00:00:08 <b_jonas> nakilon: yes, and I think that's exactly the problem that fizzie wants to solve by commands in esowiki
00:00:19 <fizzie> shachaf: General thing. But with that "of sorts" qualifier. It's the thing that https://cacm.acm.org/magazines/2016/7/204032-why-google-stores-billions-of-lines-of-code-in-a-single-repository/fulltext hints at (if you search for "Mercurial").
00:00:28 <nakilon> then have a look at https://github.com/whitequark/irclogger
00:00:33 -!- ubq323 has quit (Quit: WeeChat 2.3).
00:01:01 <fizzie> Not interested in switching to a third-party system for the logs.
00:01:20 <shachaf> I knew Facebook was doing these things and I guess I heard vague rumors elsewhere.
00:01:43 <nakilon> not sure though why there is no Search field in the link that is in repo description but it's here for example https://freenode.irclog.whitequark.org/ruby/2021-03-26#;
00:02:38 <b_jonas> nakilon: if you want to host a searcher, you can always download the logs from the server and make your own
00:03:02 <zzo38> I think the prefix can be e.g. "hajfbhalfjblajdfbot: " to publicly command hajfbhalfjblajdfbot, and to privately command a bot you can send a private message and no prefix is needed.
00:03:45 <zzo38> (And, if a potential loop is detected, it should use a NOTICE command to respond, in order to break the loop.)
00:04:00 <b_jonas> zzo38: I assume that's just an example, and you're aware that Freenode's nick max length is 16
00:04:18 <zzo38> b_jonas: Yes, it is just an example.
00:04:58 <b_jonas> I do wish more bots responded to their name too besides their short prefix, so you can more easily find out what their short invocation prefix is
00:05:01 <nakilon> oh you reminded me I wanted to make a service for Steam
00:05:23 <fizzie> I was writing a Z80 emulator bot at one point, and for that I was planning to use that schema (of just "nick[:,] " as a prefix for public messages, and nothing for private), but never got around doing that.
00:05:25 <b_jonas> fungot does at least respond to its name
00:05:25 <fungot> b_jonas: send to someone else have already liked it, since our project is due to an another option will be faster.,
00:05:34 <fungot> fizzie: done with class me ladki ne bra me chupa liya. decimal inch of snow. enjoy ur crab was running abt getting food before u go in? i miss you
00:05:44 <fizzie> In a very useful way, yes.
00:05:50 <b_jonas> though I think I didn't make termbot respond to its nick, so I'm guilty too
00:06:31 <j-bot> b_jonas: |value error: help
00:06:51 <b_jonas> yeah, it technically responds but not in a useful way that lets you discover anything
00:06:53 <nakilon> there is a char limit for nickname in Steam that is often exceeded when it's in Russian but I've managed to fir the nickname I needed by sqeezing the spaces, capitalizing the words and swapping the letters that look the same, and I noticed that usually it's too hard for people and they cut and ruin their initial ideas
00:07:29 <b_jonas> nakilon: is it often exceeded in Russian because it's measured in utf-8 bytes?
00:07:51 <zzo38> If you need information to discover, you could include it in the real name field maybe
00:07:56 <fizzie> multibot's way of hooking up the scripts to respond to things is pretty weird. Like, the ` prefix for HackEso is set up by having a file called multibot_cmds/PRIVMSG/tr_60.cmd, because '`' is 0x60.
00:08:31 <nakilon> b_jonas, yep, every russian char is twice large
00:09:58 <nakilon> irb(main):003:0> 'ИисусХристосПорноЗвезда'.bytes.size
00:09:58 <nakilon> irb(main):004:0> 'ИиcycXpиcтocПopнo3вeздa'.bytes.size
00:14:46 <nakilon> also sometimes I have a feeling like there are two interfaces for change the nickname in Steam one of which is cutting harder than another one
00:22:45 -!- delta23 has joined.
00:48:51 <nakilon> modern competitive programming fans are something
00:49:11 <nakilon> all these people who put the words "dynamic programming" in every message
00:49:50 <nakilon> I shared the link https://blog.cryptohack.org/twitter-secrets and they decided it's spam since it's in English )
01:32:52 <esowiki> [[Esolang:Community portal]] https://esolangs.org/w/index.php?diff=81562&oldid=81550 * Baidicoot * (+18) /* Others (Unofficial) */
01:54:10 -!- FreeFull has quit.
01:55:58 -!- delta23 has quit (Ping timeout: 240 seconds).
02:02:35 -!- arcsor5 has quit (Quit: My MacBook has gone to sleep. ZZZzzz…).
02:06:35 -!- hendursa1 has quit (Quit: hendursa1).
02:06:52 -!- arcsor5 has joined.
02:06:52 -!- hendursaga has joined.
02:10:41 -!- ArthurStrong has quit (Ping timeout: 240 seconds).
02:15:37 -!- zseri has quit (Quit: zseri).
02:25:36 <nakilon> if median absolute difference between original histogram and histogram of gauss blurred (1) image in the saturation HSV band is somewhere between 0.18 and 0.4
02:26:04 <nakilon> then it's a nice looking high resolution image
02:46:13 <arcsor5> competitive programming fans on quora ask stupid questions like "is iq 145 too low for programming" and "is 20 years old too late to become good at competitive programming"
02:50:40 <esowiki> [[User:Zseri]] https://esolangs.org/w/index.php?diff=81563&oldid=53103 * Zseri * (-18)
02:52:44 -!- ArthurStrong has joined.
02:55:49 -!- delta23 has joined.
03:11:39 <nakilon> arcsor5 attention whoring? idk, I rarely hit quora but there is a subreddit that sounds like your quotes
03:12:11 <nakilon> https://www.reddit.com/r/mensa/
03:31:26 -!- rodgort has quit (Quit: Leaving).
03:41:59 -!- rodgort has joined.
03:52:04 -!- arcsor5 has quit (Quit: My MacBook has gone to sleep. ZZZzzz…).
03:55:30 -!- arcsor5 has joined.
04:01:00 -!- iovoid has joined.
04:03:18 -!- moony has joined.
04:04:20 -!- Bowserinator has joined.
04:35:47 -!- arcsor5 has quit (Quit: My MacBook has gone to sleep. ZZZzzz…).
04:50:13 -!- delta23 has quit (Quit: Leaving).
05:01:04 -!- iovoid has quit (Ping timeout: 258 seconds).
05:01:28 -!- moony has quit (Ping timeout: 258 seconds).
05:05:04 -!- iovoid has joined.
05:05:12 -!- moony has joined.
05:23:53 -!- nfd has quit (Ping timeout: 258 seconds).
05:41:50 -!- ArthurStrong has quit (Quit: leaving).
05:46:59 -!- sprock has quit (Ping timeout: 252 seconds).
06:45:19 <esowiki> [[Esolang:Community portal]] M https://esolangs.org/w/index.php?diff=81564&oldid=81562 * Palaiologos * (-11)
07:01:32 <esowiki> [[Esolang:Community portal]] M https://esolangs.org/w/index.php?diff=81565&oldid=81564 * Palaiologos * (+94)
07:05:49 -!- mich181189 has quit (Read error: Connection reset by peer).
07:06:12 -!- ocharles has quit (Read error: Connection reset by peer).
07:06:14 -!- glowcoil has quit (Read error: Connection reset by peer).
07:06:18 -!- mich181189 has joined.
07:06:27 -!- ocharles has joined.
07:06:38 -!- glowcoil has joined.
07:13:58 <esowiki> [[Esolang:Community portal]] M https://esolangs.org/w/index.php?diff=81566&oldid=81565 * Citrons * (+0) s/below/above
07:31:28 <esowiki> [[Esolang:Community portal]] M https://esolangs.org/w/index.php?diff=81567&oldid=81566 * LyricLy * (-2) Fix wording
07:35:20 -!- Sgeo has quit (Read error: Connection reset by peer).
07:59:32 -!- adminn has changed nick to zeroed.
08:07:08 -!- hendursa1 has joined.
08:09:57 -!- hendursaga has quit (Ping timeout: 240 seconds).
08:26:58 -!- ais523 has joined.
08:27:19 <ais523> <kmc> i should be able to run any program with --dump-source and get all the source code straight out of the binary ← I wrote the Windows installer component of aimake (used by, e.g., NetHack4) to be a quine in that sense
08:27:29 <ais523> it can install the source code for both the program it's installing, and the installer itself
08:27:42 <ais523> although it isn't a proper quine, it just reads its own source from disk while preparing the installer
08:36:49 -!- ais523 has quit (Remote host closed the connection).
