« first day (1927 days earlier)      last day (3299 days later) » 

00:00
@0celo7 It's two pages, I'll need longer to figure out whether you could possibly need this than you need to read it :P
Unlike you I don't understand proofs instantly.
wtf is a dyadic rational number?
@ACuriousMind Are manifolds completely regular?
user54412
@0celo7 Meh. If I wrote a specialized book (say, on fluid dynamics in GR) and I only expected to sell a few thousand copies, and I only got 10% of the gross, and I needed to take a year off from work to write it, ...
@0celo7 I don't know what "completely regular" means
user54412
people complain about expensive books, but honestly the reason I don't intend to make a textbook is because I can't afford to do it
@ACuriousMind T$_{3\frac{1}{2}}$?
00:08
@0celo7 lol, you don't seriously think I have any of the $T_i$-separation axioms memorized?
Uh...I did
They rarely come up outside of intro and highly specialized topology
@ACuriousMind def 9.5 on page 26 of Bredon then
@ChrisWhite hmm
@0celo7 Well, you can answer your question yourself. Is that a local or a global property?
(I.e. can it happen that this succeeds in a neighbourhood of $x$, but fails on the entire space?)
what
00:12
The defining property of manifolds is that they're locally homeomorphic to the reals. So any local property that holds for the reals will also hold for manifolds
I know
but I don't know if it fails or not
@0celo7 Say I give you the statement: For every $x\in X$ there is an open neighbourhood $U_x$ such that $U_x$ is completely regular. You should be able to prove that then $X$ is completely regular, too.
@dmckee What do I need to know when picking a graduate school if I aim to be able to get postdoctoral positions after my PhD? What are the important factors?
@GPhys Shouldn't you have thought about that...before choosing where to apply? :P
@ACuriousMind I'm talking about between schools I've been accepted to
00:16
@ACuriousMind I have no idea
As in, what should I look into when visiting, etc
why do you assume I can do these PhD level proofs?
user54412
@GPhys The factor is where past graduates have ended up.
@ChrisWhite ?
user54412
Where the graduates of the program have ended up tells you where you can end up. If they can't get the postdoc positions you want, you probably won't get them either. If they can, you probably will.
user54412
00:19
There's no other indicator that works so well.
user54412
I suppose it's possible that you go to a really outstanding program that opens the doors you want, but where nobody chooses that route.
user54412
In any event, the trick is getting an unbiased answer to what jobs people have gotten. In my experience departments have very selective memory. They'll tell you about so-and-so who is now tenured and doing well, but they'll forget about that guy that left the field in frustration.
@ACuriousMind seriously, I don't see what you want me to do
user54412
You should ask the current grad students to list they're immediate predecessors, and give you hard numbers about how many went where.
>they're
00:23
I couldn't find anything online for Stony Brook for past students
user54412
Actually in general you should ask the current students tough questions. About everything. How happy are they? Do they regret their choice? Do they intend to keep working with their advisors once they leave? How many hours do they work?
user54412
@GPhys yeah, department websites will at best list a few examples of success stories
@ACuriousMind Maybe $C$ can be taken to be the union of closed sets contained in the various $U_x$s?
@0celo7 Because it is quite straightforward: Let $C\subset X$ be closed, and $x\in X$. Then $C\cap U_x$ is closed in $U_x$ and since $U_x$ is completely regular, there is a function $f: U_x\to\mathbb{R}$ with $f(x)=0$ and $f(C\cap U_x) = 1$. Now just define $f(y) = 1$ for every $y\in X-U_x$ to extend $f$ to all of $X$. This is still continuous because the preimage of every open set that doesn't contain the 1 is unchanged and for those that contain the 1 their complement is unchanged.
