FREAK Vulnerability -- Any comments from our IT pros?

Only the comment that this is the result of government regulation.
 
Only the comment that this is the result of government regulation.

This. But server admins who neglect to update their openSSL implementations aren't blameless, either.

Rich
 
Just yet another encryption vulnerability with a fancy name and webpage. Anything that matters that's ran by competent folks will be patched.
 
Just yet another encryption vulnerability with a fancy name and webpage. Anything that matters that's ran by competent folks will be patched.


This.

But there's a lot of stuff that isn't run by competent folks. Enjoy the Internet!

By the way, FCC will fix it all, just ask any politician. They know.
 
Elucidate please

Yeah, I was under the impression that it was people behaving badly that were responsible. Oh no, what was I thinking, we don't blame people for their actions, we blame the enablers. Everybody is a ****ing victim in our society, and it's government to blame,
LIBERAL! government.
 
Last edited:
Yeah, I was under the impression that it was people behaving badly that were responsible. Oh no, what was I thinking, we don't blame people for their actions, we blame the enablers. Everybody is a ****ing victim in our society, and it's government to blame,

LIBERAL! government.


So who you planning on blaming for this one? The rest of the statement is just babble.

What penalty should an OpenSSL coder writing in their free time and giving the code away for free, be penalized with?

You like government. Should government be the enforcer of penalties for bad code?
 
So who you planning on blaming for this one? The rest of the statement is just babble.

What penalty should an OpenSSL coder writing in their free time and giving the code away for free, be penalized with?

You like government. Should government be the enforcer of penalties for bad code?

lol, Henning obviously has never written complicated software.
 
Why do you say this? What makes this attack special?

Elucidate please

It's an ad hominem against government regulation.

There is a morsel of truth: part of the issue here is that (at one point) the US government had restricted companies from "exporting" or selling products to the overseas market that included stronger encryption capabilities (number of bits in the encryption key). That restriction has since been lifted. The restriction was intended to ensure that certain US agencies could crack the encryption by requiring that it be weaker.

So browser makers and others built browsers that included the lower grade encryption capability. And many websites implemented only the weaker encryption capability as they didn't want to lose customers that that were limited by the encryption restriction.

The limitation was lifted a number of years ago. But many browser makers and web sites failed to implement stronger encryption that was available to all.

And then there are the OpenSSL issues.... which again are caused by web site operators that have failed to upgrade.

The government restriction issues are amplified by the current debate between certain government folks in the law enforcement sector (who don't want strong encryption on mobile devices) and device/software makers (like Google & Apple) that are setting device encryption as the default on their devices. The topic wasn't helped by the claims that TrueCrypt dropped their software over government concerns.
 
So who you planning on blaming for this one? The rest of the statement is just babble.

What penalty should an OpenSSL coder writing in their free time and giving the code away for free, be penalized with?

You like government. Should government be the enforcer of penalties for bad code?

I'm not blaming anyone about it, what the hell do I know?:dunno: I do know I blame the need for security on people who decide to be crooks.
 
lol, Henning obviously has never written complicated software.

That's my point, I'm trying to figure out WTF the government has to do with writing complicated software and how you blame defects in it on them?:dunno:
 
I'm not blaming anyone about it, what the hell do I know?:dunno: I do know I blame the need for security on people who decide to be crooks.

Strong encryption is pretty much a requirement these days because catching the crooks with the way the internet works is nearly impossible.

The problem is that the government does not like the strong encryption because they don't have a good way to decrypt the strong encryption.

The other problem is that writing strong encryption software is really really really really ****ing hard and you need to be incredibly smart. Smarter than most any developer is. So we all utilize common libraries, like OpenSSL to do it with. It's best that way because if everyone tried to write it themselves there would be nothing strong about it.

The problem arises when a problem in something like OpenSSL is found, because, then pretty much all software is vulnerable to the problem since that's what everyone uses.

Many of these attacks though in the real world just aren't very practical at all and very hard for an attacker to actually pull off. The guys that find these issues have really big heads and like to make them sound a lot more serious than they actually are...

