Editing EPrints
There are various reasons why you may want to edit an existing eprint. Full text may need to be added when it becomes available, alterations may need to be made to the full text when reviewers corrections are factored in, metadata may need to added when the publication details are provided by the publishers or errors may need to be corrected in the metadata.
Three Ways to Edit An EPrint
Editing an eprint follows exactly the same process as adding a new eprint. However, before you can make corrections to an eprint record or add full text to it, you need to understand the role of the workspace and buffer in an EPrint archive.
The most important rule is that eprints can only be edited when they are in a user's workspace. That means that to update an eprint, the user must either
- get someone with editor privileges to move the original eprint into their workspace and edit it in situ
- or create a new copy of the original eprint and edit that instead.
Of course, your institution may have abandoned editors and disabled the buffer. In that case, you will be able to directly edit eprints in the public archive. Such privilege brings great responsibility - be very careful!
The advantage of the first approach is that it is direct and simple. The disadvantage is that it leaves no record of the previous state of affairs — the archive ceases to be a genuine mechanism for stable, long-term storage. It may have repercussions for external users whose Web pages link to a particular record, as they may find its information changes without warning. For example, an unpublished technical report may suddenly become a journal article with an updated title.
The disadvantage of the first approach is that you may have to wait some days for the editor to respond to the request.
The advantage of the second approach is that it maintains the records in the database, and allows external users to see that a particular journal article is a later version of a previous technical report. The disadvantage is that it potentially leaves multiple records per publication, both of which will show up in a search.
The method that you use depends on the results that you want for external users. You are advised to use the first method only for small revisions of or additions to the metadata. The second method is much better for major revisions to the article itself. You should also consider whether the original record should be deleted from the archive to avoid it showing up as an apparent duplicate in search results.
The easiest way to copy an eprint is to clone it. (You can just create a new blank eprint and type in all the data from scratch, but this is a last resort.) You can either clone an eprint which is already in your workspace, or from your User Homepage, you can click on the "Review Your Documents in the Archive" button and clone any eprint from that list. The cloning process creates a blank eprint, copies all the original's metadata, and marks the new eprint as a successor to the old one.
Having cloned an eprint, you must then decide whether the original is still valid. This might be the case if the article has undergone very significant revisions (from a conference presentation to a journal paper, for example). If the edit represents only a change in the metadata, the original eprint should probably be deleted. Although you can delete eprints in your workspace, deletions in the archive can only be performed by a user with editor privileges, so to have the original record deleted by the editor click on the "Request Removal" button next to the "Clone" button in the "Review your Documents in the Archive" list.
Finally, note that you can only clone eprints which you deposited in the archive in the first place. If you need to update someone else's eprint (you may even be an author of the paper in question) you will need to create a fresh eprint and manually copy all the information across, ensuring that you mark the new eprint as a successor to the original in Step 2 of the addition process.