Slashdot Mirror
Google's Obfuscated TCP
agl42 writes "Obfuscated TCP attempts to provide a cheap opportunistic encryption scheme for HTTP. Though SSL has been around for years, most sites still don't use it by default. By providing a less secure, but computationally and administratively cheaper, method of encryption, we might be able to increase the depressingly small fraction of encrypted traffic on the Internet. There's an introduction video explaining it."
Firefox isn't helping the lack of SSL on the web by throwing a ridiculous warning when using self signed certs. Browsers should treat self signed certs as 'unsigned with the added bonus that communications can't be eavesdropped' instead of freaking out that you might not know who you are talking too.
self signed certs aren't appropriate for processing credit cards... but not every site that has forms needs that... and simply removing eavesdroppers would be a step in the right direction.
The video starts out saying that increased encryption is needed thanks in part to warrantless government surveillance. It then goes on to describe a system that assumes no MITM attacks can exist. The fact is, however, that governments are entirely capable of performing MITM attacks, as can telecommunications companies; and if it becomes popular we may see more techniques that allow individuals to perform MITM attacks. While this algorithm has significant merit, care needs to be taken to avoid a false sense of security.
>
Are we supposed to like Google or not now?
I'm confused :(
Realizing that large corporations consist of many separate interests might help alleviate your confusion :-)
Project owner's page is at: http://www.imperialviolet.org/ if you wanted more info.
By providing a less secure, but computationally and administratively cheaper, method of encryption, we might be able to increase the depressingly small fraction of encrypted traffic on the Internet.
If the encryption is computationally cheaper, then the decryption is computationally cheaper. I'd rather people know that what they're sending over the 'net can be sniffed than have them think that because example.com uses Rot13 encryption their traffic is private.
So, basically we have the same concept as SSL, except instead of trusting the CA signature on the certificate, we trust DNS.
Forging a CA signature on a certificate would be a BIG DEAL.
Forging a DNS entry, especially with ISP cooperation(read government snooping), is DEAD SIMPLE.
So we replace real security with, well, a CPU hog that's only a smidge better than running everything in the clear. It only keeps out the MOST casual, lazy, and uninterested snooper.
we might be able to increase the depressingly small fraction of encrypted traffic on the Internet.
I agree that this would indeed be a good thing for several reasons. An encrypted message in a medium where most everything is plaintext may attract the attention of attackers or, worse, be seen as "suspicious" by a government. (Certainly the U.S. and the PATRIOT Act spring to mind, but let's not forget the truly oppressive governments such as China's and any number of third-world dictatorships.) If online privacy via encryption comes to be a right that everyone gets used to enjoying—much like how almost all mail is sent in sealed envelopes, whether or not its contents are sensitive—then it will be that much harder, for technical and/or social reasons, for an authority to take away. If Obfuscated TCP is even a token step in that direction (and it seems to be a bit better than that), then it is probably a good thing overall.
Someone earlier today on Slashdot was plugging Cory Doctorow's Little Brother, and I'm going to follow that example (you can read it for free!) as part of it advances the same idea.
Anybody remembers what hapenned to RFC 2817 ? It tried to address this very pb by introducing the "Upgrade: TLS/1.0" header and the "426 Upgrade Required" status code, but I don't think any browser or server implement them.
Implications?
but it can assuage untargeted, dragnet sniffing of backbones
Can you read the subtext there? Snap, how's that for an implication? Privacy. (from the government.)
<xml><I><am><so><damn>Web 2.0</damn></so></am></I></xml>
At least watch the video before asking questions that it answers.
"By providing a less secure, but computationally and administratively cheaper, method of encryption, we might be able to..." give people a false sense of security.
Remember, weak encryption can be worse than none as Mary Queen of Scots found out at the cost of her life (see http://www.nikon.com/about/feelnikon/light/chap04/sec01.htm).
Running your web sites on non-standard ports is a great way for your web site not to be accessible to users accessing the internet through firewalls that limit egress traffic based on TCP destination ports.
It may be slightly off-topic but the parent has a VERY valid point. Self-signed sites are encrypted but best of luck trying to get people to use them thanks to the 3-clicks required and SMALL text. When I used the new firefox release I was even confused at first.
Now back to obfuscated TCP: This is on par with using NAT to fix the lack of IP addresses. Just fix the damn thing properly and stop screwing around with time wasting half-fixes (yeah they admitted it).
About the only thing this is going to do is make troubleshooting problems with Ethereal or other packet sniffers a pain in the a$$. Thanks.
Just make SSL cert cheaper and get rid of all the multiple server licensing and crap.
Make the damn thing ran by a non-profit organization and cut the cost.
