SoylentNews is people
SoylentNews
posted by
on Sunday February 12 2017, @08:28AM
from the can't-get-caught-if-you're-not-cheating dept.
from the can't-get-caught-if-you're-not-cheating dept.
A French businessman is suing Uber for 45 million euros, for destroying his marriage.
It seems that he installed the Uber app on his wife's phone, used it once, and then logged out. Later, when using the app on his own phone to arrange tête-à-têtes with his mistress, his wife received Uber notifications, and figured out what was going on. Uber attributes this to a bug in their software specifically related to an older version of iOS.
What do soylentils think generally about the liability of tech companies for bugs in their software? Some say liability is needed to force some responsibility; others say it would be the death of the software industry as we know it.
This discussion has been archived.
No new comments can be posted.
The Fine Print: The following comments are owned by whoever posted them. We are not responsible for them in any way.
belief, n:
Something you do not believe.
(Score: 4, Insightful) by dyingtolive on Sunday February 12 2017, @10:18AM
That sounds fine with me as long as you come up with the way to functionally prove any code will have an expected outcome, no matter the outcome or complexity.
Building a bridge is simpler than the shit management asks the dev team to do on a weekly basis. And I'm the QA team. I build shit code to test the shit code that parses the shit input. Want us to prove an infinite complexity of bullshit while still gracefully preserving content? Sure. Don't make next week the deadline and you might want to allocate some more resources.
Fuck, at least a bridge can't accept non-integer values of mass or be expected to interpret it in 20+ different ways individual developers felt like (only to be further interpreted by developers on down).
Don't blame me, I voted for moose wang!
(Score: 5, Funny) by PiMuNu on Sunday February 12 2017, @10:58AM
> at least a bridge can't accept non-integer values of mass
Erm, mass is a float...
(Score: 0) by Anonymous Coward on Sunday February 12 2017, @12:56PM
Another example to illustrate that programmers aren't real engineers.
(Score: 2) by BsAtHome on Sunday February 12 2017, @01:09PM
If the mass of the brigde is a float, then there is the possibility of a fractional atom's mass. That sounds very very wrong...
The approximation using a float may be "good enough" for most cases, but the fact remains that it is an approximation.
(Score: 2) by maxwell demon on Sunday February 12 2017, @03:47PM
A bridge consist of different types of atoms. An arbitrary atom's mass is not an integer multiple of a common base mass. The approximation using an integer may be "good enough" for most cases, but the fact remains that it is an approximation.
The Tao of math: The numbers you can count are not the real numbers.
(Score: 2) by dyingtolive on Sunday February 12 2017, @09:46PM
I meant to say "non-numerical".
At any rate, that's just why I shouldn't post at 4 am drunk.
Don't blame me, I voted for moose wang!
(Score: 1) by nitehawk214 on Monday February 13 2017, @05:42AM
Only on pontoon bridges.
"Don't you ever miss the days when you used to be nostalgic?" -Loiosh
(Score: 5, Interesting) by anubi on Sunday February 12 2017, @11:03AM
The analogy I see comparing bridge engineering to software engineering is at least the bridge engineers can verify the specifications of their materials... that is, if one is shipped a load of faulty rebar that fails to have the tensile strength claimed, this can be tested and demonstrated.
Same if somebody mixes a load of out-of-spec concrete. Or, in my case, bad parts.
I can test them. My test equipment does not change every time somebody designs a round of new semiconductors.
I have tried to use the tools you software guys are forced to use. I *hate* them. I have no idea whatsoever what is going on under the cover. I got lost back leaving pure old HTML 4 and Borland C++.
You guys aren't given near enough time to understand what you are doing. Before you even begin to understand the how to actually use your tools, they are yanked out from under you, and given a new set of tools, which are claimed to make you "more productive", but are little more than terribly insecure eye candy generators that allow you to quickly code pretty junk. While the people producing your tools hide behind "hold harmless" clauses.
No wonder you guys live in a world full of stuff that you don't know what its really doing. Its too much like taking medicine... one has to trust someone else that the stuff does what its claimed to do. Yet every day goes by, seems like not only things getting more complex, more and more law being passed to keep us all ignorant of the inner workings of said tools. Its as if I had to use my tools yet be ignorant of what was inside them. Now, how the hell am I going to be proficient at my tools if I don't know how they work? I believe you Linux guys are the last widespread group of knowledgeable professionals on the planet... the rest of us are becoming nothing than replace-the-box junkies who have no idea what's in the box. Like some modern auto mechanics that can run up several thousand dollars of car repair bill, whereas one who knew how the thing worked would go right to the bad connection that tripped off all the OBD codes.
Believe me, I could not build a bridge that would hold together either if my tools were as flakey as those you guys have to use.
I could not with a straight face apply to do the things you guys are asked to do. I am too much of a perfectionist.
You guys are forced to work in a world where the pretty secretary who can't be trusted to do anything trumps the old hag that is as trustworthy as the day is long.
Working in a world so driven by appearance simply drives me nuts.
"Prove all things; hold fast that which is good." [KJV: I Thessalonians 5:21]
(Score: 4, Insightful) by Arik on Sunday February 12 2017, @01:54PM
If you want computers to be reliable and trustworthy then you don't want anything to do with what amounts to the animating paradigm for software development in this century. And that's a very sad thing to me. But obviously not everyone shares my values.
If laughter is the best medicine, who are the best doctors?
(Score: 0) by Anonymous Coward on Sunday February 12 2017, @11:42PM
Computers are already much more reliable than your average human. Computers basically do what humans can do without computers only faster and more reliably.