08:37:02 -!- ais523 has joined.
08:37:10 <ais523> <fizzie> Sort of on that note (Nash, not git), UK just published the final design of the new £50 note that features Turing ← I'm not sure why we still even have £50 notes, if you withdraw a large amount of money from the bank it usually comes as £20s
08:38:09 <ais523> (I've been present for bank transfers that were implemented using the old-fashioned method of asking for a large amount of physical cash from one bank and taking it to a different bank in person; this may still be the fastest method, although it's somewhat vulnerable to thieves and I wouldn't do it myself)
08:40:36 <ais523> hmm, do we really need [[Category:Stupid family]]?
08:41:11 <ais523> I have half a mind to just speedy it as "unapproved category", but there's no real urgency and it's probably worth discussing over IRC first
08:41:47 <nakilon> when I get cash I choose the largest banknote and it's 5000 RUR that is around 60 ƒ I guess
08:42:10 <nakilon> though not everyone has a change for it, yeah
08:42:28 <nakilon> but fastfood always have a change
08:42:37 <ais523> looks like 100 RUR has a very similar value to £1 (marginally higher)
08:43:01 <ais523> so a 5000 RUR note is a close equivalent to a £50 note
08:43:30 <myname> ais523: i assume it's from the same guy who made the only 2 entries?
08:43:32 <ais523> many shops here don't accept £50 notes, though, there's a perception that £50 notes are only used by criminals, which means that most non-criminals choose not to use them
08:43:52 <nakilon> there are 5 RUR banknotes in theory https://ru.wikipedia.org/wiki/%D0%A0%D0%BE%D1%81%D1%81%D0%B8%D0%B9%D1%81%D0%BA%D0%B8%D0%B9_%D1%80%D1%83%D0%B1%D0%BB%D1%8C#%D0%91%D0%B0%D0%BD%D0%BA%D0%BD%D0%BE%D1%82%D1%8B
08:44:10 <myname> i don't even know if the languages in particular are that special
08:44:12 <nakilon> but in practice you'll find only 50 RUR because lower is replaced by coins
08:44:30 <myname> like, why is stupidbasic stupid? it's just a basic dialect
08:44:33 <ais523> in the UK, the largest commonly used coin is £2 and smallest banknote is £5
08:45:04 <myname> ais523: here, shops have to accept every note if it's within reasonable margin of purchase
08:45:14 <ais523> (technically £5 coins exist, but they're basically just novelties to celebrate special occasions, despite technically being real coins)
08:45:41 <myname> like, if you purchase something for 5€, the shop _has_ to accept a 50€ note
08:45:47 <nakilon> "there's a perception that £50 notes are only used by criminals" -- the way you non-verbally behave makes a great difference in how more likely you'll be considered a criminal
08:46:06 <ais523> the rules here are complex, you have to accept any note to cover a debt, but if you haven't agreed to a sale yet, you can put more or less whatever requirements you like on what currency you'll accept in exchange
08:46:18 <nakilon> (also maybe I'm a criminal?)
08:46:50 <myname> ais523: also, regarding the category, i like how it has exactly 2 entries and the description is "some languages don't share some of the aspects"
08:46:50 <ais523> (although the credit card companies generally require the businesses they deal with to agree not to discriminate against credit card users, as part of their license to be able to accept credit cards)
08:47:15 <ais523> myname: I noted that, it's likeable in an amusement sense but dislikeable in the being-a-reasonable-category sense
08:50:05 <ais523> on a UK keyboard, shift-3
08:50:15 <ais523> it's a pretty common character over here so we have an easy shortcut for it
08:50:28 <ais523> € is altgr-4 because all the reasonable shift-combinations were already taken
08:50:32 <nakilon> oh I'm using US layout I guess
08:50:34 <ais523> at the time the euro was invented
08:51:02 <ais523> it's between space and right ctrl, replacing the right alt
08:51:07 <ais523> because you don't normally need two alt keys
08:51:25 <nakilon> oh £ is opt-3 on my layout
08:51:37 <myname> here, on altgr + numbers, i have ¹²³¼½¬{[]}\
08:51:54 <ais523> on Windows it appears to be equivalent to ctrl-alt in nearly all contexts; on X11 it gives access to a huge range of special characters
08:51:59 <myname> also being the main reason people hate that layput for programming
08:52:44 <nakilon> shift -- !@#$%^&*()_+ ;alt -- ¡™£¢∞§¶•ªº–≠ ;alt+shift -- ⁄€‹›fifl‡°·‚—±
08:52:47 <ais523> oddly, Windows appears to have placed é on ctrl-alt-e / altgr-e on UK keyboards at some point, and some keyboards will even have an é on the bottom-right of the E key to illustrate that
08:52:52 <nakilon> (have fun whoever parses this day log)
08:53:50 <myname> the swiss keyboard is weird. it's the only one i know where caps lock + shift + letter does not necessarily equal to the letter without modifiers
08:54:06 <ais523> myname: oddly, I have {[]} on altgr-7890 despite the fact that I also have physical [ and ] keys (which are shiftable into { and })
08:54:20 <nakilon> I guess using Alt on Windows is something new at all
08:54:39 <ais523> UK keyboards actually have one more physical key than US keyboards do, between ' and Return
08:54:51 <ais523> it produces # unshifted and ~ shifted
08:54:55 <ais523> (where is ~ on a US keyboard?)
08:54:56 <myname> like, caps lock + a is ä, caps lock + shift + a is Ä instead of a
08:55:21 <ais523> nakilon: on Windows (also most Linux programs) Alt is used for keyboard navigation of GUI programs
08:55:23 <nakilon> \|«|» -- this is the key between ' and Enter for me
08:55:40 <ais523> there used to be underlined letters all over the interface (although nowadays the underlines normally only become visible if you hold Alt)
08:55:55 <ais523> holding alt and pressing an underlined letter is equivalent to clicking on the letter, or whatever it labels
08:56:10 <nakilon> ais523 yep, that's why I was surprised to hear about ctrl+alt+e
08:56:17 <myname> now i use alt to switch between irc channels, like god intended it
08:56:25 <ais523> nakilon: that still works with the *left* alt key
08:57:15 <nakilon> myname btw I changed Capslock to switch ENG/RUS
08:57:25 <ais523> I rebound caps lock to compose
08:57:31 <myname> i changes capslock to a global push to talk
08:57:51 <ais523> I still have a caps lock (which is useful on occasion), but it really doesn't deserve a single key of its own so I put it on shift-shift
08:57:54 <nakilon> capslock is my Jump in games
08:58:10 <myname> ais523: not very intellij friendly
08:58:25 <ais523> what does intellij use caps lock for?
08:58:44 <ais523> what does it use that for?
08:59:03 <nakilon> clear RAM and stop lagging for a minute?
08:59:16 <ais523> what a bizarre binding
08:59:21 <ais523> I'd normally expect that to be on Ctrl-something
08:59:27 <ais523> (or Command-something on a Mac)
08:59:31 <myname> nakilon: no, that's alt+f4
08:59:52 <ais523> in most programs, alt-f4 causes additional lag, rather than removing it
08:59:58 <myname> ais523: i find it pretty reasonable, because it's something you are going to use very often
09:00:04 <ais523> this may be a consequence of them doing something like trying to recurse over all memory to free it
09:00:17 <ais523> myname: I find shift-shift fairly hard to press, compared to ctrl-letter
09:00:41 <myname> how so? ctrl is lower than shift and you have to reach the other key
09:00:42 <ais523> also, is this a window that lists open files, or a window that opens files?
09:01:14 <ais523> I move my hands quite a lot when typing, they move onto each other's halves of the keyboard frequently
09:01:27 <ais523> shift-shift forces me to move them much further apart than I'd normally have them, making it slower to type
09:01:32 <nakilon> the O letter isn't far from right Ctrl
09:01:48 <myname> what kind of maniac uses right ctrl?
09:01:56 <ais523> e.g. in the word "type", I type the "y" with my left hand because the right hand is covering "p" which is at the far end of the keyboard, that means my left hand is nowhere near left shift
09:02:05 <nakilon> idk, I don't open files on Windows at all
09:02:15 <nakilon> I mean I use only mouse there
09:02:51 <ais523> when using an IDE, typically I have a tree view of all the files in the source directory in a corner of the screen somewhere, and use a mouse to open files by clicking on that
09:03:22 <ais523> or even more commonly, use implicit open via search, "go to definition", etc.
