01:13:44 ZZT itself is now Free software (the source code has been reconstructed), but you will still need Turbo Pascal to compile it, so it is still "trapped". You will also need LZEXE. Doing this allegedly produces the same ZZT.EXE file as the original; the only difference may be how the source code is formatted, and the comments and possibly variable names and so on. 01:17:43 -!- ais523 has quit (Quit: quit). 01:18:00 -!- oerjan has quit (Quit: Nite). 01:19:16 zzo38: didn't Borland make the Pascal compiler free at some point? Or do you need a specific older version? 01:21:11 (that's free as in beer, though) 01:22:04 Yes, it is freeware now, but not free software. (Still, it is a DOS program, so it is possible that you might not care.) 01:22:21 yeah, it sounds like it's abandonware 01:29:04 I used to care... but now it's just for a couple of games. 01:29:34 (though actually I do have a working dosemu installation, which is probably unusual these days) 01:32:39 Hooloovo0: nah it was properly released... not just exploiting the fact that nobody's coming after you for your copyright violation 01:32:57 oh, that's cool 01:37:31 hrm. CTRL-F9 Kill dosbox. 01:37:42 whose bright idea was that ;) 01:38:19 You could disable it if you do not want that, I suppose. 01:38:51 (CTRL-F9 runs programs in the Borland Pascal IDE) 01:39:53 I wanted to make it to use the "context menu" key instead of CTRL for this purpose, although if you save that into the configuration file, then it won't work. Does the new version fix this? 01:40:30 -!- FreeFull has quit. 01:40:36 I don't know. 01:41:26 I don't see myself actually using this stuff ever again... definitely not in the near future. 01:43:20 The stuff is there because I've copied it from the previous computer, and it was there because I copied it from the computer before that... I think I'm at the 5th iteration now? 01:47:13 -!- imode has quit (Ping timeout: 246 seconds). 01:49:11 -!- imode has joined. 01:50:09 int-e: do you have a particular use case for dosemu? 01:50:41 -!- shachaf has left. 01:56:52 kmc: no. it used to be a matter of speed. 01:57:25 ah 01:57:37 but every modern CPU can run a plain emulator like dosbox faster than any original DOS machine, right? 01:57:56 Yes. 02:00:20 The BP IDE doesn't even run in the dosemu... some dpmi problem. Works in dosbox though. 02:01:36 huh 02:03:12 http://int-e.eu/~bf3/tmp/bpc.png <-- looks weird 02:05:00 But who knows. I may have misconfigured something, and I cannot really be bothered to debug this. 02:06:32 Especially given the alternative that works. 02:29:31 Someone said they were considering updating NNTP for 2020, and posted a proposal to the IETF. I suggested increasing article numbers from 31-bits to 63-bits, and adding the WHOAMI command, both of these features I have implemented in my own software. Maybe the IETF will make this official. 02:30:38 (They also called Usenet "the world's largest social media network".) 02:31:08 -!- ais523 has joined. 02:32:24 hearing that Turbo Pascal is now available in perpetual free-trial form is pretty encouraging 02:32:28 I still have some old Pascal code from a while back 02:32:46 I remember translating it into C but the C ran more slowly; that surprised me at the time and still surprises me now 02:32:54 maybe the Pascal had a better optimiser, or faster supporting libraries, or something 02:34:59 oh wow, this pascal source is so old that the file ends with a literal ^Z character 02:35:06 how long ago was it that text files stopped needing those? 02:35:55 I'm not sure, was it when we abandoned CP/M? 02:35:56 Maybe Pascal is faster because it is DOS. 02:36:04 the C compiler was also DOS 02:36:08 I think, at least 02:39:07 oh wow, the executable I compiled from the Pascal source still runs (in DOSBox) 02:39:08 Yes, although Pascal may be more suitable for DOS, at least in real mode. 02:39:13 I guess that's not a surprise 02:39:35 the executable I compiled from the C source does not run – it's missing dependencies – and I guess the problem is that those dependencies appear to be manually linked 02:39:44 which could be very slow given that the pixels are written individually with function calls 02:40:02 What graphics mode? 02:40:43 it looks like 16-color 640×480 VGA 02:41:02 CP/M was a good guess, it seems; I vaguely remembered the record-based file access (which DOS mimics in its FCB interface for backward compatibility). "The CP/M file system only recorded the lengths of files in multiples of 128-byte "records", so by convention a Control-Z character was used to mark the end of meaningful data if it ended in the middle of a record."... 02:41:08 ...https://en.wikipedia.org/wiki/End-of-file 02:42:27 Yes, although I don't know why it should be ^Z, since that isn't the ASCII code for end of file; ^D might make more sense 02:42:45 But Z is the last letter of the alphabet. 02:43:04 . o O ( It's the latin omega. ) 02:44:09 Yes, although it stands for "substitute" and not "end of file", isn't it? 02:45:08 Whatever that means. 02:46:44 the intended use was as an equivalent of REPLACEMENT CHARACTER 02:46:46 I think "substitute" means there is no valid character, so it substitutes ^Z in its place. 02:46:50 `unicode REPLACEMENT CHARACTER 02:46:52 ​� 02:47:19 but it was mostly used for other purposes, and Unicode had to add a separate such character at U+FFFD 02:47:52 Well, I remember that ^A and ^B displayed smileys. That's not in line with ASCII either. I don't think CP/M and DOS developers cared all that much. 02:48:16 int-e: code page 437 assigns glyphs for characters in the C0 and C1 regions, in addition to G0 and G1 02:48:22 (Or the PC BIOS developers which may be the actual people responsible?) 02:48:24 Yes, in the PC character set, the graphics at those points corrsepond to faces. 02:48:26 the C0 characters are very hard to write on most output devices 02:48:48 the C1 characters are incompatible with most of the relevant standards, but are used often enough that terminals often need to find a way to output them anyway 02:49:25 That is true of all of the control range; there are also graphics there on a PC, so that if you write directly to the video memory, they will be visible. 02:50:08 Whatever. CP/M and by extention DOS used ^Z to mark the end of files. That's really where the discussion should've ended. 02:50:33 policing topicality here can be kind-of hard :-D 02:50:35 Unless you have a time machine. 