« first day (721 days earlier)      last day (4306 days later) » 
00:00 - 14:0014:00 - 00:00

2:01 PM
@JasperLoy Sorry about that. You have not offended me. I merely put you on ignore from time to time because your output volume is huge. Chat tires me out if there is too much stuff going on. So: no offense and I do get some of your messages if you address me, just possibly not all of them. I apologise.
3
 
Matt is just very busy I suppose. Exams are more important than chatting 8-).
 
user19161
@MattN. OK, thanks and noted. Take care and good luck!
 
@JonasTeuwen posted
 
@OldJohn That's quick! :-). Thanks!
 
@JonasTeuwen No problem
 
2:04 PM
Hello hello helo
Silent huh
 
@JasperLoy Thanks.
 
2:27 PM
did not get nearly enough morphine to get high
I see you are quoting Jasper's statement to Matt, but I don't know why you are addressing it to me.
 
That was a joke :-D
to see if you were high
 
2:47 PM
 
3:13 PM
When I first saw morphine given to injured soldiers, I thought it was a medicine, then later after watching more movies I came to conclusion that it is an effective pain killer. Now I realize that it also gives "high".
 
@MattN. that gives me an idea how to deal with messages from J.Lo
 
@JonasTeuwen Yes, that's right. Well, maybe "stressed out" is more accurate. I am busy, too, but that is usually not a problem.
@Ilya Oh my, what have I done... : /
@JasperLoy Btw, "huge" is a bit of an exaggeration since I have a bunch of others on ignore with a much bigger output volume. So rest assured: it's nothing personal.
 
@Ilya Hi, I think your message is a bit insensitive. You could have silently put him on ignore whenever you wish. Why to say it in public, that too in a message to a third person, unless ofcourse if you wanted to insult him.
 
@MattN. I guess, you're also ignoring me sometimes - based on this criterion :)
 
I ignore myself sometimes :-)
The problem with people is that they don't have a sense of humor. Out in the real world you will not survive without some sort of sense of humor.
Hi @HenningMakholm
example^
 
3:30 PM
Hey @Skull : nice to see you after a while
 
@RajeshD Nice to see you too Raj :)
BTW @HenningMakholm is an example of someone with a good sense of humor.
 
3:44 PM
gtg bye
 
@Ilya No actually, you're rather quiet.
 
bul
 
?
 
I'm as quite as a fish
bul
 
: D
 
3:54 PM
:)
 
Sorry, shouldn't be here, got to go again. bbl
 
@Ilya What does a fish know of the water that he swims in all his life?
 
Jackass is being willfully disruptive now, I think. Bleh. Is it so bad if he leaves?
3
 
@J.M. Jackass?
 
@Ilya Well, it's the second or third thread in meta that has been opened about him...
I mean, it can't be that he's stupid...
 
4:07 PM
I'm off MSE these days
almost completely
 
Oh, sorry about that. :|
This last meta thread just set me off.
(Unrelatedly, I'm cutting down too, for a while. Posting on the site is starting to feel more like work than fun these days.)
 
Hello everyone
 
@Szabolcs Hey.
 
I was wondering, given a "nice" scalar function in 2D, $f(x,y)$, it there a simple way to construct from it some vector field with zero divergence?
The emphasis being on simple.
 
4:14 PM
@Szabolcs This is for that hedcut thing, I take it...
 
@J.M. You're good :-)
I was wondering, if I take the gradient vector, normalize it, and rotate it by 90 degrees, is that going to be fine (zero divergence)?
okay, I think I answered my question
 
This sounds like something they'd be studying in fluid simulations...
 
yes, exactly
isn't that how you guessed it was the headcut thing? I want to distort a point pattern with an interesting flow that's related to the image.
 
I haven't checked, but the CFD literature is the first place I'd be checking for hints.
 
@J.M. Agreed - he seems to have become a deliberate troll in the last few days :(
 
4:20 PM
@Szabolcs Yeah, have the line detection algorithms work on the image first and take it from there. Fine strategy.
 
