Minetest logo

IRC log for #minetest-dev, 2016-04-03

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

All times shown according to UTC.

Time Nick Message
00:20 paramat closed #2896
00:20 ShadowBot https://github.com/minetest/minetest/issues/2896 -- Useless arithmetic operations on player position in a save file
00:40 paramat closed #3012 we now have this ability
00:40 ShadowBot https://github.com/minetest/minetest/issues/3012 -- Client side way to turn off player names
00:54 iangp joined #minetest-dev
00:57 Dragonop joined #minetest-dev
01:05 Taoki joined #minetest-dev
01:22 paramat closed #3097
01:22 ShadowBot https://github.com/minetest/minetest/issues/3097 -- Animated textures should be separately synced depending on a flag.
01:22 hmmmm why did you close it?
01:23 hmmmm there's no reasoning given whatsoever
01:31 paramat ok will add, as discussed the random flag is impossible
01:40 paramat closed #3164 as now in advanced settings menu
01:40 ShadowBot https://github.com/minetest/minetest/issues/3164 -- FOV - change via GUI
01:43 paramat any comment on #3176 ?
01:43 ShadowBot https://github.com/minetest/minetest/issues/3176 -- Don't have Schematic sources assign schematics to globals
01:56 pozzoni joined #minetest-dev
02:21 hmmmm i don't quite understand it
02:31 Miner_48er joined #minetest-dev
02:36 paramat i certainly don't
02:37 hmmmm perhaps you should ask him to clarify what he's asking for
02:49 paramat sure
02:51 hmmmm enjoying your job as a business analyst?
02:51 hmmmm ;|
02:51 hmmmm "i have people skills, dammit!  what the hell is wrong with you people!!"
02:58 paramat heh
03:36 domtron joined #minetest-dev
03:46 asl97 joined #minetest-dev
04:14 domtron_ joined #minetest-dev
04:16 hmmmm wow!  594 issues
04:16 hmmmm that's like 30 closed in a day
04:16 hmmmm I can't tell if this is a good thing or not lol
04:17 hmmmm then again if you start with the oldest first i guess it can be pretty easy to groom
04:18 hmmmm focus on metrics too much and you'll be a pointy haired corporate executive though
04:19 VanessaE yeah but if he becomes a PHB we can make fun of him ;)
04:21 hmmmm dunno, typically in my own experience your manager is a lot more in the know than they're portrayed
04:21 hmmmm it's the next level up where that stereotype really begins to apply
04:22 VanessaE heh
04:22 hmmmm they always have some kind of "Director" title
04:22 hmmmm which is basically the bottom level position on the executive job track
04:23 hmmmm i find that they often engage in something i like to call 'keyword matching', where they see one keyword pop up in one group, and a similar sounding keyword pops up in another group, and then they try to 'synergize' without any deep understanding
04:24 hmmmm they also don't get it when you tell them they're two different incompatible things, and they keep sending you people from another team - "uhh <big boss> told me you're the guy to talk to about X"
04:26 domtron__ joined #minetest-dev
04:26 paramat i've gone through 23 out of 26 pages but it's taken a week and many hours. there were quite a few obvious closes, but i've documented closes in this channel so they can be checked
04:29 sofar anyone know where the texture are for the hotbar? can't seem to find them somehow.
04:30 sofar oh, didn't expect them in minetest_game
04:30 * paramat websearches PHB
04:35 hmmmm paramat:  about #3945, you can modify somebody's PR by simply 'git commit --amend'
04:35 ShadowBot https://github.com/minetest/minetest/issues/3945 -- Add syntax fix for torch generation in dungeons (developer feature) by Kinokoio
04:36 hmmmm that's what I do on my own to avoid having to ask people to fix their atrocious (much worse than that) commit titles
04:36 paramat ok
04:36 hmmmm or other misc. whitespace errors
04:36 hmmmm cooperation with other people == slowness
04:36 sofar yup
04:36 hmmmm if you can do it on your own it's better
04:36 paramat will do
04:36 sofar or use git rebase -i HEAD~<number> and do fixup commits in case it's a series
04:37 sofar that way you can sometimes trivially fix commit 3 out of 5
04:47 kaadmy joined #minetest-dev
05:36 hmmmm what do people think about adding a label to PRs "Awaiting contributor feedback"
05:36 hmmmm we have one like it but it says user instead which is intended for issues
05:37 sofar isn't there a generic "need info" tag?
05:38 hmmmm no
05:40 sofar hmmm I'd love it if we could make the default skybox settable by texture pack....
05:40 hmmmm keep adding more issues why don't you
05:40 hmmmm just got below 600
05:41 sofar I'm not a prolific bug filer :)
05:41 sofar I tend to file bugs when I know that I can fix them myself, maybe not immediately, but soon enough
05:43 ssieb hmmmm: odd, I almost always put spaces between operators...  And I never take chances with operator precedence ;-)
05:44 ssieb I have never really coded in C++.  Are the boolean operators lower precedence than comparisons?
05:44 hmmmm foo != bar && blah isn't ambiguous to me... it's when you start mucking around with & and so on
05:44 hmmmm logical operators have lower precedence than comparison operators.
05:44 ssieb ok
05:44 VanessaE sofar: I argued for that (default skybox) when they were first re-introduced.  my arguments, whatever they actually were I don't remember, were not heeded.
05:44 hmmmm I agree it's good to be explicit, but c'mon, we're not coding lisp here...
