Just when you think it is safe to go back in the water ...
I just wrote about the migration of journals from print to digital and that one of the major problems with this move is the issue of really long term archiving and preservation of the content. JSTOR has been a real bright spot in the midst of this whole issue, digitizing complete runs of leading journals and licensing access to libraries. But doing that in such a way that libraries felt that we could rely on those back issues being around for a long time.
Now the AAAS's Science -- the grand old man of American scientific scholarly publishing -- has decided to withdraw from JSTOR because it fits its business model to drive traffic to its own site, and only its site. Read about it here.
In the history of journal publishing (in fact publishing in general) publishers have not taken the role of archiving their content seriously. Old issues or copies are pulped and print runs are short (but copyright terms aren't) when it makes business sense. They have left preservation of back issues to libraries. Now we are supposed to believe that in the digital world they will take it seriously, when digital preservation is a far more active and therefore expensive process that print preservation.
AAAS will probably do a fine job of preserving their content. But they are setting a lousy example and it is completely unnecessary. I understand their agreement with JSTOR was non-exclusive.