09:03:37 <myname> that would mean i know where every file is
09:03:46 <nakilon> and on Macbook any Cmd+ combinations are handy because the Cmd is on both sides of Space so you press it with the 1st finger, not 5th
09:03:50 <ais523> I usually do, even at work
09:04:09 <myname> the project is far to big for that
09:04:25 <ais523> when I'm using Emacs, I will usually try to have all the files open simultaneously and just use C-x C-b to switch between them, or search
09:04:34 <nakilon> also on macOS I don't really close files -- I keep them all opened forever
09:04:38 <ais523> although it sounds like myname's projects would be too big for that, too
09:04:45 <nakilon> so it reduces the amount of openings too
09:05:16 <ais523> generally speaking, though, I like project structures which don't have many levels of hierarchy, just a few main directories and lots of files in each
09:05:22 <ais523> so there aren't many places I'd have to look for any given file
09:06:04 <myname> opening a couple thousand files at once sounds like a bad ide
09:06:48 <nakilon> oh btw, while on Windows you may open the modal file browsing dialog and click with your mouse on macOS you press Ctrl+Cmd+G in that dialog and it opens the small submodal window with a single field to chose the path by typing with autocomplete just like in Terminal
09:07:11 <nakilon> so I almost never click in the file browsing dialogs
09:07:39 <nakilon> oh, I mean, I navigate to the folder with that submodal and then click
09:08:23 <nakilon> would be cool if they also add the ability to chose the file in that submodal to make it even more handy
09:09:54 <nakilon> myname I don't edit thousands of files; if I only need to read them I read them in terminal, not IDE
09:10:21 <myname> that sounds like a weird way of programming
09:11:14 <nakilon> and I have terminal and ide opened the same amount of time because two displays
09:11:37 <myname> i have terminals open way more
09:11:49 <myname> how would i irc otherwise
09:16:39 <ais523> hmm, generally speaking it makes sense to have all the text/source files in the project you're working on in memory simultaneously
09:16:57 <ais523> a) text files generally aren't that large, so there's room, b) grepping an entire project is a common operation
09:17:15 <myname> yeah, but you usually don't know which files you are going to work on
09:17:32 <ais523> and c) if you load a file into memory, modern OSes will treat the file itself as swap backing the memory, so there's no actual cost in terms of how much else you can fit into memory
09:18:08 <ais523> hmm… do you think the source code for Windows, minus any binaries / assets / large data files / images / resources etc., comes to more or less than 1GB?
09:18:13 <ais523> I'd guess less, but I'm not sure
09:19:25 <ais523> grepping 1GB of data that's already in memory shouldn't take that long on a modern computer
09:21:18 <ais523> hmm… I wonder whether nontemporal reads would make a significant improvement for that sort of workflow?
09:22:46 -!- kspalaiologos has joined.
09:23:23 <nakilon> I tried to force myself to use the "Projects" in Sublime Text that is kind of opening a new window per project with its folder tree opened on the left
09:23:38 <ais523> modern processors are so fast that most algorithms are bottlenecked on either memory write or memory read, so explicit cache control hints can cause noticeable improvements
09:24:20 <nakilon> but: 1) if there is a folder in it with thousands of asset files it wastes time on rereading the file tree 2) it's harder to navigate between multiple windows rather than one opened all the time
09:25:31 <nakilon> yeah I guess you can blacklist the subfolder in the Project settings but that's a new config file
09:26:42 <ais523> asset files are such an issue for everything programming-related :-(
09:27:04 <kspalaiologos> ais523, that greatly depends on the storage complexity of the algorithm
09:27:05 <ais523> the best solution I know is to have them in a different tree from the source (e.g. assets/ and src/ as subdirectories of your main project directory)
09:27:34 <ais523> kspalaiologos: of course, but what I'm mostly focused on is making the linear-time algorithms faster
09:27:52 <ais523> I've been spending weeks looking at low-level processor performance in order to write the world's fastest FizzBuzz
09:28:03 <ais523> and I think I know how to do it, now, just can't be motivated to actually sit down and write the program
09:29:03 <ais523> but the basic idea is to have a small bytecode program in L1 cache, where each byte of program generates a byte of output, and where a full loop of the program generates 2400 lines of output
09:29:07 <kspalaiologos> oh, lol, i think i made the world's fastest fizzbuzz for the codegolf challenge
09:29:23 <ais523> I had the idea a while ago, forgot about it, then saw it on codegolf and got obsessed
09:29:37 -!- LKoen has joined.
09:29:52 <ais523> all the answers there seem to use write(2) for output at the moment, which inherently bottlenecks you to a fairly low speed
09:30:35 <ais523> because it forces a memcpy of your data inside the kernel, which a) doubles your cache pressure and b) the copy loop itself is slow beacuse you can fizzbuzz as fast as a memcpy
09:30:46 <ais523> I've been looking into using vmsplice(2) but it's much harder to set up
09:30:54 <kspalaiologos> yep, i'm aware of that. i thought that we could probably make this thing bootable
09:30:59 <ais523> the idea is to make sure that the data never leaves the L2 cache
09:31:16 <ais523> oddly, that could actually be *faster* because you could make sure that the data never left the L*1* cache
09:31:23 <nakilon> codegolf.stackexchange.com is such a toxic place
09:31:48 <ais523> the reason the L2 cache needs to get involved in the first place is that kernel overhead is too high if you're just sending an L1-cache's worth of data with every system call
09:32:28 <nakilon> they were copying the puzzles from another website that had a rule "don't solve it publicly/collectively, don't share solutions" -- I mentioned it and they just started insulting me for that
09:32:33 <ais523> nakilon: how so? I really dislike Stack Exchange itself (the general attitude of the admin, and the software running it), but I've never had a problem with CGCC's community as people even if I disagree with many of their rules
09:32:53 <ais523> kspalaiologos: the trick is to make sure that you don't touch the page containing the written data until after it's already been read, otherwise you trigger a copy-on-write page fault
09:33:02 <nakilon> then I posted my 99 bottles solution that was the shortest among them but I censored it partially -- they deleted my submission, lol
09:33:21 <nakilon> kspalaiologos that was like 6 years ago
09:33:42 <kspalaiologos> nakilon, i can assure you that the community has changed a lot during these 6 years, most of the old regulars are gone now and the community is fairly fresh
09:33:57 <ais523> also, vmsplice is faster the simpler the page table is (it does a pagewalk), which means you ideally want to be using hugepages, and that makes it even harder to not touch the page containing the written data
09:34:00 <kspalaiologos> which is a bit of a shame, because i liked some of the old regulars, but i guess that it is how it is.
09:34:04 <nakilon> the stage when it's a phase of discussing the rules of the site, but they didn't want to discuss, they wanted to have fun only, kids
09:34:17 <ais523> Stack Exchange admin drove most of the better golfers away from the site, I think
09:34:57 <ais523> a partially censored solution would, I think, get deleted because it isn't a correct soultion
09:35:14 <ais523> problems copied from elsewhere would normally be copyright violations, though, and should be deleted because of that
09:35:21 <nakilon> but the correct solutions didn't have to exist there in the first place
09:35:31 <nakilon> because the website where they took the puzzle from is prohibiting it
09:35:46 <kspalaiologos> you shouldn't have posted anything then, and flagged the post to a moderator.
09:36:01 <nakilon> that's what I did and I got insults for that
09:36:23 <nakilon> on SO I even get bans for flagging
09:36:24 <kspalaiologos> well, this drama is 6 year old now, scratching the wound won't help probably
09:36:33 <ais523> 6 years ago is 2015, that would predate most of the current mods and most of the mods before then
09:36:35 <kspalaiologos> well yes, if you flag content that doesn't break the rules.
09:36:45 <nakilon> yep, that's why I don't visit that toxic place for years
09:36:49 <ais523> actually that might predate the first mod election?