05:45 hmmmm ((((((()(()())()(()()())()(
05:45 sofar I don't want to change the default, I just want texture packs to provide alternative skyboxes
05:45 VanessaE https://xkcd.com/859/
05:45 sofar if the server sets a skybox, end of game, the server wins
05:45 sofar but I'd like to see my texture pack skybox first
05:46 VanessaE sofar: at the time, it was decided that basically, some mod should handle even the setup of a skybox
05:46 sofar it's too bad, because I can get a ton of nice skyboxes from MC resourcepacks (legally)
05:46 VanessaE so unless something has changed, there literally isn't a default skybox for you to apply a texture to at all.
05:46 sofar but I can't ever put them in a mod and distribute
05:47 sofar ah well
05:49 ssieb hmmmm: The code I was replacing had the foobar/255.0f, I was just copying it. :-P
05:53 ssieb sofar: do you know if there's a way to update the commit time when using --amend?  I see there's a --date option, but it would be nice to just have it use the current time like usual
05:53 ssieb oh! haha, just use "now"
05:55 hmmmm phew
05:55 hmmmm i got the number of pull requests down to 130
05:55 hmmmm i thought we said we'll never let it go above 100 again
05:55 ssieb hmmmm: I didn't get an email for each comment...  Reading the others now.
05:55 hmmmm oh you get emails for every comment i make?  *shudders*
05:56 hmmmm what i say is not that important
05:56 hmmmm sorry for bugging you
05:57 ssieb No, those were good points.  The special case for 0 alpha is because there is nothing to do in that case.  Although I will double-check that to make sure now.
05:58 hmmmm i had to like triple check that function because the conditions were confusing
05:59 ssieb I'll comment the conditions to explain which cases they are covering
06:01 hmmmm from how i understand it it's like:  if we had a valid color string number then use it as the ratio between the src image and the color
06:01 ssieb The only use for 0 alpha would be to create a fully transparent texture.  But would that actually be any use?
06:01 hmmmm otherwise, we either had an invalid number string, or our ratio is 255, which means the color is solid
06:02 hmmmm so therefore if we want to keep the alpha of the color specified, use that
06:02 hmmmm otherwise, blit the full pixel as long as the destination's alpha isn't 0
06:03 hmmmm (like a mask)
06:03 hmmmm that's not confusing at all
06:03 hmmmm pshh
06:03 ssieb The second parameter is actually supposed to be a number.  That's the ratio to use.  If the second parameter is not a number or missing, then the alpha from the color is used as the ratio.
06:09 ssieb Actually it's worse than that...
06:10 ssieb If the colorize color alpha is not 255, the alpha is used as the ratio and it also interpolates between the destination alpha and the color alpha :-/
06:11 hmmmm pretend I'm somebody who has never read apply_colorize
06:11 ssieb But whatever, the new code works the same way as the old, so it is what it is...
06:11 hmmmm but I want to use this function
06:11 ssieb as a modder or coder?
06:11 hmmmm how would you explain it to me given the clusterfuck of logic branches with color, ratio, and keep_alpha
06:12 ssieb Did you read the doc change I made as well?
06:12 hmmmm yes
06:12 ssieb If that doesn't explain it , then please tell me what's not clear so I can update it.
06:15 hmmmm personally, what I'd do in this situation for clarity above anything is make a separate function that determines the blit strategy based on the intermingling of all these logical conditions
06:15 hmmmm that also makes it easier to add a unit test for
06:16 ssieb You mean separate functions for the different cases?
06:16 hmmmm well you wouldn't *need* to make separate functions for each
06:17 hmmmm again that's only what i would do if i were in the same situation
06:17 hmmmm i'm not telling you, that you need to do it this way
06:17 * ssieb did consider that at first
06:18 ssieb and one of those branches is only an optimization
06:19 hmmmm the keep_alpha branch?
06:19 ssieb no, that's the whole point of me doing this :-)
06:19 ssieb the else part of the if(keep_alpha) branch
06:20 hmmmm the keep alpha branch being separate from the else of the complicated condition if are the two that are functionally the same
06:20 ssieb not sure what you mean there
06:20 hmmmm me neither
06:21 hmmmm right
06:21 hmmmm lines 1805 through 1808 has the same effect as 1812 through 1818 when alpha == 255
06:21 ssieb right
06:22 ssieb just avoiding all the calculations
06:22 hmmmm oh dear, a subtraction, addition, and a multiplication
06:22 hmmmm :/
06:23 ssieb no, the repeated color.getInterpolated
06:23 hmmmm that's what i'm referring to
06:23 ssieb for every pixel in the texture
06:23 hmmmm interpolation isn't really that bad as it's linear
06:23 ssieb I suspect this is not a performance sensitive bit of code, so if it's a problem, I could just remove that part
06:23 hmmmm it's not, but as long as it's already been done, it's welcome
06:24 hmmmm aren't these functions only called once per textuer on game startup?
06:24 hmmmm once per texture with a colorize modifier
06:25 ssieb yes
06:25 hmmmm yeah then again the cases we should optimize for are vanessae's things
06:25 * VanessaE hides
06:25 hmmmm she has like several thousands of colorized textures from all those "same exact thing now in 101 colors" mods that are each 512x512
06:25 ssieb This new code also avoids an image malloc for every colorize as well
06:25 hmmmm "why is it taking 30 seconds to load!!!"
06:26 VanessaE now now, my mods use 16px textures
06:26 hmmmm worst case scenario
06:26 VanessaE (usually)
06:26 hmmmm oh yeah I see it
06:26 hmmmm the IImage initialization
06:27 hmmmm no reason to create a brand new image object for something like this.
