Re: [quicwg/base-drafts] New connection ID message (#232)

[ianswett <[e-mail]>]

Ok, now I understand exactly what you're saying, and it was originally what I had in mind, and I'm concerned about two potential issues:
1) You can't safely increase the packet number by more than ~2^31 today, because we only have 4 byte packet numbers on the wire.  Ideally you'd like the packet number gap introduced to be totally random within the full packet number range?  Or possibly we're ok with a single bit of correlation?
2) If you tried to migrate a second time before receiving an acknowledgement for the first time.  I guess that ends up being a sub-case of 1, where the amount you can further increase the packet number becomes limited because the peer hasn't acked the first migration.  Going back to the baseline is dangerous because you don't know if any packets on the intermediate path were received, and it would take some carefulness to avoid re-using packet numbers.

-- 
You are receiving this because you are subscribed to this thread.
Reply to this email directly or view it on GitHub:
https://github.com/quicwg/base-drafts/issues/232#issuecomment-290725692