NOTE: The galeon mailing list is no longer active. The list archives are made available for historical reasons.
Lorenzo, Thanks for following up on this. There is also the question of applying to IANA to formally register mime type(s?). If we were to do so, it's not clear whether to apply for just application/netCDF or whether, as you suggest, we should go for netCDF-3 and netCDF-4. One of the complications is that the netCDF classic data model can be used in conjunction with the netCDF4/HDF5 encoding. Is that netCDF-3 or netCDF-4? It might be worth considering just registering the application/netCDF and then specify qualifiers for the data model and for the encoding. Now that I think about it, there is also the issue of the ncML and ncML-GML encodings. Perhaps the encoding should be kept as a modifier. Maybe the key issue here is how these MIME types and qualifiers will be used. Can someone who has experience actually using MIME types provide a quick summary of how they do so? Thanks again, Lorenzo. -- Ben On Wed, Oct 12, 2011 at 7:38 AM, Lorenzo Bigagli < lorenzo.bigagli@xxxxxxxxxxxx> wrote: > Dear colleagues, > > At the last OGC TC Meeting in Boulder, we discussed the issue of > registering a MIME type for netCDF. > As I remember, someone (at Unidata?) may be drafting a proposal, possibly > following what's in use for xml. > > I've given this an extra thought, while working on UncertWeb and the > Uncertainty Conventions, that I'm sharing here. > > First, as a temporary solution, we are using the "application/x-netcdf" > MIME subtype. > The biggest issue is with netCDF conventions, since a single netCDF file > may comply with multiple conventions. > This gets in the way with a possible hierarchical structure of MIME > sub-typing, like e.g.: > x-netcdf-<convention1>-<convention2>... > > We may take the xml approach of a "+netcdf" suffix, but this wouldn't > solve/address the problem. > So I would suggest an optional "conventions" parameters with the same > syntax and semantics as the global "Conventions" attribute in netCDF. > The content-type of a netCDF-U dataset may look like: > application/x-netcdf; conventions="UW-1.0 CF-1.5" > > We may take the same approach with the version, although I'd probably go > for an "application/x-netcdf-4" subtype, in that case. > I would be interested in comments on this, also to align with what is being > drafted. > > Best regards, > Lorenzo > > > Il giorno 30/set/2011, alle ore 23:18, Ben Domenico ha scritto: > > Hi all, > > The CF-netCDF SWG session was held Thursday morning September 22, 2011 at > UCAR in Boulder. A summary is given below. If anyone who attended the SWG > has additions or corrections, please let us know. > > -- Ben > > ========================================== > > *CF-netCDF SWG Session at the September 2011 OGC TC Meetings in Boulder* > > *Agenda* > > The agenda consisted of: > > 1. CF extension to the netCDF core standard (Stefano Nativi) > 2. CF-netCDF extension to the WCS core standard (Stefano Nativi) > 3. Enhanced Data Model extension to the netCDF core standard > (Ben Domenico) > 4. Uncertainty model for netCDF-CF (Lorenzo Bigagli and Stefano Nativi) > 5. Determine how to deal with errors in existing netCDF binary standard > noted by Simon Cox (see next slide) > > Item 5 was taken care of in brief discussions with Carl Reed and Simon Cox > outside the SWG session. They both agree that the changes are not > substantive, so the typos can be corrected in the existing spec and a phrase > can be inserted clarifying that Requirement 1 is a special type of > requirement, namely, a dependency. > > The presentations and draft documents are available at: > > > https://portal.opengeospatial.org/index.php?m=projects&a=view&project_id=82&tab=2&artifact_id=45016 > > *Issues* > > A couple additional issues came up during the discussion > > - There is a need to register a mime type for netCDF with IANA. The > question is whether to try to come up with all the modifiers (e.g. > netCDF-3, > netCDF-4, netCDF-CF,…) before submitting request. The sense of the group > is that a base mime type necdf should be registered with modifiers for > netCDF-, netCDF-4, CF-netCDF, etc. > - With the new modular approach to specifications, OGC is creating a > bit of a Humpty Dumpty Problem of having too many modules in specs, too > many > conformance classes within modules, etc. As a consequence, potential new > users can be overwhelmed. It isn't always clear how to assemble fragments > into a coherent, working whole. Primers with an overall overview are a > help by not a solution. One new possibility is to emphasize the need for > overview information in Profile specs where the application to a particular > community is documented. > > *Action Items* (all internal to the SWG) > > - Make non-substantive edits to existing netCDF binary encoding spec. > See note in Agenda above. (Ben Domenico) > - In the CF Extension to netCDF core, some items that are optional for > CF are mandatory in the proposed OGC specification. A list of these items > will be circulated with the call for comments. (Stefano Nativi) > - The CF-netCDF overview and planning documents should indicate that > XML-encoding is addressed in CF-netCDF extension to WCS. (Ben Domenico) > - Discussion paper for netCDF uncertainty conventions will be captured > in a discussion paper for next TC. Will include netCDF3 and netCDF4 > options > and will pursue with CF conventions community in parallel (Lorenzo Bigagli) > > *Motions* > > There were no motions at this SWG > _______________________________________________ > galeon mailing list > galeon@xxxxxxxxxxxxxxxx > For list information, to unsubscribe, visit: > http://www.unidata.ucar.edu/mailing_lists/ > > >
galeon archives: