« first day (738 days earlier)      last day (4579 days later) » 

00:00
@JasperLoy I no has books.
@HenryT.Horton Hey, I am just wondering what does $\textrm{GL}(V^\ast)$ look like?
@JonasTeuwen Not sure what that would mean, but he's been there since 1998. My understanding is he's basically the masthead of Microsoft Research Cambridge.
that's V dual
well I know that $V^\ast$ consists of row matrices
but how does a linear transformation act on a linear functional in $\textrm{GL}(V^\ast)$?
user19161
@JonasTeuwen After I finish my three books I will send you a copy.
@HenningMakholm Do you know how this happens?
00:01
@HenningMakholm I hope to see Windows 9 written in Haskell in that case.
user19161
@ben Did you check out Bredon for AT?
@JasperLoy I am now using elements of AT
OK, everybody stop.
user19161
@BenjaLim OK. Use whatever you like.
00:03
@BenjaLim How what happens? Do you mean why Microsoft would employ programming language researchers?
@HenningMakholm No, I mean how does a matrix in $GL(V^\ast)$ act on a linear functional in $V^\ast$
because the dimensions don't match up
a matrix is $n \times n$
an element in $V^\ast$ is a row matrix of dimension $1 \times n$
I can't exactly multiply them together unless the vector is multiplied on the left
which is what I am guessing...
If $V$ is finite dimensional, then they act via the transpose
I would know why I would work for Microsoft if possible - they appear to pay well.
it is now finite dimensional
@HenryT.Horton meaning?
@BenjaLim Yes, but GL(V*) is the space of linear mappings V*->V*.
00:05
yes
invertible
user19161
@JonasTeuwen I used to work in a job that paid like shit, made me work like shit and made me do shit I did not like. I shan't say the details.
If $T: V \to V$ is linear, its transpose is a map $T^\ast: V^\ast \to V^\ast$ defined by $(T^\ast \varphi)(v) = \varphi(Tv)$ for all $v \in V$
@HenningMakholm so now what henry is saying is that
@HenryT.Horton ah so now I have the row vector on the left
@JonasTeuwen I think I would have some moral reservations, though.
which is your $\varphi$
00:07
@HenningMakholm But perhaps they call you back with: "Is money the problem?" - I had that once.
In finite dimensions $T \mapsto T^\ast$ is a isomorphism between $\mathrm{End}(V)$ and $\mathrm{End}(V^\ast)$
ok
@HenryT.Horton So in that case any element in $GL(V^\ast)$
arises as the dual of some invertible linear transformation in $GL(V)$
And $T \in \mathrm{GL}(V) \iff T^\ast \in \mathrm{GL}(V^\ast)$ unless something is horribly wrong
user19161
The worst thing is to do a job that requires you to go against your own moral conscience.
we are in finite dimensions?
00:08
Yes
but I think what you wrote above makes sense
Things do go horribly wrong in infinite dimensions
If you have a linear functional $\varphi$
and $T^\ast \in \textrm{GL}(V^\ast)$
Then $T^\ast(\varphi)$ becomes the linear functional
such that when you act on a vector $v$ it acts as follows:
@JonasTeuwen I've had employers ask me that. But no, I don't think money would be the problem there. (Since I work for a startup angling for market leadership in our niche, I have contemplated how I would react if we were bought out by MS).
$(T^\ast\varphi)(v) = \varphi(T(v))$ @HenryT.Horton
That is what I understand from what you wrote above.
00:10
Yeah matey
@HenryT.Horton
@HenryT.Horton all right no probs
just helping some friends understand the contragradient representation
@HenningMakholm Hmm... Well, then probably you mean "within reasonably boundaries which they could offer", right? What would you say, if I may ask (and get response)?
@HenryT.Horton Yeah I was just curious about $GL(V^\ast)$
@BenjaLim Oh, what's a friend?
I wanna apply a matrix to a linear functional, aka row matrix
but the dimensions don't match up :D
I was like
@HenryT.Horton Well they're learning representation theory of finite groups
something I've dabbled in before
00:13
@JonasTeuwen I'm not sure, actually. But there comes a point where more money is not that powerful a motivator. If I'm reasonably sure I can go find a job elsewhere that will pay my bills and allow me to sleep soundly at night, then what comparative difference would more digits on my bank statement mean?