06:29 ssieb I've pushed an update to the pull request addressing your comments. Do the extra comments help?
06:29 ssieb oh, how do I update the pull request with the updated commit?
06:30 sofar force push to the branch
06:30 sofar assuming you amended
06:30 ssieb I did that already
06:30 ssieb but I don't see the changes in the pull request...
06:30 hmmmm you just did the comment?
06:31 sofar did you push to the right remote and right branch?
06:31 ssieb yes
06:31 sofar which PR?
06:31 ssieb one sec, double checking
06:32 ssieb oh, that's weird.  It says I committed it 4 minutes ago, but doesn't have the changes
06:33 ssieb oops
06:33 ssieb I didn't put -a in the commit, so it just amended the comment :-(
06:33 ssieb ok, updated now
06:34 ssieb actually, you were right about the alpha 0.  I'll remove that check
06:36 ssieb It's a pointless optimization anyway.  Just don't do that. :-)
06:38 ssieb hmmmm: that's better?
06:38 hmmmm sure
06:46 Krock joined #minetest-dev
06:53 Hunterz joined #minetest-dev
08:15 nrzkt joined #minetest-dev
08:31 cd2 joined #minetest-dev
08:41 Zeno` joined #minetest-dev
09:15 technics joined #minetest-dev
09:30 nore #3895 #3948
09:30 ShadowBot https://github.com/minetest/minetest/issues/3895 -- Make node timers more efficient by Ekdohibs
09:30 ShadowBot https://github.com/minetest/minetest/issues/3948 -- Escape more strings: formspecs, item descriptions, infotexts... by Ekdohibs
09:30 nore do we merge 3895 for 0.4.14, or do we wait for it?
09:31 nore also, 3948 is compatibility code for the future, so I'd like to get it merged for 0.4.14
09:31 nore so we can get client-side translations and colored chat in 0.4.15
09:36 nrzkt nore i think we should wait release for 3895. 3948 is trivial i don't see a problem
09:37 nore nrzkt: fine for me
09:43 nore nrzkt: do you intend to rebase #2411, or do you want me to try that?
09:43 ShadowBot https://github.com/minetest/minetest/issues/2411 -- Colored chat for FreeType builds by nerzhul
09:44 Calinou joined #minetest-dev
09:45 nrzkt nore you can try it if you want, but take your time :)
09:46 nore I'll try that maybe then :)
09:48 asl97 seem like hmmmm isn't around atm, can anyone else could suggest a title change for #1218 so that it suit the reason that it was keep open since the original issue was fix?
09:48 ShadowBot https://github.com/minetest/minetest/issues/1218 -- huge error shouldn't be printed in game
09:59 nore nrzkt: there seems to be some interesting PRs that have been closed when dev-0.5 was deleted
10:00 nore like sapier's statbars
10:00 nore maybe we should try to relive them
10:23 jin_xi joined #minetest-dev
10:33 est31 joined #minetest-dev
10:40 est31 "Action needed" label, what's that for?
10:40 est31 who'se action is needed?
10:40 est31 if you dont specify who's action is needed, nobody feels responsible
10:41 nrzkt this is a useless label...
11:05 blaze joined #minetest-dev
11:08 Calinou_ joined #minetest-dev
11:14 celeron55 joined #minetest-dev
11:24 celeron55 joined #minetest-dev
11:27 asl97 joined #minetest-dev
11:34 celeron55 joined #minetest-dev
11:38 iangp joined #minetest-dev
11:44 celeron55 joined #minetest-dev
11:51 nore nrzkt: I've got working colored chat! :)
11:51 nore this still needs a lot of cleanup though
11:53 nore first change I want to do is ColoredString -> EnrichedString, which can be extended later, for example for bold text, italics, or this kind of things
11:54 nore also, make that class make the drawing instead of what is currently done
11:54 nore and finally, see if I can manage to get it to work in items descriptions or formspecs too
11:54 nore (that's a lot of things to do...)
11:55 celeron55 joined #minetest-dev
11:57 nrzkt yes gl :)
11:59 nore last one should not be too difficult once I've got the first one working
12:00 nore anyway, this is heavily WIP
12:01 nore gtg now, maybe I'll open a pull tonight so that you can test it/report what you think is not good/whatever
12:05 celeron55 joined #minetest-dev
12:16 celeron55 joined #minetest-dev
12:24 yang2003 joined #minetest-dev
12:38 Fixer joined #minetest-dev
12:45 nrzkt take your time, it's not time for this
13:25 kilbith joined #minetest-dev
13:25 kilbith i get some compile warnings because of latest commit : http://paste.ubuntu.com/15598570/
13:26 kilbith translation : "comparison between signed and unsigned integer expressions"
13:29 STHGOM joined #minetest-dev
13:34 Gael-de-Sailly joined #minetest-dev
13:41 Zeno` s32 should probably be u32... don't have the source handy
13:46 kilbith Zeno`, it fixes the warnings, wanna make a PR ?
13:47 Zeno` for that? Nah, someone should just commit it :)
13:47 kilbith but there's also an unused function in that file now
13:47 kilbith causing a warning
13:54 Fixer !seen gregorycu
13:54 ShadowBot Fixer: I saw gregorycu in #minetest-dev 3 days, 0 hours, 34 minutes, and 27 seconds ago saying "Off to bed for me"
13:57 damiel joined #minetest-dev
13:58 turtleman joined #minetest-dev
14:07 Zeno` which commit caused that?