"Because it is quite straightforward"
I really don't know why you wold bother saying that
It's like authors that throw in "obviously" and "clearly" in various places
00:32
@0celo7 You literally take the function on the $U_x$ and glue a 1 on the rest to them, I don't know how much straighter the strategy could be
Hm, okay, there is a slight issue with the preimage of open sets that contain the 1
Btw according to Wiki, manifolds are completely regular
What it is?
mah peeps
@0celo7 That's what I was trying to get you to show :P
Well I can't do PhD level topology!
And I don't see why your new $f$ is continuous
@ChrisWhite But that means working for Bob Wald is a bad idea :(
since most of his past grad students haven't ended up in permanent GR research positions or even physics academia in general
user54412
00:36
@FenderLesPaul Do his students not do well?
I think only one of them has in fact
Sam Gralla
he got a tenure track at Arizona
all the other ones I've seen are either in industry or do random things in math
user54412
well, the majority of my adviser's students are not in academia at all -- not sure you'll ever have high numbers there
user54412
it's not about absolute numbers, but comparisons between what options you have
oh I misunderstood what you meant
plus my example was probably bad
user54412
also, what matters most is where they end up with their first position afterward
00:38
since even if your advisor is Bob fucking Wald, there are next to zero post-doc and faculty positions for pure/classical GR research (theoretical, not numerical or astrophysics)
it's basically a dead field
GR is pretty worthless.
user54412
::waits for ACM to post chat log about trolling::
Who's trolling?
user54412
@FenderLesPaul saw that edit
user54412
it's okay, I have a job already
00:39
@ChrisWhite Well maybe I would change my mind if my adviser didn't tell me to get a fucking $300 GR book
@ChrisWhite you got a postdoc position?
@ChrisWhite yeah sorry I meant like Bob Wald type research
user54412
yes in fact :)
Congrats!!!
@FenderLesPaul Go into pure math, do geometry
@0celo7 I'll pass thanks
@ChrisWhite in fact do you remember David Nichols?
00:41
@ChrisWhite where
He applied for post-docs this year as well for theoretical GR
and he's had a really rough time
user54412
@FenderLesPaul yeah -- my TA for a few courses
I think his best option is Penn State with Ashtekar
but that involves taking a foray into LQG which he isn't too keen on
so yeah theoretical GR is probably not the best example since it's an almost dead field
user54412
@0celo7 joint offer by Berkeley and Santa Barbara
@ChrisWhite does your post-doc area involve warm weather? :)
OMGGG
Duude we can bro out!
:)
user54412
00:43
:)
so how does a joint offer work?
user54412
the first 1.5 years up north, with visits to KITP, and then I reverse for the next 1.5 years
user54412
at least that's the plan
Wait which one is more north?
user54412
we'll see how it plays out
user54412
00:45
@FenderLesPaul wow you really are east coast
@ACuriousMind What the heck is a sequence that is dense in a space
@ChrisWhite :p
@ChrisWhite I don't know either, but I'm also not sure if NYC is above Boston or not.
@0celo7 The rationals in the reals.
Oh Berkeley is more north
user54412
00:46
I bet ACM knows US geography
@ACuriousMind no, I mean the definition
aww that means I won't see you until 1.5 years later?
Oh but you have visits to KITP
yay
I know what the definition is
user54412
00:46
yeah
@ACuriousMind If $X$ is totally bounded, why can we find a sequence $(x_n)$ that is dense in $X$.
There we go
@ChrisWhite I don't know German geography either
@ChrisWhite are you going to be at KITP anytime soon?
@ChrisWhite I'd not be willing to take an exam on it :P
user54412
00:48
@FenderLesPaul probably not before sep/oct at the earliest
Ok cool
