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.
<SnoopJeDi>
...is the character ♠ meaningful in any numerics/sorting context? Just discovered that the {{val}} template on Wikipedia puts one in every instance, leading to a comical "correction" by the EmojiOne plugin
<UmbralRaptor>
!u ♠
<Qboid>
U+2660 BLACK SPADE SUIT (♠)
<Iskierka>
I've noticed it in google searches before (and screwing up text matching) but it never renders so I assume it's not meaningful. just some nonsense
<UmbralRaptor>
Fairly low number for something with emoji presentation.
<SnoopJeDi>
well the plugin just finds anything it can replace with an SVG, that character predates emoji by a good bit but
<UmbralRaptor>
¯\_(ツ)_/¯
<SnoopJeDi>
it seems like maybe the code driving the {{val}} template (which is apparently Lua, TIL wikipedia uses any) is putting them there
<UmbralRaptor>
!u ⚡
<Qboid>
U+26A1 HIGH VOLTAGE SIGN (⚡)
<soundnfury>
!wpn UmbralRaptor
* Qboid
gives UmbralRaptor a [DATA EXPUNGED] domain
* soundnfury
is still trying to figure out why equifax exists in the first place
<SnoopJeDi>
$
<icefire>
????:)????
<soundnfury>
like, there's three (?) companies that have a special privilege from the government which allows them to basically collect data on everyone and use it to determine who has access to capital, and because they have this special privilege they end up with an oligopoly, so they don't have any need to be even slightly good at it
<soundnfury>
and somehow this is just a normal thing that happens and neither the anti-government nor the anti-corporation activists have burned _any_ of them down?
<UmbralRaptor>
soundnfury: AIUI, they've had the FBI ad a major customer since J Edgar Hoover was young.
<soundnfury>
UmbralRaptor: oh, so it's just normal governments-being-totalitarian-in-creepy-new-ways. Gotcha.
<icefire>
there's easier targets to ridicule that gets you more attention
<bofh>
UmbralRaptor: so thinking about Equifax at this point consistently gives me a headache, it's kind of almost neat.
<bofh>
(also it says a lot about me that my first response to that was "uh why did you forget the FEC?")
icefire has quit [Remote host closed the connection]
<UmbralRaptor>
Also, while we're screaming in terror, CCleaner.
<bofh>
so I honestly don't understand what the use case for CCleaner is.
<SnoopJeDi>
the Registry is a cruel master
<UmbralRaptor>
Hypothetically, if all hope were lost and the only remaining allies were the Ruinous Powers, which would you choose? Asking for a friend.
* UmbralRaptor
is tempted to tweet that line. As a poll with the obvious 4 choices.
<Fiora>
bofh: windows generates insane numbers of temp files it never deletes
<Fiora>
or, at least, in the win7 days it did
<Fiora>
and winxp
<Fiora>
and win98
<Fiora>
and win95
<bofh>
okay true, but those don't really affect much other than taking up disk space.
<kmath>
<Newpa_Hasai> Classical Mechanics is convincing me I don't understand basic Calculus.
<UmbralRaptor>
FluffyFoxeh: Given things like that Intel management engine, and admin/admin Equifax login, I feel like nagging users about passwords is; t addressing the biggest problems.
<UmbralRaptor>
egg|zzz|egg: relatedly(?) I'm starting to see why people like Einstein notation.
<egg|zzz|egg>
UmbralRaptor: yeah you might gain from that too, since it also forces you to be consistent with your functionals vs. vectors
<UmbralRaptor>
egg|zzz|egg: Fun fact: I learned about the math term "functional" ~7 hours ago.
<egg|zzz|egg>
UmbralRaptor: form?
<egg|zzz|egg>
UmbralRaptor: coveggtor?
<egg|zzz|egg>
UmbralRaptor: that thing that is linear and maps a vector to a scalar
<UmbralRaptor>
Classical Mechanics lecture
<UmbralRaptor>
Well, to a limited eggstent. Specifically the "function of a function" part.
<egg|zzz|egg>
UmbralRaptor: so the proper way to do derivatives of a scalar-valued function in multiple arguments yields a form; the gradient is obtained by transposing the form (and requires an inner product)
<egg|zzz|egg>
UmbralRaptor: ah no I don't mean functional in that sense
<egg|zzz|egg>
UmbralRaptor: so eventually, df = g dt makes sense again, it just means something completely different from the naive thing with infinitesimals! (and you can say things like df = g dx + h dt or whatever)
<UmbralRaptor>
"any vector space V has a corresponding dual vector space (or just dual space for short) consisting of all linear functionals on V"
<UmbralRaptor>
er?
