The h Bar

@DanielSank I liked the first five or six chapters of Reif rather a lot. But then I stopped reading because I knew that half the thermal physics class hadn't had enough QM to use it. I was sad.

So I tried to get on with Zemansky and Dittman (because the bookstore already had it), but I wasn't good enough to teach out of that book.

Perhaps someone else will teach that course for a few year and I can work on things I'm better suited to.

Thought I'd like to take another crack at it eventually.

DanielSank

6:33 PM

@dmckee the best thing about Reif, in my opinion, is that chapter 1 is a pretty good discussion of the random walk.

This respects what I think is the right way to teach physics: you have to start basic but solid.

Reif makes damn sure you understand the random walk so that as you do the rest of statistical mechanics your brain can always go all the way down to the ground floor without getting confused or feeling like you missed something.

guest

"Solid" is important.

user116211

@DanielSank: I'm not an expert on all this but Feynman quite clearly explained random walk in his 1st lectures.

dmckee

Does anyone know of a good statistical physics treatment that gives center stage to the classical treatment of the system?

DanielSank

Right, how many of us learned some hand wavy thing about "root N statistics" in some lab class and never had any damned idea what was going on?

I sure did.

Learning the random walk cleared that up completely.

dmckee

^Me too.

DanielSank

6:36 PM

@user36790 I would tend to disagree with the description of it being "clearly".

user116211

Clearly meant to say intuitively.

dmckee

@user36790 DS has high standards of "clarity", and rigorous demands on "basic".

DanielSank

Feynman repeatedly does something that drives me crazy: he explains something and tells you that it's simple and doesn't need fancy math. He convinces you that he must be right and that you shouldn't worry about the details. But he never gives you the tools you need to address those details, so when you go home and try to extend what you learned even a little bit, you're stuck.

I absolutely, absolutely loathe that style of teaching.

"Teach a man to fish" and all that.

user116211

@dmckee: The best part of DS is that it makes contrast on entropy is energy dispersal and entropy is disorder.

DanielSank

@dmckee DS?

guest

6:39 PM

I agree @DanielSank

DanielSank

@guest I had a professor in grad school for mechanics whose explanation of how to do wedge products of differential forms was "follow your nose".

dmckee

@DanielSank You. which means I'm usdure what DS @user36790 had in mind.

user116211

:P

DanielSank

@dmckee Ah yes. I was confused by @user36790's usage.

user116211

_/_

dmckee

6:41 PM

@DanielSank Like I said above, Feynman's approach was driven by intuiting the answer and then finding the simplest calculation that would let him go on. I assume he was a strong proponent of Wheeler's remark "Never make a calculation until you know the answer.".

2

If you prefer rigor you're never going to be comfortable with that.

user116211

@dmckee: Great quote; it's gonna take place in my quote list.

dmckee

Instead you should be looking to people like Bethe and Gell-Mann.

DanielSank

@dmckee I've noticed in my life that people who go around saying that are all incredibly skilled calculators.

2

dmckee

^ LOL.

DanielSank

It's no accident.

dmckee

6:43 PM

I'm sure.

DanielSank

Good physics means merging calculational prowess with intuition, but each one is improved by the other.

user116211

@dmckee: Do you know about the name whose query you've answered at Meta?

dmckee

I used to think I was a good calculator. Now I know I'm merely better than the average physics major, which makes me stand out in the general population but completely average among post-docs and professors.

@user36790 No. Why?

DanielSank

Feynman sure did use intuition, he also said that you should know how to solve every problem that's ever been solved, meaning that you should just know the answers to a lot of stuff.

However, at the same time, he was proud of his ability to do hard integrals. This is no accident.

user116211

So, let me introduce the man who left phd just because his supervisors believed in black holes.

user116211

6:45 PM

Stephen J. Crothers

DanielSank

Wow, I've got 6 stars on the board. In your face, @Danu :D

guest

:D

user116211

Are Stephen Crothers' claims legitimate?

DanielSank

Get rid of that space in the square bracket.

There ya go.

user116211

Rare species!

guest

6:47 PM

@DanielSank Feynman also said what I cannot create, I do not understand :P

DanielSank

Yes!

I relate to that so much. You have to do the calculation yourself.

guest

Exactly.

user116211

okay, where is ACuriousMind?

