The h Bar

Anonymous

12:00 AM

The chat remains way too silent on the weekends

danielunderwood

Enjoy Thanksgiving!

And yeah it is quite slow here on weekends

Not as much work to avoid doing

Secret

6:01 AM

>
In 2017, two research groups from ETH Zurich and from MIT reported on the first creation of a supersolid with ultracold quantum gases. The Zurich group placed a Bose-Einstein condensate inside two optical resonators, which enhanced the atomic interactions until they start to spontaneously crystallize and form a solid that maintains the inherent superfluidity of Bose-Einstein condensates.[8][9] The MIT group exposed a Bose-Einstein condensate in a double-well potential to light beams that created an effective spin-orbit coupling. The interference between the atoms on the two spin-orbit cou…

What exactly is light, why it can do so many amazing scifi defying things

'Oumuamua Is Not Aliens

►

The guy who wrote the article has done nothing wrong, but the journalists...

> oversimplification distorts things massively, like a black hole

>pun intended

rob

6:16 AM

@Secret Well, that's an easy question. Light is our name for excitations of the electromagnetic field in the frequency range which produces electrical impulses in our nervous system by exciting, but not destroying, retinal pigment molecules. Sometimes we use the same name for excitations of the electromagnetic field with any frequency.

I mean, doy.

:-)

Secret

well, if it is really that simple, then it is amazing how so many exotic condensed matter phases can be realised by configurations of optical lattices. It seems light has the ability to simulate any conceivable material and states of matter, something that no other material shared

Almost every new quasiparticle produced with apparent properties that reminds of scifi has something to do with optical lattices

rob

@Secret That seems more like a paucity of genuinely new ideas in scifi

Secret

from negative reduced mass, to bose einstein condensates, to monopole defects and so on

rob

But yes, optical lattices are cool

Secret

but why optical lattices are so versatile, while other matter or fields don't seemed to be as diverse in the phases of matter they can produce?

e.g. we cannot produce as many varieties of quasiparticles with e.g. magnetic fields, high temperature, low temperature, high pressure, microgravity etc.

rob

6:28 AM

@Secret I think this is because we have a very good model of electromagnetism.

That's different from condensed matter, where there are emergent many-body complications that we don't understand.

Secret

I see

rob

@Secret There are many-body effects in QED and the electroweak interaction as well. But because the fine-structure constant $\alpha \ll 1$, you can mostly ignore them, and you can decide which ones to not-ignore a few at a time.

Captain Bohemian

9:52 AM

what is a Lie three-algebra?

Slereah

context?

Captain Bohemian

"Embedding the group-theory structure in a Lie three-algebra, we derive Kleiss-Kuijf-like relations for bi-fundamental matter theories in general dimension. We investigate the three-algebra color-kinematics duality for these theories. Unlike the Yang-Mills two-algebra case, the three-algebra Bern-Carrasco-Johansson relations depend on the spacetime dimension and on the detailed symmetry properties of the structure constants."

Slereah

Three generators, maybe?

Captain Bohemian

but a usual Lie algebra, like SO(3), can also have 3 generators.

Captain Bohemian

10:18 AM

does NIKHEF have openings now?

Blazar

hello

Physics Meta

11:06 AM

0

Q: Is the demise of Stack Exchange (as we know it) ineluctable?

I feel concerned about a thought that I had recently. I am wondering if the current policy about duplicate questions may become problematic in the (very) long term for the health of Stack Exchange. Let me explain: as far as I understand, questions are marked duplicate when they have already bee...

discussion

Secret

1:50 PM

newscientist.com/article/…

Being able to track when and how exactly what lead you to feel some emotions is very useful. That's my and many people's example of consciously observing the unconscious

For thinking that quick unconscious association is irrational left a danger of us less responsible on our conscious thinking, and thus allowing bias to seep in in the thinking process

result in slow thinking that give wrong results

lılostafa

2:58 PM

I had a well-known professor on my radar as a potential phd advisor. He was at UT Austin. But I recently realized that he has moved to CUNY! users.ece.utexas.edu/~aalu

Why should someone like him (h-index=68) leave a top10 university and go to a very low ranked one?

