Instead of Journals with peer review.
Maybe have a distributed system of trust and file
storage.
Where instead of 'peer review panels', you have public
cryptographically signed endorsement by other scientist.
The system of trust is a bit more top down, in that the
root
trust is university
institutions (or any other root certs you
implicitly decided to trust).
You can decide to endorse a paper, or you can endorse a
scientist.
If you want to submit a new paper for review, you need
to
already have reviewed and endorsed/reject enough
papers.
Every endorsement/rejection can contain a comment by
the scientist on why it was accepted or rejected.
Any paper that was accepted by enough peers will then
have a trail of endorsement/rejection by the scientific
community. The university can then have a choice of
saving or deleting it after a deadline (except for the
historical ledger).
Multiple universities can host other universities papers,
in
exchange for cross endorsing them (Web of trust applies
to
university level).
-----------
Summary:
* University institution (or other trusted institution) can
attest a scientist or professor
* Scientist can provide their attestation to a new
document/scientist, and their attestation is recorded in
a
distributed ledger in various repositories
* Each repositories can then obtain a copy of a journal
based on how well attested it is. E.g. Maybe a elitist
journal repository will only store journals that has been
attested by more than 3 tenured professors from ivy
league universities. While another may decide that only
100 scientist and a single professor from any recognised
universities will be needed to be stored as a local copy.
* To submit the presence of a new journal to the
network, you need to have already attested or reject
enough journals.
* Every attestation/rejection can contain a timestamped
comment by institutions or scientist. (Twitter sized?)
This would provide a way to have a public feedback for
the original author.