The h Bar

peterh

00:33

You are developing a new OS?

@BernardMeurer Your link is bad, github gives 404.

Bernard Meurer

@peterh I made some repo changes

github.com/bemeurer/daedalos

There you go

Yep, I'm making my own OS

peterh

Well, kernel/x86_64 is a good start for the source tree :-)

Bernard Meurer

I try to keep it simple :p

peterh

Does the loading with grub already work?

Bernard Meurer

Yep, you can test it on your own machine

I can walk you through it if you want

peterh

00:38

Wow!

Bernard Meurer

although it literally just prints garbage to the screen right now :p

peterh

Rather I would test it in a qemu virtual box.

You could port the parrot virtual machine in it.

Bernard Meurer

It has that automated through make already

peterh

That would also solve the memory management problem.

Bernard Meurer

just do make run and it will run it on QEMU

and make run-vb on virtualbox

peterh

00:42

Ok, I do a clone!

Bernard Meurer

you will need to install xorriso for it to work :)

It's needed for creating the bootable ISO that goes into QEMU and VB

peterh

Ok, but why?

Bernard Meurer

b/c QEMU doesn't take direct kernels built with Multiboot 2.0, just 1.0 and I don't want legacy support

so you need to make an ISO with GRUB and boot from that

and grub-mkrescue (which makes the ISO) depends on xorriso

peterh

First I need an apt-get install nasm...

Bernard Meurer

but for some reason it doesn't install as a dep automatically so you gotta do it manually

Ah, yeah, and nasm

peterh

00:45

Yeah, I am on a 64k connection, it won't be fast

Bernard Meurer

You'll be the second person ever to run this lol

peterh

Uhm, also I have some github repos, which won't be ever cloned/forked by anybody...

Maybe nobody wants to play Icaros' role :-)

Bernard Meurer

FINALLY someone gets it!

You're literally the first person to understand it

peterh

My biggest hobby-project is now at around 200k source, I think the few people checked it didn't even tried to understand, what is this whole thing about. Well, maybe I had to write some doc :-)

Bernard Meurer

Link?

peterh

00:54

I hope you could note it

Bernard Meurer

Nope, too fast

try again

OK

Got it

peterh

Essentially, it is a C++ API around the select() system call.

Bernard Meurer

Didn't get this last one

peterh

But I forked your project, and you got probably a note about it in the github.

Bernard Meurer

Yep

Ah

I see

Lol yeah you need some docs

I have no clue what this is

good readmes are crucial

peterh

00:59

Yes, but first it should compile... now the problem is that although it is in C++, I don't use templates. I do everything with multiple inheritance, some of my final classes have a 5-7 level deep inheritance diagram from around 30 classes. Now I try to use some static code analysis tool to find the diamonds.

It is a re-implementation of my master thesis, where the main problem of my teachers was that I had to talk with a lot to understand, what is this whole thing about :)

Well, daedalos runs in virtualbox

at least in the Makefile

Bernard Meurer

DID IT WORK?!

make run-vb

peterh

I use qemu on my client machine, virtualbox doesn't have a working direct pci access, but I like to run w$ 3d things on my secondary vga card...

Bernard Meurer

then just make run!

The excitement!

peterh

First I download grub-common... xorriso is already done

Bernard Meurer

::heavy breathing::

peterh

01:27

::recompiling qemu::

Bernard Meurer

Dang it

I'm almost done setting up the xcompiler

which should make it a whole lot more stable

But I've had enough kernel dev for the day

peterh

You could make a VPS-only OS running a parrot VM in the kernel

It could be easy to implement, and it would generate surely a lot of clone & fork on the github :-)

it runs

...the qemu runs

It can't find a bootable device

I debug your Makefile

Could not read from cdrom: error 0004

Bernard Meurer

01:50

Hmm

That's awkward

I think it's b/c there's no xcompiler right now

I could be wrong

peterh

The bios boot code of the qemu tries to read the 11th sector of the iso image, it checks its first byte, it should be zero. If it is non-zero, it gives back an error 0x0004. I think the iso image creation wasn't parametrized enough well.

