| Time |
Nick |
Message |
| 00:00 |
|
MoNTE48 joined #minetest-dev |
| 00:08 |
|
octacian_ joined #minetest-dev |
| 00:08 |
|
Fixer joined #minetest-dev |
| 00:08 |
|
octacian joined #minetest-dev |
| 00:08 |
|
Drigon joined #minetest-dev |
| 00:28 |
|
Grandolf_ joined #minetest-dev |
| 00:32 |
|
Miner_48er joined #minetest-dev |
| 00:39 |
|
paramat left #minetest-dev |
| 00:41 |
|
imr joined #minetest-dev |
| 00:46 |
|
Devy left #minetest-dev |
| 00:51 |
|
bigfoot547 joined #minetest-dev |
| 01:18 |
|
Devy joined #minetest-dev |
| 01:18 |
|
Devy left #minetest-dev |
| 02:00 |
|
kaeza joined #minetest-dev |
| 02:31 |
|
RiverOtter joined #minetest-dev |
| 02:40 |
|
paramat joined #minetest-dev |
| 02:47 |
paramat |
celeron55 please could you consider this critical issue? #5915 this could potentially make a mess of multiplayer MT |
| 02:47 |
ShadowBot |
https://github.com/minetest/minetest/issues/5915 -- CSM: Allowing a server to protect itself against clients running client-provided clientmods |
| 03:03 |
|
Devy joined #minetest-dev |
| 03:03 |
|
Devy left #minetest-dev |
| 03:15 |
|
Hunterz joined #minetest-dev |
| 03:22 |
|
Grandolf_ joined #minetest-dev |
| 03:23 |
|
Grandolf_ joined #minetest-dev |
| 03:25 |
|
Grandolf_ joined #minetest-dev |
| 03:26 |
|
Grandolf_ joined #minetest-dev |
| 03:28 |
|
Grandolf_ joined #minetest-dev |
| 04:08 |
|
KaadmY joined #minetest-dev |
| 05:18 |
|
Hunterz joined #minetest-dev |
| 05:41 |
|
Raven262 joined #minetest-dev |
| 05:54 |
|
DS-minetest joined #minetest-dev |
| 06:26 |
|
SG9kb3I joined #minetest-dev |
| 06:44 |
|
Puka joined #minetest-dev |
| 06:49 |
|
troller joined #minetest-dev |
| 06:50 |
|
KaadmY joined #minetest-dev |
| 06:50 |
|
Fritigern joined #minetest-dev |
| 06:51 |
|
Hunterz joined #minetest-dev |
| 07:04 |
|
Gundul joined #minetest-dev |
| 07:18 |
|
nerzhul joined #minetest-dev |
| 07:35 |
nerzhul |
hi, pushing a trivial fix, we have a SharedPtr class which is unused, removing it this will never be used (std::shared_ptr<> is there is C++11 if needed later) |
| 07:43 |
nerzhul |
merging #5909 docfix |
| 07:43 |
ShadowBot |
https://github.com/minetest/minetest/issues/5909 -- Fixed #5907 Documentation for screen_h and screen_w by Dumbeldor |
| 07:46 |
|
DS-minetest joined #minetest-dev |
| 08:35 |
|
kilbith joined #minetest-dev |
| 09:13 |
|
YuGiOhJCJ joined #minetest-dev |
| 09:38 |
red-001 |
#5852 |
| 09:38 |
ShadowBot |
https://github.com/minetest/minetest/issues/5852 -- Improve the path select GUI by red-001 |
| 09:42 |
|
DS-minetest joined #minetest-dev |
| 09:59 |
|
Hunterz joined #minetest-dev |
| 10:24 |
SG9kb3I |
https://github.com/minetest/minetest/blob/master/builtin/mainmenu/tab_credits.lua#L60 |
| 10:25 |
|
nerzhul joined #minetest-dev |
| 10:25 |
SG9kb3I |
What's the go there? Others (e.g. Zeno) did as much if not more optimization than gcu (you can check in the logs) |
| 10:27 |
Thomas-S |
nerzhul, could you please check if the HybridPtr thing in #5895 is ok for you? (I'm asking because you removed SharedPtr today.) Thanks for your efforts! |
| 10:27 |
ShadowBot |
https://github.com/minetest/minetest/issues/5895 -- Meta-set nodedef: Part 1 by Thomas--S |
| 10:29 |
|
kilbith joined #minetest-dev |
| 10:30 |
kilbith |
SG9kb3I, go troll elsewhere, Zeno |
| 10:30 |
red-001 |
if you are going to make it so obvious you are zeno you might as well change your nick |
| 10:36 |
SG9kb3I |
*sigh* |
| 10:38 |
|
Zeno` joined #minetest-dev |
| 10:39 |
Zeno` |
Last failed attempt from: |o|!red gateway/shell/elitebnc/x-tafhsqvroqtxbhac on Jun 03 19:10:28 2017. |
| 10:39 |
Zeno` |
I wonder who that could have been |
| 10:39 |
kilbith |
what's that |
| 10:39 |
Zeno` |
it's someone who attempted to login as me |
| 10:40 |
kilbith |
looks like red-001 |
| 10:40 |
Zeno` |
* [red-001] (redgateway/shell/elitebnc/x-tafhsqvroqtxbhac): red-001EliteBnc |
| 10:40 |
Zeno` |
yes, it does |
| 10:40 |
red-001 |
huh |
| 10:40 |
Zeno` |
-NickServ- Last failed attempt from: |o|!redgateway/shell/elitebnc/x-tafhsqvroqtxbhac on Jun 03 19:10:28 2017. |
| 10:40 |
kilbith |
the same one who threat servers to shut them down, right? |
| 10:40 |
Zeno` |
*shrug* that's what nickserv says |
| 10:40 |
red-001 |
werid |
| 10:41 |
red-001 |
I didn't do that |
| 10:41 |
red-001 |
it was your main account right? |
| 10:41 |
red-001 |
wait maybe my bnc was compromised |
| 10:42 |
Zeno` |
I could ask Fuchs I guess |
| 10:42 |
kilbith |
you won't admit you did that, naturally |
| 10:42 |
red-001 |
oh come on, you are trolling right> |
