The h Bar

EnlightenedFunky

00:25

Hello

Captain Bohemian

07:11

it's raining again. It turns out the afternoon rain only stopped yesterday.

if it rains, the air temperature wouldn't be as high as the usual.

but I don't like raining.

also, I feel the moderation of hotness due to rain not completely favorable. I am afraid the air temperature drops too much.

Captain Bohemian

07:28

earlier an earthquake occurred.

this kind of downpour would make my trousers all wet if I go out.

I wear slippers these days, otherwise if shoes get wet, it's harder to wait them dry.

also, carrying an umbrella is very troublesome. On Sunday, I left an umbrella in MRT. When going back to looking for it, it was not at where I put it. I guess some administrator salvaged it.

Rudi_Birnbaum

07:53

Hi there! Anyone interested in discussing a question on gauge transformation?

ACuriousMind

09:13

@Rudi_Birnbaum It's easier to judge that after we've heard the question ;)

Rudi_Birnbaum

I try to get a clear concept of the following procedure: Say you have a gauge independent global scalar property $j$ and a set of gauge transformations $G: j \to j'$ such that $j=j'$. This I like to call "normal" gauge transformation. Now if the property is not a global property but a local one which has a certain value in say ever point in space $\Bbb R^3$, and the values of $j$ in all these points are separately gauge independent.

Then instead of doing the same gauge transformation over the whole space $\Bbb R^3$, one could do a different gauge transformation $G(\vec{r})$ in each point $\vec{r}$ in space.

ACuriousMind

Are we doing discrete mechanics or are we doing field theory here?

Rudi_Birnbaum

@ACuriousMind I cannot see how its relevant. In my case it arises in quantum mechanics

to be more specific in magnetic response theory

ACuriousMind

@Rudi_Birnbaum It is relevant in that in discrete mechanics a gauge transformation depends only on time, while in field theory it depends on spacetime.

Rudi_Birnbaum

In my case time is not engaged at all.

its only about time-independent problems

magnetic vector potentials

time independent

So if one does these "spatially dependent" gauge transformations one arrives at a family of experssions for $j(\vec{r})$ instead of one expression for it all.

All I am out for is to fully understand how that scheme formally relates to a "normal" gauge transformation. Also if this procedure is used somewhere in other fields of physics.

in case some literature on it.

ACuriousMind

09:27

I'm not really sure what you're doing yet. If you have magnetostatics, then your gauge transformations are $A\mapsto A + \mathrm{d}\alpha$ for any function $\alpha$.

Rudi_Birnbaum

yes!

ACuriousMind

The $\mathrm{d}$ is a derivative.

I have suppressed the functional notation, you could just as well write $A(r)\mapsto A(r) + \mathrm{d}\alpha(r)$.

Rudi_Birnbaum

your $d \alpha$ is in my case always a "shift of the origin"

ACuriousMind

That's a very specific form of that general gauge transformation, but okay

Rudi_Birnbaum

Well instead of using any gradient of a scalar function we use only a shift of the origin and we only use vector potentials in the form $\mathbf{A} = \frac{1}{2}\mathbf B \times \vec{r}$

ACuriousMind

09:32

Then you've done what one calls "partially fixing the gauge"

Rudi_Birnbaum

We have only constant homogeneous fields as well

Yes sure

but now comes the trick..

So we have $\mathbf{A}' = \frac{1}{2} \mathbf{B} \times (\vec{r} - \vec{d})$.

for any constant vector $\vec{d}$ is a gauge transformation.

OK?

the trick is now to say we use a different $\vec{d}$ for any point $\vec{r}_0$ in space

where we want to compute our quantity.

So that we get $$\mathbf{A}' = \frac{1}{2} \mathbf{B} \times (\vec{r} - \vec{d}(\vec{r}_0))$$

ACuriousMind

But that's not a gauge transformation, unless your $\vec d(r)$ fulfills the differential equation $\vec\nabla \alpha = \vec B \times \vec d$ for some function $\alpha(r)$.