09:37:09 <kspalaiologos> i respect your stance, but it's nice to give people another go
09:37:09 <nakilon> codegolf on SE has died for me when it was prohibited on SO
09:37:26 <kspalaiologos> codegolf is mildly recreational so it doesn't fit SO all that well
09:37:31 <ais523> anyway, I think a golf website could only reasonably have the rules of "please don't discuss these problems/solutions elsewhere" if the problems are fairly unique/unusual
09:37:46 <ais523> otherwise they'd be likely to be recreated by chance
09:37:49 <ais523> I don't know whether or not that was the case here
09:38:20 <ais523> kspalaiologos: fwiw, the SE *software* is so bad for a codegolf site that it's a good reason to not go there even if/when there's nothing wrong with the community
09:38:47 <ais523> I'm protesting against it by intentionally avoiding reputation, locking my account to 11 reputation permanently
09:39:01 <ais523> because the incentives for reputation go against what good golfing is
09:39:44 <nakilon> did they even realise finally that half of the problems they are posting isn't codegolf? I mean those problems that have the shortest solution as "for (;;) { if test( rand() ) break }"
09:40:14 <ais523> that would seem unlikely to be the shortest solution to something
09:40:18 <ais523> do you have an example?
09:40:57 <int-e> well, sometimes people count tokens
09:41:01 <ais523> the tersest solution to a problem is often very inefficient because "generate all possible answers then test to see if the answer is correct" is often shorter to express than actually trying to solve the problem efficiently
09:41:19 <ais523> but that's not quite the same
09:41:19 <kspalaiologos> ais523, how do you achieve locking yourself at 11rep
09:41:25 <nakilon> was years ago too; the point is that the problem does not fit the codegolf if it can be solved in just an hour of generating random solutions in a loop
09:41:31 <ais523> kspalaiologos: never post questions + community-wiki every answer
09:42:02 <ais523> int-e: there's a whole genre of esolangs that boils relevant keywords down to single bytes, that's pretty much replaced token-counting
09:42:24 <ais523> the hard part is trying to get the number of bytes for the plumbing down, too, things like parentheses and variable names can be very costly at that sort of scale
09:42:26 <kspalaiologos> nakilon, that's because of the golfing languages which i (and lots of other people) dislike
09:42:59 <ais523> I actually think golfing languages tell you more about programming and communication than regular languages
09:43:06 <nakilon> yeah that thing was making me leave too
09:43:32 <ais523> a) they tend to have relevant concepts as built-ins in the languages rather than needing you to write them out every time or use a library, meaning there's less busywork/copy-and-paste
09:43:47 <esowiki> [[User:Palaiologos]] https://esolangs.org/w/index.php?diff=81568&oldid=79799 * Palaiologos * (-86) deadlink removal
09:43:49 <ais523> b) they try to strip communication down to its essentials, trying to capture the essence of just what it means to describe a problem
09:43:58 <nakilon> perl, python, ruby, haskell, J, K -- it wasn't enough for them, they made the stupid golfscript, that is kind of the same as HQ9++
09:44:17 <ais523> nah, HQ9++ is generally disliked, and golfscript is actually very verbose as golfing languages go
09:44:35 <nakilon> golfscript didn't even have a standard
09:44:47 <nakilon> some guy just made it up and was editing it every day, lol
09:45:01 <ais523> kspalaiologos: your opinion is close to mine on this
09:45:20 <kspalaiologos> i even myself wanted to make an ultimate golflang which in the end failed because it required too much effort :P
09:45:48 <kspalaiologos> like, coming up with the right set of builtins and syntax is hard, because for example your usual stack-based code probably isn't _the_ smallest as it gets
09:45:48 <ais523> although, the way I see it, sometimes the interest is in developing the tersest algorithm, and seeing how hard it is to express the algorithm in a golfing language is a good way to define how complex the algorithm is
09:46:08 <ais523> kspalaiologos: I've actually been working on one in parallel with this conversation
09:46:21 <kspalaiologos> but some people on CGCC literally admitted to run a bruteforcing tool to find solutions to problems.
09:46:35 <ais523> I've done that too, but only in very low-level esolangs
09:46:54 <ais523> I have a brute-forcer Brain-Flak
09:46:56 <nakilon> oh, when I was saying about finding the solution in a loop I didn't mean they generate the code in a loop
09:46:57 <kspalaiologos> i did an elaborate bruteforce for Malbolge, and probably nothing else.
09:47:12 <ais523> that solves problems like "given x above y on the stack, find code that replaces them with y above x+y"
09:47:14 <nakilon> the code they submit is a loop that generates the STDOUT solutions to STDIN
09:47:42 <ais523> nakilon: there used to be a lot of cheating by not actually answering the problem in the past, but that's mostly been banned nowadays
09:47:43 <kspalaiologos> well, codegolf is really specific and somewhat unfair, yeah, that's why each language has it's own leaderboard
09:47:50 <ais523> the main cheating that happens nowadays is in I/O conventions
09:48:11 -!- LKoen has quit (Quit: “It’s only logical. First you learn to talk, then you learn to think. Too bad it’s not the other way round.”).
09:48:28 <int-e> I'd imagine brute force to be fair game.
09:49:04 <int-e> if a whole challenge can be solved by brute force alone then it's too simple.
09:49:13 <ais523> I actually think that the fact that brute force algos (i.e. brute force at runtime, not compile time) are often terser than more efficient algorithms is an indictment of compilers being bad at optimizing
09:49:29 <nakilon> no, you can just make a universal "solver" for every codegolf in the world in this way
09:49:35 <int-e> oh, that kind of brute force
09:49:36 <ais523> very few challenges are simple enough to be solved by brute-forcing all posisble programs until you find one that works, even intelligently
09:50:08 <nakilon> for (;;) { string = rand_string(); break if md5(string) == "83yn8237hr8374hr" }
09:50:16 <int-e> I once submitted an O(2^n) time sorting algorithm (O(2^n/n) average time if my analysis back then was correct) to a golfing competition and got away with it.
09:50:17 <nakilon> just change the hash for every puzzle
09:50:17 <ais523> nakilon: that doesn't work
09:50:40 <ais523> if your program is longer than 16 bytes or so, the brute-forcer will almost certainly find a second preimage before it finds the actual program you wanted
09:50:49 <nakilon> ais523 that's how they solve 99-bottles-like puzzles
09:50:52 <esowiki> [[Talk:Zirconium]] https://esolangs.org/w/index.php?diff=81569&oldid=81513 * Strohtaler * (+291) /* Example: Fibonacci stopping after number count */ new section
09:51:02 <ais523> nakilon: that *doesn't work* in that it will *produce the wrong answer*
09:51:45 <ais523> hash collisions are a thing, e.g. SHA-256 has a 1 in 2²⁵⁶ chance of any random string happening to collide with the hash you give
09:51:52 <nakilon> ais523 do you mean the md5 collision? that's not the point, that's the principle -- make a regex instead of md5, whatever
09:51:52 <int-e> . o O ( head -n NNN /dev/urandom )
09:52:07 <ais523> and the actual desired output will take way more than 2²⁵⁶ trials on average to find
09:52:23 <ais523> so, it will almost certainly find a second preimage before it finds the answer you want
09:52:29 <nakilon> and they were submitting the solutions that work for a week to find it
09:52:34 <int-e> Somehow I'm reminded of http://golf.shinh.org/setpid.html
09:52:44 <ais523> if someone submitted that as an answer to the 99bob problem then it would get deleted due to being incorrect
09:52:57 <ais523> (the length of time it takes to produce the wrong answer is way more than weeks, but irrelevant)
09:53:30 <nakilon> but they didn't delete, they insulted me for that
09:53:45 <ais523> I doubt that would happen nowadays, with people having a better understanding of golfing
09:54:08 <ais523> int-e: anagol eventually found a better fix for that, the recheck button
09:54:23 <ais523> which runs the program again and disqualifies the submission if it produces the wrong answer
09:54:38 <ais523> so, programs that only sometimes print the correct answer are no longer allowed
09:54:54 <nakilon> I don't believe in places becoming better while the number of people in it increases, it's always the opposite
09:54:57 <ais523> anagol isn't the only golfing site that was susceptible to this, incidentally, similar tricks have been used elsewhere
09:55:00 <int-e> though a bit sad in the case of language changes
09:55:09 <ais523> nakilon: but the number of people has decreased on CGCC
09:55:21 <int-e> (does it keep a record of the disqualified solutions?)
09:55:27 <ais523> I'm not sure of the details
09:56:04 <int-e> I mean it's one thing to have a program produce random output
09:58:23 <int-e> and another to have a program that relied on an older version of the language in question (I'm aware of this for Haskell, which removed n+k-patterns...)