| 10:42 |
red-001 |
8? |
| 10:42 |
red-001 |
*? |
| 10:43 |
Zeno` |
I'm not. That is what nickserv sent me upon logging in |
| 10:43 |
red-001 |
huh |
| 10:43 |
red-001 |
maybe I did and I forgot |
| 10:43 |
kilbith |
sure, you forgot now |
| 10:45 |
|
EDAKIRI joined #minetest-dev |
| 10:47 |
Zeno` |
please don't try and login as me :/ |
| 10:47 |
Zeno` |
I dunno how you'd "accidentally" do it |
| 10:47 |
red-001 |
wait so this really happened??! |
| 10:48 |
red-001 |
wtf |
| 10:48 |
red-001 |
how the hell do I not remeber.... |
| 10:48 |
red-001 |
this is 2 days ago... |
| 10:49 |
red-001 |
brb reviewing my logs |
| 10:50 |
Zeno` |
red-001, it's ok |
| 10:51 |
Zeno` |
red-001, after discussion you may have tried to login (accidentally) with one of my linked nicks |
| 10:51 |
Zeno` |
one of them is kinda desirable so I can see how it may have been an accident |
| 10:52 |
Zeno` |
maybe you tried \o` or something as a nick... who knows |
| 10:52 |
red-001 |
I think I tried \o/ a while back |
| 10:54 |
red-001 |
Jun 03 20:10:28 *You are now known as |o| |
| 10:55 |
Zeno` |
yeah, it's probably owned by me (in a way) which may have caused the nickserv message. We'll forget it |
| 10:56 |
red-001 |
yeah ns confirms it's your nick |
| 10:56 |
Zeno` |
ok, mystery solved |
| 10:57 |
DS-minetest |
i don't want to annoy you but i think it would better to continue this conversation in #-hub |
| 10:58 |
kilbith |
you annoy us, insect |
| 10:59 |
|
Fixer joined #minetest-dev |
| 10:59 |
Zeno` |
huh? The conversation is over. |
| 11:00 |
DS-minetest |
yep, i took too long to write >_< |
| 11:02 |
Zeno` |
I have some moral/ethical objections to some things in place on -hub but the best place to discuss them is in -hub |
| 11:03 |
Zeno` |
so I'm stuck in an infinite loop |
| 11:03 |
Zeno` |
(well infinite recursion really) |
| 11:04 |
Raven262 |
http://i0.kym-cdn.com/photos/images/newsfeed/001/246/322/9b2.gif |
| 11:04 |
Raven262 |
That? |
| 11:06 |
|
YuGiOhJCJ joined #minetest-dev |
| 11:17 |
nerzhul |
Thomas-S, why do you use a reference to a hybridPtr, it's crazy |
| 11:17 |
nerzhul |
oh an now we use C++11, #ifdef MYHEADER #define #endif can be removed for new files |
| 11:18 |
nerzhul |
just use #pragma once after the license header |
| 11:18 |
Thomas-S |
Actually, I only use it because Ekdohibs and celeron55 used it too in their meta-set nodedef versions. I mainly only rebased that part. |
| 11:19 |
Thomas-S |
What do you suggest me to use instead of HybridPtr? |
| 11:19 |
nerzhul |
what is the purpose of this hybrid ptr ? |
| 11:20 |
Thomas-S |
Actually I don't really know why this is used. It seems to be the same as the old SharedPtr, but with an additional never_delete field. |
| 11:21 |
nerzhul |
i don't understand really why the nodewithdef has a pointer to contentfeatures |
| 11:21 |
nerzhul |
we have a repository to check that if needed |
| 11:22 |
Thomas-S |
I think in the pointer to contentfeatures the meta-set properties are stored. Maybe nore knows more? |
| 11:23 |
nerzhul |
contentfeatures is a structure defining node static properties, meta should be in another structure |
| 11:24 |
nerzhul |
interesting CSM mods, who_plus is nice :) |
| 11:59 |
|
CalebDavis joined #minetest-dev |
| 12:09 |
|
DS-minetest joined #minetest-dev |
| 12:53 |
|
octacian joined #minetest-dev |
| 13:09 |
|
numzero joined #minetest-dev |
| 13:15 |
|
RobbieF joined #minetest-dev |
| 13:22 |
|
kilbith joined #minetest-dev |
| 13:25 |
|
imr joined #minetest-dev |
| 13:25 |
|
MoNTE48 joined #minetest-dev |
| 13:34 |
|
cx384 joined #minetest-dev |
| 13:35 |
|
juhdanad joined #minetest-dev |
| 13:38 |
|
DI3HARD139 joined #minetest-dev |
| 13:39 |
Thomas-S |
nerzhul: nore explained the purpose of HybridPtr: |
| 13:39 |
Thomas-S |
<nore> Okay so IIRC the idea is that when the ContentFeatures are in the nodedef, you want to free them |
| 13:39 |
Thomas-S |
However, when they are global, you absolutely do not want that |
| 13:41 |
nerzhul |
i see the concept but not hybridptr itself |
| 13:43 |
Thomas-S |
So, you wanna know where HybridPtr is used? For this, please take a look at part 3: #5903 |
| 13:43 |
ShadowBot |
https://github.com/minetest/minetest/issues/5903 -- [WIP] Meta-set nodedef: Part 3 by Thomas--S |
| 13:46 |
|
antims joined #minetest-dev |
| 13:55 |
|
DI3HARD139 joined #minetest-dev |
| 13:55 |
|
Warr1024 joined #minetest-dev |
| 13:58 |
|
cx384 joined #minetest-dev |
| 14:11 |
|
lisac joined #minetest-dev |
| 14:12 |
Thomas-S |
