|
Members in Shoutbox
None.
Shoutbox Search
Shoutbox Commands
/w [name] > Whisper
/r > Reply to last whisper /me > Marks as action Shoutbox Information
Moderators may delete any and all shouts at will.
|
Global Shoutbox
Please log in to shout.
[2018-7-11. : 1:35 am]  O)FaRTy1billion[MM] -- NudeRaiderNudeRaider shouted: \r = carriage return - used in electronic typewriters, for a pc (afaik) this moves the cursor to the start of the line. \n is newline (self explanatory), why you need a \r before a \n I'm not sure, but I think it's used to mark a new paragraph whereas \n alone is just a new line. no, for pc it means you're using Windows and it just arbitrarily has that but if you don't have that then all of your shit is broken[2018-7-11. : 12:04 am]  Lanthanide -- NudeRaiderNudeRaider shouted: Suicidal Insanity but \n doesn't move the cursor down, it creates a new line. By default new lines are empty, so it makes no sense to have a cursor anywhere other than the beginning, or otherwise if you were to start writing characters in the middle of the new line, how would that be represented in memory? For the char to appear in the middle of the line it would have to pad with spaces or something. But there is nothing telling it to pad. So it would actually write other things than there are in memory. Tbh that doesn't sound like what it's doing. you're assuming that the data is being written to memory. It may just be displayed on the terminal, until an action occurs that writes it to memory (perhaps a "page down" key press), at which point the text shown on the screen is serialised to memory/disk[2018-7-10. : 11:47 pm] Lanthanide -- Mini Moose 2707Mini Moose 2707 shouted: Lanthanide Lanthanide Having used a number of consoles, I haven't seen any implemented that way yet. It would be interesting to see if there are any that had widespread use (ie. non-academic/egdy/pedantic purposes) I'm thinking of like 60s and 70s tech here[2018-7-10. : 11:32 pm] Moose -- Seen typewriters where \r vs \n is important, that's about it. (and I think the ones I've used still had a key that combined them)[2018-7-10. : 11:31 pm] Suicidal Insanity -- http://photos.arlinghaus.org/Animals/Dierenrijk/2018.07.10-ESD_8142-QHD-%7B11B1B8E1-632C-4467-A1CE-C03CB881BDCF%7D.jpg <- Kitty ![2018-7-10. : 11:31 pm] Moose -- LanthanideLanthanide shouted: \n is treating output like normal writing - you start the beginning of the next line, whereas \r\n is treating the output like a position grid of text LanthanideLanthanide shouted: \r\n makes sense if you have fixed width displays that are always 80 characters (or whatever other fixed number), but otherwise doesn't make any sense Having used a number of consoles, I haven't seen any implemented that way yet. It would be interesting to see if there are any that had widespread use (ie. non-academic/egdy/pedantic purposes)[2018-7-10. : 11:31 pm] Suicidal Insanity -- http://photos.arlinghaus.org/Animals/Dierenrijk/2018.07.10-ESD_7965-QHD-%7BD4BC31BA-A040-4CDB-95DA-E42174384325%7D.jpg <- Fly ![2018-7-10. : 10:49 pm]  NudeRaider -- NudeRaiderNudeRaider shouted: Suicidal Insanity but \n doesn't move the cursor down, it creates a new line. By default new lines are empty, so it makes no sense to have a cursor anywhere other than the beginning, or otherwise if you were to start writing characters in the middle of the new line, how would that be represented in memory? For the char to appear in the middle of the line it would have to pad with spaces or something. But there is nothing telling it to pad. So it would actually write other things than there are in memory. Tbh that doesn't sound like what it's doing. *doesn't only move the cursor down, it creates a new line AND moves the cursor there.[2018-7-10. : 10:48 pm] NudeRaider -- Suicidal InsanitySuicidal Insanity shouted: NudeRaider One moves the cursor one line down( so +1 Y), one all the way to the left (so x = 0). Hence you need both but \n doesn't move the cursor down, it creates a new line. By default new lines are empty, so it makes no sense to have a cursor anywhere other than the beginning, or otherwise if you were to start writing characters in the middle of the new line, how would that be represented in memory? For the char to appear in the middle of the line it would have to pad with spaces or something. But there is nothing telling it to pad. So it would actually write other things than there are in memory. Tbh that doesn't sound like what it's doing.