Tim:

I've written a new style tgcm netcdf file to try out
in vis5d format. The file on mss is:

-rw----rw-  1 FOSTER  28100039  169514708 Sep  8 13:33 
/FOSTER/tgcm/netcdf/CCMS510_2740000-2750000.nc

Here is an ncdump -h:

netcdf CCMS510_2740000-2750000 {
dimensions:
	lon = 73 ;
	lat = 36 ;
	lev = 43 ;
	time = UNLIMITED ; // (25 currently)
	mtime = 3 ;
	hvol_len = 80 ;

variables:
	float lon(lon) ;
		lon:long_name = "geographic longitude (-west, +east)" ;
		lon:units = "degrees" ;
	float lat(lat) ;
		lat:long_name = "geographic latitude (-south, +north)" ;
		lat:units = "degrees" ;
	float lev(lev) ;
		lev:long_name = "vertical log pressure" ;
		lev:units = "ln(p0/p)" ;
	long time(time) ;
		time:long_name = "total model time (day*24*60+hour*60+min)" ;
		time:units = "minutes" ;
	short MTIME(time, mtime) ;
		MTIME:long_name = "model times (day, hour, minute)" ;
	char HISTVOLS(time, hvol_len) ;
		HISTVOLS:long_name = "mss paths to history volumes" ;
	long YYDDD(time) ;
		YYDDD:long_name = "year and day of year (yyddd)" ;
	float F107D(time) ;
		F107D:long_name = "daily f10.7 cm solar flux" ;
	float F107A(time) ;
		F107A:long_name = "80-day average f10.7 cm solar flux" ;
	float HP(time) ;
		HP:long_name = "Hemispheric power (Gw)" ;
	float CP(time) ;
		CP:long_name = "Cross-tail potential (Volts)" ;
	float BYIMF(time) ;
		BYIMF:long_name = "IMF BY component" ;
	float TN(time, lev, lat, lon) ;
		TN:long_name = "NEUTRAL TEMPERATURE" ;
		TN:units = "DEG K" ;
		TN:type = "SCALAR          " ;
		TN:nlev = 43 ;
		TN:blev = -16.f ;
		TN:dlev = 0.49999997f ;
		TN:vtype = "ZP      " ;
	float UN(time, lev, lat, lon) ;
		UN:long_name = "NEUTRAL ZONAL WIND" ;
		UN:units = "M/S" ;
		UN:type = "VECTOR          " ;
		UN:nlev = 43 ;
		UN:blev = -16.f ;
		UN:dlev = 0.49999997f ;
		UN:vtype = "ZP      " ;
	float VN(time, lev, lat, lon) ;
		VN:long_name = "NEUTRAL MERIDIONAL WIND" ;
		VN:units = "M/S" ;
		VN:type = "VECTOR          " ;
		VN:nlev = 43 ;
		VN:blev = -16.f ;
		VN:dlev = 0.49999997f ;
		VN:vtype = "ZP      " ;
	float O2(time, lev, lat, lon) ;
		O2:long_name = "MOLECULAR OXYGEN" ;
		O2:units = "CM3" ;
		O2:type = "DENSITY         " ;
		O2:nlev = 43 ;
		O2:blev = -16.f ;
		O2:dlev = 0.49999997f ;
		O2:vtype = "ZP      " ;
	float H2O(time, lev, lat, lon) ;
		H2O:long_name = "WATER VAPOR (H2O)" ;
		H2O:units = "CM3" ;
		H2O:type = "DENSITY         " ;
		H2O:nlev = 43 ;
		H2O:blev = -16.f ;
		H2O:dlev = 0.49999997f ;
		H2O:vtype = "ZP      " ;
	float NO(time, lev, lat, lon) ;
		NO:long_name = "NITRIC OXIDE" ;
		NO:units = "CM3" ;
		NO:type = "DENSITY         " ;
		NO:nlev = 43 ;
		NO:blev = -16.f ;
		NO:dlev = 0.49999997f ;
		NO:vtype = "ZP      " ;
	float O3(time, lev, lat, lon) ;
		O3:long_name = "OZONE" ;
		O3:units = "CM3" ;
		O3:type = "DENSITY         " ;
		O3:nlev = 43 ;
		O3:blev = -16.f ;
		O3:dlev = 0.49999997f ;
		O3:vtype = "ZP      " ;
	float O1(time, lev, lat, lon) ;
