« first day (2341 days earlier)      last day (2589 days later) » 

2:00 PM
SEASON THREE IS HERE
 
@AccidentalFourierTransform ALL OF IT?
I ONLY HAVE EP1
 
nah
ep1
that felt very 2011
 
Trying to debug this CPU is ridiculous
I gotta debug by looking at individual bits lol
literally looking
 
>R registers
Damn MIPS
 
Lol
MIPS is easy, that's the advantage
I only get to design my own x86 CPU on Advanced Computer Architecure
and that's in like 3 years
 
2:07 PM
plz
Does MIPS even have a floating point unit
 
Mine doesn't :D
Who needs float anyway
 
good luck doing math
 
I will implement floating point one day
Right now I just need my SUBI instruction to work
I'm either stupid and implement the instruction wrong or I am stupid and do not understand what this instruction should do
 
@BalarkaSen Just very vaguely
Something about a homomorphism from the fundamental group of the leaves to some diffeomorphism group?
It was too technical
@BernardoMeurer btw don't ever say the c word around kat
 
2:24 PM
Or C++?
or Just C
 
@BernardoMeurer If I flag that you're gone
Instantly
better delete it
 
@0celouvsky Thanks for that, also, who the hell flags just a word
Anyway, was that the C-word?
 
the one you deleted
I just remembered Kat really dislikes it
just an fyi
Michelle doesn't like it, but does't really care either
Kat dislikes it
 
I don't get what people have with that word
But that's good to know, thanks :)
@Slereah Do you know VHDL?
 
I think that's the reason people use Contacto(M) to denote the group of contactomorphisms, instead of Cont(M)
@0celouvsky Group of germs of diffeomorphisms, yes. Otherwise it's just a pseudogroup.
 
2:31 PM
There's a girl in the study room about 3 desks away who just got a Blue Screen of Death
This is what happens when you run proprietary malware, people
 
@BalarkaSen "contactomorphism", that's even worse than "symplectomorphism", and that's already a pretty terrible word :P
 
it's fun to spell
 
@BernardoMeurer I know of it
Contactomorphism is probably one of the worst word
 
I guess cont(M) can also mean continuous functions on M
for a mathematical reason
 
@Slereah You wouldn't know how to debug a MIPS CPU in VDHL I suppose?
@0celouvsky is Kat married?
 
2:36 PM
@0celouvsky what if my friends nick name is cunty?
can I say that?
 
Not a clue
 
@AccidentalFourierTransform I advise against it.
 
I feel like that kinda violates my rights
 
@AccidentalFourierTransform You complain about ACM violating your rights he'll violate something else
Your argument, he'll violate your argument, with logic
Oh god
I wish VHDL was a mandatory course for physicist
my life would be easier
 
Go to stack overflow
 
2:41 PM
They suck at VHDL
Also, it's a lot of code
 
and i'm just being lazy tbh
 
what the hell is wrong with you
keep things to yourself for once
 
@0celouvsky Hm?
 
@AccidentalFourierTransform See, that's the danger of making references - people might not get them.
(I didn't.)
 
2:47 PM
apologies accepted
 
Ok, if I understood correctly from here and also what I was taught back in my 3rd year QM and in susskind's, the direction of the vector on the bloch sphere corresponds to the direction I must align my detector in order to get 100% detection of the electron. In addition, the bloch sphere maps a ray in hilbert space to a vector on the sphere. Therefore the angle that appeared as the parameters on the Bloch sphere
are half of that for the corresponding ray in the hilbert space. Now my confusion:
 
@ACuriousMind I apologise, on behalf of you, to me for deleting his reference
 
The "spin vector" is defined to be $\langle \vec{S}\rangle =(\langle \hat{S}_x\rangle \langle \hat{S}_y\rangle \langle \hat{S}_z\rangle)$ where $S_i$ are the spin operators given by the pauli matrices under the standard basis. Because of the uncertainty principle, only one component is ever found with certainty thus we cannot speak of the "angle between two spin vectors" Now,
 
@Secret You should make a blog
 
@BernardoMeurer This time is different, I am actually asking a question, I just have not tag acuriousmind in yet because I had not finished writing the question
 
2:54 PM
@Secret I'm just speaking in general
 
suppose on the bloch sphere I have plotted the states $\lvert 1\rangle$ and $\frac{1}{\sqrt{2}}(\lvert 0\rangle+\lvert 1\rangle)$, therefore in the bloch sphere, the angle substended by these two vectors is $\frac{\pi}{2}$
 
So here's a stupid question
I sometimes see $U(n)\subset SO(2n)$ stated casually
But the determinant is not always going to be one. What's going on here?
On the other hand, I guess elements of $U(n)$ preserve orientation and are orthogonal w.r.t. the inner product induced on $\Bbb R^{2n}$ by identifying it with $\Bbb C^n$, so they should be in $SO(2n)$.
 