Rudi_Birnbaum

No, thats a misunderstanding.

What we do is to calculate each point in space separately.

So in each point in space its gauge transformation.

And the whole results we get are the correct results.

What we do is the obtain a family of expressions indexed by the spatial coordinates.

ACuriousMind

Ah, I see

Rudi_Birnbaum

instead of using one fixed expression for all points in space

nice :-)

So I kind of understand roughly ho it works but I want to find a/the clearest way to describe/understand that.

In this way you can obtain nice expressions.

You can e.g. make the diamagnetic part everywhere 0 (or the paramagnetic), formally.

Its a bit like using a potential energy function that is everywhere =0 while it still has the correct derivatives ..

ACuriousMind

09:43

Well, here's how I'd say it: You have a family of gauge-equivalent fields $A'_{r_0}(r) = B\times(r - d(r_0))$, and you have some function $f_{r_0}(A)$ that computes a value at the point $r_0$ from the gauge potential. If the result of the function is gauge invariant, then it follows immediately that you can also use $f_{r_0}(A_{r_0})$ to compute the same value, since all the $A_{r_0}$ are gauge-equivalent

Rudi_Birnbaum

Hey, thats nice! Thank you very much!!

Are you aware of anything similar in the literature? How much "standard" is the trick? Do you think it requires much of an introduction if one uses it?

ACuriousMind

I have a feeling I've seen it before, but I can't put my finger on it

Rudi_Birnbaum

Hm yeah. There is "a legend" in our field that the guys who introduced it got it from "particle physics". But their publications nowhere refer to anything like that.

ACuriousMind

I feel there is some standard computation where you "gauge away" some annoying term just for the point you're currently interested in, but I can't figure out what it is

The heat has probably cooked my brain :P

Rudi_Birnbaum

@ACuriousMind yes yes. And in that way you may even "gauge away" more than usually ... though it can be very confusing to handle these equations then. E.g. (to me) surprisingly you have to calculate the derivatives of $d$ when the computation of $f$ affords deriving by $\vec{r}$ ...

I somehow have the feeling there a should be a more abstract way to describe the relations between "normal" gauge transformations and this "trick".

Slereah

11:12

Engineering quesiton

How can I block out the sun

ACuriousMind

Slereah

Oh Mister Burns

ACuriousMind

11:36

the man has his flaws, but this idea is starting to look really good

Captain Bohemian

there was no sunshine here. It was overcast---the whole sky was white and looked heavy. It has been nightfall now.

the rain has stopped. I wonder whether the sky will clear soon. Now the moon won't come out until after midnight. The brightest celestial body in the evening will be Jupiter if the the sky clears.

Emilio Pisanty

13:03

guess who's on the Nature Photonics front page :-D

8

JMac

@EmilioPisanty Congrats! Is your paper "Topological Optics - Torus Knot Beams"?

Emilio Pisanty

@JMac yep

Jun 10 at 18:39, by Emilio Pisanty

Knotting fractional-order knots with the polarization state of light. E Pisanty et al. Nature Photonics, doi:10.1038/s41566-019-0450-2 (2019). Courtesy eprint here.

JMac

That has to be extremely satisfying

Emilio Pisanty

it is :-D

it was originally submitted as a proposal for the cover, which we didn't get, but it's still great =)

JMac

@EmilioPisanty That'll have to be your next paper I guess :D

Emilio Pisanty

13:09

I'm unsure that I can top this, but I guess you never know

JMac

You mean you're not willing to just arrogantly assume you're going to find something even more publishable? You're being way too rational about this.

Gyromagnetic

14:32

Gratz mate ;)

Jake Rose

16:43

@DanielSank just some of the stuff you said. It seemed quite pointed towards getting an admission of the EU wants to control everything. I may have completely misinterpreted and don’t mean any offence. It just seemed what you were questioning about was an already flawed view

@EmilioPisanty did you learn a lot of the topology stuff at university? Or afterwards?

vzn

17:20

. chat.stackexchange.com/transcript/message/51163396#51163396

ACuriousMind

@vzn Jeez. Can we please not drop (sensationalist reporting of) suicide into this chat room without any warning or prior discussion related to it?

vzn

17:39

wow, trigger warning(s) for this chat room now? guess its your rules as youve expressed in a meta post... it came from a blog written by a semifamous CS chairman etc... the link has the word "suicide" in it so yeah, it discusses it...