[2018-7-10. : 10:01 pm] Lanthanide -- \r\n makes sense if you have fixed width displays that are always 80 characters (or whatever other fixed number), but otherwise doesn't make any sense[2018-7-10. : 10:01 pm] Lanthanide -- \n is treating output like normal writing - you start the beginning of the next line, whereas \r\n is treating the output like a position grid of text[2018-7-10. : 9:59 pm] Lanthanide -- jjf28jjf28 shouted: Suicidal Insanity is this like in the console? yes[2018-7-10. : 9:55 pm] jjf28 -- Suicidal InsanitySuicidal Insanity shouted: NudeRaider One moves the cursor one line down( so +1 Y), one all the way to the left (so x = 0). Hence you need both is this like in the console?[2018-7-10. : 9:49 pm] Suicidal Insanity -- NudeRaiderNudeRaider shouted: \r = carriage return - used in electronic typewriters, for a pc (afaik) this moves the cursor to the start of the line. \n is newline (self explanatory), why you need a \r before a \n I'm not sure, but I think it's used to mark a new paragraph whereas \n alone is just a new line. One moves the cursor one line down( so +1 Y), one all the way to the left (so x = 0). Hence you need both[2018-7-10. : 5:27 pm] jjf28 -- most browsers don't respect line ending characters and demand <br /> or other formatters[2018-7-10. : 5:23 pm] jjf28 -- \r\n = standard windows line ending , some other systems use just \r or just \n (just \n is preferable wherever possible), some even use form feeds or vertical tab nonsense[2018-7-10. : 2:04 pm] NudeRaider -- \r = carriage return - used in electronic typewriters, for a pc (afaik) this moves the cursor to the start of the line. \n is newline (self explanatory), why you need a \r before a \n I'm not sure, but I think it's used to mark a new paragraph whereas \n alone is just a new line.[2018-7-10. : 2:10 am] O)FaRTy1billion[MM] -- or maybe like a newline that doesn't actually move it to a new line[2018-7-10. : 2:09 am] O)FaRTy1billion[MM] -- if I'm reading this right, it looks like <1A> is meant to be a zero-width space?[2018-7-09. : 11:42 pm] Lanthanide -- Suicidal InsanitySuicidal Insanity shouted: I was told that SC should use UTF8 for EUD maps, and if I have an example where that does not work I should forward it. That is why I was asking about the EUD + unicode space issue a week or so ago. if you have an EUD map, then SC won't display unicode characters (like quad M space or whatever)[2018-7-09. : 11:10 pm] O)FaRTy1billion[MM] -- I guess all the used fonts are 33 to 255, with 127 to 160 being weird things[2018-7-09. : 11:04 pm] O)FaRTy1billion[MM] -- er, I meant most because not all of them have every character after 0x80[2018-7-09. : 11:02 pm] O)FaRTy1billion[MM] -- NeivNeiv shouted: Suicidal Insanity Aren't the 1.16.1 mpq font files just 256 characters in 1252 order? Korean text in 1.16.1 gets an exception and is rendered through GDI but I thought otherwise the handmade fonts were used most of the fonts only include characters >0x20 and a few just after 0x80 are a bit messed up, but otherwise yes[2018-7-09. : 8:19 pm] Suicidal Insanity -- NeivNeiv shouted: Suicidal Insanity If they do that then it seems to be pretty bad heuristic.. Every map that I've seen to use high ascii 1252 characters, even if it's only one á or í for ~style~, doesn't get drawn with 1252 in SCR It is a fairly simple function. Doesn't UTF-8 set the high bit to indicate a UTF sequence? That probably is messing it up, since that may get misdetected as a Korean sequence.[2018-7-09. : 8:18 pm] Suicidal Insanity -- NeivNeiv shouted: Suicidal Insanity Aren't the 1.16.1 mpq font files just 256 characters in 1252 order? Korean text in 1.16.1 gets an exception and is rendered through GDI but I thought otherwise the handmade fonts were used Closer to 200 chars, yes. I used to have a Korean 1.16 install, never had Russian or Japanese so I don't know[2018-7-09. : 6:54 pm] Neiv -- Suicidal InsanitySuicidal Insanity shouted: remastered parses strings to try to determine if they are plain ANSI / UTF-8 encoded, or korean encoded. If they do that then it seems to be pretty bad heuristic.. Every map that I've seen to use high ascii 1252 characters, even if it's only one á or í for ~style~, doesn't get drawn with 1252 in SCR[2018-7-09. : 6:52 pm] Neiv -- Suicidal InsanitySuicidal Insanity shouted: but I thought SC always uses system locale Aren't the 1.16.1 mpq font files just 256 characters in 1252 order? Korean text in 1.16.1 gets an exception and is rendered through GDI but I thought otherwise the handmade fonts were used[2018-7-09. : 5:30 pm] Suicidal Insanity -- If you want to be compatible to 1.17 and below, you need to convert to 949[2018-7-09. : 5:29 pm] Suicidal Insanity -- I was told that SC should use UTF8 for EUD maps, and if I have an example where that does not work I should forward it. That is why I was asking about the EUD + unicode space issue a week or so ago.