On Thursday 14 July 2016 20:55:26 Carsten Ziegeler wrote:
> > On Thursday 14 July 2016 16:02:46 Carsten Ziegeler wrote:
> >> I think we should have the service user concept and getting a service
> >> user is pretty easy. So I don't see the need for something need,
> >> competing with service users.
> >> 
> >> We can discuss about a scripting service user which gets the required
> >> access rights though
> > 
> > +1, see my comment in SLING-5252 from yesterday
> > 
> > There should be a service user for scripting with limited read access to
> > /apps and /libs (and maybe /etc).
> > 
> > I see Radu's point in having a ResourceResolver provider in Scripting
> > Core.
> > 
> > All Scripting implementations (everyone!) could ask this provider for a
> > "script reader" ResourceResolver which would call
> > ResourceProviderFactory#getServiceResourceResolver(null) – instead of
> > defining a service user mapping for every scripting implementation.
> 
> All that is required is a configuration per bundle and that's it. Do we
> really need an API for that?

No, +1 for scripting service users (subsystems: script reader, cache writer) 
and mappings.

> In addition, what if for whatever reason want to use a different service
> user for the jsp engine than for the sightly engine?

You could still add dedicated service users and ACLs.

> With the proposal we're tying something together which does not
> necessarily belong together just for the sake of saving a single OSGi
> configuration.

Right. I'm not convinced from that approach either.

O.

> Carsten

Reply via email to