Closed rsignell-usgs closed 7 years ago
We have it. It is listed under AOOS however. I don't know the reason behind that, other than the AOOS/Axiom connection. Does that make sense?
Axiom wrote the script, so put it in their WAF. But they should be listed under "NCEP", as these are all forecast models from NCEP.
Unidata acquires this data from the IDD/LDM "NOAAPORT" feed from NCEP, and puts them on thredds so we can access them as a Forecast Model Run Collection (FMRC). Axiom set up the crawler to access only the best time series datasets from the FMRC.
@lukecampbell @dpsnowden We should decide what Organization should own these data in the Catalog and move the WAF in the Registry (https://registry.ioos.us/waf/Unidata/) to the appropriate CKAN Org and harvest configuration.
Either NCEP or Unidata sounds reasonable. Any thoughts?
How are you using the organizations now? Are they intended to mean "this is where the data can be accessed" or "this is the originator of the data"? If it's the former then Unidata sounds fine. If not, then we may need to create two organizations because there are also several Navy (FNMOC) models in that collection.
Hmm, maybe AOOS is still appropriate then, since they are the ones gathering the ISO records?
The pathway these data are getting into the IOOS Catalog are kind of stretching the limits of our concept of data provider a bit.
Is Unidata acquiring the data from NOAA LDM for IOOS specifically, or is this a standard service they provide to re-serve the NCEP model data to users? If that's the case, I'd say using Unidata as the Org makes sense (and it's also the easiest to implement). We can just add an Org for Unidata (here)[https://dev-catalog.ioos.us/organization] and switch the harvest of https://registry.ioos.us/waf/Unidata/ to belong to it.
My humble recommendation is not to twist attribution with the catalog's concept of organization. Proper attribution for who funded the dataset, created the dataset, maintained and distributed it, are still in the original XML documents and represented in our catalog.
@lukecampbell I can't tell if your comment is to recommend for or against Micah's recommendation. I think his rationale makes sense.
I'm in support for his recommendation.
Unidata definitely doesn't reserve those datasets specifically for IOOS. They are for anyone.
I agree with Luke that all the info you should need is in the metadata. Unidata
is fine.
Ok, closing this and making a new issue with the action spelled out for the next sprint @lukecampbell can work on Catalog issues.
Maybe you have this already -- easy issue to close if you do! http://thredds.axiomdatascience.com/iso/unidata/