Living with the surveillance state
Living with the surveillance state
Posted Oct 29, 2013 14:54 UTC (Tue) by ms (subscriber, #41272)Parent article: Living with the surveillance state
Huh? I use open source. Facebook uses open source. How does this have any impact on the NSA hacking Facebook and spying on me?
Posted Oct 29, 2013 16:06 UTC (Tue)
by drag (guest, #31333)
[Link] (68 responses)
Closed source software like Skype, Adobe Flash, Facebook Android clients, and things of that nature are actively being used to spy on individuals.
And obviously online services use open source software, but they are not open source themselves. Using your own hosted services and P2P protocols that eliminate middle men services are also going to help a lot.
Posted Oct 29, 2013 16:17 UTC (Tue)
by ms (subscriber, #41272)
[Link] (67 responses)
Equally, if facebook open sourced all its software today, I would bet it wouldn't make a blind bit of difference to the NSA.
The hacks that are always going to be available are hacks that rely on a combination of software, a combination of configurations and a combination of events that will always escape capture by mere "more eyes". Absolutely, we can talk about mitigation and stopping random unsponsored hackers making off with troves of data and supplying a black market. But realistically, the complexity of modern software (coupled with the utter lack of interest by 99.99999% (underestimate) of all programmers in proving correctness of their code) means for a well sponsored state-backed agency, there will *always* be ways in.
That really, is the core of it all.
Posted Oct 29, 2013 16:38 UTC (Tue)
by drag (guest, #31333)
[Link] (53 responses)
The NSA doesn't depend on hacks. It depends on blackmailing corporations and using legal threats to get what it wants. Along with that they work to undermine the use of secure encryption technologies whenever possible.
It _CAN_ and do use hacks and such, but that's not how they operate primarily. That's expensive and only going to be used with specific targets, I believe. NSA, and the equivalents are cooperating with your governments to undermine your security and safety. You don't need to rely on hacks when you have the support of the military, police, and the politicians.
Also security does not have to depend on everybody getting everything right 100% all of the time. Security is done by good design, layers, good protocols, good encryption, and good sense. You cannot eliminate the threat, but you can massively reduce it. You can make it so that it's difficult, if not impossible, to simply stick a pipe into a ISP or internet backbone and suck up all the email and online activities of users in one fell swoop.
> If you care about the type of interactions that facebook provides then there is no point being part of some open-source non-facebook because the absence of other users you care about means you can't take part in the interactions you want - the platform holds no value.
Facebook displaced MySpace. MySpace displaced LiveJournal. And newer online services are displacing Facebook.
Facebook isn't going to be around forever and, believe it or not, a great majority of the population has no desire to use Facebook at all and would happily jump ship if something better came along. It's not the be all and end all.
And, frankly, all these 'social media' services are built on a house of cards. Their modus operandi specifically revolves around gathering as much information on users as possible , packaging it, and selling demographic groups for the purposes of online advertising. Once the businesses that depend on these advertisements start realizing that the 'views', 'likes', 'clicks' and the rest of the metrics are all a complete fabrication then I expect to see a massive contraction in the industry.
So in case that opportunity arises then the Open Source/Free software community has to be there with already established and mature ways to not only put yourself out there to be found, but to be able to find and communicate with other people in a distributed and P2P fashion. The trick is that not only does it need to be open source, it needs to be better, easy to use, and gives uses the ability to not only decide, but directly control how much control they should have over their information.
Posted Oct 29, 2013 16:57 UTC (Tue)
by hummassa (subscriber, #307)
[Link] (10 responses)
Care to elaborate on this? Calling all the metrics "a complete fabrication" is kind of incompatible (IMHO) with "businesses that depend on these advertisements"... or I didn't parse it right.
Posted Oct 29, 2013 18:43 UTC (Tue)
by mathstuf (subscriber, #69389)
[Link]
Posted Nov 1, 2013 4:26 UTC (Fri)
by drag (guest, #31333)
[Link] (8 responses)
Hrm.
'businesses that depend on these advertisements'. I mean like toilet paper companies, vacuum cleaner salesmen, car companies, movie producers, and other people that purchase ads online and provide the money that 'social' websites need to thrive off of. They depend on advertisements to sell their products. They give money to advertising agencies that then buy space on popular websites.
That's the money that pays for all the bandwidth, servers, etc that companies like facebook use to attract the demographics that the advertisers want.
One thing to always keep in mind with these companies is that the primary business of companies like Google or Facebook or Twitter or whatever isn't the online services they provide you. Their primary business is selling you, the user and every bit of personal info they can get their hands on, to the advertisers. Bundling you up and creating packages that the advertisers can pick and choose from.
I used to work for a company that did this sorts of stuff successfully pre-internet. They depended on mortgage companies selling your personal data. State governments selling your personal data. Drivers license info, credit card spending habits, and all that stuff for tracking people and carving them up into demographics and worked with the Post Office to make sure that they had accurate information on people living at various addresses. Used it for junk mail.
Now that information combined with your online habits and email history they can paint a much more complete picture of you and figure out how to bundle you with other people and sell you.
It seems likely to me that there is a widespread and epidemic practice of generating false metrics in order to drive up prices for advertisers. Not just by people like Google or whatever, although they are part of it, but all the people that get kick-backs from google. Youtube users, people advertising crap on facebook, people trying to drive traffic to their blogs, etc etc. It goes top to bottom. Ranging from small time BS, to organized crime and botnets.
Once the advertising agencies, or the companies that spend the money on the advertising agencies, figure out how to accurately gauge the effect of those advertisements on the buying habits of the public then I figure there will be a significant constriction in the online service industries.
