3PHealth Blog
Privacy: X marks the spot where…
Sunday, February 26th, 2012
In the beginning there was the Internet… and the Internet was HTTP
The Hypertext Transfer Protocol (HTTP) is an application-level protocol for distributed, collaborative, hypermedia information systems. It is a generic, stateless, protocol which can be used for many tasks beyond its use for hypertext, such as name servers and distributed object management systems, through extension of its request methods, error codes and headers [47]. A feature of HTTP is the typing and negotiation of data representation, allowing systems to be built independently of the data being transferred. HTTP has been in use by the World-Wide Web global information initiative since 1990. (Source: RFC: 2616)
Those “request methods” mentioned above can contain something called an X header. They are fields in the request that begin with an X and allow for either non-standard or proprietary add-ons to the regular fields in the HTTP header.
So what’s an example of an X header?
x-h3g-network-quality – this tells you the type of network. “3G” for example. It’s a very straightforward idea on how you can add some data to the request that a browser makes to a Web server. The Web server now knows that the phone that the browser made the request from is on a 3G network.
Why are X headers so important?
Well because they’re a “standard” way to add “non-standard” data to the HTTP protocol. Pay special attention to the word “standard”. Everything is about standards.
So why all the fuss about “X” marks the spot.
Well there’s something else I need to tell you about how the Internet works. Remember our familiar picture – two cans and a piece of string. One can (the browser) talks to the other can (the Web server). Ok. Now we have to introduce a new term – Caching. A web cache is a mechanism for the temporary storage (caching) of web documents, such as HTML pages and images, to reduce bandwidth usage, server load, and perceived lag. (Source Wikipedia). The biggest caching engine in the world is probably Google or maybe Bings. But there are tens of thousands of cache sites out on the Web, virtually every ISP uses them to accelerate content. Browsers can connect with a Web caching server vs. going directly to the “origin” Web server. The cache serves as an intermediary.
So what’s the big deal?
Well some of those caching engines have critical problems when it comes to headers. (Source: link) and so if they see something they don’t understand they strip it out.
So what?
Well Do Not Track is a header – it’s not an X header (there’s no X_DNT) which they wouldn’t strip out. So older caching engines may not respect that new header (HTTP_DNT=”1”) and simply strip it out.
Ok… so why’s that a big deal?
Enforcement of the Do Not Track standard – that’s the big deal. How will the origin server ever know if the browser really sent a header that said “Don’t Track Me”? If an intermediary has removed it then there is no knowledge, and therefore it becomes unenforceable. Both Caching and Web servers will need to be upgraded to support a new standard header that is being transmitted in a non standard way i.e. not as an X-header, but as a regular header (that was never planned for in the original HTTP design).
And that is going to be the biggest problem facing the Do Not Track standard. Unless it’s implemented as an X header vs. a standard header then “EVERY” caching engine will have to be updated to support the new standard (to make sure it doesn’t strip anything important out). There are tens of thousands of caching servers out there that will have to be modified and upgraded. Normally no one would care about this… but now due to Privacy they will have to.
Therefore X marks the spot where in this case something is enforceable or not.
Privacy on the Internet is NOT binary – But what if it was?
Saturday, February 18th, 2012
Back at the end of October last year I wrote a blog post on why Privacy on the Internet is NOT “binary”. So I thought I would follow up with another blog post discussing the alternative viewpoint – What if it was binary?
Well first we need a little context about the Internet and how it works. At it’s core it’s basically two cans and a piece of string. The can (client) on one end talks to the other can (server) on the other end. The string acts as the communication layer (protocol) between the two cans. The protocol that runs across the string is “stateless”.
In short that means that there is no means of storing a users data between requests. The work around for this is…. Cookies which allows the server to track requests from the originating client (can). So how we do we stop one of the cans tracking the other one?
Obviously we cannot get rid of cookies otherwise the Web is going to grind to a halt. So what we need is another “binary” method of allowing me to control my privacy.
