[moby] Re: [MOBY-l] Re: [MOBY-dev] lease versus agent for registry updating

[Mark Wilkinson]

On Wed, 2005-08-17 at 10:16 -0600, Paul Gordon wrote:

> Keeping a cookie around somewhere for a month, then 
> manually using it to renew the lease (remember, automating the lease 
> renewal is just as bad as having an RDF file that you need to change) is 
> bad.  If I lose my cookie, or it is intercepted, I'm toast.

Yes... in fact, that solution makes me lose my cookies! ;-)


> If we can add a "push" like method to the registry saying 
> "check my RDF ASAP", and maybe an optional expiry for people who like 
> leases (the default would be turf the service if the test case fails 
> repeatedly) we should have the best of both worlds...

This is easy enough to do given that we are moving to RDF as our Service
Signature - it just becomes another metadata tag.  However, we would
then have to add some code to the registry that stores lease times, and
this then makes some assumptions about what the underling registry has
to know about.  Given that we are now trying to support both a MOBY
Central and a myGrid/Feta underlying store we cannot make "arbitrary"
decisions like this anymore...  A non-lease mechanism does not require
the registry to "think", only to store the data that it needs and
respond to "come hither agent" calls.

M


-- 
"Ontologists do it with the edges!"

Mark Wilkinson
Asst. Professor
Dept. of Medical Genetics
University of British Columbia
PI in Bioinformatics
iCAPTURE Centre
St. Paul's Hospital
Rm. 166, 1081 Burrard St.
Vancouver, BC, V6Z 1Y6
tel: 604 682 2344 x62129
fax: 604 806 9274