Especially if at around the same time we enter into a new stage of 'recession' in the economy. As long as people have big budgets then sometimes the main problem is just figuring out how to spend it. However that can change if corporations start having to penny pinch.
Posted Nov 1, 2013 7:54 UTC (Fri)
by klbrun (subscriber, #45083)
[Link] (1 responses)
Posted Nov 1, 2013 9:50 UTC (Fri)
by khim (subscriber, #9252)
[Link]
Of course. Was there any doubt? When Google just started effectiveness of ads on it's search pages was off-the-charts. It was ten or maybe hundred times more effective then TV ads (per dollar spent). Of course such thing brought marketing guys in droves, ads filled less and less relevant pages and effectiveness of ads went down. Guess what exactly limits said process? Right: other forms of advertisement. Internet spending grows till it starts wasting more or less the same percentage as other mediums. This, again, shows how wrong drag is: short-term cheaters win, but medium-term mediums with better metrics win (and long-term we are all dead which makes this case not very interesting).
Posted Nov 1, 2013 9:43 UTC (Fri)
by khim (subscriber, #9252)
[Link] (5 responses)
Google is not part of it. Not because they are all that “altruistic” or “fair”, but because all such shenanigans can only ever provide temporary boost and Google does not need temporary boost: it makes more then enough money short-term and it's goal is to convince advertisers to continue to spend money on them long-term. That means that when Google discover some large cheats it usually cracks on them and “miss the expectations” that quarter. Small cheaters can get away with their schemes for awhile, alas. LOL. Nope. The effect will be the exact opposite. You think Google business is big? TV ads business dwarfs it by a huge margin. It's budgets are slowly moving to the Internet because it already easier to gauge the effect of the ads on the Internet. If someone will find even better way to more accurately measure effects of the ads on the Internet then Internet will get bigger slice of the advertisement fee. Wrong again. We are not in the 'recession', we are in the first stages of extremely large depression (thing Great Depression… only bigger). All the corporations are hurting because buyers are just not there (and buyers are not there because they don't have money). What does it mean? If you'll start to “penny pinch” then you'll go under even faster. Which will probably mean that medium-term ads will become even more important. Long-term, yes, situation will be different (if all your competitors go bankrupt and you are left alone then you don't really need more ads, right?), but this stage is many years removed from today.
Posted Nov 1, 2013 22:02 UTC (Fri)
by nix (subscriber, #2304)
[Link] (4 responses)
Posted Nov 2, 2013 0:50 UTC (Sat)
by khim (subscriber, #9252)
[Link] (3 responses)
Do you need theory or evidence? Evidence is there if you look for it, situation with theory is much harder because last century was spent in building nice mathematical models which explained how you can achieve infinite growth on a finite planet. They apparently don't work, but we have no new ones just yet. As for evidence… it's there if you know where to look. These are all smokes and mirrors. They are supposed to be “early indicators” for the future employment rates, but they no longer work that way. If you'll take a look on the the actual situation with the labor force then there are no improvement. Official explanation of difference between this rosy picture and the sad reality which non-easily-falsifiable metrics gives us is “oh, that's all about baby boomers, you know they are retiring and there are fewer young workers”, but if you'll visit the appropriate site you'll find out that number of workers above 65 was 64.54 million five years ago, 78.78 million year ago and 81.97 year today. IOW: these pesky baby boomers are not retiring, instead they work till they drop! What goes down instead are workers between 35 and 44 years. This basically means that this actually-not-so-rosy picture is completely artificial: government just writes off millions of people (they apparently like to live on subsidies). This four years after the end of recession, remember? Europe? Don't make me laugh. The only country which is in good shape is Germany and they don't have resources to bail everyone else out.
Posted Nov 3, 2013 11:37 UTC (Sun)
by kleptog (subscriber, #1183)
[Link] (2 responses)
However, I think your point is more to the long term. The thing is, our ability to produce things is indeed limited by a finite planet, but most of the economy (80%) is in services, not goods and there no particular limit to the number of services that can be provided. I can see production of goods stabilising (if it hasn't happened already).
That's not to say there aren't challenges. Fossil fuels will run out and we need to replace them with some other energy source and drastically improve efficiency. But I'm a glass half full kinda guy and there are signs of movement. Our economy is 20 years will look radically different, but hey it looked radically different 20 years ago too.
That said, I'm not entirely sure about the US. They have a serious problem at the political level and it's not clear they look far enough ahead to make the necessary adjustments for a smooth transition.
Posted Nov 3, 2013 14:51 UTC (Sun)
by dlang (guest, #313)
[Link] (1 responses)
the government has surprisingly little influence on business in the US, especially on the direction of what businesses do.
Posted Nov 29, 2013 9:13 UTC (Fri)
by jospoortvliet (guest, #33164)
[Link]
Posted Oct 30, 2013 10:51 UTC (Wed)
by NAR (subscriber, #1313)
[Link] (41 responses)
I guess this implies lots of data upload which doesn't work that well with current asymmetric wired connections and would "needlessly" drain the battery of mobile devices. I mean for this to work, you have to also convince people to pay for keeping and serving other people's private data (which might be child porn for all we know) in exchange for having their private data stored by somebody else. Additionally, for a centralized community network to succeed, it "only" needs to have enough "friends" registered. For a distributed community network, not only friends are required, but enough well-connected (in this case, technically) friends. That's an additional hurdle to clear.
An other problem is that if we accept that some surveillance is reasonable, the government will want to have a backdoor and then we're back to square one - what if the three letter agencies abuse the backdoor?
In my opinion, this is a social problem, not a technical, so a social solution is needed, not a technical.
