The h Bar - 2018-12-17 (page 1 of 2)

« first day (2965 days earlier) ← previous day next day → last day (1947 days later) »

1:43 AM

@danielunderwood I tried using Astrophysics in a Nutshell once and it was just awful. Many things left unexplained or worse, even explained in a misleading way. I don't think the series has consistent quality, so you'd have to judge more by the author.

7 hours later…

8:44 AM

hello

9:29 AM

I have the loveboat themesong stuck in my head

1 hour later…

10:30 AM

Listening to the song helps to get it out of the system

10:41 AM

Come aboard

We're expecting you

3 hours later…

Anonymous

1:37 PM

@ACuriousMind Do you have any canned comment for answers based on papers published in non-reputable sources (or say they provide a self-produced YouTube video as reference)? I need to expand my canned comments list

1:59 PM

do you not trust vixra

danielunderwood

2:54 PM

Ahh I just took a look at vixra and George Rajna still seems to be at it

Which one is that again

danielunderwood

Somebody who seems to just continually put stuff on there. I think they're usually summaries, but it's still beyond my ability to explain the amount

Anonymous

6

Q: Who or what is George Rajna?

I don't know maybe this question seems off topic but anybody get the chance to look at vixra.org archive? It seems it is blown up by some manuscripts from a guy, which is called "George Rajna", and more interestingly all of them are about some ridiculously (excuse me I didn't find another appro...

online-publication

Anonymous

> George Rajna, a Hungarian-Israeli physicist, computer scientist, and Chess International Master, in 1987 among the Top 100 Players of the World .

Anonymous

Whew

2:58 PM

"they seem to be a reasonably well-educated person with extensive experience working for IBM (or so they claim), so not your average crank."

that guy doesn't know the Salem hypothesis

Anonymous

Aug 25 at 7:27, by Blue

Spheroid Theory of Everything

Anonymous

Lol, I just remembered this one

danielunderwood

3:24 PM

I prefer the sphere theory of everything

Who needs something as complicated as spheroids?

Anonymous

@Slereah It sounds wonderful! :)

Anonymous

It has the 80's vibes

Nidana is a Sanskrit word that means "cause, motivation or occasion" depending on the context. The word is derived from ni (down, into) and da (to bind, dana). It appears in the Rigveda, such as hymn 10.114.2, and other Hindu scriptures, wherein it means "primary or first cause, linked cause"; in other contexts such as Rigveda 6.32.6, nidana refers to a rope or band that links, binds or fastens one thing to another, such as a horse to a cart. == Buddhism == Nidana doctrines of Buddhism are attributed to Shakyamuni Buddha. It has two specific meanings. The more common use refers to the Twe...

Who needs the Spheroid Theory of Everything when there is Nidana

O and the Buddhist cosmology is super awesome to read because of its many many layers of reality, each more abstract than the previous

I mean, Naivasaṃjñānāsaṃjñāyatana is sooooo cool

too bad only the human realm is relevant to science

3:47 PM

@Blue actually listening to this one

Intro - (*ﾟーﾟ)

danielunderwood

I'm mildly concerned about people github.com/eeeeeeeeeeeeeeeeeeeeeeeeeeeeeeee/…

4:09 PM

@danielunderwood e.ee isn't up

not very professional

danielunderwood

4:27 PM

Yeah I saw that too. No startup funding from me!

Such a shame since they could jump on the EaaS market

hmmm

the pharma company in NY wants me to come in to do a "thought exercise"...I hope they realize I'm in LA and they'd have to pay for my travel to do so...

danielunderwood

5:07 PM

I mean surely it's come up that you aren't local?

yeah, but the way this email was sent didn't seem to acknowledge that lol

5:43 PM

I think I'd like to know the salary range before I commit to going to New York...

should I just straight up ask about salary range...

6:05 PM

alllright pytorch now supports cuda10

@Slereah there's a Salem Hypothesis and it doesn't have anything to do with the town of Salem, Massachusetts?

@danielunderwood @Blue @Slereah The papers by Rajna are curious enough, but none of them beats this one vixra.org/abs/1507.0081

danielunderwood

oh boy the first sentence has a grammar error!

he says it's peer reviewed though

@enumaris oh, no

it's a peer review

Nassif is reviewing Rajna's paper

so Rajna's paper must now be considered to be peer-reviewed

INdeed

6:13 PM

(or so, I imagine, thinks Nassif)

time to publish

danielunderwood

> I am a Computer Engineering and IT student at the time of this paper, and know that I will have to learn about Theoretical Physics

Anonymous

Lol, keeps getting better :D

danielunderwood

I sure do like Theoretical Physics

"This is related to my field in many big ways. I am a Computer Engineering and IT
student at the time of this paper, and know that I will have to learn about Theoretical Physics,
Big Data, Cisco and IBM technologies, etc. throughout my field. Quantum Computing is
obviously related to this and is a subject that I may want to later explore more about."

danielunderwood

6:15 PM

Don't you know about Theoretical Quantum Big Data Computing?

