UmbralRaptor changed the topic of #kspacademia to: https://gist.github.com/pdn4kd/164b9b85435d87afbec0c3a7e69d3e6d | Dogs are cats. Spiders are cat interferometers. | Космизм сегодня! | Document well, for tomorrow you may get mauled by a ネコバス. | <UmbralRaptor> … one of the other grad students just compared me to nomal O_o | <ferram4> I shall beat my problems to death with an engineer.
icefire has quit [Quit: Leaving]
<soundnfury> APlayer: it's fine as an adjectival phrase, not as a verb
<soundnfury> you can say "I am deep in / deeply into [x]", or "I want to get deep into [x]", but not "I want to deep into [x]"
<UmbralRaptor> "Each session is limited to 10 participants; register on Eventbrite via the button below."
<UmbralRaptor> "Research Exchange –
<UmbralRaptor> Open to all COS faculty, postdocs, and PhD students
<UmbralRaptor> "
<UmbralRaptor> Okay.
<Iskierka> message them to say trump hasn't dismantled higher institutions yet
<UmbralRaptor> Dear Colleagues,
<UmbralRaptor> I am pleased to announce dates for the College of Science Research Exchange. These meetings are for the informal sharing of ideas and initiatives, and a way to network amongst our colleagues to discuss opportunities to work together.
<UmbralRaptor> Faculty, postdocs and PhD students performing research within our college are encouraged to attend. This is a great opportunity to explore ideas!
<UmbralRaptor> WHY WOULD YOU LIMIT THIS TO 10 STUDENTS AND REQUIRE PREREGISTRATION?!
<kmath> <iva_momcheva> Hey #GHC17 if you ❤️ CS and science ask me about the @stsci undergrad internship! Applications start Jan. 1st open to international students
<soundnfury> !wpn UmbralRaptor
* Qboid gives UmbralRaptor an invariant corpuscular [REDACTED] with an adjective attachment
* UmbralRaptor gives soundnfury a ☕ instead of ?.
APlayer has quit [Ping timeout: 186 seconds]
<bofh> "Meanwhile, according to a source close to the recent dogfight, the F-35 “turns like a garbage truck. It might be faster than the triplane, but that doesn’t matter in a stall fight.”"
<bofh> oh lol it's a parody site. oops. still, it'd hardly surprise me, especially since there's nothing on a Fokker triplane for the F-35 to get a missile lock onto.
<soundnfury> bofh: I was gonna say "I think there are many layers to this story"
<soundnfury> "try peeling a few of them away"
<UmbralRaptor> duffleblog is a military version of the onion.
<Fiora> it's obviously parody
<Fiora> the real f-35 probably would be stuck on the runway
<bofh> LOL
<kmath> <johnregehr> @oshepherd @shafikyaghmour @mcclure111 obviously optimizing for a processor that executed semicolons, newlines, and curly braces slowly
<soundnfury> the real giveaway was when they claimed the dr-1 could climb faster. I mean, I know the f-35's no EE P.1B, but still
<soundnfury> (now that's a far more reasonable comparison to make. English Electric Lightning vs. LockMart Lightning II)
<bofh> UmbralRaptor: well pretty much... I mean putting that all on one line is okay, but fuck's sake, use an if, curly braces and semicolons.
<Fiora> i mean okay but a monkey can also climb faster than an f-35 that's grounded
APlayer has joined #kspacademia
<awang> APlayer: ferram4: Iskierka: That was a very illuminating and confusing conversation :D
<egg|phone|egg> !u ☕?
<Qboid> U+2615 HOT BEVERAGE (☕)
<Qboid> U+1F375 TEACUP WITHOUT HANDLE (?)
<egg|phone|egg> !Wpn UmbralRaptor
* Qboid gives UmbralRaptor an oxidizing git
<egg|zzz|egg> (via whitequark)
<egg|zzz|egg> <APlayer> But why add a spike and say "it does not significantly worsen things" if you could simpky nit have added it?
<egg|zzz|egg> <Iskierka> although I'm unsure if it has another functional purpose here. I don't think the vent needs to be a spike
<egg|zzz|egg> <Iskierka> ^ because aerodynamics is consistent.
<egg|zzz|egg> <APlayer> So it basically abuses the air to change its aerodynamic shape depending on speed?
