raptop 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> egg|nomz|egg: generally if your eyes are dewing over, that's not the weather. | <ferram4> I shall beat my problems to death with an engineer. | We can haz pdf
<egg>
!acr -add:NAQU Notice Advisory to QZSS Users
<egg|zzz|egg>
"In an SDR, hardware and software technologies are used in parallel Part C | 13.2 to generate a reconfigurable solution. The goal is the ability to update the system by software changes. In any case, an analog front-end down to the ADC will be present: on the processor side, an field programmable gate array (FPGA), an MP or a general purpose computer or a mix of such a processor could be used"
<egg|zzz|egg>
this acronym list is a nice mix of aerospace acronym and eleggtronics acronym
<egg|zzz|egg>
!acr -add:SDR Software Defined Radio
<galois>
Definition added!
<egg|zzz|egg>
whitequark: "In GNSS software receivers, a general purpose processor is used: e.g., an Intel PC or an advanced risk machine (ARM) processor. All signal processing is programmed in a higher programming language like C/C++" << I don't think that's the right spelling of "RISC"
<whitequark>
lol
<egg|laptop|egg>
ECOM?
<galois>
ECOM: Empirical CODE Orbit Model
<egg|laptop|egg>
UmbralRaptop: mofh: oh no, there's a GPS draconitic year
<egg|laptop|egg>
which is about 14 days shorter than the sidereal one
<egg|laptop|egg>
!acr -add:MCAS Maneuvering Characteristics Augmentation System
<galois>
[WIKIPEDIA] List of Galileo satellites | "This is a list of past and present satellites of the Galileo navigation system. As of February 2019, 22 satellites are declared operational, 2 satellites placed in non-nominal orbit are declared in testing, and 2 satellites are currently not usable. The two GIOVE prototype vehicles were retired in 2012..."
<egg|laptop|egg>
... amusingly the one that has yet to be launched is the GB one
<UmbralRaptop>
Will it launch before 2019-10-31?
<egg|laptop|egg>
no, scheduled 2020
<egg|laptop|egg>
there's no Croatian name, because the drawing competition was before 2013
<mofh>
egg|laptop|egg: meow
<egg|laptop|egg>
mofh: CDDIS
<egg|laptop|egg>
mofh: if you host a mirror of CDDIS and friends on ~pbarfuss, is it WatDIS
<mofh>
LOL
<egg|laptop|egg>
(of course the server then has to be written in WATFOR)
<mofh>
...I think I might still know a physics prof there that might be interested in th--ROFL
<mofh>
I keep forgetting that's actually a think that eggsisted once.
<egg|laptop|egg>
also,
<egg|laptop|egg>
WATFIV?
<galois>
WATFIV: WATerloo FORTRAN IV
<UmbralRaptop>
Ꙩ_ꙩ
<egg|laptop|egg>
mofh: anyway, WatDIS,
<egg|laptop|egg>
mofh: also underlying-1-step-method
<egg|laptop|egg>
!wpn котеньке
* galois
gives котеньке a adobe floppy disk
<egg|laptop|egg>
mofh: derp, I accidentally referred to 2000-01-01T00:00:00 (UTC) Check failed: (year - 1972) * 2 + 1 < leap_seconds.size()
<egg|laptop|egg>
s/2000/2020/
<galois>
egg|laptop|egg meant to say: mofh: derp, I accidentally referred to 2020-01-01T00:00:00 (UTC) Check failed: (year - 1972) * 2 + 1 < leap_seconds.size()
<kmath>
<johncarlosbaez> When I first went to Paris I fell in love with the city, like most people do. And I started trying to visit all th… https://t.co/Fx9I2l7s4R
<egg|laptop|egg>
!acr -add:IGGBSPM Institute of Geodesy and Geophysics BeiDou Solar Pressure Model
<galois>
Definition added!
egg|laptop|egg has quit [Ping timeout: 198 seconds]
egg|laptop|egg has joined #kspacademia
<egg|laptop|egg>
!acr -add:CNSC China Satellite Navigation Conference
<galois>
Definition added!
<egg|laptop|egg>
mofh: whitequark: I feel like I am going down a very deep rabbit hole of GNSS madness
<egg|laptop|egg>
mofh: also I think I'm going to use the entire reduced ECOM model instead of just radial bullshit, it's just as easy,
<egg|laptop|egg>
oh hey these CNSC proceedings talk about PANDA
* UmbralRaptop
tries to get an understanding of Christoffel symbols to solve a problem.
<mofh>
reduced ECOM model?
<mofh>
egg|laptop|egg: I've been to a good chunk of those streets, at least the ones in 5e
* UmbralRaptop
sees in the solution that no Christoffel symbols are used, and the answer is completely different
<_whitenotifier-3d18>
[Principia] pleroy commented on issue #2134: Random Crashes - https://git.io/fjmKz
<egg|laptop|egg>
ECOM?
<galois>
ECOM: Empirical CODE Orbit Model
<egg|laptop|egg>
mofh: ^ an SRP model
<egg|laptop|egg>
UmbralRaptop: what is the problem
<egg|laptop|egg>
UmbralRaptop: also try repeatedly swatting mofh for physics help
<B787_300>
UmbralRaptop: so if we can combine observations from radio waves to be able to see stuff super far away (the EHT) why cant we do it with optical obs?