So I'll see you before the end of the year at least
:)
user54412
by the end of the year I intend to have met 3-4 people from this chat in person
@0celo7 Just take the sequence that consists first of all the centers of an $\epsilon=1$-cover, then of the centers of an $1/2$-cover, then of the centers of an $1/3$-cover...
user54412
also, KITP mentioned in their offer that they have funds to travel especially to MPA-Garching, so...
@ACuriousMind Proof by contradiction to show that all points appear in the sequence?
00:53
@0celo7 "dense" does not mean that "all points appear".
I don't know if this question is worth posting as a question on the site
Is it okay if I ask physics questions in the chatroom?
Or is it better off if I just post a homework question on the site?
Only topology is allowed in here
Okay, no dank memes?
@skully That's not allowed
@skully Unless you have read our homework policy and made sure your question is not off-topic, don't post it on the site
00:54
Oh :(
@ACuriousMind What does it mean
@0celo7 I thought you knew the definition?
A dense set is one whose closure is the entire space.
@ACuriousMind Well I have no clue how to define $\operatorname{cl}(x_n)=X$
specifically the LHS
I have no clue what it means
@0celo7 You're in a metric space.
So I take the limit, right
00:55
wat
The limit would be a point, if it exists
Yes
Ok so $(x_n)=X-\lim x_n$?
You want the closure. The closure are all points which are infinitesimally close to the set.
@0celo7 No
That is, the closure consists of every point that is either in the sequence or for which there is a subsequence converging to it
Yes, of course.
By making a sequence that consists of the centers of continually shrinking balls, it is ensured that for every point in the space, there is such a subsequence
Ok...so how is the sequence you defined dense in $X$?
What
This should be obvious, he does not prove it
This is not obvious
00:59
@ACuriousMind certainly the first condition was not necessary :P
user54412
@skully which means you can ask things in chat, but while someone might be willing to help with homework, no one here is interested in just doing it
@0celo7 For every $x\in X$, and every $n\in \mathbb{N}$, there is a ball of radius $1/n$ such that $x$ is inside it. The sequence of the centers of those balls obviously converges to $x$, and the sequence I built contains every such sequence for every $x$.
I would challenge the "obviously" line
@GPhys Huh? $x_n= 1/n$ has no subsequence converging to $1/3$, for instance.
But we've been down that road before, you like using words like that
01:00
@0celo7 Those centers get arbitrarily close to $x$, that's exactly the definition of convergence.
@ACuriousMind is your metric you are finding the closure of the union of those?
@ACuriousMind seems pretty sketchy
@0celo7 The "obviously" is has a dual meaning: It means both "I will not bother writing this out." and "If you do not think this is obvious, think about it until you agree".
@GPhys I cannot parse that sentence. Please rephrase.
I agree, but I don't know how to justify that I agree.
@ACuriousMind :27544080 I was intending to respond to what I read was a claim about the nature of the points in the closure of set. I do not understand how your response dealt with this, so I am asking for clarification.
01:05
@GPhys Chris' remarks are extremely pertinent. The more so as they represent a factual question rather than one of simple opinion.
That said the data could be hard to come by.
@ACuriousMind I will try to come up with a proof of your statement for $\mathbb{R}$...
It's not "obvious"
@GPhys I had two conditions: 1. The point is in the sequence. 2. The sequence has a subsequence converging to it. You claimed the first was not necessary, I countered your claim by exhibiting a sequence which has points to which no subsequence converges.
::pulls earplug out of pocket::
What
@ACuriousMind Pretty sure I'm going insane.