ACuriousMind

@vzn If you want to start a coherent discussion about the content of that post, be our guest. Just dropping the link and a stream of consciousness like "_academia + Phd research ~~worst case~~ horror/ nightmare/ tragedy scenario!_ o_O" in here, however, is not the way to do that.

Ryan Unger

what

Slereah

I should probably also do a list of math proofs on my site

So that I can refer to them in other pages

Without having to use BIBLIOGRAPHIES

vzn

it was my way to do it. dont regard any of it as sensationalist so far. but it could easily happen in the near future. seems to be a breaking story as they say in the news biz. yes, touchy/ edgy/ dark (etc) subj but this chat room has seen very many over the years.

Ryan Unger

@vzn the linked chinese article is absolutely insane, what the actual fuck

> I considered all the cases, I really can’t work them around. In order to make up for the fault, I decide to suicide.

is this real?

vzn

17:52

@RyanUnger it is a very regrettable apparently real story, suspect it will soon be reported. have some thoughts on it. suspect there is something about the students background that contributed to this. maybe mental illness, depression, code of honor from his culture, etc... there is more background in chinese in the msgs, am curious about the case now. the identity of the anonymous medium writer will come up to question, how do they know so much? private IMs? one wonders...

Semiclassical

an article about the "hidden story" behind someone's death last month...and that's not sensationalism?

vzn

@Semiclassical the blog seems to be based on facts and would not personally regard it as "sensationalistic".... how about "aggrieved" etc...? yes the deeper details may seem lurid or sordid...

Slereah

18:05

Important theory question

The metric on the worldvolume of a brane

$\gamma$ or $h$?

bolbteppa

What is $\gamma$

Ryan Unger

$\gamma$

Emilio Pisanty

@ACuriousMind CC @DanielSank I think an important distinction got lost in that conversation. Does Germany actually ban any (new) organizations, or is it just bans on attempts to revive certain well-codified ones?

i.e. is it just this

Strafgesetzbuch section 86a

The German Strafgesetzbuch (Criminal Code) in section § 86a outlaws "use of symbols of unconstitutional organizations" outside the contexts of "art or science, research or teaching". The law does not name the individual symbols to be outlawed, and there is no official exhaustive list. However the law has primarily been used to outlaw Nazi and Communist symbols. The law was adopted during the Cold War and notably affected the Communist Party of Germany, which was banned as unconstitutional in 1956, and several small far-right parties. The law prohibits the distribution or public use of symbols of...

or are there other mechanisms?

Semiclassical

it's interesting to see the clear differences between US vs. German free speech rights

Emilio Pisanty

hmmm, no, I do see the banned organizations listed under 'symbols affected', but they all seem to be along very clearly defined lines

Loong

18:13

@EmilioPisanty more this: gesetze-im-internet.de/englisch_gg/englisch_gg.html#p0121

> (2) Parties that, by reason of their aims or the behaviour of their adherents, seek to undermine or abolish the free democratic basic order or to endanger the existence of the Federal Republic of Germany shall be unconstitutional. The Federal Constitutional Court shall rule on the question of unconstitutionality.

Emilio Pisanty

@Semiclassical well, "unconstitutional organizations" is a term defined by the constitution, when it was written in 1948

that's a... different... historical context to when the US constitution was written

ACuriousMind

@EmilioPisanty That one doesn't ban organizations, it prohibits the use of symbolism associated with organizations that are already banned.