DanielSank

So when those people get up at a lecture and tell me to not worry about the calculation, I get a little bit... annoyed.

dmckee

I just noticed a bigger problem with my answer then the identity of the OP: the OP didn't actually have access to the comments system.

user116211

6:51 PM

@dmckee; He also avoided it in his plea.

dmckee

@DanielSank Too many of my students are dependent on CASs to work any integral that requires more than one transformation to solve. I'm trying to break them of it.

At least the math department is moving away from TI-89 based teaching. Yeah!

guest

Your math department needs a complete change in philosophy :P

DanielSank

@dmckee I'm of two minds on this. In some sense, why use a human brain as a lookup table if a computer can do it?

4

On the other hand, having those things in you brain changes how you feel and think when you see an equation.

user116211

Agreed.

DanielSank

You're a better physicist if you feel some kind of organic/aesthetic notions when you look at an equation rather than feeling like you need a computer.

guest

6:57 PM

You need them in your brain to build on.

dmckee

@guest They teach a real math major, but for the last decade their service teaching has been built around the damn calculators. Which means that physics major come to me able to do without aid only part of the calculus that I want to rely on them being able to handle.

Now they've decided they can justify stopping that nonsense.

And there was much rejoicing.

user116211

@DanielSank: How can one make himself a calculator?

user54412

@dmckee "does anyone know of a good statistical..." nope

user54412

been looking for years

guest

^

dmckee

7:00 PM

@ChrisWhite Can I ask you to get busy writing one? In your copious space time, of course.

user54412

how about I write a page for every bug in our codebase you find?

dmckee

Fair enough.

guest

Is there enough paper for that :P

dmckee

That's basically the problem of course. You almost have to have tenure and a lack of pressing research to write a book.

Or a sabbatical year which is what my father-in-law does. But he's in history.

user54412

and even then most people I know who wrote one did it on sabbatical

dmckee

7:03 PM

Great minds.

guest

Seven years is a long time to wait for a book.

Danu

@DanielSank Remember that time when I had 9? ;)

@DanielSank Sethna's book has the same approach (and is freely available online)!

@DanielSank Couldn't agree more; I read parts of those lectures and they were so useless to me

guest

He's at 9 now :P

user54412

@guest what's seven years away? me getting tenure?!

Danu

The most useful thing I got from it is intuition as to what Euler's constant $e$ really means

@guest Shitttttt :D

No, I only see 4

Nothinggggg

guest

7:09 PM

@ChrisWhite I meant between sabbaticals :)

Not everything that can be counted counts :P @Danu

Danu

It's about 9 different starred message @guest, on the board simultaneously.

guest

Thanks for the clarification :-)

I was just counting total stars.

Danu

(I think I, at one point, had 9; It could perhaps have been 8)

DanielSank

@Danu No, I do not. ::covers eyes and sings "Lalalalala"::

guest

In your system, multiple starred messages are not significant?

user116211

7:17 PM

@DanielSank; Feynman's 2nd volume has a separate chapter dedicated on electromagnetic mass.

Danu

@guest Not really.

Only if it's like 5+, does it get raeally interesting.

guest

I see, interesting.

user116211

@DanielSank: I don't think this concept is still accepted.

guest

In which case, I will have to change my previous statement to not everything that counts can be counted. :P

DanielSank

7:33 PM

@user36790 What does it mean?

Danu

I think the electron self-energy something something

guest

When does 0celo7's suspension end?

I saw JD drop by...

... perhaps to throw some salt on his wounds :P

dmckee

@guest Feb 8 '16 at 15:14. I think that is UTC, but I'm not positive.

guest

Thanks.

He should be back by the end of the Superbowl

Danu

@0celo7 2 minutes ;)

dmckee

7:49 PM

@Danu Check the date. I think that's tomorrow.

guest

Stop teasing @Danu :P

Danu

@dmckee I replied to a different message

Y'all are confused

user54412

0

Q: Rigorous proof of Ampère's law from the Biot-Savart law

Let us assume the validity of the Biot-Savart law for a tridimensional distribution of current:$$\mathbf{B}(\mathbf{x})=\frac{\mu_0}{4\pi}\int_V \mathbf{J}(\mathbf{y})\times\frac{\mathbf{x}-\mathbf{y}}{\|\mathbf{x}-\mathbf{y}\|^3}d^3y$$ where $\mathbf{J}$ is the density of the current distribute...

user54412

This isn't the first time I've seen such a question.

guest

and it won't be the last ;)

user54412

7:51 PM

Why does everyone want to start with the conceptually derived equation, and rederive the more primitive one from it? Maybe I don't get it because I never took high-school E&M, and therefore never really saw Biot-Savart.

dmckee

Mathematicians pull that "turning the development around" thing a lot. And they get somethig out of it because t lets them rebuild a field on the more sophisticated foundation.