4:34 PM
In fluid dynamics, the Stokes stream function is used to describe the streamlines and flow velocity in a three-dimensional incompressible flow with axisymmetry. A surface with a constant value of the Stokes stream function encloses a streamtube, everywhere tangential to the flow velocity vectors. Further, the volume flux within this streamtube is constant, and all the streamlines of the flow are located on this surface. The velocity field associated with the Stokes stream function is solenoidal—it has zero divergence. This stream function is named in honor of George Gabriel Stokes. Cylin...
It seems that's only for cylindrically symmetric flows
 
Yes, a bit limited...
 
@J.M. Who?
 
@MattN. That guy being discussed in meta for quite a while now...
 
I just figured that if I want to talk to you I can do it in private : D
I see.
 
@MattN. Bah, you're exploiting my abilities... :D
 
4:45 PM
@J.M. : )
bbl
 
5:06 PM
Hm. Live updating magically started working again. That's too bad.
 
What browser and OS are you on, Dylan?
 
Well I just updated to Safari 6. Maybe that's done it.
Still on OS X 10.7. Apple seems to have botched sending out upgrade codes for 10.8.
 
Oh. When I thought the live updating was gone, I was on Opera. When I used Firefox again, then that feature showed up.
I guess it's a browser thing.
 
@J.M. Did you see my meta post?
@DylanMoreland When did it?
 
@robjohn Right after installing the new Safari. Maybe they messed something up WebKit-wise.
 
5:18 PM
@DylanMoreland Perhaps I need to update Firefox. Mine seems to have stopped yesterday.
 
Hahhh. So I noticed that some of my teaching reviews were clearly intended for another section being taught at the same time. And they were horrible. The registrar claims that there's nothing they can do to fix this.
 
Oh joyous, 10.8 finished downloading.
 
@ZhenLin Ah, cool. I'm still waiting on my code...
I enjoy new operating system installation to an unreasonable degree.
 
A fancy new Mac for you then?
 
@J.M. So in this starred post are you referring to a certain Mortal Kombat user
I feel like he's gotten slightly more agreeable. Still shotguns questions up a bit much, I'd say.
 
5:23 PM
@DylanMoreland Nothing they can do? Really? So you have to live with the after-effects of a mistake, and that's all they can say?
 
@robjohn It's pretty cool.
 
@DylanMoreland I would be more upset than you seem.
 
I'm getting there :)
 
@J.M. I assume you've seen that I answered the question you pointed out to me.
 
@ZhenLin Yeah. It had been five years. I figure once I get two versions behind on the OS (couldn't install Lion on it, already) it's time to move on.
 
5:33 PM
@robjohn I have upvoted your fine answer when I first saw it.
@robjohn which?
 
@DylanMoreland Five years is a long time. And the latest round of hardware looks very nice. Good timing!
 
@J.M. this one.
It's just what Dylan and I were talking about.
 
@DylanMoreland Yes. "Slightly" is a bit much, methinks...
 
Asaf gave a very useful suggestion :-|
 
Well, now I have. As I said, at least on Ubuntu, the live updates work on Firefox, but not on Opera.
(which is why I'm using Opera now)
 
5:35 PM
It works for me in Firefox 14.0.1.
 
@ZhenLin I wonder how big the update is. I am tethered on my iPhone, so I don't want anything too big.
 
user19161
@robjohn Some people like to stay on the ESR. But using FF10 does not give full functionality to some sites like say Blogger.
 
@JasperLoy I'm using 11.0
 
The full download is 31MB.
 
user19161
Also, anyone who feels my output volume is too big, please feel free to ignore me! But if I make you upset, please tell me!
 
5:38 PM
@ZhenLin That's not bad at all.
 
I remember the days when Firefox was hailed as being small... I think back then it was less than 10MB!
 
user19161
Chrome is quite a big download too, and on Linux you need to install a whole bunch of dependencies to make it run.
 
Hm hm, someone is asking lots (> 1) of Riemann surfaces questions today.
 
@ZhenLin Wow. That was a loooong time ago. About six years, eh?
 
@JM I've used Firefox since two names ago! :p
 
5:47 PM
@ZhenLin Well, there's still a Mozilla apparently, so Firefox is more a fork than anything, right?
 
user19161
On Debian it is called Iceweasel instead of Firefox. The logo is different too.
 
@JasperLoy That's just because of clashes in licensing... kinda nuts, I think.
 
