessed, or c). any one of the misc. functions it internally uses can't be accessed |
| 05:00 |
est31 |
I've learned that the hard way |
| 05:07 |
hmmmm |
btw why haven't you merged 3033 yet |
| 05:08 |
hmmmm |
well celeron left a comment saying that he'd rather you use g_config->getS32 instead of getFloat |
| 05:10 |
hmmmm |
oh by the way est, I looked at the PcgRandom comment a bit better and it is correct, but it demonstrates the math without the divmod optimization |
| 05:10 |
hmmmm |
so the comment and the implementation is not 1 to 1 |
| 05:11 |
est31 |
ok |
| 05:12 |
est31 |
getS32 isn't better than getFloat either, no? |
| 05:12 |
est31 |
its all below 2^52, so everything goes well precision wise |
| 05:13 |
hmmmm |
the max exponent on a float is 2^37 |
| 05:13 |
hmmmm |
37 i mean |
| 05:13 |
hmmmm |
it's just hacky to use a float where the thing you're using is actually an integer |
| 05:14 |
est31 |
modern cpus have fast floats |
| 05:15 |
hmmmm |
float to int conversion is very slow |
| 05:16 |
est31 |
well its not important at all |
| 05:16 |
est31 |
settings are slow already |
| 05:17 |
est31 |
i think it takes 10 microseconds to request a key from the settings map |
| 05:17 |
est31 |
but i might be weong |
| 05:17 |
est31 |
wrong* |
| 05:17 |
est31 |
is there no __CLASSNAME__ macro? |
| 05:18 |
est31 |
typeid(*this).name() |
| 05:18 |
est31 |
too long to be worth it |
| 05:19 |
hmmmm |
i don't think that's a static expression |
| 05:19 |
hmmmm |
it'd be evaluated on runtime, but even more, it'd force the compiler to use rtti |
| 05:19 |
est31 |
this is hardly static |
| 05:20 |
est31 |
I see |
| 05:31 |
sfan5 |
<hmmmm> whereas sfan, not so much |
| 05:31 |
sfan5 |
me not really doing any dev work (except reviewing PRs) is nothing new |
| 05:32 |
sfan5 |
also i have been "active" at making minetest builds and maintaining servers.minetest.net |
| 05:36 |
hmmmm |
oh |
| 05:36 |
hmmmm |
i mean i haven't seen you around lately |
| 05:37 |
hmmmm |
in any case, are you able to fix forums.minetest.net/mmdb? |
| 05:37 |
sfan5 |
no, c55 hosts that |
| 05:38 |
sfan5 |
also i don't really have that much of an idea how mmdb works |
| 05:40 |
|
Hunterz joined #minetest-dev |
| 05:51 |
est31 |
hmmmm, turns out the texture trick doesnt work |
| 05:52 |
hmmmm |
alright then |
| 05:57 |
est31 |
I'll prepare a PR then |
| 06:02 |
est31 |
hrrmm, our server count is much larger than our player count |
| 06:02 |
est31 |
which is good :) |
| 06:23 |
hmmmm |
it is? |
| 06:23 |
hmmmm |
anyway |
| 06:23 |
hmmmm |
est, you know what you have to do to add a hud flag? |
| 06:23 |
hmmmm |
i realize that minimap isn't officially part of the hud but it would be cleanest just to use that interface anyway |
| 06:24 |
hmmmm |
and set VolatileRunFlags::show_minimap based on the presense of that flag |
| 06:25 |
est31 |
hud flags are inflexible |
| 06:25 |
hmmmm |
so add a brand new packet why don't you |
| 06:25 |
est31 |
well yes one can add all the info as flags for the hug |
| 06:25 |
est31 |
hud* |
| 06:25 |
est31 |
every bit one flag |
| 06:25 |
est31 |
just then you'll need lots of bits |
| 06:26 |
est31 |
which means lots of flags |
| 06:26 |
hmmmm |
i thought this was just being turned on/off |
| 06:26 |
est31 |
thats the blocker part of the issue |
| 06:26 |
est31 |
but only because its a blocker doesnt allow us to hack a solution |
| 06:26 |
est31 |
instead of doing it properly |
| 06:27 |
hmmmm |
we're in a feature freeze |
| 06:27 |
est31 |
the problem with adding this as hud flag is that we'll have to support that hack for eternity |
| 06:28 |
est31 |
and if people ask why, we'll have to explain that back in 2015 we had a feature freeze |
| 06:28 |
hmmmm |
not necessarily |
| 06:28 |
hmmmm |
i think it could be implemented as a hud item later on |
| 06:28 |
est31 |
and rather did a quick hack than solving it properly |
| 06:28 |
hmmmm |
it's better than the alternative (making a new set of packets for a single feature) |
| 06:29 |
est31 |
the key press is instantly |
| 06:29 |
est31 |
it should not rely on lag |
| 06:29 |
hmmmm |
? |
| 06:29 |
est31 |
and if you allow mods to access the f9 key |
| 06:29 |
est31 |
then you open the door for even more problems |
| 06:29 |
hmmmm |
erm |
| 06:29 |
est31 |
they'll code sprint under that key |
| 06:29 |
est31 |
just because |
| 06:29 |
hmmmm |
I don't get what you mean at all |
| 06:29 |
est31 |
and it will be totally laggy |
| 06:30 |
hmmmm |
my idea was to add HUD_FLAG_MINIMAP_VISIBLE |
| 06:30 |
est31 |
hmmmm, you suggest the server turns minimap on/off? |
| 06:30 |
hmmmm |
yes isn't that what you wanted?? |
| 06:30 |
est31 |
the client sends whether there is a keypress |
| 06:30 |
est31 |
and the server decides |
| 06:30 |
est31 |
thats your solution, no? |
| 06:30 |
hmmmm |
erm no |
| 06:30 |
hmmmm |
that isn't my solution at all |
| 06:31 |
hmmmm |
i have no idea how you interpreted my words that way |
| 06:31 |
hmmmm |
here is my idea: |
| 06:31 |
hmmmm |
in l_object.cpp, add HUD_FLAG_MINIMAP_VISIBLE to es_HudElementType, add it to l_set_flags and l_get_flags |
| 06:32 |
kahrl |
<est13> [...] we'll have to support that hack for eternity |
| 06:32 |
kahrl |
couldn't we just mark it as // EXPERIMENTAL now and later deprecate it? |
| 06:32 |
hmmmm |
then in game.cpp just set flags.show_minimap = !!(localplayer->hud_flags & HUD_FLAG_MINIMAP_VISIBLE); |
| 06:33 |
hmmmm |
i think that's a 10 line change |
| 06:33 |
|
nrzkt joined #minetest-dev |
| 06:33 |
hmmmm |
oh of course you'd also have to add a check when F9 is pressed, if hud_flags & HUD_FLAG_MINIMAP_VISIBLE == 0, then the server must have disabled it |
| 06:34 |
hmmmm |
and so change the message to |
| 06:34 |
hmmmm |
"This server has disabled the minimap." |
| 06:34 |
hmmmm |
