Time  Nick          Message
02:36 Megaf_        us_0gb, why your IRC nick is us_0gb and everywhere else is 0gb-us?
02:38 us_0gb        Megaf_: I never use a dash unless I have no other option. "0gb.us" is my preferred name, but when dots are not allowed (such as on IRC), I use "0gb_us", and when leading digits are not allowed (also such as on IRC), I use "us.0gb". The combination of these limitations on IRC names results it the name "us_0gb".
02:39 us_0gb        So ... a lack of name legality standards, to put it simply.
02:44 Megaf_        I mean, 0gb*us vs us*0gb
02:44 us_0gb        It's because some platforms do not allow leading digits. Sadly, my own operating system also has this issue.
02:45 Megaf_        hm
02:45 Megaf_        ok
02:52 ShadowNinja   TS servers reserve identifiers starting with a number for SIDs and (E)UIDs.  But -> #minetest.
03:28 ShadowNinja   Any progress on libmtmap?  Perhaps it should be written in C++, since we'll have to implement a libmtmapc++ if not.  That will mean no python libraries though, I think.
03:28 ShadowNinja   kahrl: ^
03:44 hmmmm         there is literally no reason why it should be in C++
03:44 hmmmm         that only serves to make it less flexible
03:57 ShadowNinja   It makes it harder to use.  Minetest currently uses classes for a lot of map objects.
03:58 ShadowNinja   We'd need to make libmtmapc++, depending on libmtmap and providing a C++ interface.
03:58 ShadowNinja   Or change a LOT of code to use global functions and structs.
04:08 hmmmm         you misunderstand the point of libmtmap
04:09 hmmmm         there is no reason why it has to create MapBlock classes and do exactly everything that is currently done the way it is
04:09 ShadowNinja   I understand it as a easy way for programs to interact with a MT DB, regardless of it's format.
04:09 hmmmm         it just needs to provide a consistent interface on how to access data from the map
04:10 hmmmm         if you think that you're not able to do this with a plain C interface, then you haven't been thinking hard enough.
04:11 ShadowNinja   Hmmm, I propose a second lib that builds on it then.  Such structures and classes will be usefull in the C++ mapper.
10:29 RealBadAngel  hi darkrose
10:30 darkrose      hi
10:30 RealBadAngel  are you doing anythin for the project?
10:32 darkrose      I'm working on -classic, some of my improvements are being backported to 0.4 by others... so kind of in a round about way
10:33 RealBadAngel  so gamemode
10:33 RealBadAngel  imho we shall make a few more games aviable
10:34 RealBadAngel  but not like before
10:34 RealBadAngel  something different should be added, like cornernote's skyblock
10:38 RealBadAngel  btw cornernote is back and said that he will start workin on SkyBlock2
10:39 darkrose      skyblock?
10:40 RealBadAngel  ure not saying that...
10:40 RealBadAngel  u rly dont know what skyblock is?
10:41 darkrose      I haven't touched 0.4 in over year, so no
10:41 RealBadAngel  SHAME on u ;)
10:42 RealBadAngel  skyblock is fun, an island in the sky, tiny
10:42 RealBadAngel  you have to build from
10:42 RealBadAngel  with limitations and goals
10:42 RealBadAngel  and being extremaly careful all the time, to not fall down
10:44 RealBadAngel  in such world cobble is rare resource
11:04 Jordach       RealBadAngel, nope
11:04 Jordach       (i had several stacks on my map naturally dug up)
20:02 sapier        I'm merging #1116 in about 15 minutes so last chance for comments prior merhe
20:02 ShadowBot     https://github.com/minetest/minetest/issues/1116 -- Add proper client initialization by sapier
20:07 ShadowNinja   sapier: Have you removed max-concurrent-connections or given it a slightly higher default?
20:16 sapier        of course
20:19 ShadowNinja   So, which one?  :-)
20:20 sapier        ok, hope noone mentioning any real issue means there aren't any ... but I think it means more "noone did look at it" ... I removed it because I don't have any interest in discussing if 2 is better then 3 100 is better then 101 ...
20:21 sapier        guess the issues will have to pop up once it's merged
20:21 sapier        merging NOW
20:21 ShadowNinja   8 or so seems like a reasonable limit.  It seemed like the function of the setting was misunderstood.