@JM Well, actually, I'm referring to the fact that it used to be Phoenix and Firebird.
I think I started with Phoenix 0.7...
 
user19161
@J.M. The problem is that backports in Debian seems broken. I can't seem to get Iceweasel 14 on Debian, only 10. So if we want 14 we may have to unpack pure FF ourselves.
 
@ZhenLin Ah. That I have forgotten... I had been using Mozilla and Navigator alternately back in the day, and then suddenly boom, hey try Firefox!
 
5:50 PM
@JasperLoy Technically, Firefox-branded Firefox can only be distributed by the Mozilla Foundation, or something like that.
Oh, I used Mozilla for only a short while. I forget for how long.
 
user19161
@ZhenLin Hmm, I think they must satisfy some requirements. Most Linux distros actually use "Firefox".
 
Perhaps they distribute a bit-for-bit identical copy, as opposed to making their own build.
 
@JasperLoy It's something in the licensing terms of Debian that clashes with the terms by the Mozilla Foundation, so...
 
I have a Firefox-branded local build on my Linux box, but I am informed that I am not allowed to distribute it...
 
@ZhenLin I think so. I like the new Airs a lot, and they're reasonable priced unlike before.
Firefox has kind of lost its way, I feel.
 
user19161
6:00 PM
@DylanMoreland What browser do you like?
 
I know of at least two people still doggedly sticking with Firefox 3.x ...
 
I like Safari. I've never used it on Windows — it might not be so good there.
And the new version incorporates two of the things that I preferred in Chrome, so that's nice.
 
Hm, 10.8 has a whole new set of desktop pictures and gets rid of the old ones. :(
But they are beautiful, at least.
 
user19161
@ZhenLin Erm, you know desktop pics can be changed right? But of course I also like the default ones to be nice. :-)
 
@ZhenLin >Not using pictures of dogs.
 
6:07 PM
I use a slideshow of the stock backgrounds. :p But they changed the stock pictures!
 
@ZhenLin nowhere else to grab them from?
 
user19161
I usually use plain blue background. My favourite is "steelblue".
 
@JM I have backups, of course.
 
user19161
I see that we have become more cryptic about some user, using a word instead of two letters to refer to him...
 
user19161
But I think it's OK to use the two letters to refer to him...
 
6:13 PM
I don't understand. I thought porton did super formal algebraic stuff that no one knows about. But he doesn't seem to know any category theory.
Everyone's different, I guess.
 
user19161
Of course, there are over 9000 branches of mathematics!
 
@DylanMoreland People who take a hands-on approach generally won't encounter category theory!
 
user19161
@ZhenLin But most grad algebra texts today cover categories quite well.
 
I wouldn't call what he does algebra...
 
user19161
The category theory is not necessarily in a separate chapter but is spilled all over the place.
 
6:16 PM
I have no idea what to call what he does.
What is a funcoid
 
Something involving a filter on a powerset...
 
This is the greatest line in any question ever:
"I have recently come across the concept of 'eigenvectors'"
 
user19161
@EdGorcenski Well, OP may not have a background in math...
 
user19161
Actually, it's not really funny at all.
 
S/he does not. However, I do not begrudge the OP that; as I have an engineering background, my co-workers and I often joke about how frequently eigenvectors come up and get used as seeming magic wands
 
6:21 PM
@EdGorcenski Didn't know about them until I was studying math on my own, so... :)
 
I recently came across the word "eigensheaf". Mind-boggling...
 
user19161
You must know MSE is for people studying math at all levels.
 
user19161
So an elementary school kid like me is here too.
 
So the phrasing of it seemed very a propos in a number of ways.
 
@ZhenLin this isn't linear algebra anymore, is it?
 
6:22 PM
@JasperLoy: Remember, when you divide fractions you flip them over...
 
user19161
@StevenStadnicki Hehe.
 
@JM I have no idea. It was thoroughly out of my league. (Geometric Langlands)
 
@StevenStadnicki my teachers always wondered why I turned my papers upside-down.
 
@StevenStadnicki give him a chance to reciprocate...
 
user19161
6:24 PM
They are less commonly called proper/characteristic vectors/values.
 
@JasperLoy That's the old English way of referring to them, yes.
 
user19161
@J.M. I think I probably got that from Halmos's Finite-dimensional vector spaces.
 
(but then people liked the German sooo much better. Now, we have eigenstuff.)
 