It will seriously affect his career as fewer top grad students will be interested in his group

@dmckee Hi. Any thoughts on this? ^

vzn

in theory salon, 39 secs ago, by vzn

The Greatest Mathematician You've Never Heard Of — Robert Langlands’s work may have no practical applications—and he doesn’t mind at all / the walrus

dmckee

3:16 PM

@lılostafa CUNY is a special case because it is in NYC. There are people who have a very strong preference for living in big, high-density, bustling metropolises (-lii? -podes?). And some of those people seem to think that New York is special.

Don't ask me why. I like my elbow room.

rob

@lılostafa You have this backwards. University rankings are a function of what faculty are at the university doing interesting things, not the other way around.

dmckee

Austin may be the capitol of Texas, but it only bustles on a few blocks, and for some big-city services you actually have to drive to San Antonio (which should be about 70 minutes, but will take almost two hours with the traffic and the speed trap at Selma-Schertz).

(Mind you the "City Jail Cafe" (not actually affiliated with the jail) in Schertz is worth the stop.)

rob

If you have a faculty member who has built a research group with a good publication record and robust external grant funding, that group could pretty much move anyplace and keep doing their thing. Recruiting such a group would be a coup for a smaller school.

lılostafa

Do you think that's a plus for a grad student? (living in a big city)
and I have to mention I myself am used to big, crowded cities (Tehran population is ~10m)

dmckee

The other thing that can cause surprising moves is the need to be close to family who have chronic health problems.

rob

3:23 PM

Take you, for instance, @lılostafa. You were considering this person as a PhD advisor. Did you find them because they were doing the most interesting work at UT Austin? Or because they were doing the most interesting work in metamaterials? One of those is portable.

lılostafa

@rob I'm confused now as I read really terrible things about CUNY and its bad funding and management online. Yet two really great professors in my research area of interest have recently moved there

dmckee

@lılostafa Same trade-off as everyone else faces: the cost of living is a killer, but the service that big cities offer just can't be beat.

Cities have museums, concerts, plays, variety and quality of cusine, quality of medical care, ... the list just goes on and on.

Small places have ... access to the countryside. And in some regions really friendly and helpful people. But a lot of small towns with significant colleges also have a town-versus-gown problem.

rob

@lılostafa Every large organization has management issues. It could be that the department there has made it a priority to focus on your area.

lılostafa

say: instagram.com/shity_college_of_new_york

@rob Certainly their work is what matters most; but if the university itself has a lot of problems (a lot of teaching for grad students, funding issues, and stuff like these that I've read about CUNY) that will affect your work as a graduate student

Lu Bu

@JohnRennie Hello, mind validating my answer to a simple newton force problem?

John Rennie

3:33 PM

@LuBu I'm about to eat my lunch and I'm afraid lunch takes priority over physics!

rob

@lılostafa If you consider a teaching load as a grad student to be a "problem," I'd suggest you think hard about your planned career path as a PhD.

Lu Bu

@JohnRennie Oh )=

@JohnRennie I'll wait, but I dont think its gonna take long to validate.

Have a nice lunch.

=)

rob

As far as your instagram link of broken things at CUNY: I've encountered broken things like that every place I've worked. Hiring a physical plant staff that keeps up with repairs is a logistical issue that gets harder as your facility gets bigger.

lılostafa

@rob Isn't teaching too much a problem for someone whose work is doing research?

rob

One year at one of the US national labs, we were getting ready for a public open house, but there was a persistent water leak at the entrance to one of the public areas.

The physical plant's temporary solution was a tarp under the leak, which fed a hose, which at least diverted the leak from the middle of the walkway/truckway into a gutter.

