←2019-05-27 2019-05-28 2019-05-29→ ↑2019 ↑all
00:10:10 -!- Phantom_Hoover has quit (Read error: Connection reset by peer).
00:33:38 <salpynx> orin: *gʷíh₃womos seems to be "we live"? 1st thematic pl. of gʷih₃wós , and the first word is .... something about stealing juice? He steals juice that we might live? Sounds Biblical.
00:57:39 -!- salpynx has quit (Quit: Page closed).
00:57:58 -!- arseniiv has quit (Ping timeout: 272 seconds).
01:31:14 -!- moei has joined.
01:52:11 -!- adu has joined.
01:52:25 -!- mniip has quit (Quit: This page is intentionally left blank.).
01:54:12 -!- imode has joined.
01:58:39 -!- mniip has joined.
02:03:42 -!- sprocklem has joined.
02:29:06 -!- Frater_EST has joined.
02:29:33 -!- Frater_EST has left.
03:53:37 <esowiki> [[Special:Log/newusers]] create * ZachAttrax * New user account
03:55:17 -!- FreeFull has quit.
04:01:27 <esowiki> [[Talk:Bitch]] M https://esolangs.org/w/index.php?diff=62801&oldid=62799 * A * (+7)
04:09:22 <esowiki> [[Bitch]] M https://esolangs.org/w/index.php?diff=62802&oldid=62797 * A * (+115) Another significant proof was ignored
04:18:42 <esowiki> [[User:A]] M https://esolangs.org/w/index.php?diff=62803&oldid=62784 * A * (+77)
05:27:33 <esowiki> [[Volatile]] M https://esolangs.org/w/index.php?diff=62804&oldid=62747 * JonoCode9374 * (+18) /* Infinite loop */
05:32:48 <esowiki> [[Volatile]] M https://esolangs.org/w/index.php?diff=62805&oldid=62804 * JonoCode9374 * (-3) /* (Another) Python 3 interpreter */ Fixed the bug where it would always come up saying 'empty stack'
05:46:11 -!- Frater_EST has joined.
05:47:16 -!- Frater_EST has left.
07:04:41 -!- Hooloovo0 has quit (Excess Flood).
07:05:58 -!- AnotherTest has joined.
07:11:19 -!- Hooloovo0 has joined.
07:23:36 -!- imode has quit (Ping timeout: 244 seconds).
08:35:16 <esowiki> [[Volatile]] https://esolangs.org/w/index.php?diff=62806&oldid=62805 * TuxCrafting * (+3) Undo revision 62805 by [[Special:Contributions/JonoCode9374|JonoCode9374]] ([[User talk:JonoCode9374|talk]]) dup'ing an empty stack is supposed to error. not a bug.
10:12:47 -!- Lord_of_Life has quit (Ping timeout: 248 seconds).
10:15:46 -!- Lord_of_Life has joined.
10:27:51 -!- S_Gautam has joined.
11:16:15 -!- Cale has quit (Ping timeout: 264 seconds).
11:29:12 -!- Cale has joined.
11:42:01 -!- wob_jonas has joined.
11:47:18 -!- copumpkin has joined.
11:51:45 -!- sebbu3 has joined.
11:55:57 -!- sebbu has quit (Ping timeout: 268 seconds).
11:59:01 <esowiki> [[Bitch]] M https://esolangs.org/w/index.php?diff=62807&oldid=62802 * Int-e * (+29) update links to new repo name
12:00:36 <esowiki> [[Talk:Bitch]] M https://esolangs.org/w/index.php?diff=62808&oldid=62801 * Int-e * (-2) update links
12:01:39 -!- Frater_EST has joined.
12:11:29 <int-e> Ah, Leela (Chess Zero) is now officially the one-eyed leading the blind... it convincingly beat Stockfish in the TCEC final with a +7 score in 100 games.
12:13:16 <int-e> (Though the zero is no longer quite accurate; as I understand it, Leela is incorporating tablebase endgame knowledge somehow. The keyword is "tablebase rescoring".)
12:14:27 -!- Frater_EST has quit (Ping timeout: 268 seconds).
12:21:47 -!- Frater_EST has joined.
12:22:13 -!- Frater_EST has quit (Read error: Connection reset by peer).
