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.
<egg|df|egg> nay, several children have been fighting with several racoons
<hattivat> well, then bind them in whatever is light
<hattivat> native aluminium maybe?
<egg|df|egg> I don't think I have any metals other than gold actually >_>
<hattivat> set up an import-export company then? ;p
<hattivat> I imagine gold is worth quite a bit of 'native aluminium'
<egg|df|egg> I could use purple spinels
<soundnfury> hattivat: so long as he doesn't go down the mercantilist road
<soundnfury> which, if he's contemplating exporting gold, he doesn't seem to be :)
<hattivat> haha, right
<hattivat> google says that purple spinel is some neat-looking gem
<hattivat> I says go for it, assuming it's not too heavy
<egg|df|egg> O_o somehow not an option, even though I have bindings made of gems
<egg|df|egg> UmbralRaptor: huh, the current mayor of starwatch of oxen is a goblin
e_14159 has quit [Ping timeout: 204 seconds]
e_14159 has joined #kspacademia
<UmbralRaptor> egg|df|egg: Are you promoting Nomal?
<UmbralRaptor> hattivat: ...native aluminum? How would that even happen?
<UmbralRaptor> (Reducing atmosphere?)
<hattivat> UmbralRaptor: No idea, but egg mentioned it as one of the bindings he has seen on imported books
<UmbralRaptor> Huh, so it *is* a thing. And I guessed correctly about the environment. >_>
<Ellied> is 'native' archaic speak for 'elemental'? as in, pure aluminum, not an alloy?
<Ellied> or compound for that matter
<UmbralRaptor> Yes.
<UmbralRaptor> I think.
<Ellied> also, if dorf fort takes place in the sort of time period it makes sense for it to, gold might actually be significantly less valuable than aluminum.
<Ellied> otoh, the real world of those times didn't have dwarves.
<Ellied> wait, actually, I think aluminum didn't get cheap until electrolysis became a thing. Before electricity, purifying it from bauxite was really hard. I bet it's valuable even with dwarves.
<hattivat> that is assuming they find it useful
<hattivat> price is not only a function of rarity
<Ellied> but who *wouldn't* find aluminum useful? its strength to weight ratio is fantastic.
<hattivat> you have a point
<hattivat> but it kind of assumes advanced metallurgy
<Ellied> this is dwarves we're talking about here. Isn't that kind of their thing?
<hattivat> I don't think a solid lump of aluminium is much more useful to a dwarf than a similar lump of iron
<hattivat> I wouldn't know, I never got much beyond the opening in df
<Ellied> Never played it in my life, I'm extrapolating from what I've heard egg say and Tolkein.
<UmbralRaptor> Also, the whole way aluminum doesn't do iron style rusting/flaking.
<egg|df|egg> UmbralRaptor: yeah, but then again it seems hard to work with with the techniques you have for iron
<soundnfury> hattivat: iirc before they developed the electrolytic method, rich gits used aluminium cutlery
<egg|df|egg> my miners aren't mining
<egg|df|egg> they're reading about opticks
<soundnfury> I guess because it was _more_ ostentatiously expensive than silver service vOv
<Ellied> true, if you jammed it in a forge for iron, it would just catch on fire.
<egg|df|egg> than gold even, see the DF wiki link
<hattivat> soundnfury: very interesting, thanks for this nugget!
<UmbralRaptor> egg|df|egg: to be fair, your dwarves probably get grants for materials. >_>
<hattivat> hahahah
<soundnfury> hattivat: ah but is it a native aluminium nugget, or just that crappy gold?
<hattivat> ...native aluminium, I think
<egg|df|egg> also they tried making coins out of aluminum
<hattivat> they still do in Indonesia
<egg|df|egg> yeah but that was back then they thought the material of the coin was its value
<hattivat> feels cheep, flimsy, and lacks that satisfying weight to it
<egg|df|egg> so they made it out of *aluminium* because that was so costly :-p
<hattivat> I get it, but it sounds impractical
<hattivat> I don't you can check aluminium's authenticity by biting it or scratching it on a magical stone ;p
<hattivat> don't think*
<egg|df|egg> not sure if they ended up being circulated, but they tried making some
<kmath> <whitequark> OH: "Therefore dogs are cats, and as spiders have 8 eyes, they must be cat interferometers."