in the amount of time it took me to explain this I could've coded it by now |
| 06:34 |
hmmmm |
:| |
| 06:34 |
est31 |
we should make it default on, too, IMO. |
| 06:35 |
hmmmm |
it is default on |
| 06:35 |
est31 |
so HUD_FLAGS_MINIMAP_DISABLED |
| 06:35 |
hmmmm |
no |
| 06:35 |
hmmmm |
stop it |
| 06:35 |
hmmmm |
you make it default on by setting HUD_FLAGS_MINIMAP_VISIBLE in the Hud ctor |
| 06:35 |
hmmmm |
look at the way all the other HUD things are done already |
| 06:35 |
est31 |
but it only works on new servers |
| 06:35 |
hmmmm |
don't reinvent things |
| 06:35 |
est31 |
I dont reinvent anything |
| 06:35 |
est31 |
I just say |
| 06:35 |
est31 |
if its sent by the server |
| 06:36 |
est31 |
then we have no way to find out whether its default off |
| 06:36 |
est31 |
err |
| 06:36 |
est31 |
we have no way to find out whether the server is onld |
| 06:36 |
est31 |
old* |
| 06:36 |
est31 |
or has disabled the feature |
| 06:37 |
est31 |
and, no we dont raise protocol version just because of this |
| 06:37 |
kahrl |
in Client::handleCommand_HudSetFlags we do hud_flags &= ~mask and hud_flags |= flags |
| 06:37 |
kahrl |
so if the server doesn't send a flag it is unchanged |
| 06:39 |
est31 |
I see |
| 06:39 |
est31 |
ok then |
| 06:39 |
est31 |
still, it should be HUD_FLAGS_MINIMAP_ENABLED rather |
| 06:39 |
est31 |
not _VISIBLE |
| 06:39 |
kahrl |
yeah I like that better |
| 06:40 |
hmmmm |
this only controls the minimap visibility, though. |
| 06:40 |
hmmmm |
you can't control if the client had initialized with the mapper |
| 06:40 |
hmmmm |
besides, it's the same as all the other hud items |
| 06:40 |
est31 |
the visibility is controlled by the client |
| 06:40 |
hmmmm |
follow the pattern |
| 06:40 |
est31 |
the other hud items are always shown |
| 06:41 |
est31 |
except if you turn off the whole hud |
| 06:41 |
hmmmm |
yeah |
| 06:41 |
est31 |
but you can't say "only show me the inventory bar" on the client |
| 06:41 |
hmmmm |
this is something that can be totally fixed in the next version when we would like to expand upon it |
| 06:42 |
hmmmm |
as it stands right now, the minimap not being visible is a specific MinimapMode |
| 06:42 |
hmmmm |
whereas HUD_FLAGS_MINIMAP_VISIBLE would control game.cpp's flags.draw_minimap |
| 06:43 |
hmmmm |
so in order to see the minimap the user has to: |
| 06:43 |
hmmmm |
- have the MinimapMode in a mode other than MINIMAP_MODE_OFF |
| 06:43 |
hmmmm |
- have HUD_FLAGS_MINIMAP_VISIBLE set |
| 06:43 |
hmmmm |
if either one of those conditions are false, then the minimap isn't shown to the user |
| 06:44 |
hmmmm |
the first one is controlled by the client, |
| 06:44 |
hmmmm |
the second one is controlled by the server |
| 06:44 |
est31 |
it should put MinimapMode to force MINIMAP_MODE_OFF on login |
| 06:44 |
est31 |
if it sees HUD_FLAGS_MINIMAP_VISIBLE |
| 06:44 |
est31 |
to be not set |
| 06:44 |
hmmmm |
sure, it can do that |
| 06:44 |
hmmmm |
not necessarily on login |
| 06:44 |
hmmmm |
stop putting game logic in weird places |
| 06:44 |
est31 |
well, check it at every server step then |
| 06:44 |
est31 |
err client |
| 06:45 |
hmmmm |
on client draw |
| 06:45 |
est31 |
players will love how fast it is |
| 06:45 |
hmmmm |
there are obvious solutions to each of these objections of yours if you were to think about it for longer than 5 seconds |
| 06:45 |
hmmmm |
stop being so negative |
| 06:47 |
est31 |
the root problem is i think that localplayer has no notifications if hud_flags change |
| 06:47 |
est31 |
notification* |
| 06:47 |
est31 |
changes* |
| 06:47 |
est31 |
it was perfectly fine for the other hud elements |
| 06:47 |
est31 |
but minimap is special |
| 06:47 |
hmmmm |
i think it's perfectly fine as well |
| 06:47 |
est31 |
we have to change the minimap mode if its turned off |
| 06:47 |
hmmmm |
there is no actual penalty to setting the minimap mode |
| 06:48 |
est31 |
well, code it, and then lets talk again |
| 07:01 |
RealBadAngel |
btw, last used minimap mode shall be saved and restored imho |
| 07:03 |
|
leat joined #minetest-dev |
| 07:07 |
|
Darcidride joined #minetest-dev |
| 07:19 |
hmmmm |
https://github.com/kwolekr/minetest/commit/ea88855ecef65e88c3aa92bbf5d75e2d6cb5b719 |
| 07:21 |
hmmmm |
and then after the release we'll change the game->mapper interface to be hud->mapper |
| 07:22 |
hmmmm |
so you can do all the same things with a minimap that you can with the statbar, healthbar, etc. |
| 07:24 |
|
celeron55 joined #minetest-dev |
| 07:36 |
|
kilbith joined #minetest-dev |
| 07:42 |
|
dzho joined #minetest-dev |
| 07:42 |
|
Ardonel joined #minetest-dev |
| 07:42 |
|
crazyR joined #minetest-dev |
| 07:42 |
|
Routh joined #minetest-dev |
| 07:42 |
|
ShadowBot joined #minetest-dev |
| 07:42 |
|
Kray joined #minetest-dev |
| 07:43 |
|
crazyR_ joined #minetest-dev |
| 07:45 |
|
ShadowBot` joined #minetest-dev |
| 07:47 |
|
Kray joined #minetest-dev |
| 07:48 |
|
dzho joined #minetest-dev |
| 07:51 |
|
Amaz joined #minetest-dev |
| 07:51 |
|
est31 joined #minetest-dev |
| 08:00 |
|
Ardonel joined #minetest-dev |
| 08:00 |
|
Yepoleb_ joined #minetest-dev |
| 08:00 |
est31 |
hmmmm, commented on your commit |
| 08:00 |
est31 |
otherwise its good |
| 08:01 |
est31 |
but please address the remarks before merging |
| 08:01 |
|
Routh joined #minetest-dev |
| 08:02 |
est31 |
but I wonder where the mask gets changed |
| 08:02 |
|
nrzkt joined #minetest-dev |
| 08:04 |
est31 |
ah I see how it works |
| 08:04 |
est31 |
sorry for being a bit rude btw |
| 08:05 |
hmmmm |
you seem like you're mad at something |
| 08:05 |
hmmmm |
just fyi |
| 08:05 |
|
rubenwardy joined #minetest-dev |
| 08:12 |
est31 |
I'll try to be more kind |
| 08:14 |
|
Darcidride_ joined #minetest-dev |
| 08:24 |
|
Calinou joined #minetest-dev |
| 08:40 |
est31 |