12:22:27 <wob_jonas> whoa, python's struct.pack puts endianness modifiers before the type marker in the pack template, whereas perl puts it after.
12:26:05 <esowiki> [[Language list]] https://esolangs.org/w/index.php?diff=62809&oldid=62726 * Unlimiter * (+12) /* P */
12:26:57 <esowiki> [[Language list]] https://esolangs.org/w/index.php?diff=62810&oldid=62809 * Unlimiter * (-12) /* P */
12:27:32 <esowiki> [[Special:Log/delete]] delete * Ais523 * deleted "[[User:A]]": using user page history to disrupt the wiki and attack people
12:27:37 -!- sebbu3 has changed nick to sebbu.
12:27:38 <esowiki> [[Language list]] https://esolangs.org/w/index.php?diff=62811&oldid=62810 * Unlimiter * (+12) /* P */
12:28:35 <esowiki> [[Special:Log/protect]] protect * Ais523 * protected "[[User:A [create=sysop] (indefinite)]]": this userpage was being used for disruption / as an attack page rather than for any beneficial purposes, thus salting
12:29:15 <esowiki> [[Special:Log/delete]] delete * Ais523 * deleted "[[User:A/asdfadsfasdfasdfasdfasdfasdfasdfadsfasdfasdfasdfasdfasdfasdfadsfasdfasdfasdfasdfasdfasdfadsfasdfasdfasdfasdfasdfasdfadsfasdfasdfasdfasdfasdfasdfadsfasdfasdfasdfasdfasdfasdfadsfasdfasdfasdfasdfasdfasdfadsfasdfasdfasdfasdfasdfasdfadsfasdfasdfasdfasdfasdfa]]": now being used primarily as an attack page
12:30:30 -!- Sgeo__ has quit (Read error: Connection reset by peer).
12:37:29 -!- ais523 has joined.
12:37:50 <esowiki> [[Point]] https://esolangs.org/w/index.php?diff=62812&oldid=62744 * Unlimiter * (-4)
12:38:27 <ais523> <arseniiv> should we leave revertion of that edit to salpynx’s page to ais523? ← a) don't revert edits just because of who makes them, b) feel free to revert edits because of their content or because you disagree with them, c) if someone is persistently acting in an obnoxious way (making attack pages, etc.) it will probably need admin involvement to stop them, but for things that don't need admin involvement you can fix them yourself
12:38:38 <esowiki> [[Point]] https://esolangs.org/w/index.php?diff=62813&oldid=62812 * Unlimiter * (+31)
12:39:59 -!- ais523 has quit (Remote host closed the connection).
12:40:13 -!- ais523 has joined.
12:41:14 <ais523> if User:A has been attacking pages in a widespread or systematic way, I can do a total or partial block, but short of that, any edits to pages created by other users should be treated on a case-by-case basis (the creator of a page doesn't have /control/ of it; but if you disagree with the meaning or style of someone else's edit, you can revert it regardless of who made the page, unless it causes a revert war)
12:41:54 <ais523> or, I guess the policy is "for small-scale problems, you can try to fix them yourself, if it becomes a large-scale problem or there's a disagreement with two people reverting back and forth, ask for admin help"
12:42:33 <esowiki> [[Bit**]] N https://esolangs.org/w/index.php?oldid=62814 * A * (+19) Redirect bit** to bitch to prevent people from offensing others
12:43:39 <esowiki> [[Bitch]] M https://esolangs.org/w/index.php?diff=62815&oldid=62807 * A * (+15) I don't want the redirect to be forgotten...
12:45:04 <esowiki> [[Bitch]] https://esolangs.org/w/index.php?diff=62816&oldid=62815 * Ais523 * (-15) Undo revision 62815 by [[Special:Contributions/A|A]] ([[User talk:A|talk]]): in the vast majority of circumstances, there is no point on linking a redirect back to the page you're currently on, and this is not an exception
12:46:59 -!- arseniiv has joined.
12:47:29 <esowiki> [[Bitch]] M https://esolangs.org/w/index.php?diff=62817&oldid=62816 * A * (+92) Okay. Clarify
12:58:51 -!- ais523 has quit (Ping timeout: 268 seconds).
