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.
* UmbralRaptor
picks glass shards out of their feathers.
* UmbralRaptor
goes back to stabbing canonical transformations
<SilverFox>
riddle me this one, how does a sphere have zero volume?
<kmath>
<unormal> This place is a message and part of a system of messages pay attention to it! Sending this message was important t… https://t.co/OpktEJyCjt
<UmbralRaptor>
uh oh
* Iskierka
wonders what Ellied shenaniganed in the mac lab
* Iskierka
also thinks the mac lab at uni is keycard locked so probably can't shenanigan
<Ellied>
ours is keycard locked but our cards work in it
<Ellied>
we tried the Sierra get-root thing
<Iskierka>
Interesting, not sure ours would, pretty sure we do need to get explicit permission
<Ellied>
they give us a surprising amount of freedom here. 24h access, including to two computer labs, and those of us who do research have actual metal keys that unlock all the labs[1].
<Ellied>
[1] any given key only unlocks one of the faculty labs, and none of the currently-issued keys unlock the Dark Lab.
<Iskierka>
Yeah, we're trusted a remarkable amount for the systems in this building, but not that much. Only most of the non-specialist labs open, not 24 hours, no physical keys (but I don't think most staff have them)
<Iskierka>
Pretty sure at minimum the mac lab and the giant AI project are locked off
<Ellied>
I have what I think is the only currently-issued key to the Laser Lab (my prof's lab) which is honestly probably the most dangerous
<Ellied>
the lab next door (one of the other faculty labs) has some chemicals and stuff in it but no high voltages, pinchy mechanical parts, or high-powered lasers.
<Ellied>
I understand the Dark Lab has asbestos in it, which is why it's dark.
<Ellied>
not surprised by the lack of physical keys, tbh. Our keys apparently cost like $200 to have ground, but I guess our wonderful IT department must charge more than that to put in card scanners
awang has quit [Ping timeout: 200 seconds]
<Ellied>
selfishly, I'm very happy to have a metal key since it's much harder to revoke (especially by accident, or during holidays) and doesn't stop working if the network gets fucked up
<Ellied>
otoh I have been scheming about putting in a homebuilt electronic authentication system on my lab door, which would be a fun (if highly illegal and dubiously advisable) project.
<bofh>
Ellied: yeah the nice thing about metal keys is at my old uni they'd let me in during the winter break when the electronic locks would often be revoked.
* UmbralRaptor
probably should have returned the UMKC metal keys at some point.
* UmbralRaptor
was a good citizen about the MSU ones, and only knows a few codes for GMU.
awang has joined #kspacademia
icefire has quit [Read error: -0x1: UNKNOWN ERROR CODE (0001)]
<Ellied>
unfortunately I don't get access to the actual building (which is card only for us students) when the university is shut down for breaks. I'm told they're actually going to shut the building all the way down for christmas break, as in turn off all the lights and keep the heating just on enough to keep the pipes from freezing
<SnoopJeDi>
Ellied, illegal?
<SnoopJeDi>
oh, state university maybe?
<FluffyFoxeh>
this paper is about pairing neutral stimuli with sexual arousal; i.e. producing paraphilia. which makes the choice of the word "bibliophile" (one who loves books... but not like that) a bit hilarious in this context http://www.fastquake.com/images/screen-bibliophile-20171129-214740.png
<kmath>
<vdignum> Please endorse the Bechdel test for academic&tech conferences: 1) two women speaking 2) in the same panel or sessio… https://t.co/7ieAdoRRIm
<FluffyFoxeh>
for some reason, this is fascinating to me and I wish there were more large-n studies on it
<SnoopJeDi>
There was a really good one about questions from an astro conference or set thereof
<SnoopJeDi>
...which unfortunately showed that women ask a lot more second, third, ... questions once a male question opens the air :|
<FluffyFoxeh>
oh I was talking about the paper I linked. but what do you mean? women tend to only start asking questions of the presenters after a male asks one first?
<SnoopJeDi>
ah I see. yea that was the observed trend
<kmath>
<allgebrah> Stay up long enough and your sleep cycle syncs with the sun again, do this often enough and the parallel Earths you… https://t.co/OhLvcV9xhy
<SnoopJeDi>
makes me think of the Russian sleep experiment creepypasta
<Ellied>
"Illegal" in this case is "enough against university policy that I would see a similar amount of trouble as I would if I actually broke the law"
egg has joined #kspacademia
<SnoopJeDi>
oh my god Dogspotting is an actual sport?
