Hacker News Reader

solinenton Why is it important for a matrix to be square? (2018)

Yeah, for the most part all of engineering is equivalent to least squares I'd say, there's always some non-linear optimization procedure that uses a norm^2 metric since it's so well studied and solved already. It's disappointing as a mathematician, but the rest of me is fine with it :)

solinenton Why is it important for a matrix to be square? (2018)

Yeah, I wouldn't have used extension for a math audience, probably just "related concept", or "application of". In the software context it usually means the opposite--something specific that makes use of some existing functionality.

solinenton Why is it important for a matrix to be square? (2018)

Good point, while I don't need this clarification others might. I shouldn't have implied that they are the same for square matrices implicitly, hence my failed attempt at using "similar".

However, they're still an extension in the sense that singular values do correspond to (square of the) eigenvalues of the square matrix A A^T and A^T A.

solinenton Why is it important for a matrix to be square? (2018)

Off the top of my head,

- Principal component analysis

- Fitting a plane to a set of points

- Linear least squares

solinenton Why is it important for a matrix to be square? (2018)

An extension of spectral decomposition on a square matrix is called the singular value decomposition where any matrix can be decomposed. The trick is to diagonalize the matrix A A^T and A^T A, in this manner we can find the left and right singular values and have an equation which is very similar to the spectral decomposition. It has a very wide range of applicibility--most of which is also highly valuable.

There's also the Moore-Penrose pseudoinverse which also has plenty of applicaitons and works on non-square matrices.

solinenton Postbank to replace 12M bank cards after employees steal 'master key'

Almost always? Source for that? Very interesting, I guess during a siege there's a tradeoff there. If your postern is on the other side of a mountain range, then definitely it's not so bad.

solinenton Postbank to replace 12M bank cards after employees steal 'master key'

Would you build a castle with a back-door?

solinenton Procrastination is driven by our desire to avoid difficult emotions, says expert

Diet & exercise have been the cure for me, simple as that.

solinenton OpenAI API

Based on my experience with non-profits, they are just like regular corps except they don't pay taxes, and they're always attached to a for-profit interest. The real community organizations don't tend to incorporate, as then you have to hire people to manage the corp or do it yourself.

This OpenAI work is almost certainly a way for these bigger corps to collude. Proving that would be impossible, though.

solinenton Looking back at how Signal works

I've been paying attention as well, I've been using it since the early days, I also have an M&A lawyer by my side.

The Signal Foundation is responsible for all costs associated with the project--without it the project wouldn't function.

OWS was also largely a corporation, but it wasn't require for TextSecure to operate--a key difference.

> And while OWS/Signal Messenger did not have formal non-profit status prior to the Signal Foundation, it was never acting as a for-profit entity[2]:

A corporation, non-profit or not, is still a corporation. Based on my knowledge of full-cycle accounting for non-profits, they tend to make more profit, they just pay less taxes.

Having a non-commercial parent doesn't mean your business is non-commercial, Signal Messenger LLC is the corp which is associated with Signal Messenger. Signal Messenger could offer private equity, since Signal Foundation is not the exclusive shareholder.

I have an M&A lawyer right here if you want me to ask him.

In any case, I doubt they have enough funding to do this in a standard donation-based non-profit, it seems like there's a non-profit which is used to funnel money to the commercial entity, Signal Messenger LLC.

It's a standard structure, I'm sure, but it's definitely not your standard non-profit.

I'm more concerned about the coloquial use of commerical in any case--does the corp make profits or not, not whether they pay tax or not. I'd say they're looking to transition into a fully non-commercial entity, but for the purposes of this thread they are still a corporation and can still be held liable for their users' actions--in this case the app can still operate without a formal legal entity, but it would be removed from app stores, I'm sure.

solinenton Zoom closes account of U.S.-based Chinese activist after Tiananmen event

With Jitsi the main thing is the browser, you should try different versions of Firefox & Chrome.

solinenton Looking back at how Signal works

I didn't mean to imply Signal wasn't open-source, just that it was based on TextSecure and can be forked.

If the commercial entity fails or is held liable, we just need a distributed profile system, should be easy enough.

solinenton Looking back at how Signal works

> Signal Messenger, LLC, is a software organization that was founded by Moxie Marlinspike and Brian Acton in 2018

Did you google it? You're simply wrong.

It is open-source as it was, I didn't dispute that, but they are liable for their users if this bill passes, and they will easily go bankrupt. If there's no commercial entity, then liability falls to the developers most likely--whose identity can be obscured since they're developing Signal hopefully.

https://www.corporationwiki.com/p/31jiai/signal-messenger-ll...

