I was only able to attend a morning panel today but it was excellent: Futures of Scholarly Editing II, chaired by T.H. Howard-Hill. The papers synthesized and clarified a number of issues I've been pondering here and there.
Andrew Stauffer layed out the issues that must be tangled with in the future of digital editing. He began by identifying two aspects of the term edition that apply to print. An edition is (1) the textual product of the act of editing as well as (2) a print run of a limited number of copies of a book. In digital editions aspect (2) is lost, which leads to potentially dire consequences.
A limited print run of a book constitutes a relatively stable, known set of errors and variations. Since a digital edition can be silently emended on the fly, changes and corrections are no longer recorded. Errors and variations, which are important markers of the reception history of a work, disappear in the digital realm and prevent textual scholars from knowing the genealogy of a text. This means that certain questions we're used to will no longer be answerable in the future.
I don't think this assessment is completely accurate. After all, wikis automatically save version histories of all changes, so it shouldn't be difficult to apply that principle rigorously in a digital edition. Besides, Karsten Kynde and Kim Ravn of the Soren Kierkegaard Research Center demonstrated a functionality in their digital archive of Kierkegaard's papers that allows readers to see the version history of particular elements. The interface did not seem very intuitive (granted, I did not have a chance to play with it), but at least it allows access to the history of mistakes or to previous versions of the archive. That is helpful, of course, if you quote from it in a certain way and then need to go back to it later after it's been revised. He made the very nice point that we owe future scholars our mistakes.
Andrew ended with a discussion of digital preservation that led naturally to the next talk, by Barbara Bordalejo. Andrew said that it is not possible to preserve all digital editions, archives, and other works. Since they'd have to be backed up on tape or redundant hard drives, access becomes a serious issue. This becomes worse as technologies become outdated and won't be able to red the auto-archived material, meaning that we'd have to be selective about what gets archived. Much would be lost.
Barbara's talk focused largely on preservation, too. She made the very good point that digital formats will eventually become outdated and unreadable by newer machines, so the old machines themselves need to be preserved. This will lead to a situation in which a rare digital book will become even less accessible than the rarest print book.
Peter Robinson gave a fascinating demonstration of an archive he's developing with people at U Birmingham, Münster, St. Catherine's Monastery in Egypt, and the British Museum. They're digitizing all the leaves of the Codex Sinaiticus MS. I don't have time to describe it in full, but they've made a beautiful AJAX-y interface that accounts for correction histories and does a number of wonderful things. Will write more on this later.