Time  Nick      Message
00:04 Fixer     nore, ok, reproduced too, now i will compile some more to test and to find commit with regress
00:04 nore      Fixer: ah, good
00:05 * nore    goes to bed (it's 2am here...)
00:05 Fixer     3am here %)
00:15 Fixer     commit with regression identified
00:20 Fixer     i guess now it is devs time to shine to fix it :}
00:20 VanessaE  well report the bug then :P
00:20 Fixer     VanessaE, don't, I've reported already and found the commit
00:20 Fixer     issue #3955
00:20 ShadowBot https://github.com/minetest/minetest/issues/3955 -- Death on login [needs investigation]
00:35 paramat   thanks good work
00:40 VanessaE  wow, old bug
00:50 Dragonop  Yep, hopefully fixing it won't create more bugs
02:39 paramat   closed #3472
02:39 ShadowBot https://github.com/minetest/minetest/issues/3472 -- Add chat command /m to send PM
02:43 paramat   .. as we now have what was asked for
02:44 * sofar   slaps paramat with a bush
02:44 sofar     :)
02:45 paramat   closed #3621 as confirmed actually working
02:45 ShadowBot https://github.com/minetest/minetest/issues/3621 -- minetest.transforming_liquid_add(pos) is not working?
02:46 paramat   *=.
02:59 paramat   closed #3949 because pointless issue
02:59 ShadowBot https://github.com/minetest/minetest/issues/3949 -- Add games that should be bundled with Minetest
04:34 paramat   now merging games 994 1003 1008 1009
04:42 sofar     well, crap, so much for making a nice cactus mesh
04:43 sofar     that would break the existing mapgen placed cactus models
04:43 VanessaE  sofar: round trees mod does round cacti also
04:43 VanessaE  (I'm thinking of the abm it uses, or is it an lbm now? I forget)
04:44 paramat   merged
04:44 sofar     that's... not what I'm looking for. I want to make it spikey by using a mesh node
04:44 VanessaE  ah
04:44 sofar     but I don't want to make the arms of the current mapgen cactus trees float
04:45 sofar     so I'd have to make the spikes stick outside of the 1x1x1 box
04:46 sofar     and then they z-fight :(
04:47 VanessaE  sounds like the only solution then would be a multi-node-sized model of something like a C. gigantea that drops several nodes' worth when dug.
04:48 VanessaE  (modeled, of course, to look like it's still made from cubes)
04:50 VanessaE  that's the route I'd take in this situation anyways.
04:52 VanessaE  another option might be to split the model into three or four pieces (central trunk, two arms, maybe a top piece) and have some active code that say, chops off one arm, then the other, then the top, finally the trunk, all pieces drawn so that they work well together but don't care if the arms look weird "separately".
04:52 VanessaE  that way you retain some granularity.  might even be seen as a gameplay feature.
04:54 VanessaE  (inb4 "no, that won't work either, not what I'm aiming for."  :P  )
05:02 sofar     lol
05:02 sofar     sorry, I'm a sucker for simplicity
05:02 paramat   it's cool you tried, but it's best left as a cube. in minetest there are many non-cubic objects represented as cubes, a mesh for cactus is going in the direction of making meshes for everything
05:02 sofar     if I can fix it with a nice mesh model.... I will
05:03 sofar     if not, then I'll pass
05:36 hmmmm     [01:02 AM] <paramat> ..., a mesh for cactus is going in the direction of making meshes for everything
05:36 hmmmm     so much this
05:36 sofar     lol
05:36 hmmmm     people are missing the point of a block world game because they're trying too hard to be like minecraft
05:37 hmmmm     a cube with spikes is not a block
05:37 sofar     slow down, it's just an experiment
05:37 sofar     it doesn't work with the mapgen cacti anyway
05:37 sofar     so, I'll probably shelf it
05:39 sofar     there's a giant gray zone between "this is an engine where anything is possible" and "let's keep everything a block"
05:40 sofar     http://i.imgur.com/sNpgTuU.png
05:40 sofar     doesn't look half-bad, does it now?