Anyways, anything important will be patched, the world will keep turning, and we'll all keep using encryption that is incredibly difficult for the government to decrypt.

If every single piece of data flying around the internet is strongly encrypted (that is the direction things are going) it's pretty much impossible for the government to make sense of that data. Maybe they could attack one small piece and eventually after burning lots of energy decrypt it..but that'd be one small piece.

Really the internet is the best thing that has EVER happened to freedom. You can't regulate it. You can try..but good luck.

The crappiest weakest forms of encryption we use these days make Enigma look like a total joke. The stronger stuff, which is basically what is used for everything, is basically impossible to decrypt (properly implemented).

Computers are so damn fast these days that the penalties of using incredibly strong encryption is nil. That really rubs the Feds the wrong way.

There is strong evidence that the government intentionally created issues in encryption products years ago so that only they would know about the bugs and they could use them to decrypt. Of course, now that stuff is being found, and rapidly removed. It's a never ending battle for them to attempt to leak attacks into this stuff in advance...

Awareness is incredibly high these days that the government is doing this (thanks to guys like Snowden). That increased awareness is causing LOTS of people to audit all the encryption software that is used looking for bugs. That will mean lots of vulnerabilities are discovered for awhile and fixed.

People finding these vulnerabilities is a VERY GOOD THING. Because there is a strong chance the Feds already knew about them and were utilizing them.

Now you might ask...why care if the Feds can decrypt your data?

You should care because if they know of a way, that very likely means that people wanting to steal your data also know of that way, and will utilize that method to steal such data. For someone like me -- who is responsible for software that protects lots of credit card numbers -- it's the sort of thing that keeps me up at night.
 
Last edited:
Let me ask this, was this vulnerability found through a detrimental act, or was it found by someone looking to get paid for finding it?
 
Let me ask this, was this vulnerability found through a detrimental act, or was it found by someone looking to get paid for finding it?

Most of this is being found by either employees of companies with strong interests in ensuring strong encryption can't be cracked (because their business relies on it) or academic folks that are just plain interested in it and trying to make a name for themselves.
 
Most of this is being found by either employees of companies with strong interests in ensuring strong encryption can't be cracked (because their business relies on it) or academic folks that are just plain interested in it and trying to make a name for themselves.

Right, so this isn't really a problem, this is a normal part of the evolution process. This was found by someone being paid to find it. Why do we have to blame anyone for anything?:dunno:

I think having an open protocol designed by some really smart people s the only practical way to go about it. You will never achieve ultimate security as long as people will look for an exploit.
 
Last edited:
That's my point, I'm trying to figure out WTF the government has to do with writing complicated software and how you blame defects in it on them?:dunno:


The simple answer? Writing good security related code is really hard.

Government made it unnecessarily harder. They did so under the auspices of wanting NSA to be able to crack the weak sauce that government would allow US companies to communicate to foreign countries with.

The natural result of making it harder? More screw ups writing it. It was nearly inevitable.
 
Let me ask this, was this vulnerability found through a detrimental act, or was it found by someone looking to get paid for finding it?


This vulnerability was found in an audit I believe but plenty have been found via actual exploit. There's already been 215 or so vulnerabilities in common software found in 2015 and it's just now March. There were 9687 in 2014.

*From the MITRE CVE database.
 
The other problem is that writing strong encryption software is really really really really ****ing hard and you need to be incredibly smart. Smarter than most any developer is.

I don't think so. In my humble opinion, if you want to write a safe and reliable encryption software, you need to be well educated in the field, be studious of previous failures, have reasonably wide ranging experience, and be dilligent. A common amount of smarts is required, but nothing extraordinary.

If you are concerned with cryptonalysis, or cryptography as such, then perhaps uncommon smarts are necessary. I'm not a judge of that.
 
I don't think so. In my humble opinion, if you want to write a safe and reliable encryption software, you need to be well educated in the field, be studious of previous failures, have reasonably wide ranging experience, and be dilligent. A common amount of smarts is required, but nothing extraordinary.