<UmbralRaptop>
B787_300: I think too dim, mainly? Chara has some silly bright magnitude limits
<B787_300>
one of the benefits of combine was getting a super large apererture right? which allowed for better resolving, so why not do that to direct image exoplanets
<_whitenotifier-3d18>
[Principia] pleroy closed pull request #2135: Avoid dangling pointer to the plugin - https://git.io/fjmit
<_whitenotifier-3d18>
[Principia] pleroy pushed 2 commits to master [+0/-0/±10] https://git.io/fjmib
<_whitenotifier-3d18>
[Principia] pleroy 649c9cb - Remove plugin_ from everywhere.
<_whitenotifier-3d18>
[Principia] pleroy 0213c4b - Merge pull request #2135 from pleroy/Crash Avoid dangling pointer to the plugin
<UmbralRaptop>
I should go hunting for the studies, but the idea exists. I *think* there's been a lack of work because of practical problems.
<UmbralRaptop>
hrm. Back of the envelope calculations suggest that you'd want at least a 10-15 km baseline for planets around prox cen
<UmbralRaptop>
Also, I vaguely recall something about needing or wanting to do this in orbit?
<B787_300>
could it be a noise floor issue?
<UmbralRaptop>
possibly
* UmbralRaptop
glares at the atmosphere
<B787_300>
like to get the noise low enough you need cyrocooled sensors (or maybe this is where you would want the super sensitive sensors like photo multiplier tubes)
<_whitenotifier-3d18>
[Principia] pleroy opened pull request #2136: Render the guidance node in the adapter and not in the flight planner - https://git.io/fjmP4
<kmath>
<bofh453> <@cgranade> getting Fortran 90 to output to YAML.... that's a fun challenge.
<mofh>
20:46:08 < egg|laptop|egg> UmbralRaptor: mofh: is there a sensible way to do simulated annealing on a continuous parameter space
<mofh>
I mean yes and no. Like there *is* a natural way but I'm pretty sure it just replicates gradient descent in terms of covergence properties and optimization rate.
<mofh>
Well, constrained gradient descent, at least.
<mofh>
As for the YAML... NWChem output.
<egg|laptop|egg>
mofh: should i use deмcmc instead
<egg|laptop|egg>
(parameter space size is 5)
<egg|laptop|egg>
I should be getting residuals of a few m, getting more like 50 with the best set of magic numbers I've tried throwing at the shitty optimizer
<mofh>
honestly I'd prolly use constrained least squares for a parameter space size that smol, lol.
<mofh>
deмcmc is likely to work better than SA I *think*
<egg|laptop|egg>
hm, the norm i'm using is a weird L-infinity-of-L2
<egg|laptop|egg>
(because the residual in position necessarily starts at 0, L2 over the day is not really a quantity i want to work with)
<egg|laptop|egg>
i guess i can write yet another deмcmc
<egg|laptop|egg>
mofh: ... or i could templatize the eggsisting on,
<egg|laptop|egg>
s/on,/one,/
<galois>
egg|laptop|egg meant to say: mofh: ... or i could templatize the eggsisting one,
<egg|laptop|egg>
mofh: what sort of population size do i want
<egg|laptop|egg>
std::thread::hardware_concurrency(), of course,
<egg|laptop|egg>
:-p
<mofh>
22:22:09 < egg|laptop|egg> hm, the norm i'm using is a weird L-infinity-of-L2
<mofh>
what the hell?
<egg|laptop|egg>
mofh: max distance from the CODE orbit, over the day
<egg|laptop|egg>
L-infinity is meaningless in position space obviously
<egg|laptop|egg>
but L2 over the orbit is really messy to interpret
<egg|laptop|egg>
I could take the distance at the end of the day but that feels wrong, I might run into weird resonances that hide stuff
<egg|laptop|egg>
!wpn mofh
* galois
gives mofh a isomorphic eigenvector with a compiler attachment
<mofh>
Huh, SIAE is happening this year at Le Bourget from Jun 17th - 21st.
<mofh>
I should figure out some way to sneak in for the non-public bit. Shouldn't be too hard.
<mofh>
Also L2 over an orbit feels like it'd be subject to possibly large error at given positions in orbit
<mofh>
Like it's kind of meaningless too tbh.
<mofh>
Or at least not any better than L1
<egg|laptop|egg>
!u
<galois>
No info for U+0001 (I only know about Unicode up to 9.0)
<egg|laptop|egg>
mofh: yeah, hence L infinity
<egg|laptop|egg>
note that I'm not optimizing the initial conditions of the orbit
<egg|laptop|egg>
so the residual mostly increases with time
<egg|laptop|egg>
which makes L1 or L2 weird
<mofh>
I mean we're only concerned with it over a single orbit, no?
<egg|laptop|egg>
two
<egg|laptop|egg>
but yes
<egg|laptop|egg>
"The article was only written, because Leos Mervart wanted Beutler to learn LaTex. In this respect, the article was a success. E-Mail by Leos Mervart commenting this slide: Remember not the sins of my youth or my transgressions (Psalm 25:7)"
<mofh>
LOL
<egg|laptop|egg>
!choose implement another deмcmc|sleep
<galois>
egg|laptop|egg: Your options: implement another deмcmc, sleep. My choice: sleep
<egg|laptop|egg>
mofh choose implement another deмcmc|sleep
<mofh>
implement another deмcmc
<egg|laptop|egg>
!choose listen to the bot|listen to mofh
<galois>
egg|laptop|egg: Your options: listen to the bot, listen to mofh. My choice: listen to mofh