02:50:36 Yes, that is probably why DOS does that, because of CP/M. 02:50:45 and retrocomputing is closer to our main topic than much of what's discussed here 02:51:08 ais523: I'm mostly annoyed at myself for participating in it beyond that point :) 02:51:09 I've actually put thought into writing an esolang for retrocomputing purposes 02:51:32 the main aim being to reduce the amount of RAM that a program written in it takes to run, as far as possible 02:51:35 Yes, and even in esolang wiki there is a few retrocomputing stuff. And, I do want to see your esolang for retrocomputing purposes once you write it. 02:52:02 using any trick available; overlays, swapping to disk, using bytecode interpreters with compressed bytecode, etc. 02:52:27 I didn't design most of the details 02:52:46 one idea I had was for it to be a byte-compiled language where the bytecode used return-oriented programming 02:52:58 as a method of compression in addition to a method of function calls 02:54:51 another idea I had was to store all values as 1 byte on the stack, if they were more than 1 byte wide then a hash table element would be tied to the stack location and contain the real value, with the stack containing information to find the element 02:55:08 you could probably do something similar with the heap 02:56:14 but these are just ideas, and that's all I have, it's not like there's some big plan that I'm hiding 02:56:33 OK 02:57:15 I like those ideas 02:58:41 (I thought of the similar idea about return-oriented programming before, too) 02:59:24 I'm actually surprised that return-oriented programming isn't done more often as a programming technique in cases where you don't have to use it 02:59:36 (of course, for the exploits against W^X code, it's often necessary) 02:59:56 I guess Underload has some similarities to return-oriented programming 03:00:05 although there are a lot of differences too 03:02:45 ais523: I imagine it's rather expensive because it messes up the builtin call-stack based branch prediction. 03:03:03 int-e: that's not necessarily the case if using a bytecode interpreter 03:03:10 (of modern CPUs) 03:03:23 but yes, it seems likely you'd have a lot of mispredicted branches 03:03:29 that doesn't matter so much in retrocomputing, though 03:04:37 I've certainly played funny tricks with the stack in golfing assembly code 03:05:19 (well, machine code is what was golfed, really, but I wrote it in assembly) really, but I 03:06:24 ... ^J (line feed = return key) is next to ^K (delete until end of line) 03:06:43 well, you normally don't linefeed mid-line anyway 03:06:44 I've had so many IRC editing accidents because of this. 03:07:09 oh, I see, you're trying to delete to the end of the line, but press ^J by mistake 03:07:17 Yes. 03:07:20 my most common accident of that nature is hitting Return instead of Backspace 03:07:42 (which are, again, adjacent, and you need to reach a long way to hit backspace, over Return) 03:07:48 Hmm. Wrong keyboard layout for this, \ is between backspace and return here. 03:07:56 I guess you could reach for the left-hand side of Backspace 03:08:02 This often happens involuntarily though:\ 03:08:13 no, I mean I could 03:08:15 to avoid the issue 03:08:35 but spoke ambiguously because basically nobody uses "one" as a pronoun nowadays and none of the common replacements are really satisfactory 03:09:39 ais523: the "you" didn't affect what I typed at all :) 03:10:18 I was just trying to relate to the problem and realized that I no longer have this particular one. 03:11:02 ais523- was using a laptop that was dropped ~3 years ago as an emergency replacement 03:11:15 the keyboard layout was very similar to my current layout except that home/end were swapped with pgup/pgdn 03:11:28 yuck 03:11:36 why do people get creative on that front... 03:11:57 (also it was using a bootable USB stick as the OS because I didn't want to substantially change the existing filesystem (for archival purposes), and its existing OS hadn't been updated in 3 years) 03:12:14 I felt that merited half an underscore 03:12:34 Is this a weak form of vendor lockin? 03:12:46 it could be, I guess 03:12:53 You /could/ use any laptop but half of them will confuse you with their keyboard layout? 03:12:58 I actually abandoned the old vendor (HP) because I didn't like its UEFI implementation 03:13:28 hmm, should the pronoun be "their" rather than "its"? I never did decide whether a company is singular or plural 03:14:37 the biggest offences of the implementation appeared to be a) making the secure boot variables read-only whenever they were edited once, until you did a full reset of the secure boot state 03:14:44 I'd probably go for "their". But I also use singular "their" for gender-neutral means. 03:15:04 b) resetting the default bootloader to Windows's bootloader whenever you went into the UEFI settings 03:15:08 Though in this case I imagine a company as a group of people. 03:15:28 I think it's fairly well established that companies don't have genders and don't use pronouns that are intended only for people 03:15:47 so gender-neutrality isn't really an issue in this context, just singular/plural 03:16:00 Oh yeah I remember having to jump through some hoops to boot Linux on my former work HP Elite notebook. 03:16:06 I guess an Agoran "eir" wouldn't be that out of place, corporations act like people in some respects after all, and "eir" is appropriate for non-human people 03:16:39 I think it's interesting that most Linux distributions don't add themselves to Windows' bootloader, it would seem like a logical thing to do 03:17:01 (so that you could use either shim+GRUB, or Windows Boot Manager, based on personal preference; and yet boot into all your OSes either way) 03:18:35 If you use both Linux and Windows, then yes it can make sense I think 03:19:20 I think it's interesting that the common Linux implementation has bootloaders stacked 3 deep 03:19:44 speaking of booting... why do I have 17 linux kernel images installed :/ 03:19:52 (firmware UEFI → shim → grub), and if you have a multiboot setup this stacks bootloaders 4 deep (because grub chain-loads Windows Boot Manager) 03:20:12 not all these levels are actually needed 03:21:03 And why doesn't debian remove them all and then rebuild the initrds once. 03:21:24 I think UEFI isn't very good 03:21:31 shim is a complicated compromise 03:21:34 zzo38: whatever 03:22:23 basically, it's a consequence of UEFI mandating an executable format that a) can only contain one digital signature and b) shipping with only Microsoft's key by default 03:22:34 b) isn't part of the standard, it's just what happens in practice 03:23:04 so it exists purely to validate a bootloader against someone else's key and then chain-load it 03:23:36 the various Linux distros persuade Microsoft to sign versions of shim that recognise their key, which is easier than trying to persuade every motherboard manufacturer to recognise their key 03:24:25 Hmm, with grubx64.efi isn't it just UEFI -> grub? 03:24:36 int-e: whose signature does grub have? 03:25:06 in my experience, distros put a signed version of grub into the UEFI bootloaders directory, but the UEFI refuses to load it unless someone manually changes the list of approved keys 03:25:08 Oh. You want actual secure boot? 03:25:16 why not? 03:25:24 It's fine. 03:25:42 And it explains the role of shim, I suppose. 