<SnoopJeDi>
This might be my roller derby moment.
egg|phone|egg has joined #kspacademia
<SnoopJeDi>
Nope, it passed, they're kinda creepy with the advertising and app.
<Ellied>
why is it so hard to find information about what frequency capacitive screens operate at
<Ellied>
all I want is an order of magnitude, but googling won't seem to reveal even that. Not even in an article specifically about touch screen EMI
<kmath>
<mekosoff> today was so nuts that the convicted bosnian war criminal who drank poison during a court hearing and promptly died… https://t.co/cEOpVfgydJ
<FluffyFoxeh>
Ellied: take oscilloscope probe, wave it near touch screen, watch FFT
<FluffyFoxeh>
:p
egg|phone|egg has quit [Ping timeout: 383 seconds]
<kmath>
<JasonKalirai> We just posted our new (somewhat technical) science sheet summarizing the capabilities of the #WFIRST mission.… https://t.co/Tb8qKGMneF
Technicalfool has joined #kspacademia
Technicalfool has quit [Remote host closed the connection]
Technicalfool has joined #kspacademia
<bofh>
!u –
<Qboid>
U+2013 EN DASH (–)
<SnoopJeDi>
oh btw bofh, since I only had to replace 晓安 that tooltip is just a hard-coded thing, it's not looking up the pinyin or anything
<SnoopJeDi>
I bet there's a Translate API I could use for that though, heh.
<SnoopJeDi>
Briefly considered including the IPA in the tool-tip and then realized that was actually across the line for the scope of this. (we have been *butchering* the name in actual play)
<SnoopJeDi>
thankfully we have not had occasion to say her PC's surname (劉) because I have historically been garbage at that one
<kmath>
<sdssurveys> Also, the attendees at #sdss5data would do a horrible job at Congressional hearings (not that we are anticipating t… https://t.co/FByhQ6Gi86
<SnoopJeDi>
I posted that Bechdel-for-conferences in our dept Slack and someone tried to argue that it's driven by a lack of qualified women in the field :|
<UmbralRaptor>
Hopefully this was them not knowing? :|
APlayer has joined #kspacademia
egg|phone|egg has joined #kspacademia
egg|mobile|egg has quit [Read error: -0x1: UNKNOWN ERROR CODE (0001)]
egg|cell|egg has joined #kspacademia
egg|phone|egg has quit [Read error: Connection reset by peer]
* Ellied
imagines a modern Death of Socrates scene yesterday in Hague
<Ellied>
hopefully this guy won't find similar immortality.
<SnoopJeDi>
UmbralRaptor, he's the sort whose knee-jerk reaction to that kind of uncomfortable truth is to rationalize it away
<FluffyFoxeh>
personally I don't see an issue unless female panelists are being specifically turned away
<SnoopJeDi>
Nice enough guy, but high-strung by nature. I'm super sympathetic to that and all, but not at the expense of...20%+ of our field :|
<FluffyFoxeh>
the mere fact that there is an underrepresentation does not strike me as problematic, it's the reason why that matters
<SnoopJeDi>
I don't agree personally
<Ellied>
the problem is that women are being pushed out of tech by men, and not always in overt ways that are easy to crack down on. There is *no* good evidence that women are somehow less good at STEM than men, but the disparity between women and men in STEM jobs is huge. That's a problem.
<SnoopJeDi>
yea more to the point this person specifically replied that the issue was recruitment/retention of women
<UmbralRaptor>
There's also a tendency for women to be underrepresented in panels relative to the number in $field. CS is particularly notable.
<SnoopJeDi>
which...I'm all for seeing more women in physics but 20% is *more* than enough that there should be more female speakers/panelists
<SnoopJeDi>
even before accounting for any kind of survivorship
<FluffyFoxeh>
all that can reasonably be done is to ensure a level playing field. but you can't force people to play
<SnoopJeDi>
"level playing field" is vague. "how many women are speaking" is very much not vague
<SnoopJeDi>
it's not perfect either, but I'd rather stumble towards a solution than offer an apoplectic shrug
<SnoopJeDi>
which is this guy's general reaction to anything of this sort (I should clarify that he's a nice guy and I don't dislike him or anything, but he is imo too quick to form his opinion)
<Ellied>
No one is saying that women should be *forced* into STEM fields. They're busily being force *out* right now. In order to be part of fixing that, you have to actively try to balance out the distribution where you can.
