SoylentNews is people
SoylentNews
from the not-what-we-expected-but-is-it-bad? dept.
http://arstechnica.com/information-technology/2015/06/is-the-internet-a-failed-utopia/
LONDON—At Shoreditch Town Hall on Thursday, at an event hosted by Intelligence Squared and Vanity Fair, the longevous British broadcaster Jeremy Paxman of University Challenge fame asked the audience of few hundred: "Is the Internet a failed utopia?" He asked us to vote on the matter by raising our hands. About two-thirds of the audience disagreed with the statement, a fair few (including myself) were undecided, and only a smattering of people actually thought the Internet was a failed utopia.
It was then the turn of four panellists, in the style of an electoral hustings or stump speech, to change our minds. In the failed-utopia camp were Andrew Keen and Frank Pasquale; in the not-a-failed-utopia faction were Peter Barron and Beth Noveck. They took it in turns to deliver quite rousing speeches.
The naysayers obviously had the harder job from the outset—we were at an event that was specifically tailored for fans of the Internet, after all—but they did a good job of reminding us that the Internet, as it stands, is not the elysium that we were all promised at its inception. Keen warned us that, while we think the Internet is an idyllic plateau where everyone is on an even footing, where two guys in a garage can compete with the monolithic, infrastructure-owning giants, we're all deluding ourselves: just like the real world, the Internet is now ruled by big corporations.
The utopian speakers, Barron and Noveck, mostly focused on all of the cool things that wouldn't have been possible before the Internet and World Wide Web were created. Noveck, who was a driving force behind President Obama's Open Government Initiative, reminded us that, with a smartphone in your pocket, you have access to more information than the president of the United States did 25 years ago. Barron, who is a public affairs bod at Google, spoke about the equality of opportunity on the Internet—and of course, about all the free services that we get to enjoy.
What does SN think?
Original Submission
(Score: 3, Insightful) by anubi on Tuesday June 16 2015, @05:26AM
Agreed. I do not think the net is the problem. It has the same characteristics as the post office. It delivers the mail. It does not verify the authenticity of it.
I believe about 50% of our problem is our software. We do not understand how it works. Hence we are unable to detect if it is malfunctioning - especially at the behest of a skillfully crafted piece of malware.
The other 50% of the problem is US. We run stuff without understanding it. We willy-nilly run stuff because we are told to - for the most trivial of "rewards". Would we wantonly sign binding legal contracts that easily? An executable to a computer is the same as a legally binding contract is to us. The laws of physics dictate that the code will be executed as written, regardless of any "good faith" the user may have. Yet, we have been loathe to insist the writer of software, especially paid-for software, is responsible for what it does.
My way of looking at it is a lot of "free" or "pirated" software is like a washing machine someone gave me or I found in a dumper. I plug it in and get electrocuted. Its hard to blame anyone. But if I know how washing machines work, I can verify the thing does what it is supposed to do and verify its safety. I can fix it if I need to. I will likely end up with a machine that lasts me a long time and provides years of service... but *I* have to know exactly how that washing machine works.
If I bought the machine from the store, and it electrocutes me, someone is gonna be held liable.
For this paradigm to work in the computer world, we need to have a simple, understandable, OS kernel that runs on understood hardware. No "proprietary" designs, No back doors. No "hidden stuff" on the public side of it.
Any proprietary programs can be as opaque as they like - however in exchange for copyright/patent protection, they should also be subject to responsibility for what it does.
We have enough trouble with this just trying to communicate with common English language. Ever critically listened to TV ads where the pitchman is carrying on, but using weasel words where lots of verbiage is passed, but nothing said to commit anything? I find those insurance commercials extremely vexing... you know the ones who implore you to buy their insurance for less than a cup of coffee a day (1)(2)... then "just answer three simple questions"(3)...
Salesmen mislead humans all the time. Now consider how much a similarly incentivized computer programmer can mislead a machine. There has to be some responsibility there for what code does. Now, if you are one of those types that download every app "they" tell you to, you are only setting yourself up for an array of upsets in your life.