Hmmmm...or are you people using "sequence" also for finite ordered tupels?
@ACuriousMind Yeah I don't understand at all. What are you defining with this? It absolutely does not appear to be what I thought I was responding to, still
01:11
@ACuriousMind A sequence is $f:\mathbb{N}\to X$.
@GPhys Those two conditions define the points that lie in the closure of the sequence (viewed as a set)
And a subsequence is $a_{n_j}$, $j\in\mathbb{N}$ where $n_{j}<n_{j+1}$
@ACuriousMind That's how we defined it in analysis class, anyway
@0celo7 Yes, that's how I see it, too
@GPhys It uses the characterization of a closed set as being sequentially closed, I'm not sure where we are misunderstanding each other
@ACuriousMind the prof said a billion times that the inequality there is strict
@ACuriousMind The claim is that if $x_n$ is the center of the $1/n$-ball containing $x$, then $\lim x_n=x $?
That's obvious
@0celo7 Yes
@0celo7 Yes.
01:19
Don't know how to prove it though
Don't prove it!
I know you love proving analysis things for me, control yourself
@ACuriousMind It certainly was partially unclear to describe the set itself as a sequence: however, that was what my first question was intending to ask: if your set you're finding the closure of was intended to be the union of the points in the sequence you gave (the answer: yes). I interpreted your second condition as being the set of points that have a sequence in the set converging to it. And, of course, for metric spaces this is exactly the closure without needing to include the first condition
@GPhys Ah, right. I see that could have been clearer.
I was certain something was misunderstood (by me or both), I was just a little confused what
Anyway, I'm headed off
time to go fret graduate school choices
@GPhys bye!
@ACuriousMind Do I need to do induction for this proof?
01:28
@0celo7 No, not that I can see
@ACuriousMind I'm having trouble with the $\forall n\ge N$ part, then.
It's clear that the balls have to be nested, but I think I need an induction to prove that.
@0celo7 No. $n\geq N$ implies $1/n\leq 1/N$.
And the balls are not necessarily "nested"
Actually, they're not nested.
@ACuriousMind uhhhhh
let me think
01:45
@ACuriousMind Ah!
It's trivial
Just needed some food to clear my mind :)
@ACuriousMind Just pick $N\ge 1/\epsilon$ and everything works out.
@ACuriousMind I was focusing on them having to be nested, which was wrong.
@ACuriousMind Oh god in PDE we're going to do wave equations with time-dependent boundary conditions D:
I can imagine the number of Fourier series in the solution doubling
02:43
@ChrisWhite Do you own Hawking-Ellis?
@ChrisWhite Alternatively, what do you know about Bianchi I spaces?
user54412
no and I never could keep track of that crazy classification scheme
user54412
it's like the separation axioms
No?
What GR books DO you own
Carroll, Wald, MTW?
user54412
of course
"Of course"
Carroll and MTW are too expensive.
user54412
02:50
probably a few lower level ones too
Why?
user54412
is HE even a textbook?
Serious question
@ChrisWhite Nope
It's a "monograph"
user54412
I only really own textbooks that were used in a class, and a couple reference books of the kind I can actually cite in a publication
You don't think HE can be cited in a publication?
user54412
02:53
let me rephrase -- something I can and will cite in some publication
user54412
nothing I do is at the level of HE
Damn chap 7
@ChrisWhite What book did you learn PDEs from?
not numerical
I want to reread HE, this time proving everything properly, but chap 7 on the Cauchy problem scares me
user54412
@0celo7 none, my PDE class didn't even pretend to use a book
I'm stuck
I despise analysis and topology
Can't do much without them, though
user54412
02:57
also, in hindsight I'm not sure how much there actually is to non-numerical PDEs
Sobolev inequalities
...that might be it
ask @yuggib, I think that's his field :P
user54412
