00:16:55 -!- tromp has joined.
00:22:05 -!- tromp has quit (Ping timeout: 240 seconds).
00:47:14 -!- oerjan has joined.
00:52:07 <lambdabot> fizzie said 16h 22m 37s ago: Good point, I didn't even consider them bots.
00:53:57 <HackEso> 11585:2018-07-21 <oerjän> addquote <Aearnus> i\'m sending this from within a computer on minecraft \ 11580:2018-07-01 <alercäh> addquote <zzo38> Please look at the new [[BackTurn]] program language see if it is good or else what other comment/question/complaint. \ 11551:2018-05-08 <oerjän> addquote <shachaf> Taneb: are you suggesting the Tanebvention joke might be getting slightly old <Taneb> shachaf, not at all <Taneb> I would never suggest that i
00:54:36 -!- XorSwap has joined.
00:58:40 -!- sjums has joined.
01:00:10 -!- sjums has quit (Read error: Connection reset by peer).
01:04:52 -!- imode has joined.
01:13:21 -!- Phantom_Hoover has quit (Read error: Connection reset by peer).
01:17:48 <oerjan> the last panel of the last oots isn't entirely reassuring.
01:32:14 <oerjan> oh thank god, there was a preference setting to hide the left navigation bar on PPCG again
01:33:23 -!- S_Gautam has joined.
01:35:15 -!- sparklefarkle has joined.
01:36:27 -!- sparklefarkle has quit (Remote host closed the connection).
02:02:23 -!- tromp has joined.
02:07:34 -!- tromp has quit (Ping timeout: 264 seconds).
02:46:38 -!- pixdamix has joined.
02:46:55 -!- pixdamix has quit (K-Lined).
02:59:12 -!- variable has joined.
03:20:52 -!- variable has quit (Quit: /dev/null is full).
03:21:44 -!- variable has joined.
03:47:22 -!- tromp has joined.
03:52:18 -!- tromp has quit (Ping timeout: 268 seconds).
03:54:43 -!- trout has joined.
03:57:26 -!- variable has quit (Ping timeout: 260 seconds).
04:25:43 -!- variable has joined.
04:28:57 -!- ForexTrader has joined.
04:29:23 -!- trout has quit (Ping timeout: 276 seconds).
04:30:33 -!- ForexTrader has quit (Remote host closed the connection).
04:57:22 -!- trout has joined.
05:00:53 -!- variable has quit (Ping timeout: 265 seconds).
05:03:09 <zzo38> If you want to make a total of floating numbers without losing precision must you put them in order, or accumulate them by log2 or whatever? (You might also have to consider the positive and negative numbers separately?)
05:28:56 -!- variable has joined.
05:32:26 -!- trout has quit (Ping timeout: 276 seconds).
05:44:11 -!- Vorpal_ has quit (Quit: ZNC - http://znc.sourceforge.net).
06:00:23 -!- trout has joined.
06:03:14 -!- variable has quit (Ping timeout: 265 seconds).
06:23:05 -!- imode has quit (Ping timeout: 240 seconds).
06:24:01 -!- oerjan has quit (Quit: Nite).
06:31:38 -!- variable has joined.
06:34:42 -!- trout has quit (Ping timeout: 260 seconds).
06:41:57 -!- tromp has joined.
06:58:35 -!- XorSwap has quit (Ping timeout: 240 seconds).
07:02:45 -!- trout has joined.
07:06:04 -!- variable has quit (Ping timeout: 265 seconds).
07:34:12 -!- variable has joined.
07:37:53 -!- trout has quit (Ping timeout: 276 seconds).
07:39:07 -!- AnotherTest has joined.
07:50:05 -!- AnotherTest has quit (Ping timeout: 240 seconds).
08:05:51 -!- trout has joined.
08:05:52 -!- erkin has joined.
08:09:23 -!- variable has quit (Ping timeout: 265 seconds).
08:18:38 -!- tromp has quit (Remote host closed the connection).
08:31:24 -!- tromp has joined.
08:37:21 -!- variable has joined.
08:40:56 -!- trout has quit (Ping timeout: 276 seconds).
09:01:10 <int-e> `learn Hodl ym bere, I'ev gto thsi!
09:01:12 <HackEso> Learned 'hodl': Hodl ym bere, I'ev gto thsi!
09:09:06 -!- trout has joined.
09:12:42 -!- variable has quit (Ping timeout: 265 seconds).
09:41:10 -!- variable has joined.
09:44:38 -!- trout has quit (Ping timeout: 276 seconds).
10:11:44 -!- trout has joined.
10:12:53 -!- S_Gautam has quit (Quit: Connection closed for inactivity).
10:14:58 -!- variable has quit (Ping timeout: 260 seconds).
10:43:50 -!- variable has joined.
10:46:57 -!- trout has quit (Ping timeout: 265 seconds).
11:15:24 -!- trout has joined.
11:18:14 -!- variable has quit (Ping timeout: 276 seconds).
11:27:35 -!- Phantom_Hoover has joined.
11:46:09 -!- variable has joined.
11:46:57 -!- wob_jonas has joined.
11:48:04 <wob_jonas> Phantom_Hoover: wow. that is something. (re Sir Robin in game of life)
11:49:14 -!- trout has quit (Ping timeout: 260 seconds).
11:52:44 -!- ululate9 has joined.
11:53:32 -!- ululate9 has quit (Remote host closed the connection).
12:07:20 <esowiki> [[RANDo]] https://esolangs.org/w/index.php?diff=57219&oldid=54869 * Kamish * (+54)
12:10:51 -!- codex2064 has joined.
12:11:11 -!- codex2064 has quit (Remote host closed the connection).
12:17:51 -!- trout has joined.
12:20:43 -!- variable has quit (Ping timeout: 265 seconds).
12:22:26 -!- MillerBoss1 has joined.
12:23:37 -!- MillerBoss1 has quit (Remote host closed the connection).
12:29:38 -!- rej5 has joined.
12:30:19 -!- rej5 has quit (Read error: Connection reset by peer).