I was walking past this with one of the staff scientists after it had been there about a week.

He pointed at it said, "Repeat after me: world-class facility."

Everyplace is dirty under the hood.

2

lılostafa

3:39 PM

:/

dmckee

@rob The pure length of time that some of the "temporary" solutions I saw around labs worked always impressed me. In one facility I where I spent a summer, there were decade-old divert-the-drip-with-a-tarp arrangements in a large experimental hall and they still worked.

rob

@dmckee We might be discussing the same tarp, then. I wonder if it's still there.

dmckee

Which tells you that whatever problems they have with capital repair budgetting they have really good people working there.

rob

@lılostafa There's a bit of a paradox here. Teaching is the primary activity of a university, and somebody has to do it. If you're at a place where the grad students are teaching many of the lower-level courses, it's because the faculty there don't want to do that teaching themselves --- a value that they will instill into the PhD-holders they produce.

It's a problem to teach too much if your primary interest is research, but it's a problem that nearly every academic research position has.

You want to realize this early in the grad school process, not after you've been job-hunting for a few years.

dmckee

4:15 PM

There are several ways to get to be a researcher without significant teaching duties.

On the "academic" side staff scientists at national labs have no regular teaching; many major universities have some arrangement for soft-money researchers (but that is a precarious way to make a living); and organizations like KITP don't require teaching.

If you are in a suitable subfield industry R&D labs are a place to do research without teaching classes.

Captain Bohemian

4:32 PM

teaching the same experiments again and again is very boring.

@JohnRennie I usually eat when I am too hungry to do physics.

graduate students are usually employed to teach the most boring courses, like undergraduate experiments.

dmckee

5:00 PM

@CaptainBohemian But it gives you an arena to work on your communications skills. Not that you are notably lacking in that area, but communicating technical ideas successfully to to least prepared student in your lab section is a whole 'nother level of skill.

It will teach you a lot about understanding where your audience is coming from.

Captain Bohemian

@dmckee but my research fields are not experiment-related.

dmckee

In my last semester as a teaching assistant (before starting as a research assistant) I got to grade upper-division E&M. That was ... quite a bit of work but I actually learned a lot more about the subject from the experience.

Captain Bohemian

and I don't really know much about electronics or the mechanisms underlying those instruments.

dmckee

@CaptainBohemian It doesn't really matter as far as "the experience of re-re-re-stating something that you thought was easy until you find a way to make it clear to this dummy" is concerned.

bolbteppa

Captain Bohemian

5:09 PM

@dmckee actually teaching experiments doesn't require a lot of communication because most of time during the course is left to students to do experiments on their own.

Anonymous

@CaptainBohemian Ideally that shouldn't be the case. Most students have no idea what they're even doing in the lab. A good communicator is always appreciated.

dmckee

You may be coming from a different experience than mine, because ... that wasn't my experience a s a student or a TA.

Captain Bohemian

experimental teaching only requires to teach for several minutes in the beginning of the course, then the following great propportion of the course is left to students to do experiments on their own.

Anonymous

Just because that's the norm, doesn't mean it's the best way to go about it. :P Oftentimes, during an experiment, I felt that I would have benefitted from a more thorough guidance throughout the duration of the experiment and not just a small lecture at the beginning.

Captain Bohemian

if you teach a lecture, you need to teach throughout the course, but lecture is always left to professors, so I have never got that experience.

Anonymous

5:14 PM

Otherwise, you might as well give them a free period to study up the background material on their own. In most cases, just a small lecture at the beginning doesn't suffice.

Captain Bohemian

I recall when I was an undergraduate student, our experiments were usually taught by teaching assistants having expertise in experiments, like electric engineering graduate students, not theoretical people.

Anonymous

@CaptainBohemian Yeah, it's similar here

Captain Bohemian

@Blue I prefer to teach theories than experiments because I don't know much about electronics, which most experiments employ.