<egg|zzz|egg> uuuuh wat
* egg|zzz|egg stabs clipboard
<kmath> <NAICobservatory> We are continuing to check our infrastructure as the flooding under the dish recedes. @robminchin and @HCMinchin ka… https://t.co/vlD0X1oOze
* egg|zzz|egg stabs hexchat for copying whatever is selected, too
<egg|zzz|egg> stabitty
<APlayer> egg: "I am not allowed to see these tweets"
<APlayer> !wpn people
* Qboid gives people a francium tofu
<APlayer> That sounds double disgusting
<APlayer> Also, I only now realized you didn't mean to paste the whole conversation, not just the last message
ferram4 has quit [Ping timeout: 383 seconds]
<Iskierka> egg|zzz|egg, there's a way to turn off auto-highlight-copy
<Iskierka> preferences -> chatting -> advanced -> auto-copy behaviour
<egg|zzz|egg> ah, nice
<Iskierka> after that I'm not sure if the copy shortcut is default or I added it but for me copying the window is shift+ctrl+c. Since it lets you highlight in both and the devs don't think you might have highlighted intentionally and know which it will copy if it was just one command on most recent
icefire has joined #kspacademia
awang has quit [Ping timeout: 186 seconds]
kmath has quit [Remote host closed the connection]
kmath has joined #kspacademia
egg|cell|egg has joined #kspacademia
egg|phone|egg has quit [Ping timeout: 204 seconds]
BPlayer has joined #kspacademia
egg|phone|egg has joined #kspacademia
egg|cell|egg has quit [Read error: Connection reset by peer]
egg|cell|egg has joined #kspacademia
egg|phone|egg has quit [Read error: Connection reset by peer]
egg|zzz|egg is now known as egg
<egg> whitequark: any котяnews (or котяpics)?
<egg> !wpn whitequark
* Qboid gives whitequark a Pauli repeater/lemma hybrid
<egg> !wpn UmbralRaptor
* Qboid gives UmbralRaptor an ill-conditioned isomorphism
<egg> !wpn Iskierka
* Qboid gives Iskierka a finite float iBod
<UmbralRaptor> !wpn egg
* Qboid gives egg a catadioptric falcon
<UmbralRaptor> Flying Kitty!
BPlayer has quit [Ping timeout: 383 seconds]
BPlayer has joined #kspacademia
<kmath> <3PSboyd> Black cats https://t.co/ne6qly3v4w
<BPlayer> Uggh
<BPlayer> Still can't get that RK4 thing to work
<BPlayer> (Yes, that one from back when I first joined here)
<BPlayer> This is correct, isn't it? k1 = f(t, Y); k2 = f(t + c2*h, Y + h*(k1*a31 + k2*a32)); k3 = f(t + c3*h, Y + h*(k1*a31 + k2*a32)); k4 = f(t + c4*h, Y + h*(k1*a41 + k2*a42 + k3*a43))
<BPlayer> (Or let me make a pastebin of that, it's cleaner
<BPlayer> )
<BPlayer> There we go: https://pastebin.com/93K6Y6L4
<egg> no your indices on your as look wrong?
<BPlayer> How?
<egg> k2 = f(t + c2*h, Y + h*(k1*a31 + k2*a32))
<egg> <<<< k2, so I should be seeing a2js here
<BPlayer> Ah, k2
<egg> also you might want to just write a couple of loops so you don't screw that up
<BPlayer> Sorry, that's a pasting / pseudo-coding error. k2 = f(t + c2 * h, Y + h*(k1*a21))
<egg> ok, then that looks correct?
<BPlayer> Humm, let me see what else I may have screwed up
<BPlayer> Okay, mistake #1 was I did Y = k1*b1 + k2*b2 + k3*b3 + k4*b4
<Qboid> [#1] title: Quantity experiment | This seems useful. It might well lead to a switch to C++.... | https://github.com/mockingbirdnest/principia/issues/1
<BPlayer> Uh what
<BPlayer> So Qboid is activated every time I use a #-number now?
<egg> yup, that's not new
<egg> #1453
<Qboid> [#1453] title: Plotting frame changes to body centered inertial when using Revert to Launch | Cauchy introduced unification of speed display and plotting frames and as such is supposed to switch to body centered body fixed frame when in atmosphere, on ground or at launch time. This works most ... | https://github.com/mockingbirdnest/principia/issues/1453
<BPlayer> But that doesn't seem good. I mean, I can totally see people using # in their messages :P
<egg> yes, to refer to principia pull requests :-p
<BPlayer> Could instead do something like i#-number
<BPlayer> Also, while I am at it. Would the burn time of a maneuver not be (1 - e^(-(dv / ve))) / fuelUsage where dv is the delta-v, ve is the exhaust velocity and fuelUsage is in kg/s?