Posted Oct 30, 2013 11:04 UTC (Wed)
by ms (subscriber, #41272)
[Link] (33 responses)
And indeed, no doubt some surveillance is reasonable. Which means that ultimately we're back to requiring laws to stop people from doing things which they technically can do. Which I find very amusing as it's essentially the same sorts of laws as DMCA and DRM. The only difference is that here we want such laws to be passed in order to protect citizens rather than protect "rights holders". Which explains everything about the order in which such laws were passed...
Posted Oct 30, 2013 11:40 UTC (Wed)
by HIGHGuY (subscriber, #62277)
[Link] (32 responses)
Usually, a technical solution is superior to any social solution.
Also, technical solutions tend to be easier to solve than social solutions.
Posted Oct 30, 2013 15:14 UTC (Wed)
by raven667 (subscriber, #5198)
[Link] (31 responses)
Woah, strongly disagree. Technical solutions are by nature inflexible forcing people to circumvent them when their needs are outside the scope of the solution and you can't enumerate and prevent every kind of badness in the world, the effort of trying to do so is madness and leads to worse outcomes than the problems you are trying to prevent.
A strong audit capability, performed out in the open, is what works, and is what concepts like the warrant provide.
Also any proposal which begins with some variation of "If everyone would just ..." is dooooomed.
Posted Oct 30, 2013 17:06 UTC (Wed)
by PaXTeam (guest, #24616)
[Link] (28 responses)
> Woah, strongly disagree.
do you carry a key chain and lock doors? if you don't then please post your home and office addresses along with where you park your car. you should not have a problem with this since you must have a social solution to this problem already ;).
Posted Oct 30, 2013 18:20 UTC (Wed)
by mathstuf (subscriber, #69389)
[Link] (5 responses)
Posted Oct 30, 2013 22:14 UTC (Wed)
by PaXTeam (guest, #24616)
[Link] (4 responses)
Posted Oct 30, 2013 22:24 UTC (Wed)
by mathstuf (subscriber, #69389)
[Link] (3 responses)
And in the general case, your statement is wrong since a subset of all numbers (uncountably infinite) can be countably infinite (integers) or finite (integers uniquely representable by a single Arabic digit).
Posted Oct 30, 2013 22:41 UTC (Wed)
by PaXTeam (guest, #24616)
[Link] (2 responses)
Posted Oct 31, 2013 1:21 UTC (Thu)
by mathstuf (subscriber, #69389)
[Link] (1 responses)
> state that a subset doesn't have the properties of the set
Did you mean to talk about *members* of the sets in question here?
What I was originally replying to is that ¬∀x.p(x) is not the same as ¬∃x.p(x). This is the conclusion you seem to have made given your reply here:
> > > Usually, a technical solution is superior to any social solution.
Posted Nov 1, 2013 22:35 UTC (Fri)
by PaXTeam (guest, #24616)
[Link]
> Did you mean to talk about *members* of the sets in question here?
yes i was being sloppy but thought it would be clear from the context, sorry if that made you misunderstand me. as for what i pointed out, it's really not hard: if you disagree with the elements of a set, you also disagree with the elements of any subsets of the set, unlike what you stated.
Posted Oct 30, 2013 19:08 UTC (Wed)
by nix (subscriber, #2304)
[Link] (11 responses)
No, what generally keeps everyone from getting robbed blind and society from collapsing is that in any system of this nature *cheating is rare* and there are systems in place to detect and punish cheaters to keep their numbers down: most of those systems are not technical but social and procedural. Among other things, just breaking a window is high-risk because there might well be someone inside who could hear you and send an alarm to a social cheater-deterrent system, to wit, the police. (Here I presume a police force consisting of thinking human beings, not a militarized horror like that in many parts of the US, which might well be considered by now a purely technical system without the ability to respond in a graduated or reasonable fashion!)
Of course, this doesn't mean that posting your home and office addresses and car location in response to a request to do so is sane: there is a low percentage of cheaters in any society, and one moderate-risk way of detecting potential targets might be to simply ask for relevant information while concealing your own identity. But just because a few cheaters exist, and that technical defences against those cheaters also exist, does not mean that the technical defences are the *primary* defences. Heck, on my street most of us have our front doors open most of the time during the summer days, sometimes even when nobody's home. Number of robberies: zero, despite the total absence of any technical measures against theft. We trust our neighbours to note any strange unshaven men leaving our houses bearing bags of swag, and any potential burglars realise this and don't try wandering in and nicking stuff. We happen to all know each other well enough that free-rider problems don't arise.
(I'm sure you've read Bruce Schneier's _Liars and Outliers_, in which he talks about all this at great length and much more clearly than you ever could. Perhaps you disagree with him?)
Posted Oct 30, 2013 19:08 UTC (Wed)
by nix (subscriber, #2304)
[Link]
Posted Oct 30, 2013 20:50 UTC (Wed)
by khim (subscriber, #9252)
[Link]
Bingo. IT world lived under different rules for so long it forgot how people interact with a real world. Think one recent hoopla. What happens if real world “security professional” (someone who tests keylocks for living) will pick a code of some Mom&Pop store (or, even worse, General Motor's HQ), visit it and make a copy of a couple of confidential documents? Just where exactly he'll be if he's not affiliate of said company? Sure, people do pick locks on safes and crack other systems regularly for different reasons—read Feynman's book, or Wozniak's one, but they absolutely do expect to see repercussions if caught. The fact that computer “security professionals” expect to see easy acceptance for such an acts is baffling to me: sure, if you want to study security precautions of some firm or a website then you need need to negotiate it in some form. It should not be advertised widely among the compnay employees or site visitors, but some people “at the top” must know about your efforts. If you go and crack different sites willy-nilly to collect information for your Phd.D. and you are caught… well, your Ph.D. will be postponed for couple of years, I guess. The whole “technical problem” vs “social problem” is false dichotomy: few problems are purely social and few problems are purely technical. All the security measures in the world can not protect you if some government feels you house must be cracked… either NSA or MSS will crack it. And it'll not matter much how many locks and how complex you've attached to your door. But if something is perceived as totally socially unacceptable then some rare individuals will still try to do that and to repeal them you need things like keylocks. Why computers should be any different? It's the same story.
