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|zzz|egg>
bofh: hmm, so the thing is I have annoyingly low resolution for the low frequencies using the 5 "predictable" years, and then if you look at the century-long graph some weird changes happen (close encounters?) so that it fills with noise
<bofh>
so I think there's some nasty chaotic behaviour in the century-long graph.
<bofh>
lowpass-filter it perhaps?
<egg|zzz|egg>
bofh: yeah I think the century-long graph is mostly worthless, I'm pretty sure there are close encounters in there that truly change the behaviour of things
<bofh>
this is "please give me my fucking money that you are holding in magical limbo for a transaction that was marked as received with a positive ebay review already"
<bofh>
like, by their own policy it should have been released yesterday, but it isn't, and the "funds will be available by" estimate is December 31st, 1969.
<bofh>
I am 100% serious.
Majiir is now known as Snoozee
<Iskierka>
Lol did they set the time to -1?
<bofh>
APPARENTLY.
<Qboid>
1d 12h 0m 0s left to event #2: Atlas 5 • TDRS M [at 2017-08-18 14:04:00]. Say '!kountdown 2' for details
<bofh>
okay after way too fucking long on hold they fixed that in 5 minutes >_>
ferram4 has quit [Read error: Connection reset by peer]
ferram4 has joined #kspacademia
awang has joined #kspacademia
<bofh>
lmfao freenode accidentally klined the entire network and crashed a few of their servers in the process
<FluffyFoxeh>
I didn't get K-Lined
<FluffyFoxeh>
:p
xShadowx|2 has joined #kspacademia
xShadowx has quit [Ping timeout: 200 seconds]
xShadowx|2 has quit [Client Quit]
xShadowx has joined #kspacademia
Moistmelon has quit [Ping timeout: 183 seconds]
<egg|zzz|egg>
!wpn bofh
* Qboid
gives bofh a poisonous irritating comma
<egg|zzz|egg>
!wpn Fiora
* Qboid
gives Fiora a dipole cardinal
egg|zzz|egg is now known as egg|afk|egg
egg|phone|egg has joined #kspacademia
egg|cell|egg has joined #kspacademia
egg|phone|egg has quit [Read error: -0x1: UNKNOWN ERROR CODE (0001)]
<Qboid>
egg|phone|egg: Description: A Протон-М with Бриз-М upper stage will launch the Экспресс-2000-based Благовест №11Л communications satellite into GEO (see http://space.skyrocket.de/doc_sdat/blagovest-1.htm).
<egg|phone|egg>
Hmm, no Роскосмос livestream?
egg|cell|egg has joined #kspacademia
egg|phone|egg has quit [Read error: Connection reset by peer]
egg|phone|egg has joined #kspacademia
egg|mobile|egg has joined #kspacademia
egg|phone|egg has quit [Read error: Connection reset by peer]
egg|cell|egg has quit [Ping timeout: 183 seconds]
egg|mobile|egg has quit [Client Quit]
Technicalfool is now known as TechnicallySleeping
<egg|work|egg>
!u ê
<Qboid>
U+00EA LATIN SMALL LETTER E WITH CIRCUMFLEX (ê)
<egg|work|egg>
!u à
<Qboid>
U+00E0 LATIN SMALL LETTER A WITH GRAVE (à)
<egg|work|egg>
bofh: spaceflight now confirms the time that I got from the livestream (also I got a bug fixed in the formatting of the time of start of livestream, it used AM/PM for all languages this week-end)
<bofh>
AM/PMis a system that honestly needs to die. 24hr 4lyfe.
Ellied has quit [Quit: WeeChat 1.0.1]
Ellied has joined #kspacademia
<egg|work|egg>
bofh: use en-GB as a language on YouTube and you will get 24 h clocks now :-)
<egg|work|egg>
bofh: you should ask Atlas to resume his FloatingPointWithAtlas series
<Iskierka>
The other day I had to filter out dates (trying to compare two filestructures using the dir output) and the regex suggested was *awful* because it used AM/PM
<Iskierka>
also dd-mm-yyyy (or possibly mm-dd-yyyy, dunno with muricans) so if they tried to order that output they'd have a nightmare
<Iskierka>
meanwhile mine was just ISO date and so much easier
<egg|work|egg>
bofh: essentially, set your YouTube language to anything but English (US) or Español (US), and you'll get 24 hour clocks (even Español (Latinoamérica) has sane clocks)
<bofh>
ooooooh. done.
<egg|work|egg>
bofh: a week ago you would have had 12-hour clocks in some places, but that was a bug and it's fixed :-p
<bofh>
thanks
<Qboid>
1d 0h 0m 0s left to event #2: Atlas 5 • TDRS M [at 2017-08-18 14:04:00]. Say '!kountdown 2' for details
<bofh>
I maintain "Darl Yaes" sounds like a character in a DnD campaign.
<bofh>
Dral Yaes*
<Iskierka>
they're selling yale?
