Dear Martin
I don't think it's practicable for the CF community to undertake to verify
applications as CF-compliant, just as we can't do it with datasets. There are
many datasets in existence which claim to be CF-compliant but are not. However
for the specific aim of compliance checking, perhaps it would be possible for
us to maintain a netCDF file containing an example of every possible
violation of a requirement or recommendation in the CF conformance document,
and maybe an example of everything which is described as legal e.g. all the
actual examples in the conventions document, and some more. Accompanying this
would be a list of the errors and warnings that ought to be found. Thus it
would be a resource for checking the CF-compliance of CF-compliance checkers!
These seems achievable in principle to me, but I'm not sure whether we have
the effort to do it, or whether it would be useful. What do you think?
Best wishes
Jonathan
----- Forwarded message from martin.juckes at stfc.ac.uk -----
> Date: Wed, 19 Jul 2017 12:37:22 +0000
> From: martin.juckes at stfc.ac.uk
> To: cf-metadata at cgd.ucar.edu
> Subject: [CF-metadata] Another CF complaince checker -- from IOOS --- with
> some issues
>
> Hello All,
>
> Following discussions with a colleague (As Stephens) I've taken a look at the IOOS compliance-checker, which contains a module for checking how files comply with the CF convention. I looked at 4 files with known CF errors, and found an average of two erroneous reports per file (listed in an issue which I raised on their github site: https://github.com/ioos/compliance-checker/issues/501 ). There are also ambiguities arising from the fact that they use priority 1 (low) to 3 (high) rather than INFO, WARN, ERROR -- but I haven't gone into all of these in the issue raised.
>
> I'm raising it here as well because I'd like to hear other views on the broader question of community tools associated with CF. It is good that people are getting engaged and working through the details of the convention, not so good if they produce and spread misleading information. In its current state, I don't think the IOOS compliance checker is one we would want to approve, but if they fix the 8 problems identified from a morning looking through the results from tests on 4 files, does that make it OK? or, since the 8 problems I've raised come from looking at a small set of files, should we assume that there are many other problems and ask them to do more?
>
> regards,
> Martin
> _______________________________________________
> CF-metadata mailing list
> CF-metadata at cgd.ucar.edu
> http://mailman.cgd.ucar.edu/mailman/listinfo/cf-metadata
----- End forwarded message -----
Received on Thu Jul 20 2017 - 07:45:09 BST