<FluffyFoxeh>
if that means selectively favouring women then you've just flipped the sign of the problem
<SnoopJeDi>
overall stats are another can of worms and altogether rather messier imo
egg|phone|egg has joined #kspacademia
<SnoopJeDi>
vs. people *already* doing incredible work not having their voices heard, which does everyone a disservice
<SnoopJeDi>
file under: ben_barres.jpg
<Ellied>
It's not "unfair to men" to favor qualified female applicants if the pool is smaller because if everything were already fair, there would be a bigger pool of female applicants that didnt get chased off before they had the chance and the personal chances of being hired for a given male applicant would be the same
<Ellied>
Sexism isn't a problem that you can just magically reverse by adjusting your hiring practices, it's institutional
<SnoopJeDi>
yea it's not a reflection because the problem statement has no symmetry
egg|phone|egg has quit [Read error: Connection reset by peer]
<SnoopJeDi>
that's rather the point of it
<SnoopJeDi>
there aren't *obvious* solutions I think, and it's hard to get it right
egg|cell|egg has quit [Ping timeout: 186 seconds]
<SnoopJeDi>
I think GitHub's ElectronConf is a perfect example of this: they did gender-blind talk proposals and ended up with zero women
egg|phone|egg has joined #kspacademia
egg|cell|egg has joined #kspacademia
APlayer has quit [Ping timeout: 383 seconds]
<SnoopJeDi>
to their credit, they decided to throw out the result because it didn't match what they wanted, rather than clinging to the result because of some misplaced notion of purity of the ruleset
<SnoopJeDi>
i.e. they cared more about the *result* than the *process* that yielded it. The Reddit Admin effect.
<FluffyFoxeh>
manipulating outcomes instead of predicates is something I can't agree with
<SnoopJeDi>
well sure, throw 10 people in a room and pose the question and you'll get 15 proposed solutions that look nothing alike
egg|phone|egg has quit [Ping timeout: 186 seconds]
<SnoopJeDi>
but I bristled rather strongly at "oh that's not actually a problem that needs solving" which is how this person's response came off
<SnoopJeDi>
"well, ACTUALLY, it's this other problem over here"
<Ellied>
you can't just say "let's just do a gender-blind application process, then we'll be completely absolved of sexism and it won't be our fault at all if they're all men" because someone else (or rather, everyone else) has already rigged the applicant pool before it even gets to you.
<Ellied>
then you're just being complicit in the problem and denying that it exists
<FluffyFoxeh>
if you correctly conduct a gender-blind application process then it is not your problem, it can't possibly be your problem, because you explicitly did not favour anyone based on gender
<SnoopJeDi>
well I think that's probably not the thought process
<FluffyFoxeh>
so where do you point the finger? everywhere? that's not actionable
<SnoopJeDi>
but the *reason* they tried it was a desire for a particular outcome
<SnoopJeDi>
and they acted according to that values statement
<SnoopJeDi>
it's the difference between saying "we're all about diversity" on your website and every so often taking a critical look at who works at your company
<Ellied>
of course you can't just jump up and fix society, no one is suggesting we try that, but you can do things that will help slowly drag us out of the constant feedback loop that's making people assume that women don't belong in STEM and forcing them out
<SnoopJeDi>
the "don't shoehorn women into WISE/etc. panels" is also a really important part of that Bechdel tweet
<SnoopJeDi>
because the burden is absolutely carried unequally
<SnoopJeDi>
or rather, "it doesn't count if it's [gimme category]"
<SnoopJeDi>
and I think those events are a pretty fair gimme at most conferences? certainly in PACs
<SnoopJeDi>
and I gather at astro
<FluffyFoxeh>
is there any evidence of an attitude that women don't belong in STEM or that people would take conscious action to keep them out of it
<SnoopJeDi>
yes but that's also another issue
<SnoopJeDi>
than the Bechdel thing I mean
<SnoopJeDi>
but if you want a recent example the Google Memo bullshit serves nicely
<Ellied>
consciously-sexist decisions aren't even the main problem. What really forces women out is constantly being treated as sub-standard by their male colleagues, who might not even notice that they're doing it, or if they do notice, not see why it's a problem.
<FluffyFoxeh>
this sounds really bogeyman-like
<SnoopJeDi>
how do you mean
<SnoopJeDi>
I line up very much with what Ellied just said, the road to hell is paved with good intentions heh.