Powered by Cisco and IBM

I think the author went through the same school as our President with The Random capitalization

apparently SAP has a cloud platform but not public like Google or AWS

they just sell it to enterprise users

I wonder if I can convince them to let me have free reign over their cloud if I go work there...

danielunderwood

> similar topics such as Computer Vision, UML, and Big data

give me ALL the GPU instances!

Anonymous

> He also meant to inform the reader with standard definitions such as Particle Duality, Quantum Enlightenment, The Heisenberg Uncertainty principle and more

danielunderwood

lol I was just about to say that

6:19 PM

mmm not just enlightenment...but quantum enlightenment...

danielunderwood

Are you going to say that you don't know the standard definition of Quantum Enlightenment?

so, I don't think I can read through pbrt's writing style without some external motivation...

danielunderwood

You should read Rajna's take on physically based rendering

Anonymous

@danielunderwood We need a Wikipedia page for that, now!

@EmilioPisanty 'fraid not

It is named after a fellow named Salem

Anonymous

6:31 PM

Unrelated: The top bar for chat script has been fixed now, by Shog9, and is available here. I'm finding it pretty helpful to keep track of main site notifications, while in chat.

why's your hat sideways

and Slereah's hat is vertical

danielunderwood

Why's your hat invisible?

Anonymous

Dec 14 at 9:37, by Blue

And `tis the right way to don the blue hat ;)

I have no hat

Anonymous

@enumaris Join any SE site or upvote any >1 year old question/answer to get a hat

Anonymous

6:35 PM

It takes around 30 minutes

Anonymous

Even visiting the Stack Overflow for Teams page, gives you a hat, I think

bam done

danielunderwood

I request that I be able to use multiple hats at the same time

I mean sunglasses aren't a hat!

Anonymous

@danielunderwood Yeppie, I want that too

Anonymous

I want to wear all my hats together :D

6:44 PM

where'z my hat

It tiny

Anonymous

@enumaris You can increase its size. Use the "show control" option

alrighty

it still tiny

hmmm

Anonymous

You're not using the control right :P

Anonymous

See the counterclockwise arrow in red

Anonymous

Use that to increase the size

Anonymous

6:48 PM

Pull it in the north-west direction (on your screen)

Anonymous

Pulling it north also does it

it got bigger on the control screen

bit it tiny here

oh it just grew

Anonymous

Yesh, now it's BIG!

Anonymous

Refresh the page

Anonymous

You're a cool green avatar now ;)

6:51 PM

hmmm so the pharma company is willing to fly me out there...for one night...that's a lot of flying for an interview

6 hours to get there on Sunday or something, 6 hours to get back on Monday

hmmm

Anonymous

@JohnRennie Just found another easy-to-get hat. You simply have to visit the SO teams page. Notifying you, in case you run out of hats. ;)

@Blue Not yet; the canned comments I use can be found at physics.meta.stackexchange.com/a/10485/50583

I have a fresh hat

Anonymous

@ACuriousMind Oh, nice list

canner

Anonymous

6:57 PM

I'll have to modify it a bit for QCSE. But seems like more or less enough for the common cases. I'll add in the "non-reputable source" comment too, in my version.

I actually find I don't often have to comment on the reputability of a source. Usually posts citing "non-reputable" material also have other problems I find more relevant.

"The choice between the two is arbitrary but has a number of implications for how some of the geometric operations throughout the system are implemented. pbrt uses a left-handed coordinate system."

w...why...

Anonymous

Well, the other problem in most cases is: "the answer has way too much b.s. for me to even consider spending time on pointing them out individually"...but I don't know how to frame that politely. :P

Anonymous

Hiding it under the "non-reputable/non-mainstream" rug saves time ;)

@enumaris From a computer graphics standpoint, it is natural to choose x "to the right" and y "downwards". Then the choice between left- and right-handedness is the choice of whether the z-axis points into the screen or out. None of these seems more natural than the other to me.

