[Foundation-l] Another look a bot creation of articles

[Magnus Manske]

On Thu, Jul 24, 2008 at 8:04 AM, Gerard Meijssen
<gerard.meijssen at gmail.com> wrote:
> Hoi,
> You may want to consider the scale of things ... when you are talking
> chemicals, proteins a number like 240 million articles can be expected. With
> such numbers you have to wonder to what extend Wikipedia can cope.

IMHO the point is database vs. free-style text annotation.

It is reasonable to expect that every human gene, in the not-so-long
run, will have loads of text annotation that doesn't fit well in a
classic database; in fact, it will have a few data points and a lot of
text. Remember, we're talking <25.000 genes in human. This is what a
wiki is best at, and pre-creating articles for them that contain the
bare facts is perfectly valid.

OTOH, millions of real/predicted/hypothetical molecules that will, for
the most part, have nothing but a few numbers with them, would fit
better in a "normal" database. That doesn't exclude the possibility of
writing about some of these molecules on wikipedia when there's
something to write about.

Magnus