4 ms·

There's been an attempt to fix this during the last two years, as the QUIC standardization talks really got going. A bunch of operator people expressed a need f

by jsnell 8y ago

There's been an attempt to fix this during the last two years, as the QUIC standardization talks really got going. A bunch of operator people expressed a need for some sort of in-path measurability, while the privacy people have expressed the need to avoid any sort of session linkage or other forms of information leakage. The most viable compromise proposal seems to be the spin bit[0] which gives RTT measurements, but it's not agreed whether a version of that will make it to the first release. [0] https://quicwg.org/base-drafts/draft-ietf-quic-spin-exp.html https://quicwg.org/base-drafts/draft-ietf-quic-spin-exp.html

pdkl95 8y ago
> The most viable compromise proposal seems to be the spin bit That's utterly useless. If the "spin bit" becomes widely used, I intend to write a trivial patch that sets it randomly on each packet. If packets could be dropped unless the bit is set to specific values at specific times, too much session information is leaking to middleboxen. More likely, the state of the bit doesn't matter so setting it randomly will discourage wasting a bit in the protocol with this kind of nonsense in the future.