To read all comments associated with this story, please click here.
Not unique to software, except maybe in degree. The same is true of stuff like car accessories - stick new wheels or a big spoiler on your car, and you're no longer dealing with a configuration the car manufacturer tested. If the car then crashes because the larger tyres rub on some part of the wheel arch, who's liable? Should the car manufacturer have tested larger wheels, because their customer base were likely to fit them?
Yeah, that's my thinking too - it introduces a whole heap of new rules and potential liabilities. It's not necessarily bad, but it's a big change, full of potential surprises.
People keep posing these questions as if they are some mystical unanswerable wisdom questions.
"Who's at fault? The plug-in -- or the browser?"
Whichever one is causing the problem. Thats what developers and debugging is for.
This problem exists already anyway. Your developing an app with .net and something throws an exception. Well good luck convincing microsoft it's them and not your software. Though 99.999% of the time is isnt the .net framework. Even if it is the .net framework, you do your best to make your product look good and work. You find a workaround.
"If the car then crashes because the larger tyres rub on some part of the wheel arch, who's liable?"
Quite obviously the owner of the car. The owner knows the size of the wheel arch and they put the big tires on anyway..
Edited 2011-03-08 13:31 UTC
Well the manufacturer cold specify minimum and maximum tested tire with and diameter in the manual, if you mount bigger/smaller tires than that you are sol.
Example of response from manufacturer when an accident is fount to be caused by tires that was outside the spec i manual. "We acknowledge that this accident was cased by the ire getting caught on the wheel arch, Subsequent investigation by the police and our representatives show that the ties fitted were outside the specification clearly stated in the manual, we are therefor not liable for any gingery or other damage caused by this accident"
PS: I'm not a loyer so this might nt hold in a trial
The day software companies are actually liable for software bugs, then we can start talking about that.
As of now, your point is moot.
The day software companies are actually liable for software bugs, then we can start talking about that.
As of now, your point is moot. "
Dude, we ARE talking about liability here. It's not moot.
Well, right now no-one is liable so how is that good for the consumer?
Neither is any corporate entity so I guess they cancel out, eh?
Well, right now no-one is liable so how is that good for the consumer? "
Competition -- not regulation -- makes software better.
Neither is any corporate entity so I guess they cancel out, eh? "
The market has certainly done a better job of delivering what consumers want than regulation.
Don't forget about compatibility with some people's digestive systems. And freshness definitely affects the compatibility with sandwich meat. And whether or not it's sliced affects the compatibility with various uses, and what kind of knives can be used. And ...
There really are a lot of compatibility issues with even "simple" things like bread, when you get right down to it.
You do understand that software is currently equated to art. Even the most braindead software developer can't believe that software is art.
Software is goods(for most part), not art.
And being classified as goods, does not imply new regulations for the software developer.
So far, I have enough trust in EU council to say it's intentions are good. And stuff in EU is already priced at a higher level, because consumer rights are wider.
As far as EULAs go, they aren't exactly cancelled, but could not be applied retrospectively or implied by the vendor.
And no one is implying that software vendors will be liable for crashes. What it does is imply, if it says that software will work - then it's a violation of your right as a consumer if it does not. Regardless of EULA.
But then again, this is draft version and will be amended 100 times.
Only in the deluded imaginations of so called software "engineers" is software more complex than physical goods. The very fact that 12 yo kids can write reasonable software shows this argument to be utter nonsense.
The average programmer is no more an "engineer" than a child making a Lego house.
Only in the deluded imaginations of so called software "engineers" is software more complex than physical goods. The very fact that 12 yo kids can write reasonable software shows this argument to be utter nonsense.
The average programmer is no more an "engineer" than a child making a Lego house. "
You both are on the extreme side of things. It depends. The answer of software's complexity is yes and no. Yes, it is complex, when you are in systems programming. Creating kernel drivers for your peripherals to whatever OS is a complex task, but not infinitely. Writing an operating system for space vehicles that _must_ be defect free is more complex than writing kernel drivers alone(being part of that task). Writing an Office Software comparable to MS Office's must be more complex than writing a piece of software to monitor your inventory stocks.
However, software is 100% defect free and very simple when you write it at the level of "Hello World" program.
Damn, this comment of yours really hit too close to home, eh?
I do agree though that software engineers continuously over-rate the complexity of software compared to other products.
But maybe they're right. Maybe creating farting iphone apps is the pinnacle of engineering efforts.
That's because anyone selling bread incompatible with human digestive system would find himself in jail.
Or shoes.
Actually you do. That's why you don't buy shoes on Internet.
Software components are infinitely more complex than your average physical good, and their interactions can cause unintended/unanticipated/unreasonable problems for the original software developer. Example: a plug-in crashes the browser. Who's at fault? The plug-in -- or the browser? It isn't clear how EU regulation would distinguish liability between the two components. On the one hand, the plug-in literally caused the crash. But that crash is manifested in the browser, and you could argue that it needs to defend itself against poor or malicious programming.
So? Car, plane, whatever else consists of various parts that have to work together. Any of those parts can break or interact badly with others. If part of jet engine fails and the engine catches fire are you going to argue that it is impossible to determine responsibility?
So, what does that mean for you, as a software developer? Answer: It isn't clear. It could make life a lot worse. You might have to deal with a lot more regulatory BS and barriers to distribution than before. Which might reduce your incentive to distribute software in the EU. Which might result in less choice for consumers, in the long run.
You could apply exactly the same reasoning to ANY industry.
Yet somehow only software vendors are absolved from any responsibility for what they sell.
I understand that regulation could also result in better software. But let's just understand that no regulation is completely benign. There will be costs for devs.
Not unique to software at all...
If you make modifications to your car, or drive it in an unsafe manner then it may crash or break down.
Car manufacturers are not held responsible for crashes as a result of poor maintenance, third party modifications or poor driving. Why would software suppliers be treated any different?
Because the interactions between software components are not easily inspectable, whereas real physical components are.





Member since:
2006-01-06
Software *is* unlike a physical good. For one thing, you can duplicate software for essentially zero cost. Second, you don't have to worry about compatibility with bread. Or shoes. Software components are infinitely more complex than your average physical good, and their interactions can cause unintended/unanticipated/unreasonable problems for the original software developer. Example: a plug-in crashes the browser. Who's at fault? The plug-in -- or the browser? It isn't clear how EU regulation would distinguish liability between the two components. On the one hand, the plug-in literally caused the crash. But that crash is manifested in the browser, and you could argue that it needs to defend itself against poor or malicious programming.
So, what does that mean for you, as a software developer? Answer: It isn't clear. It could make life a lot worse. You might have to deal with a lot more regulatory BS and barriers to distribution than before. Which might reduce your incentive to distribute software in the EU. Which might result in less choice for consumers, in the long run.
I understand that regulation could also result in better software. But let's just understand that no regulation is completely benign. There will be costs for devs.
Edited 2011-03-08 00:18 UTC