pcplantdb

[Chad Knepp <pyg@galatea.org>]

Richard Morris writes:
 > Hi All,
 >      Afew thoughts on schema, specifically how we divide up
 > the information about plants. Sort of inspired by ideas on Folksonomies,
 > and a piece I read discussing informal/foraml specs.
 > 
 > Basically we can divide info into two main groups
 > 
 > - those for which there is a well established formal/informal spec
 >      e.g. botanical names
 > 
 > - those for which there is no good spec
 >      most of what we have
 > 
 > How we organise data depends on which group things lie in.
 > 
 > Formal entries: in just makes sense to use a scheme close to formal 
 > spec, or easily convertable to formal spec.
 > 
 > Informal entries: use a folksomies structure where users can create 
 > their own headings/catagories. We do not impose our own incomplete 
 > ontologies (clasification system) onto these.
 > 
 > Stuff to go in formal spec: (just the spec for these not the actual data)
 >      Botanical names
 >      Botanical Families
 > 
 >      Common Names - whilsts informal, this is such a recuring theme
 > 
 >      Locilisation - Countries/Country codes/Lat-Long
 >              Very accepted spec for country code.
 > 
 >      Book references - generally accepted schemes for citation
 >              Author/Title/Publisher/Date/etc.
 > 
 >      Web references - URL - Title
 > 
 > Informal elements
 >      Plant uses - there can be a great many diferent plant uses
 >              do we wish to restrict our users to just those, we put  
 >              in or do we allow users to create their own?
 > 
 >      Habitats - Also a great many
 > 
 >      Division of page into sections - whilst we have some good section 
 > definitions (Edible Uses/Medicinal Uses/Cultivation details/...)
 > I can see a need where many other setion might be required.
 > For example a user might want to divide cultivation/propagation into
 >      stratification - stimulation of seeds
 >      pruning -
 >      grafting
 >      ...
 > Also other major headings
 >      history
 >      folklaw
 >      where to get seeds
 > 
 > I don't think its wise to restrict the type of sections allowable.

Absolutely, I think we all more or less agreed to an *open vocabulary*
in regards to this type of info.

 > Other things I'm not sure how to place
 >      Mainly plant properties such as height, flowering times
 >      Probably in the formal spec bit.
 >      Could potentually allow users to add a property.
 > 
 > Thoughts?

This is approximately parallel to my own thinking.  I've sort of
pictured it as a tree structure with the root being the name[s]
object.  Everything else attaches to this and becomes more loosely
structured as it branches away from the trunk, eventually becoming
very free form.

-- 
Chad Knepp
python -c 'import base64;print base64.decodestring("cHlnQGdhbGF0ZWEub3Jn")'