05:40 sofar     so, it's not a bizarre thought. someone else probably already did it
05:41 sofar     yup, I can find 1 google image that shows something similar, lol
05:42 VanessaE  I would still define that as "mostly a cube"
05:42 VanessaE  enough that honestly, I don't think it violates MT's general nature
05:43 sofar     there's one reason to do this
05:43 sofar     right now, you can't make cacti actually *hurt* a player
05:43 sofar     since it's 1x1x1
05:43 sofar     you can set damage_per_second = 1 for it, but a player could never "enter" the block
05:43 sofar     if the cactus block was *smaller* than 1x1x1, then that would work
05:44 VanessaE  the solution for that in the past has... yeah
05:44 sofar     so, there's a solid reason... for making it a mesh
05:44 VanessaE  a 12x16x12 nodebox for the body and the needles extending out to the 1x1x1 boundary
05:45 sofar     I'd just do 14px x 14px
05:45 sofar     let's not make it super skinny
05:45 VanessaE  well close enough
05:48 sofar     oh, hmm, damage_per_second doesn't work for a smaller collision box
05:49 sofar     wait, typo
05:53 sofar     well that's an ugly hack. you have to almost stand entirely inside the cactus for it to hurt you
05:53 sofar     I'm at 3/16 and then it starts to tick damage :(
05:55 sofar     meaning, I can walk in between 2 cacti and not get hurt even if they're adjacent
05:57 sofar     yeah, no wonder, yikes, that calculation in environment.cpp
06:23 * nore    wonders how 8787d2e7 could have added the problem
06:24 sofar     well, it stop players from moving
06:24 sofar     so now they're stuck in an invalid position
06:25 sofar     the trick is probably to stop damage until that condition is reached
06:47 sofar     well, all that playing with cacti at least resulted in possibly a nice addition
06:47 sofar     VanessaE: I got cacti to damage players standing on, walking into them
06:47 VanessaE  cool
06:48 Zeno`     Why to the cacti walk into people?
06:48 sofar     phrasing error, likely ;)
06:48 Zeno`     =)
06:49 sofar     the grammar really hurts the ccti
06:49 sofar     argh
06:49 sofar     obvious signs i need to go to be
06:49 sofar     d
06:49 Zeno`     heh
06:58 Zeno`     can someone explain the branches to me please?
06:58 Zeno`     Most of them didn't exist last I was active
06:58 sofar     all irrelevant except master?
06:59 Zeno`     ok, so the same as before. That's cool :)
06:59 Zeno`     what is dev-0.5?
07:00 sofar     march 5 2015 last commit... old stuff :)
07:00 sofar     I dunno maybe there's stuff in there that's still useful, but I doubt it
07:01 Zeno`     ok thanks
07:09 Obani     sofar, https://github.com/minetest/minetest/pull/3961
07:09 Obani     Will it be possible to be touched from the head ?
07:18 sofar     possibly, yes
08:23 nore      ok, maybe wait a bit before merging #3948, I will probably change the escape code to \x1b
08:23 ShadowBot https://github.com/minetest/minetest/issues/3948 -- Escape more strings: formspecs, item descriptions, infotexts... by Ekdohibs
08:29 nrzkt2    nore if we can use standard codes for escaping, could be good
08:29 nore      nrzkt2: you mean, like ansi terminals?
