Luanti logo

IRC log for #minetest, 2016-09-09

| Channels | #minetest index | Today | | Google Search | Plaintext

All times shown according to UTC.

Enable nick filtering
Time Nick Message
24 more elements. Show/hide.
04:08 Chinchou My, my.  https://wiki.allegro.cc/index.php?title=Header_file
04:09 Chinchou Explained and understood in five minutes, while a year-and-a-half class not only failed but /gave the wrong ideas/.
2 more elements. Show/hide.
04:12 Hawk777 Only thing that would make that article better is if it explained the reason for including a.h in a.c (to avoid bugs caused by accidentally defining the function in a way that mismatches its declaration).
04:12 Hawk777 The example doesn’t do that.
04:13 swift110-phone hey
04:15 Chinchou Hmm...  Ah, right.
04:15 Chinchou Took me a moment to understand what you meant.
04:16 Chinchou I'm not sure if that'd cause a compiler error, a segfault, or just weird behaviour during run...
04:16 Chinchou Now I'm curious.  I'm gonna go try that.
04:17 LionLAD joined #minetest
04:18 swift110-phone hmm
04:28 Hawk777 Unknown. Depending on the nature of the mismatch it could do any of the above (actually in C++ a parameter type mismatch would cause a linker error because overloading, but this is still true of a return type mismatch).
04:29 Hawk777 Likely not a compiler error as it wouldn’t know. Segfaults and unexpected behaviour are both possible.
04:30 Chinchou That was my guess.  I tried it in C just now, and it did indeed compile fine...
04:30 Chinchou Somehow, that's a bit amusing to me.
04:31 Chinchou ... Oddly, upon running it, I didn't get any errors...  But my test had the real function taking nothing and the declaration taking an int.
04:31 Hawk777 The compiler has type information but sees one translation unit at a time, each of which are internally consistent. The linker sees both object files but has no clue about type information so says nothing, as all it cares about is that the name is right.
04:31 Chinchou If I did it the other way, the test function would definitely cause an error upon trying to use that data.
04:32 Chinchou Aye.
04:32 Hawk777 Yes, in many cases nothing bad will happen. A function taking a single int will have it passed in a register on most architectures, so your call would just pass the value in a register that the target function doesn’t look at.
04:32 Hawk777 Doing it the other way would likely have a random value there, whatever was left in that register from prior calculations.
04:33 Chinchou That makes sense, aye
04:33 Chinchou Wouldn't that be looking at something out of the programme's allocated memory, though?
04:33 Chinchou I'm not too clear on how addressing works.
04:34 Hawk777 Nope. Most simple integer parameters are passed in CPU registers. CPU registers aren’t part of memory; they always exist.
04:34 Hawk777 Even if the parameter were passed on the stack, most likely what would happen is the function would look at a part of the stack memory that is indeed allocated, just to some other purpose in a further-out function.
04:35 Chinchou Hmm...
04:35 Hawk777 Not likely that it would be outside memory altogether; memory tends to be handed to programs in fairly large blocks, so moving just a few bytes one way or the other from something you’re supposed to be looking at usually doesn’t segfault.
04:35 Chinchou Ah, so you'd have to do a pretty big offset to actually be outside the programme's memory?
04:37 Hawk777 Right.
04:37 Chinchou I'd been wondering about this for a while, glad to have it finally make sense.
04:38 Hawk777 I don’t know what OS you’re on. If Linux, you can go cat /proc/<somepid>/maps to see which pieces of memory are allocated to the process with process ID <somepid>, along with their permissions and some other bits and pieces.
04:38 Chinchou Debian Jessie, yeah.
04:38 Hawk777 You will likely see nothing near zero, and mainly large chunks.
04:39 Hawk777 I tried it with a bash process and I see a few blocks for /bin/bash itself (i.e. mmaps of the file /bin/bash), one for [heap] (where malloc gets small bits of memory from), one for [stack] (where locals, return addresses, etc. live), and a pile of others for shared libraries, unknown stuff, and [vvar] and [vdso] which are to do with kernel interfacing.
04:41 Chinchou Oh, now that's interesting.  I see far more maps for my IRC client than I'd have expected.
04:41 Hawk777 Mostly shared libraries I would guess?
04:41 Chinchou Aye.
04:42 Chinchou A bunch of entries for .cache-4 files...
04:42 Hawk777 If it’s a graphical IRC client, you will be pulling in all of whatever windowing toolkit it’s written in, Xorg libraries, video drivers, etc..
04:42 Chinchou Aye, that makes sense.