@ACuriousMind Since in the original hilbert space where the spin states live, angles between rays are not well defined due to the spin vector can only ever have one component without uncertainty, how should I correctly interpret the physical meaning of the angle $\theta = \frac{\pi}{2}$ substended by the two aforementioned vectors on the bloch sphere?
 
Hi all
 
3:06 PM
What would be a good (hopefully basic) definition of a 'coherent quantum state'?
 
@YashasSamaga math.stackexchange.com/questions/1896422/… so 2+2=4 in every base
 
3:21 PM
@BernardoMeurer If you have two vectors $v_1,v_2\in\Bbb R^4$, how would you go about constructing a $4\times 4$ matrix $A$ with $\ker A=\vee\{v_1,v_2\}$?
I would use Gram-Schmidt to orthonormalize $v_1,v_2$, then construct a projector onto the orthogonal subspace.
 
@0celouvsky What's this operator $\vee\{v_1,v_2\}$ ?
 
But that's not allowed.
@BernardoMeurer (Closed) linear span.
 
Ah
Hm
Yeah, I would use GS, orthonormalize and barabim-barabom
 
So $A (\alpha v_1 + \beta v_2) = 0$?
 
@Slereah Yeah
And likewise $Ax=0\implies x=\alpha v_1+\beta v_2$
 
3:25 PM
I'm sure there is some property we're missing
Can we assume $A$ is invertible?
 
How could it be invertible if it has a 2-dimensional kernel?
 
i.e. that $v_1, v_2$ are non-null
Oh
Yeah, true
 
$v_1=(1,-1,3,2)^t, v_2=(2,0,-2,4)^t$.
 
Is there a point in using smooth manifolds rather than analytic manifolds
Are there any spacetimes with a manifold that is smooth and not analytic
 
@0celouvsky Nice problem, tell Reb I dunno
 
3:27 PM
@0celouvsky Maybe it has identical eigenvalues
 
I'll see if DS knows
 
@Slereah Yes. Analytic manifolds are totally a lot harder to deal with
 
I do know how to solve it, but they haven't done GS yet
 
There's no partition of unity. Can't guarantee Riemannian structure, or literally whatever
 
So there must be an easier way that's conceptually more difficult
 
3:27 PM
Whaaaat
Aren't all analytical manifolds smooth, anyway?
Aren't they a subset of smooth
 
Yeah but you can't have analytic partitions of unity
 
Sure, so what?
 
You always define functions to have maximal regularity
 
@0celouvsky Well
Since she knows the values of the vectors
 
You have to forget the analytic structure if you want to do smooth manifold theory. That means you're ignoring the analytic structure
 
3:28 PM
she can write the equation for the kernel and do it my hand
it will just take a while
 
@0celouvsky Oh, is it due to the compact support
 
If you want partition of unity to be compatible with analytic structure you're screwed
 
@Slereah Yes
 
Yes
 
ic
 
3:29 PM
An analytic function that is zero on an open set is zero everywhere.
 
No analytic bump function
 
I'm talking to JD and he's telling me that "the universe is a black hole in reverse"
 
Well outside of $0$
 
I'm confused: Besse's book says that every compact, connected manifold has a metric with constant, negative scalar curvature.
 
What about the sphere
 
3:30 PM
@Danu Is that the Kazdan-Warner theorem?
 
Doesn't that violate the Gauss-Bonnet theorem in dimension 2 for the sphere?!
 
yeah sounds pretty odd
 
@Danu Oh, lol
 
@0celouvsky ehh.. no? Idk.
 
@Danu page?
 
3:30 PM
123; thm 4.32 (i)
 
what is the dimension bound here
I'm pretty sure you need $n\ge 3$
 
I think you want dimension > 2
 
Doesn't mention dimensions...
 