Posted Oct 30, 2013 22:33 UTC (Wed)
by PaXTeam (guest, #24616)
[Link] (8 responses)
yet you failed to post a single address. i think that fact alone speaks for itself (and against everything you said ;) quite well.
as for Schneier, i have over 2k rss feeds, his isn't among them. that you should tell you something.
Posted Oct 30, 2013 23:13 UTC (Wed)
by khim (subscriber, #9252)
[Link] (6 responses)
Well, it says something, all right. It shows that people trust their anonymity (which is form of their social protection) more than they trust their locks (which is form of their technical protection). In what kind of world this information can be used as some sort of confirmation for your crazy position I just don't know.
Posted Oct 30, 2013 23:37 UTC (Wed)
by PaXTeam (guest, #24616)
[Link]
as for the topic itself, if one doesn't value technical measures and believes in the power of some 'strong audit capability, performed out in the open' (i trust you did read the post i replied to, didn't you?) then surely disclosing addresses protected by those pointless technical measures should be fine? also not disclosing addresses is not anonimity, it's fear of getting owned (broken into) despite all those so effective social measures.
Posted Nov 1, 2013 21:52 UTC (Fri)
by nix (subscriber, #2304)
[Link] (4 responses)
I am not a moron and will not compromise my safety to prove something to an anonymous blowhard like PaXTeam. (I note that PaXTeam is trying to get me to post my address when his name and indeed number remains opaque. Hypocrite.)
Posted Nov 1, 2013 22:46 UTC (Fri)
by PaXTeam (guest, #24616)
[Link] (3 responses)
Posted Nov 1, 2013 23:11 UTC (Fri)
by nix (subscriber, #2304)
[Link] (2 responses)
Posted Nov 2, 2013 8:05 UTC (Sat)
by HIGHGuY (subscriber, #62277)
[Link] (1 responses)
Well, maybe this statement missed some necessary nuances to make it acceptable for most of you.
The first would presumably be that any technical solution must be backed by a supportive social "contract". If really everybody is fine with the NSA spying on them, then you should not instate cyptography that makes it hard(er).
The second would be that ultimately the social solution (when followed by everyone) and the technical solution have the same effect.
This statement actually has its roots on the workfloor. When you worked out a procedure that people should follow to prevent breaking things for everyone then applying technical measures to guide/force them into that procedure is better than relying on education only.
My opinion is that the same thoughts can apply to society as well, in some cases.
In this last case you could say that this would mean that the cryptography in use should be strong enough to withstand mass cracking, but weak enough to allow case-by-case cracking. Which is a hard problem too, of course.
Posted Nov 12, 2013 21:29 UTC (Tue)
by filteredperception (guest, #5692)
[Link]
I was going to respond "not so hard, just traditional spying with picked locks and video or other bug capturing keys as and when they are used by the user". But that works onlysomuch when you have mathematically unbreakable crypto available, which is not a 100% for all time assumption one can make. So you are right, it is a hard problem. Because the first thought that comes to mind is that powers-that-be can (and I suspect do) try to solve it by making the methods of breaking the crypto a kind of orwellian 'unknowledge', that they will establish as such by truly any means necessary.
It's a jungle out there kids...
Posted Nov 1, 2013 21:49 UTC (Fri)
by nix (subscriber, #2304)
[Link]
Posted Oct 30, 2013 19:31 UTC (Wed)
by raven667 (subscriber, #5198)
[Link] (9 responses)
In any event the fanciness of your lock isn't what is keeping people out, it's the risk of social consequences which prevent bad actors from taking action much of the time. Having the ability to investigate incidents and increase the risk of consequences provides a ton of disincentive for bad actors.
There will still be incidents, you can't prevent that.
Posted Oct 30, 2013 22:23 UTC (Wed)
by PaXTeam (guest, #24616)
[Link] (8 responses)
and i'm still waiting for those addresses, actions speak more than words do, you know... no addresses = you believe in technical measures, simple as that.
as for what is an absolute technical measure, try to pick your own locks. i bet you can't. along with 99.9% (seems to be the random going measure here) of humanity. that makes locks an 'absolute' measure for 99.9% of humanity (including every single poster here ;). i wish we had anything close to that in other areas of life, computers or not.
Posted Oct 30, 2013 23:10 UTC (Wed)
by khim (subscriber, #9252)
[Link] (2 responses)
They have much better effect. The number one protection against burglar is privacy. If burglar knows where someone lives and knows that someone does not use two turns of key to lock the door every time (or, even better, if s/he knows that someone does not lock keys at all), well… this information is incredibly valuable for a burglar. THIS is why people don't publish it on websites. What does it change? You don't need to pick a lock. To pick a lock is akin to high-level rootkit which is totally stealthy and invisible. If you just want to take something from the apartment then you only need to have a strong scredriver: insert it into a lock hole and turn it with excessive force. All done. Often you can use just a flat screwdriver to move bolt. I think 99.9% (seems to be the random going measure here) of humanity can do that. Wow. Just wow. What kind of logic is that? Let me repeat once more: in a world with reliable locks (where technical measures dominate) this information will be absolutely worthless. Lock can not be picked up anyway, so why not publish it's location? In our world where lock is just a side-show and social aspect is the primary one… of course one will not give up their primary form of protection so easily! FWIW I've seen plenty of people who don't use large bolts on their doors and lock them only with a small latch. IOW: a lot of people are ready to neglect “technical measure of protection”. I've seen very few guys who post notes about their absence on a public website along with the address of apartment. On the contrary: a lot of guys arrange for the with neighbors pick of mail, periodic checking, etc to make sure it's not easy to notice that apartment is temporarily abandoned. IOW: they spent a lot of efforts on their “social measure of protection”. What does it say about relative merits of two approaches?