7:03 PM

how dare they not conform to my physics conventions

3

7:41 PM

hmmm pbrt is going through constructing vector classes

3D vectors and 2D vectors

surely C++ has some built in vectors or something?

or a library like numpy to do that...

also, what if I ask this guy to book me a flight and hotel and all that and then I complete negotations with another company...do I just ask him to cancel it all...will he be on me for refunds or what the deal...

danielunderwood

There are some libraries like glm, but nothing built-in as far as I know

There is a built-in vector, but that's a vector structure not a math vector

I see...

danielunderwood

And just use your own right-handed coordinate system

That's what I did, but then I had to flip all the models I loaded I think

I'm not at the level of building my own rendering system yet

danielunderwood

Better yet, write things not with just vectors, but forms too

7:47 PM

I'm not looking forward to flying out to NYC one day and then flying back to LA the next day...gonna be tiresome...

6 hour flights man...

ugh...

danielunderwood

Yeah that's a lot. I sleep on planes, but still find them tiresome

somebody else give me an offer quick!

so I don't have to do this ._.

danielunderwood

I had a flight back from China with a crying baby the whole way and some random girl falling asleep on my shoulder...interesting times

D:

was she attractive tho?

could be a perk or a anti-perk

danielunderwood

I think she was, but didn't really speak English all that well

7:49 PM

I've had some random girl fall asleep on my shoulder on a train ride in Japan

lol

ppl be tired man

danielunderwood

lol I don't know how she managed it over the crying baby

Though I have a friend that seems to fall asleep on any car ride we go on even if other people are talking. I'm jealous

I think I've accidentally attracted a type of recruiter on LinkedIn. Two seemingly different people with taglines of "Connecting extraordinary talent with exceptional opportunities" and "Connecting brilliant people with extraordinary opportunities"

Oh look they both work for the same recruiting company

@Blue OK, so I'm a sucker for hats, I'll give it a try :-)

danielunderwood

@JohnRennie your current hat is...something

8:07 PM

@danielunderwood it's a secret hat. I still don't know how I got it.

Hmm

Perchance you might have done some deleting?

It's one of the still secret secret hats

danielunderwood

Yeah I was thinking deleting of some sort

I just realized why do we get a duck hat on discarding a question

rubber duck debbuging

@danielunderwood I suppose once a person falls asleep, what they fall asleep on is not entirely under their control

danielunderwood

8:23 PM

Yeah that it true I suppose. And I feel like I've fallen asleep on a stranger's shoulder at some point as well

@danielunderwood lol

danielunderwood

btw @DavidZ did you make your profile image? I've always found it to be neat

@danielunderwood Thanks, but actually no, it was made by a friend of mine a very long time ago

based on a photograph

@DavidZ got your email, thanks for following up :D

no worries about it taking a while, these things happen

@enumaris no problem :)

8:25 PM

now I'm internally debating whether I want to go to NY for this interview or not...

so frickin far away -.-

also it's cold

(which I see as a bonus but not everyone does)

(I do not see it as a bonus)

an anti-bonus maybe

danielunderwood

@enumaris do you even have cold weather where you are?

lol

it gets chilly sometimes bro

I gotta wear a hoodie...

lol

8:29 PM

Pfff. It finally snowed here yesterday

danielunderwood

It snowed here last weekend, which was pretty bizarre. We rarely get snow in December and I don't remember it happening near the beginning of December as far as I can think

I am surprised how little I recognize this C++ syntax even though I've worked with it before...

danielunderwood

C++ has a lot of things and it just keeps growing

There are at least 3 ways just to initialize things I believe

If you get to messing with it a lot, I found Effective Modern C++ pretty useful/interesting

could be useful, thanks :D

amazon.com/Effective-Modern-Specific-Ways-Improve/dp/1491903996/…

this one?

danielunderwood

Yeah

8:35 PM

so is it like if you know C++ and now you want to do stuff in C++11 and C++14 you use this?

danielunderwood

There's also an Effective Python or similar that I've liked

I'd probably need something that is more beginner C++ friendly

danielunderwood

I'd call it a guide to all the ways to do things. I didn't use it when I was a beginner beginner, but I was pretty new when I picked it up

hmmm

"This book is the most challenging programming book I have ever encountered. Having spent many years programming in C++ I was hoping to get up to speed on the latest features in C++11/14.

The book examines extreme corner cases in C++ in which you could get into trouble. It assumes you already a highly experienced C++14 programmer. Language in the book is incredibly difficult, and it starts in the beginning. For example on page 16 there is some convoluted example, followed by a comment: "Yes, the syntax looks toxic, but knowing it will score you mondo points with those few souls who care.""

this review doesn't inspire confidence...