I check the grub-rescue

Bernard Meurer

Hmmmm

What's the output of lscpu | grep Architecture?

peterh

x86_64

Bernard Meurer

BZZZZZ

peterh

linux. Debian. stretch.

No, no, the iso is not bootable, at least with the emulated qemu bios

Your Makefile is very silent, how you made it? It doesn't show the executed commands.

Yeah, the V=@

Bernard Meurer

01:59

Just do make whatever V=

it will start talking

I gotta go home, brb

peterh

I commented out

ok

I play a little bit with the grub-rescue

Bernard Meurer

Keep me posted!

Ulthran

IT Crowd - Have You Tried Turning It Off And On Again?

►

peterh

okay :-)

@BernardMeurer This is the file structure on the generated iso: ./boot ./boot/grub ./boot/grub/grub.cfg ./boot/grub/roms ./boot/kernel.bin . Are you it is okay?

2physics

02:15

what time is it?

peterh

@2physics Funny question on a global internet chat :-)

2physics

@peterh yup didn't have anything else to ask :D

peterh

@2physics But I have :-)

2physics

@peterh well go ahead and ask

peterh

@2physics Sorry for the silly question. I try to calculate the weak hypercharge of nuclei by summing their components. But it is different for left- and righthanded quarks. I think, in a nuclei, we can calculate as if they would be 50% left-handed, and 50% right-handed. So, can I calculate with the average?

2physics

02:26

@peterh sorry, my knowledge of particle physics is elementary and can't help to answer that :)

Bernard Meurer

@peterh that seems correct, yeah

Which QEMU are you trying to run this with?

peterh

@BernardMeurer Is it the same on your system, too? Are you sure some tricky boot sector or any similar isn't missing?

@2physics Thanks, no prob!

@BernardMeurer 2.6.0, own compilation, from pure upstream

Bernard Meurer

@peterh I'm on the bus home, I'll check once there

peterh

@BernardMeurer I tracked the bios code in the qemu. It tries to read the 11th lba sector, then checks if it first byte is zero. To be able to go further, it must be zero. If it is non-zero, returns with error 0x0004.

2physics

@peterh what do you study?

peterh

02:32

@2physics programming, german & english language and hobby-level physics :-)

@BernardMeurer Now the whole iso is nearly empty with zeros until 0x8000. At 0x1be there is a little, 14 byte long binary data.

2physics

@peterh cool and what's your native language?

peterh

@BernardMeurer It seems as if any bios boot code would be entirely missing.

@BernardMeurer Could you upload your iso image to a gist?

@2physics hungarian

2physics

@peterh does it look like German?

peterh

@2physics No. It is farther from German as Persian.

2physics

@peterh didn't know..

@peterh I have a question , maybe you know about it.

peterh

02:44

@2physics Most similar language is finnish, but even it is quite different. For example, both languages use suffixes instead the prepositions, but even the suffixes meaning the "in" preposition in German and English aren't even similar.

2physics

@peterh you mean there aren't prepositions in finnish and hungarian?

peterh

@2physics None in Hungarian, and probably also none in Finnish.

2physics

@peterh you know in persian(it's called Farsi or Parsi also) you can add the subject to the verb, like suffixes

peterh

@2physics Everything goes with suffixes, and these suffixes can be combined. For example, plural has also a suffix. Now the translation "in the schools", gets first the suffix of the plural, and then the suffix for the "in".

@2physics Wow, also it seems tricky :-)

2physics

@peterh for example in english you have to use e.g, I+go separately. but in persian any verb has its subject with it. means you can say a word while it's a whole sentence :D

@peterh seems it's kinda harder than persian :D

peterh

02:52

@2physics yeah

2physics

@peterh also you know written persian is quite different from spoken. and if you try to speak like what you write, it sounds funny :D. the letters are written Arabic , although words are quite different. I mean forexample it's like writing hungarian by english letters.

@peterh btw, you know there are some storage spaces dedicated to internet users freely, to store their data freely. such as on gmail or upload websites and etc.

@peterh my question is about the capability of processing the data or installing programs remotely on some strong processors. Isn't there such a stuff on internet which is free of charge?

peterh

@2physics Is it possible, that you are writing farsi with latinic script? For example, if there are no arabic letters in your actual environment?

