HCLS/Banff2007/Informal F2F

Banff2007 Informal F2F

~20 participants

These items are based on notes from the Informal Meeting at WWW2007 May 17

The topics raised at the informal gathering of 20 people inlcude:

• converging/coordinating RDFizer activities
• syntax of URI's: http://auth.org/db/id , http://purl.org/auth/db/id , http://someOrg.org/hcls?urn:lsid:auth:db:id
• who should have authority for a given set of URI's?
• explicit contracts of what is constant in a resolvable URI per specific-URI, and what can change over time, e.g., annotations, resource links
• how to retrieve additional formats from a resolvable URI

Some Comments by Participants

Alan- issue of content negotiation; different URI's; need a theory for data records

   - solve data record issue first! return default RDF

Francois - why not adopt the uniprot:XXXX model? (part of Francois' 'Banff Manifesto')

Amit- Glycoproteins with different id's; need id mappers similar to 'same_as'

Mark Wilkinson - Promise always to keep data concept the same

1. what is named by the URI
2. namepsace listing
3. Versioning

Alan- Contract space for what stays the same, what can/will change

Chris- Aggregates/Views and atomics

Alan- agree to follow up on these issues

Vote: split on Authority vs Community as the 'owner' of the URI

Additional items

• purl.org + urn -- redirection issues
• identifying what it would take to move toward common data URIs,

including reaching out to some data authorities

• sharing and comparing some core life sciences semantics already

defined in projects (protein, publications, diseases, pathways, etc)