I mean, there's the standard couple of tricks for turning them into ODEs, and there's the 3 special cases for 2nd order linear systems, and I can't remember anything else
well somehow one has to prove that the EFE can be solved
otherwise your field would be wrong
user54412
yeah, those 3 cases are where you get a degeneracy where the Cauchy problem isn't trivially solvable
what are you talking about
user54412
03:00
I feel like you could write a book on the details, but ultimately 2nd order linear PDEs don't do anything you don't expect
The EFE are nonlinear...
Choquet-Bruhat wrote a whole book on the Cauchy problem in GR from an analytical perspective
user54412
@0celo7 leading terms are linear -- close enough
so the Cauchy problem has pretty much been figured out?
user54412
no, for GR
user54412
03:06
I mean, I count local existence as sufficient
local existence does not care about topology, which is the interesting part
user54412
also the impossibly hard part
hmm, there seems to be a correlation between impossibly hard and interesting...
03:22
@ACuriousMind I tend to end up having too much quantum stuff whenever I think about GR and always forgot it is 100% classical
@0celo7 For me, impossibly hard=alien=novel=interesting=new territories
if I had a cube shaped planet gravity would not operate ?
only spherical planets have gravity?
Some of my professor in my chemistry department kinda foreshadow this as they talked about the problem of the "pay to open access" system
@Secret - I guess you do not know the answer to my question
hence the attempt to digress
I am not digressing, I am talking to the other users that will be on 13 hours later.
For this case, I actually don't know the answer
0celo7 might be a better guy as he studies GR
curvature is the cause of gravity right @Secret ?
so if the curvature is removed what happens to gravity ?
adios ?
03:30
@gansub what are you talking about
doesnt GR say gravity is due to curvature ?
so if the curvature of the planet is removed then ?
What?
i dunno you tell me
How do you remove the curvature?
it is just a gedanken experiment bro
03:32
Physics can't tell you about unphysical situations
The curvature is the curvature of spacetime, not the curvature of the shape of the planet. As long the planet has a nonzero stress energy tensor, it should be able to bend spacetime and hence giving the effects of gravity
though it would not be a simple $1/r^2$ like potential due to its cubic shape
ok big difference
thanks for clarifying my doubts
> not the curvature of the shape of the planet.
what
Why would it be that curvature?
if i understand GR right space time is curved only because of astronomical bodies am i correct?
03:36
@gansub You understand it wrong
ok then enlighten me
Spacetime is curved by nonzero energy/momentum/mass
This means pretty much anything curves spacetime
so is empty space - where there is nothing is that curved as well?
Sorry, what?
say a location outside the solar system where there is "nothing"
no mass whatsoever
is that curved ?
03:38
Yes
Because in the solar system there is matter
i said a point in space where there is no matter whatsoever
Spacetime will be curved if there is matter anywhere in the universe
hmmmm
so the effects of matter are field like ?
Just like we can have an electrical field far away from the source
so the spacetime curvature is felt like a field ?
far away from it's source ?
03:40
Yes!
what about at the beginning
of time
GR can't tell us about the beginning, only what happened after it
so immediately after the beginning spacetime becomes curved ?
@gansub well...I'm not sure if "spacetime" is even the right word for stuff before the beginning
what i am saying is just after the beginning
say when hydrogen atoms form
by that time is spacetime curved
03:46
ok, spacetime was always curved
look up the FLRW metric
we believe that the "background" spacetime is curved by default
but I'm not up to snuff on my big bang phenomenology
ok thanks you made me understand some hard concepts
 