Hi ... how to find the integer solution for this system $$ a^b = 2^{2 c + 1} + 2^c + 1 $$
 
Eigenstoff, surely!
 
6:26 PM
@JasperLoy Ah. If you'll look at old literature, you'll also see "latent roots" and "latent vectors".
 
Huh, I never made that connection. I have an old DiffEq book somewhere that talked about characteristics and figured it was something I'd come across eventually; I filed it away and never realized that eigenvectors were exactly the same thing.
 
any relevent questions on this??
 
@experimentX exponential Diophantines tend to be difficult. Try asking on main.
 
@experimentX : if it helps, your right side is $2x^2+x+1$ for a very special $x$...
(Obvious presumably, but still bears noting)
 
i guess i'll have to ask main site ... i'm running code for half an hour ... only got 2 solutions
still running!!
 
6:28 PM
On the other hand, I've only seen two or three books that use "eigenpolynomial" instead of "characteristic polynomial". I wonder...
 
@experimentX that sounds pretty typical for exponential Diophantine equations, I think - they mostly seem to have a finite set of solutions
 
Hold on ... i'm posting a question on main site
 
@OldJohn Don't forget the part where it's a pain to actually prove anything about them. :-)
 
@StevenStadnicki Oh Yes! - I was trying to keep that bit quiet :)
 
user19161
6:40 PM
@StevenStadnicki Perhaps the same may be said of differential equations.
 
It seems to be one of the classic principles of number theory. You can add, or multiply, but start doing both and you're in serious trouble.
 
@StevenStadnicki Yes - I remember reading some author trying to explain that is likely to be difficult as the only connection between the two operations (from the axioms) was the distributive law, but I don't remember being very impressed with his argument at the time.
@experimentX For that equation, I can't imagine there are going to be any more than the 2 simple solutions - I would be amazed if there are any more.
 
nod At the same time, proving anything formal along those lines is virtually impossible. I'd love to see something like the lack of any asymptotic statistical correlation between factors of n and n+1 (beyond the obviously necessary lack of shared factors) but it's not even clear where a result like that would begin.
 
how to prove that those two are it's solutions??
 
@StevenStadnicki Indeed - I fear that I will not live long enough to see any serious breakthroughs along those sorts of lines
@experimentX You have already proved that they are solutions - you JUST need to prove that there aren't any more :P
Where did the problem come from?
 
6:49 PM
@OldJohn oh original problem was posted on this site
i got the soution running code ... but isn't that cheating??
now i'm running from 1 to 100000 for an hour now
 
@experimentX It is not possible to prove using code that there are no more solutions
 
@experimentX Can you find a link to the original problem? That site seems hard to navigate...
 
If it was a problem set as part of a course, there might be a clever way to tackle it, but if it one which arose naturally, it is likely to be extremely tough
 
i guess the second case ... it wasn't my problem.
it seemed interesting ... so i spent few hrs
 
6:54 PM
It's (meta-)interesting because it's very close to Catalan-style problems that at least have a good amount of formal weight behind them, but not quite.
 
@StevenStadnicki yes
 
e.g., if it were $2^{2n}+2^{n+1}+1$ then you could write as $a^b=(2^n+1)^2$, pull a factor of 2 out of b and be at $a^c-2^n=1$ with $c=b/2$.
 
@StevenStadnicki Yep - and then quote Catalan?
 
Exactly.
 
and can you tell me why wolfram doesn't evaluate this integral ... I'm asking it out of curiosity
 
7:01 PM
Somewhere I have a book entirely devoted to exponential Diophantine eqns - but not read much of it
 
xx is supposed to be x^x
 
That's odd experiment
 
odd integral??
 
weird
it will do the integral for some bounds
 
sorry ... the link was broken ... the correct link is here
 
7:05 PM
Yes, I saw your correction.
 
bound is from 1 to infinity
 
It will compute the integral from 1 to 10, 1 to 10^2, 1 to 10^3, 1 to 10^4
but not 1 to 10^5
and neither 1 to infinity
but for the bounds that it will compute it gives .70417
I can't imagine the value changes as you go off to infinity either
 
but it's possible to calculate for 1/x^2 why not for 1/x^x
 
because you need to integrate e^{-xln(x)}
 
doesn't all bounded integral can be calculated??
 
