[Platforms] Roy enters the fray

[Roy Lowry]

Hi Suzanne,

At last I have found some time to start rationalising our local ship code table with your master list.  What I have done so far is to load your list into Oracle and do a 2-way code difference between your codes and ours.  This turned up a number of errors in our system that I have now sorted out.  This leaves 85 codes in your list that we do not have.  Of these, 82 seem to be ships that I will pass by Fran and then load up. 

The other three worry me a lot.  These are:

74PT PROFILING FLOAT
74VK MINISTRY OF DEFENSE CHARTERED VESSELS
74VV ROYAL MARINE AUXILIARY VESSELS

As these stand, they are names of platform classes (i.e. groups of like platforms) which is totally different from the entity we are calling a ship (unique combination of hull, name and ownership/governance).  I am therefore not at all happy about adding them to the list. The question is what should we do about populating any fields in databases or data files where these codes have been used.

In the case of profiling floats (Argo I guess) I think we should certainly be using a the float identifier, which is maintained by WMO.

In the other two cases what we are saying is that we know the something about the platform, but not exactly what it is.  My current missionary-like zeal for clean and accurate metadata says that these should be coded '7499', which currently carries the semantics 'Unknown UK ship'.  However, another possible view is that we should go for a richer 'vocabulary of unknowns' such as 'Unknown Royal Marine Auxiliary vessel' and 'Unknown MOD charter vessel' .  Anyone any arguments/feelings either way?

Cheers, Roy.