69 lines
4.9 KiB
Plaintext
69 lines
4.9 KiB
Plaintext
|
%META:TOPICINFO{author="LeeBelbin" date="1258685123" format="1.1" reprev="1.10" version="1.10"}%
|
||
|
|
%META:TOPICPARENT{name="ClosedTopicSchemaDiscussionSDD09"}%
|
||
|
|
---+!! %TOPIC%
|
||
|
|
|
||
|
|
Main.KevinThiele 25 Sep 2003: we currently have:
|
||
|
|
<verbatim>
|
||
|
|
<Resources>
|
||
|
|
<TaxonNames>
|
||
|
|
<TaxonName key="Viola_hederacea">
|
||
|
|
<FreeFormDescription>Viola hederacea Labill</FreeFormDescription>
|
||
|
|
</TaxonName>
|
||
|
|
<TaxonName key="Viola_someothercea">
|
||
|
|
<FreeFormDescription>Viola someothercea Labill</FreeFormDescription>
|
||
|
|
</TaxonName>
|
||
|
|
</TaxonNames>
|
||
|
|
</Resources>
|
||
|
|
</verbatim>
|
||
|
|
What's meant by <nop>FreeFormDescription? We need to split the epithet and author in the name e.g.
|
||
|
|
<verbatim>
|
||
|
|
<Resources>
|
||
|
|
<Taxa>
|
||
|
|
<Taxon key="Viola_hederacea">
|
||
|
|
<TaxonName>Viola hederacea</Name>
|
||
|
|
<TaxonAuthor>Labill.</TaxonAuthor>
|
||
|
|
</Taxon>
|
||
|
|
....etc
|
||
|
|
</Taxa>
|
||
|
|
</Resources>
|
||
|
|
</verbatim>
|
||
|
|
is better?
|
||
|
|
---
|
||
|
|
Main.BobMorris - 26 Sep 2003: Ah, this arises from our desire to recognize that (a)Resources of all kinds may often have external definitions (e.g. for a <nop>TaxonName, it might be a URI designating an authority for the name) and (b)when elments have a lot of commonality, they should be required to be declared as being of a named datatype. In this case the draft Schema calls for <nop>TaxonName, Specimen, Publication, and maybe others TBD, to be of type <nop>ResourceConnectorType, which is a datatype that provides for two kinds of external reference (details irrelevant for this discussion) and a bit of text meant mainly to be passed on to humans. That's the text enclosed in the tag <nop>FreeFormText. The tag name is chosen to be mnemonic I think.
|
||
|
|
|
||
|
|
However, your point is that a <nop>TaxonName deserves more structure--and maybe even should make the <nop>FreeFormText either optional or prohibited, requiring instead more structure to the name. This can be done pretty easily by making a type derived from <nop>ResourceConnectorType in a way that would support your alternative. I'll modify my guerrilla-action Schema to do that and edit this page when done. As you observed below, it remains for bioligists to discuss---perhaps after the Lisbon Taxonomic Names session---exactly what should be the structure of a <nop>TaxonomicName.
|
||
|
|
---
|
||
|
|
Main.KevinThiele 25 Sep 2003: We will need to conform with TDWG Taxon Names eventually, but they will surely have something like this.
|
||
|
|
I've used <Taxa> and <Taxon> rather than <TaxonNames> and <TaxonName> because surely we will later add further resources to the taxa, and name is only one part.
|
||
|
|
|
||
|
|
Having <Taxon> as a child of <Resources> and <Taxon> also as a child of <Descriptions> won't cause problems, will it?
|
||
|
|
---
|
||
|
|
Main.BobMorris - 26 Sep 2003: No it won't. _But_ the idea here is that a _<nop>TaxonName_ is a resource to be used elsewhere---perhaps in many places---whereas a Taxon is one of the things that can be described. So while it won't make any difference to a parser examining an instance document, it could confuse any discussion that omits the context. This is a general ProblemOfResolvingIdentifiers which in my opinion is best sidestepped by using two different identifiers in this case.
|
||
|
|
|
||
|
|
BTW, I doubt we do (can?) enforce in the Descrptions section, that there is at most one Taxon referring to any given <nop>TaxonName (i.e. loosely, only one description of any given Taxon). If that's right, we need a specification of which one counts or discuss whether it's OK.
|
||
|
|
---
|
||
|
|
Main.KevinThiele 25 Sep 2003: Also, is it sensible for the taxon list to be part of <Resources> but the characters list to be part of <Terminology>. It seems to me that they are pretty equivalent things. Maybe we should have 4 base elements
|
||
|
|
<verbatim>
|
||
|
|
* <Features> - the character list
|
||
|
|
* <Taxa> - the taxon list
|
||
|
|
* <Descriptions> -the descriptions
|
||
|
|
* <Resources> - associated resources e.g. images, references etc.
|
||
|
|
</verbatim>
|
||
|
|
That way, <Resources> can be optional but <Taxa> required - clearly, it's important to be able to query a project for its taxon list.
|
||
|
|
---
|
||
|
|
Main.BobMorris - 26 Sep 2003: Well, that might be OK if Taxa were the only things ever described. But we intend others, such as Specimens.
|
||
|
|
Now, aside from the "minor" elements like <nop>ProjectDefinition, the current situation is only _three_ things:
|
||
|
|
<verbatim>
|
||
|
|
* <Terminology> - which includes both what you are here calling Features and Taxa
|
||
|
|
* <Descriptions>
|
||
|
|
* <Resources>
|
||
|
|
</verbatim>
|
||
|
|
(Here you mean Taxa you mean <nop>TaxonNames ...)
|
||
|
|
But Terminology includes much more than just <nop>TaxonNames. Notably it presently includes <nop>ModifierDefinitions and a few other global things involved in the specification of states.
|
||
|
|
---
|
||
|
|
Main.BobMorris - 26 Sep 2003: Hey wait a minute. Just when I thought that after all these years I'd finally understood the difference between a Taxon and a Taxon name, now you are confusing me again...
|
||
|
|
|
||
|
|
---
|
||
|
|
|
||
|
|
%META:TOPICMOVED{by="GregorHagedorn" date="1085765904" from="SDD.TaxonNamesInResources" to="SDD.ClosedTopicTaxonNamesInResources"}%
|