←2023-04-12 2023-04-13 2023-04-14→ ↑2023 ↑all
00:02:40 -!- Lord_of_Life_ has joined.
00:02:55 -!- Lord_of_Life has quit (Ping timeout: 265 seconds).
00:04:02 -!- Lord_of_Life_ has changed nick to Lord_of_Life.
00:33:15 -!- user3456 has quit (Ping timeout: 248 seconds).
02:30:37 -!- razetime has joined.
02:31:12 -!- razetime has quit (Client Quit).
03:25:27 <esolangs> [[User:Heptor]] https://esolangs.org/w/index.php?diff=108013&oldid=87503 * Heptor * (-63) Blanked the page
03:26:49 <esolangs> [[Special:Log/newusers]] create * Balt * New user account
03:29:08 <esolangs> [[Esolang:Introduce yourself]] M https://esolangs.org/w/index.php?diff=108014&oldid=107994 * Balt * (+176)
04:14:46 -!- slavfox has quit (Quit: ZNC 1.8.2 - https://znc.in).
04:19:19 -!- slavfox has joined.
04:43:34 -!- user3456 has joined.
04:53:07 -!- bgs has joined.
05:48:01 -!- m5zs7k has quit (Ping timeout: 265 seconds).
05:49:00 -!- m5zs7k has joined.
05:59:56 -!- bgs has quit (Remote host closed the connection).
06:20:44 -!- fowl has quit (Ping timeout: 246 seconds).
06:28:30 -!- fowl has joined.
06:43:42 -!- fowl1 has joined.
06:45:20 -!- fowl has quit (Ping timeout: 248 seconds).
06:45:20 -!- fowl1 has changed nick to fowl.
07:15:57 -!- Sgeo has quit (Read error: Connection reset by peer).
08:45:06 -!- wib_jonas has joined.
10:06:57 -!- __monty__ has joined.
10:17:52 -!- ddccdd has joined.
12:17:00 -!- wib_jonas has quit (Quit: Client closed).
12:18:19 <esolangs> [[LogicGate]] https://esolangs.org/w/index.php?diff=108015&oldid=107870 * Taokyle * (-13) /* Runner */
13:48:06 <esolangs> [[Special:Log/newusers]] create * Peepoostpin * New user account
14:29:19 -!- Sgeo has joined.
15:02:57 -!- wpa has joined.
15:14:51 -!- wib_jonas has joined.
16:13:49 <esolangs> [[IBSA]] https://esolangs.org/w/index.php?diff=108016&oldid=107988 * Simple9371 * (+245) Complete Turing-completeness proof
16:27:59 <esolangs> [[IBSA]] M https://esolangs.org/w/index.php?diff=108017&oldid=108016 * Simple9371 * (-24) /* Computational class */ Comment on successful halt
16:40:09 <esolangs> [[User:Simple9371]] https://esolangs.org/w/index.php?diff=108018&oldid=107946 * Simple9371 * (+31) Add IBSA
16:47:26 <ddccdd> what is the difference between a stack-based language and a language that passes arguments in the stack?
16:47:40 <esolangs> [[IBSA]] M https://esolangs.org/w/index.php?diff=108019&oldid=108017 * Simple9371 * (+3) /* A complete code */ obj0.! > obj0.01
16:51:08 <esolangs> [[IBSA]] M https://esolangs.org/w/index.php?diff=108020&oldid=108019 * Simple9371 * (+3) /* Documentation */ Oops... applied previous edit to other parts
17:06:20 -!- wib_jonas has quit (Quit: Client closed).
17:51:58 -!- FreeFull has joined.
18:32:13 <zzo38> Can GCC be told to refuse to load files with non-ASCII characters in them (including files loaded by #include, but not by #embed)?
18:48:08 <esolangs> [[Infinite state machine]] N https://esolangs.org/w/index.php?oldid=108021 * Stkptr * (+2954) Create page
18:58:50 <b_jonas> zzo38: I don't think so, but you can ask gcc to list all the files that it's including, and then read those files to check for that condition. and yes, there's a race condition in that.