* soundnfury gives egg|df|egg a gegenschein regenschirm
<UmbralRaptor> egg|df|egg: \o/
<egg|df|egg> !wpn UmbralRaptor
* Qboid gives UmbralRaptor a POSIX rocket-like oscilloscope
<egg|df|egg> !wpn Nomal
* Qboid gives Nomal an einsteinium semigroup
<egg|df|egg> how is Nomal doig
<egg|df|egg> s/ig/ing
<Qboid> egg|df|egg meant to say: how is Nomal doing
<egg|df|egg> "So I overindulge sometimes"
<UmbralRaptor> Is she sliding into alcoholism. =S
<egg|df|egg> she felt anxious after discussing foraging behaviour, existence, pulmonary medicine, technological evolution, time, diseases, the screw, conical water clocks, water displacement, bodily fluids, dictionaries
<egg|df|egg> right now she's discussing reproductive behaviour
<egg|df|egg> UmbralRaptor: she is dejected after a lack of decent meals for too long
<egg|df|egg> but then again she's right next to a room with food
<egg|df|egg> should I put food inside the amphitheatre?
<egg|df|egg> I'm not sure whether she's actually trying to get the decent meals or just complaining about their nonexistence
<UmbralRaptor> Quick, check if her stipend covers food.
<egg|df|egg> dinosaurs \o/
<egg|df|egg> the inky contingents, a performance troupe, is visiting
<kmath> YouTube - I Don't Want To Set The World On Fire-The Ink Spots
<kmath> <RupertMyers> Out-shaken https://t.co/OQr5Q2uaA8
<egg|df|egg> Ellied: UmbralRaptor: Al is indeed more valuable than Au in DF
<egg|df|egg> though probably useless
<hattivat> UmbralRaptor: that infographic managed to make flying rats look appealing, quite an achievement
<egg|df|egg> ah you can make furniture http://dwarffortresswiki.org/index.php/DF2014:Aluminum
<hattivat> (pigeons, I loathe the bottom-feeder bastards)
<UmbralRaptor> hattivat: pet flying rats!
<hattivat> yeah, still flying rats
<hattivat> and not even rats with good PR like squirrels
<hattivat> just plain flying rats
<hattivat> anyway, dawn is approaching, high time to go to sleep O_o
<hattivat> g'night folks!
hattivat has quit [Quit: WeeChat 1.8]
<UmbralRaptor> Hey, Darwin had an entire chapter on them!
<UmbralRaptor> bah, no more hattivat.
kmath has quit [Remote host closed the connection]
kmath has joined #kspacademia
<egg|df|egg> UmbralRaptor: nomal was satisfied upon improving observation!
<egg|df|egg> UmbralRaptor: also anxious after discussing the windlass, perception, processes, exacerbation, reproductive behaviour
<egg|df|egg> she does a lot of that whole discussing thing
<egg|df|egg> !wpn
* Qboid gives egg|df|egg a type
* egg|df|egg stares at the time
<egg|df|egg> !wa time in zurich
<Qboid> egg|df|egg: current time in Zurich, Switzerland: 4:34:31 am CEST | Friday, May 26, 2017
<egg|df|egg> >_>
egg|df|egg is now known as egg|zzz|egg
<UmbralRaptor> egg|zzz|egg: Research leads to discussion. Discussion leads to anxiety. Anxiety leads to research.
<Ellied> ;wa hubble's constant in Hz
<kmath> Ellied: convert 1 H_0 (Hubble constant) to hertz: 2×10^-18 Hz (hertz)
<Ellied> not sure why I didn't do that yesterday
icefire has quit [Read error: Connection reset by peer]
<whitequark> ;u ☡
<whitequark> !u ☡
<Qboid> U+2621 CAUTION SIGN (☡)
<Iskierka> Ellied, aluminium's straight strength-to-weight is actually rather unremarkable. Aluminium itself is actually quite weak (literally never used), alloys (what's actually used) are actually comparable to, and occasionally slightly better than, steel. The advantage is the bulk density being lower means significantly more torsional strength for the same compressive and tensile strengths, which require a significantly larger
<Iskierka> object
<Iskierka> if you want superior strength to weight you *have* to go to titanium or composites, or at high temperatures inconel, and steel actually still holds top spot for most strength & hardness in one material, which is what makes suitable weapons and tools.
<whitequark> *stares at Iskierka*
<whitequark> it seems really unfair to compare "aluminium alloys" to "steel" especially given that tensile strength of aluminium alloys strongly depends on the temper and there are literally thousands of steels
<whitequark> A36 and 2014-T6 are comparable in both yield strength and UTS, for example
<whitequark> (which means that per unit weight 2014-T6 is about three times better...)