Posted Oct 30, 2013 23:52 UTC (Wed)
by PaXTeam (guest, #24616)
[Link]
and i thought you just said it was anonimity. make up your mind 'cos the two are different things. and never mind that it's also false as you clearly explain (and contradict yourself) in the rest of your sentence, good job ;).
as for picking a lock and whatnot, you clearly have zero experience with real life locks (and rootkits and other buzzwords, these things have about nothing in common) so maybe stay away from the topic, pretty please? ;)
as for the logic... it's really simple. if you state that you don't believe in technical measures yet you rely on them (=afraid of disclosing where exactly you do) then that's a clear case of hypocrisy, simple as that. my point is that the world isn't black and white where one or another measure dominates everything else, rather it's a careful balance that one has to adapt to his own circumstances (in different parts of the world you'll get away with a different mix of social/technical/etc measures).
Posted Nov 1, 2013 21:54 UTC (Fri)
by nix (subscriber, #2304)
[Link]
Posted Oct 31, 2013 4:49 UTC (Thu)
by raven667 (subscriber, #5198)
[Link] (2 responses)
I don't see people commonly going around testing doors, and when there are home invasions I don't see basic door locks being a factor.
> and i'm still waiting for those addresses, actions speak more than words do, you know... no addresses = you believe in technical measures, simple as that.
That's ridiculous, but whatever, I guess I'm too dumb to back down, whois raven667.org
> as for what is an absolute technical measure, try to pick your own locks. i bet you can't. along with 99.9% (seems to be the random going measure here) of humanity. that makes locks an 'absolute' measure for 99.9% of humanity (including every single poster here ;). i wish we had anything close to that in other areas of life, computers or not.
I don't see how that is relevant since 99.9% of people aren't commonly trying to break into my house. The risk can be increased if there are more people willing to transgress, if they are desperate for example, and if there is a failure of investigation and remediation, police don't come to your neighborhood for example, but that just makes my point that the strength of societies norms comes from the consequences of violating them, not from technical and authority systems which could prevent you from violating them if you desired to.
Posted Nov 1, 2013 22:56 UTC (Fri)
by PaXTeam (guest, #24616)
[Link] (1 responses)
Posted Nov 2, 2013 20:05 UTC (Sat)
by raven667 (subscriber, #5198)
[Link]
Thanks man, I love you too. 8-)
> why did you post a pointer to some data
Because I know that information is out there if you have two brain cells to rub together to find it, you can also find out where I work, how much I am paid and what my house is worth among other things. I know that I'm not truly anonymous when I speak online unless I have gone to significant effort to create an anonymous identity separate from my "normal" identity which I have not done.
I think the root of the disagreement is in the perception of risk. You seem to believe that my risk of a home invasion, or something bad happening to me, has been materially changed in some way and I disagree with that assessment. I also don't think you are actually going to jump on a plane and steal my toaster, or that our local drug addled poor are just waiting to read the lwn.net comment section to figure out which houses to rob. You could of course try and pull some juvenile prank which might change my risk assessment slightly but that would also say more about you than me and I am presuming that you are an adult.
A risk assessment which includes means, impact, and most importantly likelihood is useful for everyday living and as humans we are naturally bad at it. All risks seem highly likely and greatly harmful when they are not.
Posted Oct 31, 2013 5:34 UTC (Thu)
by Cyberax (✭ supporter ✭, #52523)
[Link] (1 responses)
Let's see if your locks are going to help you.
Posted Nov 1, 2013 22:48 UTC (Fri)
by PaXTeam (guest, #24616)
[Link]
why would i want to contradict myself?
Posted Oct 31, 2013 18:22 UTC (Thu)
by HIGHGuY (subscriber, #62277)
[Link] (1 responses)
You are right that not all problems lend themselves well to being merely technological problems (with no social impact) and that sometimes what you're protecting against has legitimate use-cases (DRM, anyone?).
The point is that by making something illegal or socially unacceptable doesn't stop it from happening. That's why a technical solution is usually superior than a merely social solution.
Posted Oct 31, 2013 22:19 UTC (Thu)
by hummassa (subscriber, #307)
[Link]
A technical "solution" does not prevent it from happening or make it impossible, either. At most the technical measure would make it harder, but the NSA has infinitely more resources than the spied entity, and an infinite number of attack vectors to obtain the sought information.
> sometimes what you're protecting against has legitimate use-cases (DRM, anyone?).
DRM, as I have reiterated many times, is neither a legitimate cryptography application (because it seriously hinders protected-by-law Fair Use) NOR a technically or mathematically sound cryptography application (because B and E are the same person.
> The point is that by making something illegal or socially unacceptable doesn't stop it from happening. That's why a technical solution is usually superior than a merely social solution.
That's where IMNSHO you have it backwards: there is never a perfect technical solution, and that's why you MUST have a social solution if you want to have any chance of making the "something" happen less.
An analogy: we will NEVER have zero murders. Currently, there is no technical protection against being murdered, but even in a Dune-like future where you can't be murdered by projectile weapons, people will murder each other with knives and poisons, or just putting each other in the pool and removing the ladder. Now, if murder is socially acceptable, there is no reason NOT to murder the people in front of me in traffic. So we make murder socially unacceptable with the objective that we have less murders.