19:06:30 <esolangs> [[User:Stkptr]] https://esolangs.org/w/index.php?diff=108022&oldid=107997 * Stkptr * (+344) Add proof target bookmarks
20:01:00 -!- ddccdd has quit (Quit: leaving).
20:02:42 -!- wpa has quit (Quit: Connection closed for inactivity).
20:14:27 <esolangs> [[Talk:4]] N https://esolangs.org/w/index.php?oldid=108023 * Stkptr * (+2022) /* Proof of Turing completeness */ new section
20:20:08 <esolangs> [[Talk:4]] M https://esolangs.org/w/index.php?diff=108024&oldid=108023 * Stkptr * (+3) Fix the general structure
20:21:37 <esolangs> [[4]] https://esolangs.org/w/index.php?diff=108025&oldid=101593 * Stkptr * (+276) Clarify computational class
20:27:04 <esolangs> [[Talk:4]] https://esolangs.org/w/index.php?diff=108026&oldid=108024 * Stkptr * (+0) /* Proof of Turing completeness */ re-sign
20:28:52 -!- chiselfuse has quit (Remote host closed the connection).
20:29:18 -!- chiselfuse has joined.
20:29:22 <esolangs> [[User:Stkptr]] https://esolangs.org/w/index.php?diff=108027&oldid=108022 * Stkptr * (+180) Add stuff
20:32:44 <esolangs> [[Talk:Cratefuck]] https://esolangs.org/w/index.php?diff=108028&oldid=107999 * Stkptr * (+0) /* Proof of Turing completeness */ Fix signature
20:34:27 <esolangs> [[User:Stkptr]] https://esolangs.org/w/index.php?diff=108029&oldid=108027 * Stkptr * (+96)
20:55:00 -!- ais523 has joined.
20:55:13 <ais523> zzo38: gcc can be, you want the option «-finput-charset=ascii»
20:56:09 <ais523> although, it seems that the errors from that don't show line numbers, because if gcc can't decode the input it can't figure out where the line endings are
20:56:15 <ais523> so you get some very abrupt error messages
20:58:18 <b_jonas> ais523: huh, good point, I haven't found that option
20:58:32 <b_jonas> good to know
20:58:59 <ais523> b_jonas: it's an option to the preprocessor rather than the compiler (which makes sense if you think about it)
20:59:20 <ais523> I only remembered it because I read the entire documentation for GNU cpp a few days ago (for an unrelated reason)
21:00:05 <b_jonas> so apparently by default gcc also assumes that your source files are encoded in locale-dependent encoding? that's crazy. I knew that javac did that and have seen a java program that can't be compiled unless you set the locale correctly, but I thought gcc would just always assume utf-8 (at least for C or C++ source files)
21:00:52 <ais523> b_jonas: default seems to be "locale, or UTF-8 if the locale doesn't specify"
21:01:39 <b_jonas> ais523: is there some way to specify the encoding in the source file, short of escaping every non-ascii character?
21:01:41 <ais523> and if you think about it, on a computer system where the locale isn't UTF-8, most files will be expected to be non-UTF-8 because programs will be using the locale encoding to read and write them
21:01:44 <b_jonas> specify utf-8 encoding that is
21:02:09 <ais523> b_jonas: not as far as I know
21:02:19 <ais523> there doesn't seem to be an equivalent to Perl's "use utf8"
21:02:27 <ais523> (which is somewhat confusingly named)
21:03:10 <b_jonas> ais523: maybe, but I'm not sure that argument applies to C or C++ sourrce files. you rarely have a locale-aware program write a C or C++ source file.
21:04:10 <ais523> this reminds me of the difference between text and binary files – theoretically a file that states "this file is encoded in UTF-8" isn't actually a text file
21:04:10 <ais523> because text files can be re-encoded into an encoding that contains all the same characters without changing their meaning
21:04:10 -!- ais523 has quit (Remote host closed the connection).
21:04:26 -!- ais523 has joined.
21:04:35 <ais523> b_jonas: aren't most editors locale-aware?
