Time Nick Message 02:26 * ANAND peeks 02:27 ANAND Hello ;) 09:19 shadowmaker Hello. Just a lil question ;) If downloaded the source code of last version because I want to add a fixed cross on the center of the screen while playing. 09:20 shadowmaker I just wanted to know if there's a reason why the texture folders don't have a .nomedia 09:23 shadowmaker Because when you decompress the zip in a mobile device the gallery creates all its thumb stuff and the images have near no weight, but they take the same room as a 512kb photo (I think that's the cluster space I've got in sdcard). As long as the gallery creates a "copy 09:23 shadowmaker It just weights 100mb more than the weight a was specting. 09:36 rubenwardy Already fixed 09:37 rubenwardy You'll need to clear the app storage for it to reextract 09:37 rubenwardy You may need to restart your phone for it to notice the .nomedia 09:37 shadowmaker Oh. I did it yesterday when I downloaded it 09:38 rubenwardy The .nomedia is at the root of the minetest folder 09:38 shadowmaker There's no need to restart. Just delete .thumnails in DCIM 09:38 rubenwardy File system issues means that the .nomedia file may not be recognised without a restart 09:38 shadowmaker Oooh. Okey, thats why it's not in the open source. Jeje 09:39 shadowmaker Thanks ;) 09:39 rubenwardy Oh right, you expected it to be in the source code 10:00 sfan5 merging #10052 10:00 ShadowBot https://github.com/minetest/minetest/issues/10052 -- Android: fix TMPFolder path by MoNTE48 12:00 rubenwardy Argh, this is so stupid 12:02 rubenwardy So, whilst the GPL does require a copy of the license to be included, putting it in the assets folder doesn't necessarily make sense, especially if the assets folder is moved to the app's storage in the future 12:03 rubenwardy This feels like a non issue not worth my time 12:04 rubenwardy #10073 12:04 ShadowBot https://github.com/minetest/minetest/issues/10073 -- Android: Fix asset upgrade by renaming license file by rubenwardy 12:04 rubenwardy Tbh, I doubt anyone would actually complain if the file was just removed 12:12 rubenwardy Merging #9943 in 10 12:12 ShadowBot https://github.com/minetest/minetest/issues/9943 -- Add warning when disabling secure.enable_security by rubenwardy 12:22 rubenwardy Do we need a meeting today? 12:23 rubenwardy I guess it's all under control 12:23 rubenwardy Minus bone issues 12:23 sfan5 bone issues? 12:24 rubenwardy There's a whole thing about bone animations breaking mods 12:24 rubenwardy I've been keeping away as models and animations is an intersection of horror 12:39 sfan5 so far only playeranim has been reported to be broken 12:51 rubenwardy Ok, I'm confused. Looks like get_translated_string already strips the translation if given an invalid, "en", or "" langcode 12:53 sfan5 not surprising considering you don't want the escapes to show up visually in the client 13:15 rubenwardy #10075 13:15 ShadowBot https://github.com/minetest/minetest/issues/10075 -- Clean up server-side translations, remove global variable by rubenwardy 13:15 rubenwardy Not for 5.3.0, obviously 13:17 sfan5 The engine could really use a function to read a file into a string 13:18 sfan5 there are slightly different solutions scattered everywhere 13:18 sfan5 in this case: std::string data((std::istreambuf_iterator(t)), std::istreambuf_iterator()); 13:18 rubenwardy Yeah 13:23 rubenwardy Done 13:23 rubenwardy I should search for uses of ifstream I guess 13:24 sfan5 there's also no error handling in that function 13:24 sfan5 and I think what Server::loadMediaFile does is more performant 13:25 sfan5 and the ifstream should be std::ios::binary 13:28 rubenwardy It's text, not binary 13:29 sfan5 other uses of std::ifstream are often binary 13:30 sfan5 and IMO instead of relying the OS to do some line-ending conversion(?) it would be better to handle \r\n in the code that decodes the data 13:30 sfan5 alternatively the function could have a bool argument for text vs. binary 13:31 rubenwardy Will there be any problems with reading a text file in binary? 13:32 rubenwardy Text encoding is black magic 13:32 rubenwardy Ideally you'd want it to accept most common file encodings and return utf8 13:34 rubenwardy Also, are you ok with #10073 sfan5? 13:34 ShadowBot https://github.com/minetest/minetest/issues/10073 -- Android: Fix asset upgrade by renaming license file by rubenwardy 13:34 sfan5 yes 13:35 rubenwardy I'll merge in 10 then 13:35 sfan5 rubenwardy: AFAIK the only thing text vs binary mode does it convert the line endings and only on Windows(!) 13:36 pmp-p what about byte order mark ? 13:37 sfan5 utf-8 doesn't need a BOM and utf-16 requires the application to specifically handle the encoding anyway 13:38 rubenwardy The code in server add media file doesn't reserve a size in the string 13:39 rubenwardy I guess it can't tell the size of the file beforehand? 13:39 pmp-p sfan5: i know but windows does not know it https://en.wikipedia.org/wiki/UTF-8#Byte_order_mark 13:40 Krock rubenwardy: well, it could know that.. it's just that nobody did that pre-allocation 13:41 rubenwardy It appears to work 13:42 rubenwardy Wait, there's a compile error. How did it work?! 13:42 rubenwardy Don't tell me android doesn't do unit dependencies correct 13:44 pmp-p it does not even link correctly so i would wary on that :D 20:14 Krock rubenwardy: field_close_on_enter was never in a stable release, but the warning outlived it 20:14 Krock precisely: the warning existed until 5.3.0-dev 20:14 Krock Déjà vu. This was already brought up once 21:28 Wuzzy what are you talking about? field_close_on_enter is there for a long ime 21:28 Wuzzy have used it in MCL2 for a good while 21:29 Wuzzy Krock: ↑ 21:29 Wuzzy long before 5.2.0, in fact 21:30 Krock Wuzzy: I meant the feature which used an additional param in some elements 21:30 Wuzzy ok 21:30 Krock I'm well aware of the dedicated formspec element which was introduced as a follow-up