They have a standard corporate structure--based in Delaware, registered as a Foreign entity in California.

solinenton Looking back at how Signal works

Signal started open-source, it was TextSecure, I'm sure there'll be an open-source alternative if the commercial entity fails, though I hope they do not.

solinenton Have you ever asked yourself “how did research get done before LateX?”

I disagree, I mean HTML would be even worse for text-flow if you rendered a PDF directly to HTML. The text-flow of HTML is not guaranteed, in the same way that the PDF text section is not guaranteed to be correct.

It's more about the social economics --but if you can find me a tool that supports the PDF export wrongly, I'll fix it myself.

> The market economics are simply not in the favour of people doing extra work,

Most of the tools are open-source. Does Word or any of the closed source tools do a bad job of this? That would be the main stopping point I guess.

It's still super trivial for them to implement, and shaming them will get their PR team evaluating the risk of not implementing such a feature. You just have to be successful at shaming them.

An HTML file also doesn't represent a document on paper, most people writing papers wouldn't switch to a electronic-only format at the moment--each page is carefully typeset and designed, HTML is not at this level even today.

solinenton Google Is Not God of the Web

Nothing that they actually made themselves after I (I'm not OP) have stopped using Google search, it's just YouTube now.

What has Google actually created that's successful, in house, other than search & gmail?

solinenton SpaceshipGenerator: A Blender script to procedurally generate 3D spaceships

Some more in Houdini: https://www.youtube.com/watch?v=cNPO_w9YRb0

solinenton Have you ever asked yourself “how did research get done before LateX?”

> But it simply doesn't work, because people can't be bothered to get the text right,

That's hardly "it's broken beyond compare." I think with accessibility formats which represent fully typeset written documents (as opposed to pre-typesetting, like Latex), it's always going to be an issue because they're mainly produced for people who are sighted initially.

We should simply start shaming programs which export to PDF wrongly, it's not really a difficult thing to get right.

solinenton Have you ever asked yourself “how did research get done before LateX?”

> PDF and TeX are broken beyond repair in terms of accessibility.

All that needs to be done is to have a accessible text section of the PDF--then Tex can just include the source of the document minus the Tex commands into the PDF, and PDF readers can have the screen reader work. The text could also just be stored per page, if the blind require to communicate where they found a particular item to a sighted person.

solinenton The Vanishing Of Ashlar: 4kb demo using GPU for sound and pathtracing

It's usually quite small--fractal geometry can be very succinctly described in a parametric way. The mandelbrot set can be described with just a small equation and a range of values to display.

They're not getting much from the GPU at all. Procedural modelling Houdini style is very common and you will find that algorithms to describe models/textures are often quite small, especially when compressed.

These days we do a lot of procedural generation of content, but then we don't make the user re-generate it because it does have a performance penalty--GPUs have plenty of RAM and tons of bandwidth, so it usually doesn't make sense.

source: Did some dev work on Houdini

solinenton No cookie consent walls, scrolling isn’t consent, says EU data protection body

These days I just inspect the page and if I can delete the overlay and find the overflow:hidden then I'm done! No accepting or scrolling required. It takes me 20 seconds to do this now typically.

solinenton Netflix now supports TLS 1.3

Good point. I think there's always a certain element of trust here--the law is more important than the implementation sometimes.

solinenton Netflix now supports TLS 1.3

We don't allow our mail to be seized and read without warrant, why should our other data be susceptible?

Netflix already had a secure system with their original business model of mailing you the content.

Finally, allowing advertisers to profile us based on our browsing data is one step before allowing the police to do so as well. Maybe it's the same step.

solinenton Ask HN: How do I overcome mental laziness?

It's interesting you mention sugar. There's also a biological component here--diet plays a huge role in how your brain works.

For example, on a ketogenic diet (no sugar and limited carbs) your brain produces ketones and many people report far less lethargy.

Sugar helps in the short-term, but if you get past the keto-flu, you no longer become glycogen dependent. Long-term ketogenic diets are now finally being studied, with many people reporting tremendous success.

Psychiatrists love giving you sugar and drugs though. Any drugs, any sugar!

solinenton Google Apple Contact Tracing (GACT): a wolf in sheep’s clothes?

> We can't trust Apple/Google to manage the anonymous data behind Contract Tracing but we can for our photos, phone calls, messages, web history, app data and GPS locations.

Just because our rights are already eroded doesn't mean we should give up the fight and disable our freedom.