So, nerzhul, is HybridPtr okay for you under these circumstances? |
| 14:13 |
nerzhul |
i didn't get time to review it |
| 14:13 |
nerzhul |
i'm working on continuous daily builds for Windows ATM |
| 14:14 |
Thomas-S |
Ok. Could you please notify me when you have reviewed it then? |
| 14:15 |
nerzhul |
i don't know when because it's a huge PR and i'm working on another parts, but i try to look at this, the main problem i want to solve is the hybrid ptr usage, if we can use STL or not |
| 14:16 |
Thomas-S |
Ok. Thanks! I'll try to fix some problems in part 3 then (and maybe start with part 4) |
| 14:23 |
|
iZacZip joined #minetest-dev |
| 14:32 |
|
MoNTE48 joined #minetest-dev |
| 15:11 |
|
glut32 joined #minetest-dev |
| 15:12 |
glut32 |
hi all |
| 15:12 |
cx384 |
hi glut32 |
| 15:14 |
glut32 |
i would like to compile the latest minetest with mingw32 but I have compiling errors about new c++11 stuffs like "std::thread"... I add -std=c++11 on CMAKE_CXX_FLAGS but I have always the same errors |
| 15:14 |
glut32 |
I added |
| 15:16 |
glut32 |
I have gcc 5.3 (released in 2015) |
| 15:17 |
DS-minetest |
is it difficult to decide what will happen to #5604? :P |
| 15:17 |
ShadowBot |
https://github.com/minetest/minetest/issues/5604 -- lua_api: Correct craft recipe param `type` to `method` by DS-Minetest |
| 15:22 |
|
kilbith joined #minetest-dev |
| 15:27 |
|
fwhcat joined #minetest-dev |
| 15:39 |
kilbith |
there's a lot of lag with sfinv when switching tab |
| 15:40 |
kilbith |
up to 1s or more |
| 15:41 |
kilbith |
my creative inventory got a cancer |
| 15:41 |
|
numzero joined #minetest-dev |
| 15:43 |
|
lisac_ joined #minetest-dev |
| 15:44 |
|
corvus_ joined #minetest-dev |
| 15:51 |
|
corvus_ joined #minetest-dev |
| 15:53 |
|
lisac_ joined #minetest-dev |
| 15:57 |
glut32 |
I would like to know who has compiled the mingw32 version of minetest? |
| 15:57 |
glut32 |
(of minetest 4.16) |
| 15:58 |
Fixer |
i did mingw64 one under linux vm and msys2 |
| 15:58 |
Fixer |
glut32: check your compilers if they are up to date, buildbot.sh script should be newest, libraries too |
| 15:59 |
glut32 |
ok |
| 16:01 |
red-NaN |
oh and disable leveldb |
| 16:01 |
red-NaN |
it breaks |
| 16:02 |
red-NaN |
you need to copy a few extra dll's too |
| 16:02 |
Fixer |
red-NaN: if you have newest libleveldb from kitsunemimi, you don't need to |
| 16:02 |
Fixer |
compiles just fine |
| 16:02 |
red-NaN |
is that diffrent from the one used by the buildbot? |
| 16:03 |
red-NaN |
well I don't really need it |
| 16:03 |
kilbith |
he does not enable leveldb |
| 16:03 |
kilbith |
(hint: he's my job mate) |
| 16:04 |
red-NaN |
and the he is the buildbot? |
| 16:04 |
kilbith |
what? |
| 16:04 |
red-NaN |
I'm really confused right now tbg |
| 16:04 |
red-NaN |
tbh* |
| 16:05 |
kilbith |
I was talking about glut32 |
| 16:05 |
red-NaN |
oh that makes sense |
| 16:26 |
|
Devy joined #minetest-dev |
| 16:27 |
glut32 |
This the output errors : https://pastebin.com/LKGVuyvD I relanch a CMake from zero (from an empty folder) and I added CMAKE_CXX_FLAGS=-std=c++11 |
| 16:27 |
|
Devy left #minetest-dev |
| 16:27 |
glut32 |
this is |
| 16:43 |
|
Krock joined #minetest-dev |
| 16:43 |
|
nerzhul joined #minetest-dev |
| 16:46 |
nerzhul |
Krock, can you review #5917 permitting to merge ? |
| 16:46 |
ShadowBot |
https://github.com/minetest/minetest/issues/5917 -- CSM: Fix documentation error for register_on_*_chat_messages by DS-Minetest |
| 16:46 |
nerzhul |
hi Krock :) |
| 16:53 |
|
CalebDavis joined #minetest-dev |
| 16:53 |
Krock |
yes but first I'll try to find a nice Git client for Linux and after that - cloning |
| 16:54 |
nerzhul |
git |
| 16:54 |
nerzhul |
git clone gitgithub.com:minetest/minetest.git |
| 16:54 |
nerzhul |
:) |
| 17:09 |
|
Raven262 joined #minetest-dev |
| 17:10 |
|
lisac_ joined #minetest-dev |
| 17:33 |
|
lisac_ joined #minetest-dev |
| 17:33 |
|
Raven262 joined #minetest-dev |
| 18:03 |
|
Wuzzy joined #minetest-dev |
| 18:07 |
Wuzzy |
What does dtime_jitter in the debug screen mean? I have 29% atm. is this good or bad? |
| 18:09 |
|
ThomasMonroe joined #minetest-dev |
| 18:17 |
Krock |
jitter is generally bad. No idea why it is measured in percent |
| 18:18 |
Krock |
perhaps the peak over/undershoot of the average dtime? |
| 18:19 |
DS-minetest |
Thomas-S: how does is_ground_content make sense in the list of #5919 ? i mean meta can only be set after generating, right? or is that only for consistence? or for mods? ¿ |
| 18:19 |
ShadowBot |