21:04:58 <ais523> I'm fairly sure yacc is locale-aware (I wrote ayacc against the POSIX yacc specification and had to include locale-handling code in order to match it)
21:05:33 <ais523> …and if you try to write a C program on an EBCDIC system, you'd better respect the locale or nothing will be able to read it
21:05:46 <ais523> (that said, I don't think EBCDIC is at all widely used nowadays)
21:05:54 <ais523> (but this used to be a real issue at one point)
21:07:25 <b_jonas> apparently python assumes that source files are utf-8 by default, but you can override this with a magic comment in the file if you really want; but stdout is assumed to be locale-encoded instead if it's a regular file
21:08:15 <ais523> Rust ignores locale, I think intentionally
21:08:41 <ais523> although it lets you get at the raw bits of things like command-line arguments, so that you can decode them yourself if you need to and don't expect them to be UTF-8
21:09:35 <ais523> what about esolangs? I've been writing a BF interpreter that reads the input in binary and assumes an ASCII-compatible encoding
21:09:43 <b_jonas> "raw bits of things like command-line arguments" => sure, it's a general-purpose low-level language so it has to allow that sort of operating system interaction, would be silly not to
21:10:18 <ais523> and I'm not sure that's what most interpreters do (I think if you compile a typical BF interpreter on an EBCDIC system it'll read the program in EBCDIC)
21:13:04 <b_jonas> dunno, I kind of don't care enough about EBCDIC to support it as a language in which you can write source files, though of course an intercal interpreter/compiler should probably support that for flavor reasons
21:13:51 <ais523> CLC-INTERCAL uses EBCDIC "natively" in some sense, I think
21:13:56 <ais523> possibly as an internal representation?
21:14:20 <b_jonas> so I'll probably specify consumer society such that the source file needs to have an ascii-compatible encoding, as in no ebcdic or utf-16, but utf-8 or cp1252 etc are allowed
21:14:48 <ais523> Perl's default encoding appears to be "unspecified ASCII-compatible 8-bit character set"
21:15:24 <b_jonas> ais523: it uses EBCDIC natively in that the bookworm operator is written with a character that isn't in ASCII
21:16:03 <ais523> ah yes
21:16:24 <b_jonas> intercal supports an ASCII replacement for it, but it's just a replacement for when you have an unusual system that isn't reading your program from punch cards
21:16:38 <b_jonas> oh, the change operator too, not just bookworm
21:16:44 <ais523> oh, this reminded me of how C-INTERCAL accepts input encoded in either UTF-8 or Latin-1 or even mixed together in the same file, there are no clashes of byte sequences for characters that INTERCAL uses
21:17:07 <ais523> but there is one clash of *character*, ¥ that does something different in C-INTERCAL and CLC-INTERCAL, and it determines the meaning based on which encoding it's encoded with
21:17:23 <ais523> * C-INTERCAL determines
21:17:36 <b_jonas> er what
21:17:41 <b_jonas> how does it have a different meaning?
21:17:54 <b_jonas> I assume one of the meanings is the bookworm operator
21:17:58 <ais523> in C-INTERCAL, all international currency symbols mean mingle
21:18:08 <b_jonas> oh!
21:18:14 <ais523> ¢ being generalized to $, but also things like £
21:18:20 <b_jonas> that makes sense
21:18:27 <ais523> and yes, bookworm is the other one
21:23:38 <b_jonas> for languages that are ascii-based like bf or unlambda or underload or consumer society, I'd just expect an ascii-compatible encoding, and the most I'd do with specific non-ascii encodings is allow some non-ascii whitespace as whitespace, such as an utf-8 byte order mark, or nul bytes for a pure-ascii file encoded as utf-16
21:24:24 <ais523> at least in BF, it's probably OK to treat all high-bit-set bytes as comments – that gives correct behaviour whether the input is UTF-8 or some 8-bit character set
21:24:50 <ais523> …and it also works if the input is the ASCII subset of UTF-16 (either of them), because the NULs are treated as comments
21:25:02 <ais523> (but not always for UTF-16 that contains non-ASCII characters)
21:25:17 <b_jonas> and if there are characters/strings to print, as in unlambda/underload/consumer society, then I'll just copy raw bytes, which works if the output has the same encoding as the source file