03:26:00 one of the many ideas I've had and will never have time to act on, is a UEFI program that lets the user manage keys manually, choosing who to trust and who not to trust 03:26:14 together with kernels that check the filesystem for integrity during boot (at least, the parts they execute) 03:26:49 you wouldn't need much more than a signed md5sums file and a more secure hashing algorithm than md5 03:28:52 I guess an extreme version would require some sort of password to set files as +x… 03:29:24 (under the control of the user, of course) 04:05:13 -!- shinh has joined. 04:30:55 hi shinh 05:04:09 question to everyone: I invented an esolang just to write one program, and have now written the program, but haven't implemented the esolang 05:04:18 should I implement the esolang, or just compile the program by hand? 05:10:31 What program and what esolang? 05:21:41 I'm talking in general terms here 05:21:53 I'll be putting it up on the wiki once I'm sure it works 05:23:22 actually, screw it, I'll upload the program 05:23:57 here's the program: http://nethack4.org/pastebin/spiral-rise.wck 05:24:22 it's an esolang interpreter (for a new esolang called Spiral Rise) written in another new esolang (called Waterfall Construction Kit) 05:24:28 WCK is meant to compile to The Waterfall Model 05:24:57 and is currently defined by that one example program, which is its purpose for existence 05:26:21 I think the program is correct, although I'm not sure, it's pretty hard to test 05:26:35 I think I'm going to try compiling it by hand 05:26:55 O, OK. I think that it is worth at least documenting WCK properly. 05:49:01 what's 310031-in-base-5 in decimal? 05:50:46 If you mean 310031 is a base five number, then the decimal is 10016. (You can figure out this by dc easily. It can also be figured out easily in JavaScript and also in PostScript.) 05:51:17 I thought dc might be able to do it, but I wasn't sure how 05:51:33 thanks for the answer, anyway, you interpreted the question correctly 05:51:56 How to do in dc is: 5i310031p 05:54:33 `dc 5i310031p 05:54:35 dc: Could not open file 5i310031p 05:54:39 `` dc -e 5i310031p 05:54:40 10016 05:54:42 aha 06:06:39 -!- mig22 has joined. 06:12:22 -!- mig22 has quit (Ping timeout: 256 seconds). 06:19:56 -!- mig22 has joined. 06:19:56 -!- mig22 has quit (Client Quit). 06:22:05 -!- Cale has quit (Ping timeout: 240 seconds). 06:25:02 `` dc -e 5i200310031p 06:25:04 791266 06:29:42 -!- Cale has joined. 07:19:15 -!- ArthurStrong has joined. 07:27:23 -!- xelxebar has quit (Ping timeout: 240 seconds). 07:40:21 ais523: DOS editors added the ^Z, not because they needed it, but for backwards compatibility with DOS 1. it's like how all windows programs and editors write crlf line endings by default, despite that most windows programs read lf just fine. 07:42:09 int-e: CP/M and DOS version 1, because DOS version 1 has CP/M's record-based file access. DOS 2 fixes this, adding unix-style open/close/read/write calls (NB. DOS 1 doesn't even have a close), and file descriptors that can point to things other than disk files, so you can redirect the stdout of programs. DOS 3 adds some more calls. 07:45:37 ais523: like storing function return address a single byte on the stack, provided that there are no more than 256 call sites to the function after inlining? 07:47:37 b_jonas: I think that might be incompatible with return-oriented programming 07:47:51 because it requires you to track what functions are currently running 07:54:01 what is record-based file access? 07:56:26 I'm pretty sure that return-oriented programming is usually unintended behavior 07:57:34 (usually) 07:58:03 btw, do we want a Planet of the Hats comic strip list, in the sense of `*list ? 07:58:33 it also seems like that would only be useful on old CPUs... it would be hell on branch predictors 08:01:23 int-e: "why do I have 17 linux kernel images installed" => I don't delete the old ones either until the space on the one gigabyte sized boot gets low 08:02:58 kmc: in CP/M and DOS 1, can't just ask the kernel to read or write any number of bytes, instead you can only read a fixed size block, which corresponds to a block stored contiguously on the disk. as a consequence for this, the file size isn't stored in byte granularity either. 08:04:49 this was the old default until unix came along, and it would still make sense for regular files, though with a variable block size that the kernel decides for you, because why the heck would you need a buffer for regular files both in the kernel and in userspace? 08:04:52 what's the block size, and is it the same for every file? 08:04:59 mm 08:05:04 kmc: on CP/M and DOS 1, they are the same 08:05:23 these days, it's only the same on one file system, might vary after that 08:05:26 so ^Z is used to indicate the actual end of the file within the block? 08:05:37 back then it was the same globally, because there was only one type of file system 08:05:41 right 08:05:42 kmc: for text files, yes 08:05:52 other files like executables can have any format the programs handle 08:06:09 it makes a lot of sense for floppy drives - the only writable sections are blocks 08:06:16 also, why ^Z 08:06:26 and in the early days of DOS and CP/M, that's all you really had 08:06:29 doesn't ASCII have a C0 control code for end of file? is it ^Z? 08:07:02 kmc: that was discussed earlier in chat, we don't know why ^Z but it would be stupid to change it now 08:07:11 it's kind of tragic how the C0 control codes, most of which are never used, occupy prime 1-byte real estate in UTF-8 08:07:14 c'est la vie 08:07:17 -!- Sgeo__ has quit (Read error: Connection reset by peer). 08:07:44 -!- Sgeo__ has joined. 08:08:19 anything >0x7f is undefined in ascii, right? 08:08:33 Hooloovo0: that hasn't changed, even these days hard disks and SSDs are written in blocks. in fact, with hard disks having controllers built onto them, you can't even do anything about it, whereas back in the old days you had some control over the format of your floppies. 08:08:52 The ASCII control code for end of file isn't ^Z; that is due to CP/M, not ASCII. 08:09:03 Hooloovo0: although file systems add an abstraction layer over that, where a logical block for storing files can be larger than the physical sectors 08:09:15 (and the even more obscure C1 codes occupy 1-byte real estate in ISO 8859, but the 8-bit Windows code pages put printing characters there instead) 08:09:32 yeah, I know that, but given the age and early status of the OSs, an there were no real alternatives to floppies... 