On Sat, 11 Sep 2004, Eberhard R. Hilf wrote [in part]:
>eh> In addition, commercial publishers do aim at the
>eh> present time to earn money and do not care about the
>eh> future, when they might no longer exist. Some of the
>eh> e-versions of my papers with Wiley are gone after less
>eh> than ten years, because the Publisher bought the
>eh> (indirect daugther) Physikalische Blaetter, but without
>eh> the e-archive.
>eh>
>eh> Of course these published e-Documents of mine are still
>eh> in our Institutional OA archive. Institutional
>eh> self-archiving, together with agreements on mirrors,
>eh> and retrieval, is a safe proposition for long-term
>eh> archiving.
The paper version (if there is one) is also a safe
proposition for long-term archiving. However, if there's no
paper version, then a stable institutional or central
archive becomes crucial. And, if the only version that
survives is a paper version, then (in my limited
experience), it becomes quite difficult for an amateur to
prepare a version (such as a PDF version) that's an exact
copy of the original version, unless the copied version is
prepared entirely as a set of images. My understanding is
that, for text embedded entirely in images, full-text
searching becomes a problem (e.g. for documents in
DSpace-based archives, such as the one recently created for
my home department). See:
https://tspace.library.utoronto.ca/handle/1807/2324
Much better to be able to archive a good-quality electronic
version of the kind that publishers prepare. But, this
solution isn't available from some publishers. Nor for many
older articles, unless the publisher has already prepared
electronic backfiles. Obviously, if the publisher has
already gone out of business, such backfiles are likely to
be missing.
If there's a reasonable chance that a paper might become a
"classic", then a copy of it clearly should be archived, to
meet the needs of scholars in the future (such as
historians), as well as those in the present. Fortunately,
the short-term goals and the longer-term goals require no
difference in behavior. One simply self-archives a
good-quality version of the article, in a stable archive.
Jim Till
University of Toronto
Received on Sat Sep 11 2004 - 15:24:24 BST