theruran ππ΄
read this today and I am again reminded of the suckers who believe C is a decent language:
https://www.edn.com/toyotas-killer-firmware-bad-design-and-its-consequences/
herein described Toyota - who ought to be intimately familiar with safety - half-assed* their C software and ended up killing and mutilating people.
you bet your ass I am pissed when I read this stuff! and not ready to listen to apologists.
* more like, 6/(93+34) = 4.7%-assed MISRA-C
theruran ππ΄
@mupan and the same applies to medical devices. but corporate lawyers and lobbyists are still able to convince legislators and judges that releasing their source code hurts their business.
mupan π
@theruran I'd rather root my car than wait for authorities. But that's of course not safe either. We needed grass root NGOs that gain enough competences to test and analyse the cars really independently. A pre-condition would be open-source firmware though. So in a first step these organisations (like consumers' or patients' association, but for car users) would try and force open-source upon cars.
That's of course difficult. It took decades to force non-proprietary internet router standards, and obliged open-source is not in sight. And the telecom industry is powerful, but I guess the automotive industry is even worse. So, a challenge, but I think, not impossible.
theruran ππ΄
@mupan this litigation happened in 2013. and I still don't see regulators requiring software audits before automobile makers push Over-The-Air (!!!) updates to customer vehicles.
just recently there was a report of a Tesla vehicle failing to respond at 83 mph on the freeway. that guy got lucky.
mupan π
@theruran Will we learn for the future, at least? I wish autonomous cars would not be allowed before every aspect is tested. But I'm sure they will force them into the market (too) soon.
theruran ππ΄
@mupan I read somewhere that modern automobiles have over 300 computers in them! what a complexity nightmare. but not to worry, because you get to have Alexa in your infotainment system-
mupan π
@theruran True.
Thinking about cars twice, I guess we will not have personal cars in the future, so this would be fruitless. I love public transport. But in less populated regions this would be on-demand car sharing (of autonomous cars of course). As to the media articles (blogs?), the concepts for organizing this don't respect data minimization principles and force to use smartphones with Google or Apple accounts. I would be out.
The Doctor
@theruran Hi. I'm not an apologist. I worked at NASA (GSFC) when we were contracted to do the forensic analysis.
We didn't find anything. It was remarkably tight code.
What we did find was that the floor mats were loose and kept slipping forward over the accelerator pedal, forcing it downward.
I'm not saying C is great. It's not. This is not a good example demonstrating that C is not great.
theruran ππ΄
@drwho I'm not fully versed on the history, but one of the responses in this thread mentions that separate, earlier issue of the physical problem with the floor mats.
This article mentions and links the NASA/NHTSA investigation. This related article says more:
Neither NHTSA, with its absence of software expertise, nor the NASA Engineering and Safety Center β to which NHTSA turned to study the Toyota problem β were able to pinpoint a software cause for unintended acceleration. Nor were they able to rule out the possibility.
The NASA researchers, who were both on a deadline and not allowed to study Toyota's source code, simply ran out of time, noted Barr.
:flan_reaper:
I think it sounds somewhat rediculous that Toyota and not the NHTSA gets to dictate how the source code will be reviewed. Yet another example of how we don't really give agencies sufficient teeth to regulate the various companies.
If a car company can't submit 100% of the source running on a vehicle for public audit or place that code in the public domain, then their vehicles shouldn't be allowed on Federal highways. Simple as that.
The Doctor
@theruran We were allowed to study their source code. Only under very strict conditions.
They basically turned a conference room into a SCIF for three weeks. And supplied all of the hardware, because we couldn't use our NASA issued laptops.
But yes. Time ran out.
theruran ππ΄
@drwho this Barr Consultancy studied the code for 18 months :O
Santa Claes πΈπͺππ°π
Santa Claes πΈπͺππ°π
Chirp! is a social network. It runs on GNU social, version 2.0.1-beta0, available under the GNU Affero General Public License.
All Chirp! content and data are available under the Creative Commons Attribution 3.0 license.