20:22 ShadowNinja   You can add that later though...
20:22 sapier        usually I drop features which noone likes and never return them ... I have valuable things to do then argue about numbers
20:23 ShadowNinja   sapier: s/pp/p/ :-(
20:23 ShadowNinja   Can you update the changelog now?  :-)
20:23 sapier        lol ... usually typos are the first thing to be found
20:24 sapier        no I don't do history rewriting, I don't have a problem with others knowing I can't write words correct ;-)
20:26 sapier        I'm more concerned about the side effects of conlock removal. I never did test a patch that hard then this one (almost a full week of continous testing) but I have a limited environment so there's still some risk remaining.
20:26 ShadowNinja   sapier: Can you check #862 and #1106 after updating the changelog?
20:26 ShadowBot     https://github.com/minetest/minetest/issues/862 -- Add the option to bind to a specific address by ShadowNinja
20:26 ShadowBot     https://github.com/minetest/minetest/issues/1106 -- Codepages support (utf-8) for all platforms
20:27 sapier        862 will be done after the skybox thingy
20:27 ShadowNinja   c55's patch?
20:28 PilzAdam      sapier, is it intentional that a local game starts in under a second now?
20:28 sapier        yes
20:28 sapier        #1106 is not a patch and I already burnt my fingers on i18n spening almost 2 FULL weeks on fixing it I decided not to touch this topic again
20:28 ShadowBot     sapier: Error: ProcessTimeoutError: Process #384 (for String.re) aborted due to timeout.
20:29 sapier        String.re?
20:29 sapier        what's that?
20:30 sapier        ohh ... what did shadowbot find there?
20:30 ShadowNinja   The re command in the String plugin. ,,(help re)
20:30 ShadowBot     ShadowNinja: (re <regexp> <text>) -- If <regexp> is of the form m/regexp/flags, returns the portion of <text> that matches the regexp. If <regexp> is of the form s/regexp/replacement/flags, returns the result of applying such a regexp to <text>.
20:30 sapier        Pilzadam under a second?
20:31 ShadowNinja   I use it to get only the first section of the issue title.
20:31 PilzAdam      sapier, yea
20:31 sapier        well not intentionall but I don't understand if this is an issue?
20:31 sapier        do you want it to be more slow?
20:32 PilzAdam      its just a way of saying "good work"
20:32 sapier        ohhhh :-) well thank's .... guess I'm not used to positive feedback any more
20:35 sapier        well there's one last thing to do to make minetest core architecture capable of using multicores
20:36 sapier        besides fixing the threading bugs that are burried deep inside of course
20:38 ShadowNinja   Do it. :-D
20:39 sapier        not now, I'm gonna focus more on mobf the next weeks/months, yet I'll review and merge pull requests
20:45 sapier        ShadowNinja what's the intention behind 1129?
20:46 ShadowNinja   sapier: Capturing the output of commands.
20:47 sapier1       what's the "success" variable good for?
20:47 ShadowNinja   For example you could do nested commands.
20:47 sapier1       can you give an example about a nested command?
20:47 ShadowNinja   For a &&-like operator.
20:48 sapier1       && in chat command? are you serious?
20:48 ShadowNinja   Well, ShadowBot has them.  for example ,,(mp list --id 1)
20:48 ShadowBot     ShadowNinja: messageparser list [<channel>|global]
20:49 ShadowNinja   Er, ,,(mp show --id 1)
20:49 ShadowBot     ShadowNinja: The action for regexp trigger "(?:^| )#([0-9]+)\b" is "echo <issue $1> -- <re m/[^·]+/ <web title <issue $1>>>"
20:49 sapier1       do you really think anyone will use those things?
20:49 ShadowNinja   That's the issue number command.  <> are used for nesting.
20:50 sapier1       well doesn't matter if you want them I guess they won't harm, but you should remove the unused success variable
20:50 ShadowNinja   Possibly not, but it makes that possible.  What I really want it for is just the output capturing to send back to IRC or show in a commandblock formspec.
20:51 sapier1       no it doesn't all you seem to need from minetests point of view is the return text. which imho isn't something to be done there too but I don't care that much about it. your chat commands can nest with parameters returned the other way round too
20:51 ShadowNinja   I can see /protect <mypos> <mypos +1 +3 +1> or something like that.
