Re: [Treebase-devel] dois or dryad identifiers in api?

[Hilmar L. <hl...@ne...>]

Just FYI, CrossRef has stated earlier this year that DOIs should now  
always be presented in their HTTP URI form. While it's perhaps not too  
surprising that publishers have been slow to follow suit, that need  
not mean that we should move at the same pace in following the  
recommendation, in particular as it goes in precisely the right  
direction in terms of making our data more Linked Data compliant.

Ryan pointed out to me in a recent conversation we've had here that  
Dryad DOIs are DataCite DOIs, not CrossRef DOIs, but while DataCite  
apparently so far hasn't taken an official position on whether it is  
in line with this change, I would argue that they have little reason  
not to, and so it would seem fair enough for us to preempt that.

	-hilmar

On Sep 26, 2011, at 10:31 PM, William Piel wrote:

> Hi Carl,
>
> Were you deciding to use our OAI service because it allowed search- 
> by-date?  If so, you might want to know that TreeBASE's PhyloWS API  
> now supports search-by-date, such as:
>
> /study/find? 
> query=prism.creationDate>"2011-08-30T05:00:00Z"&format=rss1
>
> This now works for prism.creationDate, prism.modificationDate, and  
> prism.publicationDate. Use the prism.publicationDate for searching  
> on the publication year of the article -- hence  
> prism.publicationDate>"2011-01-01T05:00:00Z" includes all papers  
> with the year 2011 and later (but a second earlier and it would also  
> include all papers with the year 2010, because in TreeBASE's world,  
> time is EST, which is 5 hours after London).  Use  
> prism.modificationDate to catch older papers that were since modified.
>
> The returned data contains article DOIs like so:
>
> <prism:doi xmlns:prism="http://prismstandard.org/namespaces/1.2/ 
> basic/">10.1006/mpev.2000.0752</prism:doi>
>
> Which presumably can be matched with article DOIs in Dryad.
>
> In our last conversation, I think Ryan said that "For the metadata  
> on the Dryad end, we still need to correct some issues with the  
> relationships to TreeBASE objects."
>
> For the TreeBASE end, we had discussed whether TreeBASE should use  
> "dcterms:source" or "dcterms:isPartOf" (prefixed with http://dx.doi.org/) 
> .
>
> I think we decided on "dcterms:isPartOf" -- I'll check that we're  
> exposing Dryad DOIs like so.  Note that we will be using the "data  
> package" DOI (e.g. "10.5061/dryad.tf48r") rather than the "data set"  
> DOI (e.g. "10.5061/dryad.tf48r/2") because only the former is  
> equivalent to a TreeBASE study.
>
> bp
>
>
>

-- 
===========================================================
: Hilmar Lapp  -:- Durham, NC -:- informatics.nescent.org :
===========================================================