Emilio Pisanty

@Loong aha, thanks.

so that's specific to political parties, not arbitrary organizations

enumaris

hmmm

Slereah

@bolbteppa the metric of the worldvolume

Emilio Pisanty

18:16

@Semiclassical also, "the US doesn't ban organizations" is a tricky statement to make, given that there are bans on terrorist and criminal organizations.

Semiclassical

true

and there are restrictions on freedom of speech in the US, e.g. you're not free to shout "fire" in a crowded theater

ACuriousMind

@EmilioPisanty There's lots of clauses allowing to ban organizations in the laws governing such organizations, e.g. paragraph 3 of the law on "Vereine" says that organizations that are either criminal, unconstitutional or (interestingly) "counter to the goal of international understanding" may be banned.

enumaris

@bdegnan hey, I finally had time to read through the chat. I found it very enlightening :D You mentioned the power-reduction benefits of using neuromorphic chips comes from their analog nature (being charge integrators with a threshold). The Loihi chip from Intel is fully digital though, so do you have any insight into the benefits of doing that? o.o

Emilio Pisanty

@ACuriousMind oh, god, legal German =/

enumaris

I got my supervisor to assign me research into neuromorphic computing as my next task lol.

Getting paid to read papers ain't too bad..

Emilio Pisanty

18:20

Feb 22 at 17:55, by Emilio Pisanty

German is Hard

ACuriousMind

You kinda asked for it in this case :P

Emilio Pisanty

Feb 22 at 17:55, by Emilio Pisanty

Legal German is Harder

@ACuriousMind I guess =/

b degnan

@enumaris I'm sorry to say that I don't have a lot of insight into the digital implementations specifically. However, the power will be higher than the analog equivalents just due to the nature of digital. For instance, when I make digital ICs, they are never "on" because I make them operate at threshold. Intel doesn't do that. They are basically a "classical digital" shop. Sadly, I don't get paid to read papers anymore so it's unlikely I'll dig into it. :P

enumaris

XD ok

b degnan

You really need to read Stephen Brink's dissertation and that link on the neuromorphic roadmap from Jen Hasler. That'll pretty explain the low power side in analox.

analog.

enumaris

18:24

will definitely look into it

But I'm not a hardware guy so it'll take me some time to understand lol

(good thing I at least have a physics background)

b degnan

It's just physics. We are actually a bunch of device physicists. If you can do physics, you can do anything, including analog, digital, and neuromorphic design.

bolbteppa

@Slereah they both are (up to a conformal factor)

b degnan

There's more art in the analog side, sadly. :/

Slereah

@bolbteppa I know

I am asking

What is the best notation

most commonly used, anyway

enumaris

That's encouraging :D

Slereah

18:26

or unlikely to conflict with another variable of the same name

bolbteppa

$h$ for Polyakov, $\gamma$ or $G/g$ for NG

enumaris

Polyakov...reminds me of Tinker Tailor Soldier Spy

Slereah

also a vodka brand

enumaris

I need to at least figure out why this digital neuromorphic chip is "better" at doing what it was designed to do than a conventional digital chip...

vzn

@bdegnan thx for the refs saw the roadmap those were great :) welcome to the Physics club :)

b degnan

18:30

@vzn I'm usually over in cryptography because I'm trying not to ruin something. Thanks for the welcome.

enumaris

oo

vzn

@enumaris youve already figured out the main reasons. also havent heard of any major ML prj breakthrus using SNNs... yet. do believe that will change soon. eg havent heard of google doing much with them but presumably theyre on it to some degree. the big "big picture" ML breakthrus seem to be mainly running on tensorflow.

enumaris

does that mean you're familiar with Homomorphic Encryption as well?

vzn

@bdegnan lol what do you mean "trying not to ruin something"

Constantine Black

@ACuriousMind Greetings. May I ask you if you have a minute: Does an interview for admission in master physics program of Heidelberg contain questions around basic bachelor stuff or do they ask also about things in your resume? Thanks for your time.

enumaris

18:31

@vzn well the main reason I've seen so far is power usage...but that seems to mostly apply to analog chips rather than digital ones...