08:30 nrzkt2    why not like ANSI, it's a well known and documented thing
08:30 nrzkt2    it's better to follow a standard than invent our standard for this
08:31 nore      yeah, probably
08:31 nore      I should look into its documentation though
08:31 nore      problem is, we do want to add a bit more escape sequences
08:31 nore      (translations for example)
08:34 nore      well, another advantage of doing that is that is should work almost out-of-the-box with terminal_chat_console
08:37 nrzkt2    of course :D yes
08:37 nrzkt2    except for windows where a translation layer should be needed, look at trinitycore or mangos project for mapping colors
08:37 nrzkt2    they have many color mappings for cross platform
08:40 nore      hm, it looks like we can't add other escape sequences easily :/
08:41 nore      well, if we wanted to add more data like informations for translations
08:42 nore      we would need to do things that are very hacky
08:42 nore      and, another problem with ANSI escape sequences
08:42 nore      is that ";" is used as a separator
08:42 nore      but it is already a separator in formspecs :/
11:31 Zeno`     rewriting all my test scripts to deal with changes :)
11:32 Zeno`     I've spent a lot of time looking at everything that's changed since my sabbatical and I have to say I'm very impressed
11:33 Zeno`     great work
11:34 celeron55 where do you even get that information in any reasonable form
11:35 celeron55 the commit log is quite long for reading
11:37 Zeno`     I read nearly every commit
11:38 Zeno`     wasn't much else to do :)
11:38 celeron55 nore: i would imagine ANSI is still among the best options even if when you have to escape the ';'s
11:38 Zeno`     I suppose it's not really "information" but an impression
11:38 nore      celeron55: yeah, but how would you add more escape sequences? (translations for example)
11:39 nore      since you have to write the string to translate, and its arguments too
11:39 celeron55 well i don't think the strings to be translated should contain those
11:40 celeron55 that would be a huge issue no matter what
11:40 nore      celeron55: how would you handle things such as "Furnace is active (42%)"
11:41 nore      the classical way of doing that is to translate "Furnace is active(#1%)"
11:41 nore      and then give it #1
11:41 ShadowBot https://github.com/minetest/minetest/issues/1 -- GlowStone code by anonymousAwesome
11:41 celeron55 and you want to color a single word in it?
11:41 nore      but we can't do that if we use ansi escape sequences :/
11:42 celeron55 can't?
11:42 nore      celeron55: no, what I mean is that the server has to send separately the string to translate and its arguments
11:42 celeron55 i'm not following
11:42 nore      and we can't pass a string as argument in an escape sequence
11:43 nore      sorry, gtg :/
11:44 celeron55 ehm, well i hope someone knows what is going on and prevents some huge mess
11:45 nore      to make it short : imagine the server has to send the text "Furnace is at x%", where x is any value
11:47 nore      the client isn't able to translate that when x is replaced by a value
11:47 nore      but it is able if you say that you have to translate a format string
11:48 nore      and you give the arguments of the format string
11:49 nore      and that means you have to send the argumznts separately
11:50 celeron55 i understand that but i don't understand how this relates to ANSI codes
11:50 celeron55 i guess it doesn't
11:54 nore      well, we want to add escape sequences that can be stripped easily and that can be extended
11:54 nore      so that if we want to add more escape sequences in the future, it can be done without a protocol bump
11:55 nore      and that is the problem with ansi escapes
11:56 nore      they are not extensible
11:56 celeron55 that is kind of a problem, yes
12:02 nore      that's why I used that non-standard format :/ but if you have a better idea, I would be interested
12:11 celeron55 well, i don't care as long as it was considered
12:20 nrzkt2    nore, why not use blizzard's method ?
12:21 nrzkt2    $|f000000 xxx |r
12:21 nrzkt2    if i remember
12:21 nrzkt2    they also added $g $r $n for gender race and number :p
12:21 nrzkt2    name*
12:30 nore      nrzkt: do you have a link?
12:48 nrzkt     http://wowwiki.wikia.com/wiki/UI_escape_sequences
12:49 nrzkt     my syntax was a little bit errored because i worked on in three years ago :) but not many
12:50 nrzkt     the hyperlink mode in wow is also interesting for us :D
12:50 nrzkt     in wow and warcraft (maybe starcraft too) you can type those in caht to add colors directly from end user
13:25 nore      nrzkt: hmm... I'm not sure we should do exactly the same, but it can be used to give some ideas
13:25 nore      the main problem is that their format is not extensible
13:28 nrzkt     what do you mean ? extensible for doing what ?