09:59:09 <ais523> on CGCC, programs that only work in old interpreters are generally considered to be in a different language from programs that work in newer interpreters too
09:59:45 <nakilon> all my currently existing 4 answers are Jan 28 '11
10:00:11 <ais523> I know I've submitted a CGCC answer which only worked in one specific old version of Java (IIRC 1.4)
10:00:23 <ais523> nakilon: oh, the site was a complete joke in 2011, I can see why you were driven away
10:04:29 <esowiki> [[Talk:Zirconium]] M https://esolangs.org/w/index.php?diff=81570&oldid=81569 * Strohtaler * (+250) /* Example: Fibonacci stopping after number count */
10:11:48 <myname> why did haskell remove n+k patterns?
10:13:53 <int-e> well they were odd, potentially confusing... they needed both arithmetic and a comparison to be implemented
10:14:17 <int-e> and in contrast to the plain numeric patterns, they were not all that common or useful.
10:15:40 <int-e> but all in all... no strong reasons, just an opinionated decision by the people writing the Haskell 2010 standard document
10:16:01 <int-e> ghc still supports them as a language extension anyway
10:16:17 <ais523> n+k seems more like an agda feature than a haskell feature
10:16:33 <kspalaiologos> when i was learning haskell i was wondering why n+k isn't a thing too
10:16:46 <ais523> because agda is implemented as-if it uses peano arithmetic and n+k patterns are genuine pattern matches there
10:17:17 <lambdabot> (!!3)<$>transpose[show$foldr(\k a->2*10^2^n+a*k`div`(2*k+1))0[1..2^n]|n<-[0..]]
10:17:21 <lambdabot> [show(sum$scanl div(100^n)[1..[4..]!!n])!!n|n<-[0..]]
10:17:28 <shachaf> Hmm, pi_10 is longer than I remember.
10:17:31 <int-e> that said, having both a comparison (the 'n' has to be nonnegative) and arithmetic in one place is what made them occasionally useful for golfing.
10:18:20 <shachaf> There was a slightly shorter variant of one of those that took way longer to evaluate than lambdabot allows.
10:18:53 <int-e> but adding {-#LANGUAGE NPlusKPatterns#-} will generally not pay off :-)
10:20:06 <shachaf> How many "golfing" competitions are there where the program size is fixed, and the goal is to find the best solution within that size?
10:20:26 <shachaf> I only know of one that was run that way, and it was barely code golf, really.
10:20:43 <shachaf> Best defined by the problem.
10:21:08 <shachaf> I'm thinking of http://djm.cc/bignum-results.txt
10:21:29 <shachaf> Though I guess the C++ error explosion competition was sort of like this too in a sense.
10:22:04 <ais523> shachaf: you see them on CGCC occasionally, normally in challenges which are directly or indirectly bignum-related
10:22:19 <nakilon> shachaf PI is always longer than you remember ..D
10:23:16 <nakilon> shachaf also there are contests for finding the best solution with fixed time limit
10:26:25 <nakilon> this https://www.codingame.com/multiplayer/optimization
10:28:20 <nakilon> for example, here your Mars lander has to use as low amount of fuel as possible https://www.codingame.com/multiplayer/optimization/mars-lander
10:28:36 <nakilon> also engines will stop when depleted
10:34:09 <esowiki> [[Talk:Zirconium]] M https://esolangs.org/w/index.php?diff=81571&oldid=81570 * Strohtaler * (+9) /* Question / Understanding */
10:42:45 -!- LKoen has joined.
10:52:57 -!- arseniiv has joined.
11:35:45 <fizzie> I think I saw a "speed golf" thing (find the 'fastest' way of doing something) somewhere too, but can't recall the details, or where.
11:41:00 <Taneb> You'd need a well-defined time measure
11:43:43 <nakilon> actually on codingame you have to use some C anyway to get to the very top of A.I. contests
11:44:12 <nakilon> they traverse deep into the tree
11:45:36 -!- Lord_of_Life has joined.
11:46:22 <fizzie> We had a (playful) competition of sorts for a university course once that I was helping to run, initially with Scheme (using SICP's metacircular evaluator as the time measurement) but later on the JVM (with a bit fuzzier process-time-based time limits), on who writes the best "AI" for a specific (non-standard) board game.
11:47:07 <Taneb> We had one of those at uni, too, in Python
11:47:23 <Taneb> It was a variant of battleships with pieces which weren't straight lines
11:48:18 -!- Lord_of_Life has quit (Remote host closed the connection).
11:48:19 <nakilon> I sometimes think about such thing when imagining an AI contest -- it should be some unique language with your own VM that sums the weights of the commands executed
11:48:20 <fizzie> This was vaguely chess-like, but with an interesting two-step non-local capture mechanism (you "target" a piece first, and then capture on the next turn if the criteria for capturing are still fulfilled).
11:49:05 <fizzie> I did the same sort of results visualizations (and a web game viewer) that we've done with bf joust, I wonder if those pages are still up somewhere.
11:49:20 -!- Lord_of_Life has joined.
11:49:32 <fizzie> (Probably not, it's been quite long. Then again, university people aren't that keen on cleaning up web servers.)
11:50:26 -!- Lord_of_Life has quit (Remote host closed the connection).
11:52:13 -!- Lord_of_Life has joined.
11:52:21 <fizzie> Hah, the university wiki page that explains how to use the framework is at least still up. But it doesn't have a link to the results pages. https://wiki.aalto.fi/display/T934400/Guide
11:52:31 <nakilon> these things we talk about now are by magnitudes more complex than what all my coding friends (not even mentioning offline friends) could even understand
11:52:31 <fizzie> Also, the link to the game's page is just a 404 now.
11:53:50 <fizzie> Also, there's spam in the comments, as with any unmaintained wiki with a commenting facility. :)
12:01:06 <fizzie> Hah! My personal user account web directory's still around. That's ridiculous, but maybe shouldn't be surprising.
12:01:09 <nakilon> wow, Atlassian instance with free registration?
12:01:10 -!- LKoen has quit (Remote host closed the connection).
12:02:27 <fizzie> Yeah, no idea... I imagine they needed separate accounts for some use cases (non-student collaborators or whatnot), but a bit surprising that they didn't just create accounts separately for that.
12:03:05 <fizzie> (Maybe it actually requires manual review for account creation, otherwise you'd think there'd be more spam.)
12:05:28 -!- LKoen has joined.
12:06:06 <fizzie> I think my favourite visualization was the "time used per game" thing, because you got really different graphs depending on how people implemented their time bookkeeping. Like, http://users.ics.aalto.fi/htkallas/ai2012/img/cpu_justice.png (the 2012 winner) is probably some sort of a "use a specific fraction of remaining time per move" kind of thing, while
12:06:11 <fizzie> http://users.ics.aalto.fi/htkallas/ai2012/img/cpu_angrymarmots.png is something more linear/situational.
12:08:44 <fizzie> Even the JS-based game viewer still works. That's making me smile. Gotta be thankful for that backwards-compatibility and "dumb" web servers vs. "smart" clients, I guess. http://users.ics.aalto.fi/htkallas/ai2012/game.xhtml?justice-phailbot
12:09:02 <fizzie> Okay, the XHTML 1.1 doctype's kind of obsolete by now.
12:14:42 -!- LKoen has quit (Remote host closed the connection).
12:26:30 -!- clog has joined.
12:46:39 <b_jonas> "<ais523> many shops here don't accept £50 notes, though, there's a perception that £50 notes are only used by criminals" => we don't have that perception, but shops sometimes refuse 10000 HUF or 20000 HUF notes because they don't have enough change for them. and the problem is, ATMs usually give you cash in the form of 10000 HUF and 20000 HUF notes, but post offices make it difficult to actually
12:46:45 <b_jonas> change these to smaller notes. so I'm quite affraid that within 10 years the whole fucking currency, including non-cache forms, will die in a cascading collapse where everyone will realize that 10000 HUF and 20000 HUF banknotes aren't actually worth 10000 HUF or 20000 HUF, so you can't reliably get money from your bank account as cash.
12:48:00 <b_jonas> "<myname> ais523: here, shops have to accept every note if it's within reasonable margin of purchase" => well yes, but that doesn't help. cash does work well if you're doing building construction or reconstruction work, but not so well for everyday grocery purchases
12:48:54 <myname> i mean, i don't use it, usually, but i did
12:49:25 <b_jonas> "<nakilon> […] the way you non-verbally behave makes a great difference in how more likely you'll be considered a criminal" => and the way you look too. and since I do look sort of like a criminal, I make efforts to conspiciously look like I'm not stealing when I'm in a supermarket or other large shop
12:50:52 <nakilon> "oh look, that suspicious dude again" -- "nah, he's just from esolang"
12:51:22 <b_jonas> "<nakilon> wtf is altgr?" => either the right alt key, or control plus the left alt key, used as a third layer besides unshifted and shifted for typing characters. you need to know about both right-alt and control+left-alt, because sometimes programs will preempt using one or the other for some shortcut so you can only type something with the other one.
