I used to be a scientist (well, I still am, but not being associated
with a university anymore.) And my first reaction I had when I was
introduced to HTML and WWW was "now, that is what the scientific world
needs." And it wasn't for the graphics, but the ability to link documents,
and retrieve it via the network. I have cursed out library many times
for having tons of journal, but the one I need. (Oh, yes, we have them,
all the issue starting from 1980. But I need the Dec 1979 one!) And
then you had to wait for a week or two so the article you wanted could
be ordered from some other library. It works, you get the article, but
it is slow, and simple browsing to see if it is interesting isn't
possible. Worse are references "Unpublished manuscript" or
"Journal, Volume [next year]". If you put articles on your own website,
the refereeing process is less of a problem.
Having said that, I seriously doubt if HTML is the right tool for
publishing scientific articles, at least in the field I worked in
(Computing Science/Mathematics). I am not sure if the current math
proposal is sufficient enough, and more important, precise presentation
*is* sometimes important, for instance if you want to use the same
notation as in the "common" literature. (That is, if my work builds
on the work of person X, I want to use the same notation as X.)
A lot is already published in (La)TeX, and that a dvi or postscript
file is as easily downloaded as an HTML file. Sure, it might take
5 minutes, but scientific articles aren't read as quick as the average
HTML page. What's 5 minutes if it's going to take a day or two to read
the article? HTML can just be used as a "frontend" to real documents.
You use HTML to navigate to the site which has the document you need,
then download it, and view it in a dviviewer, ghostview or print it
out on a laser printer.
++ Big archive and mirror sites such as Info-Mac, are working on the basis
++ of collaborative feeding of central servers, can we imagine to set up
++ something equivalent for scientific publications ? Can we hope that the W3C
++ consortium will focuse on scientific usage, for publication and knowledge
++ diffusion, of the Web now that its commercial usage is in good shape :-) ?
++ Why not opening the W3C to pool of scientists (such as Rzepa's ChemMIME
++ group of researchers) ?
That would be real cool and valuable.