1 hour later…
user116211
05:52
@Qmechanic: Good to see you back! PSE without you is like the cold winter; Spring came back ;D
10:40
Hi, web.pdx.edu/~pmoeck/pdf/Mermin%20short.pdf says that we can build an impossible device using famous Bohm's setup in the appendix. I see something about couple of detectors rotated by angles various angles and formula $sin^2(\theta/2)$, which looks similar to Mermin's $cos^2(\theta/2)$ but it is too hardcore for me in Harris. I do not not have a clue. Is there explanation for mortals, starting from how do you detect spins?
Is there a variation of the experiment with simple light polarization which can explain everything in electrodynamics?
 
2 hours later…
12:13
@0celo7 Maybe the omega comes from THE ALPHA AND THE OMEGA
(PRAISE THE LORD)
You know I'm not quite sure what would be the metric of a cubic distribution in spacetime
not sure it's been done
http://chat.stackexchange.com/transcript/message/27549980#27549980
this stuff reminds of entanglement but I think I still have not fully understood the maths of qunautm mechanics to investigate this further

Thus right now, besdies writing my molecular laser photochemsitry thesis, I am reading this

https://en.wikipedia.org/wiki/Mathematical_formulation_of_quantum_mechanics
The problem being that a cube basically has no continuous symmetries
@Slereah Well, it can in princple be modelled, but I don't think there's an analytical expression for it. IT has to be doen numerically
12:18
Oh you never know
Plenty of weird spacetimes have been done anaytically
But one with no symmetries (outside of time), not sure
It's probably doable in the thin shell limit
I remember that Visser did cubical wormholes
$$|\Psi\rangle=\sum_i|A_i\rangle\otimes|B_i\rangle$$
$$ \hat{M}|\Psi\rangle=\hat{M}\left(\sum_i|A_i\rangle\otimes|B_i\rangle \right)$$

To be continued...
@Slereah Why would cubic wormholes be interesting?
Because a wormhole with flat faces has no tidal forces
makes sense
Well, the tidal forces are concentrated on the edges
don't bump your head on the edge
Although really a cubical wormhole has pretty unrealistic sources
(To prevent anyone from asking WTF is that): some attempt at modelling a general entanglement system and then measured it using $\hat{M}$. The aim is to see how computing this thing shows how the entangled states changed by the measurement operator

However I need to revise how t do tesnro procdtucts
 
3 hours later…
15:30
Man, gravitational wave questions really are drowning out everything except the ever-persistent homework questions :P
Also, Skeptics is being really skeptical:
60
Q: Do researchers receive no income from revenues arising from their published papers?

user69715According to this article Meet the Robin Hood of Science (about the creator of a website to bypass paper access restrictions and paywalls), Elbakyan made a point that will likely come as a shock to many outside the academic community: Researchers and universities don’t earn a single penny...

I mean...that's not really a question about "scientific skepticism" is it?
@ACuriousMind I can remedy that with some topology questions :)
I'd ask a question but nobody ever answers my questions :p
@Slereah Yeah, that happens :/
15:51
hello
@Slereah lets hear it
Well I have a bunch of unanswered questions if you want
Although for some I found the answer later on
was it interesting?
Well physics is always interesting :p
I concur
:P
give us one question
The one I found an answer for btw is that the chronology protection conjecture cannot be circumvented by a spacetime with non-geodesic CTCs because those are always present in the Cauchy horizon
2
Q: How to deal with boundary conditions for path integrals?

SlereahFor non-relativistic quantum mechanics, the boundary conditions are rather simple to deal with, they are just \begin{equation} \langle x_1, t_1 \vert x_2, t_2\rangle = \int_{x_1(t_1)}^{x_2(t_2)} \mathcal{D}x(t)e^{\frac{i}{\hbar}S[x(t)]} \end{equation} And then we can solve the path integral by...

This one still doesn't have any satisfying answers, too
6
Q: Homotopy proof of the lack of foliation of the Gödel metric

SlereahA common proof of the lack of foliation of the Gödel universe, apparently mostly copy pasted from Hawking and Ellis, goes thusly : A closed timelike curve must cross a spacelike hypersurface without boundary an odd number of times A continuous deformation of the curve can only change the numbe...

Sorry Timaeus
15:56
@ACuriousMind Holy crap that $300 book treats Morse functions on Lorentzian manifolds :o
gimme a sec ....
also
2
Q: Full form of the Pauli-Fierz action

SlereahIn Deser's paper on the fully interacting version of the Pauli Fierz theory, he does a rather simple method of treating the Pauli Fierz equation without going with infinite sums, just by treating the metric and the connection separately, hiding the infinite sum in the dynamics linking the (invers...

@Slereah
-4
Q: What will a true law of gravitation look like?

VixillatorNewton's law of gravitation and Einstein's general relativity (GR) are empirical laws of gravitation that are strong on description and prediction but lack deterministic and/or causal rigor, which makes both of them incompatible with the fundamental laws of physical reality. Newton's action-at-d...

lawl
@ACuriousMind.....what do you make of @Slereah 's questions?
they're complex!
15:59
I know

« first day (1927 days earlier)      last day (3299 days later) »