Anonymous

That's understandable, obviously. I don't know why they'd even employ the physics theory grads to teach EE experiments at your place.

Captain Bohemian

there is even no the course electronics in our physics department.

Anonymous

5:21 PM

@CaptainBohemian Even at the undergrad level?

Anonymous

Weird.

Captain Bohemian

@Blue well, because teaching those experments is very boring and paid very low so that not many people would be glad to teach them.

Anonymous

I see

Captain Bohemian

@Blue in undergraduate we had basic electronics but it only taught something very basic and was not a required course.

Anonymous

Makes sense

Captain Bohemian

5:25 PM

the main boring part of teaching experiments is that you need to grade students' experiment reports.

Anonymous

Lol

Anonymous

Half of them would be "duplicates" and another half - "terribly written" :P

Captain Bohemian

when I was a full-time teaching assistant, I had two adjunct teaching assistants to help me grade those experimental reports.

Anonymous

At my place, I'm not even sure they ever check those reports, lol. I always see them lying in a huge pile at the side of the room. :P

Captain Bohemian

when I was a part-time teaching assistant, the lecturer graded those experimental reports and I only graded test paper, that's far easier.

Anonymous

5:30 PM

They mostly grade us on basis of our viva perfomance

Captain Bohemian

what's viva?

Anonymous

@CaptainBohemian viva-voce (oral exam)

Captain Bohemian

@Blue do experiment courses have oral exams?

Anonymous

@CaptainBohemian Yes, all of them

Anonymous

We have oral tests only for the experimental courses

Anonymous

5:36 PM

And written tests for the theoretical courses

Captain Bohemian

we (both when I took the course and when I taught others) had both operational exams and written exams for experimental courses.

Anonymous

Ah. We have 80 students in a batch, so conducting an operational exam would be a bit time-consuming. Also, our experiments are sometimes way too complicated/lengthy - some take an entire 3 hour period to complete.

ACuriousMind

@dmckee The proper Greek plural of polis is poleis.

Captain Bohemian

@Blue the class which contains too many students is divided into two parts in an experimental course.

Anonymous

They say the Latin plural is metropoles. Who wins - Greek or Latin? ;)

Anonymous

5:44 PM

@CaptainBohemian We were divided into three groups, yeah

Anonymous

But still, it would be too time-consuming to ask all 80 students to conduct an experiment during the exam. We only have one 30-student-capacity lab for each subject, etc

ACuriousMind

@Blue Greek, because polis is a Greek word that Latin borrowed.

However, loan words can of course form their own plurals according to the rules of their new language. So English would be at liberty to talk about "metropolises"

Using the Latin plural in English would be the worst of both worlds, though :P

Captain Bohemian

@Blue I think 3 hours is normal; our experiment course usually persists 3 hours (3-course time) but how long each experiment takes to finish varies; some may only require 1 hour and some may require more than 3 hours.

Anonymous

Apparently, people don't care about word-origins and word-supremacy :P

Anonymous

5:50 PM

@ACuriousMind Interesting

Captain Bohemian

some students can't finish experiments even before it's very late so you need to arrange them to do the experiments the other day when you don't have courses, so teaching experiments sometimes is very troublesome.

Anonymous

Did you take Latin/Greek in school?

ACuriousMind

@Blue Yes, I learned Latin and Ancient Greek

Anonymous

Nice!

ACuriousMind

I've forgotten much, but I can still recite the beginning of the Odyssey by heart :P

Captain Bohemian

5:53 PM

I think last time I meant to say Greek letters but missaid it into Latin letters.

Anonymous

Hehe, I vaguely remember being taught an English translation of the Odyssey (a small part)

Anonymous

Are the Greek grammar/spellings more difficult or easier to learn than English?

Anonymous

(I guess that would be subjective, but whatever :P)

Anonymous

I do hope to learn some of the classic Western languages someday. Should be a nice experience

ACuriousMind

The alphabet isn't hard to learn. The grammar is similar to Latin, and both are "harder" than English in that these languages still retain markers for all the different cases/tenses/voices/person/...

