SuchAnonMuchWow

Don't you mean USB-A instead of USB-B ? I don't think I've ever seen a laptop with USB-B ports.

and the rendering engine is still based on chromium underneath ...

I think those cases are a minority, and in most cases you would have been fine with just slightly larger ints. It doesn't seems to cause any issue in python for example, because the cases where integers grow exponentially (which is the same as making their size grow linearly, so its still slow memory-wise) and without bounds are not all that common.

For example, an overflow bug was famous in java for being present in the standard library for a very long time: in a core function performing a binary search, the operation computing the middle of the two bounds was a simple average (max + min)/2, which overflow when max and min are big enough, even if the final result always fit in a standard integer.

In a lot of other cases, integer are used to address memory (or to represent array indexes, etc), so in most code they will naturally be restricted by the available amount of memory, even if intermediate computations can temporarily overflow.

> I'm pretty sure if we all suddenly went offline, half the internet would seem to keep going as usual.

Seems like a reasonable estimate if half the internet is porn.

... and Frenchs.

damn I was wrong, that's indeed the pictures I had in mind !

The feature is only supposed to be about pictures stored on iCloud, so the landlord comparison is more correct.

When Apple will start analyzing pictures on iphones and not in the cloud, then your substitution will become correct.

There are cases where pranksters painted a road on a wall, and drivers were fooled by it. So it definitely happens with humans too.

We are just too much used to the existence of optical illusions for humans that we fail to consider them a possibility for attacks. Adversarial attacks on neural networks are simply that: optical illusions, but designed for neural networks instead of humans.

No, that's inflation slowing down. But its still inflation, because new coins are being created.

And the pdf is probably not named "For-R-Tinfoilers.pdf"

The point is not that 0 must be easily identifiable, but also O: If you see a O without 0 as a comparison, can you identify it as such ?

In every monospaced font I know, there is a way to identify 0 (the dot or the slash), but not O which could be easily mistaken for a 0.

I found ubuntu mono to be one of the best in term of identifying those characters, with l and 1 being immediately identifiable, but O is only slightly larger than 0 so it can still be mistaken for a 0 if you are not careful (0 has a dot so it can not be mistaken with O).

I don't know of any monospace font where O can easily be identified as such and not mistaken with 0.

edit: see a sample here: https://fonts.google.com/specimen/Ubuntu+Mono?category=Monos...

Every character is easily identifiable by itself, except for O. I'm also curious if anyone has a monospace with better identifiably for O and 0.

Slight rant about a mistake I see every times this is mentioned:

> a user needs to easily differentiate between 1, I, |, and l along with 0 and O.

Most of the times, a font present those characters side by side, to show their differences and to illustrate the fact that they can be easily differentiated.

But what we need is not differentiate between them, but identify precisely every single one of them when presented without the other as a comparison point.

Almost every font presented in the article makes that mistake: for example, 0 is almost always the exact same shape and size compared to O, except for a dot or a slash across the 0. This allows you to recognize a 0 at first glance, but not a O. The same issue is also present in a lot of the presented fonts with l that can be mistaken for a 1.

Burn it at the site of extraction. That way, nobody will be able to use it !

Some of them are stretched examples, others comes from other languages/constraints (floating point, octal, ...), but some other are legitimately weird and error prone:

[1, 2, 3] + [4, 5, 6] // -> "1,2,34,5,6"

[,,,].length // -> 3

In theory, it is also broken.

It practice, it appears to be slightly harder to break than RSA for the same security level as we define it in non-quantum computing, but not by much.

I don't think we knows that for sure either.

If the universe is expanding indefinitely, then it means it has unlimited potential to store information, for example as the position of matter inside the universe.

For example imagine a spacecraft going away from earth, with a mirror toward the earth. On earth you can send a stream of bits with a laser toward it and it will come back. As it goes farther and farther away, you can increase the amount of bits in transit in the stream, making it a way to store a potentially infinite amount of bits.

