3 ms·

GPL software in cars doesn't kill people (2016)

33 points by tzhenghao 9y ago

reificator 9y ago
I had no doubt that it does. But this article has so little substance that maybe I should reevaluate my opinion.
- reificator 9y ago
  For context, the original title was something like "Security Through Obscurity Kills People (2016)"
runningmike 9y ago
Good statement of the very respected Bradley M. Kuhn from the Software Freedom Conservancy (https://sfconservancy.org https://sfconservancy.org). But my current view nowadays is that the automotive industry is more and more open. Foss unless it seems for new developments in this sector. Of course the battle is definitely not over, due to fad discussions on security and open source. Nice one on this subject is http://security-and-privacy-reference-architecture.readthedocs.io/en/latest/ http://security-and-privacy-reference-architecture.readthedo...
jfv 9y ago
Proprietary software vendors market their software as being more secure through FUD in general. This isn't specific to the car industry. What's sad is that this article uses exactly the same rhetorical tactics. The failures of Therac-25 had nothing to do with open vs. closed codebases. It had to do with quality control at the level of the engineering team. > At least one person has already been killed in a crash while using a proprietary software auto-control system. This is just a silly argument. If there were an adequate sample size of open auto-control systems you could make a case. > Meanwhile, there has been not a single example yet about use of GPLv3 software that has harmed anyone. Has there really not been an exploit used on an GPL webserver that hasn't been used to "harm" someone in some way? Not much software has the impact to directly physically harm someone in the same way a car does. But user information and data? There's so much more wrong with this article that it would take an even longer article to explain it. This is literally the stupidest thing I've read in years.
- reificator 9y ago
  > There's so much more wrong with this article that it would take an even longer article to explain it. Yep, I started writing up a rebuttal and realized I'd be here all day. It's incredibly frustrating to see articles supporting your opinion, but in a fallacious manner that's easy to rebut.
- eeZah7Ux 9y ago
  > > At least one person has already been killed in a crash while using a proprietary software auto-control system. > This is just a silly argument. If there were an adequate sample size of open auto-control systems you could make a case. You are taking the sentence without context, making it look like a silly argument. Kuhn was juxtaposing it to unfunded claims that open source is known to be more dangerous than closed source. Later on the discussion takes a rational approach: "until you can prove that proprietary software assures safety in a way that FLOSS cannot..."

33 points by tzhenghao 9y ago

reificator 9y ago

I had no doubt that it does. But this article has so little substance that maybe I should reevaluate my opinion.

reificator 9y ago
For context, the original title was something like "Security Through Obscurity Kills People (2016)"

runningmike 9y ago

Good statement of the very respected Bradley M. Kuhn from the Software Freedom Conservancy (https://sfconservancy.org https://sfconservancy.org). But my current view nowadays is that the automotive industry is more and more open. Foss unless it seems for new developments in this sector. Of course the battle is definitely not over, due to fad discussions on security and open source. Nice one on this subject is http://security-and-privacy-reference-architecture.readthedocs.io/en/latest/ http://security-and-privacy-reference-architecture.readthedo...

jfv 9y ago

Proprietary software vendors market their software as being more secure through FUD in general. This isn't specific to the car industry. What's sad is that this article uses exactly the same rhetorical tactics. The failures of Therac-25 had nothing to do with open vs. closed codebases. It had to do with quality control at the level of the engineering team. > At least one person has already been killed in a crash while using a proprietary software auto-control system. This is just a silly argument. If there were an adequate sample size of open auto-control systems you could make a case. > Meanwhile, there has been not a single example yet about use of GPLv3 software that has harmed anyone. Has there really not been an exploit used on an GPL webserver that hasn't been used to "harm" someone in some way? Not much software has the impact to directly physically harm someone in the same way a car does. But user information and data? There's so much more wrong with this article that it would take an even longer article to explain it. This is literally the stupidest thing I've read in years.

reificator 9y ago
> There's so much more wrong with this article that it would take an even longer article to explain it. Yep, I started writing up a rebuttal and realized I'd be here all day. It's incredibly frustrating to see articles supporting your opinion, but in a fallacious manner that's easy to rebut.
eeZah7Ux 9y ago
> > At least one person has already been killed in a crash while using a proprietary software auto-control system. > This is just a silly argument. If there were an adequate sample size of open auto-control systems you could make a case. You are taking the sentence without context, making it look like a silly argument. Kuhn was juxtaposing it to unfunded claims that open source is known to be more dangerous than closed source. Later on the discussion takes a rational approach: "until you can prove that proprietary software assures safety in a way that FLOSS cannot..."