https://github.com/minetest/minetest/issues/5919 -- [WIP] Meta-set nodedef: Part 4 by Thomas--S |
| 18:20 |
Thomas-S |
Actually, I asked the same question to myself when I wrote down the list. I decided to write it down for now (for consistency, as I listed everything from `read_content_features`). But I think this will never be really implemented. |
| 18:22 |
DS-minetest |
ok |
| 18:24 |
Wuzzy |
Krock: fun fact: This number can get >100% |
| 18:24 |
Wuzzy |
i have no idea what that's supposed to mean |
| 18:48 |
red-NaN |
csm_player_q |
| 18:48 |
red-NaN |
?? |
| 18:48 |
red-NaN |
#5921 |
| 18:48 |
ShadowBot |
https://github.com/minetest/minetest/issues/5921 -- Remove old network code by red-001 |
| 18:52 |
|
paramat joined #minetest-dev |
| 18:58 |
kilbith |
https://lut.im/aQt5ubrNKY/cvt551m8aYS7byx5.png |
| 18:58 |
kilbith |
subtle bloom |
| 19:00 |
kilbith |
https://lut.im/6qrW5JOh5f/hXksGZY6QM5xzKxY.png |
| 19:00 |
kilbith |
you can see better on the clouds |
| 19:07 |
celeron55 |
Wuzzy: it's % of dtime or something like that |
| 19:08 |
Wuzzy |
thats a pretty vague description ;) |
| 19:08 |
celeron55 |
i often say something like 50% is acceptable |
| 19:08 |
Wuzzy |
so 50% dtime_jitter means … what? |
| 19:08 |
celeron55 |
it means something has caused a frame to be 50% longer than the average frame length |
| 19:09 |
Wuzzy |
ah. so MT averages the dtimes out over the past and the jitter is just how “far away†it is now |
| 19:09 |
celeron55 |
no |
| 19:09 |
Wuzzy |
hm |
| 19:10 |
Wuzzy |
Do I want a low jitter? |
| 19:10 |
celeron55 |
jitter means what the highest peak is compared to average |
| 19:10 |
kilbith |
https://github.com/minetest/minetest |
| 19:10 |
kilbith |
wth |
| 19:10 |
celeron55 |
if one frame is 40ms and others are 20ms, the number should show 100% |
| 19:10 |
Wuzzy |
okay |
| 19:11 |
celeron55 |
i don't think the average is any particular time, it's just an avg = avg * 0.99 + current * 0.01 type thing, i.e. just a lowpassed value (probably a few seconds or something) |
| 19:12 |
Wuzzy |
celeron55: the "highest peak"... i guess you mean a peak value over the last few second? |
| 19:12 |
celeron55 |
hmm... i think the collected peak value is just decremented each frame |
| 19:13 |
celeron55 |
it's just a quick way to see how choppy the framerate is |
| 19:13 |
Wuzzy |
so roughly spoken: high jitter = framerate bounces up and down like crazy |
| 19:14 |
|
Hunterz joined #minetest-dev |
| 19:14 |
celeron55 |
not really; more like high jitter = frametime has high peaks |
| 19:14 |
paramat |
github status "We are currently experiencing major service outages." |
| 19:14 |
celeron55 |
high peaks in frametime are the most annoying thing you can get |
| 19:14 |
celeron55 |
that's why it's trying to show them |
| 19:15 |
Wuzzy |
ahhh. o 50% jitter means the peak drawtime in last few second was 50% higher than average (or fake average?) drawtime |
| 19:15 |
celeron55 |
yes |
| 19:15 |
Wuzzy |
okay, i see now why this is useful |
| 19:15 |
celeron55 |
i like this fake peak and fake average terminology |
| 19:15 |
celeron55 |
:D |
| 19:15 |
Wuzzy |
i was almost saying that this value is useless |
| 19:19 |
celeron55 |
and actually not drawtime but frametime or dtime (same thing); it's fine if drawtime has jitter as long as it's masked by sleep or other processing than drawing |
| 19:19 |
celeron55 |
altough usually the minetest client's bottleneck is drawing so usually there's no practical difference |
| 19:20 |
celeron55 |
but you can get frametime jitter from deserialization of large amounts of data in packet processing or whatnot |
| 19:21 |
|
DS-minetest joined #minetest-dev |
| 19:21 |
Wuzzy |
celeron55: can you please take a quick look on <http://wiki.minetest.net/Debug> and tell me if this is correct? |
| 19:22 |
celeron55 |
dtime isn't a shorthand for drawtime |
| 19:23 |
celeron55 |
as far as i can remember |
| 19:23 |
|
cx384 joined #minetest-dev |
| 19:23 |
celeron55 |
the name "dtime" comes from delta time, i.e. time from last frame |
| 19:23 |
celeron55 |
which includes other things than drawing |
| 19:24 |
celeron55 |
i wonder if drawtime in the debug view is actually drawtime... |
| 19:24 |
Wuzzy |
o_O |
| 19:24 |
Wuzzy |
lol |
| 19:25 |
celeron55 |
i need to check |
| 19:25 |
celeron55 |
looks like it's a lowpass filtered drawtime that doesn't include other things in a frame |
| 19:26 |
celeron55 |
this is just totally random |
| 19:26 |
celeron55 |
well that's what it is, better document it 8) |
| 19:27 |
celeron55 |