Enter the “Do Not Track” option. This is a browser setting that tells the first can (client) to send a message to the second can (server) that it doesn’t want to be tracked. (We’ll discuss how you enforce this in a later blog). Great – I now have a method of communicating with a Web server telling it that I want to remain Private.
Well here’s where it gets interesting or should I say “binary”. Most people think of binary as either a Zero or a One (0,1) however there is actually a third state in the case of the Do Not Track header and it’s called “not set”. So in essence the absence of a Zero or a One is another “state”.
So what does this mean. Actually quite a lot. If the user turns on his Do Not Track option then the answer is obvious. The server MUST not track him/her. If the Do Not Track option is turned off then the server can track away. However if the option is NOT set at all what is the server to do?
Here comes the really big problem – what if the Web service providers are forced either by legislation or by the W3 standards group to start honoring the Do Not Track option? It’s really easy to spot a Zero or a One as part of the communication request. But what if you don’t see anything? Well it means that you have “insufficient data” to make a decision. Which in turn means that you have to send another message back to the client (can) and ask it for more data.
So how does this all work? Well I have to figure out how to send a pop up message to that device and ask the user if they want to be tracked. Well you can guess what the answer will be – no thanks! So in essence we’ve wasted bandwidth, processing cycles, battery life (on a Mobile phone). And at the same time we’re ruining the users experience. The only thing this third “binary state” (which actually doesn’t exist) has done is draw attention to the problem. Think of all the code that is going to be written to solve this problem. What a waste of resources!
Everybody is trying to solve this “Do Not Track” problem through the lens of the “Web server”. (Because it’s where all the content is stored). What no one is thinking about is solving it through the lens of the Browser user i.e. ME. It’s pretty clear that Privacy cannot be binary – there’s simply not enough information to make real decisions which effect my rights.
By making it binary the Web is risking billions and billions in revenue. Why? Because people want a Choice – if forced to pick between Privacy and NO Privacy which one would you pick.
Pretty obvious really.
Privacy: Does it Scale
Friday, February 17th, 2012
Well it sure has been a busy day on the Privacy front. However as I watch all of this, I can’t help but ponder if people really understand the magnitude of the problem the Web faces. Lets talk about “Scaling”.
On a thermometer today would have Privacy scaling above boiling point. Something is going to happen for sure. Browser companies will have to make changes, Web services will have to make changes – but does it scale?
Let’s think about it for a moment in the context of the “Do Not Track” header which is proposed standard to help curb the Privacy problem. If the Do Not Track header is on then Web services “MUST” observe the customers requests and not track them. There is no ambiguity in the word MUST. It’s a have to do vs. a should or may do.
But what are the implications? Well no one really knows, but Web servers, ISP, all tiers of the Internet will be affected. They will have to add code, a lot of code, to accurately support this measure. We’re talking years of coding efforts in some cases.
Up until now we’ve been living in a free for all world. Then Mobile came along and the world changed. Now Privacy is Personal and that means everybody is going to have to honor my request to not be tracked.
And lets not forget “location”. What if I get on a plane and access a Web site from a foreign country. Before they can really service my request there’s a lot of context they have to determine. And that translates into code, which translates into bandwidth and processor issues.
People want a Choice – because without it there is no Privacy. Today was just a glimpse at what is coming down the pike. The temperature is rising, and it’s rising fast.
Privacy: Clash of the Titans – Where Humans battle Algorithms
Friday, February 17th, 2012
Well it seems like a epic battle is about to get underway. To the North are the Humans – demanding that their Privacy be protected and to the South are the Algorithms – constantly “searching” to undermine my Privacy in the name of a better experience.