<Ellied>
that's why it's so hard to see when this is happening, because you can just sit there and go "is my professor sexist? of course not! He has female research students, and he'd never reject someone as an applicant to grad school just for being a woman!" but just because they've never done anything overtly sexist around you doesn't mean they haven't spent decades using unconscious double standards
<Ellied>
It doesn't have to be a really obvious fork in the road (e.g. "male applicants get in, female applicants don't"), it can just be stuff like male students getting more help from a professor, more leniency during grading, or more encouraging feedback after success or failure
<FluffyFoxeh>
meh, I'm not convinced. but I don't need to be, I don't make these or any other relevant decisions
<Ellied>
Good, I hope not.
<FluffyFoxeh>
that sounds pointed, I don't appreciate it
<SnoopJeDi>
the way I see it, everybody's part of the ecosystem, so everybody's behavior *does* matter
<SnoopJeDi>
but yea granted there are places where the lever arm is a lot longer
<SnoopJeDi>
...like, for example, panels :)
egg|phone|egg has joined #kspacademia
egg|cell|egg has quit [Read error: Connection reset by peer]
<egg|afk|egg>
!wpn Ellied
* Qboid
gives Ellied a Christoffel extensional vial of Y-909
<egg|afk|egg>
!wpn UmbralRaptor
* Qboid
gives UmbralRaptor a samarium descending McShipFace with an impenetrability condition attachment
<kmath>
<stephentyrone> @bofh453 Maybe 15 years ago, my grandparents next-door neighbors renovated their kitchen. The end wall of the house… https://t.co/QwZKlKYOh7
<UmbralRaptor>
egg: a portable distribution function. Or a probability document format.
<UmbralRaptor>
(As in it's probably a document, and not malware)
<Qboid>
[#1351] title: Accumulate gravitational interactions in DoublePrecision | Different behaviour due to non-associativity being amplified by the uncorrected stock system's terrible Ляпунов exponent is expected, but Aardvark being stabler than Kerbol which is stabler than Sun p... | https://github.com/mockingbirdnest/Principia/issues/1351
APlayer has joined #kspacademia
* UmbralRaptor
shrugs.
egg is now known as egg|nomz|egg
<UmbralRaptor>
Hm. Goldstein has a part on precession of satellite orbits.
* UmbralRaptor
eggspects to mainly see problems involving tops instead.
<kmath>
<vi3k6i5> @jgerity @freeCodeCamp @jgerity BTW: All benchmark codes are linked in the Medium post, if you want to review the c… https://t.co/y4XKXRBhTc
<SnoopJeDi>
half-gross reply to "I think maybe your benchmark is unreasonable"
<SnoopJeDi>
(original post was good though, tl;dr don't use regex if something more specific exists)
<SnoopJeDi>
The benchmark I was questioning was https://git.io/vbtTS, jamming all your patterns into a regex joined by '|' seems...a little unfair to `re`
<kmath>
<FioraAeterna> There was an engineer who swallowed a fly; I don't know why she swallowed a fly - perhaps she'll die! There was an… https://t.co/LH9229xH16
<SnoopJeDi>
JS-hating kitsch being what it is, I think reveal.js is a very nice slideset tool and has the neat bonus that your slides can be directly viewed on e.g. a GitHub Page
<egg|nomz|egg>
Fiora: so at some point you'll have catbots
<kmath>
<diodelass> 2010: Everyone uses facebook because it's convenient 2016: Facebook is a hell but now we're trapped 2017: THE ALGOR… https://t.co/eFb6XHypFg
* UmbralRaptor
just approves of the perversity in writing slides in vim. >:D
<egg|nomz|egg>
!wpn UmbralRaptor
* Qboid
gives UmbralRaptor a gastrointestinal hedgehog
<egg|nomz|egg>
!wpn Fiora
* Qboid
gives Fiora an equinoctial profuct
egg|nomz|egg is now known as egg
<UmbralRaptor>
That sounds very painful.
<egg>
true
icefire has joined #kspacademia
<Ellied>
APlayer: huh?
<APlayer>
Ellied: Asking if that tweet is legit?
<APlayer>
That is the information content of this tweet :P
<Ellied>
I mean I can't predict the future but the thing about facebook calling first responders automatically if their AI thinks you're suicidal is apparently a real thing
<APlayer>
Dang, that sucks. Fuck facebook. (This may have actually been my first time swearing on IRC like this, but I mean every single word of it.)
<egg>
!u U+f6c1
<Qboid>
U+F6C1 PRIVATE USE-F6C1 ()
<SnoopJeDi>
UmbralRaptor, writing Markdown in vim is nice!