2physics

@peterh of course. it's called Finglish (Farsi+English) :D

peterh

@2physics Ok, but English isn't really consequent in the actual pronouncation of the letters.

2physics

@peterh you just write the way words sound

peterh

03:02

@2physics But, at least it has wowels. :-)

@2physics It doesn't depend on the cpu, it depends on the OS you are using on it.

2physics

@peterh yeah, for example there are some consonants which are not exist in english like : kh, gh, zh(like in mesure, or bonjure)

@peterh no no, I'm not talking about my computer's hardware. I asked, for example you imagine google has a service which allows you to use its computers remotely to process data, or to install a software and run it on their computers.

peterh

@2physics Amazon has a cloud. It starts to be a "well-going" industry. As if anything could be well-going in the sysadm world...

2physics

@peterh is it free?

peterh

@2physics No. I don't think any free service would exist for this. Running a virtual machine requires a lot of hardware.

Bernard Meurer

@peterh Okay, I'm home

let's see

2physics

03:13

@peterh okay..

Bernard Meurer

@peterh Okay, works here

I'll up the .iso

peterh

@2physics There are small providers in your home, and in Russia, China... you don't need to pay a lot of hard dollars for big U.S. companies, they have already far more than they should have. And, as I know, in Iran it is not so simple to pay on the net. But it is simple if you pay for an iranian provider.

@BernardMeurer okay! Let know it :-)

Bernard Meurer

@peterh drive.google.com/…

I've put the bin and the iso there

2physics

@peterh cloud service providers you mean?

peterh

@2physics yes

Bernard Meurer

03:16

Tested and works on VirtualBox too

2physics

@peterh yes you can't pay for international services here through iran. international creditcards are not supported here.

peterh

@BernardMeurer I am on a 64k link. Could you give me a simple link, where I can simply download the binary with a wget?

Bernard Meurer

@peterh Sure, one second

2physics

@peterh didn't know we have such a thing here lol :D just heard of some Grid Computing Groups

peterh

@2physics Don't think it would be a problem. In the EU, it has a significant effect to the budgets that people pays a lot to the ebay, amazon and similar US companies. In Iran, there is the possibility for this money to remain in the country.

Bernard Meurer

03:19

@peterh Why are you on a 64k link?

peterh

@2physics I am surprised, maybe with Russia and China, there should be a possibility of the international pays.

@BernardMeurer Because I have a hobby-project to make me independent from the "cloud".

2physics

@peterh that's right. but when you need you have no other option

peterh

@BernardMeurer Thus, I use a 64k mobile contact for everything, and use a lot of offline downloading things and localnet solutions. I don't use gmail, I load my mails in the background with an offline imap client, and use my own local mail server. For example.

Bernard Meurer

@peterh wget "https://files.fm/down.php?i=mxqkugn2&n=DaedalOS-x86_64.iso" -O DaedalOS-x86_64.iso

@peterh Hardcore

2physics

@peterh yea there are some with the russia but they are not so useful and as prevalent as paypal and etc

peterh

03:22

@2physics Why do you need a virtual server 10000 km far from you? I think, if you want to do something which is better to do out of the country, so or so you can solve it. In your place I would use a small russian provider for the task.

Bernard Meurer

Hope that one works

peterh

@BernardMeurer Thanks, I can parametrize wget :-) Your image is much bigger, 19MB... my is only 380k.

Bernard Meurer

@peterh I wonder what was going wrong...

Why was yours generating such a tiny image

peterh

@BernardMeurer You said the file structure on my iso looks okay. Could you please compare to yours again?

2physics

@peterh okay you are right, thankyou for the advice, I'll look for some in Russia.