20:53 ShadowNinja   Yes, the order doesn't matter, but success, output seemed better.
20:53 sapier1       unused variables are ugly even if they're local. Maybe that's a c coders view but I guess this applies to lua too somehow
20:53 ShadowNinja   You could also disable a commandblock if it failed, or show a red inducator.
20:54 sapier1       you could do everything but is this really usefull to have in basic features?
20:54 ShadowNinja   I guess that's a valid reason to reverse it.
20:55 ShadowNinja   Better to have the API available for advanced things and not use it then to need such an API and not have it.
20:56 sapier1       I'm absolutely positive about concatenation of help messages. But I'm not convinced about adding chat messages as basic handling feature of chat handlers. It's not a task to hard to handle within command handler itself
20:57 sapier1       It's not better to have an api handling everything because it's hard to see what really happens
20:58 sapier1       well that's my opinion, no more no less
20:58 ShadowNinja   The alternative is to override minetest.chat_send_player before and restore after the command (A string in a table should work).  And there's no way to check success.
20:59 sapier1       do you really expect users not to call chat_send_player in a command handler?
20:59 ShadowNinja   Another advantage is that it's all sent in one message, unless Minetest splits it server-side or combines them.
21:00 Sokomine      sapier: good to hear that mobf will progress further :-) your network fixes will certainly help as well
21:00 sapier1       I already said that I'm positive about it. But I know about modders they will continue to use chat handler
21:00 ShadowNinja   They can call chat_send_player, but text sent to the caller should be returned.
21:00 sapier1       And I veto against not sending messages on calling it
21:00 sapier1       it's crap to have a function behave different dependent on WHERE you call it
21:01 ShadowNinja   sapier1: What do you mean?
21:01 sapier1       your override idea
21:01 ShadowNinja   What?  The override is the hack that you need without this patch.
21:02 sapier1       we're talking about different things right now
21:02 * ShadowNinja isn't sure what he's talking about anymore
21:02 Sokomine      what do you intend to do to chat_send_player anyway?
21:02 ShadowNinja   I'm talking about my patch and how it's beneficial.
21:03 sapier1       Maybe first find a common base for talking. I've seen three things in your patch 1) add two new return values success and text
21:03 ShadowNinja   Sokomine: Nothing.  Avoiding a hack involving it.
21:03 sapier1       2) text is sent to player instead of command handler sending it itself
21:03 sapier1       and 3) consolidation of help messages
21:03 sapier1       did I miss something?
21:04 ShadowNinja   If by 3 you mean that it's all send in one chat_send_player, then yes.  That's about all it does.
21:05 ShadowNinja   I might have fixed a typo or two or fixed some bugs/cleaned up some code while writing it though.
21:05 sapier1       well ... maybe the only thing missing is evaluation of success in main chatcommand handler
21:06 ShadowNinja   I don't think there's anything for it to do with that value though.
21:07 ShadowNinja   Adding Error: or something like that maybe.
21:07 sapier1       the only thing I can imagine is do a log entry ... but if it's that hard to find a use for it should be removed
21:08 ShadowNinja   You proposed reversing the order.
21:08 sapier1       yes in conjunction with dropping it where not needed
21:09 sapier1       no need to pass a value if it's not used
21:09 sapier1       if you have custom commands returning more parameters they wont fail on reversed order
21:10 sapier1       and we don't bloat api with superfluss parameters
21:13 sapier1       well wait for some other persons opinion
21:17 proller       sapier1, please show all you pulls to celeron55 before merge
21:17 sapier1       can someone test this on windows https://github.com/minetest/minetest/pull/1130
21:17 sapier1       proller please just be quite ;-P
21:17 proller       and recieve his agreement about every line
21:17 proller       please dont commit shit
21:19 sapier1       proller I'm really anoyed about your generall oposition without spending even a single thought. If you keep talking shit only here I may really have to ignore you, a thing I never ever have done to anyone.
21:19 ShadowNinja   Excuse me for commiting the falacy of tu tuoque...  ;-)
21:20 ShadowNinja   He'd be ID 1 on my list, if I was sure I could get the backlog later of I needed to.
21:20 ShadowNinja   if*
21:21 ShadowNinja   I guess I could use irc.minetest.ru, but I'd like to haveit in my DB.