Sure, not much. And I would not be "wanted" by software companies, it were more engineering firms...
And they seem to pay less.
@JonasTeuwen have you sorted out your accomodation?
Except Shell, but that has some moral implications as well.
I can ask pierre this wednesday
@BenjaLim Not yet, I am waiting for Pierre.
@BenjaLim No need, will be fine!
If he does not respond timely enough I will call him :P.
00:15
oh wow
skype
Wow?
No, just plain old phone.
EVEN BIGGER WOW
You know, with a hardware number pad and such.
user19161
@jonas I have been thinking of what to tell you after that email, but I still don't know.
What... wait... what?
@JasperLoy No worries.
@BenjaLim I just call him in his office - don't want to disturb him in private. But I prefer e-mail.
00:17
I am drinking coffee now :D
I am so not drinking the coffee you are drinking.
Son of a crack, already past 2 AM again.
Why does time go so fast when I am sleeping through the day?
user19161
@JonasTeuwen Go to sleep bro!
ok bye guys
going to work!
user19161
@BenjaLim Bye.
00:19
Need to work man.
I need Haskell on my iPad.
user19161
@JonasTeuwen Can't work without sleep, think.
Can't sleep without Haskell on my iPad either.
user19161
Why do you need the iPad? Just throw it away!
@JasperLoy For portable Haskell for one.
user19161
@JonasTeuwen I know exactly zero programming languages.
user19161
00:21
But if I were to learn one, I would do Python.
Depends on your definition of "knowing" I know between 0 and about... 500?
Ah, Python, I love Python like my brother.
user19161
Python seems to be commended everywhere.
What? Why?
user19161
I don't know. People like it? I think a lot of Debian software is written in Python as well.
It is easy and it somehow forces good programming practices.
So no Perl/PHP shitty.
user19161
00:26
Well perl-tk is used for GUIs.
GUIs are for homos.
user19161
TL installer and manager GUI use perl-tk.
Also, I know a perl guru which is homo. QED.
user19161
latexmk uses perl.
Yes, I know. I am rewriting it. In Haskell.
user19161
00:27
I am hetero, in case you forgot.
Is it my idea or is pdflatex really bad with pipes? Probably my idea.
user19161
I can still use pstricks with pdflatex.
@JonasTeuwen What do you want to pipe where?
Everything -- including the .log to my program.
I want to add some signing to the .pdf and such.
user19161
Don't waste your time. Just do math Jonas.
00:30
And be better at determining enough runs.
@JasperLoy I am not always able to do math, then I do programming! Like put R on the iPad. I did it and then figured out somebody else did it... but I did it better!
user19161
@JonasTeuwen So how did you do it?
@JasperLoy Just -- beter, it works together with the GUI (Springboard whatever it is called). Opens the pdf in the viewer etc.
@HenryT.Horton what you wrote above is not correct
@JonasTeuwen YOu can try making a named pipe called whatever.log, but it's not unlikely that pdftex will just unlink it and create a regular file in its stead.
user19161
@JonasTeuwen Just use TeXworks with latexmk. End of story.
00:32
@HenryT.Horton You should have $(T^\ast\varphi)(v) = \varphi(T^{-1}v)$
@HenningMakholm Yes, but named pipes are already a bit... well erm, not how I would like it.
I've tried it (very briefly) and it gets unlinked, I can do ugly stuff like pretend it was able to unlink it and it was actually not.
Or... I sandbox it and fix it through that. Must think about that. First Haskell!
@JonasTeuwen Pretend you were able to unlink it? Are you doing userspace file systems or something like that? Or hacking the source code.
@JonasTeuwen Signing has in PGP signing?
user19161
00:33
@PeterTamaroff Watching.
user19161
@JayeshBadwaik GPG
@JasperLoy Same Thing.
Almost (tm)
@HenningMakholm For example - or use pdftk on it.
@HenningMakholm No, I would not hack the source code, but use FUSE.
@JonasTeuwen Pipe is such a bad thing for pdftex, it has so many output streams.
So actually pdftex is bad.