<egg|zzz|egg>
UmbralRaptor: yes, the dual of the space of |egg>s is the space of <egg|s
<egg|zzz|egg>
UmbralRaptor: see also the conversation with rqou above
<egg|zzz|egg>
UmbralRaptor: so a linear functional, linear form, 1-form, covector, whatever you want to call, on V (which is a vector space on the field K), is a linear function from V to K
<egg|zzz|egg>
UmbralRaptor: i.e. if λ is in V*, λ(v) is a scalar (in K) for all v in V, and λ(αv+w) = αλ(v)+λ(w) for all v and w in V and α in K
<egg|zzz|egg>
UmbralRaptor: in coordinates, that's a row vector: you multiply it by the usual (column) vectors
<kmath>
<Newpa_Hasai> Classical Mechanics is convincing me I don't understand basic Calculus.
<egg|zzz|egg>
bofh: how do you write the divergence using the covariant derivative in a coordinate-free manner? using the eggsterior derivative you have ⋆d⋆(F♭), but I don't see a nice way to say ∇_i F^i coordinate-freely
<kmath>
<stephentyrone> @FioraAeterna @rygorous @barrelshifter Nothing could be more punk than 2s-complement wrapping. Except graph algorit… https://t.co/7KMoGH75Ge
egg|zzz|egg is now known as egg|afk|egg
egg|cell|egg has joined #kspacademia
egg|phone|egg has quit [Ping timeout: 195 seconds]
<egg|cell|egg>
Bofh: oh it's just the trace of ∇F right
egg|phone|egg has joined #kspacademia
egg|cell|egg has quit [Read error: Connection reset by peer]
egg|phone|egg has quit [Client Quit]
egg|phone|egg has joined #kspacademia
egg|cell|egg has joined #kspacademia
egg|cell|egg has quit [Client Quit]
egg|phone|egg has quit [Read error: Connection reset by peer]
SilverFox has quit [Ping timeout: 383 seconds]
SilverFoxy has joined #kspacademia
SilverFoxy is now known as SilverFox
ferram4 has quit [Ping timeout: 383 seconds]
ferram4 has joined #kspacademia
<egg|work|egg_>
!wpn whitequark
* Qboid
gives whitequark a hypersonic vertex
<whitequark>
!wpn egg
* Qboid
gives egg a Pauli moth
* egg|work|egg_
fries the moth and gives it to whitequark
egg|work|egg_ is now known as egg|work|egg
<egg|work|egg>
!wpn котя
* Qboid
gives котя an actinium wombat/embedding hybrid
<soundnfury>
egg|work|egg: ones-complement is punker
<soundnfury>
!wpn egg|work|egg
* Qboid
gives egg|work|egg a diamagnetic git-rebase
<UmbralRaptor>
Today in being Nomal, my advisor commenting on impostor syndrome (wrt my struggles with classical mechanics) .
awang has joined #kspacademia
<APlayer>
Uh, is it appropriate to ask for advice on an EE topic here?
awang has quit [Ping timeout: 204 seconds]
<UmbralRaptor>
Absolutely. We might not be able to help, though.
<bofh>
go ahead
awang has joined #kspacademia
awang has quit [Ping timeout: 183 seconds]
awang has joined #kspacademia
awang has quit [Ping timeout: 200 seconds]
<APlayer>
Uh, sorry, had to go AFK. Anyway, I am back now.
<APlayer>
So I have my Raspberry Pi 3 hooked up to a powerbank. And now I want to be able to turn it all on and off. When it's all off, I just press the button on my powerbank and it starts delivering current, powering up the RPi setup.
<APlayer>
However, when I want to turn it off, I shut down the RPi, but it keeps drawing some current for the display and some low power CPU mode, unless I physically cut power somehow. And my powerbank thinks something still needs current and keeps delivering it.
<APlayer>
I tried to build in a button that breaks the flow when pressed, but I need to hold it down for about 12 sec to turn off the powerbank.
<APlayer>
The question is, is it possible to create a relay or something circuit that does this sort of thing for me? I also read somewhere that the RPi has a GPIO pin that goes low when shut down.
regex has quit [Ping timeout: 204 seconds]
<Iskierka>
the screen is drawing power, so yes, the powerbank is correct that something needs current
<Iskierka>
interrupt the supplies for them when pi powers off to disable it all easily
regex has joined #kspacademia
<Iskierka>
it's basically measuring continuity and/or current. Until one or both of those is interrupted, it figures okay, I'm connected, I'm needed
<Iskierka>
(would guess more likely current so it will shut off when a phone is fully charged)
<APlayer>
Well, the question is, how would such a circuit look like? A momentary or permanent interruption both seem easy enough, but what about a short time?