21:22 sapier1       I'm thankfull for any found bug in my commits. I'm sure there are bugs in, but it's useless to have to talk about nonsense every day.
21:23 ShadowNinja   sapier1: That looks good, if the asignments are atomic, I think.  I know very little about what's threadsafe and what isn't.
21:24 sapier1       it's based uppon a suggestion of kahrl some days ago. Even if those aren't atomic they wont cause real issues
21:25 sapier1       the worst thing to happen may be a thread doing another loop while "stoprequest" is already beeing handled ... well if the stop request would've been issued 1 instruction later the thread would do that loop too
21:33 ShadowNinja   I had an idea for case-insensitive player names that still allows players to use capitalization. (Like IRC)  Simply deny access if there is another player with the same (case-insensitive) name in the list and you are not it.  This would select the first case-insensitive player name in the list as the main one.  There might be some users that are not first in the list but should control the account, this will have to be sixed manually
21:33 ShadowNinja   by server admins.  This would also fix the ugly directory iteration hack since case-insensitive filenames will work fine.
21:33 ShadowNinja   That way case sensitive comparisons that mods do will still work properly.
21:34 ShadowNinja   The compat may be a bit ugly, but it should work fairly well.
21:34 VanessaE      +1
21:36 VanessaE      also, sapier's patch has been staged for my servers (and deployed to 30003 for early testing, the others to follow tomorrow at 1:00pm UTC)
21:36 VanessaE      s/patch/commit/
21:44 xyz           ShadowNinja: ugh, ignoring other devs is no good at all
21:48 ShadowNinja   xyz: At this point he might as well not be a dev.  He hasn't contributed anything usefull recently, in either code or discussion.
21:49 ShadowNinja   I've reversed the order of the arguments on my pull.
21:49 xyz           ShadowNinja: same could be said about many other people who're devs
21:49 xyz           ShadowNinja: I think you're cherry-picking
21:51 xyz           okay, maybe not many
21:51 ShadowNinja   xyz: They haven't been actively unhelpfull though.
21:51 xyz           what do you mean?
21:52 ShadowNinja   I can't remenber proller saying anything halpfull within the last few weeks.
21:52 xyz           I think his words to not merge pull requests because he thinks they're shit are helpful, no?
21:52 ShadowNinja   But he trolls, and annoys other devs.  This dicsussion isn't really worth it though.
21:54 xyz           I see
21:55 xyz           maybe it'd be better if we left
21:56 ShadowNinja   You're still helpfull, but proller isn't.
21:58 VanessaE      xyz, I can't remember the last time I saw proller actually make a constructive comment in here that both included a pull request or a helpful tip, and didn't also include some snarky remark, attempt to troll, or outright "NO!  REVERT!!!" type of comment.
21:58 VanessaE      and just about every pull/commit sapier has made regarding the network code, for example, has led to a net improvement in performance
21:59 VanessaE      (pun honestly not intended)
21:59 xyz           ShadowNinja: I don't agree with the course of action and this drama shit; I don't think mt will lose much anyway, there are still people much smarter than me
22:00 PilzAdam      I see it the same way as ShadowNinja and VanessaE
22:01 VanessaE      wat?  PilzAdam agreeing with me? O_O
22:01 VanessaE      :)
22:01 VanessaE      everyone else here at least tries to work through the problem in some sane manner, even if we argue.
22:06 VanessaE      so I don't mean to shit on proller, as he can obviously code...
22:07 VanessaE      but how is it that you usually put it, xyz?  "don't shitpost"?
22:10 xyz           VanessaE: maybe
22:22 ShadowNinja   Does this look better?  http://pastebin.ubuntu.com/6851883/
22:23 PilzAdam      no
22:23 ShadowNinja   PilzAdam: What's better/what's worse?
22:23 PilzAdam      it looks the same as the current one
22:24 ShadowNinja   Well, it's just some tweaks, not a rework.
22:24 PilzAdam      its longer
22:24 PilzAdam      thats not really good
22:24 ShadowNinja   It's far under the length limit.
22:26 PilzAdam      I dont really care
22:45 celeron55     the topic is intentionally as short as possible so that there is a larger chance that people actually notice what it contains
22:46 celeron55     if you can make it shorter without removing information, that would be good