Anonymous

5:57 PM

Ah. Sounds similar to the Sanskrit derivatives in that sense

ACuriousMind

Greek may be a bit harder because it has a tense - the aorist - that isn't present in many other European languages

Captain Bohemian

@Blue when I was in undergraduate, I took the course basic Japanese, but now I feel taking a western language course may be more useful.

ACuriousMind

@Blue Since Sanskrit is also a Indo-European language, that would not be a coincidence.

Anonymous

Right. I remember you explained it to me once :P

Anonymous

@CaptainBohemian How's Japanese btw?

ACuriousMind

6:01 PM

I find the history of languages and the relations between them really fascinating

Anonymous

Same here (except that I know very little about it) :). At our schools in India, we hardly ever have a focus on the languages (and when we have - it's mostly mugging up poems or preparing for a literature test - memorization!). Almost always the STEM subjects took preference, and thus my knowledge of languages is rather lacking. Hopefully, I'll get to take up some interesting language classes in the future

Captain Bohemian

@Blue I didn't follow well when I just started to learnt it; because my memory is poor I often remember wrong word, like remember cat to mean dog. This occurred when I just started to learn English in junior high school, too; back then I usually earned the lowest mark in English in my class.

Anonymous

That's again one reason, I'm tending towards the European side rather than the US. The languages in Europe are rather varied and it would be fun to learn a new language in a new country. :D

Anonymous

@CaptainBohemian Ah, it would be difficult initially

Anonymous

I can understand

Captain Bohemian

6:09 PM

Japanese also takes a lot of memory to memorize words.

Anonymous

Hehe :P

ACuriousMind

@Blue Ugh. I hated analyzing poems (German or English, no difference). I loved learning what the Romans and Greeks considered poetry., and what stories they told in them.

Captain Bohemian

I don't know why my schools have never taught us English poems or English literature.

maybe they think knowing how to use English in a prosaic way is sufficient.

I did try to learn English poems from English teaching magazine but never get the clue of how to compose a poem.

danielunderwood

6:52 PM

I seemed to understand poems at one point as a kid, but don't get them at all these days

dm__

My legacy --
What will it be?
Flowers in spring,
The cuckoo in summer,
And the crimson maples
Of autumn...

-Ryokan

danielunderwood

My interpretation: writer hates winter

dm__

XD

Captain Bohemian

I also hate winter, which brings desolate atmosphere. I like summer the most.

danielunderwood

Evidently I don't know C as well as I thought I did...didn't know varargs messed with the stack stackoverflow.com/questions/8638881/…

I think I like winter better than summer. Summer is just miserable weather for me

Spring/autumn are my favorites though

dm__

7:04 PM

can always put on more layers; only so many you can take off

Captain Bohemian

I like the place where all year is like summer.

Anonymous

Summer is ~40°C here. So no, thanks. :P

dm__

summer in Alaska is perfect

Anonymous

Winter, at worst is 10°C

Captain Bohemian

when autumn begins, I feel depressed obviously.

danielunderwood

7:06 PM

Winter can get pretty nasty here, but it's fairly rare. And I think the nastiness is more with people not knowing how to handle snow/ice than the amount of it

dm__

when it snows people seem to completely forget how to drive

Anonymous

@danielunderwood You have the subways opened during the snowy season, no?

Captain Bohemian

it never snows here so can't imagine it.

danielunderwood

Pretty much no public transportation where I am

Anonymous

Uh, that would be bad :/

dm__

7:09 PM

Bury me when I die
beneath a wine barrel
in a tavern.
With luck
the cask will leak

-Ikkyu

alarge

@danielunderwood Everything needs to mess with the stack when that's the calling convention (I think for vectorized code you can go through xmm/ymm as the compiler can generate specialized versions of the functions with their own calling conventions, and nothing at all is needed when inlined, but there you go)

danielunderwood

I suppose that makes sense. I never really wondered about how varargs worked until I found out that you can do printf with format specifiers but not give arguments

alarge

This is not about varargs, but any function

ACuriousMind

I think we need to be more specific about what "messing" with the stack means here :P

danielunderwood

fooling around with!