08:09:36 Hooloovo0: correct. ASCII is a 7-bit code 08:09:59 Hooloovo0: sure there were, they were casettes, only those are much less reliable than floppy disks 08:10:13 then how can you have C0/C1 codes? that's 8 bits 08:10:26 technically, sure 08:10:28 C0 and C1 are not meant to be hex numbers here 08:10:32 ohhhh 08:10:32 they're just called that for some reason 08:10:33 and ROM cartridges of course, which are _more_ reliable than floppy disks, but you can only write them in huge production runs 08:10:52 * Hooloovo0 is currently in the middle of a tape backup 08:11:05 C0 codes are 00-1F hex and C1 codes are 80-9F 08:11:24 https://en.wikipedia.org/wiki/C0_and_C1_control_codes 08:11:33 > ASCII defined 32 control characters, plus a necessary extra one for the all-1 DEL character (needed to punch out all the holes on a paper tape and erase it). 08:11:34 :1:36: error: parse error on input ‘,’ 08:11:43 oh! so that's why DEL is 0x7F 08:11:45 lol 08:11:53 yeah, it's because of paper tape 08:11:56 always wondered why they put it at the end 08:12:00 kmc: and poor \x7F would be an orphan out of blocks if it weren't for that it used to be one of the more useful control characters for 7-bit paper tape 08:12:14 yeah, you were faster 08:13:55 08:16:12 Some people (such as myself) still use ASCII, though. 08:17:02 did you get my BELs? 08:17:50 Yes, although it only displayed them because my IRC client filters control codes 08:18:31 I didn't, nor are they in the logs 08:19:13 "rice-\a\a-a-roni, the san francisco treat" 08:19:57 They are in the logs. (However, maybe you are looking at a different file than I am looking at.) 08:20:04 who posted BELs >:( 08:20:05 oh I meant my local logs 08:20:12 * moony yells at Hooloovo0 08:20:20 -!- imode has quit (Ping timeout: 246 seconds). 08:20:29 kmc: O, OK. 08:21:06 Hooloovo0: a lot of older IRC clients (NOT mine, i'm just messing with you) will actually cause the terminal "bell" (usually a beep) to go off if one is sent in a channel 08:21:09 * Hooloovo0 BELs at moony 08:21:10 be careful :P 08:21:16 ceci n'est pas une ␇ 08:21:27 yeah, I know 08:21:59 I've IRC'd from a vax with an appropriately-aged terminal 08:22:10 I got a good amount of bells 08:22:26 apparently unicode fucks it up a lot 08:22:31 nice 08:22:35 * moony may be getting a chance to remote into a VAX tonight, on that note. Moony is happy. 08:22:36 does it have a mechanical bell 08:23:03 my terminal doesn't 08:23:15 but I know a couple people with ASR33s... 08:24:54 hmm, I wonder if I can figure out how to make BEL make asterisk do a single ring... 08:25:38 there's currently 3 physical bells on my desk, and one not far 08:27:28 rofl 08:27:30 that'd be a fun hack 08:27:36 do you collect old phones or something? 08:27:53 yeah 08:28:06 I've got a couple channel banks 08:28:23 I've been looking for a decent non-asterisk pbx 08:29:07 I'm more into phone systems, but it's hard to be into that without also collecting phones 08:30:45 -!- LKoen has joined. 08:30:50 -!- LKoen has quit (Read error: Connection reset by peer). 08:31:12 Did you make a telephone with a bell? 08:31:35 -!- LKoen has joined. 08:31:35 -!- LKoen has quit (Read error: Connection reset by peer). 08:31:52 no, bell made telephones with bells 08:32:10 `? Hooloovo0 08:32:13 Hooloovo0 is a calculator brain surgeon. 08:32:21 Huh. 08:32:26 -!- rain1 has joined. 08:33:28 hi 08:33:54 https://beatriceco.com/bti/porticus/bell/telephones-princess.html if you're unfamiliar with the Bell System 08:34:52 cool 08:35:04 I have a rotary phone that i bought at a hamfest 08:35:11 I took it apart and then didn't do anything with it :/ 08:35:13 feel kinda bad 08:35:24 i could give the parts to someone who wants them 08:35:38 there was supposed to be another hamfest on Saturday, but it got cancelled :/ 08:35:47 I took apart a working IBM PC keyboard when I was (much) younger. Still regret it, perfectly good keyboard ): 08:35:49 as did the big Hamvention in Ohio in may 08:36:03 i feel taking things apart is virtuous, though 08:36:07 in general 08:36:14 yeah, taking things apart is good 08:36:21 especially if you learn something 08:36:38 if a piece of electronics gets taken apart, it died a good death 08:36:39 yeah 08:36:46 All I learned was that the keyboard controller was a Zilog made one 08:36:54 apparently 08:37:13 (Was probably a clone's keyboard? I forget, parts are long gone) 08:37:30 I've got a couple rotary dials sitting around in a box 08:37:39 came from the junk store 08:37:52 I kind of wish they just sold the phones... 08:42:32 I might suggest to make a telephone out of those parts, then. 08:45:16 I wish I had time for that 09:04:47 -!- xelxebar has joined. 09:14:45 -!- LKoen has joined. 09:17:06 [[Pxem]] https://esolangs.org/w/index.php?diff=70300&oldid=68572 * YamTokTpaFa * (+5) Googled what is to be stack-oriented paradigm. 09:53:14 -!- LKoen has quit (Read error: Connection reset by peer). 10:02:51 -!- ArthurStrong has quit (Quit: leaving). 10:09:43 ais523: how many underscores would this count in your scale: " I've IRC'd from a vax with an appropriately-aged terminal" 10:09:58 I want to calibrate that scale, even though I'm not using underscores that way myself 10:10:13 -!- cpressey has joined. 10:11:23 " there's currently 3 physical bells on my desk, and one not far" => there are three in my drawer, but they're not electric, they're tiny decorational cowbell style bells that you can only ring by hand. plus there's a wind bell hung next to me at the bookshelf, also ringable by hand only. 10:12:08 the three on my desk are all in telephones, as is the other 10:12:34 I can also ring the one in the kitchen 10:13:13 * Hooloovo0 only has one desk drawer, and it's filled with calculators 10:13:25 oh yeah, I have an electronic doorbell, an electronic bell in my home phone, and an electronic bell in the door phone 10:13:29 like, I have too many calculators that thy don't all fit 10:13:30 b_jonas: it depends on who owns the vax; if you don't have admin rights there and are using telnet or something, then 2-3; if you can install your own IRC clients on it, probably only 1-1½ 10:14:01 yeah, I installed my own IRC client 10:14:10 the drive as I got it was useless 10:14:21 but the drive worked! 