21:26:47 <b_jonas> though consumer society also has the rule that you can put any byte in a part of the code that's never evaluated, in those parts of the code only the balanced nesting of ascii bracket characters that matter
21:26:50 <esolangs> [[User:Stkptr]] https://esolangs.org/w/index.php?diff=108030&oldid=108029 * Stkptr * (+90)
21:26:58 <b_jonas> I'd do the same for underload parenthesis too
21:27:48 <b_jonas> only when you actually try to evaluate the function with unknown bytes you get an error
21:28:20 <b_jonas> this lets you write comments in any ascii-compatible encoding, as well as use syntactical extensions conditionally on them being supported
21:29:06 <b_jonas> or write polyglots of course
21:35:49 <esolangs> [[User:Stkptr]] https://esolangs.org/w/index.php?diff=108031&oldid=108030 * Stkptr * (+162)
21:37:30 <b_jonas> but there are modern esolangs like Brachylog and Jelly which do use non-ascii characters in an essential way in their syntax, so for those you must figure out what encoding the source files are in
21:37:45 <ais523> b_jonas: Brachylog and Jelly each define their own encodings
21:38:01 <b_jonas> yes, and they do that because they want to be golf languages, unlike intercal
21:38:05 <ais523> and at least Jelly's isn't fully ASCII-compatible IIRC, I think it has newline in a weird place
21:38:11 <b_jonas> that's why they use non-ascii characters too
21:38:28 <ais523> yep: https://github.com/DennisMitchell/jellylanguage/wiki/Code-page has newline at 0x7F
21:38:31 <b_jonas> so that they can have more than just 128 characters to work with
21:38:40 <ais523> whereas the usual 0x0A is ½
21:38:41 <b_jonas> ais523: what's at 0x10?
21:38:47 <b_jonas> um
21:38:50 <b_jonas> at 0x0A, yes
21:39:00 <ais523> 0x10 is Ñ but i think you mixed up decimal and hexadecimal
21:39:17 <b_jonas> yes, 0x10 is control-P in ascii
21:47:42 <b_jonas> https://github.com/DennisMitchell/jellylanguage/wiki/Code-page says "The character ¶ and the linefeed character can be used interchangeably, although it is advised to use a linefeed to separate links, and a pilcrow inside string literals.", where ¶ is the one encoded as 0x7F
21:48:55 <b_jonas> oh, maybe not, maybe that sentence applies only when the program is utf-8 encoded
21:51:07 <zzo38> I think that many file formats should not need to care about encoding a long as it is extended ASCII, but sometimes it might matter. (And, in some cases, restricting it for security might be helpful; this is useful in C if your program does not use any non-ASCII literals/comments (you can still use non-ASCII text if you put them in separate files and load them using #embed or at runtime))
21:52:52 <zzo38> I could assign code page numbers for Jelly, etc if needed; I have already assigned one for Brachylog. (These code page numbers all exceed 65535, so that the lower numbers will be those assigned by IBM)
21:54:21 <zzo38> Will C-INTERCAL or any other esolang have implementation of any variant of TRON code (such as TRON-8, TRON-16, or EUC-TRON)?
21:56:34 <b_jonas> zzo38: is that the kind of code page numbers that windows uses, or the kind that the IANA tracks in https://www.iana.org/assignments/character-sets/character-sets.xhtml ? or are those two in the same namespace actually?
21:57:28 <ais523> Microsoft's numbers are usually 4 digits, and IBM's normally 3 digits
21:57:35 <zzo38> b_jonas: Windows uses Microsoft code page numbers, some of which are different from IBM code page numbers.
21:57:52 <ais523> I wonder whether that's an intentional choice to avoid clashes, or whether there's some other reason
21:58:04 <ais523> although I think Microsoft have a 5-digit code page number for UTF-8?
21:58:08 <zzo38> (Although some of them are the same, such as 437 for the PC character set.)
21:58:34 <ais523> 437 is one of the few codepage numbers that many people have memorised even today, it's very well known
21:58:36 <b_jonas> I think there was like one number that is used as the name of two different character sets, but probably only as a name, not in one of those numbering plans
21:58:49 <ais523> UTF-8 is codepage 65001, apparently
21:59:02 <zzo38> Yes. Microsoft uses 65001 for UTF-8, and IBM uses 1209.