But science has to start from observation so it is not always useful.

guest

Did you do SR in high-school @ChrisWhite?

gonenc

@guest why is he suspended?

user54412

@guest yes

guest

Dunno @gonenc

He will be able to chat in 19 hours

hubot

7:58 PM

My family doesn't accept my interests - physics.

Danu

"doesn't accept" as in what?

guest

Leave home :P

user116211

@hubot: Kill your family ;{

Danu

...

hubot

user36790: I understand that it is irony?

I argued with my mom that I'm trying solving Feynman exercises.

guest

8:00 PM

You are correct @dmckee he will able to chat in 19 hours.

user116211

@hubot: Ok, serious, why the hell your family is saying so? What's their point of view?

hubot

My mom and aunt think that I should go according school programme.

HDE 226868

@ChrisWhite Because conceptual stuff is often easier to use in basic problems? Not that "conceptual" is that intuitive in some high school E+M problems, but still.

user116211

@hubot: Huh?

guest

What grade/age?

hubot

8:02 PM

user36790: Because they think that I should go with school programme.

user54412

^ that sentence doesn't really make sense, or at least doesn't convey whatever you think it does

user116211

What School programme?

hubot

I have duty. Be right back

user116211

@DanielSank:Is the electromagnetic mass real?

user54412

I'm not entirely sure everyone is talking about the same thing on that page.

dmckee

8:13 PM

@user36790 As usual, the real question isn't if you can group terms together in that way but if it is a good idea to do so.

user54412

I'm also inclined to believe Feymann knew what he was talking about but everyone missed some subtlety and took one sentence out of context.

Danu

I found a sign error in his book ;)

user54412

In the right limits, I'm sure you could find exactly the quoted EM mass as a needed additional inertia working against changing the velocity of a charged sphere.

Danu

Definitely

And I definitely buy Feynman talking about electrodynamics, because much of his earliest work was on it (time-symmetric ED with Wheeler).

...let alone QED ;)

user54412

@Danu have you brought it up with a certain curator of the Feynman lectures? :p

Danu

8:17 PM

@ChrisWhite Yup.

I'm on the errata list :D

Fun fact: It was an original mistake, i.e. one really made by Feynman himself, present in the first edition. I was pretty proud.

It's also in the very last chapter of the third volume, which may explain why nobody found it yet!

dmckee

@Danu Nice.

Danu

> Dear Danu,

Thanks very much for your attention to detail, and for taking the time to write to us about the sign error you found in FLP Vol. III Chapter 21. That is a very old one too: it is present in the first edition. Congratulations on finding it! I have added the corrections to our posted list of newly reported errata for FLP Vol. III. These corrections will appear in all future editions of FLP: printed and electronic. They have already been made in the HTML edition at feynmanlectures.info.

user116211

@Danu: Hats off.

Danu

Thank you

I remember being afraid I'd messed up after all---and found a correct "error" ;)

dmckee

I had a graduate professor who would ask questions in class and then wait until someone answered. About once a week he'd look right at the person who spoke up and ask "Are you sure?" without regard to the correctness of the response.

user116211

8:25 PM

Witty.

dmckee

It made you think carefully about things.

I've taken up the habit with my upper division classes.

user116211

Many teachers do follow that.

Danu

Once a week? That's about how many lectures half of my courses have :P

hubot

What is more important - social or science?

user116211

Also, I felt ashamed when I couldn't conceive Feynman's lectures; after all he made these intuitive lectures for dumb like me :(

user116211

8:27 PM

Social Science.

dmckee

@Danu Because it was occasional, it felt like an accusation. I got a deer-in-the-headlight look the first time he did it to me.

hubot

user36790: You mean about Cognitive Sciences?

Danu

@user36790 You know what's really humiliating? Electromagnetic duality for children.

hubot

And History?

dmckee

Took me about twenty second to decide that I was sure. And I was right.

