|
19 August 2004 Teleconference
Present:
Jessie Kennedy, Robert Kukla, Trevor Paterson, Dave Thau, Bob Peet, Xianhua Liu, Aimee
Stewart, Robert Gales, Joana Trajkova
Agenda: A response to the following points put forth in an email from Robert and Aimee. Most other points raised in the last 2 months have been agreed upon or tabled in emails.
- KU Email
- Given your concession to allow any type of relationship outside of a TaxonConcept, I think we can make our system work with your schema. We are thinking of this as purely a vehicle for transfering data, but I get the feeling that there are some assumptions that go hand in hand with it - if so, could you make those available?
- Teleconference
- JK agreed that yes, Napier is considering that there are some assumptions/rules on how different groups (i.e. SEEK) would interpret the data from the schema. They will document these assumptions for later discussion.
- KU Email
- The Assertion element name should be changed since it is already established as something different in Taxonomer. Perhaps something less ambiguous - Relationship, Link ...?
- Teleconference
- Napier agrees that this is a problematic element name, but would prefer to address this issue after more discussion at the TDWG conference in October.
- KU Email
- Metadata (and sub elements Creator and CreationTime) should be required
- Teleconference
- Fine with Napier.
- KU Email
- Define elements before they are referenced - so put Vouchers and Publications right after Metadata.
- Teleconference
- Fine with Napier.
- KU Email
- Add Repositories (before Vouchers) and AccordingTos (after Publications) and allow both of them to be referenced.
- Teleconference
- KU does not see these as elements receiving GUIDs, only elements that may be reused in a document. Repositories sound fine to Napier, but they are not convinced that AccordingTos are an element that will be reused much. KU would like AccordingTos, but does not see it as a key issue. We agreed to revisit this after more reaction at TDWG.
- KU Email
- Add the ability to transfer an existing TaxonConcept by its GUID. The GUID would be sufficient, no data would be allowed, to prevent a GUID being sent with incorrect data.
- Teleconference
- Napier explained that the GUID of an existing TaxonConcept (in the TaxonConcepts element) can be transfered in the TaxonConcept id attribute with the type attribute equal to "empty". KU now understands the thinking behind this, though would prefer for GUIDs in this element to be treated similarly to the GUIDs in the FromTaxonConcepts and ToTaxonConcepts elsewhere in the schema. There may not be an elegant way to prevent the transfer of a GUID with incompatible data, so we will table this disagreement for the time being.
- KU Email
- Add the ability to transfer an existing "Assertion" by its GUID alone.
- Teleconference
- KU explained that this would allow the transfer of GUID-labeled relationships between data providers, just as a provider might wish to transfer a single GUID-labeled taxonConcept. Napier suggested that this could be done by putting the GUID in the Assertion id attribute, but acknowledged that the current solution is not ideal because there is no "type=empty" attribute that would indicate this element is a GUID and needs no further data.
Additional Discussion
Robert Kukla offered to create an updated version of the schema, with the agreed upon modifications for internal use by SEEK Taxon. The publicly presented schema will not change for the time being. With the new internal schema, Aimee will return to coding the parser for use in the EnterProviderData API.
Robert Kukla and Jessie will be leaving October 2 for TDWG in New Zealand. They will be presenting the schema as a global solution to the problem of exchanging taxonomic data. They would also like to represent the SEEK viewpoint (at the same conference, but at a separate time) on the definition of a taxonomic concept, and how the SEEK Taxon system will interact with the community.
Jessie will go directly from New Zealand to Kansas, joined by Trevor Paterson, for the SEEK All-hands meeting.
The next conference call is tentatively scheduled for Thursday, 2 September 2004, 15:00 GMT.
|