<whitequark> which is, what you say might well be true but the argument is extremely unconvincing to me
<Iskierka> whitequark, broadly speaking, the top steel alloys retain about the same S/W as the top aluminium alloys, meaning for maximum strength you don't automatically prefer aluminium; you need to ask whether tensile or torsional or hardness is what's desired, although for most structural uses torsion is. But steel does still see use in a few airliner components, even among aluminium, titanium, and inconel
<Iskierka> (and where weight must be minimum)
<whitequark> right, that sounds much more sensible to me
<whitequark> I imagine cost, corrosion, and manufacturing operations possible are also quite important
<Iskierka> Corrosion is the main one for inconel, which fails the other two hard, but steel is an affordable one that's hard-wearing for physical considerations at least. Doesn't deal so well with exhaust above melting point
<Iskierka> (well, inconel as a material is comparable to steel in cost, but glhf keeping it that low after manufacturing with it)
<whitequark> would we even use steel for aircraft wings if it had the same density as aluminium?
<whitequark> (as in, some steel, suggest your own)
<Iskierka> well, easiest answer; now, no, it's all composites. Otherwise, good question, how much thicker does it need to be for the same torsion and an excessive tensile strength?
<whitequark> sure, it's cheaper, but then mild steels rust, and high-chromium steels have the same fatigue issue as aluminium alloys but *also* harder to work with and pretty expensive
<Iskierka> I think probably yes but it wouldn't be close to the same as the density reduction. And cost is usually not a large object; they'd probably either just work with high chromium and take the cost or find a way to protect the mild steel
<Iskierka> the latter is certainly accepted as a technique for making inconel work at even stupider temperatures
<Iskierka> weight is top requirement and cost is not a major object beyond development, so long as the thing works and has more performance for more mfg cost it'll usually go through. Unless it's more costly to maintain airlines seem usually willing to take the hit for a new aircraft; new models certainly haven't gotten any cheaper
<Iskierka> maintenance cost they do care more about though and is Rolls-Royce's biggest advantage on the engine front, though they also have lower noise (for related reasons too). Fuel consumption is typically around 0.5% higher in comparable engines though
<whitequark> mmmm
<egg|zzz|egg> !wpn whitequark
* Qboid gives whitequark a photon hippomobile snake which strongly resembles a warhead
<egg|zzz|egg> whitequark: oh the ☡ is nice
<egg|zzz|egg> whitequark: in french it's read as "dangerous turn", and its usage by Bourbaki is very close to that https://en.wikipedia.org/wiki/Bourbaki_dangerous_bend_symbol
ferram4 has quit [Ping timeout: 204 seconds]
<egg|zzz|egg> !wpn
* Qboid gives egg|zzz|egg a Fréchet photon
icefire has joined #kspacademia
<egg|zzz|egg> !wpn Iskierka
* Qboid gives Iskierka an organometallic shoe
<egg|zzz|egg> !wpn Thomas
* Qboid gives Thomas a stabbity goto
<egg|zzz|egg> !wpn UmbralRaptor
* Qboid gives UmbralRaptor a rhenium hexagon
<kmath> <scanlime> https://t.co/bswtE3PXlE
icefire has quit [Read error: Connection reset by peer]
<egg|zzz|egg> !wpn bofh
* Qboid gives bofh a liquid wolf
* egg|zzz|egg pets the wolf
<egg|zzz|egg> !seen bofh
<Qboid> egg|zzz|egg: I last saw bofh on [24.05.2017 21:00:37] in #kspacademia saying: "yaey, typos "
icefire has joined #kspacademia
<egg|zzz|egg> !wpn icefire
* Qboid gives icefire a tunneling trihexaflexagon which strongly resembles a demi-cannon
<egg|zzz|egg> so, the mean anomaly from r and tg ν is reasonably well-conditioned in both arguments for |ν| in [π/2, π] at all eccentricities
<egg|zzz|egg> I suspect that the upper bound for the condition in tg ν tends to π as e tends to 1 amusingly
<egg|zzz|egg> hmmm, that's only if r is expressed in units of semilatus rectum though
<egg|zzz|egg> and the semilatus rectum is ill-conditioned if the flight path angle nears π/2...
ferram4 has joined #kspacademia
<egg|zzz|egg> !wpn ferram4
* Qboid gives ferram4 a peculiar 2N3906
<egg|zzz|egg> sarbut sarbian
<Sarbian> Hey egg|zzz|egg
* egg|zzz|egg pokes Kepler's equation with a stick
<kmath> <tsboyajian> A14: We have Doppler data and should be able to tell. https://t.co/C5uFe1wGQt
<egg|zzz|egg> Iskierka: ^
<UmbralRaptor> Yes, *that* Boyajian.
