If you put a paper on the web, please consider that it will be
downloaded over the net, printed out, and read by people. You do them a
favor if you
(a) provide also an abstract (in HTML), so that poeple can
get a quick summary of your paper,
(b) provide also compressed versions
of your paper to accelerate downloading,
(c) try to keep the overall
size of the paper within reasonable limits, e.g. by reducing the resolution
of images,
(d) maybe provide also a version where text and figures are
split in different files, if the figures are large
(e) include figures
and tables in the text, instead of appending them at the end,
(f)
choose 10pt font size and single spacing to save paper, and
(g) choose
the textheight and -width such that it can be printed on letter-format as
well as on A4-format.
Another important issue is that of copyrights. Several journals now ask you to sign copyright transfer forms in which you explicitely give up the right to make the paper available in electronic form. Many authors, on the other hand, ignore that and put their paper on the web anyways. I prefer to be legal and found so far that journals are often willing to compromise on this point if you insist on your right to put your paper on the web (before you sign the transfer form). One possible compromise you can offer after a while of negotiation is, for instance, that you put a draft version on the net for the time being and one year after publication you are allowed to put the final version on the net. Try it, or send your paper to journals that don't restrict you in that way, e.g. those by IEEE.