Re: [ferret_users] mc files don't open anymore in v6.84

[Ansley Manke <ansley.b.manke@xxxxxxxx>]

Hi Hella,
This is due to an undocumented change. Prior to Ferret v6.84, the 
descriptor file was the default data type for SET DATA, so if you did

   yes? set data 1D.mc

Ferret would assume the file was a descriptor file.  Starting with 
Ferret v6.84, the default type is netCDF, and a descriptor file is 
recognized only if its extension is .des.  Ferret has never used the .mc 
file extension to determine file type. (Previous to Ferret v6.84 you 
could have named it "1D.ansley" and it would have behaved in the same way.)

For completeness, it might be nice if we added a file format type,

   SET DATA/FORMAT=descriptor

I'll add that to our list of enhancements, and we will certainly clarify 
the documentation.

Ansley

On 5/29/2013 1:15 PM, Hella Riede wrote:
> Yes, that's it! Thank you so much for this easy solution!
>
> I am surprised that after years of ".mc" tradition - at least in my 
> working environment - this should be something of the past now. After 
> all, it saved some directory listing space.
>
> So the change is possibly a side effect of the "Enhancement to 
> descriptor file definition" 
> http://ferret.pmel.noaa.gov/Ferret/documentation/release-notes/version-6-8-4-release-notes#descript
>
> Thanks again and best wishes,
> Hella
>
>
>
> On 29/05/13 21:07, Andrew Wittenberg - NOAA Federal wrote:
> > Hella,
> >
> > Could it just be the file name of your descriptor?  Renaming your 
> > "1D.mc"
> > to "1D.des" makes it work for me.
> >
> > I see the same results for your file in v6.72 as I do in v6.85, so 
> > I'm not
> > sure anything along these lines has actually changed in Ferret.
> >
> > Andrew
> >
> > On Wed, May 29, 2013 at 1:50 PM, Hella Riede <hella.riede@xxxxxxx> 
> > wrote:
> >
> > > Dear ferreters,
> > >
> > > I cannot open multi-netCDF files (mc files) in ferret anymore. They are
> > > created with nc2mc 
> > > (http://www.pa.op.dlr.de/~**PatrickJoeckel/nc2mc/index.
> > > **html <http://www.pa.op.dlr.de/~PatrickJoeckel/nc2mc/index.html>) on
> > > Linux. It works fine in v6.72, but in v6.84, the following error 
> > > message is
> > > issued:
> > >
> > > **netCDF error NetCDF: Unknown file format (OPeNDAP/netCDF Error 
> > > code -51)
> > > Data set: <whatever mc file>
> > >
> > > Please find attached a very simple example I made by hand containing 
> > > only
> > > a time dimension and 6 time points. Both single files are recognized
> > > normally, but the mc file causes the above error in v6.84.
> > >
> > > So it must be a change in ferret between v6.72 and v6.84. From the
> > > changelogs, I gathered that between these versions:
> > >
> > > 1) the number of dimensions was increased from 4 to 6
> > >
> > > 2) all coordinate axes are now treated as double precision.
> > > http://ferret.pmel.noaa.gov/**Ferret/documentation/release-**
> > > notes/version-6-8-release-**notes/<http://ferret.pmel.noaa.gov/Ferret/documentation/release-notes/version-6-8-release-notes/> 
> > >
> > >
> > > 3) "Ferret adheres more closely with the CF conventions for axis
> > > directions"
> > > http://ferret.pmel.noaa.gov/**Ferret/documentation/release-**
> > > notes/version-6-8-2-release-**notes/<http://ferret.pmel.noaa.gov/Ferret/documentation/release-notes/version-6-8-2-release-notes/> 
> > >
> > > http://ferret.pmel.noaa.gov/**Ferret/documentation/release-**
> > > notes/version-6-7-3-release-**notes/<http://ferret.pmel.noaa.gov/Ferret/documentation/release-notes/version-6-7-3-release-notes/> 
> > >
> > >
> > > 4) "Ferret v6.8.2 is linked with NetCDF 4.2"
> > > http://ferret.pmel.noaa.gov/**Ferret/documentation/release-**
> > > notes/version-6-8-2-release-**notes/<http://ferret.pmel.noaa.gov/Ferret/documentation/release-notes/version-6-8-2-release-notes/> 
> > >
> > >
> > > To 3) I first guessed that mc files don't work anymore if the contained
> > > single files include an axis that is decreasing-ordered (which is 
> > > the case
> > > for our large 3-D files), but my simple example without decreasingly
> > > ordered axes seems to prove that theory wrong.
> > >
> > > Many thanks in advance for your help.
> > >
> > > Best wishes,
> > > Hella
> > >
> > >
> > > P.S: Ferret documentation for descriptor files
> > > http://ferret.pmel.noaa.gov/**Ferret/documentation/users-**
> > > guide/converting-to-netcdf/**CREATING-A-MULTI-FILE-NETCDF-**DATA-SET<http://ferret.pmel.noaa.gov/Ferret/documentation/users-guide/converting-to-netcdf/CREATING-A-MULTI-FILE-NETCDF-DATA-SET>