editing already submitted publications

[taylor13]

Many of the publications currently in the CMIP5 list need to be updated (by the authors). I'm wondering whether it would be too difficult to implement an automated system to grant permission to co-authors of a paper to edit the information (rather than having them ask one of us to add their names)? I guess the steps would be:

1. co-author is required to register and logon.
2. co-author finds (on the the view or search page) the title of a publication that needs updating and copies it.
3. co-author goes to the "submit" page and (along with the list of papers submitted recently by him/her) finds a box to paste the title of an old publication that needs updating. The box could be labeled "enter title of existing publication that you, as a co-author, request editing privileges"
4. A script is initiated that locates the existing publication and checks that the user requesting edit privileges has a last name that is included in the list of authors of the publication. Once verified, the user name becomes associated with the existing publication, and co-author is informed that edit privileges have been granted. If user is not a co-author, a message is returned indicating that the user is not listed as a co-author and may not edit the the publication.

If permission is not granted, the user should be advised to email the administrators.

How difficult would this be?

thanks,
Karl

[mattben]

This is doable. Since this site is for our users it seems they are trust worth. We could have a search box on the edit page for Title. If the current users last name matches any of the last names of the Authors of that publication then we will permit editing. I see a few possible issues.

1. The Authors last name was mis spelled which would not allow the correct user to make edits.
2. A malicious user could create a dummy account with the last name of the author they wish to cause havoc to. Thus giving them access to edit anyones paper with the basic understanding of creating accounts based of the last name chosen.
3. A generic last name like 'Woo' would give access to all Woo's wether or not its the correct author and again we would have to put trust in the user to only make edits to their actual publications.

I don't see any venerability's or security risks with this, and in short it is doable as long as you find the reward greater than the risks. I also agree that this is the easiest must straight forward solution to allow the community to make changes, with out making us (the admins) do all the work.