Bernard Meurer

03:25

@peterh lemme mount my iso

@peterh AHA! gist.github.com/779dffff12cf5792d5eb8f10e12fb0d0

I was wrong about the file structure

peterh

@BernardMeurer First I have to download it. But I pasted the files on my iso, here is an echo $(find) output: . ./boot ./boot/grub ./boot/grub/grub.cfg ./boot/grub/roms ./boot/kernel.bin

2physics

@peterh btw sometimes I think it's not so hard to make a powerful pc using some rather cheap CPUs in a parallel structure.. at least it can solve my problem with my laptop :D I may try to make one and see if it works

Bernard Meurer

Yours is missing all the grub stuff

peterh

@BernardMeurer It seems, grub-mkrescue isn't a success stroy, at least the debian one...

Bernard Meurer

Very curious

Can you try manually running it?

Just manually do what's being done under the iso target at the Makefile

peterh

03:35

hehe...

0celo7

check your snapchat

peterh

make iso, relevant output line: grub-mkrescue -o ./build/x86_64/DaedalOS-x86_64.iso ./build/x86_64/iso

Heyy, makefiles aren't for microsoft users! Make them verbose, programmers want to see, what is currently going.

And don't remove intermediate files... now I need to move the rm -rf $(BUILD_DIR)/iso into the clean target.

2physics

@peterh nice talking to you. take care. bye for now

peterh

@2physics Good night!

@BernardMeurer It works also manually, but grub-mkrescue doesn't copy any grub data into the iso.

@BernardMeurer I debug it, it seems hilarious to me.

Bernard Meurer

Funky

This is why I don't like Debian :p

peterh

03:44

@BernardMeurer yes, I am also thinking to switch to an usable distro

@BernardMeurer I tried suse, because Germany rulez, but it is far worse as debian

Bernard Meurer

I run Arch and am very happy with it

I've heard good things about Void Linux as well

peterh

I've thought on Gentoo but it is not in the actual implementation phase yet :-)

Bernard Meurer

Okay, I gotta leave again, thanks for testing DaedalOS! See ya!

peterh

@BernardMeurer Your image looks as an efi thing...

@BernardMeurer I probably found the grub-mkrescue problem with a strace, it looked for its files in /usr/lib/x86_64-efi , to have this directory I have to install another grub package :-) Of course it spared me even a simple warning :-) Good night! :-)

Secret

Last night dream = super awesome because it is a rather cohesive time travel story. I have so far obtained 3 time travel ideas from it. The dream can also sound like author Mary Sue near the end though because throughout the whole dream the dream and I were battling for the control of the story's ending, despite I am only semilucid

It is however, not free of time travel plot holes, and it seems to challenge my own time travel model a lot as I was seen constantly thinking about it as the events transpired in the dream

David Z

04:33

@DanielSank About that custom flag you cast (hopefully you know the one): NAA is the option you're looking for.

0celo7

talking in code?

David Z

@DanielSank Well, we can't be sure. I did consider your idea of completely removing the HW-like close reason for a short time to see what people do instead. I suspect that most people would take it as a sign that homework-like questions are now to be considered on topic, or they would just move on from the question after not seeing an applicable standard close reason; either way, it would lead to many homework-like questions being left open.

Secret

05:24

http://physics.stackexchange.com/questions/276794/vectors-in-elementary-physics-vs-vectors-in-linear-algebra

Direction in an algebraic sense, is really the interrelation of the vector components. One can see this easily when attempting to visualise the computation of eigenvalues

For example while graphically you will have some solutions to the kernal of some matrix be vectors biting head to tail in a closed loop, or just pairs of vectors cancellign each other out, when these are presented algebraically, they are both done using 3 numbers

So what looks like numbers on paper can actually be very complex objects when drawn geometrically on paper

the pitfalls of visualisation in some context is that as soon you represent the concept on $\mathbb{R}^n$ distortion of the original concept can arise in the form of constraints introduced by the topological and geometrical nature of points in $\mathbb{R}^n$

one example is that you can end up with things that is recognised as a triangle shape while the concept itself does not say anything about triangles

DanielSank

05:44

@DavidZ Ok, noted.

@DavidZ Let me see if I understand the main points of your most recent meta post:

1) Our research thus far says that people use the homework close reason for whatever they want, i.e. there's no clear trend,
2) so you're asking people to come forward and either just say what they use that reason for, or propose a way to gather that information via further research.

Hello @dmckee.