14:08 kilbith latest one
14:08 Zeno` #01ae43 ?
14:08 kilbith yup
14:08 Zeno` oh yeah, I see
14:12 Zeno` hmm, it shouldn't really be u32
14:13 Zeno` but I'm not sure what the best approach is there. hmm
14:14 kilbith dst_pos is s32 btw
14:14 Zeno` yeah
14:14 Zeno` so I'm wondering in what cases one of the coords might be < 0 (if any)
14:15 Zeno` if they can never be < 0 then two asserts should be added and u32 used
14:16 Zeno` well, one assert: e.g. assert(dst_pos.X >= 0 && dst_pos.Y >= 0);
14:16 Zeno` if either CAN be -ve then things will break anyway
14:19 Zeno` if they can be negative then they must be promoted (well, demoted) to s32 properly with some kind of out-of-range check
14:19 Zeno` because there is a narrowing of range
14:23 Zeno` s/narrowing/change
14:24 Zeno` kilbith, maybe open an issue. It's a trivial fix but *shrug*
14:24 kilbith will do
14:27 kilbith wtf, i can't click on the "submit new issue" button
14:27 Krock fix your mouse button
14:29 kilbith https://lut.im/akeZm3J9Zb/7aRWH9rYhDpoVXRZ.png
14:31 kilbith oh yeah, did not set a title
14:55 Player_2 joined #minetest-dev
15:05 PilzAdam joined #minetest-dev
15:29 asl97 joined #minetest-dev
15:34 Krock lol
15:43 Niebieski joined #minetest-dev
15:45 Niebieski Hey guys, in the 'change keys' menu "Toggle: F8", should it state what it actually toggles ?
15:45 Niebieski for me it toggles the cinematic camera.
15:47 sfan5 it most likely should
15:48 Niebieski Let me try fix that. xD
15:48 Niebieski It should be in guiKeyChangeMenu
15:51 Niebieski It is actually written on the source code.
15:51 Niebieski in*
16:10 Krock F8 turns cinematic on/off my default
16:10 Krock seems like it's poorly described
16:12 Obani joined #minetest-dev
16:38 hmmmm joined #minetest-dev
16:55 hmmmm you know, FarMap isn't some kind of fascinating, brilliant new idea.  all games have them and they're called LOD systems
16:56 hmmmm that being said I wonder if it makes sense to make it work more like a typical LOD system, where the level of detail decreases with distance, instead of staying a constant 8x8x8 in size
16:58 hmmmm also from what i've seen in screenshots, farmap blocks cast shadows?? is that really necessary??
16:59 Zeno` normally distant objects don't cast shadows
16:59 Zeno` there is no point
16:59 hmmmm what makes this LOD system unique is that it has to transfer the information over the network to the client
16:59 Zeno` I mean they do in real life, but in a game it serves no purpose
16:59 celeron55 cast shadows? what the hell are you talking about :P
17:00 hmmmm watching this https://www.youtube.com/watch?v=gQ8fo5VBKEM
17:01 celeron55 are you talking about the fact that they include some lighting values?
17:02 hmmmm yes
17:02 celeron55 if you don't do that, then they look utterly stupid at the transition from regular to farmap rendering when there are shadows
17:02 celeron55 eg. a large cliff or cave
17:02 hmmmm do you realize how much cpu power that takes up
17:02 celeron55 what exactly takes cpu power?
17:02 hmmmm lighting calculations
17:03 celeron55 it's not anything extra at this point when farmap has no shortcut for getting its data
17:03 celeron55 i mean, the only way for it to get data is to actually generate the part of the world
17:03 celeron55 only networking and rendering is affected in this version
17:04 celeron55 of course when there's some shortcut for getting the LODed data, then the shortcut probably doesn't result in accurate lighting
17:05 hmmmm looking right now to see you do it
17:05 celeron55 there's some calculation necessary for figuring out the appropriate light level in each "farnode" of course
17:06 celeron55 but it's not that much more than is necessary anyway for squeezing up the nodes otherwise
17:08 celeron55 the farmap protocol supports the server sending any splits of nodes instead of 8x8x8, but i decided that for the server and client to perform best, it's better for it to just handle one LOD level; otherwise there's quite some overhead in handling all the levels especially on the server which has to handle players at arbitrary positions in the world
17:08 celeron55 i mean, 4x4x4
17:10 est31 joined #minetest-dev
17:10 iamafriend joined #minetest-dev
17:10 celeron55 the current client implementation doesn't quite render a selection of different LODs sent by the server correctly but it can be done if that's what's actually wanted
17:10 hmmmm this is way too big
17:10 celeron55 the textures just get quite weird
17:10 hmmmm i need to set aside a week or so to review it
17:11 celeron55 also, the farmap branch still has nasty bugs, even design bugs; it's not wise to merge it in its current state
17:11 est31 farmap does good on stuff like landscape
17:11 celeron55 but looking through it would be a good idea; i don't think anyone has taken a proper look at it yet
17:12 est31 but its bad for houses
17:12 hmmmm how do you pump out code so fast?!
17:13 celeron55 do i?
17:13 est31 yes
17:13 est31 at least I get the impression, and hmmmm too
17:14 celeron55 well... maybe it's one reason i managed to start a minecraft clone and not someone else...
17:14 est31 true
17:15 VanessaE when you sit down to code something, celeron55, you DO pump the code out pretty damned fast.