Posted Oct 30, 2013 15:04 UTC (Wed)
by raven667 (subscriber, #5198)
[Link]
I think we should accept that some is reasonable but only with heavy public oversight, not in secret, that is what the whole concept of warrants is trying to achieve. Any kind of government intervention or surveillance should be done in the open as a matter of public record so that we can independently scrutinize it's justification and methods.
I don't think that should require devices or services to have a backdoor, there is no requirement to make it easy or convenient to perform surveillance, I think it should be exactly the opposite. I would prefer data retention rules to prevent service providers from storing un-redacted logs and encourage them to design systems where they don't have the capability to access private keys and decrypt customer data. Safety mechanisms which protect against insider attack or data breaches should also protect against lawful surveillance.
You can still search a persons stuff with a warrant, you can still follow them around with a microphone to see who they communicate with, without jimmying all the worlds technology with backdoors.
Posted Oct 30, 2013 16:04 UTC (Wed)
by niner (subscriber, #26151)
[Link] (2 responses)
That leaves the social part which is also the hard part. Even with a perfectly distributed system like email, people like going to the big names.
Posted Oct 30, 2013 21:14 UTC (Wed)
by NAR (subscriber, #1313)
[Link] (1 responses)
The fact that it's trivial to add an USB hard drive to a wireless router does not mean that people would be willing to put up with its extra costs. For example my mother turns of her router when she turns off her laptop in order to save on the electricity bills.
Posted Oct 30, 2013 21:40 UTC (Wed)
by nybble41 (subscriber, #55106)
[Link]
Not all P2P systems have the narrow focus of Bittorrent. In Bittorrent, peers sharing a particular file do not interact with peers sharing other files, which is why it's hard to torrent rare stuff. To expand this to handle data for something like Facebook, you would have to make sure peers have an incentive to hold and distribute data which they aren't directly interested in on behalf of other users (with the expectation that other users will do the same for them). That's closer to the FreeNet model, though as far as I know FreeNet lacks an incentive system similar to Bittorrent's tit-for-tat prioritization.
Perhaps Bitcoin could be leveraged to provide a more stable and universal form of incentive for participation, with the clients acting as autonomous agents. Users would keep data for each other because someone will be willing to pay to access it. I'm not sure even Bitcoin's fees (about two cents per transaction at the moment) are low enough to make that scalable, though.
Posted Nov 1, 2013 4:42 UTC (Fri)
by drag (guest, #31333)
[Link] (2 responses)
The data is already being uploaded, no?
Also it does not need to be a all or nothing situation. If you don't care about controlling your information then you can use whatever service. Right now all the social media stuff is 'walled gardens'. If you do things P2P and open protocols then anybody can provide any service they like and users can use whatever software they like.
Also the amount of data that people like Facebook save in their 'big data' clusters is a hell of a lot more than people actually care about or want shared. If all you want shared is your posts or pictures or links to this or that then that really isn't a whole lot.
> For a distributed community network, not only friends are required, but enough well-connected (in this case, technically) friends. That's an additional hurdle to clear.
Yes. This is the big problem.
Needs to be something like Email, that is very distributed, but have a built in way to make sure the communication is always coming from the same person/persons.
The actual identity of the person controlling the account can be confirmed or discovered through side channels if that really matters to you and the person you are communicating with. Just have to make sure that the messages are unadulterated and whatnot.
> An other problem is that if we accept that some surveillance is reasonable, the government will want to have a backdoor and then we're back to square one - what if the three letter agencies abuse the backdoor?
Screw them. I don't think that surveillance is reasonable, but I do think it's unavoidable. As long as governments continue to give these bozos money they will continue to use it to undermine our security. But that's their problem. So let them figure out how to do their job. They don't need our or anybody else's help.
Posted Nov 1, 2013 4:48 UTC (Fri)
by drag (guest, #31333)
[Link] (1 responses)
Email is all about 'push'. You don't know what you are getting until you get it.
If instead it's subscription services then you won't have all the same problems with spam and whatnot. A user will actually need to subscribe to companies or services in order to get information from them... that is have their server actively subscribe and pull the data from them. I figure this will go a long way to cut down on the shenanigans and be more in line with the way web services work.
maybe a more elaborate system based on something like:
https://bitmessage.org/wiki/Main_Page
I donno.
Posted Nov 1, 2013 16:50 UTC (Fri)
by mathstuf (subscriber, #69389)
[Link]
[1]http://arstechnica.com/business/2013/10/silent-circle-and...
Posted Oct 29, 2013 17:44 UTC (Tue)
by khim (subscriber, #9252)
[Link] (1 responses)
Do you imply that not even 100 such programmers exist in a whole world? I kind of understand you point, but you are way too pessimistic. 99.9% or may be 99.99% is probably more realistic estimate.
Posted Oct 29, 2013 17:54 UTC (Tue)
by ms (subscriber, #41272)
[Link]
I started with more 9s and then stopped and had a think. I suppose 1 in 1e7 is too low though finding how many programmers there are in the world is tricky. 1bn seems too high but 10 million seems too low but I've no idea why I think that. Maybe somewhere around 100million? I'd be surprised if there are 10000 programmers who regularly think about proving properties of their code. Which would give 99.99%. But I am pulling all these numbers out of my posterior, it's true.
Maybe I'm just naturally pessimistic ;)
Posted Oct 29, 2013 21:00 UTC (Tue)
by k8to (guest, #15413)
[Link] (9 responses)
Yea, there are a lot of things where you can formally show that your code conforms to some formal description, but I know of no way of showing that the formal description is correct provably.
In practice, I am a person who makes small corrections to a large codebase. When I make such changes I try to imagine all the possible scenarios, and attack my change from all of those angles in both localized ans systemic fashion. If I find a problem, I either find a solution or give up on my change.