I have Aubin' book
Checking
 
$$n \geq 3 \iff n > 2$$
 
3:32 PM
We know. We typed it simultaneously.
 
what if $n=2.5$?
 
What is $\Bbb R^{2.5}$
Does the transition map have to be smooth or is that just for smooth manifolds?
 
The sketch of the proof does seem to make $n\geq 3$ at some point
 
@Danu You need $n\ge 3$ for these things. I bet he puts a dimension bound in chapter 1 or something
 
K-W for dim 2 only says any function which is not everywhere negative appears as Gaussian curvature of some metric on $M$.
 
3:34 PM
Does a $\mathscr C^k$ manifold have a $\mathscr C^k$ transition function
 
By definition, yes.
 
@0celouvsky I don't think so.
 
Alright
 
@Danu In Aubin's Geometric Analysis book, where he proves the theorem, he has $n\ge 3$.
I don't know what to tell you besides that.
It's not unimaginable that there's a typo in Besse.
 
Like I said, the proof does seem to bring in the assumption anyways.
 
3:37 PM
Hm, what's a $C^0$ manifold
is the cube a $C^0$ manifold
 
Surface of a cube or the filled in cube?
 
Surface
 
Yes. You can even give it a smooth structure.
 
Of course it's a topological manifold @Slereah---it's just a sphere.
 
How very uncubic
Well yes, but I am wondering if there's a $C^0$ manifold that isn't smooth
 
3:38 PM
Yes, there are.
 
Yep. I think there's examples in dimension 4
 
Indeed
 
There are definitely examples in dimension 11 or something
 
Although I vaguely recall a theorem about making $C^k$ manifolds smooth
that might be 2D though
 
4 is the smallest dimension where it happens
 
3:39 PM
For any $k\geq 1$ you can make it smooth
 
Yeah, for all $k > 0$
 
But the question is, who actually knows how to construct such a thing?
 
I don't know why though
 
@Slereah It's in Hirsch, but the proof is just nasty
 
@0celouvsky I might know some if I go through my gauge theory notes
 
3:40 PM
@Danu yo what are these appendices he refers to
Appendix 37 or whatever
 
What madman has 37 appendices
2
Q: Manifold that is NOT smooth

user111970Could someone provide an example of a manifold that is not smooth? All manifolds that come to mind are smooth! By a manifold, I mean a hausdorff, second countable locally euclidean space.

It's not a very telling example
 
@0celouvsky ??
 
page 124
He quotes Appendix 37 for something
 
Hm
$\mathcal C^k$ is too plain and $\mathscr C^k$ is too fancy
 
idk
 
3:43 PM
Is there a C in between
 
@Danu Ah
theorem 37 in Appendix
 
@Danu I think the point is you can approximate C^k diffeomorphisms by C^k+1 diffeomorphisms for all k > 0. Should follow from some sort of smooth approximation.
 
@0celouvsky Nope. JEE isn't over; the first stage is over. JEE Advanced will be the next stage (Exam Date: May 21st).
 
O'neill uses $\mathscr C$ while Wald uses a plain $C$
 
@0celouvsky goes mum after hearing that JEE isn't over.
 
3:46 PM
@Slereah traditionally, $\mathscr C$ denotes $C$ with a topology
 
jee is horrifying. i dunwanahearaboutit
 
I would prefer a $C$ with just one top loop but not whatever goes on down there
Not too fancy
$\mathfrak{C}$
 
Let $\alpha$ be a $C^r$ differential structure on a manifold $M$, $r\ge 1$. Then for every $s>r$, there exists a compatible $C^s$ differential structure $\beta\subset\alpha$, and $\beta$ is unique up to $C^s$ diffeomorphism.
The proof seems to require more analysis than is healthy
 
What's a healthy amount of analysis
 
3:50 PM
@Slereah zero
2
 
Sadly I must agree
@ACuriousMind did u see that picture
 
@0celouvsky the one with the definition of a group in the middle of horrifying analysis? Yeah, a bit odd, but not extraordinarily so, imo
 
Can you define anything interesting on $C^0$ manifolds?
Or is it all gonna be shit
 
Depends on what interesting means.
 