00:35
@JonasTeuwen Yes, that was my middle assumption. Hacking the source would probably be at least an order of magnitude easier.
user19161
@PeterTamaroff Beautiful. There is a lot of geometry in algebra and complex analysis that is not taught in courses.
Well... I would prefer a wrapper so I don't have to go back hacking if I need a new version or if I want to install it elsewhere.
@JonasTeuwen To quote robert pike, (thought I myself won't believe it that much), the era of a program that could do one thing well was destroyed with the birth of perl :P
How is pdftex written, anyway? Does it have a PDF engine written in Web, or is it bolted onto the TeX kernel as a cweb extension thing?
@HenningMakholm I am really not sure about that, it never occurred to me to go that route as I would prefer to leave it like it is. It works fine and if I write a wrapper I at least totally understand how it works - or that is the idea.
@JayeshBadwaik :P. After Perl it could do many things... nongood?
So perhaps the FUSE thing is the easiest in that case. Or see why it exactly unlinks them.
00:40
@JonasTeuwen Yeah. Would it be so bad to let it drop the .log in the file system and then just clean up afterwards, though?
Anyway, going to bed. TTFN.
No. It is not, but other people have done that already.
@JonasTeuwen Not saying anything, talking about how multiple streams / single stream is something we can see.
BTW, what do you exactly want to do?
@HenningMakholm Good night!
@JayeshBadwaik Wrapper. To ljonastex file.tex. Gives me .pdf and no other shit (no intermediate). Also, do some signing or whatever on the .pdf.
Do enough runs of pdflatex so it gets out fine.
@JonasTeuwen OIC, and why is a bash/sh/zsh script not an option?
3 runs are enough for everybody, you can add a 4th for safety
and by the way, you would want to remove all the aux and outs only after the final version, else you will have to generate them everytime, slowing down your compilations
@JayeshBadwaik But you might need biber in between.
@JayeshBadwaik I am fine with some precompiled stuff. But not 9 files.
And it would be nice if I could put it in some subfolder called .auxillary-files.
Which is not possible with TeX Live I thought because it also puts the .pdf there.
@JasperLoy Fuq dis. I go to bed, good night Bro!
@JayeshBadwaik Thanks for the advice. I need to sleep. It is 3 AM.
00:57
@JasperLoy Good night
01:18
@HenryT.Horton I have only one chin, and I keep it to a goatee since I can't grow a decent beard.
user19161
@JonasTeuwen Good night bro!
user19161
@JayeshBadwaik Good night. I am going to bed soon.
@JasperLoy Good night.
user19161
@J.M. I shave my goatees when they become goatees, so no goatees.
@J.M. You got me editing and commenting on the new answers to this question.
I remember working on that problem while at Big Bear.
01:26
@robjohn "I will refrain from using Euler-Maclaurin..." - feels a bit like swimming while wearing a wool coat, I thought. :)
@J.M. I guess I should probably just remove that part. I had been doing a lot of stuff with EM and so I included that where it really didn't belong.
@J.M. However, I think part of the reason was to emphasize that the question was tagged as
No, I meant you could do it without EM, but it's somewhat slower... :)
...and yes, you wanted it to be at the level the OP can parse.
@robjohn Let it be. The OP will probably look up the thing if it interest him. It is always good exposure I will say.
(On a more serious note, swimming while wearing a shirt was one of the ways I got trained to build endurance. It's surprisingly grueling.)
@J.M. However, the Riemann sum proof that I added to a comment is pretty nice, too.
01:31
@J.M. @JM Next time, try swimming with sneakers! :D
@J.M. I usually do that so that I don't get sunburned.
@robjohn Oh, certainly.
@Jeff No, that's just nuts. :P
@JM Yeah, I know. I jumped in a lake once with my sneaks on. They now reside at the bottom of it. :D
@Jeff does that allow you to swim up on someone and catch them at unawares?
Tiptoe through the seaweed...
@robjohn no, it allows you to not swim at all. you basically have to kick them off if you want to have any control of yourself at all.
01:34
@Jeff I know; I was just joking.
