Time Nick Message 01:46 VanessaE Is there any practical reason why the default map save interval is only a few seconds, versus say 1 to 5 minutes? 01:46 VanessaE other than the increased risk of losing map data during a catastrophic crash I mean 01:46 ShadowNinja ^ 01:47 ShadowNinja The data-loss is the biggest(only?) reason. 01:47 VanessaE wouldn't it make sense for the server to do a last-minute save on the way down? 01:47 VanessaE (obviously not possible in the event of a segfault or similar) 01:50 Exio4 what about it 30 seconds, or some "middle value"? 5 minutes seems high to a software like MT that is crashy :P 01:51 Exio4 i don't know what is the actual value anyway 02:16 kaeza Exio4, https://github.com/minetest/minetest/blob/master/src/defaultsettings.cpp#L193 06:12 celeron55 VanessaE: the more rare the interval is, the more big of a stop the server must make to write all accumulated stuff 06:12 celeron55 also, not all changed things are written at the interval 06:12 celeron55 only important ones 06:12 celeron55 or at least, that's how i originally made it to work... could have changed completely already 8) 06:14 celeron55 the interval should be low enough to allow the changes to tricke to the disk in small steps without causing any noticeable stops in server responses 06:14 celeron55 trickle* 06:16 celeron55 or performance otherwise 09:39 VanessaE celeron55: that makes sense, but as said in my case it's saving large quantities every several seconds - not enough to impact overall performance, but enough to register as a pretty large, repeating spike in CPU and disk activity. 09:39 VanessaE I suggest there may simply be more being saved than is necessary. 11:03 jojoa1997 With compiling in Linux do I just have to do the build32.sh thingy in buildbot? 11:07 jojoa1997 Sorry wrong channel 11:40 celeron55 VanessaE: possible 11:40 celeron55 VanessaE: at some log level some details about what is saved and why is logge 11:40 celeron55 +d 15:11 VanessaE https://github.com/minetest/minetest/pull/943 15:11 VanessaE I've tried this, it works a treat. 15:12 Exio4 from 32 to 256? 15:14 ShadowNinja Yep, 128 seemed a bit low. 65000 was fun too. ;-) 15:15 ShadowNinja It could even be pushed to 512, but that would probably look like too much for fragile nodes like leaves. 16:17 celeron55 ideally it would be different for different nodes to get the right feel of mass 16:20 VanessaE indeed so, but this does look nice at least 16:22 celeron55 also there should be a burst of particles when the node gets actually removed, so that the volume of the node kind of crumbles down 16:23 VanessaE agreed 16:23 VanessaE sfan5's particles mod used to do that (with entities) 16:23 VanessaE if by "burst" you mean like a small explosion 16:24 celeron55 well anything in where the geometric cube turns into a volume of particles which then somehow end up not existing anymore 16:25 celeron55 the rest is artistic stuff 16:25 celeron55 8) 16:26 celeron55 (some people won't like that though, then it's worth a setting) 16:26 celeron55 +if 16:26 Exio4 like terasology? 16:30 celeron55 terasology's effect is terrible 16:32 VanessaE celeron55: http://www.youtube.com/watch?feature=player_detailpage&v=-QvtHw1i-EY#t=23 16:33 VanessaE that's sfan5's particles, seom a while back 16:33 VanessaE s/seom/from/ 16:33 ShadowNinja That is what my pull does, there are a bunch of particles when the node breaks, but they don't form in a cube, more like a circle. 16:34 ShadowNinja I am not terribly familiar with the way nodes are managed in C++ but I beleive there are solideness and visual_solidness variables. 16:39 celeron55 VanessaE: well, that's kind of hyper-lame too; probably would need some experimenting 16:43 VanessaE celeron55: heh, well I didn't write the mod :P 16:55 celeron55 http://i.imgur.com/xfxAxE2.gif <- that kind of smoothness from cube to particles one should aim for (might be completely unplayable with MC-like digging though; nobody can know without trying) 16:56 VanessaE that's a bit slow but I see what you're getting at 17:31 VanessaE is there any reason why the F5 debug cuts off at 800 pixels? 17:31 VanessaE e.g. doesn't follow the width of the window 17:42 PilzAdam increasing particles has a huge effect on FPS 17:44 PilzAdam celeron55, I want to play soilnar, now! ;-) 17:46 celeron55 PilzAdam: impossibru, i'm reworking the whole thing to be very different in gameplay 17:48 VanessaE PilzAdam: the dig animation has a far greater FPS impact than some momentary particles. 17:51 ShadowNinja Hmmm, 65000 locked it up, but 256 didn't seem to have much of an effect. Maybe the particles have to be optimized. 18:57 VanessaE 15:01:25: ERROR[main]: ERROR: An unhandled exception occurred: ServerError: LuaE 18:57 VanessaE rror: error: ...st/.minetest/games/vanessae_game/mods/default/player.lua:125: attempt to index local 'anim' (a nil value) 19:03 VanessaE just happened again. 19:04 VanessaE I have noticed lately that all players seem to be animating at "sneak" speeds despite their actual velocity 19:05 PilzAdam is it reproduceable? 