Can you even define the tangent space properly?
Since the tangent space definition usually involves derivatives of charts
 
4:00 PM
 
where did you encounter that?
 
Nowhere. I just know these are things
 
Smoothing and microbundle theory is actually a complicated arena of topology I have no idea about
 
@BalarkaSen has a priori knowledge
 
4:07 PM
How would one find a vector in $\Bbb R^4$ orthogonal to $(1,-1,3,2),(1,0,-1,2),(-2,0,0,1)$?
 
dot and solve linear equation
 
yeah p. much
 
That's three equations for 4 unknowns
2
 
Well then you will get a whole subspace of 'em
lucky you
 
ofc, cause there are lots of vectors orthogonal
a whole R^1's worth of
 
4:08 PM
Aha
 
I mean pick the unit vector, I guess?
Add the condition $|v| = 1$
 
why's everyone beating me to stuff I am saying today
 
You gotta step up your game
 
@Slereah nah it doesn't matter
whichever one has less fractions is better
 
4:20 PM
$(1,20,5,2)$
that was too much work, damn
 
Is there a Long Circle
The compactified long line
 
who knows
 
I assume it would be fairly long
 
one pt compactification of the long line is the long ray iirc
 
@ACuriousMind @BalarkaSen Are you familiar with the notion of an "almost periodic function"
 
4:24 PM
no
 
nope
 
dangit
 
16
Q: Why is the Long Line not a covering space for the Circle

JSchlatherI know of several reasons why the long line can't be a covering space for the circle, but I'm more curious in what exactly goes wrong with the following covering map. Let $L$ be the long line and define $p: L \rightarrow \mathbb S^1$ by wrapping each segment of the line around the unit circle o...

heh
"every continuous map from the long line to the real numbers is eventually constant."
 
@Qmechanic Updated redaction on this one is much appreciated.
I would normally just go for "What happened to [user]?", which is shorter and feels more formally correct, but that's just a quibble.
 
4:43 PM
@Slereah Give me a quick hand, I'm trying to translate my MIPS ASM into C to do some testing of my implementation
I have this assembly code:
SUBI  R3,R0,#1
SUBI  R6,R0,#-1
SUB   R1,R3,R6
SUBI  R4,R0,#3
B.EQ  R4,R0,#7
SUB   R2,R3,R6
SUB   R3,R0,R6
SUB   R6,R0,R2
SUBI  R4,R4,#-1
ADD   R1,R1,R2
B.NE  R4,R0,#-5
SUB   R2,R0,R1
B     #0
Which I turned into this C code
 int main(){
     int r1 = 0;
     int r2 = 0;
     int r3 = -1;
     int r4 = -3;
     int r6 = -(-1);

     r1 = r3 - r6;

     printf("r1 = %d\nr2 = %d\nr3=%d\nr4=%d\nr6=%d\n\n", r1,r2,r3,r4,r6);

 loop: if(r4 == 0) goto end;
         r2 = r3 - r6;
         r3 = -r6;
         r6 = -r2;
         r4 = r4 -(-1);
         r1 += r2;
         printf("r1 = %d\nr2 = %d\nr3=%d\nr4=%d\nr6=%d\n\n", r1,r2,r3,r4,r6);
     if(r4 != 0) goto loop;
     r2 = -r1;
 end: printf("\nr1 = %d\nr2 = %d\nr3=%d\nr4=%d\nr6=%d\n", r1,r2,r3,r4,r6);
Is this right?
 
How would you write $z = 2t-t^2$ in terms of $t$?
 
It's not a bijection so I'm guessing you can't
 
@Slereah plox
 
I dunno m8
 
Oh no
@dmckee Come save me
 
4:50 PM
Oh quadratic formula of course.
 
@EmilioPisanty : Please make whatever edit you see fit.
 
I don't even remember what the long line is
 
@0celouvsky Of what?
 