12:59:09 <esowiki> [[Talk:Bitch]] https://esolangs.org/w/index.php?diff=62818&oldid=62808 * Int-e * (+1418) /* [Complete] Equivalency between bitch and Home Row by User:Helen */ revive a comment, and add a few more
12:59:45 <esowiki> [[Talk:Bitch]] M https://esolangs.org/w/index.php?diff=62819&oldid=62818 * Int-e * (+0) /* [Complete] Equivalency between bitch and Home Row by User:Helen */ (probably) correct indices
13:03:27 <esowiki> [[Volatile]] M https://esolangs.org/w/index.php?diff=62820&oldid=62806 * A * (-47) /* Infinite loop */
13:05:15 <wob_jonas> `python3 -cimport sys; print(sys.getrefcount(8))
13:05:16 <HackEso> 62
13:05:24 <wob_jonas> `python3 -cimport sys; print(sys.getrefcount(0))
13:05:25 <HackEso> 443
13:05:55 <wob_jonas> those numbers don't match what my python says. does that mean I have a fake python?
13:07:11 <int-e> "The count returned is generally one higher than you might expect" ;-)
13:07:35 <int-e> clearly by induction the function should return infinity.
13:08:42 <esowiki> [[Bitch]] M https://esolangs.org/w/index.php?diff=62821&oldid=62817 * A * (+7) /* Computational class */ : Uh huh, so that was currently invalid. I will uncomment it if [[User:Helen|@Helen]] successfully proved that.
13:09:02 <wob_jonas> admittedly it's a different version of python
13:09:19 <wob_jonas> `python3 -cimport sys; print(sys.version)
13:09:20 <HackEso> 3.5.3 (default, Sep 27 2018, 17:25:39) \ [GCC 6.3.0 20170516]
13:09:25 <esowiki> [[Bitch]] M https://esolangs.org/w/index.php?diff=62822&oldid=62821 * A * (+0) /* Computational class */ : Sorry, I covered a matching bracket
13:09:32 <int-e> but anyway, no. openat(AT_FDCWD, "/usr/lib/python3.7/__pycache__/sitecustomize.cpython-37.pyc", O_RDONLY|O_CLOEXEC) = 3
13:09:50 <int-e> so I bet it depends on the python packages that are installed.
13:10:47 <int-e> python2 '-cimport sys; print(sys.getrefcount(0))' => 394; python3 '-cimport sys; print(sys.getrefcount(0))' => 246 is what I have here atm.
13:13:40 <esowiki> [[Talk:Bitch]] https://esolangs.org/w/index.php?diff=62823&oldid=62819 * A * (+7) /* Memory equivalence */ : Wow, it is really cool putting the signature into code snippets! (I will remove that though)
13:21:56 <esowiki> [[Talk:Bitch]] M https://esolangs.org/w/index.php?diff=62824&oldid=62823 * Int-e * (+0) /* Memory equivalence */ move </code> to intended place (sorry for missing that in the first place!)
13:25:07 <esowiki> [[Talk:Bitch]] M https://esolangs.org/w/index.php?diff=62825&oldid=62824 * A * (+212) /* [Complete] Equivalency between bitch and Home Row by User:Helen */
13:27:00 <esowiki> [[Talk:Bitch]] M https://esolangs.org/w/index.php?diff=62826&oldid=62825 * A * (-46) /* [Complete] Equivalency between bitch and Home Row by User:Helen */
13:27:56 <esowiki> [[Talk:Bitch]] M https://esolangs.org/w/index.php?diff=62827&oldid=62826 * A * (-167) Undo revision 62826 by [[Special:Contributions/A|A]] ([[User talk:A|talk]]): No, that was wrong!
13:41:15 <john_metcalf> That Bytemark link at the bottle of the esolang wiki leads to a dead page
13:43:27 <int-e> bottle :)
13:44:19 <int-e> Hmm, fizzie's Bytemark's esolang contact?
14:11:26 <arseniiv> I had an impression N ⋊_ϕ H ≅ (N ⋊_id im ϕ) × ker ϕ (group theory) but now I think it’s wrong
14:11:36 -!- imode has joined.