12:35:45 -!- Turner92 has joined.
12:41:05 -!- Turner92 has quit (Ping timeout: 240 seconds).
12:48:46 -!- variable has joined.
12:52:29 -!- trout has quit (Ping timeout: 276 seconds).
13:15:35 -!- SopaXorzTaker has joined.
13:19:09 -!- erkin has quit (Quit: Ouch! Got SIGIRL, dying...).
13:20:42 -!- trout has joined.
13:23:12 <esowiki> [[MIX (Knuth)]] https://esolangs.org/w/index.php?diff=57220&oldid=53097 * B jonas * (+147) clarify that this is not an esoteric language
13:23:41 <esowiki> [[Game of Life]] https://esolangs.org/w/index.php?diff=57221&oldid=30898 * B jonas * (+410) credit Conway with Turing-completeness; add like ten categories; link LifeWiki
13:24:02 -!- variable has quit (Ping timeout: 265 seconds).
13:38:03 <esowiki> [[Game of Life]] https://esolangs.org/w/index.php?diff=57222&oldid=57221 * B jonas * (+187) mention still active research
13:44:39 -!- Lean1 has joined.
13:44:51 -!- Lean1 has left.
13:47:09 <esowiki> [[Talk:RAM0]] N https://esolangs.org/w/index.php?oldid=57223 * B jonas * (+612) a mistaken statement
13:52:02 -!- variable has joined.
13:55:32 -!- trout has quit (Ping timeout: 276 seconds).
14:05:40 <esowiki> [[Game of Life]] https://esolangs.org/w/index.php?diff=57224&oldid=57222 * B jonas * (+1346) explicit definition; program size
14:12:19 -!- pikhq has quit (Ping timeout: 260 seconds).
14:17:44 -!- wob_jonas has quit (Quit: http://www.kiwiirc.com/ - A hand crafted IRC client).
14:18:27 <esowiki> [[Game of Life]] https://esolangs.org/w/index.php?diff=57225&oldid=57224 * B jonas * (-4) consistent spelling of neighbor and behavior
14:21:12 <esowiki> [[Von Neumann's 29-state cellular automaton]] N https://esolangs.org/w/index.php?oldid=57226 * B jonas * (+1478) Created page with "'''Von Neumann's 29-state cellular automaton''' is a [[Cellular automaton]] published in 1966. The state of the automaton is a square grid filling the whole plane, in which..."
14:24:03 -!- trout has joined.
14:24:08 -!- pikhq has joined.
14:26:51 -!- variable has quit (Ping timeout: 240 seconds).
14:28:07 <esowiki> [[Cellular automaton]] https://esolangs.org/w/index.php?diff=57227&oldid=45557 * B jonas * (+482) history
14:29:18 <esowiki> [[Language list]] https://esolangs.org/w/index.php?diff=57228&oldid=57168 * B jonas * (+48)
14:34:51 <esowiki> [[Cellular automaton]] https://esolangs.org/w/index.php?diff=57229&oldid=57227 * B jonas * (+454) /* Relation to esoteric programming */ clarify that every CA is an esolang
14:37:05 <esowiki> [[Cellular automaton]] https://esolangs.org/w/index.php?diff=57230&oldid=57229 * B jonas * (+2) /* Relation to esoteric programming */
14:39:47 <esowiki> [[Prehistory of esoteric programming languages]] https://esolangs.org/w/index.php?diff=57231&oldid=49933 * B jonas * (+406) add cellular automata
14:48:06 -!- wob_jonas has joined.
14:50:09 <wob_jonas> I was wondering what the oldest esolang with still active research is. Game of Life is definitely old and has a large active research community, and dc might qualify and might just predate it. But then I realized that lambda calculus, or lambda calculus with some specific evaluation order, probably trumps anything else.
14:51:52 <wob_jonas> Them and Turing-machines. They were both first published in 1936.
14:55:13 -!- variable has joined.
14:57:25 -!- wob_jonas has quit (Quit: http://www.kiwiirc.com/ - A hand crafted IRC client).
14:58:42 -!- trout has quit (Ping timeout: 260 seconds).
15:16:35 -!- variable has quit (Quit: Found 1 in /dev/zero).
15:31:31 <esowiki> [[Rosa Parks]] https://esolangs.org/w/index.php?diff=57232&oldid=57091 * Plokmijnuhby * (+327)
15:35:34 -!- Phantom_Hoover has quit (Ping timeout: 256 seconds).
15:42:26 <esowiki> [[Rosa Parks]] M https://esolangs.org/w/index.php?diff=57233&oldid=57232 * Plokmijnuhby * (+1)
15:43:52 -!- imode has joined.
15:46:01 -!- imode has quit (Client Quit).
15:47:55 -!- rosseaux26 has joined.
15:49:25 -!- rosseaux26 has quit (Remote host closed the connection).
16:05:52 -!- missnomer2 has joined.
16:06:05 -!- missnomer2 has quit (Remote host closed the connection).
16:13:27 <zzo38> 2HI have implemented a SQLite extension to download files from the internet by using libcurl, but it involves a rather klugy way to detect if sqlite3_interrupt() has been called.
16:19:08 -!- TriangleSausage has joined.
16:20:42 -!- TriangleSausage has quit (Remote host closed the connection).
16:24:10 -!- AnotherTest has joined.
16:24:22 -!- duckgoose15 has joined.
16:30:14 -!- AnotherTest has quit (Ping timeout: 244 seconds).
16:30:21 -!- duckgoose15 has quit (Ping timeout: 264 seconds).
16:31:15 -!- Nick` has joined.
16:37:21 -!- Nick` has quit (Ping timeout: 248 seconds).
16:37:35 -!- Ckat16 has joined.
16:37:56 -!- Ckat16 has quit (Killed (Unit193 (Spam is not permitted on freenode.))).
16:47:07 -!- Ryuzaki has joined.
16:53:48 -!- Ryuzaki has quit (Ping timeout: 265 seconds).
16:57:07 -!- dh1284 has joined.
