[Bug 484] - eml-attribute changes needed

[bugzilla-daemon@ecoinformatics.org]

http://bugzilla.ecoinformatics.org/show_bug.cgi?id=484





------- Additional Comments From peter.mccartney at asu.edu  2002-09-03 17:17 -------
In responding to a request from Chad to provide a sample metadata file from our 
Xylographa tool, im finding a number of stumbling blocks as i go through 
filling in the blanks to make the file valid. Many of these are in attribute 
and result from the fact that unit, measurementScale and attributeDomain are 
all mandatory. This is forcing me to put in nonsense information for some 
attributes (in fact, in order to get my sample file to validate a couple weeks 
ago, this was exactly what Matt put in for these fields - nonsense). I dont 
see "dimensionless" as a choice under unit, although its valid if i just make 
the element but leave it blank. I can answer nominal for measurement scale, but 
how many people are going to recognize this as the right choice for a field 
they feel doesnt "measure" anything? finally, many attributes dont have a 
domain other than the general topic that the information is supposed to be 
about. I could fill in a textDomain entry that simply says "any text" but i 
would regard someone that did this with the same contempt i do people that 
put "N/A" in spread sheet cells. I think all of these need to be optional or we 
come up with an easier and meaningful way to generate default entries. If there 
really is no domain information, id rather the element not be there. same with 
unit and scale.

im also drawing blanks for things i should be able to put in unit - dates and 
times, photosynthetically active radiation (umol/m2/s), pressure in 
kiloPascals, and of course lots of ratios. we're going to need to have a really 
rich list as well as some primers on how to fill in custom unit definitions.
,