Fancy a physics puzzle?

dmckee

@DanielSank It's late here, but sure.

DanielSank

@dmckee Well hopefully this is really easy.

There was a post today asking where the formula for the thermal deBroglie wavelength comes from.

The formula being $\lambda = n / \sqrt{2 \pi m k_b T}$.

dmckee

Yeah. I saw that post, and I didn't get it. The oddity seems to enter at "effective kinetic energy" which is not a concept I'm familiar with.

$E_K = \pi k_b T$ ?!?

DanielSank

Well ok but let's reason this out.

(sorry for delay, I lost connection to the site)

The Maxwell Botzmann distribution tells us that the mean momentum of a particle in a thermal gas is

$\left \langle p \right \rangle = \sqrt{8 k_b T m / \pi}$.

So, if we just take the deBroglie wavelength as $\lambda = h / \left \langle p \right \rangle$, then we get

$\lambda = \sqrt{\pi h^2 / 8 k_b T m}$.

However, that expression disagrees with Wikipedia.

So something is amiss.

Is it wrong to use $\lambda = h / \left \langle p \right \rangle$?

Secret

06:05

Are there any kinetic model of ideal gas that is independent of whether the gas is maxwell boltzmann, fermi or bose, bcause wikipedia said the expression applies for different types of gases depending on how the wavelength compared to the interparticle spacings?

DanielSank

@Secret I'm just doing classical statistics right now.

Secret

I am (roughly) guessing the starting point in deriving the effective kinetic energy might have nothing to do with assuming the nature of the ideal gas being a maxwell boltzmann distribution

DanielSank

I don't even know what "effective kinetic energy" means.

Let's go at this another way.

We know that the distribution of any one component of the momentum is $P(p) \propto \exp(-p^2 / 2 m k_b T)$ from the standard Boltzmann factor.

That's a Guassian with $\sigma = \sqrt{m k_b T}$.

The total momentum is $p^2 = p_x^2 + p_y^2 + p_z^2$.

Therefore, the distribution of $p^2$ is a chi square distribution of order 3.

Right?

...and the distribution of $p$ is therefore a chi distribution.

...which we can is the case from the Wiki article giving the Maxwell Boltzmann distribution.

Secret

06:30

ok, so we can start with maxwell boltzman, and our attempt seemed to be somehow missing a $\frac{4}{\pi^2}$ factor

Secret

07:11

http://physics.stackexchange.com/questions/201262/kinetic-energy-as-pi-k-b-t
https://www.usna.edu/Users/physics/mungan/_files/documents/Scholarship/ThermalDeBroglie.pdf

I find conflicting information here on how p is evaluated

one gives $\sqrt{\pi}$ another gives $\frac{1}{\sqrt{\pi}}$

David Z

08:32

@DanielSank yeah, that's basically the idea except for one thing: with respect to point 1, our (my?) research says that among the factors I used in the survey, there is no clear trend in what people use the HW-like close reason for. There could be a clear trend and it's just something I didn't collect data on. It's worth noting that actually being a homework question, or having educational value, were not on the survey, so I don't believe my results have anything to say about them.

In particular, if one wanted to test the hypothesis that people use the homework-like close reason on questions that they do not believe to be homework, or on questions that they do not believe to be of an educational nature, I don't think this research has anything to say about that hypothesis.

@DanielSank I think it'd have to be $\lambda = h\langle\frac{1}{p}\rangle$, which may be quite different if the distribution isn't sufficiently peaked

Secret

10:17

1

Q: what is the connection between two phenomenals?

what is the connection between two phenomenal states in the smallest of scale while they are divided by plank constant time or choronon with a similar of sense? even they are not really action as divided by them seft while actually been divided by the plus effect of camera of human,then they must...

Even a weird mind has a comprehension limit!

Mind=screwed

MonaLisaOverdrive

10:58

'lo all

Last week @Secret suggested textbooks by Shankar and Sakurai to supplement a Susskind text I'm reading.

Secret

does it help?