<BPlayer> Taking the Tsiolkovski Equation in the form of m * e^-(dv / ve) = (m - fuelUsage * burnTime) and solving for burnTime
<BPlayer> Wait, no, I definitely screwed that up too
APlayer has quit [Read error: Connection reset by peer]
APlayer has joined #kspacademia
ferram4 has joined #kspacademia
BPlayer has quit [Ping timeout: 383 seconds]
BPlayer has joined #kspacademia
<egg> bofh: uuuh, doing my (3 months late) semesterly Bulletin C pull request, the earth orientation products got *very heavily* modified
<egg> bofh: 2704013a162d0f06e2b3b7707024d43e23fa690d
<egg> uh
<egg> how does Qboid work again
<Qboid> [27040] title: update EOP headers. by Robin Leroy | Additions: 12960 | Deletions: 12686 | https://github.com/eggrobin/Principia/commit/2704013a162d0f06e2b3b7707024d43e23fa690d
CPlayer has joined #kspacademia
APlayer has quit [Read error: Connection reset by peer]
<bofh> egg: oh fucking joy
<egg> EOPC04Entry(1984'01'01, 0.3957[-2-]{+4+}1[-8-]{+4+} * Second),
<egg> and basically 12 kilolines of that
<egg> same for the C02 entries too
<egg> they even added entries to C02? I'm confused
<egg> what's mjd 57853
<egg> oh right C02 does go up to the modern day?
<egg> huh it's no longer experimental?
<egg> bofh: I have to applaud their format stability though, because apparently they went out of experimental for C02 and significantly changed the data, but my shitty awk script still works :-p
<soundnfury> \o/ awk \o/
<soundnfury> !wpn egg
* Qboid gives egg a Yukawa array
<egg> bofh: haha the test no longer compiles because UTC - UT1 changed (and this was checked by static_assert of course,
<egg> )
<egg> (UTC - UT1 at J2000 UTC that is)
awang has joined #kspacademia
<egg> bofh: I wonder whether I should also use BIPM TT instead of assuming that TT runs parallel to TAI,
<egg> bofh: just so I can have more horrible text files from which to generate humongous header files,
<bofh> BIPM?
<Qboid> bofh: [BIPM] => Bureau International des Poids et Mesures
<bofh> so how about no :P
<egg> bofh: why
<bofh> (this reminds me of fitting high-order polynomials to ephemerides to avoid having humongous header files in a doppler correction utility, rofl)
<bofh> b/c large header files? well okay I guess .rodata is effectively free
<egg> bofh: well look at the size of eop_c04.generated.h
<egg> bofh: gotta do that UT1 to TAI conversion at compile time,
icefire has quit [Quit: Leaving]
<bofh> I am now curious if eop_c04 has a nice analytic approximation (I swear it should but I'm sadly blanking on it atm)
<egg> bofh: wait "EOP C 04 is no longer being updated and is presented here for historical reasons only. See EOP 05 C04 for the new version of this series.
<egg> "
<egg> aaaaa
<egg> *sob*
<bofh> LOL
<bofh> rip
<egg> bofh: wait no, I was looking at the wrong EOP C 04 metadata, here I'm seeing EOP 14 C04 is updated two times per week.
<egg> bofh: and the last entry is from 2017-09-05 https://datacenter.iers.org/eop/-/somos/5Rgv/getTX/223/eopc04_14.62-now
<egg> so everything is fine?
<egg> (like, except that all the data changed)
<BPlayer> Okay, let me try to rubber duck debug here, with the added benefit that someone could interject and tell me if I do things wrong.
<BPlayer> For that RK4 method, I have a function f(t, Y). It gets the first part of Y for the position and the second for velocity. Then it proceeds to calculate the acceleration vector, combines velocity and acceleration and returns that.
<egg> bofh: So, 2000-01-01T00:00:00 UTC was formerly estimated to be 2000-01-01T00:00:00,3555094 but the current estimation (following EOP C04) is 2000-01-01T00:00:00,0.3554730
<egg> bofh: who retroactively kicked the Earth,
<BPlayer> The only thing I may have screwed up here is the acceleration calculation. It takes the parent body to be at the origin of a 2D euclidean plane. Then it calculates engine acceleration and gravity acceleration, and adds those vectors fo the total acceleration.
<BPlayer> Since all thrust is assumed to be tangential, I use the normalized velocity vector and multiply that by the trust in newtons. Then I divide that by (initialMass - fuelUsage * t) to get the engine acceleration
<BPlayer> For the gravity I take the normalized position vector and invert it to get the normalized gravity vector. I multiply that by the body's mu and divide by the squared magnitude of the position vector
<BPlayer> Are those formulas correct?
<BPlayer> That is: norm(velocity) * thrust / (mass - fuelUsage * t) + -1 * norm(position) * (bodyMU / |position| ^ 2)
<BPlayer> Where norm() normalizes a vector
<BPlayer> Should the k-parameters be generally similar?