Long line is basically $\mathfrak{c}$ copies of $[0,1]$
 
it's a bad space. who cares
 
5:02 PM
Topologists
 
only the polish guys
 
let's make a long spacetime
it will be very long
 
so, will that long line be in the time direction or the spatial drections? I reckon the time direction will be more interesting
 
@BalarkaSen you make fun of point set topologists but point set topology is basically all of analysis :o
Although the long line is a bit pathological even for analysts
 
Cantor sets are weird af
AC must be wrong
 
5:07 PM
@0celouvsky not making fun of them.
also you probably don't know point set topologists to say point set topology is all of analysis :)
they study really pathological spaces
 
@BalarkaSen I meant that analysis consists of mainly point set topology
 
Lol, my processor does $-1-2 = 4$
Clearly my son
 
Sure. But point set topology is far broader than the stuff analysis uses
 
and it contains non hausedoff and non T1 spaces, both are essential in describing time travel topologies
 
@BalarkaSen Ok, but when you said "polish guys" I was thinking of the 1920s guys
Banach and friends
 
5:10 PM
Here's an example of a point set topology paper : matwbn.icm.edu.pl/ksiazki/fm/fm101/fm101110.pdf
Banach and friends were analyst.
 
@BalarkaSen Urysohn?
 
@Qmechanic nah, it's good as is
it's just the voldemortsy formulation that really bugs me
 
@BalarkaSen that's pretty bad
My linear algebra TA did his dissertation on point set topology
 
I feel like Urysohn's work is rather close to analysis than topology. Anyway, he was Russian.
Contemporary Russian point set topology too was rather... ad hoc. Until Gromov's advisor arrived
I forget his name
Rokhlin
 
Never heard of him :/
 
5:13 PM
That's when the golden era of topology as in "geometry and topology" started.
 
@EmilioPisanty I have a question about PR boxes vs entanglement. I understood that PR boxes have stronger correlations than entanglement because the CHSH correlation function of PR boxes can go up to 4, beyond the Tsirelson bound of bell states. But physically speaking, how are they more nonlocal than entanglement, like what extra kind of correlation they can do that entanglement cannot?
 
@0celouvsky It's apparently a seminal paper on continuum theory :)
 
geometry is too hard
 
@0celouvsky Does bob know assembly?
 
He used to
probably not any more
 
5:14 PM
Cruel world
 
@0celouvsky He's famous for Rokhlin invariant.
Gotta get dinner
 
@BalarkaSen Ah, that's out of my realm
 
5:32 PM
Huh, apparently one can get Arzela-Ascoli on merely separable, totally bounded metric spaces.
 
 
1 hour later…
6:36 PM
kinda dead the chat right now, innit
no they're not
 
I'm here
what do you want to talk about
 
about my friend cunty
 
who dat
 
can I ask you a question
 
it's going to be mean I bet, but sure
 
6:43 PM
Ban @AccidentalFourierTransform
 
it's a certain type of von Neumann algebra
 
I thought we were alone
it looks like we have company
Ill ask you some other day
 
sounds like a plan
@Slereah what are you up to these days?
back to GR?
what happens to nonstandard analysis?
 
Still on GR yeah
I tend to go back and forth between things
 
7:00 PM
@Slereah have you seen GR people use pseudodifferential operators?
 
I have not
 
@JohnRennie Are you there?
 
7:21 PM
@ACuriousMind I need algebra halp
 
$x+1 = 2$
plz help with this algebra
How does one even prove that $x + a = b$ has a unique solution
 
using Weierstrass approximation theorem
 
I guess the usual $x_1 - x_2$ way
 
you just construct a suitable polynomial, and use induction
 
basically, yeah
you can prove that in any ring, such an equation has a unique solution
where is ACM when you need him
 
7:46 PM
@yuggib are you around?
 
Hello
Does anyone know of a program where I can input some potential data and have it plot the allowed wavefunctions?
I made a program that does this but it's not working and I need to compare it against one that works
 
@Secret I don't understand the question
CHSH is a measure of correlation
that is a kind of correlation where they do "more" than entangled states can
what else do you need?
 
@EmilioPisanty What is your feeling about the proton anti proton question question?
 
@Anonjohn as in, do I think a bound state exists?
yes, I do
 
Right! I don't have commenting rights yet
But i did have a comment
In all likely hood, bound states do exist. If you were to look at the Ground state energy as a continuous function of the internuclear distance, starting from $R_ to \infinity$ you would need a length scale where the ground state energy vanishes
 
7:57 PM
pre-latex science was the worst
 
Not to sure about the symmetry you were talking about though.
 
@ACuriousMind I died because of you
 

« first day (2341 days earlier)      last day (2589 days later) »