MonaLisaOverdrive

Would these be the correct books? amzn.to/2bQZ4My and amzn.to/2bJK9QF

@Secret: Haven't purchased books yet…wanted to verify I'd the correct titles

And wow. Textbooks are expensive

Secret

They looked different to mine, perhaps older/newer editions. I guess you better wait until the other h barers get on and ask them to be sure

MonaLisaOverdrive

Sounds like a plan. Meanwhile I'm off to slog through more Susskind

Secret

13:12

https://en.wikipedia.org/wiki/Conway%27s_Game_of_Life
That weird question caused me to revisit the concept of cellular automaton. Now suppose we have rules (be it probablistic or deterministic) does it account for everything we observed so far in the universe. What does phenomenon without rules look like, what exactly is no rules, is it even well defined...?

Gemini -- Self-replicating oblique spaceship in Conway's Game of Life Cellular Automaton

►

IT seems peopel still studying it, it is still fresh

Secret

13:33

https://en.wikipedia.org/wiki/Cyclic_cellular_automaton
Well, I guess nature love loops and repeating patterns

Dragonsheep

15:41

So I asked a question on stack exchange andit was flagged as too opinion base, so I was linked to here.

Do I just reask it here?

John Duffield

@Dragonsheep : you can try. I'm not sure you'll get any definite answer to your question. But if you asked me how should physicists read a math text differently from mathematicians? I'd say the physicist should have an eye on the underlying reality. He should ask himself How does this mathematical text relate to the real world?

Sir Cumference

15:57

Howdy

@JohnDuffield Say, you've said that we cannot know whether the Universe is infinite or not, right?

John Duffield

@SirCumference : not quite. I've said we have no evidence that the universe is infinite. And that there's a non-sequitur wherein people claim a flat universe is an infinite universe. I've also said that I think that the expansion of the universe is evidence that it is not infinite.

Sir Cumference

@JohnDuffield Hmm...but you define the Universe to be all of space and time, right?

As do most astronomers

John Duffield

@SirCumference : no. All of space. Space is expanding. The universe is expanding. Spacetime isn't. It's a abstract thing that models space at all times, and is therefore a static arena. It's typically depicted as shaped like a flared champagne flute without a stem.

Sir Cumference

@JohnDuffield The general consensus is that the Universe is all of space and time and its contents

Essentially everything that exists, has existed, and will exist, everywhere

John Duffield

@SirCumference : no it isn't.

Sir Cumference

16:08

Universe

The Universe is all of time and space and its contents. It includes planets, moons, minor planets, stars, galaxies, the contents of intergalactic space, and all matter and energy. The observable universe is about 28 billion parsecs (91 billion light-years) in diameter. The size of the entire Universe is unknown, but there are many hypotheses about the composition and evolution of the Universe. The earliest scientific models of the Universe were developed by ancient Greek and Indian philosophers and were geocentric, placing the Earth at the center of the Universe. Over the centuries, more precise...

John Duffield

If it was the universe wouldn't be expanding.

Sir Cumference

That first sentence alone has like 5 sources

@JohnDuffield Space is expanding

ACuriousMind

@SirCumference "universe" has different meanings depending on who says it. E.g. "Observable universe" only denotes a certain spatial region, but no extent of time.

Sir Cumference

@ACuriousMind We're clearly talking about Universe with the capital U

As in, everything that exists, has existed, and will exist

John Duffield

@SirCumference : space is expanding, the universe is expanding. The universe is all of space and its contents.

Sir Cumference

16:09

Space ≠ Universe

Space + time + matter = universe

ACuriousMind

@SirCumference Yeah, that's one meaning of the word. But, for instance, when people say things like "the universe has no center" or "the universe is expanding", they mean a purely spatial extent. You can insist on your "correct" definition all you like, it doesn't change the actual usage of the word at all.

Sir Cumference

@ACuriousMind I'm insisting on the general consensus

You're right, it can be situational

But if we say the Universe has an edge, it must include all space and matter, at least

@JohnDuffield That said, there any appreciable difference between an infinity or a naught beyond the observable boundary to us at the present, does there?

ACuriousMind

@SirCumference Sure. Note, however, that being finite and having an edge are two completely different properties

A sphere is finite but has no edge, a half-plane is infinite and has an edge.