[2018-7-09. : 4:54 pm] jjf28 -- so assuming there was an EUD map with korean text... I have to convert the utf8 back to that codepage before saving strings, right?[2018-7-09. : 2:12 pm] Suicidal Insanity -- NeivNeiv shouted: jjf28 I'm saying that don't bother supporting compiling a version which calls the A variants of winapi functions, if you have a unicode one working. It'd be an inferior build which handles (or breaks on) character values above 127 differently depending on system. The one advantage of having an ansi build is it matches what staredit does. But I agree, if all text conversion routines are in place, use unicode.[2018-7-09. : 2:11 pm] Suicidal Insanity -- If it doesn't use system locale, I don't need to worry about russian, japanese, chinese, etc, and I'm much closer to being done than I thought[2018-7-09. : 2:10 pm] Suicidal Insanity -- remastered parses strings to try to determine if they are plain ANSI / UTF-8 encoded, or korean encoded.[2018-7-09. : 1:06 pm] Neiv -- And remastered is apparently utf8 sometimes? Though it seems to read old maps always as 949-encoded, so I don't know when it actually uses utf8[2018-7-09. : 1:04 pm] Neiv -- (To be exact, BW only special cases korean codepage 949 for text encoding if the system language is korean, and otherwise just uses 1252, so it's not really always the system codepage)[2018-7-09. : 1:03 pm] Neiv -- Though guess you can argue that a BW editor can do less work by writing map strings using system's codepage, as BW kind of reads them back using system's codepage, even though I'm disgusted by that[2018-7-09. : 1:01 pm] Neiv -- jjf28jjf28 shouted: or maybe you're suggesting that my support for compiling without defining UNICODE in which I just treat UTF-8 text as 8-bit ASCII is the wrong approach - I hadn't really thought about what happens in this case, I just assumed any compilation intending to support unicode would define UNICODE and that if someone wanted to compile without that define they would only be doing 8-bit ASCII (BAU) I'm saying that don't bother supporting compiling a version which calls the A variants of winapi functions, if you have a unicode one working. It'd be an inferior build which handles (or breaks on) character values above 127 differently depending on system.[2018-7-09. : 11:52 am] lil-Inferno -- Mini Moose 2707Mini Moose 2707 shouted: How bout those kids in Thailand training for Cave Wars O_O LMAOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO[2018-7-09. : 7:48 am] Suicidal Insanity -- Basically every week I switche dto Unicode, fixed stuff for an hour, got frustrated, switched back and worked on something more interesting  [2018-7-09. : 7:47 am] Suicidal Insanity -- All of the file handling stuff has been using wchar_t s for the past year, and the last dialog to be updated was the classic trigedit UI which I did back in feb or march[2018-7-09. : 7:44 am] Suicidal Insanity -- jjf28jjf28 shouted: "New build up which is compiled in Unicode mode." ohey, you're going through the pain right now too Ya I finished converting all of the UI to use TCHARs a month ago, and then had to update ~50 locations which touched string data to use utility functions to convert from SC format to TCHAR [2018-7-09. : 2:36 am] jjf28 -- I guess it's anglo-centric of me to assume an application without UNICODE defined is using the.. shall we say, default codepage, and that assumption would hurt other languages in UTF-8, so I'll rethink (or at least document) that a bit[2018-7-09. : 2:05 am] jjf28 -- if I were to pursue Unicode support without defining UNICODE then your suggestions make total sense[2018-7-09. : 2:04 am] jjf28 -- and since UTF-8 is backwards compatible there should be no issue with that approach[2018-7-09. : 2:03 am] jjf28 -- or maybe you're suggesting that my support for compiling without defining UNICODE in which I just treat UTF-8 text as 8-bit ASCII is the wrong approach - I hadn't really thought about what happens in this case, I just assumed any compilation intending to support unicode would define UNICODE and that if someone wanted to compile without that define they would only be doing 8-bit ASCII (BAU)[2018-7-09. : 1:54 am] jjf28 -- windows only supports Unicode in UTF-16, so if you're in UTF-8, doing the conversion is mandatory; UTF-8 to UTF-16 conversion should in theory be lossless[2018-7-09. : 1:54 am] Pr0nogo -- Mini Moose 2707Mini Moose 2707 shouted: How bout those kids in Thailand training for Cave Wars O_O HAHAHA[2018-7-09. : 1:51 am] Dem0n -- wow I can currently buy a raffle ticket for 10 minerals for a chance to win 5 minerals O_O |
Members Online: humberto025, DarkenedFantasies, JohnnyTheWolf, anoeth47
DarkenedFantasies, JohnnyTheWolf, anoeth47