Hacker Newsnew | past | comments | ask | show | jobs | submit | Rubu's commentslogin

Seems to be the wrong link, I end up at the newsletter subscription page. The article can be found here:

https://shecancode.io/blog/7-tricks-to-level-up-your-tech-ga...



Hmm OK, then it's odd that I couldn't find it, but I guess I'm overestimating by browsing-fu :P

Thanks!


And so was he. The first talking you hear is them telling the kids to stay put until they're sure no one is going to be rear ending them.

Does anyone know if these braking systems account for people driving behind you? And if a collision is inevitable, what do they "choose"?


I have no idea if it looks behind, or uses that information for anything. Ideally, it would time it such that when it was rearended, it was touching the car in front. It would significantly increase the mass and decrese the risk of whiplash.


Most Tesla's on the road don't have rear facing radar. I don't think the rear facing camera is used for anything other than parking. So, that leaves ultrasound, which has limited range.

I would think in theory they could use the camera alone to detect an approaching vehicle, and take some remedial action.


When the car is motionless (e.g. immediately after it has come to a stop from the emergency breaking), does the computer engage the breaks just soft enough so the car stays put?

And would this reduce the impact from behind if it were to happen?


I'm curious if you really want the brakes on full when you get hit from behind. It seems the cars would start to crumple in place until firmly together and you overcome friction, getting launched. With the brakes off, you would immediately accelerate, but perhaps with a lesser initial burst.

That's the decision I had to make one day. I saw a car coming to hit me from the rear. I took my foot off the brake and gripped the steering wheel, leaned back and relaxed. It worked out well, but I've always wondered if the brakes make any difference one way or the other.


I think you misread me. I was suggesting the computer don't put the break on full when stationary, but only enough to keep the car still (as the road may not be flat etc).


I was just using your comment as a starting point.


The opposite, you should release the brakes to reduce the effect of the impact so you actually acted correctly.


Thanks, the scientist part of me said this was correct. That just left a conflict between the experimentalist part of me and what my wife will allow.


Well I would hope it would choose getting rear ended, particularly because that is the fault of the driver behind you. If you choose to slam into someone instead of letting someone slam into you, you are now at fault.


I dont think the driver would like to have a whiplash...


Cool stuff. There's an open-source TCP stack available in Verilog on OpenCores[0], but that is actually C code compiled to Verilog using Chips[1]. Had no idea this kind of stuff existed, but then again, it has been a while since I've done anything in VHDL.

[0]: http://opencores.org/project,tcp_socket [1]: https://github.com/dawsonjon/Chips-2.0


Pretty sure that was just an example to demonstrate that skrub uses the -rf options by default.


Perhaps something on your end? LibreOffice (4.2.8.2) doesn't have a problem with the file on my system.


Interesting. This is LibreOffice 5.0.2.2.


Pretty sure that driving for Uber is a way to rake up more miles (training data) rather than actually test the system.



Yeah, I think so too. There's a list [0] over here with common mobile GPU specs which states the same (actually denotes T764 as an alternate name for T760 MP4). Not sure what his/her source is though so take it with a grain of salt.

[0] http://kyokojap.myweb.hinet.net/gpu_gflops/


I'm pretty skeptical as well. Can't say I know a lot about going rates for 0days like these, but something tells me 1M isn't even that much. If this "anonymous hacker collective" doesn't mind selling to the highest bidder, why not take it to the market themselves?


There are two big issues I see:

1) trust problems between the seller and the buyer. either the seller needs to trust that the buyer will deliver a working exploit or the seller will give the chance for the buyer to evaluate and the seller then needs to trust that the buyer won't steal the vulnerability.

2) issues of access. sellers don't know how to get in touch with buyers.

Also, some buyers may want to purchase an exploit library instead of negotiating for each sale and middleman can offer value here.

Also it is possible zerodium offered over the market value (or over what the usually pay) in this situation in order to generate publicity. Other companies that buy vulnerabilities have offered larger one off bounties in the past.


Politics? Could be a stolen technique - maybe it's easier to hack the hackers and steal their hacks to sell than develop your own.



the company that payd $1m for the exploit can make multiple sales of the same exploit to various governments and agencies. The hacker collective either dont have the connections or dont want to deal with the NSA, US Govt, GCHQ et al.

Would you want to end up on their list of interesting people?


Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: