[MOBY-dev] RegistryCache

[Paul Gordon]

Hi folks,

Yes, the "Fetching ... ontology" does exactly what Martin describes 
below...builds the definitions for the static 
MobyDataType.getDataType(), MobyNamespace.getNamespace() etc. methods by 
downloading the complete RDFs from Moby Central.  RegistryCache 
coordinates when to update the cached copies, etc..  This is the caching 
mechanism used in Seahawk, which minimizes the number of http connects 
required to gather all the ontology info the program may need to 
access.  If you use this mechanism, you should not need to access the 
cache directly, MobyDataType.getDataType(), etc. handle it transparently 
for you.  If you only need to access little bits of the ontologies, 
CentralDigestCachedImpl may still be your best bet performance-wise.

Regards,

Paul
> But I do not know who (what software component) is using this caching, 
> and what else it requires from Moby Central. It may be the RDF files 
> describing services (and their parts) - they are not cached by the 
> client CentralDigestCachedImpl. For long time, I wanted to 
> re-implement this caching mechanism to get data from the central not 
> by calling its API methods but by getting the RDFs (actually I even 
> believe that Eddie wrote some classes for me to help with that). So 
> far, it has not happened - but when/if it does I may consider to keep 
> in the cache (and make it available) also the RDF files (they may have 
> richer contents than the results gotten by the API methods).
>
> Cheers,
> Martin
>
> -- 
> Martin Senger
>    email: martin.senger at gmail.com <mailto:martin.senger at gmail.com>
>    skype: martinsenger