<Iskierka> the verbiage meaning that it should have already been apparent, or that now that it's being suggested they're compiling the data to figure it out?
<UmbralRaptor> I think the latter.
<SnoopJeDi> I would assume that paper has been in prep for a little while anyway and was turned into a rush job because of the new dip
<SnoopJeDi> Not that it is any worse for the wear, it's a good read! Um, aside from the 1σ 130-170 M_♃ bit
<Iskierka> suddenly "we don't need to guesstimate the period from the 60 degree cluster, publish with exact day number!"
<Iskierka> or the disc significantly larger than the star?
<SnoopJeDi> ...huh, I'd missed that. I, uh...
<UmbralRaptor> SnoopJeDi: uh, that would be… an M6V star? o_O
<SnoopJeDi> I don't know what that means \o/
<SnoopJeDi> ah, spectral type + mass?
<Iskierka> pretty much. + assumption of stable phase
<SnoopJeDi> V is a luminosity class?
<Iskierka> and we'd *probably* be able to see that, so like ... unless it's a really tiny black hole or something?
* SnoopJeDi is more or less completely ignorant of stellar taxonomy
<UmbralRaptor> SnoopJeDi: it is. Main Sequence.
<SnoopJeDi> spiffy
<UmbralRaptor> Also I was wrong. Probably more like M4 or M5. >_> Which, uh, doesn't help.
<egg|zzz|egg> sooo the mean anomaly has abysmal condition as a function of the cartesian coords when the true anomaly is a right angle??? O_o
<egg|zzz|egg> or... it depends?
<egg|zzz|egg> aaa
<egg|zzz|egg> so if e < 1 and ν > π/2, M(x/ℓ,y/ℓ) is well-conditioned
<egg|zzz|egg> if e >> 0 (including e > 1) and ν < π/2, M(x/a,y/a) seems well-conditioned too
<egg|zzz|egg> (not that I have any idea how to evaluate this 2-parameter functions correctly, but one bit of madness at a time)
* UmbralRaptor isn't sure that the cat can help.
* UmbralRaptor is also unclear on using mean anomaly for e ≥ 1.
<UmbralRaptor> Speaking of things with tails that do what they want, comets disguised as asteroids! https://arxiv.org/abs/1705.09086
<egg|phone|egg> UmbralRaptor: it is imaginary
<UmbralRaptor> ah
<egg|phone|egg> UmbralRaptor: well that is the reasonable thing
<egg|phone|egg> Typically people make it real
<egg|phone|egg> It is unbounded
<SnoopJeDi> "If Pluto were in the orbit around the sun Earth has," Tyson explained, "it would grow a tail. And that's no way for a planet to behave!"
* SnoopJeDi is very ambivalent about NDT
* egg|phone|egg shares the ambivalence
<whitequark> nondestructive testing?
<egg|phone|egg> !w
<egg|phone|egg> Uh
<egg|phone|egg> !Wpn whitequark
* Qboid gives whitequark a pi quark
<egg|phone|egg> ...
<whitequark> !wpn egg
* Qboid gives egg a ground mechanism
<SnoopJeDi> egg|phone|egg, at least he's not Michio Kaku?
<egg|phone|egg> True
* UmbralRaptor stares at evaporating exoplanets.
* egg|phone|egg pets UmbralRaptor
<SnoopJeDi> I was rather furious that they brought him on the TV news circuit when Fukushima Daiichi happened
<SnoopJeDi> although I think maybe that was *somewhat* unwarranted, Dan Carlin thinks highly enough of his book on nukes that I have to reconsider on that topic in particular
<SnoopJeDi> And he *did* build a cyclotron in his parents's garage...