kaeza, can you double tap jump perhaps a bit faster? |
| 08:40 |
est31 |
does that work? |
| 08:40 |
est31 |
I can't reproduce #3043 |
| 08:40 |
ShadowBot |
https://github.com/minetest/minetest/issues/3043 -- In-game key configuration dialog ignores "double tap space to toggle fly". |
| 08:40 |
est31 |
the limit is 0.2 seconds |
| 08:42 |
kaeza |
nope, it ignores the setting completely |
| 08:43 |
kaeza |
i.e. if I disable it and click 'Save' I can still double tap |
| 08:43 |
est31 |
ahh |
| 08:43 |
est31 |
and if you restart? |
| 08:43 |
kaeza |
it works if I restart, yes |
| 08:44 |
est31 |
then I'll make the cache update |
| 08:44 |
est31 |
that should work |
| 08:44 |
est31 |
we'll perhaps get faster settings after 0.4.13 |
| 08:44 |
est31 |
but right now we have to cache them |
| 08:44 |
est31 |
and users think the setting is ignored |
| 08:48 |
|
err404 joined #minetest-dev |
| 08:55 |
kilbith |
so i tested client_mapblock_limit = 10 and the RAM is well restrained now |
| 08:55 |
kilbith |
thanks a million, est31 |
| 08:55 |
est31 |
10 is a bit low |
| 08:56 |
est31 |
but you can set it as you want :) |
| 08:56 |
kilbith |
i prefer don't reach over 1 GB of RAM |
| 08:56 |
kilbith |
simple, light and low-ressource |
| 08:57 |
est31 |
I don't think a mapblock takes 1 MB RAM |
| 08:58 |
est31 |
it would take at least 500 kb, if you get more than 1 GB ram with client_mapblock_limit = 1000 |
| 08:58 |
kilbith |
oops, i was watching at the overall OS consumption |
| 08:58 |
kilbith |
nvm then |
| 08:59 |
est31 |
happy time playing around kilbith :) |
| 09:00 |
Calinou |
I set my view range to 256, and don't want to be restricted by that limit :) |
| 09:00 |
kilbith |
minetest sticks under <350 MB |
| 09:01 |
kilbith |
not as strict as MC, but good enough |
| 09:28 |
|
T4im joined #minetest-dev |
| 10:07 |
est31 |
#3044 |
| 10:07 |
ShadowBot |
https://github.com/minetest/minetest/issues/3044 -- game.cpp: Update cached settings by est31 |
| 10:09 |
|
FR^2 joined #minetest-dev |
| 10:28 |
est31 |
interesting, our launchpad builds fail on 15.10 |
| 10:30 |
|
rubenwardy joined #minetest-dev |
| 10:53 |
rubenwardy |
#3011 |
| 10:53 |
ShadowBot |
https://github.com/minetest/minetest/issues/3011 -- Periodic "Lua: error in error handling" crashes |
| 10:54 |
rubenwardy |
I know how mmdb works and I am a developer for it |
| 10:55 |
rubenwardy |
Well, contributor with write access |
| 10:55 |
est31 |
can you get it back up? |
| 10:55 |
rubenwardy |
no, celeron55 has to do that |
| 10:55 |
rubenwardy |
there's documentation in the readme |
| 10:56 |
rubenwardy |
It's an unrealistic project as it requires every mod to be registered. Minetest-Bower combined with MTPM Lists is much better as it fetches from the forum. |
| 10:56 |
rubenwardy |
It also is a pain to update mods on |
| 10:56 |
rubenwardy |
an absolute pain |
| 10:56 |
|
Lunatrius joined #minetest-dev |
| 10:57 |
est31 |
rubenwardy, agree. |
| 10:57 |
est31 |
so you vote for removing the button? |
| 10:57 |
rubenwardy |
(MTPM Lists is a forum parser) |
| 10:57 |
rubenwardy |
I don't know, one day it could be made better |
| 10:57 |
est31 |
well, as we have sandboxing, we can finally install untrusted mods |
| 10:57 |
est31 |
yaay |
| 10:58 |
est31 |
for that we have to turn sandboxing on by default |
| 10:58 |
est31 |
but all these things are functional changes |
| 11:17 |
T4im |
does it supply a secured implementation for require(), btw.? or did that end up blocked now |
| 11:18 |
rubenwardy |
T4im, minetest.request_insecure_environment() is the only way |
| 11:19 |
T4im |
rubenwardy: I know the workaround, I'm talking about a fix |
| 11:19 |
T4im |
;) |
| 11:19 |
T4im |
we talked about that a while ago, just wondered if that was dropped or added |
| 11:22 |
RealBadAngel |
bower is an excellent thingy |
| 11:22 |
RealBadAngel |
i just fucking love it |
| 11:23 |
RealBadAngel |
hail cornerote for it |
| 11:23 |
T4im |
last time I checked bower installed dependencies without checking signatures of the installments for any form of injection/manipulation |
| 11:24 |
T4im |
hope they fixed that by now :( |
| 11:26 |
T4im |
oh wait, you're talking different bower, my bad |
| 11:29 |
est31 |
wb T4im :) |
| 11:29 |
T4im |
hey est31 :) |
| 11:30 |
est31 |
ShadowNinja has supplied a require() impl, but it didnt get merged |
| 11:30 |
est31 |
because replicating all the properties of the function is very hard and complicated |
| 11:31 |
T4im |
oh ok |
| 11:33 |
est31 |
also, we have depends.txt, we dont need require |
| 11:34 |
|
kilbith joined #minetest-dev |
| 11:34 |
T4im |
require can be helpful within a mod, or if you still remember mtt when including external libraries that make use of it (that otherwise would go through mod_security, if it weren't for the require) |
| 11:36 |
kilbith |
est31: something i don't understand with mapblocks limit, you set the limit to 1000 (1000*16 nodes = half of the map) |
| 11:36 |
est31 |
half of the map? |
| 11:36 |
kilbith |
but before you needed less than that for overwhelming your RAM |
| 11:36 |
kilbith |
with 4 GB of RAM |
| 11:36 |
kilbith |
yes, a mapblock = 16^3 nodes |
| 11:37 |
kilbith |
1000 mapblocks = 16000^3 nodes |
| 11:37 |
est31 |
no |
| 11:37 |
|
ElectronLibre joined #minetest-dev |
| 11:37 |
est31 |
1000 mapblocks == 1000 * (16^3) nodes |
| 11:37 |
est31 |
1000 * (16^3) != (1000 * 16)^3 |
| 11:38 |
kilbith |
this is what i'm saying |
| 11:38 |
T4im |
I think it's also a radius, so you have 1000 into each direction from the center, don't you? |
| 11:38 |
kilbith |
you allow caching half of the map |
| 11:38 |
kilbith |
it's still way too much |
| 11:38 |
est31 |
no |
| 11:38 |
est31 |
the map is much more |
| 11:39 |
est31 |
1000 mapblocks, thats even less than a mapgen chunk |
| 11:39 |