10:14:41 (had useless stuff on it, not broken) 10:15:16 installing 4.3bsd isn't too trivial when you don't have a tape drive 10:18:59 b_jonas: the best way to think of it is that it's a scale based on how much you hate your setup, which means that an underscore-laden setup only happens if you have no real ability to change it (otherwise, you would have changed it rather than using a setup you hate) 10:19:01 hello 10:20:45 [[Spiral Rise]] N https://esolangs.org/w/index.php?oldid=70301 * Ais523 * (+4759) now that I'm sure this can be implemented in The Waterfall Model (I ran some compiled programs against a simulation in a spreadsheet and compared the internal states), it's worth adding to the wiki 10:20:49 I see 10:20:52 b_jonas: I'm imagining you signing in as b_____jonas 10:20:54 I see 10:21:02 well, The Waterfall Model is TC 10:21:09 but I meant, implemented with a small program 10:21:37 [[Language list]] https://esolangs.org/w/index.php?diff=70302&oldid=70278 * Ais523 * (+18) /* S */ +[[Spiral Rise]] 10:21:47 Taneb: I don't use that scale. if I sign in as b_jonas_ or wib_jonas_, it's only because the main nick is still connected. 10:21:55 and even b_jonas__ means only that. 10:21:55 [[User:Ais523]] https://esolangs.org/w/index.php?diff=70303&oldid=67193 * Ais523 * (+17) +[[Spiral Rise]] 10:22:00 I... use completely different nicks 10:22:01 and b_jonas-2 10:22:11 -!- ais523 has changed nick to callforjudgement. 10:22:12 I use nc_jonas and such things for stupid setups 10:22:15 -!- callforjudgement has changed nick to ais523. 10:22:29 (I think atriq is currently set as my fallback, which is rot13 of approximately my initials) 10:24:30 I use variations of the main one 10:24:40 Hoolootwo, Hooloo42 10:24:42 incidentally, something I noticed recently: a program in The Waterfall Model with n non-halt waterclocks can be compiled into a FRACTRAN program with N+1 fractions 10:25:48 the basic idea is to use a prime for each waterclock (including halting waterclocks), then multiply together all the primes but one in order to implement a zeroing trigger 10:26:36 then if none of those fractions can run, a final fraction does the steady decrement 10:27:01 halting waterclocks aren't given fractions, but are still given primes, so if one of them zeroes, none of the fractions can run because their prime is multiplied into every denominator 10:27:48 hmm, I should post that on the Waterfall talk page 10:28:39 wow, very cool! 10:28:45 [[Talk:The Waterfall Model]] https://esolangs.org/w/index.php?diff=70304&oldid=58808 * Ais523 * (+863) /* Relationship to Fractran */ new section 10:31:18 -!- wib_jonas has joined. 10:31:48 I haven't been logged as b_jonas and w?b_jonas simultaneously for a very long time 10:34:13 my vax is unfortunately not with me right now 10:34:25 hmm, might be able to get my 486 box in here 10:37:03 -!- rain1 has quit (Remote host closed the connection). 10:43:14 -!- Hoolootwo has joined. 10:43:40 -!- LKoen has joined. 10:43:48 Client: HexChat 2.12.4 • OS: Gentoo Linux (stable) • CPU: 486 DX/4-WB (0MHz) • Memory: 45.4 MiB Total (20.4 MiB Free) • Storage: 208.0 kB / 96.6 MB (96.4 MB Free) • Uptime: 2d 9h 49m 32s 10:44:56 that's a surprisingly empty disk 10:45:08 do you have just a kernel and an IRC client and nothing else? 10:45:18 it's netbooted 10:45:34 mounting an nfs? 10:45:57 I'm pretty sure hexchat's sysinfo script doesn't know how to handle that 10:46:00 yeah, hnfs 10:46:04 nfs, even 10:54:23 [[Spiral Rise]] https://esolangs.org/w/index.php?diff=70305&oldid=70301 * Ais523 * (+3664) this is almost certainly TC, but I guess that technically, I haven't proven it 11:01:05 argh, where do I change the font antialiasing settings on windows 10? 11:01:22 this windows is set to English and yet I can't find the fucking setting 11:02:37 Technically I never proved Burro is TC either. That's on my TODO list for this year sometime. 11:02:47 wib_jonas: does searching for "ClearType" in the various control panels help? 11:03:08 it can take a while to find all the settings programs the one you can get at by right-clicking the start menu is fairly comprehensive though 11:03:54 ais523: ah yes, now it worked. I typed it to the control panel search first, and it didn't find anything. now I typed it to the Settings, and it did find the option, and opened a control panel app. wtf. 11:06:23 and yes, searching apps in the start menu for "ClearType" finds it too, that's where I should have started 11:06:49 I wonder if "antialiasing" also works, or whether you have to use the marketing name 11:08:06 "antialiasing" doesn't, but there are probably other keywords that find it 11:10:03 oh 11:10:06 "subpixel rendeering" 11:10:11 *"rendering" 11:10:17 I doubt that works either, but it's a more accurate name 11:10:44 [[Spiral Rise]] https://esolangs.org/w/index.php?diff=70306&oldid=70305 * Ais523 * (+2325) comparison with High Rise 11:11:35 MTGBusyBeaver42 (if you're logreading), and anyone else interested: I'm now pretty much convinced that 6 waterclocks + 1 halt waterclock (which can of course flood) is enough to create a UTM in The Waterfall Model 11:14:13 [[User talk:Ais523]] https://esolangs.org/w/index.php?diff=70307&oldid=70205 * Ais523 * (+543) /* The Waterfall Model UTM */ see [[Spiral Rise]], which can be implemented with a 6+1-waterclock model and is almost certainly TC 11:15:21 corollary: Fractran is almost certainly TC with 6 fractions, although the size of the fractions is probably going to be astronomical 11:18:06 The coronavirus liability cascade has reached our boss, so he told us to work from home. This just two workdays after we could have started to work more efficiently in the office, because all the interns who normally come in only a few days a week can now work full-time since universities are closed, and this office happens to have few parents who 11:18:07 have to stay home because schools are closed. 11:18:48 So I'm setting up the work notebook for use at home. The monitor and lighting conditions are better there, but I have to change the ClearType settings, and possibly also the gamma setting. 11:19:45 I'm glad I brought the work mouse home. It's worse than my home mouse, but my home mouse is for PS/2 port, and the notebook doesn't have a PS/2 port. 11:20:08 Though I'll have to check which direction passive converters work, I think I have two in my drawer 11:20:52 Yeah, I remembered right, the passive converters let newer USB mouse to plug into PS/2 ports. That doesn't help me. 11:21:04 Oh well. 11:22:36 Anyway, this results in the unprecedented state that I'm not only connected as b_jonas and wib_jonas at the same time, but also they are NATed to the same public IP address. 11:48:54 -!- Lord_of_Life_ has joined. 11:50:02 -!- Lord_of_Life has quit (Ping timeout: 240 seconds). 11:50:12 -!- Lord_of_Life_ has changed nick to Lord_of_Life. 12:01:00 -!- ais523 has quit (Quit: quit). 