14:19:06 <esowiki> [[User talk:A]] M https://esolangs.org/w/index.php?diff=62828&oldid=62783 * A * (+1727)
14:19:19 <esowiki> [[User talk:A]] M https://esolangs.org/w/index.php?diff=62829&oldid=62828 * A * (+17)
14:19:31 <esowiki> [[User talk:A]] M https://esolangs.org/w/index.php?diff=62830&oldid=62829 * A * (-1744) Blanked the page
14:21:54 <arseniiv> tada
14:26:04 <esowiki> [[Special:Log/move]] move * A * moved [[User talk:A/asdfadsfasdfasdfasdfasdfasdfasdfadsfasdfasdfasdfasdfasdfasdfadsfasdfasdfasdfasdfasdfasdfadsfasdfasdfasdfasdfasdfasdfadsfasdfasdfasdfasdfasdfasdfadsfasdfasdfasdfasdfasdfasdfadsfasdfasdfasdfasdfasdfasdfadsfasdfasdfasdfasdfasdfasdfadsfasdfasdfasdfasdfasdfa]] to [[User:A/asdfadsfasdfasdfasdfasdfasdfasdfadsfasdfa
14:46:22 <wob_jonas> `? ninja
14:46:24 <HackEso> ninja? ¯\(°​_o)/¯
14:46:25 <wob_jonas> `? samurai
14:46:26 <HackEso> samurai? ¯\(°​_o)/¯
14:46:32 <wob_jonas> `? pirate
14:46:33 <HackEso> Pirates are humourously nautical persons. Their grammar is friendly and plural.
14:53:16 -!- ais523 has joined.
14:56:49 -!- ais523 has quit (Client Quit).
14:57:02 -!- ais523 has joined.
15:08:25 -!- LKoen has joined.
15:34:37 <esowiki> [[Special:Log/move]] move * Ais523 * moved [[User:A/asdfadsfasdfasdfasdfasdfasdfasdfadsfasdfasdfasdfasdfasdfasdfadsfasdfasdfasdfasdfasdfasdfadsfasdfasdfasdfasdfasdfasdfadsfasdfasdfasdfasdfasdfasdfadsfasdfasdfasdfasdfasdfasdfadsfasdfasdfasdfasdfasdfasdfadsfasdfasdfasdfasdfasdfasdfadsfasdfasdfasdfasdfasdfa]] to [[User talk:A]]: user talk pages shouldn't be move
15:34:39 -!- esowiki has quit (Remote host closed the connection).
15:35:40 <ais523> err, esowiki, are you OK?
15:37:01 <int-e> this clearly is too much drama for the bots.
15:37:24 <ais523> it's not even drama really, that'd require multiple people on each side of the argument
15:37:48 <ais523> it's just someone being disruptive and me gradually turning their permissions lower and lower in response to them demonstrating that they can't be trusted to use them correctly
15:38:06 <int-e> . o O ( Shakespeare made a lot of drama all by himself. )
15:39:10 <shachaf> Now they overflowed the bot's buffer.
15:40:18 <ais523> it's annoying because I normally use the bot to monitor esowiki's recent changes (often via the logs), now I have to check them directly
15:40:56 <wob_jonas> like at https://esolangs.org/wiki/Special:RecentChanges ?
15:41:06 <ais523> the funny thing is that I actually missed that A's user talk page had gotten renamed (as I was concentrating on the userpage rename), then A's further screwing about made it obvious what had happened and so I could fix it
15:41:07 <ais523> wob_jonas: yes
15:41:29 <ais523> being able to read #esoteric and the recent changes at the same time saves on pages to open, also the stalker-mode logs update automatically
15:42:15 <shachaf> Did you see the bit where they joined the channel, pasted criticisms of themselves from the logs into it, and then quit?
15:42:43 <shachaf> imo what's the deal
15:43:04 <ais523> they're trying to be passive-agressive but aren't very good at it
15:44:34 <shachaf> hm
15:44:37 <ais523> they could be a good user if they stopped trying to test the boundaries of what's allowed and stopped intentionally antagonising people…
15:44:51 <ais523> (and stopped jumping to conclusions about computational class0
15:44:53 <ais523> s/0/)/
15:45:02 <shachaf> if you're going to be a complete jerk you could at least be witty about it or something
15:45:24 <shachaf> that would be a fair deal
15:45:36 <int-e> . o O ( and we all know how that goes )
15:45:57 <shachaf> Do we?
15:52:07 <wob_jonas> ... and counts too
15:52:35 <wob_jonas> python's struct.unpack takes the count and endianness modifier before the type letter in the pattern
16:05:02 -!- LKoen has quit (Remote host closed the connection).