If you are concerned with cryptonalysis, or cryptography as such, then perhaps uncommon smarts are necessary. I'm not a judge of that.

If you use an existing library to write it like OpenSSL l agree. If I decided to write JesseEncryption and didn't use any existing known encryption libraries i would be asking for trouble. Not to mention it would be a giant red flag in a level 1 PCI audit.
 
Here is a non-hysterical explanation of the Freak attack from an expert:

FREAK: Security Rollback Attack Against SSL

The government used to restrict export of encryption software under laws that controlled export of munitions. This was back in the early 90s.

Somewhere I still have a t-shirt with an encryption Perl encryption script encoded as a bar code. The shirt has a label that says 'This shirt is a munition!'

These restrictions no longer exist, but several commonly used protocols still have rollbacks that date back to those bad old days under the Clinton administration.

Freak really isn't a big threat to the average POA participant.

What IS the big threat is ransom-ware. That's why I install CryptoPrevent on all my Windows boxes. [Note: I had to use tinyurl because the real domain name has a POA forbidden word in it.]

Even the free version does one thing that will thwart a lot of malware. It alters the registry such that files with suffixes like ransomware.pdf.exe or ransomware.jpg.bin will not execute.

I have no connection with CryptoPrevent except as a satisfied customer.
 
Here is a non-hysterical explanation of the Freak attack from an expert:

FREAK: Security Rollback Attack Against SSL

The government used to restrict export of encryption software under laws that controlled export of munitions. This was back in the early 90s.

Somewhere I still have a t-shirt with an encryption Perl encryption script encoded as a bar code. The shirt has a label that says 'This shirt is a munition!'

These restrictions no longer exist, but several commonly used protocols still have rollbacks that date back to those bad old days under the Clinton administration.

Freak really isn't a big threat to the average POA participant.

What IS the big threat is ransom-ware. That's why I install CryptoPrevent on all my Windows boxes. [Note: I had to use tinyurl because the real domain name has a POA forbidden word in it.]

Even the free version does one thing that will thwart a lot of malware. It alters the registry such that files with suffixes like ransomware.pdf.exe or ransomware.jpg.bin will not execute.

I have no connection with CryptoPrevent except as a satisfied customer.
The tshirt has the pgp encryption algorithm on it. I still have mine and wore it to my 3rd year law course on Computer Law. The prof knew what it was, so we had a great time explaining to the rest of the class what it was.
 
I have a couple of sites here I have to leave an export cipher available on due to a customer requirement that I'm stuck supporting :(

We have some boxes in the field that connect back to our datacenters via ssl that aren't easily updated (recertification is going to take forever... one is part of an avionics suite, and another is separate from the avionics, but still on the a/c) that got all ****y when we disabled SSL3 and would only play nice with TLS1 with a weak cipher.
 
I have a couple of sites here I have to leave an export cipher available on due to a customer requirement that I'm stuck supporting :(



We have some boxes in the field that connect back to our datacenters via ssl that aren't easily updated (recertification is going to take forever... one is part of an avionics suite, and another is separate from the avionics, but still on the a/c) that got all ****y when we disabled SSL3 and would only play nice with TLS1 with a weak cipher.


That's another significant problem. Old stuff that works that no one wants to pay to update. Every company has some if they're not a startup. And very few companies budget to maintain old stuff like that or pay to properly rebuild it over time.
 
That's another significant problem. Old stuff that works that no one wants to pay to update. Every company has some if they're not a startup. And very few companies budget to maintain old stuff like that or pay to properly rebuild it over time.

Yep... and it's particularly fun when our PCI scans and other automated audit scans keep carping at them, but we can't do anything about it.
 
Yep... and it's particularly fun when our PCI scans and other automated audit scans keep carping at them, but we can't do anything about it.


You can write business exception documentation and have execs sign it, to maintain the fantasy that PCI actually works! ;) ;) ;)
 
Back
Top