« first day (2446 days earlier)      last day (874 days later) » 

7:00 PM
How many times have you seen Rick Astley singing in outer space?
 
1 time.
 
That's dank. That's original.
That's why people like it so much.
 
I don't think that "I've never seen this before" is enough for me to like something.
 
@ACuriousMind yeah...while this gem has only 316k views
 
In fact, I can imagine a great many things I've never seen before and can live happily ever after without ever seeing them.
 
7:03 PM
You seem like a bad meme consumer.
Try this absolute art from Jimmy Fallon.
If you don't like that, you're not friends with me.
 
@BalarkaSen ::shrug:: I see your midly funny clip mishmash and raise you a Beatles & Star Wars crossover.
 
I lose.
That is great.
 
@ACuriousMind What was that for?
@BalarkaSen Because it was deleted.
Not by my hand.
 
Oh. Well don't post stuff which gets deleted
 
@Mostafa it took more than just tech to make the polio vaccine work, to think of how to use the technology, I mean...seriously. That's kind of a silly statement.
 
7:14 PM
@ACuriousMind I still stick to the fact that your taste in humor is missing the dank, on-high feel that for example is apparent in this
 
@0celo7 Derogatory usage of the word "autistic" is not appropriate.
 
@ACuriousMind I was not using it in a derogatory fashion.
 
@BalarkaSen I wouldn't even find that funny if I were high, believe me :P
 
man you are a robot
 
@BalarkaSen dat ASMR doe
 
7:18 PM
@BalarkaSen I thought that was established :P
 
@0celo7 Right?
Try
This is the most epic find from that list
I am not going to link anymore of those videos because this is clearly the best
 
@ACuriousMind What are these videos?
@BalarkaSen ok dude
 
Apparently what people with a very different kind of humor think people on drugs find funny :P
 
Consider the rectangle $[-2,2]\times\cdots\times [-2,2]\times[0,2]$ in $\Bbb R^n$
Let $g$ be a function in $C^\infty(\Bbb R)$ with support in $[-1.5,1,5]$ and $g=1$ on $[-1,1]$
Now consider the function $G(x)=\prod_{i=1}^n g(x^i)$ on $R$
I think this is the function I need.
It is zero along each part of the boundary except for the plane $x^n=0$, where it is one
and it is constant in a neigborhood of that plane, along the $x^n$ direction
And it gets killed well before any of the other boundaries...so you can have a nice Neumann condition on one side, and Dirichlet and Neumann conditions on the other sides
 
7:37 PM
@AccidentalFourierTransform €550 is a bit outside the envelope I'm looking for (but still OK), but mostly I'm looking for indefinite stays, so single-month rooms are not much of a solution
=/
@0celo7 I'm around
MMA still bugging you?
 
Yeah.
(Sorry for the delay... totally forgot to answer)
@EmilioPisanty MMA=mixed martial arts?
 
@EmilioPisanty Yeah. My ODE has a term of the form $xw_x$. I assume MMA tries to solve the equations for the derivatives, so it runs into $1/x$ when $x=0$.
 
@Mostafa not particularly, no
@0celo7 hmmmm
 
I can prove that the solutions are actually regular there, but the numerics get screwed up.
 
is there a simple MWE?
 
7:39 PM
My solution right now is to put the initial conditions at $x=0.001$ and solve on $[0.001,b]$
@EmilioPisanty MWE?
 
@0celo7 minimal working example
 
@ACuriousMind Minimal working example?
 
Minimal working example.
 
minimal working example
 
7:40 PM
Minimal working example?
 
An example that has as few lines of code and clutter as you can manage but still demostrates the issue.
 
if your solution is regular, isn't there a transformation to a dependent variable for which the equation is explicitly regular?
@ACuriousMind what's an examüle?
 
@ACuriousMind examüle?
 
that has to be the strangest typo ever
 
A German typo :P
(p is next to ü)
 
7:41 PM
@EmilioPisanty Boy if you could find it, then you can write our paper for us!
 