16:06:24 -!- LKoen has joined.
16:07:13 -!- S_Gautam has quit (Quit: Connection closed for inactivity).
16:08:01 -!- wob_jonas has quit (Remote host closed the connection).
16:48:54 -!- S_Gautam has joined.
16:49:42 -!- arseniiv_ has joined.
16:50:25 <fizzie> We're in HEXHAM. So exotic. One could even say, esoteric.
16:50:53 <fizzie> `? hexham
16:50:54 <HackEso> Hexham es la ciudad mas importante de programación esotérico.
16:52:08 -!- arseniiv has quit (Ping timeout: 245 seconds).
16:52:44 <shachaf> fizzie: help
16:52:47 <shachaf> what are you doing
16:53:18 <shachaf> you're breaking everything
16:53:36 <ais523> fizzie: esowiki the bot is down
16:54:05 <ais523> also the bytemark link in the footer is a 404
16:54:37 <ais523> also, did you go to hexham because #esoteric or did you somehow go there for an unrelated reason?
16:56:08 <fizzie> A bit from column A, a bit from column B. We're mostly driving around Northumberland, but staying a night in Hexham was #esoteric-motivated.
16:56:26 <fizzie> Not sure what's wrong with the bot though.
16:56:35 <ais523> OK, because if it were coincidence, I'm not sure I could deal with it :-D
16:56:59 <ais523> the bot may have had a buffer overflow? User:A created a page with a ridiculously long name and the bot crashed when I tried to clean up after it
16:58:09 <shachaf> Is fizzie the only person in this channel in Hexham?
16:58:42 <ais523> we used to have multiple channel regulars in Hexham, which was an absurd coincidence beause it has a population of ~12000
16:59:26 <fizzie> It's supposed to just cut long messages. But there could be an bug.
16:59:46 <rain1> i cant believe IOCCC is done
16:59:50 <rain1> it feels too soon
17:00:05 <rain1> has it really been a year since last IOCCC
17:00:37 <ais523> even taking the birthday paradox into account, you wouldn't a channel of 85 people to have two unrelated regulars living in the same small town
17:04:13 <fizzie> The bot had indeed died of SIGSEGV. Sounds buggy.
17:05:38 -!- LKoen has quit (Remote host closed the connection).
17:05:56 <fizzie> Trying to restart it makes it die of the same thing.
17:06:17 <fizzie> I don't think I'll get this fixed before dinner. Will have a look later.
17:08:44 -!- LKoen has joined.
17:13:26 -!- LKoen has quit (Remote host closed the connection).
17:41:57 -!- b_jonas has joined.
17:42:02 -!- LKoen has joined.
18:05:51 -!- Hooloovo0 has quit (Ping timeout: 248 seconds).
18:12:14 -!- Hooloovo0 has joined.
18:12:53 -!- arseniiv_ has changed nick to arseniiv.
18:13:33 <b_jonas> fizzie: is the channel log collecting process down?
18:14:31 <b_jonas> for the logs at esolang.org
18:15:57 -!- LKoen has quit (Remote host closed the connection).
18:16:44 <orin> *sokʷh₂yóteh₂ti *gʷíh₃womos : We live in a society in PIE
18:16:45 -!- LKoen has joined.
18:18:52 <ais523> b_jonas: apparently, it must be esowiki who's collecting the logs because they stopped when it segfaulted
18:26:22 <b_jonas> ais523: oh, there could be a connection
18:31:01 <b_jonas> ah I see, fizzie already knows about the problem
18:31:08 <b_jonas> (jsut checked the logs at tunes)
18:33:33 -!- arseniiv has quit (Read error: Connection reset by peer).
18:33:50 -!- arseniiv has joined.
18:46:23 -!- Hooloovo0 has quit (Ping timeout: 248 seconds).
18:47:15 <b_jonas> the advancement of solid state is amazing. when I was a child, little girls had shoes with red leds built into the side of their soles, flashing red at every step they took. now little girls have shoes with multi-colored leds built into them, which can flash in various colors including blue.
18:48:22 <b_jonas> if one of those were transported back in time, the little girl who wore them would be an envy of all their classmates
18:48:37 <b_jonas> s/an envy/the envy/
18:49:02 <arseniiv> b_jonas: :D
18:52:19 -!- Hooloovo0 has joined.