04:42 Hawk777 Ah, I see a couple of font files in mine. Yes, anything that is brought in with mmap() will show up there.
04:42 Chinchou So it'd pull in GTK stuff, fonts...
04:43 Chinchou Er... not GTK, is it?  Pango, cairo, gdk...
04:43 Chinchou The mind, it boggles.
04:43 Hawk777 GTK would be the higher-level windowing toolkit built on top of glib, gdk, pango, and cairo.
04:44 Chinchou Oh, so it is GTK.
04:44 Hawk777 So very likely yes
04:44 Chinchou That clears up some confusion I had about how those were related.  Nice.
04:45 Chinchou This is why I enjoy talking with those more competent than I - I learn a lot.
04:46 Chinchou Hmm...  Apparently it even pulled in libxcursor stuff.
04:46 Hawk777 Any X11 program is big and complicated.
04:48 Chinchou I think I'm starting to grasp that now...  It mostly gets hidden even from the developer, huh.
04:48 Chinchou I'm trying to catch up with Allegro 5(I used 4 previously), and here I was thinking that it was already complicated enough.
04:51 Hawk777 I think I once looked at Allegro many, many years ago when it was a DOS game development library.
04:51 DMackey joined #minetest
05:03 MinetestBot [git] SmallJoker -> minetest/minetest: Anticheat: Use the physics_override modifiers aswell c0cd7aa https://git.io/vi4nY (2016-09-09T01:03:36-04:00)
17 more elements. Show/hide.
11:14 Megaf Hi all
11:14 Megaf !server Megaf
11:14 MinetestBot Megaf: Megaf Server v4.0 | mt.megaf.info:30003 | Clients: 0/20, 0/1 | Version: 0.4.14-Megaf / MegafXploreNext | Ping: 6ms
12 more elements. Show/hide.
13:55 Megaf Channel is very quiet today
13:55 aix yeah
13:56 Megaf Well, I temproraly solved the disk space issues on my server
13:56 Megaf !server Megaf
13:56 MinetestBot Megaf: Megaf Server v4.0 | mt.megaf.info:30003 | Clients: 0/20, 0/1 | Version: 0.4.14-Megaf / MegafXploreNext | Ping: 6ms
17 more elements. Show/hide.
15:44 everamzah PvP Areas mod: https://forum.minetest.net/viewtopic.php?f=9&amp;t=15480&amp;p=232766
15:48 MinetestBot [git] everamzah -> minetest/minetest: Return nil on empty get_area() (#4508) 403dada https://git.io/viB8f (2016-09-10T01:47:13+10:00)
15:51 engineer-pearl joined #minetest
15:54 Megaf aix: Hows your server doing?
15:55 Megaf !server aix
15:55 MinetestBot Megaf: No results
5 more elements. Show/hide.
16:35 aix !server Balance
16:35 MinetestBot aix: Lightning (Balance) | tau.aix.ovh | Clients: 1/32, 0/4 | Version: 0.4.14-dev / balance | Ping: 7ms
16:35 aix Megaf: ^
14 more elements. Show/hide.
17:48 MinetestBot Krock: Sep-09 12:27 UTC <Jordach> meow
17:48 Krock double meow?
17:48 Krock !tell Jordach eww. if it only was LE:ET time
17:48 MinetestBot Krock: I'll pass that on when Jordach is around
17:49 Krock <3 MinetestBot
17:49 MinetestBot <3 Krock
4 more elements. Show/hide.
18:22 agaran Hello
18:22 Krock hi agaran
18:23 agaran how goes?
18:23 Krock goes fine.
18:23 Krock my feet are a bit hurt but apart from that, fine :)
18:23 Krock how about you?
18:23 agaran tired after work but still alive
18:23 Krock good to hear :P
3 more elements. Show/hide.
18:32 agaran Krock: building anything or busy with other stuff?
18:32 Krock busy, sadly.
18:32 proller joined #minetest
18:33 Krock had two hard days now and am not really up to build anything. Let's see how it goes tomorrow
18:33 agaran yup, sure I get feeling
3 more elements. Show/hide.
19:07 whitephoenix https://vid.me/tIDQ modding at its finest
19:17 nore joined #minetest
19:20 Krock whitephoenix, very good. Pressing the play button does nothing
19:20 Krock doesn't even highlight correctly on hover
19:23 Krock the embed site doesn't work too, same problem
19:23 Krock but at least I can see that you died in fire
5 more elements. Show/hide.
20:38 YvesLevier I got a "Control logic unit" in my inventory.  Cant put anywhere to test it and didnt find any help in forums.  Someone knows?
20:39 YvesLevier Also, the craft_guide mod didnt change my craft window at all.  Cannot see how to use it.
8 more elements. Show/hide.
21:19 agaran YvesLevier: control logic unit might be from homedecor, used to make other things.. it is just crafting resource
21:19 agaran YvesLevier: craft guide.. I was placing it on wall then rightclick
21:20 LNJ2GO left #minetest
21:20 YvesLevier agaran: thx alot :)
21:20 agaran yw
21:20 agaran good night
21:23 Gael-de-Sailly joined #minetest
21:30 Calinou !server Calinou
21:30 MinetestBot Calinou: Calinou [Europe/Paris] (survival - minimal mods) | hugo.pro:30002 | Clients: 2/32, 3/5 | Version: 0.4.14-dev / minetest | Ping: 12ms
21:58 Hijiri CLU can be used to upgrade MV/HV technic machines with pipe output
10 more elements. Show/hide.

| Channels | #minetest index | Today | | Google Search | Plaintext