Measurement and data analysis in sprint 14

[bassosimone]

Following up from #581. Need to think about:

1. MITMEngine?
2. extract TTL, IP ID and RTT and/or Impossible Cipher 44.32%, Impossible Extension 42.09% , Impossible TLS Version 5.63% , Impossible Quirk 4.88% , Other 3.09% (xref MALCOM?)
3. use eBPF/create PCAP

We should in parallel use Jafar to provoke similar failures. Do we collect all the information we need? Does the usage suggest we can post-process measurements differently?

---

• blog post re: additional insights wrt SNI blocking [some progress]
• blog post examining blocking of DoT
• consider simplifying tls handshake to make reasoning on errors easier
• can we tag each resolved IP with its corresponding ASN?

---

This is part of our efforts to perform testing and QA and cycle back to increase software quality. As part of this effort, we may also be writing about interesting findings we have found.

[bassosimone]

Inflated effort and removed effort from #646 accordingly.

[bassosimone]

Borrowed 8 story points from https://github.com/ooni/probe-engine/issues/614, which has been moved to a future milestone. This story's size now roughly reflects the total effort employed.

Closing because everything has been done here.