18:53:09 <b_jonas> blue leds and blue lasers, totally magical
18:53:52 <kmc> yeah
18:54:08 <kmc> I remember when blue LEDs were expensive and so they were a mark of high quality equipment
18:54:18 <kmc> then they became cheap and now every cheap shit from china is covered in unreasonably bright blue LEDs
18:54:34 <kmc> perhaps "high quality" is the wrong word but "expensive" anyway. think early 00s Sony
18:55:04 <kmc> in the past I had a computer case with a blue LED and I had to put a resistor in series to dim it
18:55:52 <kmc> b_jonas: what's also pretty cool is that you can buy RGB (or RGBW) leds like WS2812 in a standard 5050 package that have a controller chip on-board, so you can build a whole string of them and control them all individually with one data line
18:56:39 <kmc> this has made projects with many LEDs tremendously less of a pain in the ass and has lead to a boom in the amount of Burning Man style blinky LED art (we have some of that in public areas in SF, it's pretty cool)
18:56:48 <kmc> in Seoul it seemed every skyscraper was covered in RGB LEDs
18:57:12 <shachaf> soles and seoul
18:57:36 <ais523> kmc: what's the protocol like? two power, one data? two power, one data, one clock? one power, one data?
18:58:08 <shachaf> jade plate, six eight
18:58:18 <b_jonas> `? cheese
18:58:19 <kmc> ais523: power, ground and async serial (meaning it has precise timing requirements)
18:58:19 <HackEso> cheese? ¯\(°​_o)/¯
18:58:33 -!- S_Gautam has quit (Quit: Connection closed for inactivity).
18:58:43 <ais523> IIRC even RS-232 can be driven without a clock
18:58:53 <ais523> I can't remember offhand how many pins are actually necessary, but it isn't very many
18:58:54 <kmc> https://cdn-shop.adafruit.com/datasheets/WS2812.pdf
18:59:03 <kmc> ais523: RS-232 is *always* async
18:59:07 <ais523> the way it works is to send small bursts of clock signals in between the data so that the receiver can resync
18:59:11 <kmc> well, ok
18:59:27 <kmc> that's still async in my book, but with a method for one side to derive the clock
18:59:39 <kmc> but you can't arbitrarily change the timing between any two bits like you can with SPI
18:59:47 <kmc> might be called pleisiosynchronous
18:59:59 <kmc> another way to do that is Manchester encoding
19:00:06 <kmc> I have used that for a project with a single unidirectional data line
19:00:14 <b_jonas> kmc: I have a sheet of paper over the blue led of my voltage spike filtering electric socket thingy that is always on because the computer and the router are plugged into it
19:00:29 <kmc> the WS2812 has built in PWM control with 8 bits x 3 or 4 channels of dimming
19:00:49 <kmc> this means a full framebuffer is actually outside the RAM capacity of many microcontrollers, depending of course on how many LEDs you have
19:01:16 <FireFly> I need to do something with the WS2812 rolls I bought last year..
19:01:22 <ais523> wow, the protocol given there is bizarre
19:01:29 <ais523> hold times are different for 0 bits and 1 bits
19:01:38 <FireFly> yeah, it seems a bit funky
19:01:38 <kmc> yeah *shrug*
19:01:43 <kmc> it's wacky
19:01:46 <kmc> but you can get libraries for most platforms
19:01:56 <kmc> you can even drive it from Raspberry Pi, taking advantage of the built in PWM and DMA hardware
19:02:44 <kmc> 1-Wire seems to also have fixed timing
19:03:25 <kmc> a few cool things about 1-wire: a) each device has a unique 64-bit ID b) you can have arbitrarily many on the same bus, limited only by electrical characteristics
19:03:31 <ais523> my guess is that the protocol simplifies the circuitry in the decoder somehow
19:03:36 <kmc> c) you can multiplex power and data onto the same line if desired
19:03:39 <b_jonas> yes, RS-232 uses a data line where a byte can be delayed by any amount of time, but within the byte, the bits are placed at fixed time offsets after the leading edge of the byte
19:03:59 <kmc> yes
19:04:04 <kmc> same with RS-485 and RS-422
19:04:10 <b_jonas> but you need hardware that can interpret data with flexible timing anyway to read disks or casettes
19:04:18 <b_jonas> and they've had those ages ago
19:04:23 <kmc> which are pretty much the same "link layer" on top of a different hardware layer (full or half duplex differential signalling)
19:04:34 <b_jonas> so it's not too surprising that it's done on communication lines like ethernet in modern hardware
19:05:16 <kmc> yeah
19:05:22 <kmc> you can expect an ethernet device to have a precise clock
19:05:25 <b_jonas> RS-232 is simpler because it uses fixed times for the bits rather than flexible times
19:05:31 <kmc> not so much for a tiny cheap micro running off an internal RC oscillator
19:05:45 <b_jonas> I mean, obviously there's some tolerance because the sending side holds each bit for a short time,
19:05:52 <b_jonas> but it's not adaptive to different speeds
19:06:00 <b_jonas> whereas the casette and disk need to be
19:07:12 <kmc> I have also seen designs for combining RS-232/485 and power on the same wires
19:08:02 <kmc> by applying a DC offset, and AC-coupling the data at the other end, while drawing power through a series inductor
19:08:12 <b_jonas> what surprises me is the "power over ethernet" protocol. the power goes through two separate wires that are in addition to the four for normal ethernet. that's not the surprising part. the surprising part is that they send an immense amount of power through thin wires and tiny connectors.