Sir Cumference

@ACuriousMind The Universe, however, doesn't seem to be spherical. It has a small curvature, but we don't even know whether it is negatively or positively curved

John Duffield

@SirCumference : yes. IMHO if the universe was truly infinite it could not expand.

@SirCumference : as far as we know it has no curvature.

Sir Cumference

16:15

@JohnDuffield You've studied GR, right?

ACuriousMind

@SirCumference So? There are shapes other than spheres that are finite without edge.

Sir Cumference

@JohnDuffield Curvature simply means that, if we drew a triangle between spacial coordinates, whether it would be more, less, or equal to 180°

ACuriousMind

And I believe we do know the signature of the curvature, at least tentatively.

John Duffield

@SirCumference : yes. By reading up on it. I haven't been taught it.

Sir Cumference

@JohnDuffield All right. The expansion of the Universe means space is dilating. Literally, distances between matter are increasing.

ACuriousMind

16:16

We seem to be living in a weakly curved de Sitter space, if I can trust the cosmologists in our seminars

John Duffield

@SirCumference : as far as we know it would be exactly 180°. We have no evidence of any curvature.

Sir Cumference

They aren't moving away from a single point. Space is literally being created between matter. That's why they can drift away faster than light.

Proper distances are increasing.

Based on this explanation of expansion, why couldn't the Universe be infinite?

John Duffield

@SirCumference : right. I think a stress ball gives a nice analogy.

ACuriousMind

@SirCumference "created" is a dangerous word. Space(time) is not a substance in the traditional sense.

Sir Cumference

@ACuriousMind You're right. I'm oversimplifying it.

It's safer to say space is dilating.

John Duffield

16:18

@SirCumference : because stress is directional pressure, and space has its innate stress-energy. If the universe was infinite this would be counterbalanced at all locations. The universe could not expand.

Sir Cumference

@JohnDuffield Are we on the same page on the meaning of expansion?

My analogy is this: imagine all objects on a grid

John Duffield

@SirCumference : yes. Space is expanding.

Sir Cumference

Suddenly the grid is getting larger. Objects are moving away.

Here's a gif I made

Imagine all matter placed on that infinite grid.

Distances are increasing. Space (the grid) is dilating. That's how it can expand.

John Duffield

No problem. We've all seen the raisin-cake analogy. The cake rises in the oven. It gets bigger. The galaxies are the raisins.

Sir Cumference

@JohnDuffield I would've called it the dog-leash analogy. The ground expands. You and your dog are being separated, but the leash (gravity) is holding you two together.

@ACuriousMind By the way, feel free to correct me if anything I say is incorrect

@JohnDuffield When you say the Universe is finite, do you mean that there is an edge from which no matter exists, or from which no space (nor matter) exists?

John Duffield

16:25

@SirCumference : note that gravity isn't holding the universe together. A gravitational field is a place where light curves and your pencil falls down. But it isn't a place where space falls down. Or up.

ACuriousMind

@SirCumference The traditional analogy is dots on a balloon being inflated, the point being that there's no center to inflation. When you expand a flat grid, you'll have a point that stays stationary.

Sir Cumference

@ACuriousMind True. I always thought that the analogy of the Universe being on the surface of a spherical balloon was unintuitive. I get many questions on Astronomy asking if we can "get inside of the balloon"

ACuriousMind

@SirCumference That is a very traditional way of overextending the analogy, yes ;)

John Duffield

@SirCumference : the cake has a centre, and an edge. We have no evidence of any higher dimensional curvature as per the balloon analogy. Ergo I have to consider the situation wherein space has an edge. If you were at the "edge of the universe" and you took a step forward, you would find that you couldn't get past this edge. There is no space beyond the edge of space.

Secret

If the grid is so big, then where's the stationary point in the grid analogy?

Sir Cumference

16:28

@ACuriousMind Yeah. People take it so literally, I prefer not to use it. Even if it covers something that my analogy misses.

ACuriousMind

The basic fact is that manifolds can exist without being embedded in something, i.e. the "balloon world" must be thought of as only the balloon, not the space it exists in

Sir Cumference

@Secret There is none. The idea is that the grid squares (proper coordinates) are getting larger.