19:05 VanessaE yep 19:05 VanessaE it just now happened a third time 19:05 VanessaE same error/line number every time. 19:05 PilzAdam after line 35 of player.lua add else and return {} with all the values above = 0 19:06 PilzAdam do you have any mods that modify the players model? 19:06 VanessaE only one, wield3d, and it hasn't seen an update since July. 19:07 VanessaE [10-02 15:12] I got many errors in terminal when playing about missing armatures 19:08 VanessaE does that mean anything? 19:12 VanessaE PilzAdam: added the lines. 19:12 VanessaE (and restarted the server)... that function now looks like: http://pastebin.com/4sCQyvDE 19:16 VanessaE ok, that triggers a bug in technic now 19:16 VanessaE or it may be entirely unrelated. 19:17 * VanessaE waits for results.. 19:26 VanessaE that seems to have gotten rid of the player.lua-related crash. 19:26 VanessaE (the other one is also stopped, as I removed the mod causing it) 19:58 VanessaE PilzAdam: just fyi, I removed the wield3d mod. Without your suggested change, the crash still happens. So your change is definitely needed. 20:02 Sokomine i do have a problem with my circular saw: if i take something out of it and place it on a stack in my inv that's already occupied with something else, it does get wired 20:03 Sokomine trouble is: the saw does not accept input in the fields the output is taken from. that's nothing special so far - many inventory slots may work the same way 20:06 PilzAdam Sokomine, isnt that a mod problem? 20:08 VanessaE another one: 16:12:39: ERROR[main]: ERROR: An unhandled exception occurred: ServerError: LuaError: error: /usr/share/minetest/builtin/vector.lua:44: attempt to index local 'v' (a nil value) 20:09 Sokomine hmmm. if you can tell me a way to catch this case? 20:09 PilzAdam VanessaE, thats a mod problem 20:09 VanessaE perhaps, but which mod? 20:12 VanessaE #923 is needed. 20:12 Sokomine ok, to put it shortly: take something out of an inventory slot of a block on the map, place that on an inventory slot in your inventory that is not empty. what happens now is that the old item stack in the inventory apparently sticks to the mouse pointer in order to be moved somewhere else. so far so good. but the place that item stack is stored is the output slot the other item was taken from 20:13 Sokomine ...which is a problem if that slot is not intended to take input 20:14 Sokomine is there a way for a mod to catch if something taken out of an inventory will end up on a slot that's occupied and then deny the moveß 20:18 PilzAdam no 20:20 Sokomine hmpf 20:20 Sokomine so i'd claim it's a bug and need fixing since it does lead to item loss 20:21 Sokomine expected behaviour: if the inventory the stack that replaces the one in your inventory was taken out of does *not* accept the stack from your inventory, then deny themove alltogether 20:21 thexyz https://github.com/minetest/minetest/pull/923 we just need debug/release modes for Lua mods 20:21 Sokomine let the player sort it out and clean up the inventory 20:22 thexyz oh no, I think this won't work 20:23 Sokomine did you understand my problem or was my presentation too chaotic? 20:23 thexyz I was talking to myself, actually 20:24 Sokomine yes, your problem/solution seems unrelated. but that's quite common on irc that there are multiple topics at the same time. sanity checks do sound like a very good idea 20:33 PilzAdam Sokomine, its basically the same thing in the furnace output list 20:34 PilzAdam but I dont see how this leads to item loss 20:34 PilzAdam the only problem is that you can glitch items into lists where you are only allowed to take from 20:35 Sokomine yes. and that's a problem if the mod relies upon allow_put where it denies such moves 20:35 VanessaE PilzAdam: sign onto my server and try it 20:36 PilzAdam VanessaE, try what? 20:36 PilzAdam Sokomine, can you create an issue on github? 20:36 Sokomine and since my saw forbids putting items in those slots, it does not expect them to appear there. the inventory is updated after taking of items (after all, there may be less available now) 20:36 VanessaE PilzAdam: the bug she's talking about. Put a stack of cobble and a stack of dirt in your inventory. Place a saw. Put the cobble in the saw. Take out say cobble stairs. Place those stairs onto the stack of dirt that's in your inventory. Bye bye dirt. 20:37 Sokomine the only solution i see currently is to add a comment to allow_metadata_inventory_put that it's not reliable 20:40 Sokomine shall the issue be attached to minetest_game or minetest? 20:41 PilzAdam minetest 20:41 Sokomine ok 20:55 Sokomine https://github.com/minetest/minetest/issues/944 (hope the error description is understandable) 21:02 ShadowNinja Better to move it to a "holding" stack. 21:04 Sokomine that would be a solution as well 21:05 Sokomine there was a time when misplaced items (=taken out and not properly inserted into another slot) where dropped as items on the ground. perhaps that would work as well 21:06 Sokomine but for this particular case, denying the move would be the best imho. the player can sort out what he/she wants to do in order to get a free inv space