ACuriousMind

@ConstantineBlack I have no idea because due to nepotism students who did their Bachelor's in Heidelberg don't have to do an interview to be admitted to the Master's :P

vzn

@enumaris yeah think you raised excellent point about whether digital SNNs eg Intel new chip are all that beneficial, seems like they would lose the main (power) advantage.

Slereah

Is there a generic term for like

members of a configuration space

like fields and particle position and whatnot

Ryan Unger

@ACuriousMind Are interviews common in Germany

Semiclassical

"configurations"?

Ryan Unger

18:34

I only had to interview for one of my things and people were surprised

ACuriousMind

@RyanUnger no

Slereah

@Semiclassical A term that covers both a field $\phi(x)$ and a particle position $x(t)$

Constantine Black

@ACuriousMind Hahahaha! Sorry, that made me laugh-maybe they do this because they know the students?.. I don't get the Uni's logic on this but anyway. Thank you again.

Slereah

Things that go in a Lagrangian

enumaris

the press release is a lot of marketing language

hard to tell what the exact benefits are lol

The press release states: " Because the hardware is optimized specifically for SNNs, it supports dramatically accelerated learning in unstructured environments for systems that require autonomous operation and continuous learning, with extremely low power consumption, plus high performance and capacity."

Semiclassical

18:35

generalized coordinates, maybe? but that may still be too suggestive of particles rather than fields

enumaris

how does it get the "extremely low power consumption" part tho if it's digital...

hmmm

ACuriousMind

@ConstantineBlack Well, I think the actual reason is that HD is one of the few universities that even has an admission process for physics for both Bachelor's and Master's, so they want to vet students that did their Bachelor's elsewhere.

Slereah

Degreeeees of freeeeeedooooom?

I don't know

vzn

@enumaris so they think they have some power advantage. honestly sometimes chips are released without anyone knowing if they will work out. its kind a like game platforms / consoles + waiting for content model.

Semiclassical

Configuration degrees of freedom, maybe?

ACuriousMind

18:36

To be fair, there's also a grade cutoff for local students - I think you have to be better than 2.0 to be automatically admitted

Slereah

bit of a mouthful

Semiclassical

yeah

ACuriousMind

@Slereah Generalized coordinates.

enumaris

The paper only mentions this on power: "First, since the activity in SNNs is
highly sparse in both space and time, the activity gating that
comes automatically with asynchronous flow control eliminates
the power that would often be wasted by a continuously running
clock. Second, local flow control allows different modules in the
same design to run at their natural microarchitectural
frequencies."

Constantine Black

@ Yes, but the grade filter applies also to internationals, although I' m not sure how to translate it to the German scale. I have seen other universities(Freiburg,Bonn-Cologne,Aachen and others) also having a procedure. What do you mean by vet them?

ACuriousMind

18:40

@ConstantineBlack It's a bit of elitism where they think other universities' Bachelor's are worth less than their own :P

but I'm just speculating, I haven't talking to anyone responsible about the reasons

Constantine Black

Are they correct you think? I mean, are their students really better trained at some level?

b degnan

@enumaris You can get rather low power in digital if you operate at threshold. I made this thing: ieeexplore.ieee.org/abstract/document/7333527 It's asynchronous so I run it to just above threshold where you are still in the log region.

vzn

@enumaris its very complex to measure performance at this point. it reminds me a lot of the QC benchmarking efforts that have been going on for years. there is a lot of similarity. its about trying to compare apples and oranges almost.

b degnan

"on" transistors are actually a lie. They are always "on".

Constantine Black

It is interesting, but being a big and known institution gives the ability to think and act that way; and maybe in some cases it is effective as long as others are not being ruled out with no reason.

b degnan

18:43

Generally, digital will never lower power than analog, but it's often easier to burn more power to go faster.

ACuriousMind

@ConstantineBlack I don't really have enough comparisons to be able to say that

enumaris

hmmm

@bdegnan thanks for the references!

b degnan

@enumaris Look at this from me on electronics.se: electronics.stackexchange.com/questions/310660/…

There's the physics for you, and graphs that I measured off 14nm.