16:57:58 -!- dh1284 has quit (Killed (Sigyn (Spam is off topic on freenode.))).
17:14:13 -!- borsin15 has joined.
17:14:41 -!- borsin15 has quit (Killed (Unit193 (Spam is not permitted on freenode.))).
17:21:10 -!- Lord_of_Life has quit (Excess Flood).
17:21:34 -!- Lord_of_- has joined.
17:27:11 -!- shikhin_ has joined.
17:28:24 -!- sebbu has quit (Ping timeout: 256 seconds).
17:28:25 -!- shikhin has quit (Ping timeout: 256 seconds).
17:28:25 -!- lifthrasiir has quit (Read error: Connection reset by peer).
17:30:22 -!- lifthrasiir has joined.
17:35:39 -!- Quokka25 has joined.
17:36:15 -!- Quokka25 has quit (Remote host closed the connection).
17:37:50 -!- catfuneral has joined.
17:38:26 -!- Lord_of_Life has joined.
17:38:56 -!- Lord_of_- has quit (Ping timeout: 256 seconds).
17:40:22 -!- paul2520 has quit (Ping timeout: 256 seconds).
17:43:36 -!- paul2520 has joined.
17:43:36 -!- catfuneral has quit (Ping timeout: 256 seconds).
17:43:39 -!- paul2520 has quit (Changing host).
17:43:39 -!- paul2520 has joined.
17:47:55 <zzo38> I think someone ask about SQLite extension to use Swiss Ephemeris (to see the example code), and I have posted it now (although some features are not currently implemented, including sunrise/sunset times, and some other stuff)
17:51:57 -!- MikeSpears15 has joined.
17:53:13 -!- MikeSpears15 has quit (Killed (Sigyn (Spam is off topic on freenode.))).
18:04:54 -!- evil has joined.
18:06:04 -!- evil has quit (Remote host closed the connection).
18:37:41 -!- Phantom_Hoover has joined.
18:43:52 -!- EdSaperia20 has joined.
18:49:35 -!- EdSaperia20 has quit (Ping timeout: 240 seconds).
19:09:30 -!- wob_jonas has joined.
19:12:47 <wob_jonas> You've all heard the tale of Scheherazade, in which an evil sultan vows that every evening he would marry a beautiful maiden and every morning he would get that maiden executed?
19:13:28 <wob_jonas> Apparently in at least some variants, the sultan actually vowed that every evening he'd marry the *most* beautiful maiden of his empire.
19:14:37 <wob_jonas> But in reality, wouldn't that cause one of those unending arms races, when eventually every remaining maiden tries to deliberately make themselves more and more ugly, to avoid this fate?
19:17:10 <wob_jonas> I mean, maybe it would, and the sultan was stupid to make such a vow no matter what and the simplest solution would have been to get him assasinated as soon as possible after people find that he is willing to follow through his vow, but still, I haven't heard of that arms race part in any retelling of the story.
19:24:48 -!- sebbu has joined.
19:25:08 -!- AnotherTest has joined.
19:39:29 -!- SopaXorzTaker has quit (Quit: Leaving).
19:55:25 <zzo38> What if his opinion who is most beautiful may differ somewhat from one of those maiden? Or if it is about equal and is difficult to know?
19:56:48 <int-e> they should just ask the mirror from Snow White.
19:57:03 <Taneb> wob_jonas, what if it was an environment with commonplace arranged marriages and the status gained from having a daughter marry the sultan is worth the high chance of her execution
19:57:25 <wob_jonas> I imagine that if beauty is subjective and its measure used by the oath is not easy to predict, that would just speed up the arms race because every maiden will want to make herself so obviously ugly that by no matter what measure she isn't among the top ten thousand most beautiful ones.
19:57:31 <int-e> Taneb: worth to whome though...
19:57:39 -!- AnotherTest has quit (Ping timeout: 260 seconds).
19:57:51 <Taneb> int-e, the people arranging the marriage (probably the poor girl's father)
19:58:28 <wob_jonas> Right, but the maidens themselves would probably find a way to become ugly without the permission of their fater. The surviving girls at least, soon.
19:58:38 <int-e> Taneb: So unfair. Almost like real life.
20:18:33 -!- foxcookie has joined.
20:19:17 -!- foxcookie has quit (Remote host closed the connection).
20:34:09 -!- KobrAs29 has joined.
20:34:50 -!- KobrAs29 has quit (Remote host closed the connection).
20:41:36 -!- ais523 has joined.
20:45:27 -!- benoliver99916 has joined.
20:45:45 -!- benoliver99916 has quit (Killed (Unit193 (Spam is not permitted on freenode.))).
20:51:00 <wob_jonas> ais523: https://esolangs.org/wiki/Talk:RAM0
20:51:10 -!- tromp has quit (Remote host closed the connection).
20:51:26 -!- tromp has joined.
20:52:00 <ais523> oh, yes, I was kind-of assuming you couldn't put an oracle there
20:52:20 <ais523> (also there's not much point linking me to talk pages of my own languages, I'm likely to notice anything put there anyway)
20:52:50 <wob_jonas> but since you asked lambdabot for @messages, I corrected him when he said no messages today
20:52:56 <ais523> what I meant was that the language is still TC if it's initialised randomly
20:53:21 <esowiki> [[RAM0]] https://esolangs.org/w/index.php?diff=57234&oldid=54505 * Ais523 * (+16) /* Data storage */ clarify ambiguous sentence
20:53:52 <esowiki> [[Talk:RAM0]] https://esolangs.org/w/index.php?diff=57235&oldid=57223 * Ais523 * (+247) clarified
20:53:55 <wob_jonas> what? then it has a randomness source... hmm wait, that doesn't increase the power of R. right.
20:54:19 <wob_jonas> although it's not obvious what "initialize randomly" means in a machine with bigints in each cell
20:54:43 <int-e> wob_jonas: it's not a source of randomness if you can make no assumptions about the memory... it might be initialized.