Nobody said we should trust Google/Apple with any of that. This tracking of our anonymous data should only be allowed to be unencrypted by warrant. All data produced by the device stored on the cloud should be encrypted, unaccessable to Google and Apple.

Instead of blindly trusting corporations, we could implement regulations, allowing oversight and preventing unreasonable searches.

In any case, I use an Android phone and Google can't access my photos, calls, messages, web history, app data, or GPS locations. I've chosen a radio firmware which completely disables the GPS.

Looks like it's about time to get rid of Google completely from my Google phone.

solinenton Lidar: Peek into the Future with iPad Pro

Yup, flash lidar has some ways to go before it can become usable, in my opinion. I don't think there's much hope. I shouldn't say more!

Instead of using the time to determine the distance as typical lidar does, it uses the phase difference, and requires a precise phase measurement. This unfortunately is not nearly as accurate, especially with such a low-powered laser.

solinenton Swift: Google’s Bet on Differentiable Programming

Differentiable programming allows one to specify any parametrized function and allows one to use optimization to learn the objective.

There is definitely some need for an EDSL of some sort, but I think a general method is pretty useless. Being able to arbitrarily come up with automatic jacobians for a function isn't really language specific, and usually much better results are obtained using manually calculated jacobians. By starting from scratch you lose all the language theory poured into all the pre-existing languages.

I'm sure there'll be a nice haskell version that works in a much simpler manner. Here's a good start: https://github.com/hasktorch/hasktorch/blob/master/examples/...

I think it's pretty trivial to generalize and extend it beyond multilinear functions.

solinenton A new form of brain stimulation relieved severe depression in a small study

> Prevention and treatment are different though.

There are different, but they are not mutually exclusive.

> A patient who is already severely depressed and anhedonic and struggling to leave the bed is unlikely to suddenly start dieting an exercising.

I had the same symptoms, I found that exercising and dieting was the only thing that gave me pleasure (other than alcohol or cannabis). During the thick of it I would do pull-ups by walking to the bar, knocking some out (if I had a RA flare-up then I would do planks), and then collapsing after my heart-rate went back down. The issue was that people kept telling me to keep eating carbohydrate, which was causing lots of inflammation (I had tons of visceral fat). I also found that direct sun exposure improved my symptoms a ton (I was talking a D3 supplement prior).

Maybe if we told people "We have no idea how to cure this--the only proven method is diet & exercise, and this often lead to an early death if you don't start now" it would work? Once I realized that myself, it was a huge kick in the ass. I was figuratively paralyzed because no one seemed to give me this most basic advice, and the only way to get to a psychologist where I live is to self-harm.

I know a cancer survivor who was told that, and she did it, at 65, and she was obese by a large margin with severe trauma from the chemo.

solinenton A new form of brain stimulation relieved severe depression in a small study

> Alcohol and depression aren't a good mix.

Again my point, obviously I don't believe this to be true, but since it does have a positive success rate (there's no study which quantifies this, strangely enough, your studies are purely epidemiological, you'd have to take someone without ASD and then prescribe them ethanol at about 20% to actually study this) it would be negligent, as per your theory/ethics, to not prescribe this, and in fact it was prescribed in the past during prohibition [1]. Now, looking back this is obviously false.

> People are working on it. Meanwhile, we have make the best of what we know right now.

Great! The best of what we know right now is simply diet & exercise, no pill can prevent the onset of depression.

I think you'll find the taxonomy of depression to be a subset of existing disease categories. The brain, gut, and nervous system are connected in a chaotic manner.

PS. I don't drink at all, zero alcohol. I agree that it's probably very much related to all existing depression categories, correlated positively.

[1] https://www.smithsonianmag.com/history/during-prohibition-yo...

I can find a better source if you want, I'd have to do some digging.

Finally, hopefully you have already read this:

https://www.sciencedirect.com/science/article/abs/pii/S01650...

Anyways, thanks for the discussion, I don't expect to convince you. Maybe you'll think twice, however.

solinenton A new form of brain stimulation relieved severe depression in a small study

https://scholar.google.com/scholar?hl=en&as_sdt=0%2C5&q=+soc...

Also depression. It took me one minute to find three sources, could you please make some effort here? I have other things to do.

And you're right, it's not a causative factor necessarily, but I think there's pretty strong evidence at this point, it's hard to find a poor person who doesn't have some condition.

[1] https://www.sciencedirect.com/science/article/abs/pii/S00223...

[2] https://psycnet.apa.org/doiLanding?doi=10.1037/0021-843X.111...

[3] https://ajph.aphapublications.org/doi/full/10.2105/AJPH.93.1...