<Iskierka> what're his general opinions on nuclear energy? Since that would be the sensible judgement on whether he should be allowed in the news for fukushima
<SnoopJeDi> "allowed," heh
<Iskierka> .... a rather ambivalent position
<SnoopJeDi> we're talking about an industry that is now soliciting the opinions of literal nazis so idk how much fruit there is on that tree but
<Iskierka> ah, right
<SnoopJeDi> I'm not super familiar with his views
<SnoopJeDi> What little I've heard is pretty realistic
<SnoopJeDi> Particularly considering it's coming from an irredeemable futurist
<Iskierka> realistic is at least a good initial sign, depending on who says it
<SnoopJeDi> Not the DOE, to be sure
<SnoopJeDi> Nuclear as a community has an official stance best described as "I can't hear you la la la la"
<SnoopJeDi> file under: ray cats
<egg|phone|egg> !Wpn whitequark
* Qboid gives whitequark a magic bombe with a traversal attachment
<egg|phone|egg> !Wpn UmbralRaptor
* Qboid gives UmbralRaptor a cetacean paraboloid
<egg|phone|egg> !Wpn Iskierka
* Qboid gives Iskierka a radio alternator
<UmbralRaptor> !wpn egg|phone|egg
* Qboid gives egg|phone|egg a Bernoulli terbium proof
<Iskierka> ... apparently he complained about cassini's RTG. This is enough to conclude he is a silly person
* UmbralRaptor attempts to measure the paraboloid's mean anomaly.
<SnoopJeDi> Yea I only know his anti-proliferation stuff I guess?
<SnoopJeDi> In truth, just the fact that Dan Carlin gives his book the time of day is a step up
<SnoopJeDi> He's a serious scientist and all, I just can't take 99% of what he says seriously
<SnoopJeDi> which is a shame because the other 1% resonates strongly
<Iskierka> anti-proliferation sensible, his nuclear power stuff is not quite so when nuclear's the only one that can provide such functionally limitless safe energy
<SnoopJeDi> A lot to unpack there
<SnoopJeDi> I definitely can't take seriously anyone who claims anything reminiscent of safety in the neighborhood of LWR
<SnoopJeDi> with the abundance of evidence that running k > 1.0 is a matter of "when" rather than "if"
<SnoopJeDi> but you've heard me go on about this many times in KSPO
<egg|phone|egg> Whitequark: special function tables!
* egg|phone|egg summons the table maker's dilemma
<Iskierka> I'll say that at least more modern designs can be operated safely but a variety of reactor types is specifically why I said can to begin with. It can be built completely safe, and uranium alone can provide enough energy for basically utopia for all for as long as the sun lasts, if you're willing to accept fuel breeding
<Iskierka> if not, at least a big chunk of it, which'll be plenty long enough to figure out fusion and then there's nothing close to a fuel shortage
<SnoopJeDi> I don't agree about modern designs, but we're singing close enough to the same tune on the matter
<SnoopJeDi> And hey, my thesis topic has completed a lap and is back onto(ish) this subject
<SnoopJeDi> so that's neat
<Iskierka> Modern designs are definitely considerably safer than the originals and those operating them are aware of the dangers. I at least doubt any have been designed recently such that a single fuel rod can reasonably go k > 1 on its own, as happened in the prior chernobyl incident
<Iskierka> I forget which way round it was but leningrad and another also had 4 and 6 fail in other incidents, so it was a known fundamental flaw of just the entire fuel rod. The famous one was just when they didn't go fast enough to contain it to a handful and it spread to all ~1000 rods
<SnoopJeDi> Safer than the originals I'll agree with at least
<SnoopJeDi> But that's a very low bar to get over
<SnoopJeDi> Fuel rods are...a mess.
<SnoopJeDi> One of the last ARPA-E solicitations we responded to was focused on the material defects that arise in confined geometries
<SnoopJeDi> Put it on the long long long list of "why fuel rods are stupid"
<SnoopJeDi> well, that's a touch too loaded. the list of fuel rod cons, then
<Iskierka> Probably true, but on the side I'd also note that even chernobyl, even with the suppression of news, even with how ridiculously bad it was, didn't cause much harm in deaths, injuries, or illness per TWh generated by nuclear energy. By the standards nuclear could reach without LWRs, yes, they're dangerous, but by practical standards of "do we actually accept this elsewhere"? LWRs should be considered safe. We should just
<Iskierka> stop with them because we can do better
<soundnfury> !wpn Iskierka
* Qboid gives Iskierka a laser-cooled Qboid
<SnoopJeDi> hard to quantify at best Iskierka :(
<SnoopJeDi> but for all the criticism I can level at LWRs, at least they're not the current gameplan
<SnoopJeDi> petrol/coal is insane
icefire has quit [Ping timeout: 201 seconds]
<egg|zzz|egg> UmbralRaptor: what should I observe
<UmbralRaptor> !choose sunflower galaxy|Bode's galaxy|Sombrero galaxy|nearest BCG
<Qboid> UmbralRaptor: Your options are: sunflower galaxy, Bode's galaxy, Sombrero galaxy, nearest BCG. My choice: sunflower galaxy
<UmbralRaptor> BCG?