21:59:02 <ais523> which is a suspiciously-valued number, being just below 65535 and having an 001 at the end
21:59:05 <b_jonas> ais523: yes, and I think they have a five-digit number for utf-16 too
21:59:30 <ais523> 65000, perhaps?
22:00:03 <JAA> 65000 is UTF-7. :-)
22:00:08 <ais523> (that's a guess based on UTF-8's number)
22:00:17 <JAA> https://learn.microsoft.com/en-us/windows/win32/intl/code-page-identifiers
22:00:23 -!- __monty__ has quit (Quit: leaving).
22:00:51 <b_jonas> https://en.wikipedia.org/wiki/Windows_code_page looks like the windows ones aren't in the same namespace as the IANA ones, in particular 37 clashes
22:01:22 <ais523> ISO-8859-1 is 28591? wow, I'd expect it to be much lower
22:01:25 <b_jonas> also https://en.wikipedia.org/wiki/Windows_code_page#Unicode-related_code_pages
22:01:30 <ais523> that's, like, the best-known 8-bit character set
22:01:54 <ais523> also I'm surprised UTF-7 has one at all
22:02:06 <ais523> that seems like a compatibility (and security) nightmare
22:02:25 <ais523> I seem to remember some security bugs related to writing text that autodetected as UTF-7 in order to get around filters
22:02:28 <b_jonas> ais523: yes, but these numbers are mostly for character sets used in terminal windows, where you usually use 437-like character sets such as 852 and 850, not ones like 1252 or iso-8859-1
22:02:31 <ais523> (because the filters were interpreting it as ASCII)
22:02:54 <ais523> iso-8859-1 is a common encoding for terminals, there's even a control code for it
22:03:22 <b_jonas> ais523: for unix terminals, sure, not for windows terminals I think
22:03:34 <zzo38> IBM uses 819 for ISO-8859-1. However, Windows uses an extension of ISO-8859-1, which is 1252 in both IBM and Microsoft, I think.
22:04:01 <ais523> yes, and the HTML standard even says something along the lines of "if a web page claims to be in ISO-8859-1, you should assume it's actually in Windows-1252"
22:04:21 <ais523> presumably because ISO-8859-1 is a subset of the printable characters of Windows-1252 and this is a common mistake in encoding declarations
22:04:45 <b_jonas> also apparently no character set registry contains the old windows-1252, from before windows made the revised version that adds the euro sign and the romanian letters
22:04:48 <fizzie> Bunch of Finnish IRC channels used ISO-8859-15.
22:05:10 <b_jonas> old windows-1250 and old windows-1252 both, the romanian letters are in the former
22:05:10 <fizzie> ("The one with €.")
22:05:39 <ais523> now I'm trying to remember which letters are Romanian: ș and ț?
22:05:54 <b_jonas> fizzie: yeah, I used to be on an irc channel that had iso-8859-2 as the standard, and then eventually switched to utf-8
22:05:58 <b_jonas> ais523: yes, those two
22:06:20 <ais523> I'm kind-of surprised that I figured out how to type them within two tries
22:06:34 <FireFly> I think I somehow purged UTF-7 from my memory
22:06:49 <fizzie> Is it just me that's having trouble parsing it, or is this sentence from https://en.wikipedia.org/wiki/ISO/IEC_8859-15 somehow weird: "This encoding is by far most used, close to half the use, by German, though this is the least used encoding for German."
22:07:26 <ais523> that is not normal English phrasing, and I'm not convinced it's even correct
22:07:43 <ais523> …should be "used most" not "most used"
22:07:49 <b_jonas> also I'm wrong about that, neither cp1250 nor cp1252 has the disunified romanian letters
22:08:01 <ais523> and even then, the grammar is ridiculously convoluted and I'm not convinced there isn't another mistake
22:08:12 <b_jonas> I think they added some character besides the euro into at least one of them, but even just the euro makes them different charsets
22:08:27 <esolangs> [[Talk:4]] https://esolangs.org/w/index.php?diff=108032&oldid=108026 * Stkptr * (-63) /* Proof of Turing completeness */ Fix PMMN dec logic
22:08:57 <ais523> there's a latin-1 variant with the euro sign in and otherwise minimal changes, isn't there?
22:09:26 <fizzie> That's pretty much what ISO-8859-15 (Latin 9) is.