kilbith |
31000/16 = ~1950 nodes |
| 11:39 |
kilbith |
oops |
| 11:39 |
kilbith |
~1950 mapblocks |
| 11:41 |
est31 |
kilbith, so you say a cubic meter of water has only 10 liters, because one liter is 10 cm ^3 ? |
| 11:41 |
T4im |
(62k^3)/16 = ~1.48 * 10^13 |
| 11:41 |
T4im |
1.49* |
| 11:42 |
kilbith |
when i wrote "31000/16" it means in all dirs |
| 11:44 |
kilbith |
so yeah, there around 1950 mapblocks on the map |
| 11:45 |
kilbith |
31000*31000*31000 nodes = 2.9791e+13 nodes |
| 11:45 |
kilbith |
1937*16 * 1937*16 * 1937*16 = 2.9791e+13 nodes |
| 11:46 |
est31 |
31000 / 16 * 31000 / 16 * 31000/16 = 7273193359 |
| 11:46 |
est31 |
thats how many mapblocks exist on the map |
| 11:48 |
kilbith |
it's correct |
| 11:48 |
kilbith |
7273193359*(16*16*16) = 2.9791e+13 nodes |
| 11:48 |
|
proller joined #minetest-dev |
| 11:51 |
kilbith |
well, miscalculation then sorry |
| 11:51 |
|
leat joined #minetest-dev |
| 11:52 |
|
proller joined #minetest-dev |
| 11:57 |
T4im |
58185546875 * |
| 11:58 |
T4im |
you only described on of 8 quadrants |
| 11:58 |
T4im |
one* |
| 11:58 |
T4im |
and I misscalculated too up there by forgetting the cube in the mapblocksize :D |
| 12:15 |
|
diemartin joined #minetest-dev |
| 12:29 |
|
amadin joined #minetest-dev |
| 12:31 |
amadin |
can i speak with developer who added redis support to server? |
| 12:58 |
|
SopaXT joined #minetest-dev |
| 13:39 |
|
FR^2 left #minetest-dev |
| 14:03 |
|
paramat joined #minetest-dev |
| 14:10 |
paramat |
nore sfan5 game#620 is ready to go please can you review? |
| 14:10 |
ShadowBot |
https://github.com/minetest/minetest_game/issues/620 -- Flowers: Add sporeless mushrooms by paramat |
| 14:17 |
paramat |
1000 mapblocks is 1000/125 = 8 = 2^3 mapchunks, a really small volume |
| 14:17 |
|
paramat left #minetest-dev |
| 14:27 |
Wayward_One |
The latest git keeps crashing without an error message for me |
| 14:27 |
Wayward_One |
Tried after a git pull and a fresh clone |
| 14:29 |
Wayward_One |
Also, debug.txt is empty |
| 14:33 |
VanessaE |
nothing at all on the console? |
| 14:33 |
rubenwardy |
gdb? |
| 14:34 |
VanessaE |
can you build in debug mode and get a gdb backtrace? |
| 14:34 |
VanessaE |
damn it, ninja's |
| 14:34 |
VanessaE |
'd* |
| 14:34 |
Wayward_One |
Trying with the console now |
| 14:34 |
Wayward_One |
So far, it's only when joining a server |
| 14:34 |
rubenwardy |
what to do... |
| 14:35 |
VanessaE |
what the... |
| 14:35 |
VanessaE |
it doesn't crash for me, but there's ZERO console output now |
| 14:35 |
Wayward_One |
Hmm, seg fault for me |
| 14:35 |
VanessaE |
vanessa rainbird:~$ minetest |
| 14:35 |
VanessaE |
<let it sit for a bit on the server, then quit> |
| 14:36 |
VanessaE |
AL lib: (WW) FreeDevice: (0x3517390) Deleting 51 Buffer(s) |
| 14:36 |
VanessaE |
vanessarainbird:~$ |
| 14:36 |
VanessaE |
this is bad. |
| 14:36 |
rubenwardy |
explicitly set log level? |
| 14:37 |
VanessaE |
my debug.txt has only this: |
| 14:37 |
VanessaE |
http://pastebin.com/GJwkJTvB |
| 14:39 |
VanessaE |
(however I can hit that URL just fine in a browser) |
| 14:39 |
rubenwardy |
doesn't happen to me |
| 14:40 |
VanessaE |
Wayward_One: file an issue. include my paste and comments |
| 14:40 |
VanessaE |
(nearly no console output and debug.txt is bad) |
| 14:40 |
Wayward_One |
Will do |
| 14:41 |
VanessaE |
I mean include my paste/comments secondary to your experience, of course :) |
| 14:41 |
VanessaE |
are you able to sign onto SC-Skyblock? |
| 14:41 |
VanessaE |
that's where I was when I tried it. |
| 14:42 |
Wayward_One |
Hmm, I'll try |
| 14:42 |
VanessaE |
("skyblock server" in the public list, stormchaser3000.noip.me:30001) |
| 14:43 |
|
nore joined #minetest-dev |
| 14:43 |
Wayward_One |
Yep, seg fault again |
| 14:44 |
VanessaE |
ok. |
| 14:45 |
rubenwardy |
Wayward_One, use gdb to find the origin :P bt full |
| 14:58 |
|
kilbith joined #minetest-dev |
| 14:59 |
kilbith |
i confirm the Wayward_One report and i caught a gdb trace on joining a server |
| 14:59 |
kilbith |
here it is : http://pastie.org/10348657 |
| 14:59 |
kilbith |
est31, if you read me |
| 15:00 |
Wayward_One |
https://github.com/minetest/minetest/issues/3047 |
| 15:11 |
|
Krock joined #minetest-dev |
| 15:21 |
rubenwardy |
kilbith, bt full ? |
| 15:21 |
celeron55 |
i tried this, and reproduced it, but once i ran it in gdb the crash disappeared |
| 15:22 |
celeron55 |
hmm i think i just got this when running in valgrind |
| 15:22 |
celeron55 |
i mean, memcheck |
| 15:23 |
celeron55 |
http://fpaste.org/254808/94793871/ |
| 15:26 |
|
est31 joined #minetest-dev |
| 15:29 |
est31 |
ill look @ the regression later |
| 15:29 |
est31 |
busy atm |
| 15:31 |
celeron55 |
it's probably caused by this: https://github.com/minetest/minetest/commit/a8e238ed06ee8285ed4459e9deda3117419837f6?diff=unified#diff-ab005ad9757130721ffb66a09833090aR1502 |
| 15:32 |
celeron55 |
when mapblock_queue is empty |
| 15:32 |
celeron55 |
if i am not mistaken |
| 15:36 |
|
rubenwardy joined #minetest-dev |
| 15:41 |
celeron55 |
can someone test this? https://github.com/minetest/minetest/issues/3047#issuecomment-130732141 |
| 15:42 |
celeron55 |
(i can't trust my tests in a timing-related issue like this) |
| 15:44 |
celeron55 |
(or, well it's probably right, but i won't push it without confirmation) |
| 15:50 |
Amaz |
celeron55, it seems to work for me. |
| 15:52 |
|
ElectronLibre joined #minetest-dev |
| 15:55 |
|
BlockMen joined #minetest-dev |
| 15:58 |
est31 |
celeron55, even if it doesnt fix the bug, it probably has to be done |
| 15:58 |
est31 |
but most likely it will fix the bug |
| 15:59 |
est31 |
I have tested my patch |
| 15:59 |
est31 |
but might have missed setting the timeout back to default |
| 15:59 |
est31 |