Would you surrender your house keys to a complete stranger in exchange for a promise to clean your toilet? That is precisely what you do when you download an app into your phone or computer... you give that program "run of the house" and you no longer have control of what it can do. The OS does. And the computer programmer makes it his business to know the OS better than you do, so guess who is in the catbird seat once you accept the download and run it?
----------
(1) Starbucks coffee.
(2) Per Unit. Often takes a couple dozen "units" to provide decent coverage.
(3) If our lawyers can demonstrate any of those three questions were misanswered, we do not have to make good on our end - AND we get to keep all the premiums you paid us. You - in your hour of need - will not have the resources to defend your position. Remember - we have lawyers.
"Prove all things; hold fast that which is good." [KJV: I Thessalonians 5:21]
(Score: 3, Interesting) by linuxrocks123 on Tuesday June 16 2015, @06:08AM
People periodically voice this opinion, that programmers should be responsible for faults in their software. It's as stupid now as it always has been.
If you want reliable software, we know how to get it. Look at the way NASA and the airplane companies write software: that's how you get software that works 100% of the time. And then look at how much it costs to write that software: 100x or more times the cost of normal software.
If you hold programmers liable for mistakes, you kill the entire free software movement overnight. No matter how charitable I am, I am not going to release software when the act of doing so can get me sued if I made a mistake in my code. You also kill almost the entire software industry overnight: only NASA, Boeing, and Airbus (and maybe a few other companies) desire 100% reliable software; for everyone else, the cost is too high, and, by making software companies liable for programming errors, you've just raised the cost of non-mission-critical software to much more than anyone is willing to pay.
In your world, we wouldn't have Windows, Linux, or MacOS. We might have Inferno and similar systems, but it would cost $100,000 for a license. In your world, no one would have a computer, except maybe programmers, because the only software available for computers that doesn't cost >$100,000 for a license is the software programmers wrote themselves, for only themselves. In your world, we never would have ever had the Internet developed.
I'm glad we don't live in your world. So are you, but you may not realize it.
Now, regarding software reliability: if you want reliability insurance for software, you can buy such a contract. It costs an arm and a leg, for the obvious reason that you're buying insurance for an event that is likely to happen (software is likely to fail). You'll have maximum liability limits, because software is likely to fail.
And if you need to have space-shuttle-level reliability, well, you can get that, too. The cost is astronomical, pardon the pun. But that's because making reliable software is extremely expensive. Which is why no one does it. Which is my point.
(Score: 0) by Anonymous Coward on Tuesday June 16 2015, @07:14AM
The majority of software problems seems to come from two sources.
One is C class languages that are really aimed at bare metal programming, as you can refer to memory addresses directly with pointers.
Another is lazy languages like JS that don't type variables.
Thing is that there exist one language already that is built specifically to mitigate these issues, Ada.
But programmers don't like it because they consider it verbose and stiff to work with.
A more recent project is Mozilla's Rust. It is closer to C, but with safeguards placed around pointers and similar.
One may hope that Rust gets traction where Ada couldn't.
(Score: 1, Insightful) by Anonymous Coward on Tuesday June 16 2015, @03:44PM
The majority of software problems seems to come from two sources.
The real source: Bad programmers. In other words, a grand majority of programmers. I don't even know how so many of these people received degrees; they don't deserve them.
Then there are languages like Java that crappy programmers use a lot. They believe that because it has garbage collection, there can't be memory leaks. Obviously false, and that's something that Minecraft proved well; that thing was horribly coded, and not just for that reason.
Just switching languages won't solve a damn thing.
(Score: 3, Interesting) by Anonymous Coward on Tuesday June 16 2015, @07:29AM
They said that those who distribute *opaque* programs should be liable. If anything, this would make open source more prevalent, not less, because providing the source would free you from such liability.
I guess the logic is "if you are the only entity who is allowed to ensure the code works correctly (because only you have the source), you *must* ensure it works correctly".
(Score: 2, Interesting) by anubi on Tuesday June 16 2015, @08:17AM
Good reply. You brought up some very interesting points. My main beef is with this overcomplicated software with all sorts of obfuscation and/or digital locks so one cannot verify/debug/trace it - and no idea what it is *really* doing.