20:55:19 <int-e> wob_jonas: oh wait. sorry, what ais523 wrote here and on the page are different things
20:55:42 <ais523> yes, I wrote it a bit more clearly on the page
20:56:10 -!- tromp has quit (Ping timeout: 264 seconds).
20:56:11 -!- AnotherTest has joined.
20:56:15 <int-e> yeah, the page looks fine to me
20:59:37 <ais523> @tell tswett re your oversized digit notation, have you seen hyperbinary? it's basically binary except you're allowed to use 2 as a digit as well, and is mostly relevant as a method of defining the stern-brocot sequence
21:02:38 <ais523> @tell Phantom_Hoover I think "elementary" in cellular automata means that it isn't a collection of smaller components; there were already known Game of Life knightships that didn't rely on universal constructors, e.g. the half-baked knightship, which relied on multiple knightship-ish components that sent beams of generalised-gliders out to stabilise each other
21:04:21 <wob_jonas> That exclamation by Phantom_Hoover made me make some improvements to the Game of Life article on our wiki, although of course it still doesn't contain much of the vast knowledge mathematicians now have of Game of Life.
21:04:49 <wob_jonas> Or more like, inspired me to make some improvements, seeing that such a popular esolang has such a bad article.
21:05:06 <ais523> there's an entire wiki for the game of life
21:05:15 -!- jrg28 has joined.
21:05:15 <ais523> so there's not much point duplicating it here
21:05:24 <wob_jonas> yes, I linked it. two wikis at least, I think.
21:05:52 <wob_jonas> It was still worth to increase the article from stub length, and may still be worth to grow it a bit, just because it's such an important language.
21:06:06 <ais523> <Phantom_Hoover> elementary = not a giant million cell machine engineered to build a copy of itself along a certain vector then take itself apart
21:06:12 <wob_jonas> There are entire books on the Lambda calculus too, but it deserves a good article on our wiki still.
21:06:57 -!- AnotherTest has quit (Ping timeout: 264 seconds).
21:07:05 <wob_jonas> And the Sir Robin article that PH linked to has the word "elementary" linked to a page where that concept is clearly described, and it is what ais says it means.
21:07:23 <Phantom_Hoover> tbqh the game of life isn't that interesting from an esolang perspective because a very small subset of what's known about it suffices for TCness
21:08:21 <Phantom_Hoover> i think the 'microscale' engineering needed to e.g. build a replicator is the real meat of it
21:08:50 <int-e> Phantom_Hoover: I don't see how this speaks against it being an esolang
21:08:55 <wob_jonas> ais523: then I was wondering if Game of Life, born in 1970, was the *oldest* esolang with still active research. I looked at dc, since people still occasionally write new obfus in it, which sort of counts as "active research", but I can't determine for sure whether dc is older or newer than Game of Life.
21:09:11 <int-e> hmm. or perhaps you just find the fiddling boring
21:09:14 <ais523> I think the game of life is interesting because instead of an explicit list of commands, it provides a number of components which feel a bit like they're randomly generated
21:09:32 <ais523> most of which are useless, some of which happen to have functionality that can be pieced together into a TC language
21:09:51 <wob_jonas> ais523: Then I decided that the question is stupid, because both lambda calculus and Turing-machines are so fucking old they probably trump everything, both having been published in 1936.
21:10:15 <ais523> what about the analytical engine's language?
21:10:17 <wob_jonas> And both lambda calculus and Turing-machines are still languages that people research.
21:10:35 -!- jrg28 has quit (Ping timeout: 240 seconds).
21:10:39 <int-e> yeah I almost wrote earlier that neither lambda calculus not Turing machines are esolangs.
21:10:42 <ais523> also, basic untyped lambda calculus is unlikely to have much research left, people mostly just make derivatives of it
21:11:55 <wob_jonas> int-e: I think they are esolangs, only without a fixed syntax, but that doesn't make them less of esolangs. the execution model is complete enough.
21:12:15 <int-e> Hmm, perhaps. You prove subject reduction, so the Church-Rosser proof carries over to whatever typed system you really study.
21:12:24 <wob_jonas> if we accept languages with no fixed definition of how they do IO or even deliver a yes-no result and they never halt, then why wouldn't we allow no fixed syntax?
21:12:48 <int-e> wob_jonas: I'm trying to say that they're not esoteric enough.
21:12:56 <wob_jonas> plus, ais523 has submitted some articles of his new esolangs that have no fixed syntax, and do you want to go against ais523's claim that those are esolangs?
21:13:27 <int-e> Well, the ambiguity isn't your fault. :)
21:13:43 <wob_jonas> I think they're esoteric in the sense that they're Turing-tarpits, being Turing-complete but its full ruleset being very small
21:13:45 <int-e> And I can hardly blame you for not reading my mind.
21:14:08 <wob_jonas> And few people use them for actual practical programming
21:14:56 <wob_jonas> and not because it's unpopular, but because they're not very suitable, because just like most other tarpits, they're not rich enough
21:15:25 <int-e> And I would probably say that Turing machines are more esoteric than the lambda calculus, because the latter is an actual viable target language for functional programming.
21:16:06 <Phantom_Hoover> int-e, i didn't say it didn't have a place on the esolang wiki just that the most interesting stuff about it involves more than computation
21:16:12 <int-e> (so clearly we disagree on the richness of lambda calculus)
21:16:51 <wob_jonas> oh, I certainly agree that Turing-machines are more esoteric
21:17:02 <Phantom_Hoover> like i haven't actually researched the history involved but it seems to me that the lambda calculus is clearly the richer, more useful computational system; but turing machines were of more scientific interest because they're very clearly physically realisable
21:18:30 <wob_jonas> PH: I think they were of more scientific interest exactly because they're esoteric. they have a strange execution time so asking how fast you can compute something on a Turing-machine is nontrivial and different from the execution time on normal machines;
21:20:01 <wob_jonas> finding the smallest universal Turing-machine or the largest number a small Turing-machine can compute gets nontrivial quickly (although there are similarly simple golf questions about finding the smallest universal combinator or something)
21:20:02 <Phantom_Hoover> i've often wondered why turing is the Big Name in the field of early formal computation when church was slightly earlier, he proved essentially the same thing and his formalism has gone on to be vastly more influential in modern computing
21:20:56 <wob_jonas> PH: so you're saying that we have to measure the esotericness of a language compared to its age, and there's not many other languages in their age to compare to?
