Be sure to only download the version of SNCTOOLS for your release of MATLAB as specified below.
If you want to live on the bleeding edge, you may use subversion to get the latest updates before a release is officially made. You might also end up with a non-functional version of snctools , but hey, you're a grownup, you're capable of making those decisions.
svn co svn://svn.code.sf.net/p/mexcdf/svn/mexnc/trunk ./mexcdf/mexnc svn co svn://svn.code.sf.net/p/mexcdf/svn/snctools/trunk ./mexcdf/snctools
* December 02 2013 John Evans-r4053 - Fixed NC_DUMP test bug in R2010a and prior, thanks to Wenyu Zhou * November 14 2013 John Evans -r4051 - Fixed "same day" opendap bug in test_nc_info, thanks to Brian Blanton. * November 6 2013 John Evans -r4049 - When dumping char variables, use char print specifier. * * October 31 2013 John Evans -r4047 - Fixed NC_DUMP for global attribute with newline (Go Rutgers!) * July 15 2013 John Evans -r4040 - Updated NC_DUMP to look more like Unidata's ncdump. * January 15, 2013 John Evans -r4033 - Fixed a bad warning on NC_GETVARINFO when java IDs being used. - Fixed indexing issue for string data retrieved by java backend - (thanks to Brendan DeTracey). - Fixed issue with grib files when intra-element ordering is to be preserved. * November 27 2012 John Evans -r4028 - SQUEEZE call sites removed. NC_VARGET will try to restore any leading - singleton dimensions that were removed because of the - row-major-col-major order issue. - Forcing integer COUNT argument in private/snc_get_indexing. * October 30 2012 John Evans -r4024 - Fixed indexing bug that crept into java backend. * October 10 2012 John Evans -r4020 - Added THREDDS_DUMP and THREDDS_INFO to top-level. - Fixed bugs in OPeNDAP string handling. * October 05 2012 John Evans -r4015 - Fixed bug in netcdf-java backend where a file on the matlab path - wouldn't be read if the full path wasn't specified. - Added debug code to force use of netcdf-java for all reading. - Minor test modifications. * September 22 2012 John Evans -r4007 - Not erroring out on zero-size variables. . . .