Constantine Black

That's fair. Thanks and have a nice night.

enumaris

tbf electrical circuit physics was never my strong suit lol

ACuriousMind

18:46

@ConstantineBlack You too :)

enumaris

I just know Kirkhoff's rules and basics like that...

b degnan

@enumaris physics is "we are conserving stuff". Still applies. No one generally likes my answers due to the math, but hey, that's the rules.

enumaris

XD

Semiclassical

electrical circuits seem like they should be fine as long as you've only got linear circuit elements

once you start worrying about nonlinear elements, idfk

b degnan

nothing is linear in circuits, not even resistors when you start talking about heat. :P it's actually nice, as you can just say +-10% of expected behavior due to processing errors, mismatch, and temperature. It's really qualitative not quantitative. My friends in particle physics don't have nearly as much slack

Semiclassical

18:51

yeah, hence why I'm okay teaching intro circuits (where I can pretend everything is nice and linear)

the only exception to that is showing off the I-V curve of a lightbulb

b degnan

I miss teaching. :(

enumaris

Hysteresis hisss hisss

Semiclassical

I'm coming to miss working with students

(grading, on the other hand, i can't say I miss)

b degnan

When it's like $10M to do IC design, they want the overhead so you don't get to spend anytime in front of students. They'd get $6.3M for being hassle and complaining that I wasn't on campus enough. I just couldn't deal. I now get put that money to fab. :)

... and Cadence. :/

btw, I see none of that. It's all fab money

ACuriousMind

Oh dear please don't use unescaped dollar signs to denote dollars

Semiclassical

18:58

\$

ACuriousMind

The MathJax script makes your message nigh-unreadable

b degnan

oops. :/

crypto.se doesn't have mathjax in chat.

Semiclassical

if you do \$ instead then it's not an issue

b degnan

sorry.

Semiclassical

we don't either by default, but we typically load a script

(see the mathjax link in the room desc)

b degnan

19:00

Summary: chip fab is stupid expensive. universities are silly and don't let you teach class. I now still do research, but the money goes to fab.

enumaris

@bdegnan is it maybe much easier to "learn" using a digital neurmorphic chip than an analog one? I'd imagine an analog chip would need to make physical modifications to the chip to change the weights and connections between neurons whereas maybe that's easier to manage using a digital chip?

Emilio Pisanty

@Semiclassical show me one person who can

enumaris

And at the end of training you'd have to save the learned weights somehow...which...for an analog chip would be some particular physical chip configuration? o.o

Semiclassical

lol, true that

in related news, I applied to an adjunct instructor post at a nearby liberal arts college last week. had a phone interview on Friday but haven't heard from them since

so...probably no

dmckee

@Semiclassical In many cases adjunct decisions are made at the very last minute. Especially when the question is “Can we get by with our regular crew or do we need the new guy who applied over the summer?”

Semiclassical

19:07

well, they had a specific class in mind when they sent out the posting via email

But yeah, hard for me to judge the context

b degnan

@enumaris neuralensemble.org/PyNN

I use that to program the networks.

I have a PYNN to fabric graph router.

It was the Ph.D. by Faiak out of Dr. Anderson's group at Tech. I'd have to dig to find his name though.

enumaris

hmmm

interesting set of API

Semiclassical

in related business, looking at indeed.com listings is (figuratively) nausea-inducing

enumaris

seems like there's a lot of things I need to look into lol

good thing I wasn't given a deadline XD

@bdegnan but regardless of the API used to program the networks, it seems like an analog chip would have to have a quite different way to handle the "learning" and the saving of weights that go on right?

b degnan

19:26

@enumaris The initial weights are set, but then the analog ICs learn through electon injection.

I assume the digital systems just use counter

In analog, I used programming of floating-gates.

vzn

@Semiclassical what class

@enumaris was looking at some specs & it seemed like there is a speed limit on initial neuron/ weight setup in SNNs that is somewhat "slow"... but presumably many apps "wont care"

b degnan

I would say from a "sane state", it'd take 10s of seconds.