<APlayer>
(I.e. 15-20 sec seems appropriate)
<Iskierka>
depends on how you power off the pi and how the display is connected and set up and what type of display
<Iskierka>
as for the pi itself the light will remain on but it is off. It doesn't have the ability to directly interrupt its own power
<Iskierka>
the only draw should be the LED
<APlayer>
The display draws power directly from the battery, and the Pi is connected via USB from the display, so it draws power through it too
<APlayer>
I just want to interrupt the flow for a few seconds and then reset the circuit. I /can/ do it by hand every time, but it seems annoying compared to a short button press or even an automatic system
<Iskierka>
and how is the pi powered off?
<APlayer>
Via the shutdown option of the OS
<APlayer>
I currently have the "backup solution" that I will use the button, but I need to press it for a long time to make sure the battery shuts down, else the whole setup powers up again once the current returns
<Iskierka>
I cannot find a GPIO pin that is meant to go low on shutdown (and really high would probably be preferable)
<Iskierka>
as in that case it seems you would just want a high-current transistor to cut off power when high, I believe
<Iskierka>
(better to poke Ellied about transistors though)
<Iskierka>
which could potentially still have a problem of not holding low for long enough to turn off the pack ... but that may be solvable by a capacitor or several to hold it high after power off
<APlayer>
My first thought was (no EE experience at all, forgive me any blunders here) to use some sort of transistor and capacitor setup for that
<APlayer>
Would that be possible? Or do I need a real timer circuit?
<Iskierka>
if you had a pin that went high after power off that would be a simple circuit that should have the desired effect (if the capacitor can hold it long enough)
<Iskierka>
though it would need slightly more involved design to determine exactly which bit needs to connect where
<Iskierka>
which is where Ellied would be useful
<APlayer>
I can't currently find the forum and post where I read that some guy found a GPIO pin going low on shutdown, and probably this is either an OS thing or individual to them, since I found no mention too
<APlayer>
No other mention*
<Iskierka>
it gets more complicated if it doesn't put any GPIO high (though that might be possible with a modified shutdown script? I don't know if GPIO holds state after shutdown), and instead holds low. This would be a problem as when the pi is powered off, it will remain low and the transistor will never allow connection
<Iskierka>
so you would need some other override until that pin gets raised in boot
<APlayer>
I am certain nothing goes high on shutdown, though
<Iskierka>
hence would be a thing to solved
<APlayer>
I could hook things up to the battery directly, and somehow initiate the circuit using a short button press?
<Iskierka>
If it holds state, you can raise something during shutdown to use that. If it doesn't, but something goes low, it should be possible to invert it using the ground and 5v, but I am tired and cannot think of what the circuit must look like to not short
<APlayer>
It seems weird to initiate a capacitor/transistor circuit using something that is supposed to be powered down using that very same circuit
<APlayer>
That is, wouldn't that interrupt itself once it cuts power to the Pi and thus to the pins?
<Iskierka>
that's what the capacitors would be for. Hold the voltage from that pin long enough to ensure that the pack isn't still trying to apply current and won't do so once the transistor opens again
regex has quit [Ping timeout: 186 seconds]
<Iskierka>
after that point, you WANT the connection to re-establish, so that you can power on with the powerbank's button press
<APlayer>
Ah, so while the circuit is running, it draws power from the capacitor, and once it runs out, it automagically resets. That's neat!
<Iskierka>
this, of course, is all for if you want to keep the powerbank unmodified. There's more options if you can open it and directly interface but that has potential dangers if you're not aware of them and we can't say how to do anything with that until we learn the internal layout and what's available
regex has joined #kspacademia
<SnoopJeDi>
don't let the magic smoke out
<Iskierka>
it may be simpler to actually modify the pack, potentially, but will need more research on the pack itself to figure out how to manipulate it
<APlayer>
I have no idea about the inner workings of the power bank...
<Iskierka>
Especially not lithium magic smoke
<Iskierka>
that magic smoke is hot and angry
<APlayer>
I only know the model and that it's from intenso, but it neither came with schematics, nor I believe I would get them if I kindly mailed the manufacturer
<APlayer>
:P
<Iskierka>
APlayer, if you're not gonna use the UART for anything that would work
<Iskierka>
just need to design the circuit that will respond to the pin
<APlayer>
If I am ever going to use the pins, I may just unpin it while I use it
<Iskierka>
and then rip apart the USB cable to inject a transistor
<APlayer>
That was the plan
<APlayer>
Well, thanks a lot! I will return here tomorrow, but for now I've got to go. Please ;tell me if there is anything more to add right now, I will meanwhile try to reseach that pin's behaviour.
<kmath>
<DSA_Phoenix> "Pass this horribly unpopular bill or we'll give you something 75% of voters are demanding" is an interesting hill… https://t.co/wJKloJPkRB
<Ellied>
somebody say transstors?