<Qboid> UmbralRaptor: [BCG] => Brightest Cluster Galaxy
<UmbralRaptor> @scrollback: Coal is better than burning trees, and not much else.
<Soozee> Unknown command, try @list
* egg|zzz|egg sees the ISS
egg|phone|egg has quit [Ping timeout: 204 seconds]
* soundnfury gives UmbralRaptor a sombrero cat galaxy
<egg|zzz|egg> tacocat?
<egg|zzz|egg> ¿tacocat?
<e_14159> egg|zzz|egg: Do you happen to have an algorithm to compute true from mean anomaly?
<egg|zzz|egg> e_14159: that's obtained by putting the true->mean formula into a solver
<egg|zzz|egg> which leads us back to the question of computing true->mean correctly :-p
<e_14159> First, you train a neural network...
<egg|zzz|egg> .............
<e_14159> egg|zzz|egg: You should not train it on "..."s exclusively
<egg|zzz|egg> stop trying to randomly throw linear algebra at transcendental functions
<SnoopJeDi> did somebody say neural network algorithms https://medium.com/towards-data-science/stupid-tensorflow-tricks-3a837194b7a0
<SnoopJeDi> I didn't know about the Thomson problem until I read that, pretty neat
<SnoopJeDi> You can tell a mathematician wrote a lot of the wikipedia article about it, because there's a huge table filled with tons of fairly useless information about solutions for multiple values of N
<e_14159> SnoopJeDi: Somehow, that article leaves me with a bad taste.
<e_14159> Nothing mentioned in the introduction is wrong per se, but it gives you a wrong idea about the whole thing.
<soundnfury> e_14159, egg|zzz|egg: konrad does a true->mean conversion at some point, in two steps (orbit.ean_from_tan() and orbit.man_from_ean())
<soundnfury> although of course it's probably numerically unstable, and it denies the existence of parabolae...
<SnoopJeDi> e_14159, oh?
<e_14159> SnoopJeDi: For example, "This is where TF excels, as it removes the drudgery of algorithmic differentiation and automagically moves the computation to the GPU". This mixes two different issues (autodiff and gpu computing), and makes it sound as if this were a unique feature of tf when it has been part of almost all DL libraries for quite some time.
<egg|zzz|egg> soundnfury: it's likely numerically crap, as principia's current Keplerian formulae are
<egg|zzz|egg> doing it right looks *hard*
<egg|zzz|egg> but interesting
<egg|zzz|egg> but *hard*
<soundnfury> egg|zzz|egg: s/likely/definitely
<Qboid> soundnfury thinks egg|zzz|egg meant to say: soundnfury: it's definitely numerically crap, as principia's current Keplerian formulae are
<egg|zzz|egg> also it's quite dark now
egg|zzz|egg is now known as egg|observing|egg
<egg|observing|egg> UmbralRaptor: scopetime!
<SnoopJeDi> e_14159, oh I thought you meant an issue re: the problem
<SnoopJeDi> I set the bar pretty low for the sort of person who goes full-bore on TensorFlow
<SnoopJeDi> IME there's substantial overlap with the "ENTREPRENEURSHIP! ALGORITHMS! BIG DATA!" mentality
* e_14159 goes full-bore on TensorFlow
<SnoopJeDi> not that neural networks aren't super frikkin' neato, just that they attract the silver bullet type I guess?
<e_14159> Agreed, though.
<SnoopJeDi> TF in particular seems pretty culty
<SnoopJeDi> maybe because they overstoked the hypetrain prior to release, heh
<e_14159> I mean, I like tf mostly because it's... well, *neat* would be the right word I guess, especially compared to theano I used before.
<SnoopJeDi> e_14159, have you read https://arxiv.org/abs/1703.10717 ?
<SnoopJeDi> really impressive facial interpolation (see figs 2-5)
<e_14159> But you do get some interesting people. We had one applicant who - after having done about two months of the udacity course - presented the udacity course project, claimed that the functions used in NNs are a good model for real neurons, and that yes, more layers is always better.
