Time Nick Message 02:22 cheapie erlehmann: Could be worse, could be "merging 1 minute ago" 09:53 sfan5 announcing merges here is not actually required 09:53 sfan5 also no I am not up at 3am 10:29 proller so https://github.com/minetest/minetest/pull/11910 can be merged right now without announce? 10:30 MTDiscord I don't think so, no 10:35 erlehmann i am only a troublemaker with an opinion, but maybe, just maybe, you could have some regular (weekly?) merge meeting or so? 10:36 erlehmann proller have you read the “what we think we reserve” paper on _t suffix and stuff? 13:17 proller erlehmann, https://github.com/minetest/minetest/pull/11843#issuecomment-987653974 13:19 erlehmann proller yes but have you read the paper 14:20 proller erlehmann, how reading will help merge process? 14:49 erlehmann that's not a yes. so i guess it is a no. 16:49 sfan5 pushing http://sprunge.us/K0qsTm?diff soon 16:58 erlehmann sfan5 are you expecting the assert to be dead code eliminated? i wonder 16:58 erlehmann i mean it is useful for future programmers 16:59 sfan5 I don't, is that relevant? 17:00 erlehmann just to figure out your intention. like, is this a hint for programmers or a safeguard for runtime errors. 17:01 erlehmann if you do not expect it to be eliminated ig you suspect a case in which the assertion can fail exists 17:01 erlehmann correct? 17:01 sfan5 no 17:02 sfan5 it's for programmers and static analyzers to make clear that we didn't forget to check layer->texture != NULL at the bottom when such a check exists at the top 17:02 erlehmann i see. but then dead code elimination should take care of it unless i am missing how that works. 17:03 erlehmann thanks! 20:52 Sokomine short question: how can i increase the amount of the cached chat messages that are displayed? 20:54 sfan5 https://github.com/minetest/minetest/blob/f2d1295fe646105f1b98b0c204f47f781336e211/src/chat.cpp#L727-L729 this looks like it 20:56 Sokomine sfan5: thank you for the quick help! i was looking at the wrong places 20:59 erlehmann Sokomine, can you make it a setting maybe? 21:00 Sokomine erlehmann: hmm. that would be a very good idea. i might not be the only one wishing that. i also had people asking for logging chat. sadly my solution is anything but elegant, and even though it's just one line, it might be even more difficult for general users to add 21:01 erlehmann logging chat sounds like a CSM issue tbh 21:01 Sokomine no, not really. there are servers that forbid csms. so it can't be done that way 21:04 erlehmann can't really forbid that lol 21:05 Sokomine the csms are forbidden for other reasons than logging chat. forbidding chat mods that have local effects only and don't i.e. allow to cheat by looking where valuable ores are is a side effect 21:06 erlehmann Sokomine can you show me such a sernver? haven't seen one so far. default CSM restictions seem to disallow a lot of stuff though. 21:06 erlehmann such a server i mean 21:06 Sokomine it might be very helpful for other players to have a larger chat buffer, chat logging and perhaps even timestamps. while it now (still compiling) may work for me, my solution isn't good enough. i fear my skills at cpp/c are too low to do it really well 21:07 erlehmann hmm, how about what if 21:07 Sokomine erlehmann: e.g. your land 21:07 erlehmann chat would be just appended to another file 21:07 Sokomine exactly 21:07 Sokomine right now i simply log it to debug because that was most convenient and because i always have to manually merge it into newer versions 21:07 erlehmann make a PR, if it is any good, it will improve 21:08 Sokomine sadly it's not good yet. the pr would be a request to write it much better i'm afraid 21:08 erlehmann yes but the requests might contain detailed advice 21:09 Sokomine ah! ok. sounds like a plan 21:09 erlehmann :) 21:09 erlehmann i see code reviews as knowledge transfers 21:09 erlehmann also timestamps are important 21:10 Sokomine hm. perhaps such an (as such) small task may help there, yes 21:11 Sokomine right now, it's a single line: actionstream << "[Chat] " << wide_to_utf8(chatMessage->message) << std::endl; 