"This is a great book with lots of nuggets of wisdom. My only complaint is that this book is written from the point of view that the reader has already worked his/her way through every other Scott Meyers book out there."

My face is showing signs of doubt

danielunderwood

lol I don't think it's all that bad

I mean it will be if you try to memorize everything he talks about, but I don't think that's the point

Basically look at the table of contents at shop.oreilly.com/product/0636920033707.do for the short version. Aside from the first 4, most are just conventions that the text goes deeper into

Though I remember some of the stuff with references being a bit tricky

8:43 PM

looks like I would need a more beginner book lol

danielunderwood

I do feel very rusty flipping through it now

your sunglass too small to see btw

Also I just clicked your profile pic and the smaller thumbnail in this chat makes you look chubbier than you are

lol

danielunderwood

fixed

wow thanks

XD

danielunderwood

8:49 PM

I should probably update my photo at some point anyway. That one was years ago

My hair is pretty much all white now as I push into the mid-20s

that's a fast change in color...

might want to see a doctor about that lol

danielunderwood

It's all downhill once you hit 22 and graduate

lol

I didn't graduate until I was 29

28 actually

danielunderwood

But people don't call me doctor

nobody calls me doctor either

lol

8:57 PM

Jeez, what's with the sudden stars?

danielunderwood

Doctors are really funny

star party

it wasn't me

danielunderwood

I mean it couldn't be us since we can't star our own things

Unless we're in a star conspiracy

@danielunderwood It could be both of you

danielunderwood

Or it could be you!

9:01 PM

conspiracy

danielunderwood

You seem to be around quite often when the star parties happen

2

The more that I look at this book, the more I think that there are better things to spend time on than C++

@danielunderwood Astronomers like to watch meteor showers, does that mean astronomers cause meteor showers? :P

2

@danielunderwood lol

danielunderwood

Clearly I've forgotten how many things are involved in C++

@enumaris Just get the first Meyers book then; Effective C++.

danielunderwood

9:04 PM

@ACuriousMind isn't there an xkcd to that extent?

I think I'll have to work more with C++ at a lower level before I try to tackle any of these books

@danielunderwood Only one I can think of is this one

@enumaris Well if you understand the basic ideas of inheritance, I think you should be good with Effective C++. That's much more beginner friendly than Effective Modern C++.

I understand the basic ideas of inheritance as it pertains to Python classes

I think my course in C++ in college was only 1 quarter so it didn't get into classes really

it covered like all the real basics... standard syntax, loops, if, else, then, file i/o...

Well, there you go, that should be enough: Effective C++ will explain why you need to do things in a certain way with C++ to achieve the same result (e.g. virtual destructor), and other best practices.

9:11 PM

hmmm

As I remember it, Effective C++ is really very basic

More Effective C++ maybe had implementing a smart pointer as a chapter, so also basic fresh grad interview question stuff.

nobody interviewing me for C++

danielunderwood

@ACuriousMind maybe it as an smbc or something instead. I can't seem to find whatever I was thinking of. Oh well

Effective Modern C++ takes a much larger leap as it already assumes that you're familiar with move semantics, and basically know how to implement std::move and then goes into the minutia of the different type inferences of C++

danielunderwood

Though I did find this funny smbc-comics.com/comic/2011-03-26

9:18 PM

@enumaris If you don't intend to use C++ at work, why learn it? I mean, sure, it's a widely used language, but I think more modern languages with more modern ideas are probably more interesting and instructive.

I dunno lol

I haven't decided to learn it

just looking into it at a superficial level for now

Why C++ instead of something simpler (in language rules, though by no means in use) than C, or something more modern, like rust? (assuming that you're looking to learn a systems programming language that is)

I don't know anything about rust lol

the only language I know well is Python

somehow my script is breaking the associativity of set intersections

idgi

breaking associativity is pretty tough, nice job :D

9:28 PM

lol

@ACuriousMind If they could cause meteor showers, then yes; there would be a considerable chance of them causing meteor showers.

C++11 in particular added a huge number of new capabilities to the language.

wow 1 year after I graduated College

Now I'm feeling old XD

danielunderwood

@enumaris news.developer.nvidia.com/…

yeah I've seen that one I think

danielunderwood

ahh I just now saw it

9:39 PM

tbh the "AI rendered image" doesn't look all that impressive lol

like a blurry version of a picture

afaik they are just using GAN's similar to the ones we saw in those papers that trained GANs on natural language to produce outputs

"The demo allows attendees to navigate a virtual urban environment that is being rendered by this neural network. The network was trained on videos of real-life urban environments. The conditional generative neural network learned to approximate the visual dynamics of the world such as lighting, and materials.