@rob oh :D
Remember, I am twisted and mean >8(
@rob well, you're not the mean square today (i can't make out what your new icon is, though)
@Jeff you've heard of Möbius and his strip?
i have - but i don't really know anything about it
01:36
@Jeff You can see the full size here
:D the mean square on a mobius strip! :D
So basically, he's twisted, one-sided, and mean. :D
3
@J.M. two-faced fits in there somewhere :-)
lol
"I'm not two-faced! If I were two-faced, I definitely wouldn't be using this one."
01:38
mobius strips are the "one-sided" strips. do they really exist in life? Or just in graphical form?
@jm ha! :D
@Jeff You can make one with a piece of paper, scissors, and tape
oh yeah? but a piece of paper has two sides. i'm gonna google instructions on that
@Jeff Give it a half twist before taping
i'll have to try that out and look at it.
@Jeff Try coloring just one side of it :-)
01:41
i will. (I'll have my niece and nephew do it).
@Jeff Cool... outreach :-)
yeah. start 'em young! :D
@Jeff See this.
what plugin do you need for that?
it tells me 'no plug in available'
Whoops, let me look for an alternative...
@Jeff Try this.
01:47
A patent for a Conveyor Belt for Hot Material?
The conveyor is in the form of a Möbius strip, if you'll notice the drawing. :)
See this as well.
@J.M. They don't mention Möbius at all in that patent
@robjohn Yes, I'm not sure why. But the drawing shows that their conveyor is one-sided.
@robjohn : How did you create your current Icon? You probably have been asked this before, so you could just direct me to the archives instead.
@J.M. yes, I see the mention of a 180° twist
01:53
@what's the significance of it being a mobius strip to it being useful for conveying hot material? IOW, couldn't it just as easily have been a two sided escalator style strip?
...and still met the requirements?
@FortuonPaendrag I used Mathemtica and its built-in Texture[] primitive.
@Jeff The wear on the conveyor is distributed evenly.
ok. so it's just for wear-and-tear.
note, though, that each "tile" of the conveyor has two sides!
01:55
@Jeff but they are not orientable. You can't say that one is on a given side of the belt.
@rob yup. if you try "coloring one side" as you said, you will eventually color both sides of the tile.
Just like the paper used to make a Möbius strip has two sides, but when connected to make the strip, both sides become one.
The non-orientability is a global feature.
what do you mean by 'global feature'?
The neighborhood of any point will look ordinary-flat. Only if you zoom out entirely, will you realize that this is not an ordinary strip.@jeff
@Jeff well, as you say, locally the strip has two sides, and disconnected from the whole strip, it is orientable. But when put together to make the whole strip, it is non-orientable.
locally it is orientable, globally it is not
I have to go take my dog for a walk. BBL
02:01
@rob ok.
see ya
user19161
Could not sleep so back here...
Why would a PhD candidate answer this?
@Jas Did you try counting sheep?
user19161
@Jeff Well, it's OK. But that does not work.
user19161
@PeterTamaroff Hmm, why not? Isn't Alex an undergrad?
02:08
you could try reading math papers. that always puts me to sleep! :D
user19161
@PeterTamaroff You can answer too if you want.
@JasperLoy I linked to "Paul" not "Alex".
@jasper @peter wow, do you guys remember everyone's level?
@Jeff I don't remember much, I just "quoted".
user19161
@Jeff Well, only those I know about. Levels don't mean too much really. I know many idiots with PhDs, not in math though.
02:09
"Two years are fled. In the third year of PHD, my major is general topology and I'm facing with graduation from PHD. "
Maybe he was trolled?
user19161
@PeterTamaroff Oh you are saying he gave a bad answer? Well, it happens.
@JasperLoy Bleh. I think the more you know about math the more you realize some stuff simply hasn't a "easy" solution.
user19161
@PeterTamaroff Yes, though some hard stuff might have a solution that seems easy after you read it.
@peter I think the more you know about math the more you realize you don't know (for me at least, so far)
user19161
@Jeff Perhaps the same is true for everything.
02:13
@Jeff Well, I would say the more you read. Knowing is a side effect.
user19161
I would say @peter is doing very well for his level. He might win the Fields medal.
@JasperLoy !
@JasperLoy that is "pretty well for his level", i would say :D
user19161
@Jeff I mean he has not even begun first year and he is already studying such stuff!
@peter what're you studying?
@Jeff Just introductory general topology.
