13:33 < heinrich5991> "unicode" strings?
13:33 < heinrich5991> utf8? utf16?
14:09 <@minus> utf-8
14:27 < rand> i though json text was supposed to be encoded in utf8 (by default)
14:41 <@minus> but there are those silly escape sequences
14:42 < rand> how do you handle utf8 string in C ?
14:43 < rand> in json, for a non ascii character, its '\u' followed by utf8_code 
14:43 < rand> in hex
14:46 < rand> so if utf8 string are integer array, it looks like printf("\\u%04x", utf8_char)
14:47 <@minus> though utf8 can be more than 16bits, right?
14:50 < rand> yes, until 32bits
14:52 < rand> hm, im not really sure about my simple printf
14:55 < rand> after one test, it seems ok :D
15:06 < rand> > http://ix.io/knr/c
15:07 < heinrich5991> rand: is wchar_t encoded as utf8?
15:07 < heinrich5991> or utf16? or utf32?
15:10 < heinrich5991> rand: you probably have to re-encode utf8 sequences that take more than two uint16s for this
15:17 < rand> right
15:17 <@minus> does it make sense to do everything as char (or byte if you so want) in your program, utf-8 encoded, and let the text renderer take care of the rest?
15:19 < rand> as soon as the program don't know what is in its char, it's okay, i guess ^^
15:20 < rand> i mean, if the program don't process char arrays as character but as raw data
15:21 < rand> i think chat messages in tw are handle that way
15:22 < heinrich5991> minus: yes, that makes sense IMO, and it's mostly* how teeworlds treats strings, as opaque byte arrays
15:22 < heinrich5991> *mostly, beecause it doesn't really at a few places
15:25 <@minus> also, char is a silly data type name for what it's used
15:25 < heinrich5991> but char in c is byte
15:25 < heinrich5991> -but
15:25 < heinrich5991> char in rust is codepoint
15:25 < heinrich5991> both not named too well
15:26 < rand> byte arrays are great :D
15:26 < heinrich5991> (although char was right at the time C was developed, so...)