Biology has years for feedback.

vzn

@bdegnan are you working on a chip design/ prj now?

b degnan

yes, but I am doing cryptography

vzn

@bdegnan cool what kind

b degnan

19:36

If you look at my recent work, you'll find that it's on hardware implementations of ciphers. I'm pretty good at extracting keys if I have physical access to the hardware.

Passively-powered ICs is currently my thing

I'm putting SIMON into them.

I used to work as the non-classified academic with the NSA's cipher group. I'm still doing work in that area.

vzn

interesting en.wikipedia.org/wiki/Simon_(cipher)

b degnan

SIMON is solid, but it's from the NSA, so people are paranoid. Cryptographers generally are not mindful of engineering issues. You often have to make engineering tradeoffs that are less than optimal than ideal from a cryptographic side.

vzn

this is all great but still long wondering about your yakuza + hells angels + collapsed parking garage story (as are others) o_O :P

b degnan

I'll let you know when the book comes out. It was terrible. I should have just stayed in Japan. :/

vzn

ok, will look for it, (think youve already sold me!) but book is probably quite awhile away

Semiclassical

19:45

@vzn algebra-based intro physics at a liberal arts college. apparently mostly biology students?

which I tried to take advantage of, insofar as I have a lot of experience with the intro physics for biology/pre-med sequence here

i dunno how I did with the phone interview, though, and I of course have no idea how many others applied

enumaris

20:18

funkay

DanielSank

20:34

@EmilioPisanty Oh oh! I wanna guess.

Is it..... an octopus?

That happened once.

Well Nature, not Nature Photonics.

Captain Bohemian

Yesterday I dreamt of visiting a research group I applied to and they are very friendly.

Novalium Company

@CaptainBohemian I dreamt of a giant minecraft zombie attacking the world on a space ship

Captain Bohemian

@NovaliumCompany I have not dreamt that kind of dramatic scenario for long.

I just dreamt that Father is sent to prison so suddenly that he doesn't have a chance to bid a farewell to me.

Novalium Company

20:51

@CaptainBohemian According to Sigmund Freud, you should examine your dreams a bit more deeply. The dream with your father reveals something about your relationship I guess

ACuriousMind

...you realize most psychologists today are not exactly fans of Freud? :P

Semiclassical

nor of dream interpretation more generally

Captain Bohemian

but I don't have relation with anyone now.

Novalium Company

Yep, but still, that guy was a legend

@CaptainBohemian Wut?

Captain Bohemian

actually I can usually guess what I make a dream from the feeling in the dream.

Novalium Company

20:55

Many times I have woken up on purpose, but I guess that's common

Many times, I think I'm awake but I realize I'm not, so I have to wake up again :D

Captain Bohemian

often when I read books about psychology and theories of Sigmund Freud, I feel them too confusing to follow.

Novalium Company

21:09

I can't find a single store in Bulgaria in which they sell ITO coated plastic layer :cc

Novalium Company

21:22

Why is it so hard to make transparent conductive material... (and expensive)

Novalium Company

21:39

I can only find a single place (and is very expensive) where they sell ITO coated PET Plastic Film, how is that possible wtf

Question (Google couldn't answer): Can I scrap up ITO film from an old touchscreen phone?

ACuriousMind

22:23

Finally a thunderstorm is coming and this heat is gonna end :)

Ryan Unger

@ACuriousMind hurray

been watching the Tour de France

it went from 100 degrees to snowing

Europe is going through some shit rn

ACuriousMind

@RyanUnger Well, there was probably some height difference between these places, no?

Ryan Unger

@ACuriousMind yeah but even in paris it was 20 degrees cooler

bolbteppa

22:38

@RyanUnger climate change

Ryan Unger

sure

bolbteppa

In other nature news: A Giant Asteroid Just Skimmed Past Earth, And We Barely Noticed in Time (simulation)

Transcript for