7:08 PM
no
You can usually approximate them
 
for example?? other than this
 
I mean for instance e^{x^2}
the integral of that gets it own function
 
but e^x^2 is unbounded
 
because you can't express it in terms of "regular" functions, but it's very important
 
doesn't e^x^2 -> inf as x->inf??
 
7:10 PM
yes
I'm just saying it's still difficult to calculate the integral on a closed interval
 
but in our previous case 1/x^x -> 0 as x->inf .. also it can be proven that integral converges
 
yes the integral converges
but just because something converges doesn't mean it's easy to calculate
 
lol ... it's not calculated due to it's complexity in nature ... even for machines??
 
I mean, I think you could approximate the integral
on an interval
and 1/(x^x) goes to zero extremely fast
 
well yes ... some how like that ... but the upper bound is infinity
atleast i was hoping WA to give me approximate value
 
7:14 PM
like I said
.70417
is the approximation it give for the integral from 1 to 10^4
I'm guessing that's pretty close
 
okay ... i'll test for more higher bounds and see how it goes
 
it won't go past 10^5
but experiment
(10^4)^(10^4)
is 10^(40000)
so the value of f at that point is 10^(-40000)
and it just gets smaller faster
.70417 is easily accurate within 100 decimal places
well
within 2 or 3 at least
I don't know how accurate wolfram's approximations are to begin with
 
atleast it seems to give some values in pi's and logs
 
hi @Gigili
 
i was hoping answers in something like that ... like area bounded by normal curve. WA seems to give sqrt(pi) exactly
i thought i would do some algebraic calculations instead of direct numerical calculation
 
7:20 PM
@skullpatrol Hello.
 
@Gigili How's it going?
 
@skullpatrol Not bad, thanks. How are you?
 
@Gigili I'm alright, thanks.
 
 
1 hour later…
8:29 PM
Just got back. That looked like a quick visit by Gigili
 
user19161
@robjohn I'm still here if you are lonely. :-)
 
@JasperLoy Glad you're still here, but I wasn't lonely. I just saw that there looked to be a short conversation and now both conversants are gone.
 
8:48 PM
Why $$A'(x)u_1(x) + B'(x)u_2(x) = 0$$ in here
anyone like to answer this on main site
 
9:54 PM
Hi Dylan
 
Hello John.
 
Spookily quiet in here this evening ...
 
$\Huge{\mathbf{BOO!}}$
 
@HenryT.Horton Hi Henry
 
Hey guy
 
10:18 PM
Does anyone know a good notation/way to talk about sequences of sequences, and subsequences thereof?
 
user19161
@HenryT.Horton Hey, I like to say boo in the other room too! We have a lot in common then...
 
I just got done typing this answer up, and found the language rather unfortunate.
 
@AlexBecker At first glance, that is hideous :)
 
user19161
@AlexBecker None that I know of, and I think I would write it that way too. But I might add a comma between n and m so that it is not mistaken for multiplication.
 
@OldJohn Indeed, hence the question. :)
@JasperLoy I tried that. but then I needed commas on top too to be consistent and for $C_{ijk}$, and it just became a comma fest.
 
10:25 PM
(pretty much what Jasper just said!)
 
@OldJohn Yes, but I ultimately took subsequences of subsequences of ...
So there'd be a very nasty tower of subscripts.
 
user19161
@OldJohn Sometimes, the subscript becomes another problem if there are subsubsubscripts ... !
 
@AlexBecker Yes, I have faced the problem, in the past - not sure there is a totally satisfactory solution
 
@OldJohn I added commas, and tried to improve the wording at the beginning a little. New version.
 
@AlexBecker I personally find that more readable
 
10:32 PM
@OldJohn I think it looks a little better. Probably especially for people with smaller screens and less young eyes.
 
@AlexBecker Like me :)
 
user19161
@AlexBecker I think the notation is fine. It's not complicated at all.
 
10:49 PM
Ironically, I've seen $C_{ijk}$ specifically used in a different context entirely, for connection coefficients and specifically as Christoffel symbols (when $\Gamma$ wasn't available for one reason or another).
Fortunately, I'm pretty sure there's no prospect for confusion here. :-)
 
@StevenStadnicki I'm guessing there are $n$ feasible notations and $m>n$ concepts. Pigeonhole.
 
