Comment by armitron
2 months ago
DJB is not being "selective", instead you're nitpicking on something tangential to the point he's trying to make and creating the impression that what you wrote detracts from his point when that's definitely not the case.
His point in this section is that people are incorrectly or unreasonably claiming that PQC adoption in hard.
He uses Cloudflare's 33% number as counterevidence against it being hard, and I'm pointing out how this is a selective argument: it was "easy" because nobody, Cloudflare included, considers just PQ key exchange to be a full adoption of PQC. The full adoption requires asymmetric PQC and a corresponding re-imagining of how to encapsulate much larger public keys over the Internet, which is decidedly not easy.
(I don't have opinions on whether QC "won't work" or not; I think it's good to pursue PQC regardless, since we should have more "hard" primitives that don't decompose to the hidden subgroup problem.)