i did choose them at some point, probably that particular value was the most useful value for some purpose |
| 19:28 |
celeron55 |
if it wasn't lowpass filtered, it would probably jump around so much you couldn't read it so there's that |
| 19:28 |
Wuzzy |
do you mean drawtime or dtime_jtter? |
| 19:28 |
Wuzzy |
i mean with the filtering? |
| 19:28 |
celeron55 |
drawtime |
| 19:28 |
Wuzzy |
ah |
| 19:29 |
celeron55 |
drawtime: an average time it's taking to render each frame, not including other processing than rendering |
| 19:30 |
celeron55 |
dtime_jitter: jitter in the time difference between frames, including all processing |
| 19:32 |
Wuzzy |
ok thanks |
| 19:32 |
celeron55 |
drawtime is kind of useless i think |
| 19:33 |
celeron55 |
could be just removed |
| 19:35 |
Wuzzy |
because we already have FPS displayed? |
| 19:35 |
Wuzzy |
yeah, makes sense |
| 19:36 |
Wuzzy |
Relevant: https://github.com/minetest/minetest/issues/5920 |
| 19:36 |
celeron55 |
well it's not even related to FPS, because FPS is measuring the time difference between frames |
| 19:36 |
celeron55 |
not the time it takes to render them |
| 19:36 |
Fixer |
celeron55: i did tests yesterday and can confirm stuttering was seriously reduced, stand deviation 3.5 against 5-6 before, I can play without that mesh gen delay now, reduced it to just 5 ms for ideal, there is still some stuttering, including rare large ones of unknown origin (up to 100ms) |
| 19:36 |
celeron55 |
nobody cares about the drawtime because the client can't run without doing other things in addition to drawing |
| 19:38 |
celeron55 |
Wuzzy: i think your description of what jitter means was useful |
| 19:52 |
VanessaE |
celeron55: fake peak, fake average? fake news! :P |
| 19:53 |
VanessaE |
(or would those be "alternative facts"? ;) ) |
| 19:53 |
celeron55 |
alternative average |
| 19:53 |
VanessaE |
haha |
| 19:55 |
Fixer |
safe space average |
| 19:55 |
red-NaN |
you may laugh now, but soon even averages wouldn't be safe |
| 19:56 |
Fixer |
i would like to see also median and stand deviation for frame times |
| 19:56 |
Fixer |
on other hand median needs all data :/ |
| 19:56 |
Fixer |
bad idea |
| 19:59 |
VanessaE |
Fixer: have you tried evaluating stutter/jitter/etc by using VE-Survival yet? |
| 19:59 |
Fixer |
VanessaE: yes, it felt better |
| 19:59 |
VanessaE |
ok. |
| 20:00 |
Fixer |
number i got were from automated fly over tests |
| 20:00 |
Fixer |
singleplayer is playable more or less, still stutter here and there but much less than before (before it was disaster) |
| 20:03 |
Fixer |
while subjectively I had impression that msys2/mingw build was much smoother, numbers did not support me, both build got same distribution of frame more or less, msys2 got slightly higher fps by little |
| 20:03 |
Fixer |
one annoying stutter is loading lots of signs |
| 20:03 |
paramat |
VanessaE shivajiva as server admin i would be interested to see your comments in #5915 |
| 20:03 |
ShadowBot |
https://github.com/minetest/minetest/issues/5915 -- CSM: Allowing a server to protect itself against clients running client-provided clientmods |
| 20:04 |
VanessaE |
Fixer: ...which boils down to either loading entities being slow to load in general, or time spent executing all the compositing operations that go into making the sign text. |
| 20:04 |
VanessaE |
paramat: in a word? absofuckinglutely. |
| 20:04 |
VanessaE |
'nuff said? |
| 20:05 |
paramat |
ok thanks please comment |
| 20:05 |
VanessaE |
done. |
| 20:05 |
Fixer |
VanessaE: that huge freeze of signs loading was confirmed with other server and mod: display_modpack or smth like that (with neat and fancy signs), same large freezes when loading those |
| 20:06 |
VanessaE |
Fixer: but is it because of the *entities* or is it because of the compositing operations used for the text? |
| 20:06 |
Fixer |
no idea |
| 20:06 |
Fixer |
i hope someone will be interested enough into looking at it |
| 20:09 |
paramat |
#5908 is ready for review |
| 20:09 |
ShadowBot |
https://github.com/minetest/minetest/issues/5908 -- (Re)spawn players within mapgen edges by paramat |
| 20:09 |
VanessaE |
paramat: maybe MT needs to use punkbuster ;) |
| 20:10 |
VanessaE |
hm, GH is starting to lag again. |
| 20:10 |
VanessaE |
probably some little baby DDoSing it |
| 20:16 |
nerzhul |
sfan5, when you get some time can you check/approve #5923 please ? |
| 20:16 |
ShadowBot |
https://github.com/minetest/minetest/issues/5923 -- Implement GItlab CI daily builds for windows platform (32 & 64) by nerzhul |
