Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
buildzr
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
19 ms
·
151.
▲
by
buildzr
7y ago
Quick, someone grep the standard library code, we can at least get some mild profanity and improve that rating!
152.
▲
by
buildzr
7y ago
I know people like this - what's the solution? Do you just cut them loose and make them figure it out?
153.
▲
by
buildzr
7y ago
> Especially since spreading rumors of a monster in the woods would be a great way to scare people away from your illegal drug operation Straight out of Scooby Doo, wonder if there are any real occurrences of this kind of thing.
154.
▲
by
buildzr
7y ago
In many ways this open source tooling is making them money. Developers tend to use what they know and are comfortable with, if you have to move off of C# say in order to write servers with better Linux compatibility or mobile apps you'
155.
▲
by
buildzr
7y ago
> Same code, name name, same culture. Counterclaims are pure PR. I don't think that's fair, there's major, objective moves in this direction: - .NET Core is fully open source and they're moving towards it being the pr
156.
▲
by
buildzr
7y ago
It's actually not too bad - you have to essentially line up the pixels on the camera's sensor with the pixels on the display, it's a challenge and might involve some careful choices of camera, display, focal length, distance,
157.
▲
by
buildzr
7y ago
If you keep something private, don't publish it or claim it's real, etc, what's the difference between making a deepfake of something and having a fantasy in your head?
158.
▲
by
buildzr
7y ago
I tend to agree with you that we're past this point and market economies offer bountiful opportunities to perform minor tweaks with major outcomes and improved efficiency, if you're more left leaning that may mean systems like bas
159.
▲
by
buildzr
7y ago
Okay, let's assume that's remotely possible. Injecting such a watermark at the camera doesn't help you - in fact, such a watermark you're describing is attempting to be resistant to things like re-transcoding so it might
160.
▲
by
buildzr
7y ago
What's the relevant application here? If you're suggesting this could be added by video editing software or display hardware, it seems if everyone knew about such watermarking it'd be trivial to strip out and turn into noise,
161.
▲
by
buildzr
7y ago
What about training an adversarial neural net to make faked videos pass that neural net? See the problem?
162.
▲
by
buildzr
7y ago
That's actually an excellent point, even if such technology worked perfectly it wouldn't solve simple problems like that. Seeing footage literally taped by a phone off a NVR display is pretty common...
163.
▲
by
buildzr
7y ago
That has nothing to do with the comment you responded to: Can you explain what you mean as the distinction between "content creators" and "some arbitrary video online" if the same technology is used to "authenticate
164.
▲
by
buildzr
7y ago
In what world is content creators needing to prove the footage they took is "real" a problem? Considering most content creators also do post editing I don't think that has any relationship at all to what's being discusse
165.
▲
by
buildzr
7y ago
Not quite as you can't necessarily clone the keys from an existing camera without some difficulty, however you can say, replace those sensors with your own inputs fed using a cheap microcontroller you wrote the firmware for. There'
166.
▲
by
buildzr
7y ago
I was under the impression that the whole point of such a system was you could trust any video from any (non-cracked) camera. If we're at the point of trusting videos only from select providers, well, that's not really much of an
167.
▲
by
buildzr
7y ago
We're talking about faked videos here though, often of things where the quality is already quite questionable. If it shows something that looks 480p good enough and gets the green checkmark it's game over for this concept of video
168.
▲
by
buildzr
7y ago
You could try but it'd still be possible to just fake it at the sensor level, essentially so the IC would see your video as the analog signal coming in and encode it as if it were any other image. It's basically the "analog h
169.
▲
by
buildzr
7y ago
It'd be too easy to fake, just replace the sensor in an approved camera with your own data feed. And that's assuming it wouldn't be somewhat trivial to dump keys from one or more cameras on the market to create a an easy tool
170.
▲
by
buildzr
7y ago
I mean, there's the old joke about programmers being machines that convert coffee into code. I think it's pretty safe.
171.
▲
by
buildzr
7y ago
Sounds like they're actually only banning delivery apps, even in Canada most (if not all?) of the delivery services being offered by Weedmaps for example are illegal. In most provinces you'd need to order from a provincially run s
172.
▲
by
buildzr
7y ago
Wonder what they mean by "facilitate"? Were certain mail order places offering apps they're banning? directly taking payments? Or are they seriously going after stuff like Weedmaps that just offers references to places that s
173.
▲
by
buildzr
7y ago
Is this actually going to be available anywhere but the US? Expensive, rare treatments like this often aren't paid out in socialized systems.
174.
▲
by
buildzr
7y ago
> In mobile, free-to-play model has solved most of the friction. Not really, this model is pretty awful as it creates motivations for developers to make the game repetitive or overly difficult in order to convince people to pay for stuff
175.
▲
by
buildzr
7y ago
It's interesting - for music this has basically been solved because the industry was under tight financial pressures and many companies including Spotify, Apple and Google were able to work out large scale licensing deals. Heck, most m
176.
▲
by
buildzr
7y ago
There's undeniably a freedom tradeoff here as facial recognition systems expand and many corporations and governments gain access to millions of capable sensors sending this data. I can't think of a single use case where doing thi
177.
▲
by
buildzr
7y ago
There's actually been a few vulnerabilities in the OpenSSH client when connecting to untrusted servers. These come to mind: https://www.cvedetails.com/cve/CVE-2019-6111/ https://www.cvedetails.com&
178.
▲
by
buildzr
7y ago
That's my bad, I was thinking of some personal scenarios in which I've had low value servers running single applications connected to the internet for years without updates at all.
179.
▲
by
buildzr
7y ago
> The thing about Zombieload/MDS (well, not really those, they're really more theoretical attacks... But Meltdown/Spectre in general, and any other local root exploit) is that they turn a remote shell, and perhaps a very l
180.
▲
by
buildzr
7y ago
Unless you have untrusted users with SSH you can get away with a lot. I've reviewed many major Linux patches for the past several years and found we weren't actually impacted by most of them. For example, I don't need Zombiel
More ›