Years ago when I was studying for my Airline Pilot’s license, one of subjects was the Inter Tropical Convergence Zone, colloquially known as the “ITCZ”. It’s that area on the planet (Source: link) that circles the Earth, near the equator, where the trade winds of the Northern and Southern Hemispheres come together. The intense sun and warm water of the equator heats the air in the ITCZ, raising its humidity and causing it to rise. As the air rises it cools, releasing the accumulated moisture in an almost perpetual series of thunderstorms.
Well that’s exactly what’s happening right now. Especially with the announcement this morning by the WSJ (link) that Google is tracking iPhone users. There is now intense heat coming to bear from the “Humans” regarding their privacy. Something is going to give. People want a Choice when it comes to their Privacy. Sort of like “Freedom of Speech”. You just can’t take it away from people and expect them to not be upset.
For those of you who are interested in Privacy I suggest you head over to this link It’s the W3.org’s Web site on Public Tracking. Take a spin through the mailing lists. It’s fascinating. Everything is hinging on one binary number… 1 This is the setting that users will be able to control in their browser to stop Web sites tracking them. Unfortunately as we’ve seen today Google et al is ignoring those settings.
This cannot end well for the Algorithms – and here’s why. The Humans (Lawyers and Privacy Rights Activists) will write legislation to reign in the ability to track. As scissors always beats paper so will the legislation. However here’s where the law of unintended consequences comes into play – Privacy on the Internet is NOT binary (as I have written about before – link). Setting a Do Not Track header to 1 is not enough context to make a really informed decision about what a customer wants.
Ask anyone these days – Privacy is about Choice not about a number. Any solution MUST offer me a Choice otherwise it will fail. In the meantime get ready for some serious thunderstorms – the ITCZ is heating up.
Digital Privacy Twister
Friday, February 17th, 2012
Yes, Twister. The fun, bright colored game where you get twisted up with all your friends and would-be teenage loves. Actually, the rules of Twister are more clear than the twisted Privacy policies that dot the Web these days – which in my opinion, are less about privacy and more about making money. The only choice I typically get is no privacy or don’t use my Web service.
It is impossible to ignore the increase in coverage regarding digital data privacy. Today’s Wall Street Journal headline about Google bypassing iPhone Privacy Settings may just be the fateful move that brings everyone playing Digital Privacy Twister crashing down to the mat.
But before jumping to any “Google is the new Evil Empire” conclusions, have a look at this hurried, yet thoughtful post by Technology Media blogger, John Battelle, A Sad State of Internet Affairs: The Journal of Google, Apple and “Privacy”. Battelle, rightfully questions whether or not the default “privacy settings” in iOS are designed to protect yours and my privacy rights or protect Apple’s advertising revenues.
Sadly, the rules about Internet Privacy are simply not clear and even those being debated and proposed by the best minds in the space have to consider the impact changes will have on established business models and legitimate uses for data sharing between organizations. Privacy is about balancing the rights of you and me, as citizens (not just consumers) and a business’s right to make money. If you mandate technology changes to stop privacy abuses, then how that impacts legitimate data use and sharing MUST be considered or you start ripping apart the very fabric of the Web – the mat holding all the brightly colored website dots together.
Maybe we are too smart for or own good. I bet if you asked your children, they would say something wonderfully simple like, “Just ask me my permission.” or, “I’ll tell you if I trust you.” Or perhaps, “Your a stranger. Until I know you better I won’t tell you my name or where I live.” Instead of trying to re-weave the Web and break what works, why not just ask the user. Give them the Choice to share or not to share with any given site. Give them a Choice about what to share – location, but not name. Device information, but not cell phone number. Privacy is not binary. Privacy cannot be “solved.” Privacy is a right. Ask permission (in a simple, straightforward manner) and then respect it. Period.
Digital Privacy should be about delivering the appropriate Web Experience base on what is shared, not taken. An anonymous experience or a rich experience, or something in between. It should be MY choice and should not be all or nothing. People like choice. When you deny that choice based upon less than transparent practices and policies, people get really angry. Angry people stop doing business with you and tell all their friends.