12:07:40 [[Deadfish]] https://esolangs.org/w/index.php?diff=70308&oldid=70254 * Chris Pressey * (+1905) /* Implementations */ I, Chris Pressey, the author of this here implementation of Deadfish, do hereby huck it into the public domain 12:08:52 [[Deadfish]] M https://esolangs.org/w/index.php?diff=70309&oldid=70308 * Chris Pressey * (+29) /* Robin */ nowiki 12:11:56 oh nice, I got academic spam from "head chief" at publisher called "International Organization of Science Research" 12:24:44 [[Deadfish]] https://esolangs.org/w/index.php?diff=70310&oldid=70309 * Chris Pressey * (+18) Make link point to WayBack Machine. Apparently the cafe closed sometime in 2015 or 2016. 12:36:40 -!- arseniiv has joined. 13:30:29 [[Deadfish]] https://esolangs.org/w/index.php?diff=70311&oldid=70310 * Chris Pressey * (+464) I, Chris Pressey, who done authored this Deadfish interpreter, hereby hurl it into the public domain. 13:33:14 Subpixel reindeering is what Santa uses to get around the world so fast. Not sure how. 13:34:51 Also working from home, incidentally. Kind of interesting, after having just swapped the Linux laptop to a ChromeOS tablet (rationale: "I only ever use this for browsing on work travel, anyway, not for real work"), which has a total of two USB-C ports, and that's it for holes in it. 13:39:16 this one doesn't have a DVI port either, but I had the sense to bring the HDMI cable from work 13:39:58 also, these fans need to be tuned, my home computer and the work computer are giving noise in disharmony. 13:40:15 (mostly I should just replace my home computer with a not so loud computer) 13:46:48 I did buy one of those USB-C dongle things (2x USB-A 3.0, 1x USB-A 2.0, HDMI, Ethernet, SD/micro-SD, DC pass-through from a USB-C charger) "on the off chance I'll need to hook up the monitor and keyboard, though that's not very likely, I don't usually work from home", which turned out to be a good idea. 13:47:42 this notebook has a docking station, but it still has very few ports 13:47:53 no DVI in particular 13:48:26 nor a microphone port 13:49:57 it's mostly useless 14:00:00 `olist 1195 14:00:11 olist 1195: shachaf oerjan Sgeo FireFly boily nortti b_jonas 14:00:52 BTW all OOTS books are on 50% discount until later today wih coupon "washyourhands" 14:01:15 Sgeo__: does that discount the shipping too? 14:01:20 because the shipping is the expensive part 14:01:42 Oh, just the e-book versions 14:01:55 ah 14:02:06 well that can be helpful, let me check 14:04:49 Sgeo__: that would actually gives an almost acceptible price for the e-books. I'll have to consider this quickly. 14:17:24 also, these fans need to be tuned, my home computer and the work computer are giving noise in disharmony. => do you have a microwave? It could alleviate the situation. Or maybe a vacuum. Or maybe both, this could end up as a neat seventh chord 14:18:46 arseniiv: I have a microwave in the kitchen, yes, and I have a vacuum cleaner too 14:18:54 also a washing machine and a dishwasher 14:19:04 now bring them all in one room, yes, and a machine too 14:19:08 and a gas oven 14:19:13 hm I don’t know how dishwashers sound 14:19:42 presumably a gas oven would be almost silent 14:19:52 that would be hard, because the oven absolutely can't be moved, and the dishwasher and the washing machine at least need to be connected to water and drain pipes while under operation 14:20:42 hmhmhmhm then you need to place a couple microphones in there and… hm too time-costly 14:20:44 so I'd have to move the washing machine into the kitchen, which would need extra plumbing parts, and I'd really rather not do the installation myself 14:20:58 taking the computers there wouldn't be too difficult at least 14:21:17 yeah at least you could take a vacuum with you 14:21:23 I don't need microphones, I can just open the doors 14:21:28 this house isn't *that* big 14:22:19 also the kitchen doesn't have a floor drain, and the floor is wooden, so running a washing machine there would be impractical 14:23:03 agree 14:23:39 on the other hand, these computers can make a variety of noises other than the fan 14:24:05 there are speakers, three sets of them (one that I normally use for the computer, one built into the monitor, and one built into the work laptop), plus hard disks 14:24:36 and even a chasis fan in the home computer, which I've disabled long ago, but I can reenable it easily 14:24:46 alas a hard disk is hard to reason with 14:45:24 On the Origins of PCs is free with the coupon I believe 14:46:53 Sgeo__: I already have Origins and Darkness and Good Deeds. I'd probably buy book 4 and 3 next. 14:47:03 It's sad that book 6 still isn't listed in https://www.giantitp.com/Shop.html 14:47:16 so you have to find a link to it from the news archive or the forum if you want to buy it 14:48:55 or find it linked on Gumroad from the other books 14:50:14 but half price for book 4 is still 2600 forints for a book that contains mostly the same strips as the web, so I'll skip for now 15:45:22 b_jonas: Well, I don't like that the kernels accumulate like that, especially because removing them with aptitude takes quadratic time. There must be a workaround for that. 15:49:53 b_jonas: e.g. http://paste.debian.net/1135151/ 15:51:48 int-e: they don't seem to accumulate just from debian for me, at least not to the amount you mention. debian only keeps one from each abi compat version, which is a level below the kernel version number but above the actual debian package version number, and supposedly increments only when the abi becomes incompatible between kernel and modules 15:52:32 wib_jonas: I'm un debian unstable, which has many more kernels (and a good reason to keep *some* around. I just wish it wasn't every single one, ever) 15:52:41 whereas they often do kernel updates within an abi version, and I usually copy those kernels to my boot partition now (but only because of this weird setup; and I don't copy the modules) 15:52:47 I'm *running 15:52:48 oh, debian unstable. that's different 15:52:59 I don't run unstable, and wouldn't run it 15:53:37 it takes quadratic time to remove them? ouch 15:54:11 it's no surprise that debian unstable has more kernel versions 15:58:24 The obvious improvement here would be to not regenerate the initrd after every single purge. Not sure how to achieve that. 15:59:49 how does that even become quadratic? 16:00:21 I mean, maybe it regenerates the initrd with the current latest kernel version's modules every purge, but that's just linear, 16:00:39 plus it could regenerate the boot loader config file after every purge, which is technically quadratic but with a very small constant factor 16:01:46 wib_jonas: because for some inexplicable reason, locating each of these images takes non-negligible time. 16:02:04 so the initrd generation becomes noticably slower with more kernel images 16:02:43 also, out of curiosity, are you running debian unstable as a hobby for home, or for some job in sysadmin or similar capacity? 