ü is like three separate keypresses
@0celo7 hah
 
@EmilioPisanty Not on a German keyboard, it isn't
 
It's a complicated analytical argument by extending solutions on intervals $(\epsilon,b]$ as $\epsilon\to 0$
 
@ACuriousMind German keyboards are abominations and they should all be burned
 
@EmilioPisanty We have made bad experiences with burning stuff in the past.
 
7:42 PM
@ACuriousMind German keyboards are abominations and they should all be burned
(in an environmentally-friendly manner)
 
@ACuriousMind If I cannot say bad things about JEE, I propose the same to be true about German keyboards...
@ACuriousMind I would have though a Holocaust joke beneath you.
 
@0celo7 have you actually seen a German keyboard?
 
@EmilioPisanty We have German lab equipment at work
The computers are German
It's literally impossible
 
@0celo7 I intended to refer to book burnings...
 
@ACuriousMind But you see how it could be a Holocaust joke?
In any case, @EmilioPisanty, suggestions?
 
7:45 PM
@0celo7 Only if you think "stuff" may refer to persons and I'm not inclined to discuss that line of thought further.
 
@ACuriousMind The connotation crossed my mind too
 
Ah, the old switcheroo.
 
but in any case
 
"My joke isn't bad, you're bad for thinking it's bad!"
 
@ACuriousMind I also thought you're refering to book burnings
@0celo7 ....
 
7:47 PM
but in any case
^ actual picture of a German keyboard
 
@EmilioPisanty The problem is that it's a system of ODE, and making one unsingular screws up the other one.
 
@0celo7 yeah, that's tricky
it's hard to think around it in terms of coding without an MWE though
 
@EmilioPisanty So without more sophisticated methods, I'm probably stuck with my .001 trick?
@EmilioPisanty I still have no idea what an MWE is.
 
it's hard to think around it in terms of coding without an MWE though
 
:/
Fine, let me get a similar ODE
see if it breaks MMA
 
Sid
7:50 PM
@EmilioPisanty German keyboards are that great, huh.
 
@EmilioPisanty Your internet seems to have some sort of hiccough
 
@ACuriousMind What is that word?
Aaaaaaand mathematica has stopped working
 
@0celo7 The internet is a newfangled invention by which people can communicate all sorts of information...
 
What a horrible software
@ACuriousMind No, "hiccough"
 
Sid
@0celo7 hiccup. Similar to covfefe. :P
 
7:51 PM
@0celo7 The archaic spelling of 'hiccup'
 
so, say, would $x f'(x) = 1-\cos(x)$ encapsulate the behaviour well?
 
@EmilioPisanty NDSolve[{x*y'[x] == y[x] , y[0] == 1}, y, {x, 0, 30}]
@EmilioPisanty That encapsulates it perfectly because I can solve it explicitly, and it is regular at the origin.
 
@0celo7 ah. how vital to your example is the y[x] on the RHS?
 
@EmilioPisanty Vital. In my actual system that y[x] is the other unknown
 
For my example, you can set up a g[x_]:= If[x!=0, (1-Cos[x])/x, 0] and set that as your RHS
for your example, that's trickier
 
7:56 PM
Actually that's a bad initial condition, you can't get y[0]==1
y[0]==0 works just as well
that's a pretty bad ODE
It has a very unstable solution space...
 
