help: add information on blind peer review recipe

[krzysztof]

This should be made part of the FAQ as it comes up often in questions:

Confidential peer-review is possible on Zenodo, and some of our
users already use Zenodo for double-blind peer-review.

To give access to your data for peer-review to a reviewer, you will
need to upload your data under "Restricted Access" conditions.
After that, the access to your record will be given based on a generated
secret URL, which will give "anonymous" access to anyone accessing it through this link. The URL should be given only to the
reviewers and never become public. The URL contains an access
token, and it's safe (in terms of confidentiality) to use - it does
not disclose the identity of the person accessing the record.

To get a secret link your reviewer has to open the Zenodo record
and request for access to it through a request form (in this case
he will have to provide his email and reason for access). You will
be informed of this by email and you will have to grant the access
to this person. The reviewer does not need to create a Zenodo
account for this, as he is providing his email in the request - the
secret URL will be mailed to him once you grant him the access.

*However, this WILL disclose the reviewer's email to you*

If you intend on having a double-blind peer-review, you will have
to request for the access to your record *yourself*, i.e. request
for access anonymously (you have to be logged-out, but you can use
the same email you use on Zenodo), grant the access to yourself and
receive an URL to your own record on your email, which you can
later give to the reviewers.

[floledermann]

FWIW I don't think this approach works any more, since the author will still be disclosed at the download page. Another approach that should work is described here: https://ineed.coffee/post/how-to-disclose-data-for-double-blind-review-and-make-it-archived-open-data-upon-acceptance

[nkdnnlr]

I think this still works for hiding the person that downloads the directory, which sometimes (e.g. single-blind review) can be quite useful.