so I've tested it so that there is always an element with mapblock_queue.top().block->getUsageTimer() > unload_timeout |
| 15:59 |
est31 |
before the end |
| 16:01 |
est31 |
+1 |
| 16:04 |
celeron55 |
pushed |
| 16:05 |
celeron55 |
it was actually a bit more than half a year since my previous commit... it sure felt like it too |
| 16:05 |
celeron55 |
no, actually a bit less; well whatever |
| 16:07 |
celeron55 |
well, i went ahead and closed the issue; it's highly likely that this was the fix |
| 16:08 |
est31 |
the crash was most likely in the top method |
| 16:08 |
est31 |
it makes very much sense |
| 16:09 |
|
ElectronLibre joined #minetest-dev |
| 16:10 |
celeron55 |
the reason why people didn't catch that earlier was probably that it doesn't occur on fast development setups; instead it occurred on slow connections and laggy servers |
| 16:10 |
celeron55 |
i guess a fast server tends to push a block to the client before the client runs that code... at least that would make sense |
| 16:11 |
|
jin_xi joined #minetest-dev |
| 16:11 |
VanessaE |
no negative effect here but console output is still (nearly) dead |
| 16:11 |
VanessaE |
(not that your commit had anything to do with that) |
| 16:12 |
est31 |
I've explained the reason I didnt catch it above |
| 16:12 |
est31 |
also I've only tested singleplayer |
| 16:13 |
Wayward_One |
Works for me as well |
| 16:19 |
BlockMen |
Do we release a 2nd RC? |
| 16:22 |
|
Robert_Zenz joined #minetest-dev |
| 16:22 |
|
Calinou joined #minetest-dev |
| 16:22 |
VanessaE |
I think paramat has a mushrooms change to go into mt_game, might want to wait for that |
| 16:26 |
BlockMen |
VanessaE, k |
| 16:27 |
|
Krock2 joined #minetest-dev |
| 16:29 |
BlockMen |
could someone test #3041? it should fix all those nodebox-sneaking issues |
| 16:29 |
ShadowBot |
https://github.com/minetest/minetest/issues/3041 -- Fix sneaking (fixes #329 and #665) by BlockMen |
| 16:31 |
|
hmmmm joined #minetest-dev |
| 16:31 |
VanessaE |
shouldn't the player only rise to the top of the collision box they're standing in/on? rather than just whatever's highest in the whole node? |
| 16:32 |
est31 |
thats what I've said too |
| 16:32 |
est31 |
take the height at the border |
| 16:32 |
BlockMen |
VanessaE, yes. it actually does now. havent updated the desc |
| 16:32 |
VanessaE |
oh ok. |
| 16:32 |
VanessaE |
you should do that :P |
| 16:33 |
BlockMen |
well, at least it worked now when i tested it today |
| 16:39 |
|
nrzkt joined #minetest-dev |
| 16:46 |
|
SopaXT joined #minetest-dev |
| 16:47 |
|
diemartin joined #minetest-dev |
| 17:15 |
|
Robert_Zenz joined #minetest-dev |
| 17:16 |
|
est31 joined #minetest-dev |
| 17:26 |
|
MinetestForFun joined #minetest-dev |
| 17:27 |
|
MinetestForFun joined #minetest-dev |
| 17:40 |
est31 |
Pushing #3044 in 10 minutes if nobody objects |
| 17:40 |
ShadowBot |
https://github.com/minetest/minetest/issues/3044 -- game.cpp: Update cached settings by est31 |
| 17:40 |
est31 |
its fairly trivial |
| 17:40 |
est31 |
just lots of code copy |
| 17:43 |
|
paramat joined #minetest-dev |
| 17:47 |
paramat |
nore sfan5 please can i push game#620 ? |
| 17:47 |
ShadowBot |
https://github.com/minetest/minetest_game/issues/620 -- Flowers: Add sporeless mushrooms by paramat |
| 17:47 |
est31 |
why mapgen mushrooms? |
| 17:48 |
est31 |
if something grows from the spores, you get the "mapgen" mushrooms too, no? |
| 17:48 |
nore |
same question |
| 17:48 |
nore |
perhaps mushroom_with_spores_ |
| 17:49 |
nore |
also, it should have not_in_creative_inventory = 1 in groups |
| 17:49 |
paramat |
yes mushrooms growing from spores will be 'mapgen' mushrooms, hm |
| 17:49 |
paramat |
i was trying to find a reasonable name, 'unpicked' 'spored' ? |
| 17:50 |
est31 |
"with spores" ? |
| 17:50 |
paramat |
sure which should be in creative inventory? |
| 17:50 |
kahrl |
you could also use the same content id for both but a different param2 |
| 17:51 |
est31 |
what if we want to make them look differently? |
| 17:51 |
nore |
paramat: the one that you can pick in normal gameplay, i.e. the sporeless should be the one in creative inventory |
| 17:51 |
kahrl |
then not :) |
| 17:51 |
est31 |
also, then the description would then need to be in the meta |
| 17:51 |
nore |
kahrl: mushrooms are plantlike |
| 17:51 |
est31 |
if it differs between spores and spored |
| 17:52 |
est31 |
and it does |
| 17:52 |
nore |
so I guess param2 would rotate them |
| 17:52 |
paramat |
nore agreed |
| 17:52 |
kahrl |
nore, oh yeah |
| 17:53 |
kahrl |
it's really confusing that those have paramtype2 = "none" yet the engine still cares about param2 |
| 17:53 |
nore |
even in paramtype2 is none ? |
| 17:53 |
kahrl |
yeah, pretty sure |
| 17:54 |
nore |
hm... |
| 17:54 |
est31 |
cares in which sense? |
| 17:54 |
nore |
I thought there was a plantlike paramtype2, but I must have confused with something else |
| 17:54 |
kahrl |
est31: rotation angle |
| 17:54 |
est31 |
thats bad |
| 17:55 |
paramat |
nore sfan5 also game#622 is nearly ready, just a final tweak to the texture to do |
| 17:55 |
ShadowBot |
https://github.com/minetest/minetest_game/issues/622 -- Default: New grass textures, new footsteps overlay texture WIP by paramat |
| 17:55 |
est31 |
I would agree that it stores the param2 in the map, so it doesnt actively ignore it |
| 17:55 |
est31 |
but using it for rotation is wrong IMO |
| 17:55 |
est31 |
but changing that should be done after the release |
| 17:55 |
sfan5 |
game#622 looks goode paramat |
| 17:55 |
ShadowBot |
https://github.com/minetest/minetest_game/issues/622 -- Default: New grass textures, new footsteps overlay texture WIP by paramat |
| 17:55 |
sfan5 |
i'll look at 620 later |
| 17:56 |
VanessaE |
+1 on 622 also. |
| 17:56 |
paramat |
nore sfan5 comments on game#581 ? i suggest steelblock and straw, any others you want in for 0.4.13? |