Danu

8:28 PM

@dmckee Yeah, that's nice :)

At my university, Mukhanov has the following variation:

Whenever I ask you guys something, that means that the answer is zero

user116211

Still, I couldn't take up the derivation of Liénard and Wiechert potential: feynmanlectures.caltech.edu/II_21.html#Ch21-S5

user116211

@hubot; What do you like?

hubot

I'm in chapter Atoms in motion

I want to discuss about some problem from Feynman exercises eg. the principle of virtual work

But I can't because my mom is observing me

user116211

@hubot: WTF!!

user116211

Is she communist?

user116211

8:33 PM

Or is she an SS?

hubot

Worse, my mom sometimes thinks that Earth is the centre of universe

user116211

Then kill IT.

dmckee

@Danu For children. In 209 pages. Nice.

hubot

My family think that god is creator of universe, the bing bang theory is false, the Darvin theory is false because Adam and Eve was first

user116211

Tell them to watch Ancient Aliens.

hubot

8:36 PM

I got to intelligent discussion

*end

dmckee

@hubot Well, then God built in all this tricky evidence for a reason: he wants us to solve his clever puzzle and scientists are just doing his will. Or something.

hubot

I must end this

Good bye

user116211

BTW, believing in God is not against science.

user116211

@hubot: RIP.

user116211

@dmckee: where are we living? A woman still believes this crackpot theory? Pretty shame. Poor hubot.

dmckee

8:40 PM

@user36790 The thing about religious belief is that if gives many people a structure and meaning for their life that they haven't found elsewhere. Who are you or I to demand that they give up that solace?

Though I would like to do without the baggage that comes with.

2

Danu

@user36790 ??? What does that have to do with anything

user54412

Is there a word in any language for that feeling that everyone is talking past one another?

Danu

@dmckee Such nice children :3

@ChrisWhite "shit"

dmckee

@ChrisWhite There should be.

user116211

Check what i got: physics.stackexchange.com/questions/138583/…

user54412

8:42 PM

@Danu ah, so French then

user116211

The same problem I faced

user116211

@Danu: So, want a history lesson?

Danu

@user36790 No, I'm just pointing out that communism has nothing to do with observing people

dmckee

WEll various communist regimes have been very big on ubiquitous survellience.

user116211

Are you from Left? I'm talking about KGB. North Korea and every bad Left.

dmckee

8:45 PM

Thought I'm not sure that are a a patch on the NSA.

HDE 226868

@user36790 Don't most countries in the world have a secret agency like that?

Danu

Various... other regimes have been very big on ubiquitous surveillance

Has there ever been a strong regime that was not?

user116211

@HDE226868: Like?

Danu

@user36790 I think that's wholly unrelated to communism

dmckee

If you want a very good fictionalized bit on it watch Das Leben der Anderren.

Danu

8:46 PM

@user36790 Germany, Holland

HDE 226868

@user36790 CIA, MI6 - from countries that were/are strongly anti-communist.

Danu

@dmckee A classic. Very nice.

Israel's is very famous

user116211

@Danu: Nazi??

user116211

SS-Far Right; KGB-Left ! Lethal!!

user54412

@HDE so how are the King potentials going?

user116211

8:48 PM

@Danu: You won! Bury these politics!! We want physics :P

HDE 226868

@ChrisWhite Fun-ish. Not mathematically challenging, fortunately. If you use numerical integration tools, that is.

dmckee

There is no point in arguing left/right. It's about totalitarianism versus a notion that the government is not the end all and be all.

3

user116211

@dmckee: That one is also going on in my quote list.

user116211

It's 2:21 AM; I gotta go; BTW, today is Mendeleev's Birthday; check the google doodle: google.co.in/webhp?hl=en

dmckee

The one time Siberia didn't mean "backwater". A great man, indeed.

user116211

8:53 PM

For now, @dmckee , @Danu good night.

HDE 226868

@ChrisWhite One thing I hadn't considered originally was that the high prevalence of binary systems means that looking at single-single collision rates is a terrible way to look at the evolution of the cluster as a whole.

user54412

indeed

user54412

they play an important part in core-collapsing, for instance

HDE 226868

Dynamical friction, right?

user54412

(cluster core collapse, that is, not supernova core collapse, though I guess binarity affects that too)

user54412

9:01 PM

@HDE226868 more than that, by having an internal degree of freedom a binary can absorb or release energy in an interaction

user54412

it's like monatomic vs. diatomic gases and their different heat capacities

4

dmckee

^ Cute.

HDE 226868

Oh, right. That makes sense.

user507974

@dmckee just peeked in here, is that comment about communism or modern politics

dmckee

Both. Seriously, classical communist make a big deal of a left right divide the perceive between their approach and that of "fascists". but they use the same methods.

Similar the major US parties make a big deal about the difference between them. And then they propose the same kinds of policies, just picking on different sins.

