From: "Guédon Jean-Claude" <[log in to unmask]>
Date: Mon, 2 Jul 2018 07:58:09 +0000

If we go back to the basic principles of journal academic publishing
(registration, certification, dissemination, archiving, evaluation),
there is no reason to see all these functions done by one signle
entity - namely, the publisher - in the digital age.

Obviously, the research site (university, research centre) can do this
best as it is closest to the researcher. The institutional repository
can certainly register a piece of work exactly when it is sent for
submission (if an institutional repository is available). The
repository can also store this submission version in a dark archive
(with a request button aimed at the corresponding author). Finally,
rather than focus on the "version of record", the repository could
also document the record of versions.

Jean-Claude Guédon
________________________________________

From: Ann Shumelda Okerson <[log in to unmask]<mailto:[log in to unmask]>>
Date: Sun, 1 Jul 2018 19:32:14 -0400

In the June 28 issue of Nature, follow the link to the letter to the
editor by Michael Keller (Stanford University Librarian) and Stanley
Prusiner (UCSF, Nobel Laureate for his work on prion diseases)
asserting the need for publishers to acknowledge that the first date
of publication in STEM is generally the date of publication of the
e-version, not the subsequent date of the physical, ink-on-paper
version, usually 6 to 12 months later.

This sets straight the confusion occasionally arising on the primacy
of a discovery.

******