19:08:46 <kmc> b_jonas: that's why it uses 48 volts
19:09:05 <kmc> so that it's only about an amp per conductor
19:09:12 <kmc> but still seems like a lot, yeah
19:09:18 <kmc> most telecom stuff uses 48V
19:09:27 <kmc> as well as server racks that do DC power with in the rack
19:09:32 <ais523> some people have claimed to demonstrate devices doing impossible things, e.g. producing more output than the power they're given as input, it's suspected that the results were faked via putting a DC offset on the live wire (which an AC-measuring voltmeter wouldn't see)
19:09:45 <kmc> mm
19:09:54 <kmc> a lot of RF amplifiers are powered through a DC offset on either the input or the output
19:09:56 <b_jonas> but sure, in general there's an obvious trend for connectors to use fewer and fewer wires
19:10:50 <b_jonas> you see it in USB, SD cards, SIM cards
19:12:08 <j4cbo> fuck the WS2812 though
19:12:22 <j4cbo> it has obnoxiously slow pwm and I can see the pwm artifacts
19:12:55 <j4cbo> b_jonas: PoE uses the center taps of the transformer
19:13:24 <j4cbo> it *can* use the spare pairs if it’s a 10/100 device... but there are no spare pairs with gigabit Ethernet
19:13:30 <b_jonas> HDMI and Displayport have fewer wires than DVI too
19:14:41 <b_jonas> j4cbo: from the side of your eyes, or even when looking directly at it?
19:14:48 <b_jonas> s/eyes/vision/
19:15:04 <j4cbo> only when moving my eyes
19:15:09 <b_jonas> ah yes
19:15:44 <j4cbo> apa102 are much better, and have an easier protocol (basically just regular spi) too
19:16:59 <fizzie> Blargh. It just immediately segfaults in std::thread::detach by jumping to 0. I think this isn't going to be something I can feasibly fix over the phone.
19:17:23 <fizzie> Won't be back home until next weekend though.
19:17:52 <b_jonas> ouch
19:18:07 <fizzie> I'll backfill the logs later, but it's going to stay down for now. :/
19:18:24 <b_jonas> thanks for trying though
19:18:27 <j4cbo> PoE is so good
19:19:09 <j4cbo> https://goo.gl/photos/vT8E1EMwJc16TFfHA
19:49:35 -!- FreeFull has joined.
20:26:07 -!- Hooloovo0 has quit (Ping timeout: 248 seconds).
20:32:49 <orin> I would prefer having more wires if it reduces latency
20:33:00 <orin> specifically for video connectors
20:33:57 <ais523> more wires normally improves throughput rather than latency
20:34:08 <ais523> although it depends on the protocl
20:34:09 <ais523> *protocol
20:34:19 <ais523> there will be a small latency improvement but typically only one words' worth
20:34:25 -!- Hooloovo0 has joined.