| 17:56 |
ShadowBot |
https://github.com/minetest/minetest_game/issues/581 -- New stairs and slabs discussion |
| 17:56 |
nore |
622 looks good to me too |
| 17:56 |
paramat |
thanks |
| 17:56 |
sfan5 |
paramat: straw? |
| 17:57 |
nore |
paramat: straw looks good, for steel I think other metal blocks should have it too |
| 17:58 |
paramat |
straw for thatched roofs, i'm not too bothered about straw |
| 17:58 |
nore |
but that would bloat creative inventory, so we need a way to fix that |
| 17:59 |
nore |
btw, sfan5, paramat: did you look at game#626 ? |
| 17:59 |
ShadowBot |
https://github.com/minetest/minetest_game/issues/626 -- Add needed tileable flags for grass nodes by RealBadAngel |
| 17:59 |
sfan5 |
oh |
| 17:59 |
paramat |
okay, but for 0.4.13 just steel, or all metal blocks? |
| 17:59 |
sfan5 |
straw = hay |
| 17:59 |
sfan5 |
straw slabs hm |
| 17:59 |
sfan5 |
I'm not sure |
| 18:00 |
VanessaE |
sfan5: cottages mod has various shapes of straw |
| 18:00 |
VanessaE |
and I think one of the mods in plantlife also adds some similar nodes. |
| 18:01 |
VanessaE |
paramat: all metal blocks. |
| 18:01 |
paramat |
yeah 626 has some odd already merged stuff in it, needs redoing |
| 18:01 |
VanessaE |
(but not diamond, that'd be as stupid as coal block stairs :P ) |
| 18:01 |
paramat |
hehe |
| 18:02 |
paramat |
finally i will prepare a PR for new junglewood and jungletree top, using Vanessa'a initial 619 texture, plus my dark junglewood shown in #611 |
| 18:02 |
ShadowBot |
https://github.com/minetest/minetest/issues/611 -- Animation blending. by blue42u |
| 18:02 |
paramat |
arg game#611 |
| 18:02 |
ShadowBot |
https://github.com/minetest/minetest_game/issues/611 -- Junglewood texture too light, boring |
| 18:03 |
VanessaE |
paramat: you mean this one: https://github.com/minetest/minetest_game/issues/611#issuecomment-129258990 |
| 18:03 |
VanessaE |
or the one right after? |
| 18:04 |
paramat |
for new grass i was thinking of making it very slightly lighter fro a closer match to the 0.4.10 texture |
| 18:04 |
est31 |
Pushing #3044 |
| 18:04 |
ShadowBot |
https://github.com/minetest/minetest/issues/3044 -- game.cpp: Update cached settings by est31 |
| 18:04 |
VanessaE |
nah what you have now seems good to me. about the footsteps: make them show some dirt also maybe? |
| 18:04 |
paramat |
the initial 619 texture that had approval form 2 devs |
| 18:04 |
VanessaE |
ok |
| 18:05 |
paramat |
(from) |
| 18:05 |
paramat |
i was thinking the footprints could be usable for other materials in future? |
| 18:05 |
VanessaE |
of course. |
| 18:05 |
VanessaE |
I just meant footprints over grass |
| 18:06 |
VanessaE |
you should be able to composite in default dirt using the footprint as a mask layer |
| 18:06 |
paramat |
well it's an overlay texture, so if i add dirt pixels it can't be used for sand etc |
| 18:06 |
VanessaE |
something like how fence inv is done |
| 18:07 |
|
luizrpgluiz joined #minetest-dev |
| 18:07 |
|
blaze joined #minetest-dev |
| 18:08 |
paramat |
so there's talk of a release in a few days, i can get game ready to go for the end of the weekend but no sooner, please dont release without talking to me first! =) |
| 18:09 |
paramat |
ah 626 is closed by RBA, would be nice to have it though for 0.4.13 |
| 18:10 |
VanessaE |
paramat: about masking, i mean make the footprints just a translucent greyscale image, use it to clip out some dirt at a very low opacity, overlay that onto the grass. If used on e.g. sand or snow, you'd just directly overlay the greyscale image without the extra masking step to get a dark footprint. |
| 18:10 |
est31 |
game#626 |
| 18:10 |
paramat |
okay thanks |
| 18:10 |
ShadowBot |
https://github.com/minetest/minetest_game/issues/626 -- Add needed tileable flags for grass nodes by RealBadAngel |
| 18:10 |
est31 |
#626 |
| 18:10 |
ShadowBot |
https://github.com/minetest/minetest/issues/626 -- fix memory leak on sound shutdown by sapier |
| 18:10 |
paramat |
heh |
| 18:10 |
luizrpgluiz |
that strange, for me the release candidate is too good to be launched, the mapgens the v5 and v7 are very good |
| 18:11 |
nore |
paramat: it can be merged anyway |
| 18:12 |
paramat |
nore game626? commit needs editing and approval from RBA |
| 18:12 |
nore |
hm... |
| 18:13 |
nore |
RealBadAngel: can we merge game#626 ? |
| 18:13 |
ShadowBot |
https://github.com/minetest/minetest_game/issues/626 -- Add needed tileable flags for grass nodes by RealBadAngel |
| 18:13 |
VanessaE |
paramat: you would need two footprint images though to make it work (one with the totally-transparent perimeter replaced by some solid color that you could chroma-key against dirt, one without, for the plain overlay on sand et al.) |
| 18:13 |
luizrpgluiz |
out that on some computers home here could not see the minimap in the game even activating it in the config minetest release canditate |
| 18:13 |
VanessaE |
you can figure it out :) |
| 18:14 |
paramat |
meh well maybe eventually, for release i'll leave it as is |
| 18:14 |
VanessaE |
ok. |
| 18:17 |
paramat |
okay i assume i have approval for the mushroom commit once i rename to 'mushroom_with_spores_..' and do the 'not in creative inventory' thing. will push later |
| 18:17 |
luizrpgluiz |
VanessaE is right for me common player the game is in good condition, outside the bugs that appeared in the git hub to be arranged also |
| 18:17 |
VanessaE |
wat |
| 18:18 |
luizrpgluiz |
Paramat just need to fix some bugs mapgen appearing on git hub |
| 18:21 |
VanessaE |
can someone please interpret that? :P |
| 18:21 |
luizrpgluiz |
hehe,sorry for my english |
| 18:27 |
luizrpgluiz |
the Paramat may have tidy bugs before releasing the candidate releace :) |
| 18:27 |
luizrpgluiz |
on mapgen :D |
| 18:27 |
paramat |