If I was starting a project from the beginning, I can definitely come up with a significant number of ways to do a better job than this, more likely to be correct and more demonstratively so, but in the constraints, I'm not sure how to do better than this.
I guess I'm part of the problem, but I don't believe there's another way from the place I'm given.
Posted Oct 30, 2013 10:44 UTC (Wed)
by ms (subscriber, #41272)
[Link] (4 responses)
That's a good point I often fail to remember - at the end of the day, you're just trying to prove some form of equivalence between one model of your program and another.
That said, you can establish some useful properties, such as "it will never infinite loop", "it will never gobble all your RAM", "it will do 'the right thing' on all possible valid inputs". I'm not sure how far things have got in terms of the security context.
Posted Oct 30, 2013 18:40 UTC (Wed)
by smoogen (subscriber, #97)
[Link]
Most security proofs I have seen always start out with "Assume that the environment is not hostile and that the user is not malicious." which is about 0% of the time when actually trying to defend against anything.
Posted Oct 30, 2013 18:51 UTC (Wed)
by nix (subscriber, #2304)
[Link] (1 responses)
Posted Oct 30, 2013 19:06 UTC (Wed)
by ms (subscriber, #41272)
[Link]
> You certainly can't establish any of those in the general case, and even doing it in special cases requires some severely stereotyped code. (e.g. in the latter property, the inputs need a grammar which permits only inputs up to some maximum length, probably a very short one. That rules out most useful programs. The problems with proving the other two in general cases are hopefully obvious to anyone reading this!)
"All possible valid inputs" is largely covered by type checking. It is possible to do static array bounds checking in many cases, though yes obviously not all.
Termination checkers are fairly well advanced, e.g. see Byron Cook's work, and arguably you can almost always happily work with a total language. Very rarely do you actually *need* unbounded recursion (though that's not to say it's easy to cope with total languages).
As ever with static analysis it's about how much you care about the "proofs" it offers, versus how much expression you're willing to give up.
Posted Nov 8, 2013 0:55 UTC (Fri)
by Wol (subscriber, #4433)
[Link]
And this is a perfect example of the trap I rail at quite often - THE MODEL IS NOT REALITY.
Just because it is proven that your software (a mathematical construct) is mathematically correct doesn't mean that it will actually work. I'll just quote two luminaries:
Knuth: "Beware of bugs in the above code; I have only proved it correct, not tried it."
Einstein: "As far as the laws of mathematics refer to reality, they are not certain, as far as they are certain, they do not refer to reality."
Cheers,
Posted Nov 4, 2013 11:31 UTC (Mon)
by bakterie (guest, #37541)
[Link] (3 responses)
Typically the formal description lacks a lot of implementation details, and is on a more conceptual level. You strive for a formal description that is "obviously" correct to a human, and then prove the equivalence between the implementation and the specification.
You are correct in that you still don't know if the specification is provably correct (for some definition of correctness), but at least you have reduced the problem from convincing someone that the implementation (with all its gory details) is correct, to convincing a human being that a much simpler specification is correct.
Posted Nov 4, 2013 23:04 UTC (Mon)
by Jandar (subscriber, #85683)
[Link] (2 responses)
Unfortunately '"obviously" correct' is in no way the same as really correct. '"obviously" correct' is akin to a secure implementation of cryptography: not yet broken.
Posted Nov 5, 2013 8:27 UTC (Tue)
by k8to (guest, #15413)
[Link] (1 responses)
Of course we also put a lot of time and effort trying to make our code coverity-clean, and that seems to be also a fairly narrow band of improvement. I guess I suspect that for hundreds of thousands of lines of code that trying to make a provably correct model isn't likely to be worth the effort.
Posted Dec 8, 2013 17:43 UTC (Sun)
by Jandar (subscriber, #85683)
[Link]
Posted Oct 30, 2013 10:35 UTC (Wed)
by eru (subscriber, #2753)
[Link]
I think Facebook is a bit poor example in this case, because its point is to seduce people into publishing data, and it makes its living from mining said data. From the privacy point of view, it is broken by design. (I use it a bit nowadays, but I never put anything there I would not mind having shouted from rooftops).
Email and other person-to-person channels are another matter and more important for privacy. It might be easier to implement changes there, judging by the way various mobile chat systems have started to supplant SMS, even though SMS is still the only system every user's handset is compatible with.
Posted Oct 31, 2013 1:10 UTC (Thu)
by fuhchee (guest, #40059)
[Link]
Living with the surveillance state
Living with the surveillance state
Living with the surveillance state
Living with the surveillance state
Living with the surveillance state
Living with the surveillance state
Living with the surveillance state
Living with the surveillance state
Living with the surveillance state
It seems likely to me that there is a widespread and epidemic practice of generating false metrics in order to drive up prices for advertisers. Not just by people like Google or whatever, although they are part of it, but all the people that get kick-backs from google.
Once the advertising agencies, or the companies that spend the money on the advertising agencies, figure out how to accurately gauge the effect of those advertisements on the buying habits of the public then I figure there will be a significant constriction in the online service industries.
Especially if at around the same time we enter into a new stage of 'recession' in the economy. As long as people have big budgets then sometimes the main problem is just figuring out how to spend it. However that can change if corporations start having to penny pinch.
Living with the surveillance state
we are in the first stages of extremely large depression (thing Great Depression… only bigger)
Do you have any evidence for this peculiar statement? I've never heard it anywhere else outside the sort of website that tells you to turn all your money into gold and beat it into gold-lined tinfoil hats to keep the chemtrails off. The US in particular is barely in recession at all any more, and many metrics (housebuilding starts, household debt ratios, etc) are rapidly improving. Even Europe is out of crisis, though hardly in ideal state yet.