		O1:long_name = "ATOMIC OXYGEN" ;
		O1:units = "CM3" ;
		O1:type = "DENSITY         " ;
		O1:nlev = 43 ;
		O1:blev = -16.f ;
		O1:dlev = 0.49999997f ;
		O1:vtype = "ZP      " ;
	float OH(time, lev, lat, lon) ;
		OH:long_name = "OH" ;
		OH:units = "CM3" ;
		OH:type = "DENSITY         " ;
		OH:nlev = 43 ;
		OH:blev = -16.f ;
		OH:dlev = 0.49999997f ;
		OH:vtype = "ZP      " ;
	float TE(time, lev, lat, lon) ;
		TE:long_name = "ELECTRON TEMPERATURE" ;
		TE:units = "DEG K" ;
		TE:type = "SCALAR          " ;
		TE:nlev = 43 ;
		TE:blev = -16.f ;
		TE:dlev = 0.49999997f ;
		TE:vtype = "ZP      " ;
	float NE(time, lev, lat, lon) ;
		NE:long_name = "ELECTRON DENSITY" ;
		NE:units = "CM3" ;
		NE:type = "DENSITY         " ;
		NE:nlev = 43 ;
		NE:blev = -16.f ;
		NE:dlev = 0.49999997f ;
		NE:vtype = "ZP      " ;
	float O2P(time, lev, lat, lon) ;
		O2P:long_name = "O2+ ION" ;
		O2P:units = "CM3" ;
		O2P:type = "DENSITY         " ;
		O2P:nlev = 43 ;
		O2P:blev = -16.f ;
		O2P:dlev = 0.49999997f ;
		O2P:vtype = "ZP      " ;
	float W(time, lev, lat, lon) ;
		W:long_name = "VERTICAL WIND (PLUS UP)" ;
		W:units = "M/S" ;
		W:type = "VECTOR          " ;
		W:nlev = 43 ;
		W:blev = -16.f ;
		W:dlev = 0.49999997f ;
		W:vtype = "ZP      " ;
	float Z(time, lev, lat, lon) ;
		Z:long_name = "HEIGHT" ;
		Z:units = "KM" ;
		Z:type = "SCALAR          " ;
		Z:nlev = 43 ;
		Z:blev = -16.f ;
		Z:dlev = 0.49999997f ;
		Z:vtype = "ZP      " ;
	float H(time, lev, lat, lon) ;
		H:long_name = "HYDROGEN (H)" ;
		H:units = "CM3" ;
		H:type = "DENSITY         " ;
		H:nlev = 43 ;
		H:blev = -16.f ;
		H:dlev = 0.49999997f ;
		H:vtype = "ZP      " ;

// global attributes:
		:model_version = "TGCM21" ;
		:missing_value = 9.9999996e+35f ;
		:number_of_fields = 15 ;
		:field_names = "TN      UN      VN      O2      H2O     NO      O3      O1      OH      TE      NE      O2+     W       Z       H       " ;
		:number_of_times = 25 ;
		:first_history = "/ROBLE/flxcm3/CCMS510, 274:00:00" ;
		:last_history = "/ROBLE/flxcm3/CCMS511, 275:00:00" ;
}

Because of the addition of the unlimited time dimension, most of
the global attributes became variables w/ time dimension, e.g.
hp, cp, by, model time, etc. I have checked this file out some w/
IVE (a U of Washington application that uses ncar graphics).
This file has 15 fields at 25 1-hour intervals (a full day).

I may add global or var attributes later, but I guess that shouldnt
confuse anything, as long as the dimensions are consistent.
The size of the file reflects about .475Mb per field per history,
which is not too different from the cray-blocked history itself.

If you could convert this to vis5d, or send me your script/program
that does the conversion w/ instructions, I'll try it out on salida.
I did get vis5d running on vishnu and can get slices, etc., but the
volume rendering is too slow. Salida should be better, but I havent
tried it yet. Thanks,

--Ben