17:15 celeron55 usually i'm incredibly unproductive though
17:15 hmmmm lol you haven't seen me at my worst
17:15 VanessaE it's just a matter of getting you started  ;)
17:15 hmmmm the past couple of weeks i've been screwing around playing grand theft auto all day
17:16 hmmmm i should be more like celeron
17:16 VanessaE I shouldn't talk though, I haven't exactly been burning up the text editor lately myself.
17:19 hmmmm past couple of times i wanted to code something i opened up the IDE, looked at it for about 5 minutes and went 'nope'
17:19 celeron55 you have a day job that includes programming though, right?
17:19 hmmmm but yeah before starting on any new huge features i'd rather get farmap in because it touches almost everything
17:20 hmmmm celeron, yes, but i don't do shit there either lmao
17:20 hmmmm or maybe i do, it's a matter of perception
17:20 celeron55 at the moment i don't have, so i can do such a ridiculous burst of programming, like 14h per day for a week, without really a problem at all
17:21 hmmmm i'd get bored after 1 hour
17:21 celeron55 but i need a proper goal before that can happen
17:21 hmmmm what about household chores and what not
17:22 hmmmm what about food?  then again, i take an abnormally long amount of time to cook food because everything turns out to be elaborate
17:22 celeron55 well, maybe more like 12h/day
17:22 celeron55 but i mean, cooking and stuff is basically coding
17:23 celeron55 you don't necessarily need to think about anything else
17:23 est31 stove is the compiler :)
17:23 est31 or heatplate or whatever its english name is
17:23 celeron55 you just move your hands around while you think about the next thing you're going to write
17:25 hmmmm you've inspired me to do things
17:25 hmmmm thank you
17:25 VanessaE est31: stove (or oven if baking something)
17:26 VanessaE interesting analogy though, c55.
17:26 VanessaE never thought of cooking in that way.
17:26 Amaz joined #minetest-dev
17:27 celeron55 lol, i didn't quite mean that but i guess it is
17:29 est31 VanessaE, thanks
17:33 celeron55 i think i could summarise the way i write a large amount of code as "move on before you get bored"
17:42 celeron55 then you just... have to be good enough to make working code before you get bored
17:55 PilzAdam <est31> farmap does good on stuff like landscape <est31> but its bad for houses <- what about only showing is_ground_content=true nodes in the far blocks?
17:56 est31 and then manually sending over the non ground content ones? perhaps worth a look
17:59 est31 awwww
17:59 est31 found another uglyness in the pathfinder
17:59 est31 why is it so ugly
18:00 est31 it first loads the entire (!) area it considers the path to be inside into RAM
18:00 est31 thats already ugly enough
18:00 est31 and i fixed it already
18:00 est31 and then it loads it again when doing the actual path search
18:01 est31 man this really isnt something sapier can be proud of
18:04 iangp joined #minetest-dev
18:05 est31 you know this code is doing such a bad job at such a solved problem
18:05 est31 i really am considering to rewrite the whole thing
18:05 est31 its just not worth it
18:08 kaeza joined #minetest-dev
18:10 est31 to keep the code
18:19 Fixer i'm still trying to finish reading python manual... for years %)
18:19 Fixer not manual, tutorial
18:20 Fixer usually at the middle something happens and I'm starting to do something else and forgot
18:41 Void7 joined #minetest-dev
18:43 kaadmy joined #minetest-dev
18:55 proller joined #minetest-dev
18:57 paramat joined #minetest-dev
18:59 paramat celeron55 Zeno` is back and wants to be a core dev again, could you invite him?
19:01 paramat nore is there any reason to delay #3895 ? could it introduce subtle bugs? we have perhaps 3 weeks before release
19:01 ShadowBot https://github.com/minetest/minetest/issues/3895 -- Make node timers more efficient by Ekdohibs
19:02 paramat some of the best coding i do is doing-the-washing-up coding
19:03 nore paramat: I don't think there will be problems with it
19:03 nore could you check #3948 too please?
19:03 ShadowBot https://github.com/minetest/minetest/issues/3948 -- Escape more strings: formspecs, item descriptions, infotexts... by Ekdohibs
19:04 paramat yeah i'l review as best i can
19:04 paramat (i'll) (review both)
19:05 paramat nore PilzAdam ShadowNinja game#999 ?
19:05 ShadowBot https://github.com/minetes​t/minetest_game/issues/999 -- Dirt should turn to dry grass in some occasions. by sofar
19:06 paramat description needs updating
19:06 celeron55 paramat: well i suppose Zeno` can ask me directly
19:06 paramat the 3 surface types spread from node to node
19:06 paramat ok
19:07 celeron55 i know that he resigned by himself and removed himself from the team by himself but still
19:07 celeron55 it's weird to be adding people without even talking to them
19:09 VanessaE he mentioned it to me also, the other day.
19:09 VanessaE maybe it's just a time zone issue, *shrug*
19:14 Fixer http://irc.minetest.ru/minet​est-dev/2016-04-01#i_4571466 "zeno wants back part"
19:14 paramat game 999 needs a small update, see comment
19:15 VanessaE celeron55:  ^^^ I would consider that a pretty-much direct request.
19:17 VanessaE paramat: a side note about _game that you may be interested in:  dreambuilder is no longer a subgame, it's back to being a (huge) modpack, to be used on top of vanilla mt_game (like any other mod/pack).  So shit better work ;)
19:18 VanessaE so I may propose an occasional option or feature for mt_game here and there that would make some part of the modpack work better, maybe.