20:34:54 <orin> ais523: I was thinking the improvement would be by having less muxing
20:35:48 <ais523> hmm, I guess it depends on how much of the throughput is actually used in common circumstances
20:36:00 <ais523> I'm assuming that all of it would be, but you'd get an improvement if the throughput is mostly not used
20:36:25 <ais523> as it allows you to use your additional throughput to save latency if you've been idle for a while and suddenly have to send messages to a lot of sources at once
20:36:31 <ais523> err, destinations, not sources
20:37:25 <orin> also, at a practical level, these thinner wires and smaller connecters are flimsy
20:38:16 <orin> micro-displayport or whatever it's called
20:39:23 <orin> I never had a video connector fray before
21:13:55 <b_jonas> `? bruce lee
21:13:56 <HackEso> bruce lee? ¯\(°​_o)/¯
21:13:58 <b_jonas> `? bruce Lee
21:13:59 <HackEso> bruce Lee? ¯\(°​_o)/¯
21:25:23 -!- LKoen has quit (Remote host closed the connection).
21:25:41 -!- LKoen has joined.
21:29:24 -!- AnotherTest has quit (Ping timeout: 252 seconds).
21:46:09 -!- Hooloovo0 has quit (Excess Flood).
21:51:23 -!- Hooloovo0 has joined.
21:55:43 -!- b_jonas has quit (Quit: leaving).
22:05:21 -!- moei has quit (Quit: Leaving...).
22:11:09 -!- Lord_of_Life_ has joined.
22:13:51 -!- Lord_of_Life has quit (Ping timeout: 248 seconds).
22:13:52 -!- Lord_of_Life_ has changed nick to Lord_of_Life.
22:16:56 -!- 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.”).
22:19:07 -!- tromp has quit (Ping timeout: 248 seconds).
22:19:55 -!- Hooloovo0 has quit (Excess Flood).
22:20:22 -!- tromp has joined.
22:21:18 -!- tromp has quit (Remote host closed the connection).
22:23:59 -!- Hooloovo0 has joined.
22:28:53 -!- Sgeo has joined.
22:40:17 -!- ais523 has quit (Quit: quit).
22:48:14 <rdococ> In my understanding, a delimited continuation could be stored as two continuations - the point where the delimited continuation starts, and the point where it returns to its invoker.
22:48:45 <rdococ> Can a delimited continuation be delimited, then, by two continuations that are themselves delimited? Does this make sense? (I'm not asking anyone, I'm just putting an idea forward.)
22:54:56 -!- Hooloovo0 has quit (Excess Flood).
22:57:00 -!- Hooloovo0 has joined.
22:59:23 -!- tromp has joined.
23:03:55 -!- tromp has quit (Ping timeout: 248 seconds).
23:13:23 -!- Sgeo_ has joined.
23:16:48 <shachaf> Can you do constant-time sorts better than a sorting network with some kind of non-comparison sort?
23:17:09 -!- Sgeo has quit (Ping timeout: 258 seconds).
23:19:15 <int-e> what is there, besides radix sort (which is obviously not an option) and comparison-based sorts (or min/max sort, if you want to have fixed control flow)?
23:20:30 <shachaf> I don't know!
23:20:53 <shachaf> You can implement constant-time min/max using clever bitwise tricks if you don't have an instruction for them.
23:21:05 <shachaf> Maybe something like that can extend to more than two elements?
23:21:45 <int-e> I guess bogosort can be implemented in a side-channel free fashion ;-).
23:22:59 <shachaf> Is there any instruction set that has a min+max instruction?
23:29:25 <int-e> I don't know.
23:30:26 <int-e> intel has vectorized min and max (separately) at least.
23:31:51 <shachaf> Yes.
23:31:56 <int-e> And I guess having a combined instruction would be hard for them... because otherwise, the architecture never has more than a single output register.
23:32:24 <int-e> So it's likely that even if there was a combined min/max instruction it would become 2 microops.
23:32:52 <int-e> (for all I know, which is not too much)
23:48:05 -!- salpynx has joined.
23:57:00 <salpynx> fungot: I have a suspicion that the esolangs.org/logs are not updating, what are your thoughts?
23:57:00 <fungot> salpynx: " fnord department for, well, one of nature's born fnord. why, hadn't he? to bes pelargic the empire's only proper seaport.
23:58:29 <shachaf> It's true. The bot crashed on that long wiki page rename.
←2019-05-27 2019-05-28 2019-05-29→ ↑2019 ↑all