At Fri, 15 Jan 2010 20:49:04 +0100, Niklas Broberg wrote:
> > Allow (registered?) users to submit links to papers that are missing? > > Yeah, I've certainly been thinking about that. > First, I will somehow have to overcome being a complete control freak. > > The very first thought that crosses my mind is > "how will I ensure that author names are well normalized?" > > Sad, isn't it?
Not at all. It takes a compulsive control freak to assemble a good bibliography. Besides, I *want* the names to be well normalized. (If you'd seen the BibTeX entries that I get from my coauthors, you would know why...)
+1
However, I still think you should let users submit links. There are really two different issues here, one is to suggest a completely new paper, another is to provide a link to an already listed paper where there's currently no link provided. A lot of the papers only have a DOI link currently. Submitting such links won't need any name normalization, as the paper is already listed.
For papers not already in your database, maybe have a two-tier system where not-yet-checked entries are marked as such when they are shown?
Providing bibtex entries would be nice as well.
The more of these features are added, the closer it gets to duplicating citeulike.org IMO and you'd be better off making use of a citeulike group. I haven't used the groups feature much but there are several haskell/fp related groups on there , e.g. http://www.citeulike.org/group/4254 . I think there are also a great deal more papers etc on there than are represented in groups (e.g. there are more FP/Haskell papers in my own 'library' than in the group I linked to). Regards, Jim
Cheers,
/Niklas _______________________________________________ Haskell mailing list Haskell@haskell.org http://www.haskell.org/mailman/listinfo/haskell