<NonlocalRaptor>
How much can I get the Dragonfly Array for?
<bofh>
NonlocalRaptor: probably at deep discount
Moistmelon has joined #kspacademia
egg|phone|egg has joined #kspacademia
* egg|phone|egg
confused self with denormals while trying to compute the binary64 ulp distance between arbitrary numbers
<egg|phone|egg>
Turns out 1074-1023 is 51 and not 52 so it's fine after all :-p
<bofh>
heh
<egg|phone|egg>
!Wpn bofh and Fiora
* Qboid
gives bofh and Fiora a hypothetical triangle
<Fiora>
wah
egg|cell|egg has joined #kspacademia
egg|phone|egg has quit [Read error: Connection reset by peer]
<egg|afk|egg>
Fiora: also, (belatedly), yes, that's roughly how I'd pronounce Atelier https://twitter.com/Helios748/status/896528697330618369 (vowels are a bit weird though, I'd say [atəlje], and the announcer seems to be saying [atɨljɛ] or something like that)
<kmath>
<Helios748> @FioraAeterna @mirihawke Interesting pronunciation of Atelier too. French pronunciation?
<egg|afk|egg>
Fiora: and yes, the announcer's vowels on "Shallie" are, uh, creative
<egg|afk|egg>
the lesser spotted vowel shift
egg|afk|egg is now known as egg
<Fiora>
egg: i am *amazed* at how he pronounced it. like. how do you screw up that bad
<Fiora>
"atelier" has many valid pronounciations so that's okay
<Fiora>
but "shallotte" is a literal onion and "shallie" is......... straightforward
<egg>
yeah
<Fiora>
also the games themselves aren't even consistent with the pronounciation of "atelier", which makes sense because neither are languages
<Fiora>
japanese, french, and english all do it differently (and inconsistently) despite it being a loanword
<egg>
Fiora: that pronunciation of shallie seems almost like swapping the vowels? it sounds like he's saying [ʃilɛj]
<egg>
Fiora: it's not a loanword in french (unless loanword from middle french counts) :-p
<Fiora>
i mean, yes. XD
<Fiora>
english phrasing is hard
<Fiora>
(the japanese also LOVE that word, like, so much. there must be at least 10,000 shops named <X atelier>)
<egg>
Fiora: I am semi-tempted to read that number as 10 given the french context :-p
<egg>
(periodic reminder that the CGPM recommendation is either , or . for the decimal mark, thin space for thousand sep, explicitly neither , nor . for thousand sep) :-p
<egg>
ah no it doesn't say thin space, it says "divided in groups"
<egg>
so the swiss-currency-style 10'000 seems to be compliant with CGPM 9 & 22?
<Fiora>
point vs comma has to be the worst international mis-standard ever
* Qboid
gives bofh, Iskierka, Fiora a slow smooth FPGA
awang has quit [Ping timeout: 183 seconds]
<NonlocalRaptor>
Someone just had to walk out because ITAR =\
<NonlocalRaptor>
Atmospheric data also
<egg>
!wpn NonlocalRaptor
* Qboid
gives NonlocalRaptor a feline king/scream hybrid
<egg>
bofh: huh, on my machine Mathematica underflows at 10^-(10^15) or thereabouts
<egg>
big eggsponents are big
<bofh>
so 2^-(3.3*10^15)... huh, that strikes me as them using an arbitrary precision lib since no hwfloat has that exponent range... but in that case, why is there underflow at all?
<egg>
bofh: dunno
awang has joined #kspacademia
<egg>
bofh: I think it uses some amount of hardware float (unless you specify a precision in which case it's some weird decimal softfloat)
<egg>
bofh: because the MachinePrecision is 15.9546 sig dec :-p
<egg>
bofh: so it has binary64 + Really Big Eggsponent
<bofh>
huh. that's... strange
<bofh>
I guess it makes sense, be fast in binary64 range and only use softfloats when exceeded
<bofh>
(also it will eternally bother me that it's not called binary53)
<egg>
bofh: well it keeps using binary64 for the mantissa afaict
<egg>
bofh: you can use softfloats, but that's separate (that's if you say N[whatever, numberOfSigDec], or WorkingPrecision->stuff etc.)
<egg>
bofh: Log2[-Log2[$MinNumber]] is eggstremely close to 52
<egg>
bofh: they're using a binary64 for the mantissa and another one for the eggsponent
<bofh>
huh, one tends to overflow, the other underflow. makes sense.
<bofh>
for x^2-y^2 vs (x-y)(x+y)
<egg>
(errors in ULPs, all evaluated to a target precision 10 sig dec. with appropriate binary64 roundings)
<egg>
bofh: uh, there's no underflow or overflow? it just shows errors
<egg>
bofh: also to get that I wrote some silly code that computes ULP distances between arbitrary numbers, since that's all done in arbitrary precision :D
<bofh>
egg: okay, but the errors concentrate either at very tiny numbers or very large ones
<bofh>
also nice
<bofh>
mpfr or some other lib?
