[Go-essp-tech] New DRS structure

[Karl Taylor]

  Dear all,

I suggest reversing the order of the compound variable identifier, 
placing the name first and the table second.  This would be consistent 
with the filename construction.

Best regards,
Karl

On 9/8/10 1:11 AM, Estanislao Gonzalez wrote:
> Oh well, I'll start correcting myself.
>
> I've missed the new "variable" name, as it is now required to be a
> compound of both table name and variable. What I don't recall we have
> discussed is how this will be done, I suppose an underscore will be
> enough so the variable name component of the DRS would be
> <cmor_table>_<cmor_variable>  e.g. 6hrLev_ps, 6hrPlev_ps.
>
> Is this ok?
>
> Thanks,
> Estani
>
>
>
> Estanislao Gonzalez wrote:
>> Hello all,
>>
>> I'm trying to avoid confusion so I'll try to sum up the new DRS
>> structure and concepts we agreed upon yesterday.
>>
>> The new DRS structure will be:
>> activity/product/institute/model/experiment/frequency/realm/ensemble/version/variable
>>
>> This structure will be generated by the drslib from cmor's "old" DRS output.
>>
>> The publication dataset level was introduced for the TDS which involves:
>> activity/product/institute/model/experiment/frequency/realm/ensemble
>>
>> Both catalog and directory (DRS) versions are equal.
>>
>> Please amend any errors you see.
>>
>> Any plan for amending the DRS document?
>> I suggest at least deleting the link to the old one as it isn't valid
>> anymore.
>>
>> Thanks,
>> Estani
>>
>>
>