<SnoopJeDi> I know only enough about NNs to know I don't know a damned thing about them, and that a lot of the most knowledgeable folks don't know all that much either ?
<e_14159> Facial interpolation on a GAN? That sounds interesting. I'd have guessed a GAN/VAE mix architecture
<e_14159> And that it's not a silver boolit.
<SnoopJeDi> If I was gonna dive into ML (and I really need to if I'm serious about this data science tack), I'd start with way simpler stuff because of the value:effort
<SnoopJeDi> k-means clustering might not be super sexy, but at least you can reason about it
egg|phone|egg has joined #kspacademia
<e_14159> SnoopJeDi: Interesting paper, but unfortunately my GAN/VAE time is long past.
<e_14159> And I agree that starting small is very good. Random forests or k-means clustering is simple, but often enough. And a simple linear or logistic classifier too, plus it gets you experience on neural networks.
<e_14159> (Come to the dark side, we have NNs!)
<SnoopJeDi> Yea, there's so much left to pick up off the table with just smart application of domain knowledge
<SnoopJeDi> Completely unrelated, but really really amazing tornado photos (my gf is a meteorology student so I collect these as catnip): http://stormhighway.com/may242016.php
<e_14159> Wow
<e_14159> (Although I am quite happy not to live in an area with tornados)
<e_14159> ((Well, except for the planes))
<SnoopJeDi> I never appreciated how interestingly wacky meteorologists are
<e_14159> SnoopJeDi: s/meteorologists/scientists
<Qboid> e_14159 thinks SnoopJeDi meant to say: I never appreciated how interestingly wacky scientists are
<SnoopJeDi> Well that I appreciated full well in general, that's why I'm in #kspacademia :)
<e_14159> Good point. Very good point.
<SnoopJeDi> I get all ga-ga eyes when she goes on about Skew-T plots and so on, because I cannot grok a single thing she's saying and that's *fascinating*
egg|laptop|egg has joined #kspacademia
<egg|laptop|egg> UmbralRaptor: live from the hill!
<egg|laptop|egg> lots of planes landing behind me
<e_14159> New knowledge often is, SnoopJeDi
<egg|laptop|egg> Jupiter, Ganymede, Io, Europa
<e_14159> egg|laptop|egg: s/nes/nets
<Qboid> e_14159 thinks egg|laptop|egg meant to say: lots of planets landing behind me
<UmbralRaptor> egg|observing|egg: Are there any tornadoes? mirages?
<SnoopJeDi> e_14159, or even old knowledge, cf. the Thomson problem
<SnoopJeDi> Or Maxwell's paper on applying Vlasov flow to the rings of Saturn
<egg|laptop|egg> UmbralRaptor: also callisto, quite faint, at higher magnification
<egg|laptop|egg> and bands
<SnoopJeDi> The negative mass instability is well-known in accelerators, but usually treated as though it's new knowledge when it is in fact quite old!
<SnoopJeDi> (and also annoyingly named)
<egg|laptop|egg> UmbralRaptor: the sunflower is at 84 deg alt
<egg|laptop|egg> might be tricky
<egg|laptop|egg> in terms of neck bending
<UmbralRaptor> SnoopJeDi: This sounds like those wacky quasiparticles that get treated as real by the hype portion of the media.
<SnoopJeDi> UmbralRaptor, s/the.*/literal scientists/
<Qboid> SnoopJeDi thinks UmbralRaptor meant to say: SnoopJeDi: This sounds like those wacky quasiparticles that get treated as real by literal scientists
* SnoopJeDi blinks at Cooper pairs
<UmbralRaptor> Not math real, but like they exist outside of the relevant physical environment.
* UmbralRaptor is sorta fuzzy on BCS, though.
* soundnfury gives UmbralRaptor a phonon
* UmbralRaptor bounces off a crystal lattice.
<soundnfury> UmbralRaptor: s/atti/ettu
<Qboid> soundnfury thinks UmbralRaptor meant to say: /me bounces off a crystal lettuce.
<UmbralRaptor> o_O
<egg|laptop|egg> UmbralRaptor: M3!
<UmbralRaptor> \o/
<UmbralRaptor> egg|laptop|egg: What if we used more power?
* egg|laptop|egg increases magnification on M3
<egg|laptop|egg> also apparently the swiss use that hill to party when the weather is good for observation
<egg|laptop|egg> strange music is being played
<egg|laptop|egg> UmbralRaptor: it's bigger! and still fuzzy!