16:03:24 This is at home. I wouldn't do it for servers. I would probably do it for my own software development needs if allowed. 16:03:54 not on live servers, sure, but are you doing it to be able to prepare for future system problems at work? 16:04:01 It's a matter of laziness really... debian unstable is much better at having the latest versions of compilers and the like. 16:04:07 I'm asking more "for home" rather than "at home", 16:04:20 I see 16:04:43 And it is surprisingly stable, given the name. 16:05:28 Make no mistake, I've seen it break badly... happens once, maybe twice a year, and *most* of the time it's easy to fix. 16:06:05 sure, but I prefer that to happen not at unpredictable times, but at the times when I decide to upgrade to the latest stable version and spend days setting it up and debugging problems 16:06:17 The last such annoyance was firefox relying on sqllite interna. 16:06:36 I don't do automatic updates anyway. 16:06:42 and for the software where I particularly care about having the latest version, which could be compilers, or gimp 2.10, or whatever, they're usually easy to install 16:06:51 at least if you're not running a too old stable version 16:07:30 Anyway, obviously this is not any kind of advice. For me, Debian unstable hits the sweet spot between stability and bleeding edge better than debian stable. 16:08:01 And that may change. 16:08:20 sure, I don't do automatic updates either, I just put the security updates on occasionally, but those rarely break anything but firefox (which is its own sort of problem, debian basically doesn't have the resources to follow its changes), whereas for unstable you can't install just the security updates every month 16:08:30 sure, it can change 16:09:46 (I mean the above for my home machine; the work windows machine does run automatic updates) 16:10:00 Also... I ran Gentoo for several years. So compared to that, Debian unstable is MUCH more convenient. 16:10:28 (I can't justify Gentoo anymore except as a learning experience into how software is built and how many things can go wrong in the process.) 16:10:38 meanwhile, IOCCC submissions have ended 16:10:49 sure, I ran Gentoo too, but that was when I was young and had time for that sort of nonsense 16:10:57 and didn't know that I should be running debian 16:10:59 [[TTTTTTTT]] N https://esolangs.org/w/index.php?oldid=70312 * PythonshellDebugwindow * (+853) Created page with "'''TTTTTTTT''' is an esolang by [[User:PythonshellDebugwindow]]. ==Etymology== This language got its name from its longest keyword, which is also the only character in its ke..." 16:11:02 At least I have one or two good stories to tell. 16:11:44 Like the point where I removed libacl because I wasn't using access control lists... and core-utils (uh, what's the old name of that again?) broke. 16:11:48 eventually I managed to break the gentoo system too much, at which point I tried ubuntu for a few weeks, decided it wasn't for me and purged it without a backup, installed debian, and I've been running debian ever since (though multiple different installations of debian) 16:11:52 I ran Gentoo when I got an amd64 machine and Debian still didn't have good support for it 16:12:00 that was in 2004 16:12:08 when I was young and had time for that sort of nonsense 16:12:10 . o O ( Debian unsatiable ) 16:12:56 kmc: Yeah I basically stopped using Gentoo when I finally got a job. :P 16:13:04 …I need to try Debian in a virtual machine probably… 16:15:47 [[User:PythonshellDebugwindow]] M https://esolangs.org/w/index.php?diff=70313&oldid=70264 * PythonshellDebugwindow * (+96) /* "Hello" program */ 16:17:35 wib_jonas: Another related compromise would be Ubuntu (basically minus all the Ubuntu specific UI crap), which is sourced mainly from Debian unstable. 16:21:31 [[TTTTTTTT]] M https://esolangs.org/w/index.php?diff=70314&oldid=70312 * PythonshellDebugwindow * (+381) 16:22:06 [[Infinite Goto]] M https://esolangs.org/w/index.php?diff=70315&oldid=70299 * PythonshellDebugwindow * (+276) /* Syntax */ clearing up Line 27 16:22:44 [[Infinite Goto]] M https://esolangs.org/w/index.php?diff=70316&oldid=70315 * PythonshellDebugwindow * (+48) /* Resources */ 16:26:02 -!- Melvar has quit (Quit: WeeChat 2.4). 16:28:11 int-e: well, it's been lots of years since I tried ubuntu, and it's changed during those years, so I can respectfully say that I'll consider that option and get back to them, and then never actually get back to them 16:31:36 -!- Melvar has joined. 16:32:47 I used to use FreeBSD, back in the day. These days I use Ubuntu. 16:33:30 i sold out and switched to windows 16:35:22 how could you 16:35:40 and do you use it through anything other than the wsl 16:37:44 I use Ubuntu, although without the desktop environment and window manager that was initially installed. 17:16:15 -!- xkapastel has joined. 17:17:42 -!- imode has joined. 17:29:06 zzo38: yeah that's basically what I meant ;) 17:29:33 OK 17:30:10 cpressey: Why FreeBSD and not OpenBSD or NetBSD? ;-) (No, I never had a dog in that fight. I don't even really know what the fight was about... just that there was one.) 17:30:58 (I use the i3 window manager, with no desktop environment. Programs should not usually be written to require a desktop environment; they should work without it, even if without a desktop environment may require some command-line options to be used which are not otherwise used.) 17:32:03 It is lamentable... the latest development on that front is that certain programs have stopped to respect their working directory. 17:33:15 All of GTK, I think. So if I start inkscape in a subdirectory and try to open a file I'll be in my home directory. Meh. 17:34:25 int-e: I don't care too much about them respecting the wd, but the gtk file dialogs are much worse than the windows ones, into which you can just paste a filename or a file from explorer 17:34:28 int-e: I actually like NetBSD better, from an aesthetic point of view, but FreeBSD has more users and a more pragmatic outlook, so it's the easier choice. 17:34:43 this is an area where windows manages to "just work" better 17:36:34 I don't know. I like shells. I use working directories as contexts for projects. It's sad to see software add obstacles to that process, just because most users use visual desktop enviroments to manage their files. 17:37:24 But it's not terrible, I complain and cope. 17:38:20 I guess one big reason is that windows is less affected by a proliferation of toolkit libraries. 17:38:38 (though it's mainly Qt and GTK these days.) 17:40:59 Yes, the Windows file dialog does work better tha GTK, although I would think is OK to just a simple text entry. 17:42:26 @tell oerjan Agatha is such a terrbible mad scientist. 17:42:26 Consider it noted. 