dmozzherin at eol.org
Wed Sep 16 12:17:40 CDT 2009
Name strings mirror exact strings supplied by databases (except
multi-spaces). And yes, some changes are trivial (and easy to hide).
Current idea is to keep them as they are so whoever has a name can
also compare it with all other renderings.
Some names are not even names, but they were submitted by someone as
scientific name category and are connected to some real data. So we
keep them, but plan to flag more and more of them in the future.
Indexing in this case presumes there is a link to database, but point
to some data.
> Well, it is indexing something, but what it indexes is not information,
> just text strings (in many cases not much better than randomly generated).
> In the act of indexing it promotes what it is indexing to become 'data',
> and in that sense, and a very real one it is, it is creating something that
> was not there before; these 'data' are then presented as "18,323,094 names".
> For a project that is presented as "GNI is meant to collect all Scientific Names"
> it pays remarkably little attention to scientific names (and their structure).
More information about the Taxacom