For a public site using non-standard ports is an easy method to shoot yourself in the foot - you immediately block all users behind proxies or firewalls that only allow communication on "standard" web ports.
He signs posts:
Adam Langley
Google Engineering
... not google's idea?
This is stupid. Nobody crawls the web by sniffing traffic. Google and everyone else connects to webservers the same way you do. For your post to make any sense we have to assume that this would make sites using Obfuscated TCP inaccessible by default, which goes against its entire design philosophy.
It should by default accept a self-signed cert transparently without any fuss. It SHOULDN'T show a big green lock. It should just be a regular connection. If the self-signed cert changes on a subsequent visit, THEN they should get a warning. That's it.
The problem is, we've tried to train users to look for the "https" or the lock, or both. Getting rid of the lock for self-signed connections is fine, but the https is still there, and it's misleading.
"Well kids, you tried your best, and you failed. The lesson is, never try."
misleading title
It's a little-known fact that "Posted by kdawson" is Slashdot-ease for "better read TFA because TFS is FUBAR".
'a';DROP TABLE users; SELECT * FROM DATA WHERE name LIKE '%'... if you're reading this, it didn't work.
If you watch the video, your brain will leak out through your ears. It's terrible. Why produce a video which seems to be a black screen with a dark blue line wiggling when the person talks? Why pick a person with a crappy British accent and a speech impediment? Who's going to understand? Why flash up a couple of words here and there like "SSL" and "HTTP"? Why produce such a steaming pile of crap and call it an "introductory video"?
Instead, whoever is the video star in this could have written down their ideas in plain text. That would allow for easy reading and comprehension by people all over the world. Maybe I can read quickly. Maybe I don't want to sit around waiting for you to lisp and stammer through your presentation. Maybe I'd understand it better if I read it than if I heard it on a crappy video. Maybe I don't want to waste my bandwidth downloading several megabytes of video, where the same information in plain text might be a few kilobytes.
It's so that it costs more to index the web
No.
Indexing has nothing to do with sniffing data. Obfuscated TCP is opportunistic, which means that it only happens if the client requests it. It is not required.
This doesn't change indexing at all, except that if you really wanted, you could encrypt your spider traffic when spidering compatible web servers.
This is NOT "security through obscurity". It is a form of what's called "Better-Than-Nothing" security. This topic is even worked on in the IETF for IPSec. (btns working group). The idea is that, with a gnashing of teeth, you should admit that the deployment barrier for proper security, which solves all your problems, is too high for general adoption. Then, after having made up your mind in that respect, try to figure a method that only solves a subset of problems, but for a significantly lower price. Obfuscated TCP seems to provide the property of confidentiality, but not endpoint authentication. You are right that you can still do MITM, but still it is better than nothing. I proposed something similar for wireless LANs to some vendor some time ago, which I called WPA-NoAuth, where the traffic between STA and AP gets encrypted, but none of the two endpoints authenticates to each other. This would typically cater for "web-portal" authentication, where the authentication happens after associating with the AP, and no proper security schemes like IEEE 802.1X or WPA-sharedkey can be used. Wasn't picked up with great enthusiasm though. Let's see if obfuscated TCP does. (Disclaimer: I have nothing at all to do with the obfuscated TCP proposal, nor do I work for Google)
Continuous positive slashdot karma since... uh, maybe next year.
You answered your own "why not?"
NSA and ISPs like to snoop, data mine and traffic shape. Traffic shaping can even be a good thing in certain situations (and I'm not talking Comcast here.) It's highly unlikely anything like obstcp will ever get standardized, since it prevents exactly what you just mentioned.
All those "cannot"...
The point of this obfuscation is to make it harder to sniff traffic, but not have the high administrative and computational cost of actually making it impossible.
You might be protecting your traffic from the wardriving kid next door, but not from your ISP, let alone NSA.
Truth arises more readily from error than from confusion. -Francis Bacon
I think you are confusing encryption with authentication.
Even if you are presented by a Man in the Middle's self signed cert and accept it, you're traffic's still encrypted and secure from eavesdropping. It's just that the receipient changed without you realizing it.
So you now have one eavesdropper, while plain unencrypted HTTP would allow an arbitrary number of them without you knowing. Security would be broken either way, but encrypted, not authenticated traffic still is preferable to neither encrypted, nor authenticated traffic.
Truth arises more readily from error than from confusion. -Francis Bacon
you're [sic] traffic's still encrypted and secure from eavesdropping
Except from the party that did the MITM attack, which is most often the party that you want to prevent watching your traffic, you know, the one that is actually interested in sniffing your traffic..
c++;