22:10:50 <fizzie> It replaces the generic currency symbol ¤ with €, and does six other substitutions, replacing less commonly needed symbols with, for want of a better word, more langugage-y letters.
22:11:31 <b_jonas> ais523: yes, ISO-8859-15 is based on ISO-8859-1 but has the euro sign and some more letters; and ISO-8859-16 is a modified version of ISO-8859-2 but has the euro sign and some more letters including the disunified romanian letters, which explains what I was confusing the cp1252 thing with
22:11:52 <b_jonas> when was the euro sign even invented?
22:12:35 <fizzie> Apparently in 1996, by being selected from a set of 30 different proposals.
22:12:40 <ais523> 12 December 1996, according to Wikipedia
22:12:47 <fizzie> "The other designs that were considered are not available for the public to view, nor is any information regarding the designers available for public query."
22:12:48 <b_jonas> yeah, wrong question, when was its invention *published*
22:12:55 <b_jonas> 1996, ok
22:13:48 <b_jonas> so the euro sign very much postdates cp1252 and cp1250 as used in windows-16 fonts
22:14:20 <fizzie> Apparently there's some dude who says they had the idea for the current symbol 25 years earlier.
22:14:52 <ais523> the € character may actually have been a useful driver for adoption of non-8-bit character sets, come to think of it
22:15:03 <b_jonas> fizzie: that sounds suspicious because that would be well before the fall of the iron curtain
22:15:25 <ais523> because it postdates them being invented and forces a change, and it might have encouraged a change to something more general
22:17:04 <ais523> huh, I just noticed ASCII also has a codepage number, and it's 20127
22:17:07 <b_jonas> I heard that one of the 90s video cards was invented by someone who wanted to use thai letters on PC
22:17:16 <ais523> seems fairly arbitrary, although the "127" may be meaningful
22:18:45 <b_jonas> because the existing video cards had the font burned into ROM and unchangable
22:18:58 <b_jonas> well, the ones available
22:19:47 <b_jonas> may be 80s instead of 90s
22:20:38 <ais523> I'm old enough to remember text-mode VGA consoles – I'm not sure whether modern computers even emulate them, though, there are definitely text consoles but they seem to be handled by the OS rather than the firmware/hardware
22:20:58 <fizzie> I've got one of those 20x4 character LCDs here, and it has 240 of its 256 characters in ROM, and then 40 bytes of "CGRAM", into which you can write 8 unique 5x8 bitmaps as extra characters. (The remaining 8 are just duplicate copies of those 8.)
22:20:58 <zzo38> But the IBM code page number for purely ASCII is 367.
22:21:50 <b_jonas> sure, I used text-mode VGA console back when my pentium CPU was too slow so updating even a fast X terminal when scrolling text quickly was noticably laggy
22:21:58 <b_jonas> also not running X would save RAM
22:22:56 <ais523> when I was brute-forcing NetHack to find an RNG seed where the character would die immediately, I was at one point using both a Linux laptop, and a Windows computer running NetHack in DOS mode, simultaneously
22:23:15 <fizzie> You could do 132x50 text mode resolutions at least sometimes, as I recall.
22:23:19 <ais523> the DOS terminal somehow ended up having bits get stuck across the entire screen
22:23:42 <ais523> so, e.g., the redness bit would be turned off for every character and so the text would display in cyan
22:23:48 <b_jonas> I have used custom text video modes (including 30x80 at 90 hertz), and custom text omde fonts (including dynamically modifying a font for smooth scrolling text)
22:24:02 <ais523> it got really weird when the 1s bit of the ASCII code got turned off, though, so e.g. 'e' would display as 'd'
22:24:09 <ais523> I have no idea what caused that sort of failure mode
22:24:25 <ais523> after that I stopped brute-forcing on the Windows computer and just used the laptop
22:24:40 <b_jonas> ais523: that sounds odd
22:24:52 <ais523> it was very odd
22:24:59 <ais523> but I didn't put much effort into trying to reproduce
22:25:27 <b_jonas> I've seen funny text mode visual glitches when I tried to determine what was the highest refresh rate that my video card could handle without such glitches
22:25:44 <ais523> I didn't see the event that caused the issue to start, just came to the computer after a few hours and saw that all the characters with odd ASCII codes were displaying as the character before
22:25:45 <fizzie> I used to have a DEC serial terminal under the table for guests to IRC on, and it could do 132-column modes too. It also had a built-in calculator, and copy-paste functionality, and a few other tricks.
