Let's Encrypt Releases Transparency Report -- All Zeroes Across The Board
from the now-let's-watch-if-anything-changes dept
We've talked a bit about the important security certificate effort being put together by EFF, Mozilla and others, called Let's Encrypt, which will offer free HTTPS security certificates, making it much easier to encrypt the web. They've been busy working on the project which is set to launch in a few months. But first... Let's Encrypt has released its first transparency report. Yes, that's right: before it's launched. As you might expect, there are a lot of zeros here:
This is actually pretty important for a variety of reasons. First, it clearly acts as something of a warrant canary. And by posting this now, before launch and before there's even been a chance for the government to request information, Let's Encrypt is actually able to say "0." That may seem like a strange thing to say but, with other companies, the government has told them that they're not allowed to claim "0," but can only give ranges -- such as 0 to 999 if they separate out the specific government requests, or 0 to 249 if they lump together different kinds of government orders. Twitter has been fighting back against these kinds of rules, and others have argued that revealing an accurate number should be protected speech under the First Amendment.
Let's Encrypt is, smartly, getting this first report out there -- with all the zeroes -- before the government can swoop in and insist that it has to only display ranges. In other words, this is getting in before any gag order can stop this kind of thing. Smart move. It's also nice to see them break down all of the different possible types of orders, rather than lumping them into more general buckets. That's an important step that it would be nice to see others follow as well.
Let's Encrypt is, smartly, getting this first report out there -- with all the zeroes -- before the government can swoop in and insist that it has to only display ranges. In other words, this is getting in before any gag order can stop this kind of thing. Smart move. It's also nice to see them break down all of the different possible types of orders, rather than lumping them into more general buckets. That's an important step that it would be nice to see others follow as well.
RSS


Reader Comments
Subscribe: RSS
View by: Time | Thread
No news is good news, eh? -- It's like the censoring here: if no one complains, must not be any!
[ reply to this | link to this | view in chronology ]
Re: No news is good news, eh? -- It's like the censoring here: if no one complains, must not be any!
[ reply to this | link to this | view in chronology ]
Re: No news is good news, eh? -- It's like the censoring here: if no one complains, must not be any!
No, we're complaining that you're never censored. Everyone else can read the rambling bollocks you post in every thread and then when everybody gets tired of your crap and asks for your messages top be hidden you complain falsely about being censored! Actually censoring you would be a fantastic boon to this site, but we never do that.
[ reply to this | link to this | view in chronology ]
Re: Re: No news is good news, eh? -- It's like the censoring here: if no one complains, must not be any!
What is funny is the tag "This comment has been flagged by the community. Click here to show it." acts like the super canary in the article. It shows both that something exists and that some people thought it was rubbish.
[ reply to this | link to this | view in chronology ]
A better (tech) option
[ reply to this | link to this | view in chronology ]
Re: A better (tech) option
[ reply to this | link to this | view in chronology ]
Re: A better (tech) option
The usual government trick won't work here, where a company can only give a range including 0, therefor making it impossible to tell if a company has received 0 orders or several, because they've already set the baseline, and any deviation will indicate a change.
[ reply to this | link to this | view in chronology ]
Re: Re: A better (tech) option
[ reply to this | link to this | view in chronology ]
Re: Re: Re: A better (tech) option
Paradox warning! That's exactly what a canary is designed for.
[ reply to this | link to this | view in chronology ]
Re: Re: Re: A better (tech) option
[ reply to this | link to this | view in chronology ]
Re: Re: A better (tech) option
[ reply to this | link to this | view in chronology ]
To further ease compliance for companies, they should just go ahead and create a single bracket: "zero or more". This would eliminate all the excessive cost associated with unnecessary reporting and save companies a zillion dollars. Moreover, it would help achieve full transparency on the topic.
[ reply to this | link to this | view in chronology ]
sdrawcaB
And all the fields were filled with a single character - usually a zero, just to keep the formatting correct.
[ reply to this | link to this | view in chronology ]
[ reply to this | link to this | view in chronology ]
https
What i am wondering if this is good or not. When everyone uses https will this lead to less secure https? Since it is worth more to make breaks? Like there were no viruses for mac...??
Greets,
Rob Veld
ValueBlue
[ reply to this | link to this | view in chronology ]
Re: https
I am not sure about this, but this is worth studying.
[ reply to this | link to this | view in chronology ]
Re: Re: https
TLS has a number of roles to play in network communications:
1) encrypt data to protect it from sniffing in-transit
2) authenticate data to verify it came from whom you expect
3) sign data so you know you got only the data you were expecting
Now here's how it breaks:
1) man-in-the-middle servers that sign with an alternate certificate. This can be done on the client (SuperFish), at the network edge (many gateway prodcuts), or anywhere upstream that has access to a trusted certificate on the client.
2) Yeah, this is broken at a number of levels, relating to item 1 -- there are many entities out there that can fake or phish the sender identity. Web of Trust helps a bit here, but the traditional methods (whitelist/blacklist) tend to fail, as the blacklists are improperly implemented in most places. How do you trust authenticity when most major governments have access to root certs?
3) This is actually still pretty safe; TLS itself has withstood most cracking attempts, and as a result, you're likely to have received exactly what the sender sent. The only issue here is that you have no way to 100% verify that the sender was who you thought it was, unless you got the signing certificate directly from them via a separate channel, and know that nobody else has access to their root certificate.
Aside from all this, verts generally work by exclusivity; the fewer organizations who have certificates, the more secure they are. If you remove the barriers to entry so that anyone can get a certificate, then that means that while a cert may be valid, it becomes more difficult to figure out if the person who owns the certificate is trustworthy in the first place.
If certificates are free, than you can rest assured that some botnet is going to have all its nodes registering bogus certificates that it can rotate through, giving the CNs all sorts of names, from "Bankof America" to "Aqqle" to "Trusted Update Pty, LLC". Then you'll have tons of signed malware coming down an encrypted pipe with a "verified" host at the other end. And you'll have all your personal data going up another pipe, similarly encrypted.
This doesn't make certificates bad, but they're not the panacea that many would believe -- they really only protect against casual sniffing and verify the data being transmitted between two (rightly) trusted points.
[ reply to this | link to this | view in chronology ]
Re: Re: Re: https
And it is the casual sniffing of governments that these certificates are primarily aimed at. If use of encrypt everything means that the Governments of the world cannot keep up with the decrypting of Internet traffic in real time, then most people's privacy improves. I do not ask that the system is perfect, just strong enough to force governments to target who they spy on.
[ reply to this | link to this | view in chronology ]
relevant precedent case law?
Warrant canaries seam like a speculatory concept at best to me, maybe there's something I haven't heard of yet though.
[ reply to this | link to this | view in chronology ]
Re: relevant precedent case law?
[ reply to this | link to this | view in chronology ]
That's a very succint explanation. Well done.
I was just reading this earlier- should be of interest to any one who would like to make conscious choices about who they trust. somewhat complex stuff unfortuantly.
https://blogs.fsfe.org/jens.lechtenboerger/2014/03/10/certificate-pinning-with-gnutls-i n-the-mess-of-ssltls/
[ reply to this | link to this | view in chronology ]
0-249 = 0
1-250 = 1
2-251 = 2
etc..
Would that run afoul of anything?
[ reply to this | link to this | view in chronology ]
Add Your Comment