Registry Iteration 02 Metadata
- Note: comments on style (if you have comments or suggestions, please feel free to edit. - Elizabeth Auden):
-
- If the element name or content includes a science or computing term that's supposed to be in all caps, put it in all caps (ie, SURF, HTML, ICRS).
- If the element name is a single word not normally capitalized, put it in all lowercase. (publisher, email, id)
- If the element name is made up of two or more words stuck together, let the first word be all lowercase and the remaining words start with a capital letter, or, if they are usually in all caps, then keep them in all caps.(serviceStandardURL, coverageTemporal, subjectList).
- QUESTION: what if the element name is multiple concatenated words, and the first one should be in all caps (for example, XMLQuery or HTTPResult)? My preference would be to put the word in lowercase (xmlQuery / httpResult), but this kind of contradicts what I've just said.
- I think most of Bob's element names start with capital letters, but so far on the Astrogrid schemas we've had element names starting with lowercase letters - let's stick to that (unless anyone objects strenuously).
--
ElizabethAuden - 12 June 2003
Material below dates from May, see more recent (June) implimentation of this in RegistryIt02Schema and RegistryUnits etc. and applications to SURF USNO-B and other archives. The Iteration 02 Schemas use the worst of both worlds e.g. serviceHttpResults but this can easily be changed once we come to an consensus in answer to Elizabeth's question above.
*
Resource Registry: latest but out of date plain text explanation
--
AnitaRichards - 16 Jun 2003
Some of these terms are the same concept but with a different name / place in the hierarchy. A "k" after the term indicates that the term comes from Keith's registry schema, and a "b" indicates a term from Bob Hanisch's document. Which terms shall we keep (and be prepared to fill in!) for iteration 2, which ones should we axe, and are there any we should add?
IDENTITY
- id k, b
- title k, b
- serviceType k
- ticker b
- identifier b
CURATION
- publisher k, b
- publisherID b
- creator k, b
- subjectList k
- description k, b
- contributor k, b
- date k, b
- comment b
- version k, b
- serviceURL k, b (called referenceURL by Bob)
- contact k, b
- contactname k, b
- contactemail k, b
CONTENT
- facility k, b
- instrument k, b
- format k, b
- rights b
- spatialresoluction k
- briefsummary k
- type k, b
- archive k,b
- survey k, b
- catalogue k, b
- bibliography k
- journal k, b
- library k, b
- outreach k, b
- education k, b
- eporesource k, b
- integrated k
- nameresolver k
- simulation b
- animation b
- artwork b
- background b
- basicdata b
- historical b
- photographic b
- press b
- wavelengthrange k
- gammaray k
- xray k
- xuv k
- uv k
- optical k
- ir k
- mmwave k
- radio k
- wavelength k
- wavelengthshort k
- wavelengthlong k
- coverage k, b
- regionname b
- cframe b
- ICRS b
- FK5 b
- FK4 b
- ECL b
- GAL b
- SGAL b
- regionofregard b
- coveragespectral b
- radio b
- infrared b
- optical b
- uv b
- euv b
- xray b
- gammaray b
- coveragespectralbandpass b - U, V, R, I, B, etc.
- coveragespectralcentralwavelength b
- coveragespectralminimumwavelength b
- coveragespectralmaximumwavelength b
- coveragetemporalstarttime b
- coveragetemporalstoptime b
- coveragedepth b
- coverageobjectdensity b
- coverageobjectcount b
- startdate k
- enddate k
- decmin k
- decmax k
- sensitivity k
- ramin k
- ramax k
_____________________
Elizabeth Auden, MSSL
_____________________
New whittling efforts, taking in Anita's suggestions:
IDENTITY
- title k, b
- ticker b
- identifier b (SEE NOTE [0])
CURATION
- publisher k, b
- publisherID b
- creator k, b
- subjectList k
- description k, b
- contributor k, b
- date k, b
- version k, b
- referenceURL
- contact k, b
- contactName k, b
- contactEmail k, b
- dataSize
- nBytes
- nPixels
- nRows
- nCols
CONTENT
- componentTablesList
- ucdList
- facility k, b
- instrument k, b
- format k, b
- rights b
- resourceType (SEE NOTE [1])
- archive k,b
- survey k, b
- catalogue k, b
- bibliography k
- journal k, b
- library k, b
- outreach k, b
- education k, b
- epoResource k, b
- integrated k
- nameResolver k
- simulation b
- animation b
- artwork b
- background b
- basicData b
- historical b
- photographic b
- press b
- coverage k, b
- coverageSpatial a
- regionSpatial b (SEE NOTE [2]) a
- box b
- cframe: ICRS, FK5, FK4, ECL, GAL, SGAL
- coordinateAMin
- coordinateBMin
- coordinateAMax
- coordinateBMax
- circle b
- cframe: ICRS, FK5, FK4, ECL, GAL, SGAL
- coordinateACen
- coordinateBCen
- radius
- polygon b
- cframe: ICRS, FK5, FK4, ECL, GAL, SGAL
- coordinateA1
- coordinateB1
- coordinateA2
- coordinateB2
- coordinateA3
- coordinateB3 ...
- cframe b (SEE NOTE [3])
- ICRS b
- FK5 b
- FK4 b
- ECL b
- GAL b
- SGAL b
- regionOfRegard b
- coverageSpectral b (SEE NOTE [4])
- coverageSpectralWaveband a
- radio b
- mmWave
- infrared b
- optical b
- uv b
- euv b
- xRay b
- gammaRay b
- coverageSpectralBandpass b - U, V, R, I, B, etc. (SEE NOTE [5])
- coverageSpectralCentralWavelength b
- coverageSpectralMinimumWavelength b
- coverageSpectralMaximumWavelength b (SEE NOTE [6])
- coverageTemporal a
- coverageTemporalStartTime b
- coverageTemporalStopTime b
- coverageDepth b
- coverageObjectDensity b
- coverageObjectCount b
- resolution a
- resolutionspatial a
- resolutionspectral a
- resolutiontemporal a
- dataQuality a
- photometryerror a
- astrometryerror a
- timingerror a
NOTES:
[0] I think id and identifier were the same thing? a
[1] I think that the specific resource types would work better as elements of a "choice" under resourceType than they would as separate xml elements or database table column names.
[2] Box, circle and polygon should be elements of a choice under regionName. regionName change to regionSpatial for clarity a
[3] The cframe types should be elements of a choice under cframe.
[4] These wavelengths should be a choice under coverageSpectral.
[5] These bandpasses should be a choice under coverageSpectralBandpass.
[6] Eventually these could be in regionSpectral like regionSpatial? And similarly for temporal coverage - ie not restricted to one chunk, specified as an array? a
a means elements/comments added in this iteration by amsr.
amsr assumptions - please correct if wrong:
For this iteration at least:
Units are to be standardised - flux in Jy, wavelength-like in metres, position in decimal degrees, date/time in ISO8601 format YYYY-MMM-DD etc
Null values as recognised by xml are allowed
All fields except the identity metadata can be null or unbounded
The dataset specified in ID is one of the componentTable if there are more than one associated tables
--
AnitaRichards - 29 May 2003
--
ElizabethAuden - 12 June 2003