For instance take excel. You technically don't need a computer to do everything that excel can do. You can do it all with pen and paper and present it to others without any calculators and using no technology whatsoever. But it will be a whole lot slower, you will use a whole lot more space to store the same amount of information (thousands of sheets of paper vs a single hard drive), and the computations and copies will have a whole lot more mistakes.
The reason computers are useful is partly because they are a whole lot more reliable than their human counterparts. They can make perfect copies of files quickly and effortlessly (compared to the old days where to copy a manuscript by hand yielded variations and imperfections. Use a copy machine, sure, that's using technology and probably incorporates computer chips right? I suppose you can use a printing press and make impressions but then making changes is a hassle for one time prints and is that really more reliable than a computer?).
Just like with self driving cars. Are they ever going to be perfect? No, but they don't have to be, just much better than their human counterparts.
When saying computers and software is unreliable you need something to compare it with. What else does what a computer can do and how reliable is it? Well, humans we can communicate via speaking but how reliable is hearsay in making perfect copies of information vs copying the original text/audio/video and spreading it around (hint, if you ever played the kindergarten game called telephone where one person whispers something in the next person's ear and it travels around the class and by the time it reaches its destination it's a completely different message then you'll know. Our reliability is horrible).
A bridge doesn't need to copy, process/compute, and distribute information. Humans do. and when it comes to that computers are much faster and more reliable than their human counterparts. If we held software up to ridiculous standards no one would make software and we would be stuck without software and hence we would be much worse off in terms of reliability since we would be forced to do everything ourselves. and even your primitive computer can do certain things a thousand times more reliably and quickly than your most qualified and most highly licensed and educated human. For instance if I see an incorrect number in excel I assume it was inputted incorrectly, I don't assume that excel is what made the mistake. You would make the same assumption because you know that most mistakes are made by humans and not software/hardware/computers.
and there is already software that can diagnose patients better than your top doctors. Computers play chess better than we do. They will continue to progress and become better and better than us at everything. Will they ever be 100 percent reliable? No, but the base standard is are they more reliable than us given the same task. If so then we should use them or else we are stuck doing things the old unreliable way by doing it ourselves.
(Score: 4, Insightful) by Arik on Monday February 13 2017, @12:29AM
"When saying computers and software is unreliable you need something to compare it with."
Yeah, you do. I'm comparing them to other programs. I've been using computers since 1980 feel free to do the math there. I'm comparing the software that's pushed today to the software that was used in past times, and I'm talking about the trajectory of development. You mention Excel. That's a specific example of a very common sort of program, a spreadsheet. There have been plenty of spreadsheets and they all do basically the same thing. If you were working rationally to produce quality software, and you're talking about a type of software that's been made for decades, you'd expect for there to be a fairly mature spreadsheet available and widely used by now.
Well that's not what happened and I've watched as these things transpire and it's very interesting what wound up happening instead. These programs are not maintained long term, perfected, and allowed to mature. Mature software is the last thing a commercial software house that makes its money selling software wants, because once everyone has bought it sales dry up! So what they do instead is avoid maturity, through constant rewrites and also through constantly adding new features but only rarely fixing old bugs. And as a last resort they simply discontinue software that has gotten too close to being a mature product and replace it with some new written from scratch POS.
So there's no mature, reliable spreadsheet that's been studied and vetted thoroughly over the years, no spreadsheet you can count on to do exactly what it's supposed to do, and no more, each time it's run. Instead there's the big one that's riddled with bugs, that's constantly receiving updates to add new features but only occasionally has a few of the worst bugs fixed, and the also-rans that spend most of their time copying that.
Hope that helped.
If laughter is the best medicine, who are the best doctors?
(Score: 1, Insightful) by Anonymous Coward on Monday February 13 2017, @03:53AM
This has little to do with licensing or the definition of the word 'engineer' and more to do with the fact that companies want to keep charging you for something they developed years ago. The issue here is the mentality that companies get from copy'right', they want to keep profiting off of something they made a very long time ago.
The problem is that they discontinue an older more mature product in favor of creating something newer that they can keep charging you for. If the old product is discontinued it will eventually stop working on newer computers or in newer systems, since the license is probably tied to a specific computer or system, and we can no longer make copies of it for others to use or for you to transfer the license over to others. So eventually the product dies. The solution is two fold.
Abolish copy'right' on discontinued products.
Reduce copy'right' laws to something more reasonable
This allows people to port old and mature products to newer systems and to keep making copies of them and improve upon them so that they can become a stable standard. This forces companies that want to make newer products to actually improve those products, to improve both the reliability and functionality by adding features. They can't simply force everyone to upgrade by discontinuing an old product because doing so automatically puts said product in the public domain.
But as long as psychopathic corporations keep writing byzantine self serving laws we're going to continue having these problems.
(Score: 0) by Anonymous Coward on Monday February 13 2017, @02:19PM
In fact you can make a very strong argument that much of the early tech matured quickly exactly because it wasn't subject to things like patents at the time. Things won't mature as easily any more because IP gets in the way.
In fact much of the engineering stuff is mature exactly because it's no longer subject to IP or was never subject to IP. Look at anything that is subject to IP (ie: pharmaceuticals) and see how long it takes anything to mature. Usually maybe after the IP expires. Math is very mature because it hasn't traditionally been subject to patents. This lets people build upon old ideas and improve them without worrying about getting sued.
But of course businesses don't want that, they want to keep profiting off of something that someone else made decades ago. and as long as we keep on letting them write the laws ...