21:23:12 <Phantom_Hoover> but anyway, especially early on when 'what will physical computers be able to do' was presumably an important new question, the fact that turing's formal machine could clearly be mapped to the operation of an (idealised) physical computer would have made results like universality far more important than they were for the lambda calculus
21:23:21 <wob_jonas> I mean, that's true, but there's still, I guess, arithmetic notations, which existed back then and could count as very limited programming languages,
21:23:26 <int-e> looking at the Z3... "Data memory: 64 words with a length of 22 bits" "Program memory: Punched celluloid tape"
21:23:43 <wob_jonas> and in particular in 1980 Hilbert considered the set of Diophantine equations, which you could consider an early esolang actually, so there are other languages to compare
21:24:00 <wob_jonas> heck, Diophantine equations might be the earliest esolang that still have active research
21:24:55 -!- imode has joined.
21:26:23 <wob_jonas> PH: yes, but didn't people seriously start to consider building physical computers later than lambda-calculus was published, unless the legends about Babbage are true (they sound as exaggerated as the Da Vinci ones)?
21:27:33 <Phantom_Hoover> i don't know that much about the actual history! i think they were seriously considering it well before they actually built any
21:28:05 <Phantom_Hoover> like i've always kind of assumed turing had physical realisation in mind when he came up with the turing machine because, well, why else would you come up with anything so painfully awkward
21:28:22 <wob_jonas> PH: but how much before? only two decades, or five decades?
21:29:36 <int-e> Phantom_Hoover: maybe the point was to have something that could obviously be built in principle (modulo the infinite tape length) :)
21:29:41 <wob_jonas> it seems like some of those legends about Babbage are true, because he built a pretty good mechanical calculator, perhaps better than any before.
21:30:04 <imode> mechanical computers were in wide use prior to LC.
21:30:18 <imode> mechanical calculating hardware was in wide use prior to LC.
21:30:19 <int-e> Phantom_Hoover: how does one build a random access memory, really? I mean, imagine you didn't know how it was done...
21:30:41 <imode> hell the earliest programmable hardware was looms.
21:31:45 <wob_jonas> imode: that's true, I have a photo of one in https://commons.wikimedia.org/wiki/File:Wien-khm-kunstkammer-calculator.jpg
21:31:57 <imode> beautiful hardware.
21:32:44 <wob_jonas> and that one is clearly a decorative one, probably not much used in reality, but was probably also working (for some value of working), so at that point mechanical calculators must have been "cheap" enough that you would consider giving one as a useless gift to a king
21:33:11 <wob_jonas> although the description said something about how it was used by court astronomers or something
21:33:13 <imode> definitely. just look at the history of cash registers.
21:33:17 <wob_jonas> so maybe it wasn't purely decorative
21:34:05 <wob_jonas> maybe it was somewhat practical, but also beautifully decorative, like many other items in that exhibition, including utensils and grooming tools
21:34:20 <int-e> https://en.wikipedia.org/wiki/File:WLANL_-_jpa2003_-_Jaquardweefgetouw_03.jpg looks impressive
21:34:26 <wob_jonas> though some of them did look so richly decorated that I can hardly imagine they're practical to use
21:35:02 <imode> hell, mechanical computers were used to calculate ballistic trajectories in the navies around the world well up through the early 1900's.
21:35:19 <wob_jonas> and the Antikythera mechanism is claimed to be the oldest mechanical computer we know of, and it's from the *antiquities*, and it's one specialized for astronomical calculations
21:36:05 <imode> I just plopped into this discussion, what's it about?
21:36:07 <wob_jonas> perhaps there were actually languages of that age to compare to, even if I know close to nil about them
21:36:43 <wob_jonas> imode: I started by asking if Game of Life is the oldest esolang with still active research in it, then wondered if Turing-machines fit the bill,
21:36:59 <imode> oh yeah turing machines fit the bill lol.
21:37:05 <wob_jonas> imode: then it followed up by a discussion of whether Turing-machines are an esoteric language, and how you even define esoteric language,
21:37:07 <imode> persistent TMs are even more obscure.
21:37:51 <wob_jonas> and then about old physically built computers
21:38:10 -!- BackUP25 has joined.
21:38:52 -!- BackUP25 has quit (Remote host closed the connection).
21:39:13 <imode> I actually wonder if an abacus counts.
21:39:19 <wob_jonas> My favourite item from that exhibition, though, is a purely decorative one: a statuette https://commons.wikimedia.org/wiki/Category:Gefesselter_Prometheus_(Hagenauer) of which I didn't manage to make good enough quality photos
21:40:29 <wob_jonas> because there were big difficulties making photos in a dark room with ugly uneven directional lighting and lots of reflective glass cabinets with all four doors open arranged through a large room and other visitors and no tripods allowed and
21:41:13 <wob_jonas> me having only a compact camera that isn't really ideal for that sort of challenge, even though I definitely really like that camera and think it was really worth its price and have served me well so far.
21:41:46 <imode> as the "oldest computer".
21:42:42 <int-e> imode: an abacus counts, adds, and with some care, multiplies as well
21:43:22 <wob_jonas> int-e: but didn't the babylonian scribes also add and multiply on clay tablets too?
21:43:23 <imode> int-e: ayyyy there's the pun. :P
21:44:15 <int-e> wob_jonas: we're quickly approaching snakes and log tables, I see.
