Live data from Hacker News

Firefox 143 for Android to introduce DoH

blog.mozilla.org

51–60 of 121 posts

Re: Firefox 143 for Android to introduce DoH

#51

DoH centralizes DNS traffic at a few DoH resolvers. Bad thing.

Actually, DoH doesn't change the situation here one way or the other, it's just a transport. It's true, that Firefox's approach to DoH ("trusted recursive resolver") does. centralize traffic some, but DoH need not be deployed this way. For example, Chrome does what's called same provider auto upgrade, which doesn't change the resolver, but just tries to use DoH if available.

Re: Firefox 143 for Android to introduce DoH

#52

Earlier quoted context omitted.

Why is DoH in the browser more private than DoH in the OS?

Because there are fewer actors to trust. In the OS you need to trust (1) the OS vendor, (2) the client vendor & (3) any VPN app or HTTP intermediary that's integrated with OS network APIs. In the client you need only to trust the client vendor.

Surely you're at the mercy of the hardware vendor and os in either case?

Granted, the os would need to read your address space, not simply supply a recording DNS API, but still...

Re: Firefox 143 for Android to introduce DoH

#53
Does anyone know how to force disable DoH on a network?

In https://support.mozilla.org/en-US/kb/canary-domain-use-appli... it says that the canary domain does not apply for users who have made the choice to turn on DoH by themselves.

I want to avoid running an sslproxy, and it seems an application level proxy on the firewalls is necessary.

Re: Firefox 143 for Android to introduce DoH

#54
post #53

Does anyone know how to force disable DoH on a network? In https://support.mozilla.org/en-US/kb/canary-domain-use-appli... it says that the canary domain does not apply for users who have made the choice to turn on DoH by themselves. I want to avoid running an sslproxy, and it seems an application level proxy on the firewalls is necessary.

Outside of IP-blocking known popular DoH hosts (e.g. https://github.com/jameshas/Public-DoH-Lists, and even then it's not the best since there's overlap with popular DNS hosts like Cloudflare), there's no good way to do it without break-and-inspect. That's because DoH is TLS traffic over 443, just with DNS inside instead of HTTP.

Re: Firefox 143 for Android to introduce DoH

#55

Not sure why it took so long for Mozilla to expose the setting on Android, it's been a 'secret' setting for a long time. In fact, sometimes they let features ride the rails for a little bit too long IMO. For Waterfox for Android I exposed the setting by default and also added an addition DNS over Oblivious HTTP setting (DoOH) which uses Fastly as the relay (they host and control it, for privacy sanitisation) and Clou…

Hey, just wanted to say thanks for your work on Waterfox!

Re: Firefox 143 for Android to introduce DoH

#56

Not sure why it took so long for Mozilla to expose the setting on Android, it's been a 'secret' setting for a long time. In fact, sometimes they let features ride the rails for a little bit too long IMO. For Waterfox for Android I exposed the setting by default and also added an addition DNS over Oblivious HTTP setting (DoOH) which uses Fastly as the relay (they host and control it, for privacy sanitisation) and Clou…

> DoOH

How is the latency?

Re: Firefox 143 for Android to introduce DoH

#58
post #31

DoH centralizes DNS traffic at a few DoH resolvers. Bad thing.

One approach to mitigate this is to spread the queries to multiple DoH providers: https://www3.cs.stonybrook.edu/~mikepo/papers/k-resolver.mad...

I'm not sure that this mechanism delivers the desired privacy benefit, and it's quite hard to make sure it does so.

For example, the paper you cite here uses consistent hashing, where you hash the domain name and then divide by K where K is the number of the resolvers. However, consider the case where you have a conceptual site (e.g., Gmail) which actually loads resources from multiple FQDNS. For example, if you pull up the network console for a naive load of X, it loads resources from at least the following domains:

x.com, api.x.com, abs.twimg.com, pbs.twimg.com, video.twimg.com

All of these are relatively characteristic of X, but in a naive design they would often be loaded from multiple resolvers, with the result that you're actually sharing your browsing history with more resolvers than if you just had a single resolver. As is suggested by this list, you might be able to improve the situation somewhat by hashing on ETLD+1, but even here there are 2 ETLD+1s, which is not an uncommon scenario.

In general, for this strategy to work you need to hash not on the domain but rather on the conceptual site, but this information is not readily available to the browser.

Re: Firefox 143 for Android to introduce DoH

#59

This doesn't address why this needs to be built in to the browser when Android already does DoH by itself. I assume there's a reason, does anyone know what it is?

Query statistics is valuable data you can sell. Client DNS queries are in that regard similar to search queries and a default search engine setting, you can sell that to the highest bidder. So browser makers are incentivized to implement their own resolver with its own set of DNS servers instead of just the system ones. Either because they want to sell those statistics themselves. Or because they want to protect thei…

the browser, being the originator of these DNS queries, already knows what website you are visiting.

Re: Firefox 143 for Android to introduce DoH

#60
post #33

>DNS query [...] in the clear. [...] (DoH) plugs this privacy leak [...] no one on the network, not your internet service provider [...] can eavesdrop on your browsing Whoever could see DNS traffic can still see the target you're connecting to...

This is correct. The right way to think of DoH is as part of a package of mechanisms (including ECH) that collectively are designed to close network-based leakage of browsing history. Used alone, it has some value but that value is limited.
Post reply on HN