Since the output is synthetically generated, a scene can be easily edited to remove, modify, or add objects."

ok. i dunno why it wasn't working before, but it's working like it should now

this part is intriguing though so they must have made some modifications to the GANs to be able to do that

@Semiclassical welcome to the world of debugging

danielunderwood

@Semiclassical Maybe you were witnessing "spontaneous associativity breaking"

lol

i mean, I know what I changed

but I don't really get why the old one didn't work

See my comment above^

9:42 PM

yeah

user image

I wonder how much you can move around in those "models"

@Mostafa wtf

My feeling based on the language used in that press release is it's a lot of hype...

danielunderwood

Yeah I wondered about it too. I imagine it may be sort of pre-rendered, allowing you to move around

9:44 PM

like what if I want to see the world from under the car

the funniest part of this is that i'm doing something in 7 steps total, and I'm timing each step

if the training set didn't have images of it

and they proceed as such:

it can't render it

danielunderwood

Can you do image embeddings?

9:45 PM

CNNs do that

danielunderwood

Like get a giant dataset of real-world images

640 ms, 641 ms, 1.5 s, 8.28 s, 51.5 s, 4min 18s

any of the layers before the final layer can be considered an "embedding"

and I'm waiting for the last one to run

@Semiclassical this image.... I don't know why with every new display/graphics model they claim exactly the same level of enhancement, but it's again all the same in the next model/version?!

danielunderwood

9:45 PM

Ahh I haven't made it to the CNN course yet

Though that sounds familiar now that you mention it

NLP word embeddings are also being rethinked

perhaps single layer embeddings is on the way out

perhaps people would rather move to whole pre-trained DL systems to "embed" words

This one is somewhat different (AI generated) but still

following the pattern in the last few times, should take less than a half hour

(hopefully)

well they'd have to be more specific on how they modified the GAN to be more than just a generic "contextual GAN" to make me feel excited...

pbr-book.org/3ed-2018/Geometry_and_Transformations/…

is this homogenous coordinates deal pretty standard across 3-D rendering software

danielunderwood

10:11 PM

I think so. Everything I've seen has skimmed over the extra coordinate without explanation though

basically to differentiate between points and vectors

make it possible to do translations

but one could imagine a different way to do things, sticking with 3D matrices

@enumaris From what I remember, I think it's standard in crystallography as well if you want a more physics-y reference.

seems like it's a pretty convenient way to do things

changed my hat haha

let's wait for it to update

Shazaam

welp. it's been 40 minutes so far

i think I can officially call my projection blown

10:27 PM

what hat is that

still fresh

pretty boring

ah

ahhhh

10:42 PM

ahhhhhhh

I've been seeing a lot of std::xxxx commands lately...and I just realized that's why in College my prof had me write using namespace std and so none of my college code had this std:: stuff in it...

it seems stackoverflow has some reasons not to use using namespace std

Ugh

Well the reason is the same you shouldn't import * in python

I was just about to say not to import an entire namespace :P

well my CS101 prof definitely was telling us to do using namespace std the whole time

10:48 PM

Maybe in a 101 you just teach for loops or something, so I guess taking shortcuts can be forgiven. That said, you shouldn't really teach C++ on such a course if you don't intend to do it properly.

danielunderwood

I used to do stuff like using vector = std::vector or similar for string. I don't know what people think about that

(or a typedef depending on version)

anyhoo...now I know what this std:: stuff is about

Maybe a bit unfortunate to abbreviate it STD though...

Well, you can use the using keyword.

To take parts of the lib

like you'd do in python

So you don't need to bring an entire namespace into scope.

like from this_lib import this_func kinda thing?

aye

10:56 PM

if I ever try to make C++ file I'm probably gonna get a ton of missing ; errors

danielunderwood

Just start adding semicolons to the end of your python lines to get used to it!

« first day (2965 days earlier) ← previous day next day → last day (1947 days later) »

Transcript for

Dec '1817

The h Bar

General chat for Physics SE (physics.stackexchange.com). For M...

join this room
about this room

00:00

06:00

12:00

18:00

all times are UTC

site design / logo © 2024 Stack Exchange Inc; legal

mobile