21:44:17 <wob_jonas> perhaps such a scribe or group of scribes could count as a computer, and programmable too if their liege orders them to do a specific calculation
21:44:46 <int-e> (beware of adders, they are venomous)
21:45:03 <wob_jonas> and yeah, some people make the case that the universe is a giant computer, and it's clearly older than anything invented by humans
21:45:19 <Phantom_Hoover> i think you basically need to discount anything that relies on the human mind as a computer
21:45:21 <wob_jonas> int-e: yeah, I heard that pun already
21:45:41 <int-e> https://en.wikipedia.org/wiki/Human_computer
21:46:04 <wob_jonas> PH: relies in what sense? all computers built so far rely on the human mind to repair and work for more than a few decades continuously
21:46:47 <Phantom_Hoover> i refuse to believe you can be confused on what i mean by this except for the sake of pedantry
21:46:56 <wob_jonas> there have been plans for computers that last for longer, but I don't think any have been built yet
21:48:31 <wob_jonas> not that I'm not impressed by how well some computers launched in space manage to work for decades under physically very limiting conditions, mind you
21:51:53 <wob_jonas> That the Voyager 1 has had a working computer and able to contact us still ever since 1977 is definitely one of the crowning jewels of human-built technology so far.
21:51:55 <int-e> Well, we *can* build computer hardware to last, and be fault-tolerant and redundant, especially when it doesn't need to be fast.
21:52:29 <wob_jonas> int-e: theoretically yes, but I don't think anyone put together the money for it yet
21:52:32 <imode> what's the longest-lasting computer we can build?
21:53:14 <ais523> wob_jonas: I thought it was sufficiently distant that its bandwidth had dropped below the rate required to accomplish anything useful with the communications?
21:53:20 <int-e> wob_jonas: I'm hard pressed to think of a market for it, beyond space crafts (and some other hardware that is hard to access. industrial, perhaps. or things we put inside undersea cables.).
21:53:39 <ais523> also, some of the computers from the early days of transistors may still be working
21:53:41 <wob_jonas> imode: there is some research on that. some people proposed ones that would work for ten thousand years unrepaired or something and is earthbound, and is just a clock (barely counts as a computer), but it's probably unrealistic.
21:53:56 <ais523> thermionic valves are really unreliable, and when people started replacing them by transistors it took a while for them to revise the computer designs
21:53:56 <imode> yeah a clock wouldn't count imho.
21:54:05 <int-e> wob_jonas: anyway, putting some random keywords together: https://www.militaryaerospace.com/articles/print/volume-28/issue-6/technology-focus/radiation-hardened-space-electronics-enter-the-multi-core-era.html
21:54:10 <imode> turing completeness is what matters.
21:54:20 <imode> smallest, longest-lasting turing complete piece of hardware.
21:54:22 <ais523> we don't have any Turing complete computers though
21:54:34 <imode> yeah yeah mr. pedant.
21:54:44 <ais523> that makes it quite hard to define a computer in real life
21:54:54 <ais523> I guess some sort of informal "usable for programming" is what we care about
21:55:16 <imode> construct me the longest-lasting linear bounded automaton you have.
21:55:46 <int-e> wob_jonas: oh and a nice closing sentence: 'Aircraft flying at altitude, at about 30,000 feet and above, also are starting to experience radiation-induced effects. "There are 500 times more neutrons at 30,000 feet than there are on the ground," points out Aitech's Romaniuk.'
21:55:48 <ais523> hmm, a physically realisable LBA would require some sort of rewritable physical medium to supply the input
21:55:58 <int-e> (near closing, if you want to be technical)
21:56:00 <wob_jonas> ais523: it doesn't look like that, if I read https://en.wikipedia.org/wiki/Voyager_1 correctly
21:56:11 <imode> I wouldn't rely on any electronic or MEMs shit.
21:56:28 <imode> I guess we're also going to talk size.
21:56:35 <wob_jonas> ais523: it's the two Pioneer probes, both of which were launched prior to the two Voyagers, that can no longer communicate with us
21:56:43 <ais523> wob_jonas: ah right, apparently we can still tell it to steer and get back information telling us it's changed course
21:57:28 <ais523> that wouldn't need a lot of bandwidth
21:58:03 <wob_jonas> int-e: yes, there's no market for computers that work for a long time, because nobody cares for such long term investments financially, which is probably why none have been built in practice
21:58:40 <ais523> especially because computer technology improves all the time, and so a computer is likely to become obsolete fairly quickly
21:59:10 <imode> I wouldn't even care about the speed or size of it, I just want the most durable and longest-lasting piece of hardware, in whatever form it may take.
21:59:13 <ais523> although a working computer from, say, the year 2000 would still be useful nowadays, even though it's much less powerful than modern computers are
21:59:25 <imode> you may be able to make a fluidic TM out of some of the hardest material on the planet.
21:59:32 <ais523> imode: I can imagine that some microcontrollers would be very durable
21:59:53 <imode> ais523: I wouldn't rely on anything electrical or explicitly mechanical. wear and tear.
22:00:04 <imode> fluidics sound like a good option.
22:00:11 <ais523> applications like computerised sensors might legitimately want to work for a long time
22:00:39 <imode> 10,000 years in space?
22:00:40 <ais523> imode: I've been thinking of a physical implementation of The Waterfall Model using actual flowing water, but it'd likely accumulate tolerance errors over time unless you had some sort of quantisation involved
22:00:44 <wob_jonas> ais523: right, and part of that is because better radio wave receivers and transmitters have been built than were possible when it (Voyager 10) was launched, although people probably anticipated back than that such will be built
22:01:07 <imode> ais523: I'm thinking of fluidic logic gates in an enclosed block of some kind of material.
22:01:12 <imode> we don't care about getting output from it.
22:01:20 <ais523> imode: fluid scour would be a huge issue with something like that
22:01:48 <ais523> I think electrical or optical computation would be the right way to go, having no moving parts is likely to increase durability considerably
22:02:10 <ais523> with electrical computation, you'd want to use some really simple, hardwearing components like BJTs
22:02:57 <imode> I dunno, I don't think any of it would last. thermals, etc.
22:03:39 <wob_jonas> further, I believe Voyager 10 has a *programmable* computer because when it was launched, its creators anticipated that future research on Earth would let them decide how to change its program for the better, and communication bandwidth limits wouldn't allow the program to just be executed on Earth
22:03:40 <ais523> I just looked it up, as long as the current stays low enough the only known failure mode of BJTs within their normal operating envelope is due to ionizing radiation
22:04:28 <imode> I still think it should be something more passively mechanical like fluidics. either using gasses or actual fluids.