19:18 VanessaE (for now, I have nothing in mind)
19:18 paramat ok
19:18 * kilbith starts to be scarred
19:20 VanessaE kilbith: I mean e.g. adding a way to disable some function or change its behavior without hacky workarounds, that sorta thing.
19:21 kilbith ok
19:21 kilbith i was thinking to a flow of mesh models :)
19:21 VanessaE hah
19:21 VanessaE no
19:21 paramat nore sfan5 PilzAdam ShadowNinja game#1008
19:21 ShadowBot https://github.com/minetest​/minetest_game/issues/1008 -- Flood snow with either lava or water. by sofar
19:22 PilzAdam paramat, please stop highlighting me today
19:24 VanessaE kilbith: that said, I'm certainly not opposed to more meshnodes in mt_game, where they make sense.  can't think of anything that isn't currently controversial though.
19:27 Gael-de-Sailly joined #minetest-dev
19:27 VanessaE (I'm sure Mushroom Man here doesn't like that idea though ;) )
19:31 celeron55 VanessaE: oh i didn't notice that; i guess i'll just invite him on github then
19:31 VanessaE ok.
19:38 paramat PA ok
20:00 kilbith ShadowNinja, once you have time, can you review a PR on one of your mod please ?
20:00 kilbith sorry for the offtopic, seems impossible to reach you in PM
20:04 Dragonop joined #minetest-dev
20:35 paramat i'll merge game#1009 game#1008 game#1003 game#994 in an hour
20:35 ShadowBot https://github.com/minetest​/minetest_game/issues/1009 -- Creative: Add myself in credits by kilbith
20:35 ShadowBot https://github.com/minetest​/minetest_game/issues/1008 -- Flood snow with either lava or water. by sofar
20:35 ShadowBot https://github.com/minetest​/minetest_game/issues/1003 -- Fix disappearing doors while underwater by tenplus1
20:35 ShadowBot https://github.com/minetes​t/minetest_game/issues/994 -- Beds: Check for buildable_to for bottom half by sofar
20:37 Fixer i'm very worried about one bug
20:38 Fixer i'm not sure what the cause, but I've experienced death on login on two heavily modded server that uses recent minetest
20:38 Fixer you login, death, you loose your bone
20:45 Gael-de-Sailly joined #minetest-dev
20:52 Obani paramat, did you take a look at Gael-de-Sailly's new wip mapgen ? :)
20:53 paramat not yet but i saw the biome diagram, nice
20:57 Gael-de-Sailly i just discovered that someone else has a very similar project
20:57 Gael-de-Sailly https://forum.minetest.net/vi​ewtopic.php?f=49&amp;t=13492
20:58 Gael-de-Sailly so I've proposed to work with him, together
20:59 paramat ooh lovely voronoi diagram
21:22 nore Yay! I've got working colored text in chat, item descriptions, etc :)
21:26 Dragonop nore: Cool! Is there a way to change the text size in item descriptions?
21:26 nore Dragonop: not yet, but this could be added, it wouldn't be very easy though (like bold or italics, I think this means the font has to be changed)
21:27 nore other planned features are clickable links, if I can figure how to do that
21:27 Dragonop Oh I see
21:28 Amaz That sounds awesome nore!
21:29 nore it requires #3948 to be merged in a stable release before it is added though
21:29 ShadowBot https://github.com/minetest/minetest/issues/3948 -- Escape more strings: formspecs, item descriptions, infotexts... by Ekdohibs
21:30 nore clients that don't have this pull will see strange effects
21:30 nore paramat: thoughts about it btw? can it be merged before 0.4.14?
21:32 paramat i can review it later, although i won't understand it much, but at least i can so a code style review
21:33 Fixer looks like we have critical gameplay bug, looks like Dragonop reproduced it, interesting
21:34 nore Fixer: marked as blocker
21:34 nore this is for too critical to allow releasing 0.4.14 with this bug
21:34 Fixer this bug is very annoying and at least few weeks old, let me see
21:35 nore Fixer: do you know if the problem is with the server or the client?
21:35 Fixer not sure, player connects -> he dies instantly, no idea
21:35 nore also, are you able to reproduce it?
21:35 Fixer Dragonop is able
21:35 Fixer he is here
21:35 Fixer interesting
21:36 Dragonop I will double check, just in case
21:36 Fixer Dragonop, we need to bisect it, can you try earlier builds were you don't die?
21:36 Dragonop Okay
21:37 Fixer Dragonop, on what build you died?
21:37 Fixer you can grab earlier build from sfans page
21:37 Fixer probably
21:37 Fixer or krocks
21:37 Fixer i will try to reproduce too
21:37 nore Fixer: I'm at c8ff11b and I don't get this problem
21:38 Fixer Dragonop, so you started server, enabled damage and seperately started minetest client and connected to server?
21:38 paramat possibly related #3483
21:38 ShadowBot https://github.com/minetest/minetest/issues/3483 -- Very slow spawn on mapgen v7(Android)
21:39 Dragonop Fixer, no, singleplayer, just make a new world and get inside it
21:39 Fixer didn't die, so we will rely on you
21:40 Niebieski Fixer, I experienced the same issue on Xanadu.
21:40 Dragonop I'm double checking
21:40 paramat Dragonop are you experiencing this in Android? what OS?
21:40 Dragonop Windows
21:40 Fixer Dragonop, can you give seed and starting position?