<BPlayer> I mean k1 - k4
<BPlayer> Because they most definitely are not
<egg> !wpn bofh
* Qboid gives bofh a field effect singularity
<egg> !wpn UmbralRaptor
* Qboid gives UmbralRaptor a silicon COME FROM
<BPlayer> Also, after the first iteration, my velocity y-parameter, which is nearly completely determined by gravity at that point, varies from 0 at k1 to around -42 at k4. The time step is 10 seconds, and this number seems pretty big given I am in a simulated 500-km circular earth orbit
<BPlayer> Or is that okay?
<bofh> egg: huh. that's... I wasn't expecting the correction to also be in the negative direction. ugh, that seems like a pain to handle.
<egg> bofh: what do you mean
BPlayer has quit [Ping timeout: 383 seconds]
<egg> bofh: TT(BIPM)! ftp://ftp2.bipm.org/pub/tai/ttbipm/TTBIPM.16 clearly I should have that in a header,
<bofh> egg: like the current estimation is < the previous estimation
<bofh> which is odd.
<egg> bofh: well why would that be odd
<egg> we have a better estimate of the rotation of the earth at that time, why does it have to be in one direction or the other
<bofh> okay, I suppose I'm being silly.
<bofh> point.
<egg> bofh: (there was a typo in that date above of course, that's 2000-01-01T00:00:00,3554730 not ,0. >_>
<egg> )
<egg> bofh: most interesting pull request :-p #1587
<Qboid> [#1587] title: Bulletin C 54 | See Bulletin C 54:... | https://github.com/mockingbirdnest/principia/issues/1587
<egg> (also very belated)
<UmbralRaptor> !choose stab Lagrangian mechanics | stab aplayer 's problem.
<Qboid> UmbralRaptor: Your options are: stab Lagrangian mechanics, stab aplayer 's problem.. My choice: stab aplayer 's problem.
<egg> bofh: but way more interesting than #1198 at least
<Qboid> [#1198] title: Bulletin C 53 | See Bulletin C 53:... | https://github.com/mockingbirdnest/principia/issues/1198
<egg> bofh: now #1089, on the other hand,
<Qboid> [#1089] title: Bulletin C 52 | Thanks to @pdn4kd for the notification.... | https://github.com/mockingbirdnest/principia/issues/1089
<egg> (yes I dual-source my reference to bulletin C)
egg is now known as egg|nomz|egg
icefire has joined #kspacademia
<bofh> rofl.
<bofh> that's sensible tho.
<egg|nomz|egg> bofh: well I might want to automate this but then again
<egg|nomz|egg> ~work~
* CPlayer 's problem stabs back at UmbralRaptor
CPlayer is now known as APlayer
* UmbralRaptor is stabbed with JavaScript.
<APlayer> Indeed
<APlayer> I may have a theory, though
<APlayer> My output was nonsensical while my forst few iterations of line-by-line debugging looked fairly nively
<APlayer> However, I sample my output every 100 iterations
<APlayer> I may have misinterpreted it as nonsensical while actually I just went round the planet significantly enough
<APlayer> Well, yes, of course
<APlayer> ;c 1000 / 60
<kmath> APlayer: 16.666666666666668
<APlayer> I have output every 16 minutes
TonyC1 has joined #kspacademia
BPlayer has joined #kspacademia
<BPlayer> Well, no. The data still makes no sense
* APlayer accidentally started his simulation witha time step of 0 and wondered why it didn't end for so long
<APlayer> ...and the numbers are way off
<APlayer> That is the numbers are way off from what they should be if the time is still 0
<APlayer> ...which leads to the discovery of another mistake, for I forgot to multiply the change of Y by h
<APlayer> Anyway, what I see makes much more sense already, but it's not yet correct. And that's enough development for today.
BPlayer has quit [Ping timeout: 204 seconds]
<kmath> <McNutcase> Just in case you need these. https://t.co/MmoNrbIPgf
<UmbralRaptor> APlayer: uh, I think you want velocity + the acceleration terms * t, not v*a*t?
<APlayer> Where do I want that?
<UmbralRaptor> APlayer: yeah, if the Ks differ enormously, something is bad/wrong. Either implementation, or timestep size
<APlayer> Well, I fixed them more or less. Seems similar now
<APlayer> Also reduced the time step
<UmbralRaptor> <BPlayer> That is: norm(velocity) * thrust / (mass - fuelUsage * t) + -1 * norm(position) * (bodyMU / |position| ^ 2)
<APlayer> (Actuall accidentally removed the time step once :P)
<UmbralRaptor> oops
<UmbralRaptor> Then again, I might be misunderstanding the context?
<APlayer> Well, that should return the acceleration only
<awang> APlayer: What are you working on?
<APlayer> And the acceleration is obtained by getting a normalized direction vector first and then multiplying it by the magnitude
<UmbralRaptor> awang: simulations of long duration burns.