22:06:42 <wob_jonas> imode: some science writer whose name eludes me did propose that a Turing-complete computer that lives forever may be physically built, but only with capabilities of civilizations much more advanced than ours
22:07:18 <wob_jonas> its purpose would be to run a mind, so that they thus gain immortality
22:07:57 <imode> asimov's the last question?
22:09:30 <wob_jonas> imode: it actually details the physics part
22:09:53 <wob_jonas> and was intended as non-fiction, although "futurology" bounds on fiction
22:10:33 <zzo38> If you play GURPS, what TL number would you have such thing of that?
22:10:46 <imode> nothing like that's coming to mind immediately...
22:10:52 <wob_jonas> I guess the Voyager 10 might count as one of the oldest computers that is still used for practical purposes today
22:13:32 <wob_jonas> As for fluidic computers, there was a supposed one exhibited in the Technisches Museum Wien that I've seen in 2006,
22:14:21 <wob_jonas> namely it was a machine made of glass and water and air but with the glass never moving, powered by a water pump and its inputs controlled by some modern computer,
22:15:36 <ais523> fluid scour is a major source of mechanical damage to things, though
22:15:42 <ais523> presumably it would be less bad with a gas than a liquid
22:16:04 <wob_jonas> that they claimed was a four-bit adder, and with a display of the inputs and outputs in binary also made of those same components with the bits drawn in water (the water was dyed for better visibility),
22:16:06 <Phantom_Hoover> yes i don't see why you'd expect fluidics to be particularly long-lasting
22:16:38 <Phantom_Hoover> i mean how would overbuilt silicon microchips compare? ais523?
22:17:07 <wob_jonas> the controlling computer had an input terminal where you would input two arbitrary four-bit numbers, and the input display did seem to work correctly,
22:17:38 <ais523> Phantom_Hoover: there are a few failure modes for silicon chips, some of which are to do with temperature, others of which are to do with buildup of electrons or holes in the wrong place
22:18:04 <ais523> and some of which are mechanical and even weird, e.g. fusible link PROMs sometimes suffer from the copper in them growing crystals and linking up to other bit of copper
22:18:15 <wob_jonas> however, the controlling computer also forced to reset the inputs to all zeros after like ten seconds of feeding your input and forced to keep it at all zeros for some time, and during those approx ten seconds,
22:18:50 <wob_jonas> for a significant ratio of inputs I tried the state of the adder didn't stabilize (the gates didn't propagate fast enough or something),
22:19:24 <wob_jonas> so I actually suspect that they never managed to debug the water computer, and just exhibited the buggy one with a workaround that makes the bug plausibly deniable
22:19:41 <wob_jonas> or at least, makes them plausibly claim that it had worked at some point
22:20:50 <ais523> hmm, I wonder if it's possible to build a magnetic computer
22:20:57 <wob_jonas> it didn't seem like it had some theoretical design bug, the mechanism seemed like it could work in theory, but I think the glassware was difficult and slow to repair, and so they just couldn't get it right in enough tries and had to exhibit it to get a government grant or something.
22:20:59 <ais523> but it's unclear what the power supply would be and if any sorts of gate exist
22:21:32 <wob_jonas> I think besides the glass it also had a non-moving metal frame to support the weight of the glass.
22:22:26 <wob_jonas> How many times have I told this story of the supposed fluidic computer on this channel yet?
22:22:54 <wob_jonas> ais523: what would a magnetic computer mean?
22:23:32 <ais523> wob_jonas: something that used magnetism to send data rather than electricity
22:24:10 <wob_jonas> yep, a metal frame and also some sort of plastic valve-thingies, but the water was the only moving part, supposedly, unless they also cheated on that
22:24:27 <wob_jonas> I think the valve thingies were for maintenance and tuning
22:25:53 <ais523> I find it hard to see how you could do logic with only water as moving parts
22:26:27 <Phantom_Hoover> well, nobody understands fluid mechanics so it must be possible
22:27:24 <imode> fluidic logic gates!
22:30:04 -!- Phantom_Hoover has quit (Read error: Connection reset by peer).
22:30:42 <wob_jonas> ais523: the water had to be flowing continuously, bits represented by either water flowing fast or barely any water flowing and the tube filled with mostly air, the trick was some syphons:
22:31:25 <wob_jonas> the gate had a part where if water entered on both inputs, then the water would be continuously syphoned out from a container through a higher overflow output pipe, and barely any exited the lower output pipe, but if water came only on one input, then the syphon would never start.
22:31:54 <wob_jonas> I imagine the difficulty is that the water pressure has to be regulated very well, and that's hard to keep through multiple depths of gates.
22:32:43 <wob_jonas> And this is also why it's so plausibly deniable: it is possible that the mechanism was designed to only give correct output if it was actually "flushed" to all zero inputs on every gate between any two inputs.
22:33:10 -!- Phantom_Hoover has joined.
22:33:10 -!- Phantom_Hoover has quit (Changing host).
22:33:10 -!- Phantom_Hoover has joined.
22:33:12 <wob_jonas> So the electronic reset needn't have been arbitrary, it might have been a necessary requirement to use the computer, enforced to visitors.
22:33:14 <ais523> you can probably write a TC language with that restriction, though
22:34:09 <wob_jonas> But because of the instability of the mechanism, I can't really be sure that this is the case. It just makes the plausible cover story so much more believable.
22:34:48 <wob_jonas> I think it must have been a private grant, not a government grant. You don't need such a good cover story for a government grant.
22:34:53 -!- shikhin_ has left.
22:35:08 <wob_jonas> A grant from a private company that is.
22:36:14 <wob_jonas> Either that, or a machine built by a single eso-computer builder glassblower who donated it to the museum when he gave up on making it work.
22:37:11 <ais523> hmm, is it possible to build an amplifier with this sort of logic?