02:16
@peter we didn't have a single topology class even offered for my degree :(. i'm still not even sure what it is (I tried reading a Wikipedia page - but that is always so general).
user19161
@Jeff Not even point set topology?
@Jeff What is your degree?
@JasperLoy nope (I don't think so). what is that?
user19161
@PeterTamaroff Mine is 36.5 degrees Celcius.
user19161
@Jeff OK, that answers it.
02:17
@peter MS in Pure and Applied Math (I only finished the course work, still working on my thesis)
@JasperLoy You mean Celsius, mein Führer.
user19161
@PeterTamaroff Oh I am sorry. I am becoming retarded.
@Jeff What are you working on?
can you describe point set topology? maybe i'll recognize it.
@Jeff That is General Topology.
02:18
@peter something to do with something called the FPU Paradox (or maybe FPU Problem). Fermi-Pasta-Ulam (three well known mathematicians).
@Jeff Wha....?
@peter: oh. :D I'm gonna have to read up on it one day. all this stuff is so fun.
user19161
@Jeff General topology was one of my fave topics! It's really fun!
@peter Well, it's kinda hard to explain in chat, but the FPU problem comes from the first ever "math experiment" that FPU did on one of the earliest computers....
@JasperLoy i want to learn Game Theory and Probability (so I can play black jack and hold 'em and backgammon)
user19161
@Jeff Hmm, dunno any game theory.
02:21
@peter they programmed a one-dimensional (or chain, or string) of masses and springs. But the springs were not linear (linear spring: F=-kx)....
@peter they ran the simulation and measured the energy in each normal mode of motion (do you know what that is? it has to do with Fourier analysis)....
@Jeff I thought dynamical systems required topology of some kind?
user19161
@JayeshBadwaik Oh you like DS? The book by Teschl is awesome!
@peter oh, they started the system with all the energy in the first mode. they expected the energy to "leak" into all the modes and spread among them equally. instead, the energy did that, but then almost all of it returned to the first mode...
@JasperLoy you mean tedeschi?
@jayesh got me? i think dynamical systems require differential equations, which I also never took (my advisor showed me how to do the diff eq necessary for this system).
user19161
02:23
@JayeshBadwaik No, Teschl's ODE and DS.
@Jeff You really need differential equations dude. SPIEGEL "Applied Differential Equations" is quite awesome.
@peter and even now, nobody really agrees on why the energy returned to the first system. My thesis is to program a 2-dimensional FPU system and learn something about it.
user19161
@PeterTamaroff See my above recommendation. Online for free.
@peter how is that book "Div, Grad, Curl"? I like the title
@JasperLoy You still need to pay for the internet. Duh.
02:25
oh wait.. .i think that's vector calculus! :D
user19161
@PeterTamaroff OK OK.
@jasper can you repeat your above recommendation?
@JasperLoy Dude, never heard of it. I did my differential equations from krantz and then birkhoff rota then fourier series from brown churchill
user19161
@Jeff Gerald Teschl's "Ordinary Differential Equations and Dynamical Systems"
@JasperLoy sounds like it's just what i need, too
@jas .... and now it's downloaded! :D
user19161
02:27
If you guys are into probability I recommend Oliver Knill's "Probability and Stochastic Processes"
@JasperLoy Ohh, I should study it. My forays into dynamical system have been trying to get synchronized chua's circuit in a controllable frequency range
@JasperLoy I recommend Feller for probability.
@JasperLoy Teschl is nice :-) Kind of redundant to me though now though, will go through in detail it sometime though.
03:10
good night everyone. try to get some sleep!
I can't find the community wiki. =/
community wiki is a mode that posts (answers and questions) can be in, it is not a separate forum
(I run the risk of failing to get a joke)
there is meta (meta.math.stackexchange) though
03:43
@Jeff I did not much care for Div, Grad, Curl; I thought it wasn't dense enough, and it still didn't really answer my questions. But I think a lot of other people like it.
@anon I posted a comment to your answer and +1
yup
@anon I posted another comment
03:59
I don't really understand what you're asking. When you look at the dual space as column vectors you apply linear maps as matrices multiplying from the left, and when you look at dual vectors as row vectors you right-multiply by the very same matrix's transpose.

« first day (738 days earlier)      last day (4579 days later) »