21:11 Sokomine that at the right place does the job, and post-processing can happen through shell tools. as you can see...lots of work needed 21:13 MTDiscord I believe chat logging should definitely be handled by logs. 21:13 MTDiscord s/logs/mods 21:14 MTDiscord out of privacy concerns alone and chat channels / PMs which are implemented by mods 21:14 Sokomine can't be done if mods are forbidden server-side and then not loaded at all. of course people might patch their client to ignore that, but...that really ought not to be the goal 21:15 Sokomine we can disregard privacy issues. sadly nobody cares about them. you're even writing from the discord bridge, and servers often use such gateways. so...privacy, nope, i would care, but hardly anybody else cares. besides, it *is* for a private log anyway 21:25 MTDiscord hehe, the discord bridge definitely weakens my point 21:25 MTDiscord but for public communications, I don't expect privacy anyways 21:25 MTDiscord (I do consider it reasonable to log public chat) 21:26 erlehmann logging should not be on by *default* 21:26 MTDiscord Still, this shouldn't be hardcoded and forced upon admins. Mods can do this just as well with more fine-grained control than a bunch of crappy settings. 21:26 MTDiscord Also note that storing such logs may be problematic for GDPR compliance or the like 21:27 Sokomine sure, mods *could* do that. if they were loaded. are they now? if client-side modes are forbidden? 21:29 erlehmann forbidding clientside mods to log chats should not be allowed for servers :( 21:30 Sokomine i agree, erlehmann. needs more fine-grain there 21:30 Sokomine or, rather: a server forbidding all should be ignored in that part. not sure if that's more work to implement than just adding logging 21:31 erlehmann i mean there is other stuff that a server has absolutely no business interfering with 21:31 erlehmann like the rumble and mumble CSMs 21:31 Sokomine what does that do? 21:32 Sokomine really need to catch up with the forum again 21:32 erlehmann the mumble CSM gives 3d positioning data to mumble, a voice chat client, so other players can sound like they are in the game 21:33 Sokomine perhaps i can at least add a setting for the chat cache size where sfan5 pointed me to. will test now. if it works fine a pr for a setting may be doable 21:33 erlehmann the rumble CSM makes your hardware vibrate if you get damaged. do you have a gamepad? we still need testing. 21:37 Sokomine ah, sorry, no. sounds intresting, but is both something i can't help with testing. my second speaker broke, and i didn't bother to buy a new one. sometimes a bit irritating when sound has a direction when turning around next to water. most of the time i have sound off. and no gamepad in sight... 21:38 Sokomine you're right: both those things are things that affect local playing only, and there's no point in forbidding such csms 21:41 erlehmann i actually think that minetest should go the extra mile and just deliver a bunch of tried and tested CSMs with the engine 21:41 erlehmann the ones where even the most scared admin understands that they are not useful for cheating 21:42 erlehmann and rumble, mumble and chat logging are kinda such things 21:44 Sokomine such mods ought to be loaded regardless of server settings 21:45 Sokomine most servers just want to forbid cheating (finding ores too easily) or scripted building or the like 21:51 erlehmann i think server admins who think that all CSMS are cheating are not only misinformed about CSMs, they are misinformed about minetest 21:51 erlehmann Sokomine what do you think of the idea of having a taint flag in lua, like WOW client scripting? 21:51 erlehmann basically, specific actions are only allowed without taint 21:52 erlehmann and you acquire taint by other actions 21:54 Sokomine erlehmann: i have no idea what wow clients do. what would trigger such a flag? 21:59 erlehmann Sokomine https://forum.minetest.net/viewtopic.php?p=407658#p407658 21:59 erlehmann How WoW did Lua taint: https://www.lua.org/wshop08/lua-whitehead.pdf