| 20:16 |
sfan5 |
sure |
| 20:18 |
nerzhul |
we have #5921 (trivial) which requires a second approval too |
| 20:18 |
ShadowBot |
https://github.com/minetest/minetest/issues/5921 -- Remove old network code by red-001 |
| 20:19 |
red-NaN |
#5924 |
| 20:19 |
ShadowBot |
https://github.com/minetest/minetest/issues/5924 -- Have the server send the player list to the client by red-001 |
| 20:19 |
red-NaN |
a less trival PR |
| 20:19 |
nerzhul |
red-NaN, i prever UpdatePlayerList |
| 20:19 |
nerzhul |
prefer* |
| 20:20 |
Fixer |
nerzhul: no gettext support? |
| 20:20 |
red-NaN |
alright I will change that then |
| 20:21 |
Fixer |
nerzhul: dled 64bit artifact or how it is called, it runs, but there are no translations to other languages |
| 20:21 |
Fixer |
nerzhul: says use_gettext=0 |
| 20:21 |
nerzhul |
Fixer, i don't know how does that work, i just ran buildbot |
| 20:21 |
nerzhul |
maybe sfan5 can help me to know if something is missing to compilation to improve the build |
| 20:22 |
nerzhul |
now i have the right process to generate artifacts (and you confirm it works :p) we can modulate |
| 20:24 |
sfan5 |
pretty sure you're doing the exact same thing i do for my windows builds |
| 20:25 |
nerzhul |
sfan5, then you don't use gettext ? i don't think there is a detection ? |
| 20:25 |
Fixer |
lol |
| 20:25 |
sfan5 |
the buildbot will automatically use gettext |
| 20:25 |
nerzhul |
okay maybe it's because i don't installed gettext |
| 20:25 |
Fixer |
nerzhul: maybe it needs to rerun cmake |
| 20:26 |
nerzhul |
i will re-verify |
| 20:26 |
sfan5 |
dunno what happens if windres isn't installed |
| 20:26 |
Fixer |
nerzhul: funny how you solved that 3 dll problem, i will use the same method :D |
| 20:26 |
sfan5 |
uh i mean msgfmt |
| 20:26 |
sfan5 |
nerzhul: https://github.com/minetest/minetest/blob/master/util/buildbot/buildwin32.sh#L133 |
| 20:26 |
nerzhul |
Fixer, i don't see better automated method heh |
| 20:27 |
nerzhul |
okay i add gettext to ensure it fix it |
| 20:27 |
nerzhul |
because we have explicit ENABLE_GETTEXT=1 |
| 20:27 |
Fixer |
32 bit version also works fine, but has no gettext |
| 20:27 |
sfan5 |
nerzhul: #5922 is pretty pointless as it will use c++11 mutextes if available |
| 20:27 |
ShadowBot |
https://github.com/minetest/minetest/issues/5922 -- Use C++11 mutexes by nerzhul |
| 20:28 |
nerzhul |
sfan5, it just drop the compat mode here |
| 20:28 |
sfan5 |
well then the pr description is misleading |
| 20:28 |
nerzhul |
yeah |
| 20:28 |
nerzhul |
seems windows build doesn't like it |
| 20:28 |
sfan5 |
you should also be able to drop the other threading compat code |
| 20:28 |
nerzhul |
sfan5, i will do it in anotehr pr |
| 20:28 |
nerzhul |
i split my PR |
| 20:28 |
nerzhul |
#5906 is mergeable |
| 20:28 |
ShadowBot |
https://github.com/minetest/minetest/issues/5906 -- C++11 patchset 3: remove Atomic/GenericAtomic and use std::atomic by nerzhul |
| 20:30 |
nerzhul |
for windows pipeline: "-- Could NOT find GetText (missing: GETTEXT_MSGFMT) " |
| 20:30 |
nerzhul |
it's exactly the problem, msgfmt is in which package, hmmm |
| 20:30 |
Fixer |
nerzhul: _build\_CPack_Packages\win64\ZIP\ has exactly same files before compression, you don't need to decompress that zip, just copy dlls to there and make new ZIP overwriting old one |
| 20:30 |
Fixer |
nerzhul: imo |
| 20:31 |
nerzhul |
Fixer, i just did it to remove the double zip |
| 20:31 |
nerzhul |
(gitlab zip & buildbot zip) |
| 20:33 |
nerzhul |
sfan5, i prefered to do a PR for each C++11 component to use to reduce the diff size to review, else it will be very very huge and less clear to read |
| 20:36 |
nerzhul |
Fixer, https://gitlab.com/nerzhul/minetest/builds/17939579 |
| 20:36 |
nerzhul |
gettext is now enabled :) |
| 20:37 |
nerzhul |
wait for packaging step to finish and can re-test it |
| 20:37 |
nerzhul |
sfan5, if you get time for #5906 could be nice heh, the description is better and build is already okay :p |
| 20:37 |
ShadowBot |
https://github.com/minetest/minetest/issues/5906 -- C++11 patchset 3: remove Atomic/GenericAtomic and use std::atomic by nerzhul |
| 20:44 |
|
numzero joined #minetest-dev |
| 20:44 |
Fixer |
nerzhul: and someone should put a good visible link on minetest.net so people can DL |
| 20:45 |
nerzhul |
when it will be available and accepted by everybody in our dev council yes :) |
| 20:45 |
nerzhul |
s/everybody/majority/ |
| 20:46 |
nerzhul |