22:25:45 <b_jonas> I've seen occasionally glitches at 100 Hz but not at 95 Hz, so I went with 90 Hz to be safe
22:25:53 <fizzie> Then it caught on fire, but that wasn't really a feature.
22:26:37 <fizzie> After the smoke came out of it, it no longer worked, as is often the case.
22:26:46 <b_jonas> but what you report doesn't sound similar to the glitches that you've seen, plus I don't think you used a higher than normal clock speed for this
22:27:33 <b_jonas> fizzie: did it come with a paper tape reader attached?
22:27:45 <ais523> I considered that it might be some sort of rowhammerish hardware memory corruption, but I'd expect that to be fixed as the memory was being continuously written to
22:27:48 <fizzie> There were also definitely DOS .com executables that reprogrammed the VGA text mode font, because I had the Descent font as one of those.
22:28:24 <ais523> so I'm expecting to be some sort of software feature that was intended to mask memory, and somehow got activated accidentally or erroneously to do something weird
22:28:33 <ais523> * expecting it to be
22:28:50 <fizzie> I don't think it had a tape reader. It was a... VT420, maybe.
22:29:39 <fizzie> An actual "hard" status line, though.
22:29:45 <b_jonas> if it was just some characters showing up in incorrect color then I'd assume it's a bug in nethack, but with the low bit of the character stuck on the whole screen that sounds unlikely
22:29:46 <fizzie> You could make `screen` use that.
22:44:06 <b_jonas> ais523: the windows computer running nethack, was that 16-bit windows in 386 mode, or 32-bit windows?
22:44:38 <b_jonas> also is nethack on DOS a real mode program or does it use a 32-bit dos extender?
22:45:32 <b_jonas> I'm just curious unrelated to the screen glitch
22:47:24 <ais523> b_jonas: it was either Windows 95 or 98 I think
22:47:43 <ais523> and it uses a DOS extender nowadays, and I think it did even at the time
22:47:49 <ais523> (that I was brute-forcing)
22:49:09 <zzo38> What happens in Magic: the Gathering if a battle has multiple subtypes? (This is not possible with current rules but might be relevant with unofficial cards)
22:50:28 <b_jonas> also totally unrelated: libcurl question. if you reuse a CURL handle for multiple HTTP queries, does it default to saving cookies from previous queries and sending them in subsequent queries (if the domain matches), or does it default to sending no cookies? and if the former, what's the proper way to ask it to not send remembered cookies, and is it calling curl_easy_setopt(curl, CURLOPT_COOKIELIST,
22:50:34 <b_jonas> "ALL") each time between queries?
22:50:44 <b_jonas> I'd probably better TIAS than ask here, mind you
22:51:10 <b_jonas> ais523: I see
22:54:03 <b_jonas> ah, there's a HTTP-COOKIES.md.gz file in the documentation that I missed, that might asnwer this
22:56:10 <b_jonas> wait... "Cookies are set to the client with the Set-Cookie: header and are sent to servers with the Cookie: header." -- what? I thought they were sent to the server with the Cookie2 header. HTTP is getting too opaquely magical for me.
22:56:30 <b_jonas> there's just so much web tech stuff that I can't follow anymore
22:57:54 <zzo38> WWW is really messy now anyways, regardless of which header
22:59:38 <b_jonas> ok, I just looked up my very old cbriver script, and it sends the Cookie header, not the Cookie2 header
23:00:34 <b_jonas> Cookie2 does exist, but is an old field that's not used anymore apparently
23:01:37 <b_jonas> and that document doesn't answer my question
23:43:54 -!- FreeFull has quit.
23:46:00 -!- sprock has quit (Quit: brb).
23:50:44 -!- ais523 has quit (Quit: quit).
23:51:33 -!- sprock has joined.
←2023-04-12 2023-04-13 2023-04-14→ ↑2023 ↑all