ACuriousMind

@Secret What has the size of the grid to do with it? In that gif, the stationary point is where the diagonals of the square meet.

dmckee

@ACuriousMind Which is, of course, the hard part when you first encounter the idea.

0celo7

@ACuriousMind they can?

Sir Cumference

16:30

@JohnDuffield So you don't think it is infinite with a boundary, but rather finite with a boundary?

dmckee

It took me quite a while to learn to take certain abstract mathematical constructs on their own merits.

Which is strange because with other—simpler ideas—I didn't have that problem.

Set were always just sets.

Secret

I tend to wrap my head around intrinsic curvature by imagining a gird with squished gridlines, so that any object that follows geodesics will have their trajectories "parallel" to those gridlines that are infintesimally close to them

ACuriousMind

@dmckee I think the more "geometric" a concept is, the harder we find it to separate our intuitive understanding from the mathematical abstraction

@0celo7 ?

John Duffield

@SirCumference : yes. As far as I know the universe began 13.8 billion years ago. I know of no higher-dimensional curvature, the Planck mission found no evidence of it. IMHO the only option left is a finite universe with some kind of edge.

Secret

that is, the whole gird might be lying on some flat plane, thus avoiding that allusion of the embedded space, but the curvature display itself by how the gridlines curve and twist around

Sir Cumference

16:34

@JohnDuffield The Big Bang is something no cosmologist will claim to understand

0celo7

@ACuriousMind are we talking about mathematically or physically

Sir Cumference

All we know is that something happened back then. Our modern theories break down at those scales.

0celo7

"something happened"

ACuriousMind

@dmckee I think you don't have an issue with "sets" because you don't have a preconceived notion of what a "set" is supposed to be. But for manifolds, you do have a preconceived notion of what a "shape" or a "geometric object" is.

Sir Cumference

Especially if you've heard that the Universe began at a singularity. That is simply a claim from extrapolation and using GR (which cannot explain those conditions, mind you. QM would do that better)

dmckee

16:35

I don't think it helps that many of the introductory examples (say a mobius strip representing a non-orientable surface) are embedable in 3D-space.

But they have to be to be useful as example for the really lost. And around and around we go.

0celo7

no mathematician visualizes manifolds intrinsically

just prove Whitney and then embed away

ACuriousMind

@0celo7 Uh, both? Neither does an abstract manifold need a space it is embedded in, nor does spacetime need a larger thing containing it

0celo7

@dmckee It's actually nontrivial to prove that you can embed all $n$-manifolds in some Euclidean space if you define them to embedded in the first place!

@ACuriousMind I don't think a "topological space" is very physical at all

Sir Cumference

Welp, a single conversation has forked into two separate ones.

0celo7

but I could imaging us sitting in some 256-dimensional real embedding space

dmckee

16:38

@0celo7 I'd heard that, but I've never chased after the proof.

ACuriousMind

@0celo7 What has that to do with anything?

0celo7

@dmckee The idea is that there might not be enough room in one R^m to contain all the possible "twisting" that could occur in an n-dimensional manifold

It turns out that 2n+1 is good enough

ACuriousMind

Anyway, I'll spend the evening playing cards now, so I'll disappear again

0celo7

This is an issue even if you define manifolds as subsets of R^N!

Secret

O, almost forgot the nonorientable manifolds. In that case if I have a space where the intrinsic curvature has a geometry of an nonorientable manifold, then it is not very clear on how my above "curved gridline painted on a plane" will help without the lines crossing in some really visually cluttered fasihion

0celo7

16:39

good god look at the time

cheerio

Secret

suppose I have a moebius strip with a grid overlaid on it. Is it even possible to project this 3D object onto a piece of paper so that the gridlines can correctly represent the curvature and nonorientability of such object?

John Duffield

@SirCumference : I don't think the universe was some point-singularity 13.8 billion years ago. Just as I don't think there's some point-singularity in the centre of a black hole.

0celo7

Got an Einstein quote to back that up

John Duffield

Yep.

Uh oh, the wife is calling. Coming down then Duffield? I have to go. Bye all.

Transcript for