Referencing macros within RIPE-DB which are defined in RADB ?

Hi Christian,

> Christian Panigl, ACOnet/UniVie +43 1 4065822-383 writes :
> >
> >GSL (Global Sprintlink) is connecting in Paris shortly, and they announce a 
> >lot of ASes. These ASes are defined in an as-macro which resides in the RADB
> >and not the RIPE-DB. Will your filter-build programs be able to pick this
> >macro up from the RADB? Or do they only scan the RIPE-DB?
> 
>     Hi Bjorn,
>     
>     hmm, good question.  We are mirroring those split-parts (rt, as) of the
>     RIPE-DB which are needed by "rlc" to our local machine.  If it turns out
>     that we also need split-parts of the RADB and they are available by FTP
>     like with RIPE, I don't see any reason why it shouldn't work, as long as
>     the RADB splits have the same format.  However, because rlc isn't able

The RIPE NCC is mirroring the RADB database. It's available in:

ftp://ftp.ripe.net/ripe/dbase/others/radb.db.gz

It is easy to split the file in separate files for every type of object
with a small perl script.

>     It's an interesting question which should be discussed by the RIPE-DB
>     working group !  Wilfried please take this, unfortunately I won't attend
>     RIPE-23 !!!  I have the following general questions in mind:
>     
>     - Is it allowed within the RIPE-DB to reference objects (macros) which
>       are defined in a different RR (e.g. RADB) without duplicating them to
>       the RIPE-DB ?  

No that isn't supported yet.

>     - Is there a mechanism which ensures that object-names (macro-names) are 
>       unique ?  

The name of the macro is not unique for all registries. One could
uniquify the name by adding the registry name as defined in the 'source:'
attribute of the database objects.

I think this might be an interesting topic to be discussed during the RIPE
db-wg session,

Kind regards,

David Kessens