server data based only on standard_name

I do have had a lot of work in the past with too different namings of parameters.

We have also build in some of our routines a translation of meanings in our old idl library. This is a quite similiar idea as on using standard_name

http://cfconventions.org/standard-names.html

By unifiying a name by a new attribute helps a lot more than investigating many files to get a never complete list of possible writings of something.

In my opininion everyone who wants to setup a server knows more about the data he wants to publish than anyone else. He just has to add the parameter standard_name if not done.

There are various easy possibilities by. eg. ncatted - netCDF Attribute Editor https://linux.die.net/man/1/ncatted

It is rather simple to not let our routine guess what is meant and also explicit by the user doing it right. Description for setup of data and server has much less special cases.

If you have good reasons to keep

https://bitbucket.org/wxmetvis/mss/src/e403243e4cd55fc751d03f43ef23d56096ae2f7c/mslib/netCDF4tools.py?at=develop&fileviewer=file-view-default#netCDF4tools.py-55

and related code please tell your story.

Comments (2)