8:15 PM
yeah, the problem with the finite-offset trick is that if the problem is unstable then you're never fully sure you can trust it
 
@EmilioPisanty I've asked her if the room can be rented for an indefinite time instead of just one month
I'll let you know
 
@AccidentalFourierTransform thanks =)
 
BTW ü is a cute, happy letter
 
ö
I'm not sure why you'd think it is a tautology, but it is not. It is, at best, a pretty natural result, but it is most certainly not tautological. (BTW, if you want to give the book a look before buying/borrowing it, you can find it here). — AccidentalFourierTransform 1 min ago
^ Is it wrong for me to suggest... ehem... downloading a book for free?
also, DeWitt is my new favourite book
 
8:29 PM
@AccidentalFourierTransform Does chapter 16 contain regularity theory for manifolds with boundary?
 
no, that's chapter 2
 
@AccidentalFourierTransform doesn't look like it from the ToC
 
what?
Chapter 1. deez nuts. Chapter 2. Regularity theory for manifolds with boundary.
its right there
 
Chapter 1 has a different title
We must be looking at different books
 
I guess we'll never know
 
8:32 PM
@AccidentalFourierTransform Relevant meta.SE post.
 
I would read Chapter 1
 
@ACuriousMind k.
but is it ethical for me to do it?
like I give a crap :-P
wat
everything I believed is a lie
 
@BalarkaSen Don't ever do PDE
 
@AccidentalFourierTransform ¯\ _(ツ)_/¯
who cares
 
The geometric normal to my rectangle in the chart is not the Riemannian normal according to the metric
I'm done
 
8:37 PM
@0celo7 Uhh... cry everytime?
 
@BalarkaSen where do I start with algebra?
 
what kind of algebra
 
by the way, how does the lawsuit against sci-hub progress (can it, with the creator of sci-hub hiding)?
 
anything that's not PDE on manifolds
 
@heather well, a fine was imposed recently
 
8:38 PM
I AM THE CREATOR OF SCI-HUB
 
@0celo7 lol
 
@AccidentalFourierTransform I would find that... surprising
 
that is what my lawyer said
 
@ACuriousMind why is that funny?
 
@EmilioPisanty but has it been determined illegal, or how does it work if you use it, etc?
 
8:39 PM
@heather what does "illegal" mean?
 
@0celo7 learn some commutative algebra
you should learn topology anyway
 
@EmilioPisanty two parts: can people other than the user use it without repercussions? and can the creator come out of hiding without repercussions, or with minimal repercussions?
 
@heather if the question is, whether a regular-(joe|jane) internet user who obtain pdfs from sci-hub is likely to face any legal repercussions
 
@ACuriousMind are you saying I am mistaken?
 
the current answer appears to be no
 
8:41 PM
I don't know right now
 
@0celo7 No, I just find the normal not being the normal funny.
 
I am quite confused by this
 
oh hey, sci-hub is up and running again
 
This problem is so much harder than at first glance
 
it was blocked here in Spain
 
8:42 PM
@AccidentalFourierTransform was it down?
 
for the past few months it was blocked
 
@ACuriousMind It's quite counterintuitive! Also the pesky $\sqrt g$ in the integral is coming to haunt me.
 
@AccidentalFourierTransform I'm pretty sure I've used it in the past few months
but it's a bit hard to tell as most of my usage is through telegram
 
then it was my internet provider or something
 
@AccidentalFourierTransform if any pesky ISP gets blocksy against scihub, go through telegram
 
8:43 PM
@EmilioPisanty what about the creator?
 
it's end-to-end encrypted so the traffic is exclusively between you and the scihub bot
 
@EmilioPisanty I've been using a proxy, so its fine
 
and also, i'm curious, does using sci-hub hurt the scientists? or really is it just a problem for publishers?
 
@heather it's to some extent up to whether Alexandra Elbakyan is OK staying in Russia or not
 
what do you mean?
 
8:44 PM
its only hurting capitalist pigs
 
is Russia considering extradition or something?
 
@heather How would it hurt the scientists? They don't make any money from their papers to begin with.
 