<APlayer> ^
<APlayer> Ninja'd
<awang> Ah, I see
<awang> Steal some code from egg, maybe? Thought Principia was capable of something like that
<APlayer> Basically what I did
<APlayer> But I was more difficult to handle than just "take this code and get working" :P
<awang> Well, not entirely surprising, I guess
<UmbralRaptor> Yeah, but C is a write only language.
<APlayer> I estimate people spent a total of a few man hours to explain things to me
<awang> Guessing the code isn't intended to stand on its own or anything
<awang> I thought Perl was the write only language
<soundnfury> APlayer: if you don't mind the world's shittiest integrator, you may also find reading konrad's burn-sim code useful
<soundnfury> that's Python, so it's perhaps more readable
* soundnfury gives UmbralRaptor a snek
<UmbralRaptor> Perl *can* be.
<APlayer> Anyway I am working with (wait for it) Javascript
<UmbralRaptor> Sneks!
<awang> egg|nomz|egg: ^Banhammer plz
<egg|nomz|egg> UmbralRaptor: Principia is emphatically not written in C
<APlayer> Well, that's not the worst. There are neither-read-nor-write-nor-software-analyse languages
<APlayer> See Malbolge
<egg|nomz|egg> (I guess if I started now, or a bit later (still need genericity on values to be happy), I might have used rust, but whitequark tells me back when I started it wouldn't have been up to the task (and I hadn't heard of it either))
<awang> I think esoteric languages are more or less a freebie for "write-only"
<egg|nomz|egg> also it taught me one of the languages I use at work, so I've got that going for me, which is nice
<APlayer> UmbralRaptor: Anyway, did that explanation of the formula make sense? Am I even doing it correctly?
<awang> egg|nomz|egg: Who needs generics? Use Go :P
* egg|nomz|egg slaps awang with the quantities and geometry library
<APlayer> Who even needs programmers and programming languages? Use advanced AI!
<egg|nomz|egg> also with the compile-time date parsing :D
<APlayer> I've recently spent some time pondering if there will be a time when human labour of any sort becomes obsolete because AI will handle it
<APlayer> I came to the conclusion that the answer probably is yes, and not too far off in the future
<awang> compile-time *all* the things!
<APlayer> When do you think we will have AI? Like what, 20 years? And it will immediately be used to manage some basic tasks for humans. And then it's not too far off to have basically everything done by AI
<awang> APlayer: Define "have AI"
<APlayer> AI as in something that can interpret human langauge, understand and process it, respond and take actions as told
<awang> Siri/Cortana/etc. arguably already do that
<APlayer> Not really
<APlayer> That's no intelligence
<APlayer> Although whatever it is, it is artificial
<awang> Uh
<awang> Doesn't that make "artifical intelligence"?
<APlayer> Well, I probably forgot things like learning, perhaps self-replicating and most of all autonomous actions
<APlayer> AI for me is a thing that is capable of /doing things/.
<APlayer> Like doing things on its own. Not showing your calendar entries when told
<awang> I'd imagine those assistants are technically capable of such things. There's just no demand for them.
<APlayer> Also, I believe - markov chains and all - we still have no kond of program with which you could have a conversation that makes the least bit of sense
<APlayer> Are you kidding, no demand for them? You have no demand for something that would do work for you on its own?
<APlayer> I'd gladly take your share, if so :P
<egg|nomz|egg> !wpn bofh
* Qboid gives bofh a metal oxide sodium distribution
<egg|nomz|egg> !wpn UmbralRaptor
* Qboid gives UmbralRaptor a discrete heisenbug
<egg|nomz|egg> !wpn awang
* Qboid gives awang a Ritchey-Chrétien crystal
<egg|nomz|egg> hmm
<APlayer> !wpn egg
* Qboid gives egg an allotropic bird
<awang> I think "work for you" might be a bit too varied
<awang> Guess it depends on what you mean by "doing things"
* egg|nomz|egg has ievan polkka stuck in his head
* egg|nomz|egg pokes UmbralRaptor with a leek
<APlayer> awang: It's hard to define just like that
<APlayer> I probably can't properly describe what I mean
* UmbralRaptor spins
<APlayer> Say, a scenario. You have to cook something and forgot to set the timer. An AI would remind you, although not specifically programmed to do this
<awang> APlayer: That would certainly be interesting to implement, given the amount of context that needs to be figured out
<APlayer> Remind you because a logic chain triggered it: If you did not set the timer, you won't know when to stop cooking. If you won't knkw when to stop cooking, you will probably burn your food
<awang> Well
<APlayer> That requires observing and noticing that you forgot to put up the timer too
<APlayer> But let's put that aside
<awang> Could also be easier if you asked for help with a recipie
<awang> But if you didn't say anything, that'd be an interesting problem
<awang> s/recipie/recipe
<Qboid> awang meant to say: Could also be easier if you asked for help with a recipe
<APlayer> Well, querying for a recipe won't be hard fkr any siri
<egg|nomz|egg> UmbralRaptor: SpinRaptor?
