Hi Bertrand We don't really disagree at all ;-)
In fact having written my previous messages, I wonder, whether our default values we provide to scripts (request, response, etc.) should not really be provided by a BVP in the "request" (or such) scope/context/usage ;-) Somehow I tend to get to the term "context" for this thing. Regards Felix Am 07.08.2013 um 09:34 schrieb Bertrand Delacretaz: > On Wed, Aug 7, 2013 at 9:04 AM, Felix Meschberger <[email protected]> wrote: >> ...But there is not currently a way that a script could declare "I am an HC >> script" or >> "I am a Request Processing script"... > > The missing piece here is that the HC scripts that I'm talking about > (and probably the workflow ones that Justin mentioned) are *not* your > usual Sling script, adapted from a Resource etc. > > In the HC case the (mini-)script is just an expression that's part of > an OSGi config (see SLING-2987 for examples), as in: > > expression = jmx.attribute('java.lang:type=ClassLoading', > 'LoadedClassCount') > 10 > > so the separation between request processing scripts and HC scripts is > obvious. They are executed in very different contexts, which warrant > different script bindings. > > In the HC case, bindings need to be provided by OSGi services so as to > be extensible - we have an existing API to do that > (BindingValuesProviders), the proposal is simply to allow that API to > indicate more specifically what the bindings are meant to be used for. > > Hardcoded bindings for HC are unnecessarily limitative, and inventing > a different bindings provider API just for HC feels very wrong, so I > have a hard time understanding the reluctance to this proposal. > > -Bertrand