Oh, without a doubt. If nothing else, there's a very small finite number of symbols to go around. And only so many concepts that can be described in 25 words or less.
 
I one says "excuses my French" could you also use it to say: "sorry for what I am about to say, take no offense"?
 
@JonasTeuwen yes - absolutely. It is a very strange English idiom :)
 
I've only really seen it in the context of 'I am about to swear vigorously and wish to offer a token apology in advance'.
 
10:57 PM
Good :-). Then I was use it before I say something outrageous.
 
It is often used to warn people that you are about to use swear words, but you don't want them to be offended
(What Steven said!)
 
user19161
@JonasTeuwen I notice you making a lot of mistakes today, maybe typos. Or you are sick.
 
Yes, sick.
 
user19161
@JonasTeuwen Are you feeling better than yesterday?
 
@OldJohn I want to call some group idiots, and add that :-).
@JasperLoy Hmm, this morning worse now better.
 
11:04 PM
@JonasTeuwen If I want to call people idiots, I don't use idioms or careful speech - I just tell them they are stupid :-)
 
'Pardon my French, but ils sont imbeciles!'
 
usually forcefully!
 
@OldJohn I would tell it to someone that agrees.
 
user19161
If I want to scold people, I just do it!
 
@StevenStadnicki Wow... I'll use that one.
 
user19161
11:04 PM
So I think the expression is quite stupid.
 
user19161
It's for cowards who want to scold people but are afraid to.
 
@JasperLoy I am not afraid. I am just talking to some ... professor.
 
user19161
@JonasTeuwen You want to tell him he is stupid?
 
Nope.
 
Time for sleep - g'night all
 
11:17 PM
good night, Old John
@AlexBecker are you there?
 
@t.b. Yes
 
I've seen your index battle.
 
@t.b. Sorry for the delay, I was reading.
Ah yes.
 
So, first I'd call elements of $C$ not $(x_n)$ but rather $x$, maybe even bold face
This frees one index for the sequence in $C$. You can then address the $n$'th entry of $x$ by calling it $x(n)$.
 
@t.b. Hmm. I was going off of the asker's notation, but that sounds cleaner.
 
11:21 PM
Now you've got your sequence in $C$, call it $\{x_k\}_{k=1}^\infty \subset C$.
The subsequence which makes the first coordinate converge is what I'd call $\{x^{(1)}_k\}_{k=1}^\infty$
Now the second subsequence would be $x_{k}^{(2)}$ and so on. Then you have the diagonal sequence $x_{k}^{(k)}$ and things look only half as nasty.
The trick is to re-index at each step and forget about carrying indices of indices of indices. After all, you only want a subsequence.
At least that's how I'd do it.
Okay, it's getting late here.
See you soon (whoever hears me :))
 
@t.b. Is this about the recent $\ell^2$ question?
 
@t.b. Hey!!
@t.b. For special topics in algebra we are studying matrix lie groups!!!
 
@t.b. Ack, I just missed you it seems :-(
 
@robjohn I think we both did :D
 
@BenjaLim If I did, you did :-)
@BenjaLim That's what we get for doing other things than chat.
 
11:46 PM
It's always a little bemusing to me what gets upvoted. I thought my answer to the recent induction question ( math.stackexchange.com/questions/174828/… ) was decent enough but I definitely would not have pegged it as 30+ good.
Anyone else have any responses (or questions, I suppose) they were surprised to see the tally so high on?
 
luck and traffic, I guess. it baffles me too.
 
I know that I don't upvote answers unless I'm pretty certain they are correct. Lower-level answers are easier to see the correctness of
Plus more users will know what the post is talking about
 
@StevenStadnicki It all depends on who reads your answer when.
 
@StevenStadnicki Still not as surprising to me as the base 10 question, which I half-expected to be closed.
 
@StevenStadnicki Of course, I don't think I've had a 30+ answer :-) Oops, one has grown to 34, but it took months.
 
11:57 PM
I was just about to say. And with an excellent picture.
 
I have a post that has slowly been getting votes... maybe someday it will be on the cover of Math.SE Magazine
 
There is nothing more puzzling than the voting habits of the MSE inhabitants.
 
00:00 - 14:0014:00 - 00:00

« first day (721 days earlier)      last day (4306 days later) »