<APlayer> However, what an AI would do is not just query, but find a suitable one based in time, difficulty and content
* egg|nomz|egg measures UmbralRaptor's spin along the Earth's axis
<APlayer> And not simply comoaring the tables, but actually making sense of what it means
<APlayer> Test
<APlayer> Uh my message got swallowed
<APlayer> Anyway, I said that that was the kind of AI I expected in about 20 years
<awang> If anything, it sounds almost more like it's a matter of someone taking the time to do the training with existing tech rather than waiting for some new tech
<APlayer> I don't think that's how existing tech works
<UmbralRaptor> egg|nomz|egg: leek-bar/2
<egg|nomz|egg> aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa
egg|nomz|egg is now known as egg
<APlayer> The main sorts of AI (Evolutionary algorithms, graph searches, ANNs, markov chains) all need some sort of clear objective
<APlayer> Once you have defined it, they can evaluate /numbers/ and give you the best match or something
<APlayer> That's, IMHO, not really AI
<APlayer> That's logic, even if arriving at the solution requires complex operations
<APlayer> Intelligence is more than logic
<APlayer> It's a mix of creativity, initiative, logic and probably much more
<APlayer> Logic is easy to achieve woth current tech indeed. Initiative is... hard but probably possible. But I don't think machines can be vreative
<APlayer> They make exactly what they are told to do and no less no more.
<awang> This sounds like it's veering into the realm of philosophy :P
<APlayer> It is
<APlayer> But firstly, I am thinking out loud and secondly, I think I managed to define pretty well what I mean by AI
<APlayer> So I call it a definition. Philosophy was trying to arrive at it :P
<awang> And you've managed to lose me along the way :P
<awang> That's what I get for dabbling in areas with which I am not familiar
<APlayer> I am not really familiar woth how AI works either
<awang> A heck of a lot more familiar than I am, at least
<APlayer> I know a few of the main algorithm's principles, but I could never make any functioning AI out of that
<APlayer> Nor implement most of them
<APlayer> Don't try to learn from what I said, though. Chnaces are I just said some total nonsense and have no idea what I am talking about
<awang> Well, this is #kspacademia after all
<awang> I'm sure if that's the case someone will be here momentarily to smack some sense into us
<APlayer> I am thinking out loud, as I said
* egg summons e_14159
<APlayer> Anyway, I think creativity can be emulated using evolutionary algorithms
<APlayer> But that seems not good enough to be actually creative. It would not think "outside the box"
<APlayer> And it's slow
<APlayer> Also, awang: In case the algorithms are where you couldn't follow. In a nutshell they are all ways to brute force but lessening the chance of ever trying a wrong solution
<awang> egg: is e_14159 a ML guy?
<APlayer> And they are mostly good at it
<egg> awang: yeah
<Qboid> 2d 0h 0m 0s left to event #11: 「みちびき4号機」/H-IIAロケット36号機 [at 2017-10-09 22:00:00]. Say '!kountdown 11' for details
* awang braces for the righteous fury of a master whose domain has been violated
<APlayer> ML?
<Qboid> APlayer: [ML] => Machine learning
<APlayer> Ah
<egg> 「みちびき4号機」!
<APlayer> Heck, I never realized what people are lurking around here
<APlayer> Some multi-talents, an aerodynamics master, an EE master, an ML master, a master of classical mechanics, all of them with extensive knowledge of maths, physics, and programming
<APlayer> I wonder what could be done if all of them were assembled into a team and give a task where eberyone is optimally performing
<egg> we have 2 aerodynamics people I think?
<APlayer> Also, I forgot to mention the quantum master
<egg> with Iskierka and ferram4
<APlayer> I considered Iskierka the multi talent. So far she was able to pitch in at every question of mine
<egg> wait who actually specializes in QM here? bofh might be competent there, but I don't recall him talking about it much
<Iskierka> that would probably be more accurate
<Iskierka> just don't ask about maths or specialised QM
<APlayer> w.hitequark?
<egg> I don't recall whitequark talking about QM here?
<egg> maybe I was asleep tho
<APlayer> His twitter says "Working on Quatum Computers for a living"
<egg> wrong pronoun
<APlayer> Sorry?
<APlayer> Which pronoun?