If I had some mod points, I would have modded you up for that post. I will revisit this thread tomorrow for that purpose. I learn a lot from good debates.
As far as the open source goes, just by the fact its open source, shared, and intended for general public under GNU in my mind should shield its creator. The code is open, modifiable, and one should be able to look at and verify ( and others more technically astute are free to do so as well ). Its hard to look a gift horse in the mouth, so to say. I even have some of my stuff released for public domain under GNU. Its wide open, I left it as an example of code that will do a certain thing, and hopefully I left enough comments in it that if someone needs something close, they are welcome to take what I left and modify it. Like you say, I do not want to take responsibility for it, but if someone finds it useful, just take it for what its worth ( free ) and have a ball with it. I see it as my way of payback for all the "free" stuff others have left for me.
And, yes, I do have some of that "Airplane" software. Its great stuff. Micrium's uC/OS-II. [micrium.com] They have written much documentation [micrium.com] on it. It is absolutely elegant code. This is the kind of stuff we *should* be using for a kernel as far as I am concerned. Its paid software; I see nothing wrong with that. Its the way Micrium is so open about exactly how it works that makes it so elegant. If something isn't right, I have a very good stab at finding out exactly what went wrong.
I was introduced to this code when I purchased a NetBurner board for some projects I was working on. This was code I felt I could trust.
Its when I see all these BSA types, putting digital locks on everything, having Congress pass all this DMCA crap, having me agree to EULA's and the like, trying to hold ME accountable. If they are gonna be that way, then I want accountability too. From them.
If the code is released as open-source, I feel I assume the risk of running it, as it is supposedly open and I take all risks to plug it in. However, if its a potion sold to me in a sealed bottle, and I have no idea what's in it, the salesman tells me it does one thing, I take it in and it does another thing. I feel I have been really taken for a patsy by not being allowed to know what's in it. I have been very frustrated at not even being able to open Microsoft DOC files without danger of infections. ( much to the delight of internet email phishers who try every trick in the book to get one to open their infected DOC's ). I thought we learned not to mix executables and data back in the old BBS days when we discovered we could place "ANSI bombs" in a text file. It was soon discovered that an ANSI-art "nudie" just might wipe your C: drive. I learned the lesson that it is quite obvious Microsoft professional programmers did not.
We may never get absolutely perfect code, but I would sure like to remove incentives to produce sloppy code by removing ways for producers of bad code to sweep it under the federally legislated rug (DMCA) to hide it from the customer. If they are gonna sweep crap under the rug, then THEY take responsibility for it. If its out in the open ( like the source code to the BIOS was printed in the early IBM-AT technical service manuals ), then its publicly viewable - not hidden, and therefore I would have a hard time holding anyone accountable for a programming error. This is not about paid software, copyright, or licensing issues... rather its about openness. I feel copyright is like a book - one does not encrypt a book so no-one can read it to protect copyright.
"Prove all things; hold fast that which is good." [KJV: I Thessalonians 5:21]
(Score: 2) by kaszz on Tuesday June 16 2015, @04:34PM
Don't get sucked into that tar pit. Only the lawyers will win in the long run. Much better to have legal protection revoked from products that are faulty or something else that makes it a less of a lawyer source of profit and more of a user benefit.
(Score: 2) by c0lo on Tuesday June 16 2015, @11:54AM
Not true. One can release software in source code with all the protection of free speech. The liability will remain with the ones that choose to execute that code.
Incidentally, this is why one can still release a code that implements a patented algorithm (oh, the horror) and be care free as long as it is source code: not only is free speech, but it will be a translation in another language of what is otherwise a public document (written in that weird language - patentese - which uses proper words ordered in sentences that maybe makes sense)
https://www.youtube.com/watch?v=aoFiw2jMy-0 https://soylentnews.org/~MichaelDavidCrawford
(Score: 0) by Anonymous Coward on Wednesday June 17 2015, @02:47AM
Its expensive because nobody does it. Market forces dictate that when demand increases, so will supply, which will drive down the cost, so if everybody starts demanding 99% reliable software, it will become cheap enough for everyone real quick. That's basic economics, I'm told.