21:41 nore Dragonop: and mapgen too
21:41 Dragonop first attempt, can't reproduce with latest minetest_game
21:41 Dragonop let me test
21:42 Fixer 0115da1 won't die in sp
21:42 Dragonop first two attempts in mgv6, and I didn't died, first attempt in mgv7, and I died
21:43 Fixer Xanadu and ESM also use v7
21:43 Fixer interesting
21:43 paramat =/
21:43 Dragonop seed 1863198836516757775 mapgen v7
21:43 Fixer Dragonop, what is game build?
21:43 Fixer from sfan? which one?
21:44 Dragonop Krock's latest
21:44 Dragonop I can test on sfan5's latest too
21:45 Fixer ok, will grab krocks latest
21:46 nore with same seed, I don't die
21:46 Dragonop Maybe it is related to how long it takes to generate terrain
21:46 nore Dragonop: possible
21:47 Fixer Dragonop, hmm, what version of mintest_game you use?
21:47 Dragonop latest
21:47 Fixer really latest?
21:47 Dragonop 1977ae1
21:49 Fixer paramat, i think tenplus1 actually updated to pretty recent MTG, after server restart, probably week ago
21:50 Niebieski It happened when my login took so long.
21:50 nore could reproduce by adding the following code: "minetest.register_on_generated(function() os.execute("sleep 5") end)
21:50 nore "
21:50 nore i.e. simulate a very slow mapgen
21:51 Dragonop tested again, seed 14752785969290718230 and I died every time I respawed too
21:52 Dragonop So it is related to how quickly the mapgen generates, this time it took from 2 to 5 seconds to generate the map where I was spawning
21:52 nore Fixer: ^
21:52 Fixer didn't reproduce it yet with those seeds on krocks build
21:52 Dragonop Even if I move it still kills me, lol
21:52 paramat seems very similar to issue 3483
21:53 nore it is probably related to how fast your machine is thus
21:53 nore that would explain why you can't reproduce it
21:53 paramat ah
21:53 Fixer interesting
21:53 paramat yes spawn is fast for me
21:53 Fixer so if server is slow
21:53 Dragonop btw 1.6GHz singlecore, 2Gb ram
21:53 Fixer everyone is fucked up (well randomly) %)
21:53 nore a 2s sleep is enough actually
21:54 Fixer hmm
21:54 paramat might also be related to recent damage bugs, why should damage be given on slow spawn?
21:54 nore with that, I die every time I create a new world
21:55 Dragonop mgv6 takes 1 second to generate when I first spawn so that's why won't die
21:55 Fixer let me search for prime95 that will load my CPU a lot
21:55 Dragonop on mgv7 I died mostly when spawning on snowy biomes
21:56 Fixer hmm
21:56 Fixer so we have sept 2015 - no bud, dec 2015 - bug reported
21:56 Fixer bug*
21:57 nore Fixer: well, now we know how to reproduce, we can bisect :)
21:57 Dragonop Should I test with sfan5's latest win32 build too, or do you think there is no need to?
21:58 Fixer try it just for luck %)
21:58 Fixer i will run prime95 and try again
21:59 Fixer it is server side bug i think
21:59 paramat mgvalleys is a slow mapgen too, perhaps try that
22:00 paramat might be worth trying all mapgens to see which are affected
22:01 Dragonop Okay, mgv7 seed 1423490487402405678 latest sfan5's win32 build, and I died
22:01 Dragonop will test with mgv7 too then, I saw this on mgv6 but it's uncommon for me. I stick with "it's related to how much it takes to generate terrain"
22:01 Dragonop mgv5*
22:02 Dragonop died on mgv5
22:02 Fixer ok, so i set lowest priority to minetest, fired prime95 to take as much of CPU as possible, didn't die in mgv7
22:03 Dragonop Oh! when I was _respawning_ after that, I took damage! about 8 damage points!
22:03 Dragonop maybe it thinks I'm falling?
22:03 Fixer well, i got 240 damage on one server when that bug striked me
22:04 Fixer some people already rage quit :S
22:04 paramat is your noclip set correctly?
22:04 Fixer noclip was off
22:04 Fixer it was singleplayer
22:05 Dragonop noclip, fly, and fast, all off, I still die
22:06 Fixer god of death
22:06 Dragonop testing with valleys
22:06 paramat i guess you're all correctly spawning at ground level
22:06 Fixer fun thing is I die on server but not at localhost :/
22:06 Dragonop spawned in jungle, noclip fly and fast off, dead. This time I spawned on a tree top
22:06 Fixer i mean on other people servers
22:07 Fixer o.O
22:07 Fixer so died in valleys?
22:07 Dragonop yes
22:07 Fixer ok, so it remains to bisect it and look on what builds you don't die
22:08 Fixer so we have sep 2015 - mar 2016, choose middle date and download krocks build
22:08 Dragonop new year?
22:09 Fixer stop-stop-st
22:10 Fixer this problem was also probably reported at 23 Dec 15
22:10 Fixer hmm
22:10 Fixer sep 15 to dec 15
22:13 Fixer Dragonop, try this? https://www.dropbox.com/s/1ex2irm9​1bq2kwn/minetest_151102_5c3546e.7z
22:13 Dragonop Okay
22:14 Fixer Dragonop, and minetest-game from that time if possible :)
22:15 Fixer or maybe it is engine bug anyway
22:15 Fixer it should be
22:15 Dragonop I think it's an engine bug
22:15 Fixer so minetest-game can be new if old is not available
22:15 Fixer just try that build
22:15 Dragonop Yep
22:16 Dragonop Or test with minamal dev test and die anyways :-)
22:17 Dragonop minimal*
22:17 Fixer you die with that too? niiice
22:17 Fixer even better
22:17 Dragonop mgv7 again
22:17 Fixer died?