Of course this is not a really good example, because I'm sure there is a lots of reasons why this scheme would not work indefinitely, such as chaos theory preventing us to target the spacecraft with the every-growing precision required.

But still, I would love to see an actual reasoning of wether or not the information we can store in the universe is actually bounded.

> Then again, no real computer is Turing complete because they all have finite memory so they're strictly speaking only big finite state machines

Interestingly, cloud computing made this limitation less obvious: one can imagine a program running on the cloud that bill the cloud provider when it runs out of disk space. And the cloud provider will buy more hard drives for its servers when needed. There is still obviously a limit somewhere, but its much farther ahead, because its more limited by factors such as how much hard drives we can build.

There is a fundamental difference between "We don't know how to solve this practically since the best solution we have would take too much time." and "It is a problem without solution. It can not be solved at all not matter how you approach it, or even how an imaginary alien specie with technology beyond our understanding would approach it".

Also, the equivalence between the different models of computation for Turing completeness (turing machine, lambda calculus, ...) isn't about complexity classes of the problems, only wether or not the model allow to solve the same set of problems. For example searching in a sorted array is O(log n) on a Von Neumann machine but O(n) on a Turing machine.

This specific game is a 2D pixelart rogue-like game, with simple UI, simple graphics, and no need for anything real-time. Vulkan is way overkill for that, and so is Unity.

The traveling salesman problem can absolutely be solved exactly. It just takes a lot of time.

Computability doesn't take into account how fast you can solve a problem: only if it is possible to solve it at all.

That wasn't sarcasm, and looking back at my comment I don't really understand what could be understood as sarcasm. What did you understood when reading it ?

edit: ok, I see how it could be seen as sarcasm.

I think we have fundamentally different opinions on what control you should be allowed to have on your privacy.

There is a similar debate about being photographed while in public: the fact that I'm walking in the street shouldn't allow anybody to take pictures of me, just because it has been decided that doing so would remove any "expectation of privacy".

That's my image. That's my house. If I don't want people to take picture of it and put in on the internet, it's my choice. And I don't think it is that unreasonable to want to control it.

I don't care about cityscape where the picture is taken from afar, where you can not discern any details of my house. But I don't want strangers to be able to see for example if I have dogs just by googling my address on internet.

Because the number of your house is there for strangers to identify your house.

Your SSID is there for you to connect to your private network. The fact that it is broadcasted is a technical limitation, and it is not intended for others beside you. Nobody besides technical people know that the SSID broadcasting can be disabled, and even less know how to use their wifi when it is disabled.

You realize that in Germany for example, most house are hidden in google street view ?

yeah, I obviously keep all my apps up-to-date with the most recent versions. The fact that it was no longer possible is precisely what made me change my phone.

I just bought a new android smartphone, and I have the current one since 2012 I never really needed to update the system: all the apps that I wanted just worked, and I didn't really needed new features from more recent versions of android.

It works still fine: performance is ok (maps is the only app that is a little laggy), the battery still last a full day.

But now the version of android on that phone is no longer supported by one of the app I use (banking app), so I'm forced to change it. I would have wanted it to last a few more years, but 9 years is still a long run.

I hope my new phone will last as long, and I choose it with that goal in mind (modern processor, 5G so I have it if I need, no samsung galaxy flagship that will be deprecated in a year, the possibility to install a custom OS instead of android if I need to, ...)

Not related to the content of the article, but am I the only one bothered with the way the article spell ranges: "six to 15 minutes", "Twelve of 18 men", "six of 24 females" ...

The two behaves differently, so you can not strictly compare the two. Most of the time stable_mean will give more accurate results, but this is not always the case. For example you can create test cases where the naive approach will return an exact result but stable_mean will have some rounding error.

Kahan summation however is always more precise than naive summation.

Rounding errors might be small, but catastrophic cancellation can make errors much larger than parts per billion. Maybe for your application it doesn't matter, but sometimes it does, and you need to be aware of these cases.

You wouldn't want those kind of rounding errors in financial applications, in rocket science applications, etc ...