Gotta say I liked last Sunday's Bloom County.

user507974

9:11 PM

@dmckee Yea, Democrats and Republicans have like 85% the same toxic platform and only discuss that 15% difference

as a means to divide people into camps

@dmckee ironically if we talk about late marx works you see this really weird pattern where communism seems to actually resemble libretarianism a great deal

the communist manifesto is such an immature piece of writing (Marx was around 20 at the time I think) and a lot of interesting work was actually put into the ideas evolution

dmckee

You can also describe utopia by simply relying on the use of Human v2.1 (the release without the assholishness).

Implementation in the real world has to take the jerks into account. That's harder.

user507974

@dmckee I feel like a lot of the time the difficulty is overstated. I'd say most of the problem is the inertia in the current system

dmckee

@user507974 It's worth reading Scheiner on how deep and sophisticated the defenses of working social structure are, because that's the standard you have to work up to. And quickly to, or your experiment will be judged a failure before it gets properly underway.

user507974

People in any position tend to overstate the difficulty of their work, heck physicists are also a bit guilty of it (though we are great at horizontal transfer of knowledge) and I find that is most the case with businessmen and politicians.

@dmckee First name on Scheiner?

dmckee

Bruce. schneier.com/blog

The detail is in some of his books. Liars and Outliers might be a good place to start. He analyzes many social structure in terms of security and finds that there is a rich structure to detect, mitigate and punish cheating at all levels of society.

user507974

9:23 PM

well some kinda of cheating are heavily rewarded, at least in the modern environment

dmckee

@user507974 There is nothing remotely new about capturing the legal system for your own ends. Indeed there is probably less of it in modern technological democracies then in previous systems despite what you hear on the news.

user507974

i feel like its more a shift rather than mitigation, humans tend to find a way you know?

Physics Meta

10:17 PM

-1

Q: Alternative Ways to Earn Reputation

What are ways to earn reputation other than "Good" Questions and Answers that do and do not strain moderators on Physics?

discussion feature-request reputation policy moderators

Bernard Meurer

Just arrived safely to LA

@DanielSank

Secret

@Slereah hmm...

If Mitchell Porter's guess is correct then it should be possible to use some equations to check it out, but the problem is cold fusion is fringe because it still cannot be reproduced so far (and is not possible using our currently established theories, if my memory serves...)

DanielSank

10:34 PM

@BernardMeurer Good job.

Bernard Meurer

11:07 PM

It was harder than I thought @DanielSank. United screwed up with my luggage and I had to use magic to get it back

user507974

how much of a benefit does learning C++ provide over just knowing python for performance in physics simulation work

ACuriousMind

@user507974 Most of the people who write simulations use neither. It will be either pure C or even things like FORTRAN, from what I hear.

user507974

@ACuriousMind oh god fortran, how is that still a thing

ACuriousMind

Or is that Fortran now? They changed the spelling, I think

@user507974 Apparently it's efficient

user507974

well machine code is efficient too

ACuriousMind

11:12 PM

Maybe they just want to be able to tell people they still use Fortran, though :P

user507974

like thats something to boast about

but has it really been this many decades and nothing meets their needs quite as well

dmckee

@ACuriousMind The thing to consider about that kind of statement is the scale of the simulations.

By pure numbers most simulations don't need to be highly optimized and are written in whatever the writer knows.

Because most simulations are in support of other work.

In terms of results that you publish as "computational physics" it's a whole different thing: those are done with an obsessive focus on performance and that means Fortran or C or very carefully built subsets of C++.

I've written several simulations in my career using either CERNLIB (and therefore fortran 77) or ROOT (and therefore C++). But those are used to understand a data set, not to produce physics by themselves, so I didn't have to be an optimizing god.

I even know people who write their supporting simulations in python, and only turn to c when they have a bad bottleneck in the python.

That has the advantage of rapid development for most codes, and if the problem is small enough the coder's time is more valuable than the CPU's.

Bernard Meurer

11:41 PM

Python is the ultimate prototyping language IMHO. It so easy to build with it and not that complicated to port to another language afterwards

DanielSank

11:59 PM

Of course, python has the intrinsic problem of not having static types, so you get to discover all your bugs at run time. If you're writing a simulation that takes more than a few seconds to run that can be incredibly annoying and time-wasting.

Python's great for some things, but I'm getting to the point now where while I still recommend it to new programmers I always point out that it is just not really appropriate for large projects.

Transcript for