<egg> "his", should be "their" or "her"
<egg> (depending on whether you go with the diapsid or synapsid profile)
* APlayer slaps himself for assuming gender
<APlayer> Ah, not to forget the dentist master
<egg> back to the subjcet at hand, yes, but afaict it's software infrastructure for research in that field? http://m-labs.hk/index.html
<APlayer> :P
<egg> rather than directly QM research
<APlayer> Okay then
<APlayer> Scrap the QM master then, indeed renaming her to dentist master :P
<APlayer> No, seriously, the combination of people here is interesting
<APlayer> I do wonder what could be achieved or built solely by people from this channel, given appropriate funding
<UmbralRaptor> Funding, hah. =\
<egg> UmbralRaptor: clearly we need to make ANBO a thing,
<UmbralRaptor> Yeah
<APlayer> Found an institute, go an make research and develop something :P
<APlayer> And how many PhDs and/or Professors are here?
<egg> uh, bofh
<egg> (possibly, I don't actually know)
<APlayer> That could be interpreted as a pun on "both"
<APlayer> :P
<UmbralRaptor> Right, egg "only" has an MSc (ignoring that those mean something in EU systems)
<UmbralRaptor> ferram4 might "only" have a BSc, though FAR is at least a master's level project.
<egg> UmbralRaptor: well if we count MScs then e_14159 too at least (PhD student), probably others
<egg> BScs, I'd assume quite a few?
<egg> (all of those qualified with "or system-dependent equivalents" of course)
<APlayer> Uh, I always thought Master's degrees are basically a requirement to get some job beyond truck driving?
<UmbralRaptor> Heh, probably lots.
<egg> APlayer: in the EU, kinda, the US differs I think
<APlayer> No, seriously, at least where I live those are not really something rare
<UmbralRaptor> APlayer: not *yet* (also jobs driving trucks at more skilled than you think)
<egg> APlayer: yeah same here slightly south of you :-p
<APlayer> But they need no MScs and are considered low qualification
<egg> UmbralRaptor: http://abstrusegoose.com/418
<APlayer> But PhDs are like one out of every few hundred
<APlayer> Thats a nice thing to put into your resume, definitely :P
<UmbralRaptor> Certs, and lots of blue collar jobs are in the "entry level, 2 years experience required" category.
<APlayer> While we are talking about experience and jobs
<UmbralRaptor> egg: hah =(
<egg> !wpn UmbralRaptor
* Qboid gives UmbralRaptor a quartz frustum-like invariant
<APlayer> Ive been planning to spend the next summer vacations in the US, spending few weeks on a foreign student internship in the realms of physics, maths and/or IT. If anyone has contacts or heard something recently, it'd be absolutely amazing if you could give me some pointers! :D
<egg> UmbralRaptor: #1587
<Qboid> [#1587] title: Bulletin C 54 | See Bulletin C 54:... | https://github.com/mockingbirdnest/principia/issues/1587
<kmath> <johnregehr> all the tired people in the house found me https://t.co/EZXssQeUbD
<APlayer> I want to get some experience and knowledge to eventually apply for the ultimate internship at some large company like Google or SpaceX. If I succeed, that'd be the killer
<UmbralRaptor> egg: Time is confusing.
<egg> UmbralRaptor: I realized the ~proper~ thing to do would be to interpret TAI-based dates using TT(BIPM)
<UmbralRaptor> APlayer: hrm. I wonder if my profs are doing REUs?
<egg> that would be ~yet another header generated from giant tables~ tho
<APlayer> REUs?
<egg> UmbralRaptor: ftp://ftp2.bipm.org/pub/tai/ttbipm/TTBIPM.16
<egg> also it's a few μs
<UmbralRaptor> egg: define giant?
<egg> UmbralRaptor: click that link?
<egg> that's not huge in term of data, but as a header file it'd be silly
<egg> then again we have C 04
<egg> but because Principia does no GR we identify TDB and TT, so we get ~2 ms periodic errors, so the microseconds between TAI TT(BIPM) can probably safely be ignored
<egg> EAL?
<Qboid> egg: [EAL] => Échelle Atomique Libre
<egg> :D
<egg> TAI?
<Qboid> egg: [TAI] => Temps Atomique International
<egg> TT?
<Qboid> egg: [TT] => Temps Terrestre
<egg> BIPM?
<Qboid> egg: [BIPM] => Bureau International des Poids et Mesures
<UmbralRaptor> The server at ftp2.bipm.org is taking too long to respond.
<APlayer> UmbralRaptor: Ah, looked it up
<egg> !acr -add:PTTI Precise Time and Time Interval
<Qboid> egg: I added the explanation for this acronym.
<APlayer> Okay, a REU seems like a nice start
<UmbralRaptor> APlayer: hopefully. I have no idea how things work for German citizens.
<APlayer> UmbralRaptor: Thanks for the tip! That'd definitely be something I am interested in, so I will have a look.
<UmbralRaptor> Do the various Max Planck institutes do student internships?
