Am Freitag, 31. Januar 2014 schrieb Felix Meschberger: > Hi Oliver hello Felix,
> Am 30.01.2014 um 21:18 schrieb Oliver Lietz <[email protected]>: > > prefer String ServiceUserMapper#getServiceUserId(Bundle, String) > > You mean lowercase "d" (as in Id) instead of uppercase (as in ID) ? Yeah, > IIRC would have done that, too. Yet we reflect the casing of > ResourceResolver.getUserID() which in turn reflects the casing of JCR's > Session.getUserID(). Yes, I mean a lowercase "d". It's even worse in Sling Models where the OSGi mark designation is reflected which results in OSGiServiceInjector (but it's impl only). We shouldn't reflect anything and stick to strict CamelCase (-> OsgiServiceInjector). > > and module > > should be placed in bundles as it's not optional (an extension) for new > > org.apache.sling.jcr.base > > You mean you would place it directly into the bundles folder ? Yes. > I agree that > it is crucial and I am ok with it moving there. Yet, I feare we are > starting to crowd that root folder with the API, Engine, > ResourceAccessSecurity, and ResourceResolver modules and optionally the > ServiceUserMapper and maybe more. We already have a mess with bundles and especially extensions (some use extensions in the package name, some don't) and we shouldn't make this situation more worse. Extension implies it's optional, no? So do we need an additional directory - maybe support - under bundles? I'm not sure yet. regards, O. > Regards > Felix