12:52:10 <nakilon> oh, it's the end of March, exactly the year of growing my beard
12:53:43 <b_jonas> "<nakilon> he must have big hands" => I do have big hands, yes. useful for a rubik's cube.
12:54:05 <myname> i would actually _not_ use cash for stuff like building constructions and use it for groceries
12:54:56 <nakilon> this epidemic is a gold for datamining
12:55:00 <b_jonas> "<myname> how so? ctrl is lower than shift and you have to reach the other key" => I press control with the left edge of my left palm. this is an idiosyncratic thing that almost nobody else does, and it restricts my choice of keyboards significantly, because on most cheap keyboards the left control key is shaped in a way that this doesn't work. it does work well on classical keyboards similar to the
12:55:06 <b_jonas> model M and more expensive keyboards though.
12:55:17 <b_jonas> I don't recommend learning this by the way, just press control with your pinky like a normal person.
12:55:23 <b_jonas> but it does work well for me.
12:55:28 <b_jonas> it may or may not require large hands.
12:56:06 <myname> pressing ctrl with the edge of your hand sounds like pressing ctrl+a very uncomfortable
12:56:18 <b_jonas> "<ais523> shift-shift forces me to move them much further apart than I'd normally have them, making it slower to type" => we used to use shift+shift to switch between Hungarian and English keyboard layouts, before windows put it onto left-alt+shift or windows+space by default
12:57:38 <b_jonas> "<ais523> e.g. in the word "type", I type the "y" with my left hand because the right hand is covering "p" which is at the far end of the keyboard, that means my left hand is nowhere near left shift" => I don't understand this. how is left shift even relevant for the word "type"? if you mean "y" instead of left shift, I still don't get it.
12:58:43 <myname> if you type with classic 10 finger system (which i hate) on qwerty, the y is supposed to be pressed with your right index finger
12:58:56 <b_jonas> I almost never use the wrong hand for letters when typing text, but I do often use the wrong shift key (left shift when right shift would be easier), and sometimes use the wrong hand when pressing individual keys rather than contiguous text
12:59:58 <b_jonas> "<nakilon> also on macOS I don't really close files -- I keep them all opened forever" => that doesn't really solve how you open them again though, or how you find the already opened instance
13:00:13 <b_jonas> even if you don't close them, you want an open command for finding the already open instance
13:02:57 <b_jonas> "hmm… I wonder whether nontemporal reads would make a significant improvement for that sort of workflow?" => I doubt it. nontemporal works for random access, but not for sequential access. you don't have large enough registers (before avx512) to load an entire cache line, so to read anything sequentially you'll need to read the same cache line twice or four times, and nontemporal makes that worse.
13:04:10 <b_jonas> "<ais523> I've been spending weeks looking at low-level processor performance in order to write the world's fastest FizzBuzz" => I feel like you have to optimize displaying the output for that. presumably with actual VGA text mode, rather than a terminal emulator.
13:05:25 <nakilon> "how you find the already opened instance" -- fuzzy file path search feature that is probably now in all IDEs
13:05:43 <b_jonas> "<kspalaiologos> and send the fizzbuzz thingy via serial" => that's even worse, serial terminal speed is the bottleneck then, isn't it? or maybe these USB3 serial port drivers make it faster? hmm
13:08:45 <b_jonas> "<nakilon> did they even realise finally that half of the problems they are posting isn't codegolf?" => how is that a problem? it started as a "programming puzzles and code golf" site, then got renamed to "code golf" when people mostly posted golf, but you can still post non-golf problems
13:09:50 <nakilon> no matter if there was "programming puzzles" in the name they always ranked answers by char length
13:10:29 <b_jonas> "<ais523> community-wiki every answer" => there's a backdoor where you can turn a question to CW if you edit it like sixteen times, waiting between so the versions aren't coalesced. does that work for answers too, and can you do it if you have reputation 1?
13:12:18 <b_jonas> "<kspalaiologos> but some people on CGCC literally admitted to run a bruteforcing tool to find solutions to problems." => as they should, if problems are so simple that that works. the solution to that is to post more interesting problems.
13:14:16 <nakilon> "if problems are so simple that that works. the solution to that is to post more interesting problems." -- there will be no "solution" because no one was interested in "solving" it, because they all considered it to be valid codegolfing -- you can't "solve" it by trying to teach a thousand of users by "showing better example of problems" because
13:14:16 <nakilon> they'll just ignore it -- this is what the voting system is now for on SE
13:14:41 <nakilon> it's a community driven censorship of those who have something to say instead of just repeating each other
13:15:02 <b_jonas> "<ais523> so, programs that only sometimes print the correct answer are no longer allowed" => ah, so $$ in perl is no longer as useful
13:16:29 <b_jonas> "<ais523> I know I've submitted a CGCC answer which only worked in one specific old version of Java (IIRC 1.4)" => I have an old post about that phenomenon at https://www.perlmonks.com/?node_id=866594 .
13:18:52 <b_jonas> "<shachaf> How many "golfing" competitions are there where the program size is fixed, and the goal is to find the best solution within that size?" => there was one old competition for writing a rubik's cube solver where your score was the product of (your program size) with (your program runtime) with (number of face turns in the solution found) or something like that.
13:19:20 <b_jonas> and there are demoscene competitions limited to 4 kilobytes and, um, 64 kilobytes? or something that fit
13:19:27 <b_jonas> but yes, they don't count as golf
13:21:41 -!- callforjudgement has joined.
13:21:42 <b_jonas> "<myname> why not?" => because the grocery purchases are of lower value
13:22:07 <callforjudgement> <b_jonas> I don't understand this. ← talking about how shift-shift is a bad shortcut because normally at least one of my hands isn't near the corresponding shift
13:22:23 -!- callforjudgement has quit (Client Quit).
13:22:25 <myname> b_jonas: ... so it is actually easier to check if the amount given is enough to pay
13:23:06 -!- ais523 has quit (Ping timeout: 240 seconds).
13:23:22 <b_jonas> "<myname> i would actually _not_ use cash for stuff like building constructions" => you don't really get a choice. basically all the people working for individual construction projects (i.e. reconstructing your home house, as opposed to big projects that the government commissions) will ask for cash. you can still use cash for groceries though, and often have to at smaller vendors, because only larger
13:23:54 <b_jonas> "<myname> pressing ctrl with the edge of your hand sounds like pressing ctrl+a very uncomfortable" => not really. it's not the easiest control key, but it's not too hard either.
13:24:08 <b_jonas> control+b is actually harder
13:24:45 <b_jonas> "<myname> if you type with classic 10 finger system (which i hate) on qwerty, the y is supposed to be pressed with your right index finger" => yes, and I don't understand ais's rationale for presssing it with the left hand
13:25:06 <b_jonas> I don't find it strange that he presses it with his left hand, that's quite reasonable, but I don't understand his specific argument
13:26:32 <b_jonas> callforjudgement: yes, but you don't need to switch between Hungarian and English layout often, it's a rare action, so it's fine to use a not too good shortcut. but yes, it's a bad shortcut.
13:28:18 <b_jonas> also cash vs card reminds me. in the US some people still use cheques. I only ever received a cheque once, as a prize for a golfing competition organized by an american company. I never wrote a cheque, it's not a thing in Europe, not anymore at least.
13:29:46 <b_jonas> we just use cash, debit cards, credit cards, various cards and paper slips with restricted usage (e.g. food only) as tax saving measures, and the occasional shop-specific gift card
13:32:39 <fizzie> I used to have a shortcut for switching between the UK and FI layouts, but for some unexplicable reason gave that up, and now write the two accented characters you need in Finnish (ä, ö) using the dead-key accents the UK layout has (altgr-[, where [ is the key immediately to the right of p, produces a dead diaeresis).
13:32:49 <esowiki> [[Baba Is You]] M https://esolangs.org/w/index.php?diff=81572&oldid=79953 * Bo Tie * (+500) Baba is not confirmed to be a rabbit. Also fixed capitalization inconsistency and added more words
13:33:29 <b_jonas> fizzie: how do you type their capital versions Ä and Ö?