but gitlab-CI is a great meaning permitting to generate automaticly official builds, it's synced from github every day, that trigger a build pipeline which generate artifacts and we can have as many linux distro as we want, i did for debian/ubuntu but you can also build on fedora, arch if needed and generate packages |
| 20:47 |
Fixer |
nerzhul: add appimage into it and that will be amazing |
| 20:48 |
nerzhul |
appimage or a docker build can be done in some specific steps yes, but i don't know how to do it and i don't get time for this now, if somebody wants after we have this mode (if accepted) as official daily builds we can review/integrate it easily |
| 20:49 |
|
kilbith_ joined #minetest-dev |
| 20:56 |
nerzhul |
Fixer, can you test https://gitlab.com/nerzhul/minetest/pipelines/8786331 win32 & win64 packages to confirm translations are okay ? |
| 20:56 |
Fixer |
nerzhul: sure |
| 20:57 |
Fixer |
googled the way on how to add those libs into existing zip :} |
| 20:57 |
red-NaN |
nerzhul, run_in_place? |
| 20:57 |
nerzhul |
Fixer, oh in existing zip, jut call zip <zipfile> <pathstoadd> |
| 20:58 |
Wuzzy |
I saw 0.4.16 brought texturable progress bar. But how do I change it for my subgame? |
| 20:58 |
nerzhul |
red-NaN, seems not enabled by buildbot |
| 20:58 |
nerzhul |
Wuzzy, you cannot |
| 20:58 |
Wuzzy |
whyyyy |
| 20:58 |
nerzhul |
progress bar is in texture packs |
| 20:58 |
Wuzzy |
:-( |
| 20:58 |
nerzhul |
just be logical |
| 20:59 |
nerzhul |
when you connect to a server progress bar is shown and you don't know and didn't loaded any media |
| 20:59 |
Wuzzy |
okay, make sense. XD |
| 21:00 |
Wuzzy |
but what about singleplayer? |
| 21:00 |
DS-minetest |
but the logo of the subgame is already loaden |
| 21:00 |
Wuzzy |
exactly |
| 21:00 |
nerzhul |
Wuzzy, all games are using the same way |
| 21:00 |
nerzhul |
all modes |
| 21:00 |
nerzhul |
singleplayer is just a name and a config on server but you load a local server and conenct to it |
| 21:01 |
Fixer |
Using Irrlicht 1.8.4 |
| 21:01 |
Fixer |
Build info: VER=0.4.16-dfd79c0-dfd79c0-dirty BUILD_TYPE=Release RUN_IN_PLACE=1 USE_GETTEXT=1 USE_SOUND=1 USE_CURL=1 USE_FREETYPE=1 USE_LUAJIT=1 STATIC_SHAREDIR="." |
| 21:01 |
nerzhul |
the logo is shown to main menu but game loader doesn't know anything about game, currently it just connect to any random server, local or not |
| 21:01 |
red-NaN |
would it be possible to make it also build non RUN_IN_PLAVE builds for windows? |
| 21:01 |
nerzhul |
Fixer, oh it seems implicit |
| 21:01 |
Wuzzy |
in other words, loading screen is very hardcoded atm? |
| 21:01 |
nerzhul |
red-NaN, what is the use case ? |
| 21:01 |
Wuzzy |
anyway, no big deal |
| 21:01 |
Wuzzy |
the default loading bar is fine. |
| 21:01 |
nerzhul |
Wuzzy, yes, it cannot predict what he didn't load :p |
| 21:01 |
Wuzzy |
i wish the rest of the GUI would match this style as well :) |
| 21:01 |
red-NaN |
I'm currectly using non-run-in-place builds for itch.io builds |
| 21:02 |
Wuzzy |
currently it just looks plain random |
| 21:02 |
Fixer |
nerzhul: can confirm, both are working, have translations working |
| 21:02 |
Fixer |
minetest_game is present |
| 21:02 |
Fixer |
game runs |
| 21:02 |
nerzhul |
Fixer, nice, it works as intended then |
| 21:02 |
Fixer |
thumbs up |
| 21:03 |
Wuzzy |
yes, i have German runnng here. runs perfectly |
| 21:03 |
nerzhul |
red-NaN, i see the use case, maybe waiting for this PR to be merged and we can branch a specific pipeline for itch.io |
| 21:03 |
nerzhul |
then sfan5 please help us to merge that and win some daily time to not build by hand :D |
| 21:04 |
nerzhul |
itch.io uses win32 or 64 ? or multiple OS ? |
| 21:05 |
red-NaN |
multiple OS's and archs |
| 21:06 |
red-NaN |
currectly I have 64bit linux appimages and 64 and 32 bit windows builds |
| 21:06 |
red-NaN |
currently* |
| 21:07 |
nerzhul |
okay |
| 21:08 |
nerzhul |
do you think we can create a Linux appimage easily in the Gitlab CI ? |
| 21:08 |
nerzhul |
i don't know if it's easy or not |
| 21:08 |
red-NaN |
not sure I'm using Calinou's builds |
| 21:08 |
nerzhul |
okay, and on which Linux OS is this based ? Ubuntu ? |
| 21:09 |
Calinou |
nerzhul: Ubuntu 14.04 preferably, Ubuntu 16.04 if you are lazy |
| 21:09 |
Calinou |
you should build on an old distro |
| 21:09 |
nerzhul |
maybe pipeline can use artifacts from Zesty and appimage consume it too to create it, in parallel of deb package |
| 21:09 |
Calinou |
build script is smething like that: https://gist.github.com/Calinou/a326b153ab4f21853e8741f11b0aa9db |
| 21:09 |
nerzhul |
Calinou, why ? if you embed stdc++ lib you don't care no ? |
