[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]
Subject: RE: [wsrp] spec-2.0-draft-05: lifetime interfaces: metadata required?
I would expect leasing to be detailed in a high level service level agreement so did not consider service description metadata necessary. Also, there is little opportunity for consumers to do much more than renew leases as requested by producers and I expect leasing schedules to evolve sensible, quite long termed values (e.g. order of weeks). We did discuss signally whether or not leasing is in use (producers returning no lifetime element demands explicit destroys) and possibly consumers signaling a desire to lease (by adding a lifetime element to their requests) once a producer signals leasing being an option via exposing the optional ports. Regards, Andre -----Original Message----- From: Richard Jacob [mailto:email@example.com] Sent: 04 March 2005 11:32 To: firstname.lastname@example.org Subject: [wsrp] spec-2.0-draft-05: lifetime interfaces: metadata required? Hi all, - the current approach forces Consumers to use the lifetime operations whenever the producer requires them (makes sense of course :-) ), i.e. if the Producer sets the lifetime in the according context structures (RegCtx&PortletCtx). Don't we lock out Consumers not supporting these operations from using Producers requiring lifetime management even if the Producer supports "infinite" lifetime? The questions I'm targetting at are: Do we need metadata in the ServiceDescription and PortletDescription about the lifetime requirements? I could imagine a lifetimeDecsription like: <LifetimeDescription> <maxLifeTime> (in seconds) <maxRefreshDuration> (in seconds) </LifetimeDescription> with maxLifeTome=null -> infinite lifetime possible, explicit destruction supported maxLifeTime=xxx -> requestor not allowed to set this longer than the period specified with maxRefreshDuration=0 -> no automatic refresh, need explicit refresh maxRefreshDuration=xxx -> automatic refresh, requestor not allowed to set this longer than specified here In this case Consumers not supporting lifetime management could still use Producers which in general support lifetime management but allow also for infinite lifetime, i.e. explicit destroy. Consumers not supporting lifetime management could opt to not use Producers/Portlets which enforce lifetime management. Otherwise we will lock out Consumers willing to support 2.0 but not willing to manage lifetime. Another question is: Producers could choose to support infinite lifetime for certain Consumers only (i.e. special registration) but for null registrations or "try-and-buy" registrations only a limited lifetime. Do we need to express this in the metadata? Mit freundlichen Gruessen / best regards, Richard Jacob ______________________________________________________ IBM Lab Boeblingen, Germany Dept.8288, WebSphere Portal Server Development WSRP Standardization Technical Lead Phone: ++49 7031 16-3469 - Fax: ++49 7031 16-4888 Email: mailto:email@example.com --------------------------------------------------------------------- To unsubscribe, e-mail: firstname.lastname@example.org For additional commands, e-mail: email@example.com