13:33:40 <b_jonas> or do you not need those for finnish? I am hazy on how finnish works
13:33:49 <fizzie> It's just altgr-[ follwed by shift-A.
13:34:19 <b_jonas> not single altgr combos. I see
13:34:38 <b_jonas> do you type a lot of finnish with accents?
13:35:09 <fizzie> Any nontrivial amount of Finnish will require ä or ö, if that's what you mean. But I don't write that much in Finnish in the first place.
13:35:32 <b_jonas> yes, but does it require the capitals?
13:36:37 <fizzie> Oh. Yes, sometimes. Though I think those two are maybe slightly less common (but still present) as the initial letter of a word, and of course only so many words start a sentence (or are proper names).
13:36:49 <b_jonas> and that's why I gave up dvorak layout when I tried it by the way: it's based on the idea that you type vowels with your left hand and consonants with your right, but that doesn't work so well for Hungarian text that needs a lot of á and é and k and some z and y but very few h
13:37:07 <fizzie> Strictly speaking you also would sometimes need Å or å, which is a bit of a problem, because I don't think the ° is in the UK layout. But it's so rare, I haven't even learned what the compose sequence for that is.
13:37:55 <fizzie> There's five dead-key accents in the five symbol keys between the letters and the enter key in the UK layout, but they're äãáâà, no å.
13:38:13 <b_jonas> fizzie: isn't it on altgr with some number?
13:38:56 <b_jonas> though I guess it's easier to just learn the unicode code point number 0xE5
13:39:05 <fizzie> No, at least on this system. The number row with altgr produces non-dead symbols ¹²³€½¾{[]}, in order.
13:39:22 <fizzie> Actually, turns out it's just compose-o-a to get å.
13:39:27 <fizzie> That should be easy enough to remember.
13:40:30 <fizzie> Not sure why altgr-7..0 is {[]} when there's dedicated [] keys (with {} as their shifted variants) almost right next to them.
13:41:32 <fizzie> And compose-o-A gives Å. Fair enough. Next time I need to mention Åland Islands to someone, I'll use that instead of copying it from the "Swedish alphabet" wikipedia article.
13:43:49 <b_jonas> `python -cprint('\u00C5land Islands')
13:44:05 <b_jonas> `python3 -cprint('\u00C5land Islands')
13:48:39 -!- LKoen has joined.
13:49:32 <b_jonas> fungot, do you contain chemicals?
13:49:32 <fungot> b_jonas: today i am in cbe only. but have to seek. hee
13:49:43 <b_jonas> fungot, do you contain chemicals known to the state of California to cause cancer?
13:49:43 <fungot> b_jonas: wat.? abt the imp. at one? if dun go dunno... i oso saw a top dat i like but din say. there. emulsified.
13:49:50 <HackEso> ruby 2.5.5p157 (2019-03-15 revision 67260) [x86_64-linux-gnu]
13:50:32 <nakilon> his response sounds like he was caught consuming chemicals
13:53:02 <fizzie> fungot is known to the state of California to contain chemicals that may cause cancer.
13:53:02 <fungot> fizzie: hey me just drop into a cafe to get a cd? if poss. tc *cricinfo alerts are like coming. shuhui also like to attend to. just wish to pass me... haha
13:54:43 <fizzie> My terminal (urxvt) has this weirdo "keycap picture insertion mode", triggered by left-control + left-shift (as a tap, with no other keys), which makes it possible to e.g. press space to get ␣ or enter to get ↵.
13:55:23 <fizzie> Never had an actual use for that, only ever triggered it accidentally (thinking of pressing ctrl-shift-something but then decided not to).
13:56:12 <b_jonas> fizzie: there's a switch to disable it. I always do.
13:57:09 <fizzie> Guess it might be a semi-convenient way to type ←↑→↓ though. (I just use compose "->" for the horizontal arrows, but not sure how I'd type the vertical ones.)
13:58:15 <b_jonas> or, for a more complete command line, (export LC_CTYPE=hu_HU.utf8; urxvt -ls -geometry 80x25 +sb -b 0 -bg \#000000 -fg \#c0c0c0 --colorBD \#ffffff --color0 \#000000 --color1 \#aa0000 --color2 \#00aa00 --color3 \#aa5500 --color4 \#0000ff --color5 \#aa00aa --color6 \#00aaaa --color7 \#c0c0c0 --color8 \#555555 --color9 \#ff5555 --color10 \#55ff55 --color11 \#ffff55 --color12 \#5588ff --color13 \#ff55ff
13:58:21 <b_jonas> --color14 \#55ffff --color15 \#ffffff -cr \#c0f0a0 -fn "x:-*-fecupboard-medium-r-*-*-20-*-*-*-*-*-iso10646-1,x:-*-terminus-bold-r-*-*-20-*-*-*-*-*-iso10646-1" --boldFont "" -vb -sbg --print-pipe false --pointerBlank --pointerBlankDelay 2147483647 ++iso14755_52 "%@")
13:58:49 <fizzie> Aw, keycap picture insert mode doesn't insert the Microsoftⓡ Windowsⓡ™ logo when used with the Windows key.
13:59:00 <fizzie> I was thinking it'd sneakily take a picture of your keyboard and figure out what's printed on it.
13:59:50 <fizzie> At least the more standard keys (⎋⇥⇧⎈⇭) work with it.
13:59:52 <kspalaiologos> b_jonas, no copying the buffers, also you operate it directly if you boot
14:09:43 <APic> For ↑ and ↓ i have in my .XCompose: „<Multi_key> <b> <b> : "↑" U2191“ and „<Multi_key> <v> <v> : "↓" U2193“
14:12:19 <APic> (The two „b“ for pointing a_b_ove, and „v“ looks like ↓ 😉)
14:13:06 <APic> Yes, but „a“ is too far away from „v“ 😉
14:15:32 <b_jonas> wordperfect directions esdx then
14:26:15 <APic> And many First-Person-Shooters wads. But for Compose we already have ← and → as <- and ->
14:26:56 <APic> Of course we could use ^| and v| or something ☺
14:27:56 <APic> Ah, v| already makes ↓ here
14:28:39 <APic> And ^| makes ↑
14:28:41 <APic> Good to know ☺
14:36:59 -!- kspalaiologos has quit (Remote host closed the connection).
14:37:24 -!- kspalaiologos has joined.
14:46:23 -!- Sgeo has joined.
14:47:24 <Sgeo> I still have "Forget is for getting" burned in my brain. I think that's shachaf's pun? Ironically enough, I've forgotten everything else about lenses
14:58:10 <b_jonas> fungot, why do you consider Cruella de Vil your favorite Disney villain?
14:58:10 <fungot> b_jonas: poda wat is personal btwn ur fingers just press wrong one lor i thk his students
15:01:58 -!- sprock has joined.
15:11:45 -!- delta23 has joined.
15:19:00 -!- copumpkin has quit (Remote host closed the connection).
15:19:55 -!- copumpkin has joined.
15:21:25 <fizzie> Thanks for the v| ^| tip. I did half-heartedly try a few v- and ^-based ideas, but didn't think of combining them with |.
15:24:02 <fizzie> Next mystery: => gives ⇒, but what about the other directions? <= is just ≤, and there's no key for ‖ to compose v and ^ with.
15:29:30 <fizzie> Also not good: on this system, both compose +- and compose -+ gives ±, which is a missed opportunity; one of them should give ∓ instead.
15:33:31 -!- kspalaiologos has quit (Quit: Leaving).
15:40:46 <esowiki> [[Talk:Zirconium]] M https://esolangs.org/w/index.php?diff=81573&oldid=81571 * Strohtaler * (-1) /* Example: Fibonacci stopping after number count */
15:53:04 <b_jonas> fungot, they're called "private" bits, not "personal"
15:53:04 <fungot> b_jonas: she was gonna ask me to die every now and then, ithink i can ' t go... would love to get married one cannot intro
15:58:48 <esowiki> [[Esolang:Sandbox]] M https://esolangs.org/w/index.php?diff=81574&oldid=81005 * PythonshellDebugwindow * (+42) Add a test
16:07:20 <APic> fizzie: Same here
16:54:51 <esowiki> [[Deadfish]] https://esolangs.org/w/index.php?diff=81575&oldid=81548 * Not applicable * (+671) add (untested) Ti-Nspire Basic one
16:58:37 <shachaf> Sgeo: That's what Forget is for.