22:37:28 <imode> https://en.wikipedia.org/wiki/Fluidics#Amplifiers
22:37:29 <ais523> something which converts a slow flow to a fast flow but a zero flow to a zero flow
22:38:20 <ais523> normally, if you have an amplifier, you can make arbitrary circuits, without one you can't
22:38:25 <wob_jonas> at least if you had a reliable flow rate on inputs and power.
22:38:49 <imode> looks like fluid amplifiers exist.
22:39:15 <ais523> it looks like the logic relies on deflection, i.e. you can point a stream of water at another stream of water and it pushes it into a different location
22:39:40 <imode> yeah, kind of like a billiard ball model.
22:39:58 <imode> only with many many more particles. :P
22:40:05 <wob_jonas> ais523: the adder is a non-monotonous circuit, so it needs non-monotonous gates. aren't those practically required to contain an amplifier, in the sense that a transistor used as a unit gate in modern digital computers is an amplifier and has a power input?
22:40:40 <wob_jonas> ais523: yes, it can't be exactly what I said, because that would only make an xor gate
22:40:54 <wob_jonas> but perhaps it has an or gate as well as those
22:40:55 <imode> y'all should look at the article lol.
22:41:49 <wob_jonas> ais523: I can show the photos if you want to reverse engineer, but they're bad quality so hard to use for that
22:41:52 <ais523> wob_jonas: I don't think a half-adder (and thus a full-adder) actually needs amplification; the truth table is 00→00, 01→01, 10→01, 11→10, and in each case the output has no more 1s than the input
22:42:42 -!- shikhin has joined.
22:43:06 <Phantom_Hoover> it is honestly kind of crazy how much stuff in the game of life comes down to the r pentomino
22:43:13 <wob_jonas> and the mechanism looks like each gate is very simple physically, and seems to have only two inputs and two outputs
22:43:27 <wob_jonas> they have two inputs and three outputs
22:43:37 <ais523> Phantom_Hoover: you can think of it as a game of life version of an amplifier: small (easily constructed) input, really big and complex output
22:44:26 <wob_jonas> I think it's only two outputs, but hard to tell from my photo
22:44:26 <ais523> there aren't many useful components with two inputs and three outputs; if it does have that structure it's probably a demultiplexer, which is the only reasonable component I can think of that works like that
22:44:49 <ais523> additionally, a demultiplexer can be made into a half-adder via an OR on outputs 1 and 3
22:45:00 <ais523> (which would just involve connecting them together, fluidically)
22:45:09 <Phantom_Hoover> conway has said that he should've invented b36/s23 instead b/c it has the nice easy replicator
22:45:14 <wob_jonas> ais523: no, if there are three outputs, then the extra output must be drainage of extra water
22:45:42 <wob_jonas> ais523: the machine is visibly built of seven gates, clearly arranged in the configuration you'd need for a four-bit adder
22:45:48 <Phantom_Hoover> but the r pentomino dies out almost instantly in that, so it would never have supported all the immense variety of r-pentomino/herschel based technology that's used in life
22:45:57 <ais523> wob_jonas: hmm, so the "fluid demultiplexer"'s truth table would be 00→000, 01→001, 10→100, 11→020
22:46:15 <ais523> presumably there's some component that reshapes a 2 into a 1
22:47:26 <wob_jonas> if the water flows only down, then it's actually the wrong configuration for a full adder. it has two gates next to each other on the top, but one in the bottom. wtf.
22:48:25 <ais523> the reversible counter machine I'm working on effectively has flows made out of different numbers, and 1→2 amplifiers / 2→1 deamplifiers are basic components
22:48:43 <ais523> although it also allows 0s to move around with meaningful semantics
22:48:51 <wob_jonas> and both input pipes from the two top level gates seem to be connected to the second level gates
22:50:16 <wob_jonas> ais523: is that a TC-proof you're making over some already published reversible counter machine, or a new esolang? because that doesn't sound like a "counter machine"
22:50:52 <ais523> wob_jonas: it's a new esolang that's intended to be high-level enough to TC-prove it but low-level enough to implement in lower-level esolangs
22:51:11 <wob_jonas> although it looks like the left hand side gates differ slightly from the right hand side gates
22:51:19 <ais523> it's a counter machine in the sense that there's a direct representation of its state in finitely many counters
22:51:59 <wob_jonas> ais523: is the representation with "flows made out of different numbers" a different one?
22:52:20 <ais523> not really, the numbers effectively just move from one counter to another
22:52:43 <ais523> (which in a lower-level counter machine would involve a copy loop, but in the high-level representation it's conceptually a single step)
22:52:56 <wob_jonas> but... are there a bounded number of counters for the whole life of the supposed universal program
22:53:41 <wob_jonas> you mean some numbers close to zero are special, just like in many counter machines, and that's where you're using 1->2 amplifiers?
22:55:07 <wob_jonas> or is that not what the 1->2 amplifier means?
22:55:13 <ais523> nah, the way it works is that you have a graph which is effectively a graph of control flow, and an instruction pointer which is on a particular graph node
22:55:33 <ais523> data and control flow move in opposite directions; the instruction pointer moves around by swapping with data in adjacent graph nodes
22:55:51 -!- S_Gautam has joined.
22:56:09 <ais523> and a 1→2 amplifier would produce 2s behind where the control flow entered, but expect 1s ahead of where the control flow exited
22:56:29 <ais523> it's quite hard to get your head around the data and control flowing in opposite directions, incidentally, it keeps confusing me when I work on the language
22:57:15 <wob_jonas> and is 1 input and 2 output the only possible state of that amplifier, or is there another valid input?
22:57:52 <wob_jonas> I assumed it would take 0 input and 0 output, because that's what I thought "amplifier" means
22:58:00 <wob_jonas> but at this point I'm no longer sure because this is a crazy esolang
22:58:50 <wob_jonas> "data and control flowing in opposite directions" => isn't that also how any circular queue esolang works by the way?
22:59:05 <wob_jonas> I guess it can be easier if there's only one circular queue, not a more complex graph
23:48:31 -!- xkapastel has joined.