there are some minor mapgen bugs but that's for after release, i'm too busy preparing mtgame |
| 18:28 |
luizrpgluiz |
:) |
| 18:30 |
luizrpgluiz |
yes, this is good because it is very important to fix bugs in the game map generator |
| 18:31 |
kahrl |
nore: IANAL but I think by posting the PR one implicitly releases the code under the project's license, which in the case of copyleft licenses can't be retroactively retracted |
| 18:31 |
est31 |
well, you can claim its gpl |
| 18:31 |
est31 |
and has been all the time |
| 18:32 |
est31 |
and if its new media files, its even easier |
| 18:32 |
est31 |
but IANAL too |
| 18:32 |
VanessaE |
thing is there is chain of evidence |
| 18:32 |
VanessaE |
the commit ID |
| 18:33 |
VanessaE |
so posting a PR and getting it merged is an explicit offer to re-license your code if it's not already properly licensed. |
| 18:33 |
nore |
kahrl: well, I said that because of PilzAdam's comment |
| 18:33 |
VanessaE |
and if someone were to sue, I'm sure github would have no trouble coughing up the commit logs :P |
| 18:33 |
nore |
but IANAL either |
| 18:42 |
|
Hunterz joined #minetest-dev |
| 18:46 |
paramat |
i'll make a stairs mod PR for steel, copper, bronze blocks and straw (not coal, gold, mese, diamond) |
| 18:47 |
nore |
paramat: gold could have stairs too |
| 18:48 |
paramat |
hm perhaps.. |
| 18:50 |
|
BlockMen joined #minetest-dev |
| 18:59 |
paramat |
okay sure gold too |
| 19:01 |
VanessaE |
also check the model itself, I think I noticed a z-fighting issue when "inside" it once (noclip). same will happen if someone (e.g. Calinou) uses the default model to make say glass stairs (say moreblocks switches to using it) |
| 19:02 |
VanessaE |
yep, it does. |
| 19:03 |
VanessaE |
solution via RBA: scale the whole model up just a tiny amount (I'd say to 100.05 percent of current size?) so that the overlapping stair faces get buried under their neighbors. |
| 19:04 |
VanessaE |
(I guess at that point, irrlicht or the driver will cull them since they're hidden?) |
| 19:05 |
VanessaE |
reference image: http://digitalaudioconcepts.com/vanessa/hobbies/minetest/screenshots/random/Screenshot_2015-08-13_15-05-05.png |
| 19:06 |
VanessaE |
doesn't happen with slabs since they're not a mesh |
| 19:07 |
|
MinetestForFun joined #minetest-dev |
| 19:15 |
|
Ardonel joined #minetest-dev |
| 19:31 |
paramat |
is it a problem if only seen when inside? seems best to have the exact proper size |
| 19:33 |
VanessaE |
for solid textures yeah |
| 19:34 |
VanessaE |
but if someone uses the stairs model to produce stairs of any translucent material, it'll be visible from the outside too |
| 19:34 |
VanessaE |
maybe just turn on basic backface culling then? |
| 19:35 |
VanessaE |
s/translucent/non-opaque/ |
| 19:36 |
paramat |
i see |
| 19:36 |
VanessaE |
(this can be done with the "special tiles" node def option, but not with regular tiles) |
| 19:40 |
paramat |
changing stairs to meshes seems to have caused problems, sometimes i think we should go back to the nodebox |
| 19:40 |
VanessaE |
nah, the best thing to do is fixed the problems instead |
| 19:41 |
VanessaE |
using one mesh means now all mods out there that create stairs can refer to it. |
| 19:41 |
VanessaE |
that means less code to plough through,t oo |
| 19:44 |
VanessaE |
(I am of course ignoring the stairs mod's API) |
| 19:46 |
paramat |
agreed |
| 19:48 |
VanessaE |
by the way, is there any kind of ref counting or indexing used to display the same model over and over? |
| 19:49 |
VanessaE |
(I mean, for example, is each node in a long staircase its own unique mesh? or does the driver/library fetch the same block of mesh data over and over for each stair?) |
| 20:02 |
|
err404 joined #minetest-dev |
| 20:23 |
|
kaeza joined #minetest-dev |
| 20:35 |
|
paramat left #minetest-dev |
| 20:42 |
|
Amaz joined #minetest-dev |
| 20:43 |
hmmmm |
does anybody know how to use the capturetheflag game? |
| 20:44 |
hmmmm |
i put it in games/, create a world in it, start the world, and i get this: 2015-08-13 16:39:22: ERROR[main]: mod "ctf_match" has unsatisfied dependencies: "ctf" "ctf_flag" |
| 20:44 |
hmmmm |
isn't the game supposed to include that? |
| 20:46 |
|
H-H-H joined #minetest-dev |
| 20:47 |
|
est31 joined #minetest-dev |
| 20:49 |
est31 |
hmmmm, can you add my ideas to the acceptance criteria? https://github.com/minetest/minetest/issues/3051#issuecomment-130836394 |
| 20:49 |
hmmmm |
well hold on |
| 20:49 |
hmmmm |
zoom is covered under 'mode' |
| 20:50 |
hmmmm |
anyway now that RBA isn't around anymore i can speak a little free-er about this... the concept of minimap modes suck |
| 20:50 |
hmmmm |
zoom should be its own variable, shape should be its own variable, radar/surface should be its own variable |
| 20:50 |
hmmmm |
the way it is now just adds repetitiveness to code |
| 20:50 |
est31 |
if its client set, we need to specify it |
| 20:50 |
est31 |
its perfectly fine what RBA did |
| 20:50 |
hmmmm |
we'll cover all this stuff when we rework the minimap |
| 20:51 |
est31 |
well, it seems you agree to what I say then |
| 20:52 |
hmmmm |
this story basically boils down to "the minimap needs a lot of work" |
| 20:53 |
est31 |
lots of areas need a lot of work |
| 20:53 |
est31 |
right now its very limited |
| 20:53 |
hmmmm |
yeah |
| 20:53 |
est31 |
just like minecraft hardcoding the nodedefs |
| 20:54 |
est31 |
we have to get it to minetest niveau |
| 20:54 |
hmmmm |
minecraft isn't better than minetest, it just has different problems |
| 20:54 |
est31 |
and send the modedefs at connection |
| 20:54 |
hmmmm |
I like to think we're doing okay for a bunch of volunteers |
| 20:54 |
est31 |
hmmmm, I've meant it the other way, minetest is better than minecraft :) |
| 20:54 |
hmmmm |
of course |
| 20:54 |