ACuriousMind

7:20 PM

Sure, but is simply pushing things on the stack "fooling around with" it?

danielunderwood

But it seems I was missing some knowledge of how the stack works. I knew function calls pushed to the stack, but didn't know what actually was pushed

alarge

Stack as opposed to heap or register (but also stack like the algorithmic structure), as in the thing you grow and you have frames of.

danielunderwood

I knew of stack vs heap, but not how data is actually organized within

What I meant by "messing with" is it seems varargs is going to pop from the stack whether you supply args or not? But I'd imagine there are marker bits in the stack to prevent that from messing with the rest of the stack

ACuriousMind

@danielunderwood The consensus of the answers seems to be that what exactly varargs does on the stack is an implementation detail.

alarge

They mean that the entire calling convention is an implementation detail as it's not a standard. There is a de facto standard though if you use x64 on Linux etc

C standard, after all, leaves a lot up to the implementation (like the sizes of types etc; as a common example, 64 bit Linux is LP64 and Windows is LLP64, and more "exotic" architectures have whatever)

danielunderwood

7:42 PM

Do people often worry about these things when writing C? I've written a fair bit, but just always trusted the compiler to do what I expected

Aside from the occasional use of int32_t and whatnot

alarge

For some things, yes. If you need to write C, then you probably are writing code where for for whatever reason, you do need to think about low level details.

The only reason I've looked at assembly personally, though, is for computing performance, and the low level bit manipulation I've done has been for something like worrying about alignment (again for performance reasons).

But there's interesting stuff you can learn at that level

danielunderwood

Ahh what I've written in C was for interoperability with other languages, but then I realized that it was easier to write in C++ and create C bindings on top of that. I remember writing inline assembly for something, but it's only happened once. There certainly does seem to be a lot of interesting stuff going on though

alarge

I recommend "Hacking - The Art of Exploitation" if you want to learn some of this stuff like in a practical manner.

Or a book on computer architecture (e.g. Patterson&Hennessy)

NAND2Tetris will also make you work your way up, but it's a toy language so you won't be told of all the magic modern computers do

And Hacker's Delight if you want to learn some bit level hacking (not that useful really, but fun anyway)

danielunderwood

Thanks! I'll take a look. The only thing I had on my bookshelf was Practical C Programming, which looks like it doesn't have anything at that low of a level

alarge

(and obviously if you want to learn assembly and all these call conventions etc., go ahead and pick up a book on say x64 assembly, but I think your time is better spent on something else)

danielunderwood

7:55 PM

yeah I've always been kind of interested in assembly, but figured that I wouldn't really get much use out of knowing it

alarge

(and in interviews I only ever ask some low level questions, like to implement an allocator aligned at 16 byte boundaries or whatever, if the candidate claims on their CV to have knowledge of this stuff)

The most immediately useful thing with assembly that you can learn very quickly is to recognize some key things that make code slow. Learn to use a profiler (like perf), then look at the hot lines, learn to recognize when the code was probably not compiled with the best flags on (e.g. lack of vector instructions in a hot loop), or when there's excessive use of slow functions (default implementations of exp from glibc, or lots of divisions or whatever)

And cache misses

My level of assembly knowledge is not much deeper than that, but with those few things I've managed to squeeze out a lot of extra performance out of all kinds of software.

Emilio Pisanty

8:13 PM

@JohnRennie what's with the weird kink in this graph on July 2016?

$\large 🤔$

(admittedly not as extreme as the one in this one at January 2018 :-P)

Transcript for