| 21:09 |
Calinou |
if you use a recent distro, it won't work on old distros; please don't |
| 21:09 |
Calinou |
no, you can't embed glibc in AppImage |
| 21:09 |
Calinou |
that would make it much larger, too |
| 21:09 |
Calinou |
join #appimage for details |
| 21:09 |
nerzhul |
then AppImage is just.. :( |
| 21:09 |
Calinou |
(or I can invite probono there) |
| 21:10 |
nerzhul |
it's the major portability problem in C++ the libc++ |
| 21:10 |
Calinou |
well, musl did not overtake the world yet :) |
| 21:10 |
nerzhul |
GNU stdlib has poor quality compared to BSD stdlib |
| 21:11 |
nerzhul |
if appimage is just a script like this, it's easy to embed to our util/ directory and then call it in CI |
| 21:11 |
nerzhul |
we can use trusty artifacts then and have 2 pipeline branch, one for deb one for appimage |
| 21:11 |
Calinou |
yeah, you can also use yml-based scripts |
| 21:11 |
Calinou |
and derive your AppImage from a .deb |
| 21:11 |
Calinou |
perhaps use the xenial .deb... |
| 21:12 |
nerzhul |
hey but you copy some system libs, don't you try stdc++ ? :p |
| 21:12 |
Calinou |
here is the list of libraries AppImage will exclude by default: https://github.com/AppImage/AppImages/blob/master/excludelist |
| 21:12 |
nerzhul |
ouch the wget for appimage helper |
| 21:12 |
Calinou |
yeah, it requires a working connection |
| 21:12 |
nerzhul |
i can't remember appimage is redhat or canonical ? |
| 21:12 |
Calinou |
you could compile it from source, but that's not exactly easy |
| 21:12 |
Calinou |
none |
| 21:12 |
Calinou |
it's an independent project |
| 21:13 |
Calinou |
Flatpak would be Red Hat(-ish), Snap would be Canonical |
| 21:13 |
nerzhul |
oh yeah :) |
| 21:14 |
Fixer |
my way of dealing with pesky dlls https://pastebin.com/raw/8eJ3yhyB |
| 21:15 |
nerzhul |
Fixer, yeah, i just extract for removing zip into zip :p |
| 21:15 |
nerzhul |
proper user experience :p |
| 21:21 |
Fixer |
Feels Good Man |
| 21:22 |
nerzhul |
less and less coredevs present on this channel these days, 0.4.16 has tired many devs :p |
| 21:22 |
|
RobbieF left #minetest-dev |
| 21:23 |
Calinou |
it's impressive how AppImage is an one-man project existing since ~2004 (in other forms), yet has so much success, many commercial ISVs like JetBrains use it already |
| 21:23 |
Calinou |
I really admire it, fixing a big issue in the Linux world |
| 21:23 |
Calinou |
(but it does get a *lot* of opposition) |
| 21:23 |
nerzhul |
it's good for portable crap :p |
| 21:23 |
Calinou |
Debian might package appimagetool in their repositories, which will be extremely ironic to see if that happens |
| 21:23 |
nerzhul |
package manager vs windows outdated dlls |
| 21:24 |
nerzhul |
chocolatey on windows is nice (but slow as a hell) |
| 21:24 |
|
MoNTE48_1 joined #minetest-dev |
| 21:25 |
Fixer |
Calinou: +1 |
| 21:31 |
Calinou |
Chocolatey is meh, I've found |
| 21:32 |
nerzhul |
if that continue we will have mor than 170 PR tomorrow :( |
| 21:32 |
nerzhul |
ShadowBot, sfan5 nore help :p |
| 21:33 |
|
paramat joined #minetest-dev |
| 21:36 |
Fixer |
It is all in your head (c) |
| 22:05 |
|
Fixer joined #minetest-dev |
| 22:20 |
|
Devy joined #minetest-dev |
| 22:30 |
|
Player_2 joined #minetest-dev |
| 22:37 |
VanessaE |
interesting point of note: 1000m view range on VE-S when way out on the west end of West St, looking east (so all of it is loaded). not a wide swath of it, just the region around the road I ran down. still, managed 16 fps just then |
| 22:38 |
VanessaE |
celeron55: any chance of revisiting farmap? might be nice to see the "sides" of that range filled in :) |
| 22:39 |
VanessaE |
http://i.imgur.com/AZzGxR7.png |
| 22:40 |
VanessaE |
(and as you can see, that's with HDX-256, no less) |
| 23:08 |
paramat |
wow |
| 23:11 |
paramat |
#5926 |
| 23:11 |
ShadowBot |
https://github.com/minetest/minetest/issues/5926 -- Autorun: Change to 'automove' to be suitable and consistent by paramat |
| 23:11 |
VanessaE |
paramat: you like? :) |
| 23:11 |
VanessaE |
that image is what happens when the GPU gets to do its job unimpeded ;) |
| 23:14 |
Fixer |
VanessaE: m? |
| 23:14 |
VanessaE |
Fixer: 1000m |
| 23:14 |
Fixer |
VanessaE: it is just small patch of land |
| 23:16 |
VanessaE |
narrow, yes |
| 23:16 |
VanessaE |
small? no |
| 23:16 |
VanessaE |
that's a full 1000m into the distance |
| 23:16 |
VanessaE |
(actually it goes a bit past 1200 but I like a nice round number) |
| 23:45 |
|
MoNTE48_1 joined #minetest-dev |
| 23:58 |
paramat |
nice road |