hmmmm |
minecraft seems to have more features and is more polished etc. |
| 20:54 |
est31 |
but yeah, it has areas where it shines and we dont |
| 20:55 |
hmmmm |
it's easy to do this when everything is hardcoded |
| 20:55 |
est31 |
^ |
| 20:55 |
est31 |
and when you can hire devs for money |
| 20:55 |
hmmmm |
right |
| 20:55 |
hmmmm |
if we were working on minecraft, then we'd already be done with the minimap and we'd move onto the next feature |
| 20:56 |
est31 |
yea. |
| 20:56 |
est31 |
we have the flag, thats all what our users need |
| 20:56 |
hmmmm |
it's what they need right now :) |
| 20:57 |
hmmmm |
in any case I don't need to really modify my own comment to add on your bullet points... do I? |
| 20:57 |
hmmmm |
just add "Additional criteria" |
| 20:57 |
est31 |
ok |
| 20:57 |
kahrl |
if we were working on minecraft, we'd be randomly shuffling packet ids around each protocol version :D http://wiki.vg/Pre-release_protocol#Use_Item |
| 20:57 |
|
kilbith joined #minetest-dev |
| 20:57 |
est31 |
minecraft is a game, minetest is an engine xD |
| 21:00 |
est31 |
and about this error showing on mod list errors hmmmm, it turns out that this stupid code just sets an empty json tree value, on the case an error occurs, and continues |
| 21:00 |
est31 |
so the only way to check it is with some emptiness check |
| 21:00 |
|
sloantothebone joined #minetest-dev |
| 21:00 |
hmmmm |
well, however it needs to be done :) |
| 21:00 |
est31 |
or to change the whole http json fetching api |
| 21:01 |
est31 |
I'd either have to return std::pair<json value, bool success> |
| 21:01 |
est31 |
or work with references/pointers |
| 21:01 |
est31 |
or use exceptions |
| 21:01 |
est31 |
or the easiest, just simply make the emptiness check |
| 21:02 |
hmmmm |
right now i'd favor simple solutions |
| 21:02 |
hmmmm |
I think that it's okay to add hacks, as long as you make an issue on github to fix it the proper way later on and promise to not forget about it |
| 21:03 |
est31 |
I'll look to where that API is used |
| 21:04 |
est31 |
I guess not much more than mainmenu lua |
| 21:04 |
est31 |
and if there is an error with fetching the data, we just return nil |
| 21:04 |
est31 |
and as the second param an error message |
| 21:10 |
|
RealBadAngel joined #minetest-dev |
| 21:23 |
|
ElectronLibre left #minetest-dev |
| 21:31 |
hmmmm |
hrm :( |
| 21:31 |
hmmmm |
well I am trying my hardest but I can't get delete_area to crash |
| 21:31 |
hmmmm |
rubenwardy says the map gets stuck after using it, but the only thing that gets stuck is the serverthread because it's so damn slow |
| 21:31 |
hmmmm |
because sqlite |
| 21:32 |
est31 |
I got it crash with a sqlite map |
| 21:45 |
hmmmm |
so rubenwardy's CTF game wants to delete 9261 blocks |
| 21:45 |
hmmmm |
each block takes anywhere from 37 microseconds to 250 milliseconds to even 130 milliseconds (probably when the db flushes to disk) |
| 21:46 |
hmmmm |
erm, i mean 25 milliseconds |
| 21:46 |
hmmmm |
130 milliseconds is the max |
| 21:46 |
hmmmm |
what he actually wants to do is delete the map |
| 21:47 |
hmmmm |
anyway the slowness can be solved by putting sqlite in async mode or just dropping the old db and deleting the file, starting anew |
| 21:47 |
est31 |
his usecase would work with deleteing the file I gues |
| 21:47 |
est31 |
s |
| 21:48 |
est31 |
but the general problem wont be solved |
| 21:48 |
hmmmm |
async sqlite could be dangerous |
| 21:58 |
|
rubenwardy joined #minetest-dev |
| 21:59 |
rubenwardy |
hmmmm: the capturetheflag game uses submodules, you need to --recursive |
| 22:00 |
rubenwardy |
my VPS uses a SSD and it takes roughly 3 seconds to delete 160^3 |
| 22:03 |
rubenwardy |
when in the capturetheflag game, you can use /ctf_next to trigger a delete_blocks. Providing you have the ctf_admin priv |
| 22:04 |
est31 |
do you have mods there? |
| 22:04 |
est31 |
mobs |
| 22:04 |
est31 |
sorry :) |
| 22:05 |
rubenwardy |
no |
| 22:06 |
|
proller joined #minetest-dev |
| 22:07 |
rubenwardy |
http://github.com/rubenwardy/capturetheflag for the subgame. Most of the code is in a modpack, which is included as a submodule in the subgame: http://github.com/rubenwardy/ctf_pvp_engine |
| 22:09 |
est31 |
can you replicate the crash in singleplayer? |
| 22:13 |
est31 |
hrmmm cant reproduce the crash either |
| 22:13 |
est31 |
but hmmmm I can remember seeing the crash on a server |
| 22:13 |
est31 |
I have issued that delete_area call |
| 22:14 |
|
rubenwardy joined #minetest-dev |
| 22:14 |
|
BlockMen left #minetest-dev |
| 22:16 |
rubenwardy |
I'm using the latest dev, and the server hasn't crashed in five days. I've stopped killing it at midnight, so it's been up all that time. The map blocks getting stuck doesn't happen either. |
| 22:16 |
rubenwardy |
i stopped the server today when i rebooted my vps to install updates. |
| 22:16 |
rubenwardy |
and forgot to restart it |
| 22:17 |
|
luizrpgluiz joined #minetest-dev |
| 22:17 |
rubenwardy |
well, latest dev as of five days ago |
| 22:17 |
rubenwardy |
remember that I posted stack tracebacks of segfaults in V's unexplained crashes thread. |
| 22:19 |
rubenwardy |
Anyway, gotta sleep now. I'll read the logs. |
| 22:24 |
luizrpgluiz |
in which the file is hosted pvp code of minetest? |
| 22:25 |
luizrpgluiz |
it is in some .cpp file or in any file .lua |
| 22:57 |
luizrpgluiz |
would be very good in some future update has just enabled pvp in just in one or more areas in minetest |
| 23:26 |
hmmmm |
i'd answer but i have no idea what hosted pvp is |
| 23:26 |
hmmmm |
est31: are you sure the delete_area crash wasn't fixed already? |
| 23:32 |
luizrpgluiz |
hehe,My idea would be to enable the pvp in certain map areas |
| 23:32 |
luizrpgluiz |
or off pvp also in areas determined by the server administrators |