17:03:45 <esowiki> [[Functional()]] https://esolangs.org/w/index.php?diff=81576&oldid=78488 * Hakerh400 * (+824) Add TIO links
17:17:24 -!- arcsor5 has joined.
17:21:59 <esowiki> [[User talk:Truttle1]] https://esolangs.org/w/index.php?diff=81577&oldid=81556 * Truttle1 * (+110)
17:28:37 -!- mla has joined.
17:30:06 -!- owocean[m] has joined.
17:31:19 <esowiki> [[Functional()]] https://esolangs.org/w/index.php?diff=81578&oldid=81576 * Hakerh400 * (+1710) Add more TIO links
17:32:19 -!- owocean[m] has left ("User left").
17:32:55 <kmc> v| and ^| don't do anything for me
17:34:19 <fizzie> Hmm. I can do v| and |v and both result in ↓.
17:36:40 <kmc> I think I'm possibly using some GTK input method instead of the base X compose feature
17:37:07 <kmc> unlike previous machines where I had compose set up, this one shows the individual characters as I type them, underlined, and with a prompt before them that looks like an overlapping square and circle
17:37:20 <kmc> then it all disappears and is replaced with the target character when I complete a sequence
17:38:01 <kmc> on Tuesday I'm getting a new laptop! haven't decided what OS to run on it
17:38:06 <kmc> I might go back to Windows + WSL
17:38:18 <kmc> there is some Windows-only software I'd like to run
17:39:14 <kmc> but I will have 8 cores and 48 GB of RAM so running virtual machines is not a big deal
17:40:27 <kmc> not sure which one is better to use as the base OS
17:40:32 <kmc> windows might have better power management
17:46:07 <fizzie> I think I'm maybe getting some sort of a visual thing like that on my work laptop too.
17:47:02 <fizzie> Here's a thing I kind of would like WSL to have: now that it's a real kernel (AIUI, in WSL2), they could expose those drives that Windows doesn't know what to do with as block devices to the Linux thing, so that they could be mounted in there.
17:47:52 <fizzie> I do dual boot on this every once every few months (for games), and every time I'm over on that side, there's some file or whatever that's only accessible to the other operating system.
17:48:42 <fizzie> I had that sort of working on a VirtualBox thing at one point, but it was still a hassle.
17:49:36 <kmc> yeah it's possible to boot an actual hard drive partition in a VM
17:49:45 <kmc> I think I had that working once, back when I dual-booted
17:50:06 <kmc> but dual booting seems less and less appealing over time
17:51:11 <esowiki> [[Parse this sic]] https://esolangs.org/w/index.php?diff=81579&oldid=81250 * Digital Hunter * (+1342) /* Computational class */ Turing-complete!
17:51:46 <esowiki> [[User:Zzo38/Programming languages with unusual features]] https://esolangs.org/w/index.php?diff=81580&oldid=81559 * Zzo38 * (+2072) Free Hero Mesh
17:53:02 <fizzie> I turned on that thing in Steam that lets you apply Proton to any game, not only the ones they've verified, and from what I've read it works pretty well for many things. Maybe that's why I think I haven't booted the Windows installation this year yet.
17:53:31 <fizzie> One unappealing part of it is that when you don't use it in months, installing all the updates is kind of terrible.
17:57:19 <kmc> what's Proton? some WINE-like thing?
17:58:23 <fizzie> Yes, I think it's technically a fork of Wine that Valve maintains? Something like that.
17:58:40 <kmc> interesting
17:58:42 <kmc> worth looking into
17:59:26 <kmc> I don't play games very much
17:59:53 <kmc> but I might want Windows for 3D CAD software or weirdo ham radio software
18:02:11 -!- kritixilithos has joined.
18:06:30 <zzo38> Usually, if the software is Windows, I just use different software instead. DOSBOX can run 16-bit Windows programs (if you have Windows 3.1), although it is slow. I cannot install Wine on my computer due to conflicts in the package manager.
18:09:36 <zzo38> Many programs are already available for many operating systems, anyways (or there is another program which does mainly the same thing, for another operating system, e.g. a NES/Famicom emulator).
18:11:51 <kritixilithos> david madore's interview on esoteric.codes https://esoteric.codes/blog/david-madore , didn't realise it's so easy to have a machine that beats oracle_n TMs for all n
18:55:00 -!- kritixilithos has quit (Quit: quit).
19:07:37 <zzo38> I think many people they will write new free software to do the same things as some other software
19:26:09 -!- adu_ has joined.
19:37:02 <b_jonas> yay! I installed the new speakers for my home computer. it's the third try, the first two (cheaper) speakers that I bought weren't suitable.
19:37:47 -!- Lord_of_Life_ has joined.
19:39:58 -!- Lord_of_Life has quit (Ping timeout: 260 seconds).
19:39:58 -!- Lord_of_Life_ has changed nick to Lord_of_Life.
20:24:39 <esowiki> [[1066]] M https://esolangs.org/w/index.php?diff=81581&oldid=78471 * PythonshellDebugwindow * (-39) /* Commands */ grm
20:51:58 -!- arseniiv has quit (Ping timeout: 240 seconds).
20:54:46 -!- iovoid has quit (Ping timeout: 240 seconds).
20:55:16 -!- moony has quit (Ping timeout: 240 seconds).
20:58:30 -!- moony has joined.
20:58:51 -!- iovoid has joined.
21:01:49 -!- arcsor5 has quit (Quit: My MacBook has gone to sleep. ZZZzzz…).
21:10:37 <esowiki> [[OISC]] M https://esolangs.org/w/index.php?diff=81582&oldid=81158 * Kefalonia * (+0) *p
21:12:24 -!- arcsor5 has joined.
21:12:43 -!- adu_ has quit (Quit: adu_).
21:14:50 <esowiki> [[SBN]] M https://esolangs.org/w/index.php?diff=81583&oldid=81160 * Kefalonia * (+45) explain
21:15:00 <esowiki> [[Fargo]] M https://esolangs.org/w/index.php?diff=81584&oldid=79648 * PythonshellDebugwindow * (+760) I/O; Truth-machine
21:24:12 -!- adu_ has joined.
21:28:41 <fizzie> Rotterdam has an area that's literally called "Cool District"? That's a little presumptuous, isn't it. (Okay, okay, it's not the same "Cool".)
21:40:17 -!- adu_ has quit (Quit: adu_).
21:42:10 -!- adu_ has joined.
21:45:29 <esowiki> [[Esolang:Community portal]] https://esolangs.org/w/index.php?diff=81585&oldid=81567 * Heavpoot * (-36)
21:49:34 <esowiki> [[Special:Log/newusers]] create * Clsource * New user account
21:51:31 <esowiki> [[Esolang:Introduce yourself]] https://esolangs.org/w/index.php?diff=81586&oldid=81518 * Clsource * (+63) /* Introductions */
21:53:38 -!- Lord_of_Life has quit (Ping timeout: 240 seconds).
21:54:48 -!- adu_ has quit (Quit: adu_).
22:10:31 -!- LKoen has quit (Quit: “It’s only logical. First you learn to talk, then you learn to think. Too bad it’s not the other way round.”).
22:14:14 <HackEso> olist https://www.giantitp.com/comics/oots1230.html: shachaf oerjan Sgeo FireFly boily nortti b_jonas
22:15:18 <esowiki> [[Ogu]] N https://esolangs.org/w/index.php?oldid=81587 * Clsource * (+809) Added Og Language
22:19:25 <esowiki> [[Language list]] M https://esolangs.org/w/index.php?diff=81588&oldid=81505 * Clsource * (+10) Added Ogu
23:16:06 <esowiki> [[User talk:Truttle1]] https://esolangs.org/w/index.php?diff=81589&oldid=81577 * Heavpoot * (+116)
23:19:42 -!- arcsor5 has quit (Quit: My MacBook has gone to sleep. ZZZzzz…).
23:23:18 <esowiki> [[Uyjhmn n]] https://esolangs.org/w/index.php?diff=81590&oldid=74574 * Heavpoot * (+29) this *is* turing complete, right?
23:24:02 <esowiki> [[Uyjhmn n]] M https://esolangs.org/w/index.php?diff=81591&oldid=81590 * Heavpoot * (+0) uncapitalize the letter c in turing complete. oops.
23:46:39 -!- arcsor5 has joined.