<Iskierka>
figuring out if a high-current one can be used to interrupt a powerbank powering a pi (and screen that's keeping the draw high, hence need for interruption)
<Iskierka>
we apparently have a pin that is high while running & not used and goes low on shutdown so that's a place to start
<Ellied>
We do? Are we sure that it doesn't go low until after the Pi is fully done shutting down?
<Iskierka>
that is a point but something we can play with once we have something that will definitely cut off the moment it goes low
APlayer has quit [Ping timeout: 204 seconds]
<Iskierka>
besides, we're already potentially looking at capacitors in there, put one to keep the power going for a few seconds for complete shutdown
<Ellied>
spose so
<Iskierka>
since I've never used a transistor before I wouldn't risk drawing the circuit diagram to do this though. Or recommending transistor choice
<Ellied>
Pi draws up to 2.5A constant, possibly more in short peaks. I'd find one with at least 5A continuous rating and 10A non-repetitive peak.
<Ellied>
That's a little more than just a 2N7000; could be a silly giant BJT or a garden-variety power MOSFET
<Ellied>
If I were setting this up I'd probably just use the IPP80N06S4L-05, which is overkill (80A continuous) but not very expensive anyway (like a buck) and guarantees a nice low on-state resistance
<Iskierka>
I mean for a one-off that would definitely be perfectly acceptable price
<Ellied>
That's probably a little excessively specific; it's easier to go to Aliexpress (or your local corner store, if you're lucky enough to have one) with some general requirements and get what you can find that's cheap and fits
<Iskierka>
And what would the circuit look like? Pin (probably indirectly*) to centre, and the other two fitted along the power wire? Does direction matter?
<Iskierka>
*and to allow the system to be able to turn on easily after, we had the thought it probably needs to be inverted to low-on high-off, so with no power it's low and will allow current & turn on?
<Iskierka>
bbl
<Ellied>
Low-on-high-off is easy, just use a p-channel transistor
<Ellied>
there are a lot of configurations that would work, I guess I'll wait for APlayer to come back to talk about them
<Ellied>
I love the number of IPP80P03P4L-07. It's like, gosh, I wonder what polarity that one is??
<Ellied>
If I ever design a pair of mirror-image FETs, they'll be something like 2N20N88N2N4N-N and 2P20P77P2P4P-P
awang has joined #kspacademia
<Iskierka>
since I've never looked at the use of transistors polarity is not a thing I knew about but I think I've learned that part of how you make it work :p
<Iskierka>
I suppose a transistor could technically be used in a similar way in my RC car for pi control, but it is a straight-connected 12V battery with 25A esc. That'll probably be A) rough on most transistors, B) very scary if I get it wrong
<Iskierka>
so prebuilt on-off switches are more convenient and good enough for me
<kmath>
<minwadhwa> Are you a woman undergrad interested in space exploration/aerospace? Here's an exciting opportunity you don't want… https://t.co/5Eh9gvMlij
<Iskierka>
not sure that would apply to me. Plus, guessing US-only?
icefire has joined #kspacademia
<UmbralRaptor>
blarg
<egg>
!wpn whitequark
* Qboid
gives whitequark a backwards-compatible snail with a diapsid attachment
<egg>
!seen котя
<Qboid>
egg: I haven't seen the user котя yet.
<UmbralRaptor>
egg: Why am I attached to a snail?
<egg>
UmbralRaptor: I don't know, maybe you tried to eat the snail? diapsids do that sometimes
<egg>
whitequark: any котяpics? or publications in J. Cat Cogn.?
<whitequark>
mmmmmmmm not really
<whitequark>
i don't take pics generally
<egg>
huh, we apparently have 75 kilolines of C++ now https://forum.kerbalspaceprogram.com/index.php?/topic/162200-wip122-and-13-principia%E2%80%94version-ces%C3%A0ro-released-2017-09-20%E2%80%94n-body-and-extended-body-gravitation-axial-tilt/&page=33&tab=comments#comment-3175940
<kmath>
<Lioncache> For example: std::variant<std::string, bool> k = "lol"; initializes the bool value portion. Not the std::string.
<UmbralRaptor>
egg: @eevee was onto something.
<egg>
UmbralRaptor: well, possibly? I mean it will make most GR calculations a special case (and also they're always done in coordinates, though Einstein notation helps with forcing you to write somewhat sensible things)
<egg>
UmbralRaptor: the sad thing is I immediately understand why it prefers that conversion
<SnoopJeDi>
I don't like the Earth flyby part for downlinking data and it's probably invoking too many tooth fairies for ADCS and optics for each individual sat, but the idea is cute
egg|phone|egg has joined #kspacademia
<Fiora>
egg|zzz|egg: i might have run out of characters