2 ms·
Briefly (?): As the Fedora bug notes, Default TLS security policies for the Web PKI belong at the OS level, where they can be set by a small number of people w
by tialaramex 8y ago
Briefly (?):
As the Fedora bug notes, Default TLS security policies for the Web PKI belong at the OS level, where they can be set by a small number of people who are really thinking about these problems, once. Microsoft has done this (not very well, but they've done it) in modern SChannel .NET support, it is finally relatively easy to write code which says "I don't know, please have the usual amount of security" rather than "I'm an expert, enable this protocol version, these exact ciphersuites, and use this key" or as it more commonly the case "Here's something I pasted from a 2010 Stack Overflow post". As much as possible, and despite all insistence to the contrary by individual developers, these preferences shouldn't live in some effectively unmaintained code in each individual component.
1. In particular if you don't actually care about the identity claimed in the certificate, you shouldn't bother verifying anything at all since obviously bad guys can trivially obtain a cert for some identity.
This was once a VERY common problem, and it's still seen in plenty of places (e.g. one of the toy operating systems doesn't check at all, so the whole OS is wide open, package updates, web browser, the whole thing - another has told users any time there's a problem to switch verification off manually, so I would expect few have it left switched on by now)
2a. Although it feels as though turning min-prime bits up to 2048 would fix things, in practice interoperability is very poor. You might get away with this, in some cases, for say IMAP and SMTP submission with TLS, but it's unlikely to be satisfactory on the general web. Try asking for Elliptic Curves instead.
2b. Everybody called it ARCFOUR (Alleged RC4) for legal reasons, it's fine, don't worry about _that_.
3. I sympathise here, in principle the user ought to have examined the cert, or at least key, they're trusting. But realistically users don't do this, so the security penalty is very small indeed. Examining the entire chain is largely futile, a lot more work and an even smaller number of users would gain anything from it.
8. There isn't really such a thing as "Certificate Transparency revocation checking" and revocation is largely a dead letter for big public CAs in the Web PKI. Almost certainly a bad guy able to attack you using a real cert isn't going to be prevented in their attack by revocation, they will block the signal and you'll fail open, because as we saw in every other item, users switch off any annoying fail-closed machinery.
10. This isn't great, but the idea behind SSL/TLS is that this ends up just being a drop-in replacement for the socket layer. With the exception of some scary features like TLS 1.3's Zero RTT no TLS expertise is needed, and the correct approach if you have no TLS expertise in the Emacs developer team is to just never use those scary features, they're supposed to be gated so that you can't use them by accident.
- wyuenho 8y agoWhat I'm working on is only tangentially related to that Fedora bug. Essentially, what the user is seeing in the GnuTLS debug log will supposedly be checked by NSM. I wasn't even aware of that bug report before I emailed emacs-devel. I'm glad I didn't tho, as that'll only add to the confusion. Emacs's use of GnuTLS already delegates something to the OS, namely getting the system's root CA certs. As for the rest, I lobbied leaving GnuTLS defaults alone instead of intentionally lowering it to accept 256 bit DHE primes only to be warned by NSM later. Also, the default cipher suite other than DHE prime bits is already left alone OTTB. But.... Emacs being all about freedom and infinitely hackable, the philosophy (philosophers?) demand that we let users to be able to configure the lower bound of acceptable TLS security in 2 different places, because someone somewhere requested it at one time, therefore his or her need trumps everyone else's default security or requests to have a consistent UI for TLS configuration. This is the context behind that fixation quote Lars wrote. Of course, I disagree strongly. The good news is, I've already returned all GnuTLS knobs to their defaults in my branch, and Lars has fixed the docs so people know they probably shouldn't touch those GnuTLS settings unless they want to bypass NSM completely, say running a elisp script for example. I'm not going to respond to the rest of your points, they require a fair bit of context to understand why I think they are problems. I'll probably explain it in a blog some time this weekend.
- oblio 8y ago> One of the toy operating systems doesn't check at all, so the whole OS is wide open, package updates, web browser, the whole thing - another has told users any time there's a problem to switch verification off manually, so I would expect few have it left switched on by now Could you say which OSes?
- tialaramex 8y agoSure, Haiku doesn't check, the Amiga systems advise people to switch off verification.
- waddlesplash 8y agoHuh? Haiku has had certificate verification for a long time, and the issues related to false-negative verifications got straightened out in 2012 or so; the web browser has always checked even before that, and for the package manager all builds past last December have had HTTPS downloads by default.