13:31 nore      I mean, we can't add more escapes in the future
13:32 nore      while being able to remove them now
13:32 nore      also, if we keep the same escape sequences, we might break mods
13:36 nrzkt     i don't understand why ? which escape sequence are we using in 0.4.13 ?
13:38 nore      nrzkt: none
13:39 nore      I added some escape sequence escaping in a recent commit
13:40 nrzkt     but 0.4.14 wasn't released you have time to make a convergent solution
13:40 nore      what I meant was that if we use the same escape sequences as they do, we might break mods that use "|" in strings
13:42 nore      and we can't make that escape sequence format extensible with unknown escape sequences stripped without modifying it
13:42 nrzkt     because MT was so permissive, i think every std escaping model could be used in our mods
13:42 Zeno``    it's great to have reliable internet
13:43 nore      so I'd say we should choose a format, but we should keep in mind it should be easy to add new escapes in it
13:44 nore      that's why I settled to \v(...), but as ShadowNinja said, maybe \x1b instead is better
13:45 nore      so I'll change that, I will also have to find a way for the server to send translations
13:46 nrzkt     create a new packet TOCLIENT_TRANSLATIONS u32 count << std::map<std::string, std::string>
14:27 gregorycu Hey all
14:28 VanessaE  hi
14:29 gregorycu Fixer: How's things
14:49 Fixer     gregorycu, hi, nice, got a nice time outside
14:49 Fixer     gregorycu, now here
14:50 gregorycu Yeah, been busy for the past few days
14:50 gregorycu I'm basically abandoning my stutter work because farmap will probably change that stuff a little
14:50 gregorycu Along with the other work to reduce stutter
14:50 gregorycu Haven't given up
14:51 Fixer     gregorycu, farmap does not solve the jitter/stutter unfortunately
14:52 gregorycu I meant the code will be different
14:52 gregorycu It will be a nightmare merge
14:53 Fixer     gregorycu, it will be nicer if farmap pr adapts to your optimisation one %) if it really helps to fix jitter/stutter
14:54 gregorycu c55 has a hard task already
14:54 gregorycu :)
14:54 gregorycu I can make it compile in your machine
14:54 gregorycu By removing all thread safety :)
14:55 Fixer     if think that way, huge farmap pr will hold a lot of things
14:56 Fixer     gregorycu, can i at least try it?
14:57 gregorycu Of course, give me a few min
14:57 gregorycu You ever think about learning C++?
14:57 Fixer     maybe if it really does help it will convince other devs to think about it
14:57 Fixer     gregorycu, not yet, i'm procrastinating too much while reading python tutorial for years apperently
14:58 gregorycu Yeah, I was doing my fair share of python today
14:58 gregorycu It's one of my languages
14:59 Fixer     i need more effort to read and learn it, everytime some **** goes in the way and distracts me
15:03 Zeno``    hi gregorycu
15:03 gregorycu Zeno
15:04 Zeno``    how are you?
15:05 Zeno``    Is minetest taking longer to connect to servers than it used to, or is it my internet?
15:06 Zeno``    it's hard to tell; my internet has been ship lately
15:09 jin_xi    sup guys just tested if this bug is still present, and it seems it is: id's of expiring particle spawners are never deleted
15:10 jin_xi    PR here for review or redo: https://github.com/minetest/minetest/pull/2602
15:11 gregorycu Fixer: My env is being a piece of shit at the moment, I don't know if I can get this out to you tonight
15:11 Fixer     gregorycu, no problem, if you can, please make a PR for it
15:12 gregorycu Well, I'm not going to make it threadsafe unless it appears to make a difference :)
15:12 gregorycu And i won't do a PR unless it's threadsafe
15:13 Fixer     ok
15:13 Fixer     gregorycu, i will test it and then you will decide
15:18 Zeno``    hmm, that's an old PR
15:19 Zeno``    seems I commented... I wonder if it can be rebased
15:20 Zeno``    jin_xi, have you tested it?