<egg>
bofh: just Mathematica
<egg>
N[whatever, 10] will evaluate whatever to 10 sig dec. assuming whatever doesn't have some Ns in it already
<egg>
(not to be confused with N[whatever], which will evaluate whatever *in* machine precision, rather than in arbitrary precision *to* machine precision)
<bofh>
Ahh, so *that's* the difference. It confused me the last time I tried dealing with Mathematica (which admittedly was in 2009)
<egg>
bofh: yeah, like much of everything in there the APIs have dubious consistency
<egg>
bofh: specifically, N[stuff, d] will evaluate to d sig. dec., and will warn if it requires too much precision in internal calculations to reach the target (by default that's 50 sig. dec., can be changed as needed)
<bofh>
tho I disagree with the last one on the bottom
<bofh>
erm nevermind
<bofh>
wrong link
<egg>
NonlocalRaptor: you should come to CH sometime :-p
<egg>
to observe the cows probably
<bofh>
egg: what's the cheapest way to get from Geneva to where you are, anyhow?
<egg>
bofh: hm, from geneva to zurich? I'm not sure; the train is certainly less annoying (trains work really well here), not sure how the aeroplane pricing is but iirc it's very cheap; I have a halbtax so for me the train is cheaper than for you probably
<bofh>
yeah, I do know trains work well, I'll have to see what itinerary I got setup for eucas & see how much bitching I need to do w/the department
<bofh>
(esp. since I also want to possibly hit videolan dev days in paris for a day or two)
<Iskierka>
;c (1/2)^(3/2)
<kmath>
Iskierka: 0.3535533905932738
<egg>
bofh: also with the train you can see the countryside :-p
awang has quit [Ping timeout: 383 seconds]
<bofh>
ooooooh that is a nice property
<Iskierka>
... if an orbit with e = 1 and finite apoapsis has about 1/3 the period of a circular orbit of the same apoapsis, does that imply they have the same average speed?
<egg>
bofh: I mean with the aeroplanes you can see the mountains* [* if there aren't too many clouds; not for long anyway, it's a tiny flight; restrictions may apply]
<Iskierka>
get your own plane and see your own mountains
<bofh>
I don't have a pilot's license!
<bofh>
(yet, in any case)
<Iskierka>
Why would that be a problem? "get" is rather generic
<egg>
bofh: half price on pretty much everything, very quickly worth it (there's also a generalabo, everything free, but much costlier, not really worth it unless you travel a lot within switzerland)
<bofh>
yeah, from what I can see it becomes a good idea if I'm even a month there, possibly less. sadly I'm there a week.
<egg>
bofh: clearly you should stay three more weeks :-p
regex has quit [Remote host closed the connection]
awang has joined #kspacademia
<egg>
Iskierka: hmm, orbits with e=1 are nasty; no idea how the average speed behaves anyway
<egg>
bofh, NonlocalRaptor?
<Iskierka>
at or at least close to e=1. Random spinoff thought from reminding myself what angle an observer sat will move by in a dive so I can plan for the crash to be visible by the spectrometer
<Iskierka>
actually
<Iskierka>
No, thought by throwing pi in I might show that it is the same average speed, but doesn't go in tidily
<Iskierka>
it does appear to be weirdly close though
oeuf has joined #kspacademia
egg has quit [Ping timeout: 204 seconds]
oeuf is now known as egg
<egg>
!wpn whitequark
* Qboid
gives whitequark a Лифшиц cipher
<egg>
!wpn котя
* Qboid
gives котя an oxygen COME FROM
<awang>
intercal?
<bofh>
egg: orbits with e=1 *are* nasty, but doesn't the Vis-Viva equation not depend on e? just GM, distance and semi-major axis?
<egg>
bofh: yes, but how does average speed even behave
<bofh>
of course this presents an issue where r *does* depend on e, which is where the fun comes in
<egg>
bofh: integrating the speed over the orbit to average it seems unpleasant
<bofh>
well semi-major axis is presumably constant by assumption, so average speed feels like it should be proportional to the average of the inverse square root of distance?
<bofh>
so this boils down to finding the average distance between the two bodies, or am I being totally off here?
Technicalfool has quit [Read error: -0x1: UNKNOWN ERROR CODE (0001)]
<Iskierka>
point was SMa does vary in the orbit I was looking at; it's exactly half (killed all velocity into freefall)
<bofh>
ahh.
<egg>
bofh: funnily enough I have completely forgotten why I wanted to evaluate arc th y/x >_>
<egg>
(it's likely that I shouldn't evaluate it because there are further simplifications upstream that are necessary for the same reason)
<egg>
but hey, at least I know how I should evaluate it :D
<bofh>
18:44 <@egg|zzz|egg> bofh: so, fun question: assuming you want to compute arc tg (y/x) close to (and below) y/x = i; should you just divide and use the arc th, or use either of the arc sh or arc ch
<bofh>
was the only context I have in logs actually