22:18 Dragonop yup
22:18 Fixer ok
22:18 Fixer let me pick older build in between
22:20 Fixer Dragonop, try this https://www.dropbox.com/s/3hcs​n6dvex1zll3/minetest_151001.7z
22:21 Dragonop kk
22:22 Dragonop Died again, mgv7
22:23 Fixer ok, picking another one
22:23 Fixer Dragonop, try this, we should be close: https://www.dropbox.com/s/wo3ovdy59​osv1n0/minetest_150920_overwrite.7z
22:24 Dragonop give me the version prior to it, it's an overwrite one
22:25 Fixer Dragonop, what do you mean by overwrite?
22:25 VanessaE is there a particular reason the F10 console now fills the entire screen?
22:25 Fixer Dragonop, this is very important version to test
22:25 VanessaE (versus the 't' version which only holds around 19 lines)
22:25 Fixer Dragonop, next one available should have no bug :/
22:26 VanessaE 10*
22:26 Dragonop Fixer, you need to replace the files in the version prior to it with this ones
22:26 Dragonop that's what I mean
22:26 Dragonop it's not the "full" package
22:26 Fixer Dragonop, ah, nice, so here the link: https://www.dropbox.com/s/5g5x​btkd7ohxh0n/minetest_150913.7z
22:26 Fixer Dragonop, download it and TEST for die, and after that we will decide if to overwrite
22:27 Dragonop Okay
22:30 Dragonop 3 tests, didn't died
22:30 Dragonop Fixer, do I overwrite and test?
22:30 Fixer nice, now overwrite it and test :)
22:32 paramat random thought: if this is somehow falling damage, perhaps try 'movement_gravity = 0.0'
22:33 Dragonop okay will do when I'm done with this version
22:34 Dragonop Fixer, nope, didn't died
22:34 Fixer ok, looks like we narrowed it down
22:35 nore joined #minetest-dev
22:35 Dragonop wb nore
22:35 Fixer so last good version is between 20.09.2015 and 15.10.2015
22:35 Fixer i mean bug is between them
22:36 Fixer Dragonop, i can compile 64 bit builds, can you run 64bit?
22:36 Dragonop nope
22:36 Dragonop I will try what paramat said in latest version
22:38 Dragonop paramat: first attempt and it didn't happened, fall damage might be the problem
22:40 Fixer i remember some PR for damage, let me see
22:41 Fixer https://github.com/minetest/minetest/commi​t/596484da4fcd301e597a851a9a54b6dc276a286a ? possible
22:41 Fixer https://github.com/minetest/minetest/commi​t/8787d2e7e9671f3b5930ca9c41c6e0867b8a5a54 thiiis <<<
22:42 Fixer @JohnWayne1986 @est31
22:42 Fixer Fix falling through nodes on world load (fixes #2784)
22:42 ShadowBot https://github.com/minetest/minetest/issues/2784 -- Player doesn't clip on half blocks on world load
22:42 Fixer some map border fixes was also present
22:42 Dragonop paramat, I'm not getting this bug with that setting, :-)
22:44 Dragonop Fixer, yes, I think the second one is causing it
22:48 Fixer i can try to build 32 bit builds on debian, hold on
22:50 Dragonop Okay, thanks
22:50 Fixer let me prepare it
22:56 rubenwardy joined #minetest-dev
22:58 Fixer Dragonop, if everything goes well i will prepare two builds, one just commit before suspect one, and with suspect one
22:59 Fixer need to install 7z :/
22:59 Dragonop lol
22:59 Fixer hold on
23:00 paramat cool
23:02 Fixer hope all 32 bit depends are satisfied
23:03 Dragonop worried about missing .dlls?
23:05 Fixer no
23:05 Fixer i'm running the bash script
23:05 Fixer worried about mingw
23:06 Dragonop Ah I see, well lets hope there are no issues
23:06 yang2003 joined #minetest-dev
23:07 Fixer hmhmh, cmake error
23:08 Fixer wonder if Krock is sleeping
23:08 Fixer he s
23:09 Fixer need to install mingw32 probably
23:09 Dragonop yep
23:11 Fixer installed, trying again
23:12 Fixer Dragonop, damned, errors
23:13 Dragonop Argh, well maybe we could ask Krock when he connects :/
23:14 Fixer Dragonop, yes, but I think developers will figure out even without us, i will ask Krock to compile it for us
23:14 Fixer Dragonop, tommorrow :)
23:15 Dragonop Okay, thanks for testing with me :-)
23:15 Niebieski joined #minetest-dev
23:15 Fixer thank you, without you to confirm that will be much longer and worse
23:15 Dragonop Haha, glad I'm able to help
23:30 Fixer paramat, yes, admin also reported smth like this, damage because of smth unknown :/
23:30 asl97 joined #minetest-dev
23:31 Fixer nore, are you 64bit windows user?
23:37 Fixer will try to check it myself with your function
23:44 nore joined #minetest-dev
23:47 nore Fixer: no, 64bit Linux
23:48 Fixer ok, will test myself anyway
23:51 Fixer nore, how exactly that mod should look? i didn't die
23:52 Fixer nore, http://pastebin.com/raw/0JLTNgXG like this?
23:54 iangp joined #minetest-dev

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