17:44:19 int-e: Agatha does also have to be a sovereign 17:44:35 Which requires... diplomacy 17:44:45 Agatha? Krosp, yes. Agatha... hmm. 17:45:23 Still having second thoughts, how does she stand it? It's so inefficient. 17:50:34 -!- Mandarina has joined. 17:52:13 -!- Mandarina has quit (Quit: IRcap 8.7 ). 18:00:02 -!- cpressey has quit (Quit: WeeChat 1.9.1). 18:02:36 Do audio tapes support index marks like VHS does? As far as I know, it doesn't, although I would think it would be a useful thing to have (or to fake it somehow). 18:16:01 "carnivorous" is an interesting word these days... needed two attempts to read it correctly :) 18:16:25 zzo38: weren't there some fancy tape recorders that could scan the tape for silence? 18:19:40 (That's how it worked when doing it manually anyway... wait for interruptions inbetween all the squeaking.) 18:23:42 Maybe there is; I don't know. I think that there are four tracks (two on each side), so possibly one track could be used for index. 18:27:08 -!- Hoolootwo has quit (*.net *.split). 18:27:08 -!- Deewiant has quit (*.net *.split). 18:27:31 -!- Deewiant has joined. 18:29:01 -!- Phantom_Hoover has joined. 18:35:45 -!- Hoolootwo has joined. 19:01:36 -!- FreeFull has joined. 19:09:08 -!- Deewiant has quit (Quit: WeeChat 2.3). 19:09:16 -!- Deewiant has joined. 19:14:13 int-e: yes, but for resource reasons I only want one instance of gimp or other GUI programs (sometimes two instances or firefox), and not restart them every time, so the working directory isn't too useful as a context for them. 19:14:52 zzo38: for audio tapes, you just write the index onto a carboard slip that you put into the casette holder 19:16:43 -!- Deewiant has quit (*.net *.split). 19:17:00 -!- Deewiant has joined. 19:44:48 -!- FreeFull has quit. 19:50:18 wib_jonas: Yes, if you use the tape counter for this purpose, I suppose. 20:01:02 -!- wib_jonas has quit (Remote host closed the connection). 21:02:34 -!- Phantom_Hoover has quit (Ping timeout: 246 seconds). 21:03:27 -!- FreeFull has joined. 21:06:17 Can be disabled in Firefox? I am fine with , but I don't like . 21:20:18 Rosewater wrote about themes in Magic: the Gathering, including color, type, subtype, supertype, zone, and other possibilities. 21:20:54 I thought, make the zone theme based on stack, ante, and command zone. 21:49:24 whatever happened to the battlefield and phasing 21:50:13 Phasing is no longer a separate zone than the battlefield, although that can be a theme too I suppose 21:57:46 I'm trying to draw a tiny bitmap font where characters are 6 pixels high. so now I have to research which accented letters languages use, because I have to make some of the following characters look identical: a-grave, a-acute, a-circumflex, a-diaresis, a-tilde, a-brevis, a-macron, a-dotabove. 21:58:14 I decided early that o-circumflex and o-doubleacute can look the same, that's the easy part. 21:58:42 And I don't think this font will be usable for vietnamese, 21:59:51 and it would be better if the same accent looks the same on most vowels 22:03:39 so I have to be concerned with french/welsh that distinguishes e-acute, e-grave, e-circumflex, e-diaresis; portugese which distinguishes a-acute, a-grave, a-circumflex, a-tilde; romanian that distinguishes a-brevis from a-circumflex; swedish/finnish which distinguishes a-umlaut from a-ring; icelandic that distinguishes o-acute from o-diaresis; 22:07:44 as well as some romanizations of non-latin scripts, in particular the ISO 9 transcription of russian which distinguishes e-circumflex from e-diaresis (ok that's not a new constraint); I can probably give up on chinese pinyin transcription just as on vietnamese; what else am I forgetting? 22:12:10 oh yeah, also czech distinguishes u-acute from u-ring. 22:15:02 I think I'll do the following: grave is left pixel, acute is right pixel, umlaut/diaresis is left and right pixels, circumflex or double acute is middle and right pixel, tilde is left and middle pixel, macron is left and middle and right pixels, ring or brevis or middle dot is middle pixel alone, and I'll figure out what to do with the caron later, it has to be distinct from the middle dot and the acute 22:15:08 obviously 22:15:57 no wait, that won't work, darn it 22:16:14 the i with brevis has to differ from the i 22:16:44 for latin 22:17:33 -!- Hoolootwo has quit (Remote host closed the connection). 22:18:37 still, brevis as the middle dot is too mnemonic to give up. I might just special case the i with brevis 22:25:37 ideally I should have a big table prepared for this that shows what letters each language and writing system that I commonly encounter use 22:29:57 I also thought a theme based on a supertype, could be world, and you can do more than what other sets have previously done, such as adding death triggers, world cards other than global enchantments, effects that add or remove the world supertype, that suppress the world rule temporarily, count world cards in some zone, etc. 22:31:41 I actually have such a table for cyrillic, with five languages using cyrillic script, the serbian latin transcription, and the ISO 9 transcription (these two agree on every letter I think, but note that romanizations of ukranian differ, because і vs и vs ы) 22:36:23 I can probably just use similar diacritics on most consonants: acute is right pixel, circumflex is middle and right pixels, dot above is middle pixel, and I think I'll mostly mark a caron with the left and middle pixel, except I'll have to figure out what to do with l/L/t/T with diacritics, those are ugly not only because there are so many diacritics that apply, but even the base letters are hard to 22:36:29 draw properly in small sizes and distinguishable from each other and i and c 22:40:10 [[Self-modifying Turing machine]] N https://esolangs.org/w/index.php?oldid=70317 * Hakerh400 * (+30214) +[[Self-modifying Turing machine]] 22:40:16 [[Language list]] https://esolangs.org/w/index.php?diff=70318&oldid=70302 * Hakerh400 * (+36) +[[Self-modifying Turing machine]] 22:40:21 [[User:Hakerh400]] https://esolangs.org/w/index.php?diff=70319&oldid=70279 * Hakerh400 * (+36) +[[Self-modifying Turing machine]] 22:44:53 [[Self-modifying Turing machine]] M https://esolangs.org/w/index.php?diff=70320&oldid=70317 * Hakerh400 * (+5) 22:51:08 oh yeah, the constraint is that this font should be somewhat readable to others, not just to me who knows what the font has, although of course my aesthetic decisions matter more 23:14:28 -!- ais523 has joined. 23:35:25 -!- j-bot has joined. 23:49:10 -!- Lord_of_Life_ has joined. 23:51:53 -!- MTGBusyBeaver42 has joined. 23:52:05 -!- Lord_of_Life has quit (Ping timeout: 250 seconds). 23:52:09 -!- Lord_of_Life_ has changed nick to Lord_of_Life. 23:55:54 -!- xkapastel has quit (Quit: Connection closed for inactivity).