Re: [Bug-gnupedia]Broad Technical Issues

[Dries van Oosten]

On Wed, 17 Jan 2001, Simon Cross wrote:

> Secondly, I'd like to raise a hand in support of the peer review for
> articles.  This is what is going to make the encyclopedia work.  I suggest
> that the search engine be able to search for articles based on who has
> reviewed them.  For instance I might want to search for articles on
> "String Theory" which have be reviewed by "Rovelli".  We might need some
> categories for reviewers to place the articles in (e.g. "Good", "Ok",
> "Bad").  Or at least a consistent rating system for the search engine to
> work with.  We also need to build peer review into the specification of
> the article format.  Since authors will have control over which reviews
> they tag on to their work the negative categories I mentioned above should
> be dumped.  No one is going to say, "Joe Bloggs said my article has as
> much factual content as a chewing gum wrapper".  I agree that reviewers
> need to be able to sign the particular version of an article that they
> review.  Provision for the signing also needs to made in the article
> format.  Readers also need some way of checking that the signature is
> legitimate.

I agree totally. Peer review is a must. If there is no peer review, you
also end up with a new world wide web (to use your phrase). You have to
have some idea of how much value you have to attach to an article.
What you don't want is that articles can be rejected on the basis of the
opinions of a few reviewers. If you submit an article to Phys Rev A, it
can be denied for publication by ONE peer reviewer and this guy might be
your direct competitor, trying to frustrate your work.
So we don't want that. Your alternative is great. Ask a few suspected
experts from the encyclopedia contributors and ask them to give their
opinion. If everyone thinks it sucks, but it should still be published,
albeit with some bad credentials. One of the problems with peer review is
that if you have a small amount of people willing to peer review, they
tend to get a lot of work. And when a peer reviewer always responds fast,
he'll get more work the next time. Avoiding these kind of logistic problem
can become a big issue if this thing really gets of the ground.

Groeten,
Dries