<APlayer> The main issue I see is that I need someone US based to take responsibility or something for me to even be eligible to receive a visum (is that how you call it in English?)
<APlayer> So I'll have to find a company or university tajt would do this for me
<UmbralRaptor> egg: yeah that's slightly silly. But have you seen FITS files?
<egg> APlayer: sadly english speakers, like french speakers, cannot into grammar, and use visa in the singular
<UmbralRaptor> visum?
<UmbralRaptor> Oh, visa
<egg> UmbralRaptor: a visum, two visa
<egg> like templum
<APlayer> So what's the correct term?
<egg> UmbralRaptor: german likes its latin plurals
<egg> APlayer: visa, in english
<APlayer> Okay, then
<egg> UmbralRaptor: see also Limes, Limites for limit
<APlayer> Will make a note to self
<egg> UmbralRaptor: (in the mathematical sense)
<APlayer> See Kompass - Kompanten, Atlas - Atlanten
<APlayer> There is a bunch of those
<UmbralRaptor> 1 limes, 2 limites?
<UmbralRaptor> mmm… lime…
<APlayer> 1 lime, 2 limes, n limites
<APlayer> The Ackermann pluralization function
<APlayer> :P
<APlayer> That sounds almost like a !wpn thing
<egg> UmbralRaptor: also Volumen, Volumina
<UmbralRaptor> Principum, Principia?
<APlayer> Limes / Genitiv... what a ridiculous concept
<UmbralRaptor> <_<
<egg> UmbralRaptor: princeps
<APlayer> UmbralRaptor: Prinzip - Prinzipien
<egg> wait but principia is the adjective in the neuter, and the noun (masculine) pluralizes as principes?
<egg> what is newton's title actually saying
<APlayer> egg: And now find the plural for "Apfelmus"
<APlayer> :P
<APlayer> Anyway, I've got to go for today
<APlayer> I've already stayed up longer than I should have
<APlayer> See you!
<APlayer> And thanks for all your help and the nice conversations! :-)
<egg> clearly it should be Apfelmures, https://en.wiktionary.org/wiki/mus#Latin
<egg> :-p
<UmbralRaptor> \o
<egg> \\o
<egg> UmbralRaptor: ah wait no it's principium, not princeps https://en.wiktionary.org/wiki/principium
<egg> !wpn UmbralRaptor
* Qboid gives UmbralRaptor a gold theatre
<egg> !wpn bofh
* Qboid gives bofh an isothermic protractor which vaguely resembles a brioche
<egg> UmbralRaptor: TT-TAI is fun, https://hastebin.com/emigikowid.txt
<egg> EAL does drift quite a bit from TT tho, nearly a millisecond off
* UmbralRaptor ? things, and submits a forgotten (late) AAS abstract.
<egg> yay abstract
<egg> UmbralRaptor: do you list ANBO in your affiliations,
<UmbralRaptor> egg: that seems a bit silly.
<egg> Raptor, U.
<egg> * Anglo-Norman Bovine Observatory
<egg> UmbralRaptor: no sillier than F.D.C. Willard
<UmbralRaptor> Kitty!
<UmbralRaptor> STScI?
<Qboid> UmbralRaptor: [STScI] => Space Telescope Science Institute
<UmbralRaptor> eggcelent
* UmbralRaptor keeps on wanting to spell it "STSciI" despite people enounciating every letter.
<bofh> "nearly a millisecond off" what's the measurement expected error?
<egg> bofh: dunno, looks like drift to me? ftp://ftp2.bipm.org/pub/tai/ttbipm/TTBIPM.16
<egg> bofh: ah
<egg> "In the 1970s, it became clear that the clocks participating in TAI were ticking at different rates due to gravitational time dilation, and the combined TAI scale therefore corresponded to an average of the altitudes of the various clocks. Starting from Julian Date 2443144.5 (1 January 1977 00:00:00), corrections were applied to the output of all participating clocks, so that TAI would correspond to proper time at mean sea
<egg> level (the geoid). Because the clocks had been on average well above sea level, this meant that TAI slowed down, by about one part in a trillion. The former uncorrected time scale continues to be published, under the name EAL (Echelle Atomique Libre, meaning Free Atomic Scale).[7]:215
<egg> "
<egg> bofh: so indeed, EAL drifts wrt TAI/TT
<egg> bofh: the error between TAI and TT as per TT(BIPM) is much better behaved
<egg> bofh: see again that TTBIPM table
<egg> !wpn UmbralRaptor
* Qboid gives UmbralRaptor a transpose feather
<egg> !wpn bofh
* Qboid gives bofh a caffeinated projection
<egg> !wpn Fiora
* Qboid gives Fiora a commutative gazebo which vaguely resembles a prime