Living with the surveillance state
Do you have any evidence for this peculiar statement?
The US in particular is barely in recession at all any more, and many metrics (housebuilding starts, household debt ratios, etc) are rapidly improving.
Even Europe is out of crisis, though hardly in ideal state yet.
Living with the surveillance state
Living with the surveillance state
Living with the surveillance state
a distributed and P2P fashion
Living with the surveillance state
Living with the surveillance state
Living with the surveillance state
There's always going to be someone crossing the line. The only way to stop that is by preventing it in the first place.
Unfortunately for this kind of problem, there aren't many technical solutions that do not carry a social impact as well, as the migrate-away-from-facebook example shows.
Living with the surveillance state
Living with the surveillance state
Living with the surveillance state
Living with the surveillance state
Living with the surveillance state
Living with the surveillance state
Living with the surveillance state
> > Woah, strongly disagree.
> do you carry a key chain and lock doors? if you don't then please post your home and office addresses along with where you park your car. you should not have a problem with this since you must have a social solution to this problem already ;).
Living with the surveillance state
Living with the surveillance state
Living with the surveillance state
Living with the surveillance state
Even there, a social solution (that a reasonable man does not burgle others' houses, and reports burglars seen burgling others' houses, and that when called to a burglary in progress the police bother to turn up) does 99.9% of the work.
Living with the surveillance state
> during the summer days, sometimes even when nobody's home. Number of
> robberies: zero, despite the total absence of any technical measures
> against theft.
Living with the surveillance state
yet you failed to post a single address. i think that fact alone speaks for itself (and against everything you said ;) quite well.
Living with the surveillance state
Living with the surveillance state
so much nasty ad hominem, i'm hurt! more seriously, why don't you get familiar with the dictionary and look up what a hypocrite is. then quote me back where you think i said something that makes me one ;). asking for your address while not publishing mine isn't it: i stated already that i do *not* believe in black&white measures (only this or only that), but in a mixture of them, so keeping information secret is perfectly fine for me, as is using locks. but if someone believes that technical measures are superflous because he lives in such a nice neighbourhood, go ahead and prove it. you have yet to back up your statement with actual action. IOW, you're just trolling as usual.
Living with the surveillance state
Living with the surveillance state
i stated already that i do *not* believe in black&white measures (only this or only that), but in a mixture of them, so keeping information secret is perfectly fine for me, as is using locks.
In that case, please stop posting until you have the ability to express yourself in a fashion that does not cause complete misunderstanding by everyone involved. Your initial response in this thread strongly implied that you agreed with the grandparent poster, that
Usually, a technical solution is superior to any social solution.
This is the arrant insanity I disagree with. From your post, I thought you agreed with it. From other responses to you it seems that I am not the only person to think so.
Living with the surveillance state
If people have legitimate reasons for doing something, there can be no social contract and thus such a technical solution should be optional at best.
If in the ideal world of the social solution nobody cracks cryptography, then the technical solution of using cryptography everywhere is superior because it actively enforces the social solution and makes offenders 'impossible'. (With the notion of course that cryptography is merely delaying it's cracking rather than outright preventing it).
Of course, some users should still be allowed to force other behavior, considering they know what they're doing in these very special cases.
When we're all in favor of banning spying, it's better to prevent it altogether through technical measures than to rely on the goodwill of the spooks. Of course, some users should still be allowed to "spy" (think og law enforcement with a warrant), considering they have a legitimate reason to do so in these very special cases.
Living with the surveillance state
Living with the surveillance state
Living with the surveillance state
Living with the surveillance state
Living with the surveillance state
let's make it simple: would your social measures (deterrents) have the same effect if you did *not* have the technical measures in place or not? yes/no?
as for what is an absolute technical measure, try to pick your own locks. i bet you can't.
and i'm still waiting for those addresses, actions speak more than words do, you know... no addresses = you believe in technical measures, simple as that.
Living with the surveillance state
Living with the surveillance state
Let me repeat once more: in a world with reliable locks (where technical measures dominate) this information will be absolutely worthless. Lock can not be picked up anyway, so why not publish it's location? In our world where lock is just a side-show and social aspect is the primary one… of course one will not give up their primary form of protection so easily!
Again you were clearer than I. Exactly so.
Living with the surveillance state
Living with the surveillance state
Living with the surveillance state
Living with the surveillance state
Living with the surveillance state
Living with the surveillance state
On the other hand, when technological countermeasures are implemented to stop the ongoing spying it makes it impossible.
Living with the surveillance state
Living with the surveillance state
Living with the surveillance state
Living with the surveillance state
Living with the surveillance state
Living with the surveillance state
Living with the surveillance state
Living with the surveillance state
Living with the surveillance state
coupled with the utter lack of interest by 99.99999% (underestimate) of all programmers in proving correctness of their code
Living with the surveillance state
Living with the surveillance state
Living with the surveillance state
Living with the surveillance state
Living with the surveillance state
That said, you can establish some useful properties, such as "it will never infinite loop", "it will never gobble all your RAM", "it will do 'the right thing' on all possible valid inputs".
You certainly can't establish any of those in the general case, and even doing it in special cases requires some severely stereotyped code. (e.g. in the latter property, the inputs need a grammar which permits only inputs up to some maximum length, probably a very short one. That rules out most useful programs. The problems with proving the other two in general cases are hopefully obvious to anyone reading this!)
Living with the surveillance state
Living with the surveillance state
Wol
Living with the surveillance state
Living with the surveillance state
Living with the surveillance state
Living with the surveillance state
Equally, if facebook open sourced all its software today, I would bet it wouldn't make a blind bit of difference to the NSA.
Living with the surveillance state
Living with the surveillance state