<SnoopJeDi>
writing is equations in...anything that isn't hand-input isn't much fun?
<SnoopJeDi>
APlayer, <politics>that's your future probably-president, show some respect</politics>
<APlayer>
SnoopJeDi: I believe we had a discussion about me ignoring any politics that goes beyond voting in my own home country... :P
<SnoopJeDi>
right you are, my US-centrism is showing again
<SnoopJeDi>
s/your/my/
<Qboid>
SnoopJeDi meant to say: APlayer, <politics>that's my future probably-president, show some respect</politics>
<APlayer>
Anyway, one more reason to boycott facebook
<UmbralRaptor>
SnoopJeDi: well, yes. But I write LaTeX in vim. o_O
<egg>
bofh: so remember that question about computing f' from two values of f and a bunch of values of f"?
<egg>
bofh: turns out it's fairly easy, but here's an amusing tidbit
<SnoopJeDi>
UmbralRaptor, same, although I write *that* as Markdown, too
<SnoopJeDi>
and let pandoc do the hard part
<egg>
bofh: if you compute f'(0) from f(-h), f(-2h), and f"(-h), f"(-2h), f"(-3h), ..., f"(-13h), the coefficient on the 14th order residual term is ~0.26
<egg>
bofh: if you compute f'(0) from f(0), f(-h), and f"(-h), f"(-2h), f"(-3h), ..., f"(-13h), the coefficient on the 14th order residual term is ~0.21
<egg>
bofh: if you compute f'(0) from f(0), f(-h), and f"(0), f"(-h), f"(-2h), ..., f"(-12h), the coefficient on the 14th order residual term is ~0.0029
<egg>
:D
APlayer has quit [Ping timeout: 383 seconds]
<egg>
bofh: also if you compute f'(0) from f(0), f(-h), ..., f(-14h), the coefficient is 1/15
<egg>
...
<egg>
translate your langdet is bad and you should feel bad
<egg>
ComputeHarmonicOscillatorAcceleration1D -> romanian
<egg>
ComputeHarmonicOscillatorAcceleration3D -> english
<egg>
IntegrationProblem -> danish
<bofh>
LOL
<egg>
nullptr -> romanian
<egg>
(I should show that to my romanian colleagues)
<bofh>
Also, uhm, that's quite the interesting numerical phenomenon. Looks a bit like a weird analogue of Runge's Phenomenon, almost?
<egg>
bofh: I don't know, but we're currently doing the 1-sided finite difference (f(0) .. f(-14h)) so we should switch to the f(0) f(-h) f"(0) .. f"(-12h) I guess
<egg>
bofh: I wonder how this all compares to the symmetric finite difference (which we can't use because it relies on the future)
<bofh>
Like I think the symmetric finite difference would obviously give you a much better result but it's obviously unusable here.
<egg>
bofh: so it's usable in a lot of similar applications, you just integrate (working purely on the positions with your multistep method) and then get your velocities
<egg>
bofh: the thing is we want to immediately get the velocities so that doesn't work here
<Qboid>
[#1351] title: Accumulate gravitational interactions in DoublePrecision | Different behaviour due to non-associativity being amplified by the uncorrected stock system's terrible Ляпунов exponent is expected, but Aardvark being stabler than Kerbol which is stabler than Sun p... | https://github.com/mockingbirdnest/Principia/issues/1351
<egg>
never seen them before, they pop up on this issue and suggest something that the comment above says doesn't work Ꙩ_ꙩ
<Qboid>
1d 12h 0m 0s left to event #16: #2/Союз-2.1б [at 2017-12-02 10:30:00]. Say '!kountdown 16' for details
<bofh>
sec.
<bofh>
"Kahan summation can yield large relative errors for ill-conditioned sums. That's one of these cases." yep I've been bit in the arse by this property before.
<bofh>
It's "fun".
<egg>
:D
<egg>
that's why we have a double-double in principia
<egg>
the multistep integrators are of that kind
<bofh>
Also I now see why you want to be able to sense instability in the integrator: for a stable solar system you're good in the current inner loop, for an unstable one you currently *will* need corrections but those pay a *heavy* performance penalty.
<egg>
bofh: eh, if it's unstable to the point that our 10 s time step isn't enough we can be pretty confident that 1. if it's a vessel it's being destroyed by non-gravitational interactions 2. if it's the system itself it's the apocalypse and we report it
<egg>
(well the system runs at a timestep of minutes, not seconds)
<egg>
(but basically we're fine with bailing out on instability)