[Cryptography] Paranoia for a Monday Morning

[Jerry Leichter]

On Oct 27, 2014, at 4:53 PM, Alfie John <alfiej at fastmail.fm> wrote:
>> It's easy to blame Adobe or the Microsoft of old for incompetent
>> programming; but even the latest IE, produced under what may be the
>> best "secure software development chain" in the world;
> Citation needed.
Widely described that way by people who I trust.  Feel free to accept or reject the characterization.

>> and Chrome, a clean-sheet, open-source implementation by a team
>> containing some of the best security guys out there; continue to be
>> found to have gaping holes.
> Clean-sheet? No. Chrome and Chromium for a long time used WebKit as the
> rendering engine.
Yes, but they put a huge effort into isolating the rendering engine to contain any problems.  And you need only look at the bugs reported and fixed *by the Chrome team* to see how seriously they looked at the code they did take.

>> At some point, you have to step back and admit that the problem
>> doesn't lie with the developers:  They are being set up to fail,
>> handed a set of specifications that we simply too hard to get right.
> Have a look at what Mozilla is doing. They developed a new language
> called Rust which has a focus on safety, and are using it to build a new
> rendering engine called Servo.
I'm developing a new, completely secure-by-design language in which I'll implement a *really* secure browser.  It's not available yet, but trust me, it will *finally* kill off all those pesky browser security bugs.

OK, Mozilla and Rust are a real effort with real publications.  But let's keep in mind that Mozilla itself started out to be, among other things, a secure browser (not just "more secure than IE6", a rather low bar, but really secure).  So did Chrome.  Both were major steps forward; both have been attacked successfully.  No claim that some new implementation *will be secure* is worth very much.  We'll see how their new effort holds up "in the heat of battle".

> It's not that the specifications are too hard, it's more that complexity
> in general is hard to manage. And with Firefox having over 12 million
> lines of code with over 3000 contributors, you're now applying Brooks'
> Law at absurd levels.
> 
>> And that, of course, raises the question:  Accident, or enemy action?
> 
> I'd put this into the paranoia basket.
You know, two years ago I would have similarly classified claims that the NSA was recording every phone call in whole countries, keeping records on every US citizen - and deliberately working to weaken cryptographic standards.  I agree with you that the NSA probably had little or nothing to do with the state of browsers - but mainly because they really didn't need to do anything, masses of attackable software were given to them for free.

                                                        -- Jerry

-------------- next part --------------
A non-text attachment was scrubbed...
Name: smime.p7s
Type: application/pkcs7-signature
Size: 4813 bytes
Desc: not available
URL: <http://www.metzdowd.com/pipermail/cryptography/attachments/20141027/89dc5d8e/attachment.bin>