@ACuriousMind Uh, scientists get rich off of blood money
 
@ACuriousMind ::shrugs:: I dunno, that's why I was asking.
 
if she does go to the US or Europe, there is a nontrivial likelihood that there are arrest warrants against her (either sealed or not)
 
8:45 PM
they live in lavish apartments and whatnot
condos
drive Teslas
 
@EmilioPisanty so, okay, if she's arrested, what's likely to happen? told to pay a fine and let off? put in jail?
 
if she doesn't leave Russia and its sphere of interest then it's unlikely that she will personally face repercussions, at least the way things currently stand
@heather that'd depend on the charges they manage to conjure up
I don't think we'll find out because I don't think she'll leave Russia any time soon
she's stated that she's quite OK with the status quo on that score so far
 
Mar 30 at 19:34, by AccidentalFourierTransform
seize the means of production!
 
@ACuriousMind It's strange, because morally I expect this to give a solution to the Neumann problem in the chart, but from previous considerations I expect it to be a solution with the Riemannian normal.
 
8:50 PM
okay, so ACM's comment makes me wonder: why can journals charge so much money? their only costs are for printing, delivering, and editing - the review boards are volunteers, scientists are vying to get published, and paper/ink are fairly cheap, and journals cost so much money - I mean, an individual issue of nature costs more than an entire subscription to some magazines. So why does anyone buy?
why does anyone submit to them in the first place?
 
@heather because universities pay it
same reason why hospitals can charge so much
 
@heather Historical inertia, and no alternatives
 
@heather honestly: because they are unregulated monopolies.
 
because desperate people/insurance pay it
 
I submit my stuff exclusively to vixra
 
8:52 PM
and universities perpetuate it by forcing scientists to publish in high-impact factor journals to get tenure...
 
Okay, Emilio's phrasing is better ;)
 
@heather Also there is no lobby for independent scientists. Anyone worth anything will end up at a university or company that will pay for it.
 
because they hold the copyright to irreplaceable papers and libraries need to provide access to the existing stock, so it is very hard for libraries to cancel subscriptions
 
@AccidentalFourierTransform but that's problematic too, because there's no review board. arXiv is a little better, but still generally the same problem.
 
because they hold the copyright to irreplaceable papers and libraries need to provide access to the existing stock, so it is very hard for libraries to cancel subscriptions
there's two sides of the question, and both are hard
 
8:53 PM
@EmilioPisanty so scientists don't even retain any ownership over their papers?
 
@EmilioPisanty What is wrong with you?
 
@heather I was joking, of course I use arXiv :-P
also, why do people write arXiv instead of arxiv?
also, $\LaTeX$ instead of latex
 
@0celo7 terrible internet access
¬¬
 
@AccidentalFourierTransform because the X is a chi
 
@heather nope
it is standard practice for publishers to demand that you hand over the copyright to your work for them to publish it
 
8:54 PM
@0celo7 $\mathrm{LaTe}\chi$
 
well that's just dandy.
 
@heather yeah, well
welcome to the world of academic publishing
 
arXiv is spelled as archive
not arksiv
 
some good initial reading
 
@AccidentalFourierTransform $\mathscr L\mathfrak a\mathcal{T}\mathsf{e}\chi$
 
8:55 PM
the recent piece on The Guardian also makes good reading on how we got to where we are now
the Cost of Knowledge manifesto is a good summary of just how f***ed things are as they stand now
 
@BalarkaSen how is Milnor coming along?
I think my copy is in a box somewhere
 
@0celo7 I read the bit on Riemannian geometry. I'll read the applications to loopspaces etc tomorrow
Working on some doCarmo exercises
 
this Tim Gowers blog post gives a good impression of how hard it is for universities to cancel subscriptions
 
@BalarkaSen Do you already know the cell attaching stuff?
 
plus there's loads of interesting publishing-industry stuff on the archives of that blog
 
8:58 PM
@0celo7 Sort of kind of. I can just go back ahead and learn it in greater depths
I know the gradient flow and how it works
It's just a side-project; I learn more by jumping around than staying still :P
 
hmm. i wonder if the publishers could be persuaded to sell the copyrights to papers.
 
but RG is priority rn
 
@heather maybe if someone had Bill Gates money
 

« first day (2446 days earlier)      last day (874 days later) »