[netCDFJava #WIK-413864]: Reading forecast dates from grib2

hi jitka:

> 
> Hi John,
> 
> I'm  back from  my holiday and I  have picked up the 
> 'grib'  issues
> again.
> 
> I have downloaded  Netcdf Java 4.3.2  Development (build on 
> 29.07.'11) 
> to see whether this version solves the issues I previously described.
> 
> Concerning issue 1) Time offset in hours isn't determined properly :
> As I have previously mentioned,  :
> ".....Starting with 20110719.00.APCP.gec00.t00z.pgrb2af258, NetCdf-Java
> determines a wrong time offset in hours compared to the reference date"
> and you have detected that:
> "... we arent multiplying by 6."
> 
> However 4.3.2 Development still gives the wrong time offset.

It is actually working in the new grib iosp, but that isnt yet the default. ill 
probably be switching to it soon.

> I have also tried 4.2.26 (build on 28.07'11) and this version gives the
> wrong offset too.
> 
> Please can you take a look ?
> 
> 
> Concerning issue 2) Different runtimes in the same file
> Do you think you can provide 'run time dimension' (two dimension time
> coordinate)   in the upcoming 4.3 Beta release ?
> 
> When the 4.3 Beta wil be uploaded, I'am going to test it, i.e. I wil
> rerun all our tests concerning data ingest from various nc/grib1/grib2
> files,  data ingest from nc using openDap, and writing nc files.
> When do you expect to publish the 4.3 Beta ?

im guessing in 2-3 weeks. Im not yet sure if i can get 2D time coordinates in 
4.3, but ill let you know when i know.



Ticket Details
===================
Ticket ID: WIK-413864
Department: Support netCDF Java
Priority: Critical
Status: Open