15:20 Zeno``    it looks pretty straightforward and if it fixes the bug then...
15:22 jin_xi    i have tested it when i made the pr, but thats a long time ago.
15:23 Zeno``    I've added rebase needed label, but will either rebase myself and test or add another comment (once I get all my test scripts rewritten)
15:23 Zeno``    from memory it was a real concern (leak)
15:24 Zeno``    bookmarked
15:24 jin_xi    yes it creates a ever growing list of ids which might be problematic if lots of explosions or similar are used
15:26 Zeno``    it looks good at a glance. I'll get it rebased and test
15:30 Zeno``    It probably should have been merged at the time :/
15:30 Zeno``    what's the deal with the feature freeze?
15:31 Zeno``    critical/high-priority bug fixes only?
15:42 Fixer     Zeno``, no new feature but bug fixes are ok (i guess)
15:42 Fixer     Zeno``, ask paramat, he knows
16:09 sofar     to test that, you have to make a mod that creates new particlespawners every N seconds and let it run...
16:10 sofar     Zeno``: I'd love to help test that. I have a torch modification that is a good test case for it too.
16:14 jin_xi    sofar, just manually place spawners with non-zero time and print out id numbers
16:31 nore      I updated #3948 with \x1b as escape character instead
16:31 ShadowBot https://github.com/minetest/minetest/issues/3948 -- Escape more strings: formspecs, item descriptions, infotexts... by Ekdohibs
16:32 nore      the translations code has not yet been updated though
16:42 lisac     nore: Are you there?
16:43 nore      lisac: yep
16:43 lisac     Awesome! I wanted to ask you about the colored chat you are working on.
16:43 lisac     If you have some time.
16:44 nore      of course :)
16:46 lisac     Ok, is there any place where I can get the code?
16:46 lisac     Or could you tell me how do I set a specific part of code to a color?
16:46 lisac     I'm working on a mod, and it is essential to it...
16:47 lisac     See at  #3924
16:47 ShadowBot https://github.com/minetest/minetest/issues/3924 -- Colored item descriptions
16:47 lisac     :D
16:47 nore      lisac: either do minetest.get_color_escape_sequence(color), and concatenate that string
16:48 nore      it will set the color of text to the color you chose
16:48 nore      or do minetest.colorize(color, string), that will set the color to that color and then back to white
16:48 lisac     Ok, thank you!
16:49 lisac     This is going to be fun when it is added :D
16:49 nore      you're welcome: )
16:49 nore      though, you may have to wait a bit before
16:50 nore      it still is WIP, and depends on #3948 being merged in a stable release before
16:50 ShadowBot https://github.com/minetest/minetest/issues/3948 -- Escape more strings: formspecs, item descriptions, infotexts... by Ekdohibs
16:50 lisac     Oh, ok.
16:51 lisac     Also, how did you solve the fact that chat lines are handled as lines, rather than characters?
16:51 Obani     nore, I guess it will be in 0.4.15
16:58 nore      Obani: colored chat, or 3948?
16:59 nore      lisac: what do you mean?
16:59 lisac     Well, last time I checked the code, I think minetest was rendering it as a line?
16:59 lisac     Not characters.
16:59 lisac     I mean...
17:00 lisac     The whole line...
17:00 lisac     Not separated.
17:01 nore      Hmm... actually, the code does render each character separately now
17:02 nore      but there might be a moment where all these chars are just one long string
17:03 lisac     Oh, ok.
17:03 lisac     Guess I was using outdated code :D
17:03 nore      well, the same thing was happening before my changes
17:04 nore      it's just that we add our own modified version of irrlicht code now: )
17:04 nore      (**** that phone autocorrect that breaks my smileys)
17:12 lisac     :D
17:19 Obani     nore, colored chat elements