<egg|laptop|egg> the song is saying something about life and nitroglycerin infused into a porous substance
<egg|laptop|egg> UmbralRaptor: hmm, while I'm in the area I guess I should try M53?
<UmbralRaptor> Probably.
<UmbralRaptor> egg|laptop|egg: Are there dwarves playing music near you?
<egg|laptop|egg> "don't you love me baby"? the human league? something about switzerland being 40ish years in the past
<egg|laptop|egg> UmbralRaptor: I don't think they're dwarves
<egg|laptop|egg> I'm wondering if eventually their playlist is going to rickroll me given the timeframe of that last song
<UmbralRaptor> hrm
<kmath> <Alex_Parker> Turns out a micrometeoroid hitting your space camera makes a 'spang' sound. [details shortly]
<egg|laptop|egg> UmbralRaptor: M53!
<UmbralRaptor> \o/
<egg|laptop|egg> UmbralRaptor: it's rather faint
<egg|laptop|egg> UmbralRaptor: I brought a tiny folding stool this time though, so I'm not breaking my back at the scope
<egg|laptop|egg> I'm being winded on a bit though
<egg|laptop|egg> what with hills
<UmbralRaptor> hrm
<egg|laptop|egg> UmbralRaptor: thing crossed the field, not on stellarium
<egg|laptop|egg> satellite? aircraft?
<Iskierka> awiens
* egg|laptop|egg wonders whether the noise behind him is aircraft
<egg|laptop|egg> no storm forecast, so unlikely to be thunder
<egg|laptop|egg> also no clouds
<egg|laptop|egg> the music has gone back to things I can't parse
<egg|laptop|egg> UmbralRaptor: M13!
<egg|laptop|egg> music has gone silent
<egg|laptop|egg> UmbralRaptor: M92, too
<UmbralRaptor> \o/
<egg|laptop|egg> UmbralRaptor: hmm, anything else to stare at
<egg|laptop|egg> (deep-sky)
<egg|laptop|egg> (this scope isn't a binary splitter)
<UmbralRaptor> M57?
<UmbralRaptor> M29?
<egg|laptop|egg> UmbralRaptor: tbh I might get back soon, 291 K isn't warm enough for the clothes I'm wearing, surprisingly
* egg|laptop|egg might try M57 though
<egg|laptop|egg> it's a cute star field around there
<kmath> YouTube - PJ06 preview
<egg|laptop|egg> UmbralRaptor: I have M57 I think
<egg|laptop|egg> much harder to spot than the globulars
<egg|laptop|egg> but yup, there's an unfocusable thing amidst all those stars
<egg|laptop|egg> needed to go to the 5mm to be sure though
<egg|laptop|egg> whereas the globulars are identifiable in my navigation 13 mm
<egg|laptop|egg> UmbralRaptor: will try M29, seems daring though, with the low altitude
<egg|laptop|egg> and getting closer to the airport in azimuth
<egg|laptop|egg> might be ok though
<egg|laptop|egg> UmbralRaptor: I have it quite clearly actually
<egg|laptop|egg> open clusters are cute
<UmbralRaptor> \o/
<egg|laptop|egg> UmbralRaptor: tbh I have only 8 stars in it :-p
<UmbralRaptor> It's a small one.
<egg|laptop|egg> getting a bit cold though, might head back
<UmbralRaptor> So if we used a sufficiently large telescope to concentrate, say, Arcturus on you...
egg|laptop|egg has quit [Ping timeout: 180 seconds]
<egg|observing|egg> UmbralRaptor: >_>
<egg|observing|egg> you mean the SLT
<egg|observing|egg> the Sufficiently Large Telescope
<UmbralRaptor> Er, yes. >_>
<egg|observing|egg> !u Ꙭ
<Qboid> U+A66C CYRILLIC CAPITAL LETTER DOUBLE MONOCULAR O (Ꙭ)
icefire has joined #kspacademia
egg|observing|egg is now known as egg
<egg> !wpn whitequark
* Qboid gives whitequark an equality with a physicist attachment
<UmbralRaptor> Conservation of physicists
<egg> !wpn UmbralRaptor
* Qboid gives UmbralRaptor a cotangent manifold
egg is now known as egg|df|egg
<UmbralRaptor> !wpn Nomal
* Qboid gives Nomal a dynamic peeled ſtabber
egg|cell|egg has joined #kspacademia
egg|phone|egg has quit [Read error: -0x1: UNKNOWN ERROR CODE (0001)]